bookcover white edit

HealthscopeThe Official Research Book of Faculty of Health Sciences Universi� Teknologi MARA

Healthscope, 2020, Vol. 3(3)

Fakulti Sains Kesihatan

i

The Official Research Book of Faculty of Health Sciences

Editorial Board

Dr Ahmad Razali Ishak Dr Emida Mohamed

Dr Siti Nor Ismalina Isa Dr Fatim Tahirah Mirza Mohd Tahir Beg

Dr Siti Sabariah Buhari Noor Halilah Buari Norhisham Haron

Suzana Yusof Mohd Hafizi Bin Mahmud

Siti Rohana Mohd Yatim Rosilah Wahab

Published by:

Faculty of Health Sciences UiTM Cawangan Selangor Kampus Puncak Alam

42300 Selangor Phone : +603 3258 4300 Fax : +603 3258 4599

Website: fsk.uitm.edu.my

Professional Healthcare Enhancing life

ii

Health Scope 2020 Volume 3(3)

The official research book of Faculty of Health Sciences

Chief in Editor: Dr Ahmad Razali Ishak

Editorial Board Reviewer committee

Dr Emida Mohamed Dr Fatim Tahirah Bt Mirza Mohd Tahir Beg

Dr Siti Nor Ismalina Bt Isa Dr Siti Sabariah Buhari Norhisham Haron

Suzana Yusof Mohd Hafizi Bin Mahmud

Siti Rohana Mohd Yatim Rosilah Wahab Noor Halilah Buari

Prof Madya Dr Maria Justine@Stephany Prof Madya Rodziah Ismail

Dr Shantakumari Rajan Dr Chua Siew Kuan Dr Ahmad Zamir Che Daud

Dr Padma A. Rahman Dr Zamzaliza Abd. Mulud

Dr Siti Khuzaimah Ahmad Sharoni Dr Kartini Ilias Dr Syahrul Bariah Abdul Hamid

Dr Nur Islami Mohd Fahmi Teng Dr Shahridah Kassim

Dr Noor Shafini Mohamad Dr Wan Mazlina Md Saad

Copyright © September 2020

Faculty of Health Sciences Universiti Teknologi MARA Selangor (UiTM)

Puncak Alam Campus 42300, Bandar Puncak Alam Selangor Darul Ehsan, MALAYSIA

All right reserved (e-ISSN: 2735-0649). No part of this publication may be reproduced, stored in a retrieval system or transmitted in any form or by any means, electronic, mechanical, photocopying,

recording or otherwise without permission of the copyright holder.

iii

Preface

Health Scope (e-ISSN: 2735-0649) is a peer-reviewed and evidence-based scientific

research book published by Faculty of Health Sciences, Universiti Teknologi MARA, Puncak Alam Campus, Malaysia. The research book's mission is to promote excellence in health sciences and a range of disciplines and specialties of allied health

professions. It welcomes submissions from academic and health professionals’ community. The research book publishes evidence-based articles with solid and

sound methodology, clinical application, description of best clinical practices, and discussion of relevant professional issues or perspectives. Articles can be submitted in the form of research articles, reviews, case reports, and letters to the editor or short

communications. The research book's priorities are papers in the fields of Physiotherapy, Occupational Therapy, Optometry, Medical Laboratory Technology,

Environmental Health & Safety, Nursing, Nutrition & Dietetics, Medical Imaging and Basic Sciences. Relevant articles from other disciplines of allied health professions may be considered for publication.

Dr Ahmad Razali Ishak

Chief Editor Health Scope

The official Research Book of Health Sciences Universiti Teknologi MARA

TABLE OF CONTENT

Healthscope 2020 Vol 3 (3)

No Title Page

1. Evaluation of thresholding and region growing techniques in segmenting lung nodules in computed tomography images: a systematic review Siti Nur Atiqah Mohamad Sabri, Noor Shafini Mohamad

1

2. Association between non-alcoholic fatty liver disease and

urolithiasis on non-contrast computed tomography: a

systematic review and meta-analysis Raisuha Syuhadah Abdul Rahim, Noor Shafini Mohamad

6

3. Early detection of pulmonary vessel nodule in low dose CT image using local adaptive thresholding Sufia Mohd Hizam, Noor Shafini Mohamad

12

4. Diagnostic accuracy of diffusion weighted magnetic resonance

imaging (DW-MRI) and transvaginal ultrasonography (TVUS) in

assessing myometrial invasion among female patients with

endometrial cancer: a systematic review and meta-analysis Sofea Sarina Mazrul, Mohd Zulfadli Adenan

18

5. Diagnostic accuracy of transvaginal ultrasound (TVUS) and diffused weighted imaging (DWI) in female patient with ovarian cancer: a systematic review and meta-analysis Ainul Shafiqah Zulkifli, Mohd Zulfadli Adenan

26

6. Association of 18F-FDG PET/CT imaging metabolic parameters with tumour size in breast carcinoma Ain Nabila Zahari, Mohd Hafizi Mahmud

33

7. Environmentally-extended input-output analysis for Malaysia Farah Ayuni Shafie, Nursyazwani Suparman, Umi Aida Adlina Mesarandi, Dasimah Omar, Subramaniam Karuppannan, Oliver Ling Hoon Leh

38

8. Caregivers’ burdens and its’ association with activities of daily living performance of individuals with stroke: a cross-sectional study Nur Syafrawa Nabihah Mohamad Razali, Siti Salwa Talib, Nor Faridah Ahmad Roslan, Ahmad Zamir Che Daud

44

9. Evaluating microbiological quality of kitchen equipment in school canteens Rosnaini P. Ramli, Hanis Syakira Salman, Lea Nursuria Sariza Sazari, Alia Azmi

49

10. Removing copper, chromium and nickel in industrial effluent using hydroxide precipitation versus sulphide precipitation Siti Rohana Mohd Yatim, Siti Nor Hazliyana Kasmuri, Hanin Nazhifah Syahjidan, Nurul Shahirazni Mokhtar, Nur Ain Zainuddin

54

11. School functional assessment: performance in functional communication among special education at school Nursyahirah Anuar, Rosilah Wahab

61

12. Nutritional status and feeding practices among toddlers in selected nurseries in Puncak Alam, Selangor, Malaysia Naleena Devi Muniandy, Nik Aishah Aliffah Mahdy, Ifrah Ruwaida Mat Saad

69

13. Development of a healthy cookie to promote toddlers’ brain development: Brainy Bites Airin Fatehah, Naleena Devi

74

14. The relationship between physical activity level with depression, anxiety, and stress among full-time housewives in Selangor, Malaysia Rafidah Abdullah Hassan, Nur Islami Fahmi Mohd Teng

79

15. Evaluation on the knowledge, attitude, and practices toward evidence-based practice in computed tomography among radiographers Nur Ezzatie Aqillah Mohd Sauti , Lyana Shahirah Mohamad Yamin

87

16. Effect of low dose ionizing radiation-induced oxidative stress on sperm count and sperm motility in male mice Nur Adira Mohamad Yusuf, Mohd Hafizi Mahmud

92

Health Scope 1

© 2020 Faculty of Health Sciences, UiTM

REVIEW ARTICLE

Evaluation of thresholding and region growing techniques in

segmenting lung nodules in computed tomography images: a systematic

review

Siti Nur Atiqah Mohamad Sabri, Noor Shafini Mohamad*

1Centre of Medical Imaging, Faculty of Health Sciences, Universiti Teknologi MARA Cawangan Selangor Kampus

Puncak Alam, 42300 Bandar Puncak Alam, Selangor, Malaysia

Abstract:

Image segmentation is an essential step in computer-aided diagnosis and treatment planning of lung

nodules. Therefore, the purpose of this study was to perform a systematic review and provide an

overview of the literature available on image segmentation algorithm, which is thresholding and

region growing method regarding the optimization (of the different methodologies developed) of

lung nodules in the lung CT scan prior for the lung nodule segmentation. This systematic review

was compiled according to the Preferred Reporting Items for Systematic Reviews and Meta-

Analyses (PRISMA) guidelines. A total of 687 articles were retrieved from the databases, and six

articles were selected for this review. The finding showed that a 3D Automatic Lung Parenchyma

Extraction and Border Repair (ALPE&BR), which consists of an Automatic Single Seeded Region

Growing (ASSRG) and a 3D Spherical region-growing method (SPRG), showed the highest

sensitivity of 98.5% and 83.245%, respectively. Improvement of the existing methods or proposing

a new one may be the best option. Standardization of the evaluation metrics is needed to allow a

direct comparison between methods

*Corresponding Author

Noor Shafini Mohamad

Email: [email protected]

Keywords: Computed Tomography, lung nodule, region growing, segmentation, thresholding

1. INTRODUCTION

Lung cancer is one of the most chronic forms of cancer in the world [15]. It accounts for 19.8% of all cancer-related mortality and with the lowest cancer survival (11.0%) in Malaysia [4]. According to the National Cancer Registry (NCR), lung cancer is the third most common cancer in the country, the second-most in males, and the 4th most in females. It has been concluded that cigarette smoking is a major etiological risk factor [4]. Lung cancer is usually being detected at an advanced stage when surgery is no longer an option [7]. According to Loh et al (2007) the reasons for refusable to be operated were patients' beliefs in traditional, complementary treatment rather than medical treatment.

Evaluation of lung nodules in the past is usually by CT scan with thin-slice thickness, contrast media, and image post-processing such as using multiple imaging techniques, including shaded surface display (SSD), volume-rendering (VR), and multiplanar reconstruction (MPR) [5]. Contrast media helps the radiologist to have extra information about the vasculature of the suspected pulmonary nodular disease. Nevertheless, a thinner slice produces a greater detail, but the system needs to increase the mAs to reduce images noise-producing more exposure to the patient. Contrast media is highly potential for an adverse reaction, especially for a patient who in advanced age, elevated serum creatinine level,

dehydration, diabetes, and renal disease. However, the detection of faint abnormal regions, such as Ground-glass opacity nodules (GGNs), can be challenging and time-consuming since GGNs are usually small and show contrast with the surrounding lung parenchyma [7].

However, these days, researchers more likely to solve segmentation problems that involve machine learning rather than the simplest yet effective segmentation algorithm [6]. Therefore, the purpose of various thresholding and region growing algorithms is to provide an overview of the literature available on image segmentation algorithms, which are thresholding and region growing methods for optimization of lung nodules in the lung CT scan. A key to better diagnosis of the pathological condition in the image can be achieved through a correct segmentation process. Therefore, the rationale of doing the literature search is to help to find a suitable segmentation method that will help the radiologist to improve for a better delineation of lung nodules effectively.

2. MATERIALS AND METHODS

The eligibility criteria and analysis in this review were performed according to the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) guidelines 2009. Research question was formulated based on the patient, index test, comparator, outcome, and study design (PICOS) criteria as the following: What is the performance of

2 Health Scope, 2020, Vol. 3(3) Siti Nur Atiqah et al.

the thresholding and region growing methods in segmenting the lung nodule in CT images, as compared with the radiologist's manual segmentation?

2.1 Literature Search

The review methods of the present studies were conducted using two primary databases which are PubMed and IEEE Xplore. A literature search was also done through two supporting databases, such as Taylor & Francis Online and Dimension. Search strings were developed in April 2020, using the following expressions:

Search string number 1: ("Threshold" OR "Thresholding" OR "Otsu" OR "Multi-Thresholding" OR "Adaptive Thresholding") AND ("Segmentation" OR "Delineation") AND ("Lung" OR "Lungs" OR "Pulmonary" OR "Chest") AND ("Nodule" OR "Nodules" OR "Cancer" OR "Tumor" OR "Tumors" NOT "Emphysema") AND ("Computed Tomography" OR "Computed Tomographic" OR "CT" OR "CT SCAN" NOT "PET/CT" NOT "Positron Emission Tomography" NOT "Radiotherapy").

Search string number 2: ("Region Growing" OR "Seeded Region Growing" OR "SRG") AND ("Segmentation" OR "Delineation") AND ("Lung" OR "Lungs" OR "Pulmonary" OR "Chest") AND ("Nodule" OR "Nodules" OR "Cancer" OR "Tumor" OR "Tumors" NOT "Emphysema") AND ("Computed Tomography" OR "Computed Tomographic" OR "CT" OR "CT SCAN" NOT "PET/CT" NOT "Positron Emission Tomography" NOT "Radiotherapy").

In total, 687 articles were retrieved in the first stage of the literature search process.

2.2 Screening Criteria

The purpose of the first stage of screening was to remove duplicate articles. The selection criteria for the publications included were the journal (research articles), articles written in English and published between 2012 and 2020. Then duplicated articles were removed. After the removal of duplicates, a total of 34 articles were prepared for the next stage, known as eligibility.

2.3 Eligibility Criteria

Data for the included studies were extracted using a standardized form according to the following attributes: (1) Data presented in an article must at least show the accuracy, sensitivity, specificity, Dice similarity coefficient, and Jaccard's index; (2) Segmented images from the segmentation method must be compared to the radiologist's manual segmentation, and (3) Evaluation must be done on CT images that are more than 10.

2.4 Data Extraction and Quality Assessment

Data regarding the accuracy, sensitivity, and specificity of the segmentation algorithm for the lung nodule segmentation are recorded as well as the Dice similarity coefficient and Jaccard's index. The total number of CT images that were stated in the articles was recorded as well as where the data sets were acquired. General information related to the interesting data were also included. Those were the name of authors, publication time, type of used scanner, acquisition and technical parameter, and image resolution.

Quality assessment was conducted based on the best evidence synthesis, were all retrieved articles eligible for inclusion have to undergo a quality assessment process during the synthesis of results. The studies that meet the predefined inclusion criteria for the review were appraised by the two people based upon predefined guidelines, and the studies are judged as being scientifically acceptable or not. This process relied on the skill and expertise of the reviewer, which might be a source of potential bias. Where a study is not regarded as scientifically admissible, it is rejected from the review process.

2.5 Data Synthesis and Analysis

A qualitative synthesis of quantitative studies was used to qualifying all selected data. Qualitative synthesis refers to the systematic review part, in which there are several aspects of the review process that requires searching, critical appraisal, extraction, and summarising with or without additional quantitative meta-analysis

3. RESULTS AND DISCUSSION

3.1 Literature search

The current research work successfully retrieved a total of 679 articles from the databases. Manual searching efforts were also conducted on other databases based on similar keywords, which resulted in an additional number of 8 articles. In total, 687 articles were then screen based on the inclusion and exclusion criteria. The criteria for the publication include (1) Journal (research article), (2) publication timelines between 2012 and 2020, and (3) published in English.

Overall, a total of 456 articles were excluded based on these criteria, leaving a total of 231 articles to be included. After the removal of 35 duplicates studies, the remaining 196 studies were examined thoroughly based on the title, abstract, and the main contents of all the articles to ensure that they fulfilled the inclusion criteria of the study selection and fit to be employed in the present study in order to achieve the objectives of the current research. A total of 20 full-text articles were assessed for eligibility, and 14 of these articles were excluded because lack of data requirements, thus, leaving only six articles that were included in this systematic review. Two studies focused the thresholding method [9] [10], one study combined the thresholding and region growing method [11], and three studies focused on the region growing method [1] [3] [7]. The study selection was summarised in Figure 1.

Health Scope 3

Figure 1. Flow diagram of the study selection process.

Table 1. The summarization of the lung nodules segmentation methods.

Author Type Method Nodule

types

Evaluation

Metric

Samundesswari et al. [1]

Region growing

ALPE&BR All types ACC: 99.4 SN: 98.5

SP: 98.8

Dice:0.982

Kim et al. [3] Region growing

SPRG Solid ACC: 99.9 SN:83.245

SP: 99.97

Dice: 0.87

Kim et al. [7] Region

growing

Seeded

region

growing

GGN Dice:

0.6290

Lassen et al. [11]

Thresholding & region

growing

Thresholding based region

growing.

Sub-solid (n=19)

Jaccard: 0.52

Sub-solid (n=40)

Jaccard: 0.50

Mukhopadhyay

[10]

Thresholding Thresholding

(-500 HU)

followed by connected

component

analysis.

Solid/part-

solid

nodule

ACC:99.00

SN: 58.00

SP: 99.00

Thresholding

(-800 HU)

followed with

anisotropic

diffusion application.

Non-solid

nodule

ACC:98.00

SN: 44.00

SP:100

Guo et al. [9] Thresholding OTSU

Thresholding

All types Jaccard:

0.586

Fixed thresholding

Jaccard: 0.601

*ACC (%): Accuracy. SN (%): Sensitivity. SP (%): Specificity

3.2 Review of Segmentation Algorithm for Lung Nodule Segmentation

3.2.1 Thresholding Technique

Some authors take advantage of the easiness of the threshold and region growing methods (Table 1). Therefore, in this section, explanations regarding the work of authors and results would be explained, along with the advantages and disadvantages discovered by the authors with their proposed methods.

In a study conducted by Guo et al. [9], Otsu and fixed thresholding methods were compared on a total of 85 CT scans images, which consisted of 111 various types of pulmonary nodules. In fixed (or global) thresholding, the threshold value is held constant throughout the image. In contrast, Otsu's thresholding chooses the threshold to minimise the intraclass variance of the thresholded black and white pixels constant throughout the image. To obtain the ideal initial contour by choosing a suitable global threshold was difficult because of the differences in the grey levels of lung nodules [2]. However, in Otsu method had a Jaccard's index of 0.586, while the fixed thresholding method had a higher index, 0.601. If the global threshold value is correctly selected, it will yield the highest value of overlap. This method offers a high calculation speed and accuracy.

Mukhopadhyay et al. [10] employed a thresholding method that involved the application of connected component analysis and anisotropic diffusion application. The authors isolated the lung nodules into solid and part-solid grouped in one category and non-solid in another since these two groups have different density distribution. A solid lung nodule usually lies within the range of -200 HU to 200 HU in the density distribution, while the part-solid lung nodule ranged between −733 HU to 295 HU [13]. The authors selected -500 HU to HU to separate the nodule from lung parenchyma since these nodules are heavily overlapped in this region. They were considering that solid or part-solid nodules are different from the sub-solid nodules. Kauczor et al. [14] reported that the non-solid nodule could be ranged between −750 HU to −300 HU for its density distribution. The selection of threshold at −800 HU, which is well below minimum attenuation for a non-solid nodule, was employed by Mukhopadhyay et al. [10] prior to lung parenchyma removal.

Next, a connected component analysis was performed to obtain the object containing nodule after thresholding. Only sub-solid lung nodules underwent the Anisotropic diffusion filter application on the grey-scale version of the thresholded image after the connected component analysis. If any holes present in the nodule, they will be filled up using morphological closing operation. Lastly, an ellipsoid fitting was performed on all types of nodules irrespective of attachment to reduce computation time.

The procedure of removal of pleural attachment was the same for solid/part-solid and non-solid nodules. The pleural surface removal technique of Kuhnigh et al. [12] is applied in the ellipsoid volume of interest (VOI). Any unwanted pleural surface portion was removed using a refinement step through the morphological dilation application compared to removing the attached blood vessel, the vasculature removal procedure for solid nodule did not apply to non-solid nodules. The solid nodule used a Vasculature Pruning Technique to remove attached vessels from the nodule. In contrast, a Non-solid

4 Health Scope, 2020, Vol. 3(3) Siti Nur Atiqah et al.

nodule used Selective Enhancement Filtering for Removal of Vessels in the ellipsoid VOI.

The accuracy and specificity values of thresholding value at -500 HU were measured as 99% and 58%, respectively. For thresholding value at -800 HU, it was measured as 98% for accuracy and 100% for specificity. Contrarily, the sensitivity showed as 44%. The authors faced a problem where about 6 out of 891 lung nodules were failed to be segmented. They identified that the pruning technique assumes there were no vascular attached to a well-circumscribed nodule.

3.2.2 Region Growing Technique

Lessen et al. [11] has proposed a 3D method starting by defining a region of interest around a pulmonary nodule using a user-drawn stroke. A threshold-based region growing approach in a 6-connected neighborhood was performed for sub-solid nodules based on intensity analysis of the nodule region and the surrounding parenchyma. The chest wall was then removed by a combination of connected component analysis and convex hull calculation. A convex hull helps to reduce the over- and under-repair when separating the attached structures. Finally, attached vessels were detached by morphological operations. The proposed method can handle subsolid nodules of different densities automatically. The authors conducted two sets of studies where different radiologists acquired the ground truths. Study A used a total of 19 subsolid nodules that yield a Jaccard's index of 0.52. Whereas in study B, Jaccard's index was measured at 0.50 on a total of 40 subsolid nodules.

The advantage of the threshold-based region growing technique is, it offers a solution for a large set of subsolid nodules (n=59) collected from many different institutes and subjects. It was found that the adjacent vasculature is still included in the segmentation of the nodule (n=?), although the vasculature removal is persistent in follow-ups.

Kim et al. [7] conducted several methods using 2D semi-automatic Segmentation Methods for Persistent Ground Glass Nodules on Thin-Section CT Scans. One of the methods included the region growing method, particularly the Seeded region growing. The ground truth is the result of the manually segmented area by two radiologists. The mean Dice similarity coefficient for this method was 0.629 and been performed on 40 persistent GGNs.

The region growing method is well-known as a tool that is efficient at extracting a homogenous region in an image. Additionally, this method is efficient and robust in dealing with attenuation variations, for instance, attenuation variations caused by the pathologic conditions and imaging artifacts by reinforcing spatial neighborhood information and a regional term. In any case, post-processing is needed after the application of the region growing method because of the noise magnitude and the precision neighborhood criteria, which can suffer from false positive [16]. Furthermore, manual interaction or another intelligent algorithm for the selection of seed may be preferred to be run before the segmentation process. The delineation process is manually select the seed to find a suitable voxel or region.

To improve the accuracy of the traditional region growing method, Kim et al. [7] performed a 3D Spherical region-growing method (SPRG) on CT lung images that contained the solid nodules. The authors explained, SPRG is a

modification of an existing region-growing method that is based on a sphere instead of pixels. Usage of the sphere in this method prevents the leakage of growing into the neighboring tissues during the nodule detection since there was some requirement to add a neighboring voxel. They obtained an accuracy of 99.9%, sensitivity of 83.245%, and a specificity of 99.97%. The Dice similarity coefficient was about 0.87. A negative point about this work concerns the validation of the system in which it is merely tested with only solitary nodules. Thus, it is not guaranteed that this system presents the same performance in other circumstances because the system was not tested with a broad range of types of nodules.

Samundesswari et al. [1] conducted a segmentation study using a 3D Automatic Lung Parenchyma Extraction, and Border Repair (ALPE&BR) consists of Automatic Single Seeded Region Growing (ASSRG) on 30 CT scan lung images acquired from LIDC/IDRI public database. The authors claimed that this is a Hybrid Segmentation and Refinement Method for Automatic Lung Cancerous Nodules Extraction. ALPE&BR consists of five main steps. Avoiding the over or under segmentation effect and easing the process of automatic seed selection are the main points why the authors divide the lung into the right and left the region. Then, the isolated type of nodule was segment using the ASSRG Algorithm. The border concavity closing makes use of the morphological operation, connected component analysis, logical rules to extract the lung lobe region based on its area.

The inner and outer boundary of the thick boundary line was detected using the canny edge detection. Calculation of the minimum area of the entire lung lobe region eventually prevents the over the segmented region of the lung wall. Then, elimination of the small vessel structures, non-cancerous nodules that is less than 1cm, or other impurities within the lung parenchyma were done through a Connected Component Analysis (CCA) and Threshold Based Mathematical Nodule (TBMN) refinement.

The acquired images consist of all types of lung nodules. The accuracy, sensitivity, and specificity values in this study were measured as 99.4%, 98.5%, and 98.8, respectively. The Dice similarity coefficient was 0.982. The only dilemma was that the authors separated the left and right lung to achieve the highest accuracy value. They assumed that the inclusion of other findings within both sides of the lung parenchymal could thoroughly affect the performance of nodules detection accuracy. Elimination of other findings in the lung parenchyma included a small non-cancerous nodule (<1cm), vessel structures, bronchioles, and other impurities in the lung parenchyma to mimic clinical situations may consume much time and inefficient to be applied [1].

Choosing segmentation methods that give the most adequate and accurate evaluation metrics eventually assists radiologists in defining the pulmonary nodule and predict the degree of malignancy through the feature extraction and classification. Strengths and flaws were the indicators of choosing the desired segmentation methods. The complexity level of the nodule directly reflects the performance of the segmentation method.

There were some promising results of the segmentation methods [1]. The ALPE&BR method that consists of ASSRG showed a sensitivity of 98.5% (Accuracy;99.4 %, specificity;98.8%) and the Dice similarity coefficient that was

Health Scope 5

near to 1 [1]. However, validation of these systems was limited to tests with up to 30 CT scans, although being tested on all types of lung nodules. The authors did state that within these 30 CT Scan images, there was variation regarding the lung nodules that consist of solid, partially solid or non-solid, dissimilar texture (shape, size, area, and perimeter) and different locations (isolated, multinodule, juxtavascular and juxtapleural) to evaluate the effectiveness of the proposed method. However, it can be among 30 CT scans; there were some nodules not selected equally in terms of quantity, which may directly affect these results. Also, the authors purposely separate the right and left lung to ease the lung nodule segmentation to achieve the highest sensitivity, although, in the clinical setting, the simultaneous segmentation method helps enhance the radiologist efficiency.

The SPRG method [3] stood the second highest to the ALPE&BR method [1] with a sensitivity of 83.245%. They only used one type of nodule, which was the solid nodules were one of the easiest to segment. Thereby, the same performance may not guarantee to present the same performance in other conditions, considering this method was not tested with a vast range of the lung nodules.

Lessen et al. [11] have achieved the satisfactory for the segmentation processing speed where they manage to segment the subsolid nodule in less than 1 second of computation time. However, not all the acquired articles have provided such information. Some authors have overlooked this information; the sufficient algorithm time is needed along with a high sensitivity segmentation method so that this information can be used and compare with other methods since robustness is needed once the methods are applied in the clinical environment.

Moreover, there was a lack of standardized evaluation metrics to assess the segmentation performance, which can curb a correct comparison between studies. A standardized evaluation metrics can ease a future researcher in developing a systematic review that is appropriately evaluated and compared among the selected publication. Besides, additional surveys on the lung segmentation methods are needed in the technical aspects. These aspects can be varied from the ability of the segmentation methods to segment the pulmonary nodule in different scanning parameters to the degree of severity of the inhomogeneous of the nodule.

Most methods achieved an accuracy that was more than 97%. However, sensitivity is more appealing to the radiologist in terms of evaluating the segmentation performance. Therefore, only works by Kim et al. [3] and Samundesswari et al. [1] showed the highest sensitivity. Nevertheless, it is necessary to improve the specificity of the purpose methods in future research.

4. CONCLUSION

The ALPE&BR method that consists of ASSRG was considered the best segmentation method in segmenting the lung nodule in CT images among the acquired articles although, this method needs to undergo several things such as using a larger amount of sample image with a variety of lung nodule, hence confirming its ability to perform lung nodule segmentation accurately.

ACKNOWLEDGEMENTS

Authors thanks the librarian for assistance in resource searching.

REFERENCES

[1] Samundeeswari, P., & Gunasundari, R. (2019). A novel hybrid

segmentation and refinement method for automatic lung cancerous

nodules extraction. International Journal of Recent Technology and Engineering, 7(6), 28–35.

[2] Xiao, X., Zhao, J., Qiang, Y., Wang, H., Xiao, Y., & Zhang, X. (2018).

applied sciences An Automated Segmentation Method for Lung Parenchyma Image Sequences Based on Fractal Geometry and Convex

Hull Algorithm. Applied Sciences.

[3] Kim, Y. J., Lee, S. H., Lim, K. Y., & Kim, K. G. (2018). Development and Validation of Segmentation Method for Lung Cancer Volumetry

on Chest CT. Journal of Digital Imaging, 31(4), 505–512.

[4] National Cancer Institute. (2018). Malaysian study on cancer survival (MySCan). In National Cancer Institute, Ministry of Health Malaysia

(Vol. 4).

[5] Nanusha Assistant professor ECE, Institue of Aeronautical Engineering, H. (2017). Lung Nodule Detection Using Image

Segmentation Methods. International Journal of Advanced Research

in Electronics and Communication Engineering (IJARECE) Volume, 6(7), 690–697.

[6] Uzelaltinbulat, S., & Ugur, B. (2017). Lung tumor segmentation

algorithm. Procedia Computer Science, 120, 140–147. [7] Kim, Y. J., Lee, S. H., Park, C. M., & Kim, K. G. (2016). Evaluation

of semi-automatic segmentation methods for persistent ground glass

nodules on thin-section CT scans. Healthcare Informatics Research, 22(4), 305–315.

[8] Soon Hin How, Teck Han Ng, Yeh Chunn Kuan, Abdul Rahman

Jamalludin and Abdul Rani Fauzi. Survival of lung cancer patients in a resource-limited country. Asia Pac J Clin Oncol 2015; 11(3): 221-7

[9] Guo, W., & Li, Q. (2014). Effect of segmentation algorithms on the

performance of computerized detection of lung nodules in CT. Medical Physics, 41(9), 1–8.

[10] Mukhopadhyay, S. (2016). A Segmentation Framework of Pulmonary

Nodules in Lung CT Images. Journal of Digital Imaging, 29(1), 86–103.

[11] Lassen, B. C., Jacobs, C., Kuhnigk, J. M., Van Ginneken, B., & Van

Rikxoort, E. M. (2015). Robust semi-automatic segmentation of pulmonary subsolid nodules in chest computed tomography scans.

Physics in Medicine and Biology, 60(3), 1307–1323.

[12] Kuhnigk JM, Dicken V, Bornemann L, Bakai A, Wormanns D, Krass S, Peitgen HO. Morphological segmentation and partial volume

analysis for volumetry of solid pulmonary lesions in thoracic CT scans.

IEEE Trans Med Imaging. 2006;25(4):417– 434. [13] Reeves AP, Chan AB, Yankelevitz DF, Henschke CI, Kressler B,

Kostis WJ. On measuring the change in size of pulmonary nodules.

IEEE Trans Med Imaging. 2006;25(4):435–450.

[14] Kauczor HU, Heitmann K, Heussel CP, Marwede D, Uthmann T,

Thelen M. Automatic detection and quantification of groundglass

opacities on high-resolution CT using multiple neural networks: comparison with a density mask. Am J Roentgenol.

2000;175(5):1329–1334. [15] Sharma, P., Mehta, M., Dhanjal, D. S., Kaur, S., Gupta, G., Singh, H.,

Thangavelu, L., Rajeshkumar, S., Tambuwala, M., Bakshi, H. A.,

Chellappan, D. K., Dua, K., & Satija, S. (2019). Emerging trends in the novel drug delivery approaches for the treatment of lung

cancer. Chemico-biological interactions, 309, 108720.

[16] Mansoor, A., Bagci, U., Foster, B., Xu, Z., Papadakis, G. Z., Folio, L. R., Udupa, J. K., & Mollura, D. J. (2015). Segmentation and Image

Analysis of Abnormal Lungs at CT: Current Approaches, Challenges,

and Future Trends. Radiographics: a review publication of the Radiological Society of North America, Inc, 35(4), 1056–1076.

Health Scope 6


REVIEW ARTICLE

Association between non-alcoholic fatty liver disease and urolithiasis on

non-contrast computed tomography: a systematic review and meta-

analysis

Raisuha Syuhadah Abdul Rahim, Noor Shafini Mohamad*

Centre of Medical Imaging, Faculty of Health Sciences, Universiti Teknologi MARA Cawangan Selangor Kampus Puncak

Alam, 42300 Bandar Puncak Alam, Selangor, Malaysia

Abstract:

There is growing evidence that non-alcoholic fatty liver disease (NAFLD) is associated with

urolithiasis. NAFLD could be a predisposing factor for urolithiasis. However, this association has

been unclear and inconsistent. Thus, this study was conducted to review the existing studies that

compared the risk of urolithiasis in patients with NAFLD to patients without NAFLD detected on

non-contrast computed tomography (CT). A comprehensive literature review was carried out using

leading journal databases including Scopus, Dimensions, Web of Sciences, and Science Direct, to

identify all related studies. The review method was guided by PRISMA statement as reporting

guidelines and standards in the current study to ensure the review reported is with a high level of

details. The literature search resulted in four studies with 2,999 individuals who fulfilled the

eligibility criteria and were included in the analysis. NAFLD significantly associated with an

increased risk of urolithiasis with a pooled odd ratio of 2.83 (95% CI, 1.14-7.00: I2 90%). A

significantly increased risk of urolithiasis among patients with NAFLD compared to patients without

NAFLD was observed in this meta-analysis. Finally, future studies investigating NAFLD and

urolithiasis and its associate factors are needed to allow broader understanding and treatment of the

patient’s condition.




Keywords: Computed tomography, meta-analysis, non-alcoholic fatty liver disease, systematic

review, urolithiasis

1. INTRODUCTION

Non-alcoholic fatty liver disease (NAFLD) is a common

disorder where the accumulation of excess fat in the liver of

people who drink little or no alcohol. NAFLD is also known

as fatty liver and hepatic steatosis [1]. Meanwhile, urolithiasis

refers to the presence of stones in the urinary tracts. It is a

universal disease that affecting people across geographical,

cultural, and racial boundaries [2]. The prevalence of NAFLD

and urolithiasis has increased significantly in the last few

decades in most Asian countries [2,3].

NAFLD has been reported to cause urinary stones, and the

advanced form of NAFLD is associated with a high risk of

urolithiasis [4]. Both NAFLD and urolithiasis are associated

with conditions such as metabolic syndrome, obesity,

dysbiosis of gut flora [5], cardiovascular morbidity, and

chronic kidney diseases (CKD) [1]. Furthermore, NAFLD and

CKD may share a possible connection in the pathogenic

processes and mutual pathophysiological mechanisms [6].

The evidence of the association between NAFLD and

urolithiasis is growing; however, the association remains

unclear and inconsistent across different studies [1,7]. There

seems evidence of incidental findings of both NAFLD and

urolithiasis in the same patient that was referred to non-

contrast CT [8,9]. A putative link and the exact mechanisms

linking NAFLD to urolithiasis has not been well established.

A systematic review and meta-analysis of a cross-sectional study were conducted to compare the risk of urolithiasis inpatient with NAFLD compared to patients without NAFLD detected in non-contrast computed tomography.


This section explains the five main sub-sections, namely

PRISMA, resources and information sources, systematic

searching strategies, quality of appraisal, and data abstraction

and analysis, which are employed in the current study.

2.1 PRISMA

The review was guided by PRISMA statement (Preferred

Reporting Items for Systematic Reviews and Meta-Analyses),

7 Health Scope, 2020, Vol. 3(3) Raisuha Syuhadah et al.

which is a published standard to conduct a systematic

literature review. It is often utilised in healthcare behavior and

intervention [10]. It allows for the search of terms related to

the association of NAFLD and urolithiasis and the risks that

link those diseases.

2.2 Resources and Information Sources

The review methods of the present study were conducted

using four journal databases including Scopus, Web of

Science (WoS), Science Direct, and Dimensions. These main

journal database delivers an overview of more than 256 fields

of studies and contains millions of publications for medical

and health sciences. They also can be a leading database in a

systematic literature review due to advance searching strategy

and function, comprehensive and article’s quality control. The

present study was carried out in April 2020 to identify all

studies that related to the association between NAFLD and

Urolithiasis.

2.3 Systematic Searching Strategies

2.3.1 Identification

The first process of systematic searching strategies is

identification. The search process began with identifying any

similar or related terms for the main keywords for this review.

These included NAFLD, urolithiasis, and non-contrast CT

throughout thesaurus dictionaries, related previous articles

and keywords suggested by journal database and

encyclopedia. Accordingly, search strings in the journal

database were developed for this present study (Table 1). A

total of 495 publications were retrieved in the first stage of the

systematic review process.

Table 1. The search strings

2.3.2 Screening

The screening process was conducted to remove the

articles based on inclusion and exclusion criteria—the first

criterion with regards to a timeline within five years from

2015 until March 2020. The second criterion was only article

journals with empirical data were selected to focus only on the

primary sources. Hence, the publication in the form of the

reviewed articles, meta-analysis, books, book chapters, and

conference proceedings was excluded in the current study.

The review only focused on articles published in English. This

process was done automatically through a sorting function

available in the database. A total of 390 articles and 24

duplicated articles were excluded based on these criteria

(Table 2).

Table 2. Inclusion and exclusion criteria Criteria Inclusion Exclusion

Timeline Between 2015- March

2020

< 2015

Literature Type Journal

(Research articles)

Review articles, systematic

review, meta-analysis,

meta-synthesis, book, book

series, book chapter,

conference proceedings

Language English Non-English

2.3.3 Eligibility

A total of 81 related articles were prepared for the third

process, which was eligibility. At this stage, titles, abstracts,

and main contents of all articles were examined thoroughly to

ensure they fulfilled the inclusion criteria and fit to be

included in the analysis. Consequently, 77 articles being

excluded due to inadequate data on outcomes of interest and

different methodologies used to identify NAFLD and

urolithiasis. All studies included for this review were cross-

sectional, case-controlled, or cohort studies investigating the

risk of urolithiasis among patients with NAFLD compared to

individuals without NAFLD detected on non-contrast CT.

Finally, four remaining articles were ready to be analysed

(Figure 1).

2.4 Quality appraisal

In ensuring the quality of articles’ content, the remaining

article was assessed using a quality assessment tool, the

Newcastle-Ottawa scale (NOS). There are three domains in

NOS, which are the recruitment of participants, the

comparability between groups and ascertain the exposure or

outcome of interest. The modified NOS, described by Herzog

et al. (2013) was used for the cross-sectional study [11]. A

‘star system’ has been proposed in which a study is assessed

in three domains: selection (maximum 4 stars), comparability

(maximum 2 stars), and exposure/outcome (maximum 3

stars).

2.5 Data abstraction and analysis

Data abstraction was guided by the research objectives

and questions. The extracted data included the following

items: authors, publication year, country, and region area on

the study and number of subjects. Statistical analysis was

Scopus TITLE-ABS-KEY

(("urolithiasis" OR "nephrolithiasis" OR "urinary

ston*" OR "renal lithiasis") AND ("nonalcoholic

fatty liver diseas*" OR "fatty

liver" OR "nonalcoholic fatty liver"))

Web of Science TS= (("urolithiasis" OR "nephrolithiasis" OR "renal

calcul*" OR "kidney colic*" OR " renal ston* " OR

" kidney ston*" OR " urinary ston* OR "urinary

calcul*") AND ("nonalcoholic fatty liver diseas*"

OR "fatty liver" OR "nonalcoholic fatty liver" ))

Dimensions ("urolithiasis" OR "nephrolithiasis" OR "renal

calculi" OR "kidney colic" OR "urinary calculi" OR

"renal stones") ("nonalcoholic fatty liver disease"

OR "fatty liver" OR "nonalcoholic fatty liver")

Science Direct ("urolithiasis" OR "nephrolithiasis" OR "renal

calculi" OR "kidney colic" OR "urinary calculi" OR

"renal stones") AND nonalcoholic fatty liver

diseases

Health Scope 8

performed using Cochrane Collaboration’s Review Manager

5.3 software. Odds ratio (ORs) were pooled with their 95%

confidence interval (CI), which was the assumption that these

were a comparable measure of the association because of the

relatively rare prevalence of urolithiasis. Cochran’s Q test and I2 statistics were used to quantify the heterogeneity between

study. A value of I2 of 0-25% represents insignificant

heterogeneity, 51-75% represents moderate heterogeneity,

and 75% and above consider high heterogeneity [12].

Publication bias was assessed with a funnel plot.


3.1 Characteristics of selected studies

A total of 495 articles were identified using the described

search strategy (214 from Scopus, 36 from WoS, 45 from

Dimensions, and 200 from Science Direct). Of these, 491

articles were excluded from the current study for reasons

reported presented in Figure 3. Finally, four cross-sectional

studies with 2,999 participants were eligible and for inclusion

in the meta-analysis. The characteristics and quality appraisal

of the included studies are presented in Table 3. Of the four

studies, one study received ten stars, and three studies

received nine stars at the NOS.

3.2 Association between NAFLD and urolithiasis

In the current study, it was found a significantly increased

risk of urolithiasis among patients with NAFLD with the

pooled OR of 2.83 (95% CI),1.14-7.00) (Figure 2). The

heterogeneity between studies was high, with an I2 of 90%.

The funnel plot was relatively symmetrical, suggested no

publication bias in the meta-analysis of the link between

NAFLD and urolithiasis (Figure 3).

Figure 2. Forest plot of all studies

The present systematic review and meta-analysis investigated

the association between NAFLD and the risk of urolithiasis.

It was found that there was a significantly increased risk of

urolithiasis among patients with NAFLD. The present study

included four studies that evaluated the relationship between

NAFLD and the risk of developing urolithiasis. An increasing

number of studies have shown consistent evidence that the

presence of NAFLD was closely linked to a higher risk of

urinary stones that were detected on CT.

Figure 3. Funnel plot of all studies.

A cross-sectional study in Israel found a 3.24-fold increased

risk of CT diagnosed renal stones among NAFLD patients

than individuals without NAFLD [9]. Similarly, another

cross-sectional study which involved 1,381 patients revealed

that the prevalence of renal stone in patients with NAFLD

washes higher than those without NAFLD (OR:5, 95% CI,

3.05-8.2) [8]. Paz et al., (2015) also found an increased

prevalence of urolithiasis and NAFLD patients compared to

patients without NAFLD (OR:5.40, 95% CI, 1.17-24.92) [13].

A cross-sectional study examining a total of 1,010 also found

an increased incidence of urolithiasis in patients with NAFLD

(OR:1.29, 95% CI, 1.10-1.51) [14].

The mechanism underlying the increased risk of urolithiasis

in patients with NAFLD is not known with certainty.

However, several classic metabolic risk factor for NAFLD

have also been recognised as predisposing factors for urinary

stones as well hypertension, obesity, metabolic syndrome and

diabetes mellitus [15]. It has been reported that obesity might

influence and have association with NAFLD and urolithiasis.

It was suggested that fatty liver may possibly act as an

intermediate factor in getting urolithiasis [9]. However, data

on the association between obesity, NAFLD and urolithiasis

are limited. Only one study that investigated such a

relationship.

Oxidative and systemic metabolisms are compatible with the

formation of the stone [16]. Reactive oxygen species can

damage the kidneys due to abundance of long chain poly

unsaturated fatty acids in the renal lipid composition and

systemic oxidative species can result in peroxidation of lipid

that caused the formation of calcium oxalate stone [16].

Health Scope 9


Figure 1. Literature review process

Table 3. Characteristics of studies included in the meta-analysis.

Study/year Badillo et al /2020[14] Paz et al

/2015[13]

Nam et al

/2016[8]

Zeina et al

/2017[9]

Country Columbia Israel South Korea Israel

Study Design Cross-sectional study Cross-sectional study Cross-sectional study Cross-sectional study

Participants 1010 individuals aged 18

years and over

100 individuals 1381 individuals 508 individuals

Confounder adjusted

in multivariate

analysis

Sex and age None Sex and age Sex, age, visceral fat and

diabetes mellitus

Quality assessment

(Newcastle-Ottawa

scale)

Selection:5

Comparability:2

Outcome:3

Total:10

Selection:4

Comparability:2

Outcome:3

Total:9

Selection:4

Comparability:2

Outcome:3

Total:9

Selection:4

Comparability:2

Outcome:3

Total:9

PRISMA 2009 Flow Diagram

Records identified through database searching: 214 from Scopus,

36 from WoS, 45 from Dimensions and 200 from Science Direct.

(n = 495)

Screen

ing

In

clu

ded

E

ligib

ilit

y

Iden

tifi

ca

tio

n

Records after duplicates removed

(n = 24)

Records screened

(n = 471)

Records excluded

(n = 390) (excluded due to

review article, meta-

analysis, book, book

chapter and conference

proceeding, Non-English,

published in <2015)

Full-text articles assessed

for eligibility

(n = 81)

Full-text articles excluded,

with reasons

(n = 77) (excluded due to

inadequate data on

outcomes, different study

design, different

methodology used to

identify, did not report the

outcome of interest) Studies included in

quantitative synthesis

(meta-analysis)

(n = 4)

Health Scope 10

Oxidative stress and reactive oxygen species also have been

implicated in the pathogenesis of NAFLD [17]. The

involvement of oxidative stress and reactive oxygen species

in the patient with NAFLD and urolithiasis represents that

they might share the same pathogenesis [1]. Then, it has been

suggested that fatty liver may result in changes in urinary

constituents and leading to stone formation [18].

CT examination can also quantify the accumulation of fat in

the liver [8]. Data on the association between the severity of

NAFLD are limited. The authors divided into absence, mild

to moderate, and moderate to severe categories and found a

progressive simultaneous increase in the percentage of

urolithiasis with 71% for those in the mild to moderate range

and 75.7% for patients that have moderate to severe category

(p < 0.001) [14].

This present study only focused on CT examination to

diagnose NAFLD and urolithiasis. CT examination has high

sensitivity and specificity for detecting fatty liver disease and

urolithiasis than other imaging modalities such as ultrasound,

which can miss the disease of the urinary stone. It has 73-

100% sensitivity and 95-100% specificity for moderate to

severe NAFLD [19] while 99% of sensitivity and 94%

specificity in detecting urinary stone [20]. Thus, the current

study may avoid jeopardising the validity of their point

estimates.

This meta-analysis has limitations. All studies that included

were cross-sectional studies. So, the temporal relationship

between NAFLD and Urolithiasis could not be established.

Future studies of the population of all comers to non-contrast

CT are needed to better prove this association due to the

studies that evaluate this association is limited. Regarding

fatty liver disease, all studies included in the lack of clinical,

nutritional, associated pathologies, medical data information,

and lifestyle of the patients that may favor the genesis of these

diseases. Lastly, future studies are recommended to consider

these factors to determinate the cause of the relationship

between NAFLD and urolithiasis more accurately.

4. CONCLUSION

This study showed a significantly increased risk of urolithiasis among patients with NAFLD. The presence of NAFLD may be an independent variable as a risk factor for urolithiasis.

ACKNOWLEDGEMENTS

Authors thanks the librarian for assistance in resource searching.

REFERENCES

[1] S. Qin, S. Wang, X. Wang, and J. Wang, “Non-

alcoholic fatty liver disease and the risk of urolithiasis

A systematic review and meta-analysis,” Medicine

(Baltimore)., pp. 1–7, 2018.

[2] Y. Liu et al., “Epidemiology of urolithiasis in Asia,”

Asian J. Urol., vol. 5, no. 4, pp. 205–214, 2018.

[3] J. G. Fan, S. U. Kim, and V. W. S. Wong, “New trends

on obesity and NAFLD in Asia,” J. Hepatol., vol. 67,

no. 4, pp. 862–873, 2017.

[4] Y. Wei et al., “Epidemiologic Association of

Nonalcoholic Fatty Liver Disease and Urinary

Calculi : A Population-based Cross-sectional Study in

Southern China,” Iran. J. Kidney Dis., vol. 12, no. 2,

2018.

[5] M. Mouzaki et al., “Bile acids and dysbiosis in non-

alcoholic fatty liver disease,” PLoS One, vol. 11, no. 5,

pp. 1–13, 2016.

[6] L. Orlić, I. Mikolasevic, Z. Bagic, S. Racki, D. Stimac,

and S. Milic, “Chronic kidney disease and non-

alcoholic fatty liver disease - Is there a link?,”

Gastroenterol. Res. Pract., vol. 2014, 2014.

[7] J. Lubarda, S. Smith, and K. Johnson, “S-1236 non-

alcoholic fatty liver disease and urolithiasis: a

systematic review and meta-analysis,”

Gastroenterology, vol. 156, no. 6, p. S-1236, 2018.

[8] I. C. Nam, J. H. Yoon, S. H. Park, J. Ryu, S. H. Kim,

and Y. Lee, “Association of non-alcoholic fatty liver

disease with renal stone disease detected on computed

tomography,” Eur. J. Radiol. Open, vol. 3, pp. 195–

199, 2016.

[9] A. Zeina, L. Goldenberg, A. Nachtigal, R. Hasadia,

and W. Saliba, “Association between nephrolithiasis

and fatty liver detected on non-enhanced CT for

clinically suspected renal colic ☆,” Clin. Imaging, vol.

43, pp. 148–152, 2017.

[10] D. Moher, A. Liberati, J. Tetzlaff, D. G. Altman, and

P. Grp, “Preferred Reporting Items for Systematic

Reviews and Meta-Analyses: The PRISMA

Statement,” Phys. Ther., vol. 89, no. 9, pp. 873–880,

2009.

[11] R. Herzog, M. J. Álvarez-Pasquin, C. Díaz, J. L. Del

Barrio, J. M. Estrada, and Á. Gil, “Are healthcare

workers intentions to vaccinate related to their

knowledge, beliefs and attitudes? A systematic

review,” BMC Public Health, vol. 13, no. 1, 2013.

[12] J. P. T. Higgins, S. G. Thompson, J. J. Deeks, and D.

G. Altman, “Measuring inconsistency in meta-

analyses,” Br. Med. J., vol. 327, no. 7414, pp. 557–

560, 2003.

[13] D. Paz and L. Guralnik, “Association of renal stone (

urolithiasis ) with non-alcoholic fatty liver ( NAFL ),”

Electron. Present. Online Syst. ECR 2015 ESR, pp. 1–

11 Health Scope, 2020, Vol. 3(3) Nursyahirah et al.

10, 2015.

[14] F. G. L. Badillo, O. L. O. Cala, S. N. Vera Campos,

and E. D. Villarreal Ibañez, “Relationship Between

Urolithiasis and Fatty Liver Disease: Findings in

Computed Tomography,” Tomogr. (Ann Arbor,

Mich.), vol. 6, no. 1, pp. 1–4, 2020.

[15] H. Besiroglu, A. Otunctemur, and E. Ozbek, “The

metabolic syndrome and urolithiasis: A systematic

review and meta-analysis,” Ren. Fail., vol. 37, no. 1,

pp. 1–6, 2015.

[16] P. Schwille, A. Schmiedl, and J. Wipplinger,

“Idiopathic recurrent calcium urolithiasis (IRCU):

variation of fasting urinary protein is a window to

pathophysiology or simple consequence of renal stones

in situ?,” Eur. J. Med. Res., vol. 14, no. 9, p. 378, 2009.

[17] M. Morita et al., “Fatty liver induced by free radicals

and lipid peroxidation,” Free Radic. Res., vol. 46, no.

6, pp. 758–765, 2012.

[18] B. Einollahi, M. R. Naghii, and M. Sepandi,

“Association of non-alcoholic fatty liver disease (

NAFLD ) with urolithiasis,” Endocr. Regul., vol. 47,

no. January, pp. 27–32, 2013.

[19] R. Alshaalan, M. Aljiffry, S. Al-Busafi, P. Metrakos,

and M. Hassanain, “Non-alcoholic fatty liver disease:

Noninvasive methods of diagnosing hepatic steatosis,”

Saudi J. Gastroenterol., vol. 21, no. 2, pp. 64–70,

2015.

[20] W. Brisbane, M. R. Bailey, and M. D. Sorensen, “An

overview of kidney stone imaging techniques,” HHS

Public, vol. 13, no. 11, pp. 654–662, 2017.

Health Scope 12


RESEARCH ARTICLE

Early detection of pulmonary vessel nodule in low dose CT image using local adaptive thresholding

Sufia Mohd Hizam, Noor Shafini Mohamad*



Abstract:

Small lung nodules are very subtle in the medical images, and less than 30% of them are connected

to the pulmonary vessels. Most of the time, the findings of this disease are not optimal at the early

stages. These small lung nodules have similar HU values as pulmonary vessels, therefore, making

it a challenge to separate these nodules. This study aimed to segment and suppress pulmonary

vessels and detected nodules to improve the accuracy of diagnosing lung cancer by using local

adaptive thresholding. This proposed framework consisted of the image enhancement process and

three segmentation stages. Contrast stretching, median filter combined with closing morphological

operator, and unsharp masking were employed to make the image more appealing. The first stage

of image segmentation was extracting lung from the parenchyma by using a fast marching method

and active contour. The second stage was to extract pulmonary vessels and nodules together using

local adaptive thresholding. Extraction of the nodule from the pulmonary vessels using local

adaptive thresholding was employed in the final stage. The sensitivity and specificity of this method

were computed by calculating the number of pixels overlapped with the ground truth images. This

proposed method presented high sensitivity and specificity for segmentation of pulmonary nodule

(0.90 and 0.99) and segmentation of pulmonary vessels (0.87 and 0.99). After the suppression of the

vessels, the mean CNR of the nodule increased from 3.27 to 3.61). The suppression and

segmentation of pulmonary vessels in CT thorax images may reduce false-positive findings and

misdiagnosis due to human error. Hence, the early discovery of lung nodules can reduce about half

of the mortality rate.




Keywords: Image segmentation, local adaptive thresholding, pulmonary vessel nodule

1. INTRODUCTION

According to the World Health Organisation report in 2014, lung cancer accounted for 19.1 fatalities per 100,000 population in Malaysia [1], the second leading cause of death after breast cancer. The Global Cancer Observatory report in 2018 mentioned that the incidence of new cases was 4, 686, the number of deaths was 4,057 [2]. In the latest 5-years observations, males accounted for the most incidence rate (13.2 per 100 000) compared to females (5.9 per 100 000) [3]. [1]. The gold standard of lung cancer screening tests for high-risk populations to detect at an early stage recommended by the Ministry of Health Malaysia [4] is low-dose computed tomography (LDCT). Due to inadequate participation from the public at large, a population-based screening test is not fully implemented yet [5]. Besides that, this disease is only detected when it has progressed, and most patients in Malaysia are identified with an advanced stage of cancer [4].

CT scan creates numerous cross-sectional slices, and the radiologist interprets these slices every day. By doing so in a repeated manner, it can cause exhaustion and unable to focus,

thus may increase the chance of imprecise lung cancer detection and diagnosis due to human observation error [6]. Consequently, a Computer-Aided Diagnosis (CAD) system was introduced to solve this problem by assisting radiologists in making accurate reports on the findings. In Malaysia, not hospitals have fully implement CAD in their radiology department to diagnose cancers. In some cases, CAD may be too expensive and take time to learn and train radiologists to adopt this unfamiliar technology [7]. Therefore, Malaysia radiologists will depend on their visual evaluation to diagnose cancer. This situation can cause misdiagnosis because of factors linked to imaging quality and human error owing to appearance structural [8].

Li et al., as cited by Gu et al. (2019), specified that pulmonary nodules are very subtle in the medical images. Less than 30% of them are connected to the pulmonary vessels, which makes them easily overlooked [6]. Lo et al. (2018) hold the view that these pulmonary vessels can be mistaken as nodules because they have similar CT values and cross-sectional shapes as nodules do [9]. Furthermore, Gu et al. (2019) reported that CAD's grey-level intensity thresholding tends to over-

13 Health Scope, 2020, Vol. 3(3) Sufia et al.

segmented these nodules as separating the vessels from the nodules are quite a challenge [6].

Mustafa et al. (2018) stated that local adaptive technique yields a precise outcome as it reduces image noise around the edge regardless of the image was caused by reduced shading, ghosting, low resolution, and non-uniform lighting [10]. Assigning a distinct threshold for each pixel, based on the neighboring pixel in greyscale data is what local adaptive thresholding can do. There is a lack of research done to prove the capability of local adaptive thresholding in the segmentation of pulmonary vessels nodules. Therefore, this study aimed to segment and suppress pulmonary vessels to detect nodules, improving the accuracy of diagnosing lung cancer by using local adaptive thresholding in MATLAB. This alternative mechanism might assist the radiologist in the detection of lung cancer concurrently, avoiding misdiagnosis.


This study was an experimental study where the early detection of pulmonary vessels nodules from 15 LDCT thorax images using local adaptive thresholding was investigated. These images were retrieved from VIA/I-ELCAP Public Access Research Database and saved to the personal laptop in DICOM format (Figure 1). The CT helical scans were collected using a slice thickness of 1.25mm at 120 kVp in a single breath-hold. It also includes the nodules annotation identified by experienced radiologists. The images size was 512 x 512 x 280 with 0.71 x 0.71 x 1.25 resolution. MATLAB R2018b was used to process the selected images in terms of image enhancement and image segmentation.

Figure 1. Research flowchart

2.1 Pre-processing steps

Most DICOM format images came out with low contrast and grainy with minimal visibility of anatomy structures. Thereby, the contrast of the image was adjusted by computing contrast stretching. This technique widened the input image has a set of pixel values to cover a more excellent dynamic range in the output image. After contrast enhancement, the low dose CT images were appeared noisy due to its low

exposure settings causing insufficient photons delivered and detected. A median filter and closing morphological operator with a size 3 square structuring element were operated to remove this noise from the input images. Subsequently, the image was sharpened by using unsharp masking (Figure 2).

Figure 2. From left, the figure shows the original LDCT image, after the image enhancement, and the segmentation of

lungs.

2.2 Image segmentation

The image segmentation framework in this paper comprised of three stages. The first stage was to obtain the lung regions from the thorax in the image by implementing a fast-marching method. Then, an active contour edge-based method was applied to get smoother edges of the lung image (Figure 2). The second stage was to segment high-intensity structures such as pulmonary vessels along with nodule attached using local adaptive thresholding (Figure 3). A sensitivity factor was chosen to a threshold as many pulmonary vessels in the image as possible. The higher the sensitivity factor, the more pixels will be selected as the foreground. For the third stage, a lower sensitivity factor than before was selected to separate the nodule from the pulmonary vessel. The nodule was extracted according to the annotation done by the radiologist in the public database. Then, dilation morphology was applied to the solitary nodule (Figure 4). As a result, two separate binary images of pulmonary vessels and pulmonary nodules alone were obtained.

Figure 3. The image shows the segmentation of pulmonary vessels and nodules from the lung image. The red arrow

points out the location of the nodule.

Health Scope 14

Figure 4. From left, a close-up image shows the nodule (red boundary by ground truth and green boundary by local

adaptive thresholding) attached to the pulmonary vessels. The middle image shows the binary image. The right image

displays the extracted nodule

2.3 Removal of pulmonary vessels

After the segmented pulmonary vessel image was acquired, it was subtracted with the lung image and filled the holes of the subtracted image. The next step was smoothened the lung whole region with the median filter to make it appear more natural and homogenous. The solitary nodule image was added to the lung image to make the nodule more highlighted than the surrounding parenchyma (Figure 5).

Figure 5. The left image shows the lung image before and after vessels (green arrow) suppression. Only nodule (red

arrow) is preserved in the lung

2.4 Objective assessment

2.4.1 Peak Signal-to-Noise Ratio

The peak signal-to-noise ratio (PSNR) of both raw data and the pre-processed image was calculated. The PSNR results between the raw data and pre-processed images were compared using Paired T-test on SPSS. It can be expressed in decibel (dB) unit as:

2.4.2 Contrast-to-Noise Ratio

The contrast-to-noise ratio (CNR) of both lung image and pre-processed image were accessed, as well as the CNR nodule before and after removal of pulmonary vessels. The comparison was computed using the Paired Sample T-test in

SPSS. CNR was calculated manually in decibel (dB) unit using this formula:

2.4.3 Confusion matrix

A confusion matrix was constructed in MATLAB to measure the performance of our method as compared to the ground truth image. Ground truth image is a gold standard that has actual outcomes. In this case, ground truth images were not available. The Otsu thresholding method was utilised to get two different masks of pulmonary vessels and nodule as ground truths. This confusion matrix consists of:

True positive (TP): the number of foreground pixels is identified as a part of both segmented image and ground truth image. False-positive (FP): the number of foreground pixels is identified as a part of the segmented image but are not identified in the ground truth image. True negative (TN): the number of negative pixels is not identified in both segmented and ground truth image (background pixels). False-negative (FN): the number of negative pixels is not identified in the segmented image but identified as a part of the ground truth image.

From these values, the sensitivity (TP/TP+FN), specificity (TN/TN+FP), positive predictive value (TP/TP+FP), negative predictive value (TN/TN+FN), DICE (2TP/(2TP+FP+FN)), and Jaccard (TP/(TP+FN+FP) were calculated. The sensitivity was concerned with several positive values that were detected as positive—the specificity considered on the number of negative values that were detected as negative. DICE and Jaccard were achieved to evaluate the similarity of area overlapped with the ground truth.


Table 1 shows the mean PSNR of original images, 19.8284±0.0398, while the mean PSNR of pre-processed images was 20.4999±0.2766. There was a significant difference between the PSNR of the original image and PSNR of the pre-processed image (-0.6715, p<0.001). This implies that there was an increase in PSNR value when the noise removal method was applied to the original image.

The mean value of the original image CNR was -14.8771±3.8146, whereas the mean value of pre-processed image CNR was 0.2366±0.0649. There was a significant difference between the CNR of the original image and CNR of the pre-processed image (-15.1137, p <0.001). This indicates that there was an increase of mean CNR value when the contrast enhancement technique was applied to the images.

Table 1. The summary of the Paired Sample T-test of PSNR and CNR of the original image and PSNR of pre-processed

image.


The median filter can remove deviations with huge magnitudes while maintaining the degree of image edges. Thus, there is less blurring in the image. An experiment was done by Muthamil Selvi & Ashadevi (2020) on evaluating impulse and Gaussian noise suppression from CT lung images by comparing different smoothing filters has shown that the median filter had higher PSNR values than the mean and wiener filter [11]. Another study on CT lung images has been done to detect edges and denoise by using a combination of various morphological operators. The findings showed that the morphology algorithm had the highest PSNR value than Canny, Sobel, and the novel mathematic morphological algorithm [12].

Table 2 shows the CNR mean value of lung nodule with the presence of lung vessels was 3.2658±0.6992, and the CNR mean value of lung nodule without the presence of lung vessels was 3.6111±0.6515. There was a significant difference between both mean CNR values (-0.3453, p= 0.004). This suggests that there was an improvement of CNR nodule value when the lung vessels were suppressed from the lung parenchyma.

Table 2. The summary of the Paired Sample T-test of PSNR and CNR of the original image and PSNR of pre-processed image.

When calculating for the nodule CNR, the background noise assigned was the lung parenchyma that included the high-intensity pulmonary vessels. This is because the background might not be entirely uniform and contain noise that resulted in low CNR (3.2658±0.6992). After the suppression of pulmonary vessels, the empty holes were filled. Consequently, the lung parenchyma became homogenous, and the nodule preserved in its location. Now, the images display pulmonary nodules, where they are the only high-intensity object presented compared to that of surrounding parenchyma (CNR=3.6111±0.6515, p= 0.004).

Gu et al. (2019) extracted pulmonary vessels from the lung parenchyma by using a two-cascaded convolutional neural network (CNN) and filled in the void with the same voxels as the parenchyma to make it looked more natural [6]. The mean CNR nodule in their study increased from 4.23 to 6.95 (39.14%). Local adaptive thresholding had an increased CNR of 9.42% (3.27 to 3.61). The two-cascaded CNN was robust and was able to extract pulmonary vessels accurately. Therefore, more natural-looking parenchyma was produced that may contain less background noise. In the current study, the 9.42% CNR improvement showed that pulmonary nodules are more highlighted when the pulmonary vessels are suppressed.

Table 3. The summary of the nodule segmentation confusion matrix.

As stated by Pont-Tuset & Marques (2016), a small sensitivity score is the product of under-segmentation and implies the inability to detect distinctive shape [13]. Small PPV score and high false detection of object boundary pixels imply the over-segmentation technique.

Table 3 shows that local adaptive thresholding yielded a sensitivity of 90.67% in segmenting pulmonary vessel nodule. Liu et al. (2019) used a robust transition region-based thresholding to extract 50 nodules images that were attached to the pleura and compared the results with ground truth images generated by radiologists [14]. This thresholding method selected the intermediate pixels of 0.85 in between the foreground and the background region. The sensitivity of their method was 86.65%. This shows that local adaptive thresholding had high sensitivity; therefore, under-segmentation was unlikely to occur. Local adaptive thresholding was able to delineate the shape of the pulmonary nodule as close as the ground truth, even if the nodule was attached to the pulmonary vessels. This is because the local adaptive threshold manipulates a local pixel around the neighborhood instead of the whole region (Figure 4).

On the other hand, our overall PPV score was 0.64. This implies that our proposed method has average PPV; as a result, slight false detection of pulmonary vessel nodule boundary occurred. The overall specificity was 0.99. This indicates that the division between background and the pulmonary nodule was accurate [15]. The overall DICE and Jaccard score of this proposed method were ranged from 0.74 and 0.60 (Table 3).

Health Scope 16

Table 4. The summary of vessel segmentation confusion matrix.

Regarding Table 4, the overall sensitivity of local adaptive thresholding in segmenting pulmonary vessels was 0.87, and the PPV was 0.60. A study by Zhai, Staring, & Stoel (2016) implemented graph-cuts algorithm to segment pulmonary vessels from CT images [16]. The authors enhanced the vessels with a Hessian-based filter and assigned foreground and background pixels by altering graph cuts parameters with a modified Gaussian function [16]. The graph cuts method held a sensitivity score of 0.73 and PPV of 0.79. Local adaptive thresholding had higher sensitivity; therefore, the delineation of the pulmonary vessel shape was as close as the ground truth image.

On the contrary, the overall PPV score was lower (0.60); as a result, false detection of pulmonary vessel boundary occurred. The sensitivity factor of local adaptive thresholding influences how many foreground objects are thresholded. Therefore, the higher the sensitivity factor, the more local pixels are selected to be foregrounds. Local adaptive thresholding with high sensitivity factor able to threshold more large and small pulmonary vessels as compared to ground truth image; because of that, over-segmentation of the pulmonary vessels occurred.

DICE and Jaccard assess any similarities between the predicted segmentation algorithm and the ground truth algorithm. Both metrics hold on a set of agreements in terms of 0 and 1. If the score of the overlapping area is close to 0, there is no overlapping occurs. If the score of the overlapping area is close to 1, the area in both the predicted algorithm and ground truth algorithm is said to be overlapped [13]. In this case, local adaptive thresholding yielded an overall DICE score of 0.70 and an overall Jaccard score of 0.55 in segmenting pulmonary vessels (Table 4). To appraise the effectiveness of local adaptive thresholding in segmenting pulmonary vessels, the current results were compared with Gu et al. (2019) [6]. These authors used two cascaded CNN that instantly convert with image data as information and automatically produce voxel-based attributes that separate the pulmonary vessels and pulmonary nodules from the surrounding parenchyma. [6] method yielded a higher DICE

score of 0.94 and the Jaccard score of 0.89, which was lower than the current study. Local adaptive thresholding selected more pulmonary vessels as foreground pixels than the ground truth method. Consequently, more areas in the predicted segmentation image did not accurately overlap with the actual segmentation image.

4. CONCLUSION

The segmentation framework in this paper utilised local adaptive thresholding to separate the nodule from pulmonary vessels. The method presented high sensitivity and specificity score for both nodule and pulmonary vessel segmentation. Local adaptive thresholding can be an alternative tool to improve the process of segmentation of pulmonary vessels nodules that are smaller than 30mm. The suppression and segmentation of pulmonary vessels in CT thorax images may reduce false-positive findings and misdiagnosis due to human error. Hence, the early discovery of lung nodules can reduce about half of the mortality rate. Further study is recommended to apply this experiment to the clinical site to assist radiologists in decision-making. A neural network can be utilised to classify lung cancers into benign and malignant. This way, the workload can be reduced, and lung cancer can be detected early.

ACKNOWLEDGEMENTS

Authors would like to thank Dr. Noor Elaiza Abd Khalid for assistance in MATLAB protocols.

REFERENCES

[1] K. C. Siang and C. K. M. John, "A review of lung cancer

research in Malaysia," Med. J. Malaysia, vol. 71, no. June,

pp. 70–78, 2016.

[2] The Global Cancer Observatory, "Malaysia - Cancer fact

sheet," World Heal. Organ., vol. 593, pp. 1–2, 2019.

[3] Azizah AM. et al., Malaysia National Cancer Registry

Report 2012-2016. 2019.

[4] Malaysian Health Technology Assessment Section, "Low

Dose Computed Tomography for Lung Cancer Screening,"

Heal. Technol. Assess. Rep., p. 122, 2017.

[5] P. Rajadurai, S. H. How, C. K. Liam, A. Sachithanandan,

S. Y. Soon, and L. M. Tho, "Lung Cancer in Malaysia," J.

Thorac. Oncol., vol. 15, no. 3, pp. 317–323, 2020.

[6] X. Gu, J. Wang, J. Zhao, and Q. Li, "Segmentation and

suppression of pulmonary vessels in low‐dose chest CT

scans." p. 12, 2019.

[7] J. Shiraishi, Q. Li, D. Appelbaum, and K. Doi, "Computer-

aided diagnosis and artificial intelligence in clinical

imaging," Semin. Nucl. Med., vol. 41, no. 6, pp. 449–462,

2011.

[8] A. Jalalian, S. Mashohor, R. Mahmud, B. Karasfi, M. Iqbal

Saripan, and A. R. Ramli, "Computer-Assisted Diagnosis

System for Breast Cancer in Computed Tomography Laser

Mammography (CTLM)," J. Digit. Imaging, vol. 30, no. 6,

pp. 796–811, 2017.

[9] S. C. B. Lo, M. T. Freedman, L. B. Gillis, C. S. White, and

S. K. Mun, "Computer-aided detection of lung nodules on

CT with a computerized pulmonary vessel suppressed

function," Am. J. Roentgenol., vol. 210, no. 3, pp. 480–488,

2018.


[10] W. A. Mustafa, H. Yazid, and M. Jaafar, "An improved

sauvola approach on document images binarization," J.

Telecommun. Electron. Comput. Eng., vol. 10, no. 2, pp.

43–50, 2018.

[11] P. Muthamil Selvi and B. Ashadevi, "Elimination of noise

in CT images of lung cancer using image pre-processing

filtering techniques," Int. J. Adv. Sci. Technol., vol. 29, no.

4 Special Issue, pp. 1823–1832, 2020.

[12] M. Kumar and S. Singh, "Edge Detection and Denoising

Medical Image using Morphology," Int. J. Eng. Sci.

Emerg. Technol., vol. 2, no. 2, pp. 66–72, 2012.

[13] J. Pont-Tuset and F. Marques, "Supervised Evaluation of

Image Segmentation and Object Proposal Techniques,"

IEEE Trans. Pattern Anal. Mach. Intell., vol. 38, no. 7, pp.

1465–1478, 2016.

[14] J. Liu, J. Gong, L. Wang, X. Sun, and S. Nie,

"Segmentation refinement of small-size juxta-pleural lung

nodules in CT scans," Iran. J. Radiol., vol. 16, no. 1, 2019.

[15] A. A. Taha and A. Hanbury, "Metrics for evaluating 3D

medical image segmentation: Analysis, selection, and

tool," BMC Med. Imaging, vol. 15, no. 1, 2015.

[16] Z. Zhai, M. Staring, and B. C. Stoel, "Lung vessel

segmentation in CT images using graph-cuts," Med.

Imaging 2016 Image Process., vol. 9784, p. 97842K, 2016.

Health Scope 18


REVIEW ARTICLE

Diagnostic accuracy of diffusion weighted magnetic resonance imaging

(DW-MRI) and transvaginal ultrasonography (TVUS) in assessing

myometrial invasion among female patients with endometrial cancer: a

systematic review and meta-analysis

Sofea Sarina Mazrul, Mohd Zulfadli Adenan*

Centre of Medical Imaging, Faculty of Health Sciences, Universiti Teknologi MARA Cawangan Selangor Kampus


Abstract:

TVUS is the first imaging technique in determining patient with suspicious of endometrial cancer

whereas the most appropriate tool for grading of endometrial cancer is MRI. Recently, DW-MRI

gradually becomes a routine in grading endometrial cancer. Thus, the purpose of this study was to

compare the diagnostic accuracy of transvaginal ultrasound (TVUS) and diffusion weighted magnetic

resonance imaging (DW-MRI) in assessing depth of myometrial invasion in endometrial cancer. An

extensive search was performed in Dimension, PubMed and Wiley Online Library guided by

PRISMA. Following screening, the eligibility of the studies was checked. Risk of bias was assessed

using the Quality Assessment of Diagnostic Accuracy Studies-2 (QUADAS-2) checklist. The search

identified 7 studies (727 patients). Overall, pooled estimated sensitivity and specificity for detecting

deep myometrial invasion were 79% (95% confidence interval [CI]=72%–85%) and 84% (95%

CI=64%–94%) for TVUS, and 84% (95% CI=69%–92%) and 89% (95% CI=83%–93%) for DW-

MRI, respectively. Low heterogeneity was found for sensitivity of TVUS but high for the specificity.

For DW-MRI, the heterogeneity for specificity was low but the heterogeneity was significant for the

sensitivity. In conclusion, the sensitivity and specificity of DW-MRI is higher compared to TVUS in

assessing the deep myometrial invasion among women with endometrial cancer. Keywords: Endometrial cancer, myometrial invasion, magnetic resonance imaging, transvaginal ultrasonography


Mohd Zulfadli Adenan


1. INTRODUCTION

Endometrial cancer also called as endometrial

carcinoma or corpus cancer starts when cells of the inner

lining of the uterus begin to grow out of control [1]. In

United States, the endometrial cancer is the most common

gynecologic malignancy and the American Cancer Society

predicts about 65,620 new cases will be diagnosed with this

cancer and 12,590 deaths is estimated in 2020[2].

Endometrial cancer can be cured by surgery especially in the

early stages[3]. The prognosis and the treatment response of

patients with endometrial cancer depends on several factors

such as the cell differentiation, pathological grade, cervical

stromal invasion and the depth of myometrial invasion[4].

Besides early detection, presurgical evaluation of the depth

of myometrial invasion is important to improve the patient

treatment management[5]. There are two categories for depth

of myometrial invasion which are superficial invasion

(invasion depth of <50% of the myometrial thickness) and

deep invasion (extension ≥50% of the myometrium).

Magnetic resonance imaging (MRI), computed tomography (CT), transvaginal ultrasound (TVUS) and positron emission tomography (PET) are the noninvasive imaging tools to assess the depth of myometrial invasion. Transvaginal ultrasound (TVUS) and magnetic resonance imaging (MRI) are currently the most common imaging technique used for evaluating preoperatively the depth of myometrial infiltration (MI)[5]. In 1999, there was a meta-analyses comparing diagnostic accuracy of several imaging techniques in assessing preoperative radiological staging of myometrial infiltration and in early 2000 there was a meta-analysis conducted that assessed the role of DCE-MR(dynamic contrast enhancement) imaging in detecting the deep of myometrial invasion. As DW-MR imaging has been gradually practical in assessing myometrial invasion in last few years, there is a need for an updated systematic review

mailto:[email protected]

19 Health Scope, 2020, Vol. 3(3) Sofea Sarina et al.


and meta-analysis including this new generation imaging. The purpose of the present meta-analysis is to compare the diagnostic accuracy of TVUS and DW-MRI in detecting myometrial invasion in female patients with endometrial cancer.

2. METHODS

2.1 PRISMA

The systematic review and meta-analysis was done

by referring to Preferred Reporting Items for Systematic

reviews and Meta-Analyses (PRISMA)[6]. All methods for

inclusion/exclusion criteria, data extraction and quality

assessment were specified in advance.

2.2 Data sources and searches

The following electronic databases were used in this

study to search the related articles; Medline/PubMed,

Dimension and Wiley Online Library. Publication years

were filtered from 1st January 2011 up to 1st March 2020 in

searching. Methodological filters were not performed in

searching to avoid potential omission of relevant studies.

The search strategy included database subject headings and

text words as follows: EC or endometrial cancer,

endometrial carcinoma, myometrial invasion or infiltration,

ultrasound, transvaginal ultrosonography, TVUS, magnetic

resonance imaging and diffused weighted image or DWI.

The abstract of all relevant articles on the database were

reviewed and their references were also checked for other

relevant publications. Language restriction in the search was

set to English.

2.3 Study selection and data collection

In order to exclude clearly unrelated articles, the title

and abstracts identified by the searches was screened first by

two reviewers. For instance, those that not strictly associated

to the topic under review will be excluded. The reference

lists of selected articles also were screened for search

completion. Then, relevant full texts were obtained to

recognize possibly eligible studies and the same reviewers

independently reviewed the full text for inclusion criteria.

The following inclusion criteria were applied to the full text:

• Type of studies: Prospective or retrospective cohort

studies

• Type of participants: Adult women with biopsy-

proven primary adenocarcinoma of the

endometrium, undergoing preoperative detection of

deep myometrial invasion prior to surgery. Patients

with any stage of the disease were included.

• Target condition: Presurgical detection of deep

myometrial invasion in primary endometrial

adenocarcinoma

• Index test(s): TVUS and DW-MRI (DWI without

referring to fusion with other sequences) were

regarded as index tests because these tests are

usually used to assess depth of myometrial invasion

in patient with endometrial cancer. In addition,

other index tests that are usually detecting

myometrial invasion were also included.

• Types of outcome: The primary outcomes were

sensitivity (SEN), specificity (SPE), positive

predictive value (PPV), negative predictive value

(NPV), and their respective 95% confidence

intervals.

• Reference Standard: Pathological assessment of the

presence of deep myometrial invasion on the uterus

removed at surgery

Disagreements that were generated in the process of study

selection were resolved by consensus. The Patients,

Intervention, Comparator, Outcomes, Study Design (PICOS)

criteria were used for describing the studies include.

2.4 Risk of bias in individual studies

Quality Assessment of Diagnostic Accuracy

Studies-2 (QUADAS-2) is a tool that used to assess the

quality of the studies and this tool had been adapted in this

study. There are four domains in the QUADAS-2 format: 1)

patient selection, 2) index test, 3) reference standards, 4)

flow and timing. The possibility of bias and concerns about

applicability were evaluated in each domain (the latter not

applying to the 4th domain) and graded as low, high or

unclear risk. The outcomes of quality assessment were used

for descriptive purposes to explore potential sources of

heterogeneity and to provide an assessment of the overall

quality of the included studies.

The quality of methodology was evaluated independently by

the two reviewers, using a standard form with quality

assessment criteria and a flow diagram; discussion had been

done to resolve the disagreements.

2.5 Statistical analysis

Information on diagnostic accuracy of both TVUS

and DW-MRI from each primary study was extracted. All

studies had a reference standard patient (at least 50% of

myometrial invasion according to surgical pathological data)

and all considered as positives. The primary outcome was

reported in term of pooled specificity, sensitivity, positive

likelihood ratio (LR+) and negative likelihood ratio (LR-)

and of diagnostic odds ratio (DOR), together with their 95%

confidence intervals (CI). P-value based on the likelihood

ratio test were provided (α=0.05, two-sided). LR+ and LR-

were applied to characterize the clinical utility of test and to

assess the post-test probability of disease. Weak evidence

was provided when value of LR is 0.2-5.0 for either ruling

out or approving the disease. Moderate evidence was

Health Scope 20


achieved when the LR is 5.0-10.0 and 0.1-0.2. For LR >10 or

<0.1, this value provided strong evidence to either rule out or

confirm the disease.

Heterogeneity for sensitivity and specificity were explored

graphically thru constructing forest plots of sensitivity and

specificity of each primary study. Then, they were plotted in

the hierarchical summary receiver operating characteristic

(HSROC) space, the latter to identify whether any

heterogeneity could be attributable to an implicit threshold

effect. Besides, HSROC curves for each technique were

plotted to illustrate relationship between sensitivity and

specificity. Means of a test on the Q statistic and I2 index

were calculated to assess the presence of heterogeneity. A p-

value <0.1 points to heterogeneity.

The I2 index was measured to define the percentage of

overall variation across studies that are due to heterogeneity

rather than chance. The I2 value of 25%, 50% and 75%

would be considered to specify low, moderate and high

heterogeneity, respectively. Comparison of effect of the

TVUS and DW-MRI for detecting deep myometrial invasion

was done using the bivariate method. All analyses were

performed using Meta-analytical Integration of Diagnostic

Accuracy Studies (MIDAS) and (METANDI) commands in

STATA version 13.0.

3. RESULT AND DISCUSSION

3.1 Search result

The search of Dimension, PubMed and Wiley

Online Library databases provided a total 201 citations.

After removal of duplicate records, 144 citations were

remained. Of these, 133 were discarded because it was clear

from the title and abstract that they did not meet the criteria.

The remaining 11 papers were examined. Finally, 4

studies[5], [7]–[9] were excluded because these articles did

not meet inclusion criteria and the remaining 7 studies[10]–

[16]were included in the review and meta-analysis. There

were no additional related studies were found from

references cited in the articles included in the review. Figure

1 below is a flowchart summarizing literature identification

and selection.

Figure 1: Flow chart summarizing literature

identification and selection.

3.2 Characteristics of included studies

A total of 7 studies published between January

2012 to December 2017 reporting on 727 patients were

included in the final analyses. Among these 727 women, 287

had deep myometrial invasion. Mean patients’ age were

reported in 6 out of 7 studies and ranged from 31-89 years.

Two of the studies were retrospective studies and the

remaining were prospective studies. Three studies used

TVUS to assess the depth of myometrial invasion and

another three studies used DW-MRI to detect the myometrial

invasion. Only one study evaluates the depth of myometrial

invasion using both TVUS and DW-MRI on the same

patients. Table 1 shows PICOS features of studies included.

All of the studies were based on the radiologist’s or

physician’s impression.

The technical aspects of MR protocols are described in Table

2. Two studies used 3T MR and another two studies used

1.5T. All studies were used surface coil. Two studies [11],

[16]performed fused T2 and DW imaging to interpret DW

and the remaining two interpret DW imaging alone (one at

Records identified through database

searching (Dimension, Pubmed, Wiley

Online Library) (n= 201)

Records after

duplicates removed

(n= 144)

Records screened

(n= 144)

Records excluded

(n= 133)

Full-text articles assessed for eligibility

(n= 11)

Studies included in

qualitative synthesis (n= 7)

Studies included in quantitative synthesis (meta-

analysis)

(n = 7)

Full-text articles excluded

(n= 4):

- Patient did not have

biopsy-proven, only

involving suspected

cancer patients in the

study [5]

-Studies not using DWI

and using 3-

dimensional ultrasound

[7]

-A systematic review

paper [8]

-Local tumors extend

evaluation, not

myometrial invasion

[10]



1.5T and one at 3T). No adverse events were reported during

MR examinations. Concerning possible drawbacks for MR

imaging, only one study described benign pathology found at

MR examination.

For TVUS, one of the four studies[12] used objective

measurement, Gordon’s approach to assess myometrial

invasion. In 3 studies [10], [13], [14]myometrial invasion

assessment was based on the examiner’s subjective

impression. The major patient features in these studies are

not provided. There was no patient underwent oncological

therapy between TVUS and surgery.

Table 1: Characteristics of included patients and studies

according to PICOS. Study Study

design

No.

of

patient

No.

≥50

%

MI

Age (years) Obser-

vers

TVUS

Obser-

vers

DW

-MRI

Mean Min

-max

(Savelli et

al., 2012)

Pro-

155 76 63 32

-88

Multi-

ple

NA

(Hori et

al.,2013)

Pro-

71 19 58 31

-82

NA Multi-

ple

(Ørtoft et

al., 2013)

Pro-

156 66 66 32

-88

Multi-

ple

NA

(Christens

en et al.,

2016)

Pro-

110 47 69 32

-85

Single NA

(Angioli et

al., 2016)

Pro-

79 27 53 47

-78

Multi-

ple

Multi-

ple

(Rodrı et

al., 2016)

Retro-

98 39 NA NA NA Single

(Guo et

al., 2017)

Retro-

58 13 60 45

-78

NA Single

Pro-, Prospective; Retro-, Retrospective

Table 2: Technical characteristics of the MR protocols used

in the included studies. Study Tesla Mark

b

value

Thick

-ness

(mm)

FOV

(mm)

Matrix T2-

fused

(Hori et

al., 2013)

3.0 1,000 4 200 96 x 96

Yes

(Rodrı et

al., 2016)

1.5 800 - - - No

(Rodrı et

al., 2016)

1.5 800 - - - No

(Guo et

al., 2017)

3.0 - - 200 256 x

320

Yes

3.3 Methodological quality of included studies

Figure 2 shows the evaluation of the risk of bias and

concerns regarding the applicability of the selected studies.

Concerning the risk of bias and patient selection domain,

only one study[15] did not report patient selection criteria.

This study was considered as risk bias since it did not avoid

inappropriate exclusion criteria. Three TVUS studies

sufficiently described the method of index test and how it

interpreted or performed while one study was unclear [14].

For DW-MRI, two studies unclearly described the method of

index test and how it interpreted or performed [14], [15].

However, another two studies adequately described them

[11], [16]. Same threshold was adopted (≥50% of

myometrial thickness) to define the deep myometrial

invasion.

For the domain reference standard, all studies were

unclear to describe the reference standard and one study was

undertaken as low bias [16]. However, all these studies

reported that the observers were blinded to imaging results.

Regarding the flow and timing domain, only three studies

report the interval between TVUS [12], [13]or MRI [16] and

surgery, the rest of the studies did not report the information.

For patient selection, one study[15] concerned that

included patient did not match the review question. For

index test domain, also one study[13] unclearly reported

whether there was concern that the TVUS, its conduct differ

from the review question of the study. Concerning the

reference standard domain of applicability, all studies were

in low risk of bias.

Figure 2 : Quality assessmentof all studies included in the meta-analysis, according to QUADAS-2 (Quality

Assessment of Diagnostic Accuracy Studies-2) criteria.

3.4 Diagnostic performance of TVUS and DW-MRI in detecting deep myometrial invasion

Individual results of TVUS and DW-MRI are shown

in chronological order on the forest plots in Figure 3.

Health Scope 22


Overall, pooled sensitivity, specificity, LR+, and LR− of

TVUS for detecting deep myometrial invasion were 79%

(95% confidence interval [CI]=72%–85%), 84% (95%

CI=64%–94%), 5.03 (95% CI=1.90–13.29), and 0.25 (95%

CI=0.16–0.38), respectively. Low heterogeneity was

establish for sensitivity (I²=32.24%; Cochran Q=4.43;

p=0.22). However, high heterogeneity was found for

specificity (I²=80.12%; Cochran Q=15.16; p=0.00). On the

other hand, pooled sensitivity, specificity, LR+, and LR−

of DW-MRI for detecting deep myometrial invasion were

84% (95% CI=69%–92%), 89% (95% CI=83%–93%),

7.60 (95% CI=4.64-12.43), and 0.18 (95% CI=0.09–0.38),

respectively. Significant heterogeneity was found for

sensitivity (I²=59.27%; Cochran Q=7.36; p=0.06) but low

heterogeneity was achieved for specificity (I²=20.93%;

Cochran Q=3.79; p=0.28) in DW-MRI.

The Figure 4 shows the summary point with a 95%

prediction region and 95% confidence region. It can be

observed that both techniques have wider prediction

contour than the confidence contour respectively. The

region of confidence is plotted from the CI around the

summary point and shows that, the ‘real value’ would be

estimated to be inside the region 95% of the time.

Furthermore, the prediction region is wider compared to

the region of confidence because it goes more than the

improbability in the presented data. Besides, TVUS has

wider area under the curve for HSROC curve than the DW-

MRI. However, DW-MRI has higher DOR compared to

the TVUS (41.75 and 20.45 respectively) but the LR shows

moderately increase of the probability of the disease. The

specificity of detection of myometrial invasion was greater

in TVUS but lower in DW-MRI. Fagan nomograms

(Figure 5) show that a positive test for TVUS and DW-

MRI increases significantly the pretest probability deep

myometrial invasion, from 12% to 41% in case of TVUS

and from 12% to 51% in case of DW-MRI, while a

negative test significantly decreases the pretest probability,

from 12% to 3% in case of TVUS and from12% to 1% in

case of MRI.

Aside from the stage, the depth of myometrial invasion is

acknowledged as one of the most crucial prognostic

aspects in endometrium carcinoma and for management

and surgical staging[17]. Preoperative information might

impact on the treatment plan[18].

In the present meta-analysis, the pooled diagnostic

accuracy of TVUS and DW-MRI has been compared and

evaluated in detecting deep myometrial invasion among

adult women with endometrial carcinoma undergoing

surgical staging. This analysis shows that DW-MRI has

higher diagnostic performance compared to TVUS with the

sensitivity and specificity of 84% and 89% respectively.

However, there is no substantial significant difference was

observed between them which TVUS has slightly lower

sensitivity and specificity (79% and 84% respectively). In

addition, it was found that the values of LR+ (5.03 and

7.60 using TVUS and DW-MRI, respectively) and LR-

(0.25 and 0.18 using TVUS and DW-MRI, respectively)

were essentially between within the range of moderates

estimates for both confirming and ruling out the disease.

According to Sundar et al., (2017)[19], in some

guidelines, MRI is currently suggested for preoperative

imaging. Thus, these outcomes might be of clinical

relevance. The heterogeneity of the studies was studied

through this meta-analysis and high heterogeneity was

observed for specificity of TVUS. In contrast, low

heterogeneity was found for the specificity in DW-MRI. In

term of sensitivity, the heterogeneity was significant for

DW-MRI but not for TVUS. The possible explanation of

this heterogeneity is the different technical aspects and

approaches used in each of the study.

DW-MRI and TVUS play important role in detecting the

depth of myometrial invasion. This review provides an

idea of the methodological quality of studies using DW-

MRI and TVUS in detecting the deep myometrial invasion.

However this meta-analysis only included small number of

studies. Therefore, the efficacy of these imaging in

evaluating myometrial invasion should be further

investigated in the future in larger scale of studies.

Some limitations had been recognized from this meta-

analysis. Firstly, there are limited numbers of studies and

sample size reported in this systematic review. This make

the possibility to observe all potential sources of

heterogeneity is not allowed. In addition, the technical

aspects of the DW-MRI are different in every study as well

as the TVUS which used different approaches in each

study. Moreover, from the meta-analytic point of view, it is

appropriate to review the studies that directly compared the

diagnostic accuracy of both imaging on the same set of

women. This method will lessen the risk of bias as the

heterogeneity among studies that cannot be adequately

controlled. This has been done in only one study [14].

From the current study, the DW-MRI technique was

superior to the TVUS technique.

The low- and high-risk patients for deep myometrial

invasion were involved in the most of the studies. This

made the applicability of both techniques affected as

preoperative evaluation is suitable in patients with

preoperative histological records representing potential low

risk. Hence, the overestimate of diagnostic accuracy for

detecting the depth of myometrial invasion cannot be ruled

out, because of the involvement of high-risk patients. This

could also enlighten the heterogeneity observed among

studies for pooled specificity.



(A)

(B)

Figure 3: Forest plot for sensitivity and specificity for each

study and pooled sensitivity and specificity for (A) TVUS

and (B) DW-MRI

(A)

(B)

Figure 4: HSROC curve for (A) TVUS and (B) DW-MRI

Health Scope 24


(A)

(B)

Figure 5: Fagan nomograms showing how pre-test

probability change when the test is performed (post-test

probability) depending on a positive or negative result for

(A) TVUS and (B) DW-MRI

4. CONCLUSION

In conclusion, this systematic review and meta-

analysis revealed that DW-MRI is superior to the TVUS

imaging in measuring the deep myometrial invasion in

preoperative endometrial cancer and the sensitivity and

specificity of DW-MRI are slightly higher compared to the

TVUS imaging. However, the difference observed was not

statistically significant. Taking into account the time

consuming and high cost of MRI, TVUS may have a role

as the first imaging in measuring the depth of myometrial

invasion among adult women with endometrial cancer.

Whenever there is limited resources in a clinical settings,

TVUS should be considered worthy to evaluate the depth

of myometrial invasion.

ACKNOWLEDGEMENTS

The authors would like to thank lecturers of Medical

Imaging Department, Faculty of Health Sciences,

Universiti Teknology MARA for their supports.

REFERENCES

[1] W. T. Creasman, “Endometrial Carcinoma,”

eMedicine.com, 2018. [Online]. Available:

https://emedicine.medscape.com/article/254083-

overview. [Accessed: 06-Apr-2020].

[2] R. L. Siegel, K. D. Miller, and A. Jemal, “Cancer

statistics, 2020,” CA. Cancer J. Clin., vol. 70, no. 1,

pp. 7–30, 2020.

[3] M. Y. Lin, A. Dobrotwir, O. McNally, N. R. Abu-

Rustum, and K. Narayan, “Role of imaging in the

routine management of endometrial cancer,” Int. J.

Gynecol. Obstet., vol. 143, pp. 109–117, 2018.

[4] J. X. Jiang et al., “Endometrial carcinoma: diffusion-

weighted imaging diagnostic accuracy and

correlation with Ki-67 expression,” Clin. Radiol.,

vol. 73, no. 4, pp. 413.e1-413.e6, 2018.

[5] D. Moghazy Mohamed, K. Abd El-Wahab abo

Dewan, and S. Mahmoud Mera, “The Diagnostic

accuracy of transvaginal ultrasonography, magnetic

resonance imaging and diffusion weighted image in

female patients with endometrial carcinoma,”

Alexandria J. Med., vol. 54, no. 4, pp. 685–691,

2018.

[6] D. Moher, A. Liberati, J. Tetzlaff, and D. G. Altman,

“Preferred reporting items for systematic reviews

and meta-analyses: the PRISMA statement,” J. Clin.

Epidemiol., vol. 62, no. 10, pp. 1006–1012, 2009.

[7] S. K. Saarelainen and K. O. Lea, “The preoperative

assessment of deep myometrial invasion by three-

dimensional ultrasound versus MRI in endometrial,”

vol. 91, pp. 983–990, 2012.

[8] J. L. Alcázar, B. Gastón, B. Navarro, R. Salas, J.

Aranda, and S. Guerriero, “Transvaginal ultrasound

versus magnetic resonance imaging for preoperative

assessment of myometrial infiltration in patients with

endometrial cancer: A systematic review and meta-

analysis,” J. Gynecol. Oncol., vol. 28, no. 6, pp. 1–

11, 2017.

[9] K. A. Brocker, J. P. Radtke, P. Hallscheidt, C. Sohn,

H. Peter, and S. Céline, “Comparison of the

determination of the local tumor extent of primary

endometrial cancer using clinical examination and 3

Tesla magnetic resonance imaging compared to



histopathology,” Arch. Gynecol. Obstet., 2019.

[10] L. Savelli et al., “A prospective blinded comparison

of the accuracy of transvaginal sonography and

frozen section in the assessment of myometrial

invasion in endometrial cancer,” Gynecol. Oncol.,

vol. 124, no. 3, pp. 549–552, 2012.

[11] M. Hori et al., “Endometrial cancer: Preoperative

staging using three-dimensional T2-weighted turbo

spin-echo and diffusion-weighted MR imaging at 3.0

T: A prospective comparative study,” Eur. Radiol.,

vol. 23, no. 8, pp. 2296–2305, 2013.

[12] G. Ørtoft et al., “Preoperative staging of endometrial

cancer using TVS , MRI , and hysteroscopy,” vol.

92, pp. 536–545, 2013.

[13] J. W. Christensen, M. Dueholm, E. S. Hansen, E.

Marinovskij, E. Lundorf, and G. Ørtoft, “Assessment

of myometrial invasion in endometrial cancer using

three-dimensional ultrasound and magnetic

resonance imaging,” vol. 95, pp. 55–64, 2016.

[14] R. Angioli et al., “Preoperative local staging of

endometrial cancer : the challenge of imaging

techniques and serum biomarkers,” Arch. Gynecol.

Obstet., 2016.

[15] A. Rodrı and S. Martı, “Preoperative Assessment of

Myometrial Invasion in Endometrial Cancer by 3D

Ultrasound and Diffusion-Weighted Magnetic

Resonance Imaging A Comparative Study,” vol. 26,

no. 6, pp. 1105–1110, 2016.

[16] Y. Guo et al., “Myometrial invasion and overall

staging of endometrial carcinoma: Assessment using

fusion of T2-weighted magnetic resonance imaging

and diffusion-weighted magnetic resonance

imaging,” Onco. Targets. Ther., vol. 10, pp. 5937–

5943, 2017.

[17] C. M. Quick, T. May, N. S. Horowitz, and M. R.

Nucci, “Low-grade, low-stage endometrioid

endometrial adenocarcinoma: A clinicopathologic

analysis of 324 cases focusing on frequency and

pattern of myoinvasion,” Int. J. Gynecol. Pathol.,

vol. 31, no. 4, pp. 337–343, 2012.

[18] D. Dogan et al., “Preoperative evaluation of

myometrial invasion in endometrial carcinoma:

Diagnostic performance of 3T MRI,” Abdom.

Imaging, vol. 38, no. 2, pp. 388–396, 2013.

[19] S. Sundar et al., “BGCS uterine cancer guidelines:

Recommendations for practice,” Eur. J. Obstet.

Gynecol. Reprod. Biol., vol. 213, no. August 2014,

pp. 71–97, 2017.

Health Scope 26


ARTICLE TYPE

Diagnostic accuracy of transvaginal ultrasound (TVUS) and diffused weighted imaging (DWI) in female patient with ovarian cancer: a systematic review and meta-analysis

Ainul Shafiqah Zulkifli, Mohd Zulfadli Adenan

Centre of Medical Imaging, Faculty of Health Sciences, Universiti Teknologi MARA Cawangan Selangor

Kampus Puncak Alam, 42300 Bandar Puncak Alam, Selangor, Malaysia

Abstract:

This review is aimed to compare the diagnostic accuracy of transvaginal ultrasound (TVUS) and diffused

weighted image (DWI) in assessing ovarian cancer. An extensive search was performed in Dimension, Google

Scholar and Wiley Online Library for studies published from January 2011 to April 2020. Following the

screening, the eligibility of the studies was checked. Then, the Quality Assessment of Diagnostic Accuracy

Studies-2 (QUADAS-2) checklist was adapted to assess the quality of each of the study being reviewed. The

information on the diagnostic performance of TVUS and DWI were extracted from each of the studies. The

heterogeneity of the studies was explored. Overall, pooled sensitivity, specificity, LR+, and LR− of TVUS for

detecting ovarian cancer were 75% (95% confidence interval [CI]=57%–87%), 92% (95% CI=80%–97%), 8.97

(95% CI=3.21–25.08), and 0.28 (95% CI=0.14–0.52), respectively. High heterogeneity was established for

sensitivity (I²=92.28%; Cochran Q=38.85; p=0.00). Furthermore, high heterogeneity was found for specificity

(I²=88.54%; Cochran Q=26.17; p=0.00). TVUS has high specificity for detection of ovarian cancer. TVUS

should be considered as good enough for being used in clinical settings with limited resources.


Mohd Zulfadli Adenan


Keywords: Diffused weighted imaging, transvaginal ultrasound, ovarian cancer

1. INTRODUCTION

Cancer incidence and mortality are rapidly growing

worldwide. Ovarian cancer is one of the most common

gynecologic cancers that rank third after cervical and uterine

cancer. It also has the worst prognosis and the highest

mortality rate [1]. Although ovarian cancer has a lower

prevalence in comparison with breast cancer, it is three times

more lethal [2]. The high mortality rate of ovarian cancer is

caused by asymptomatic and secret growth of the tumor,

delayed onset of symptoms, and lack of proper screening that

result in its diagnosis in the advanced stages [1]. In 2015,

ovarian cancer was diagnosed in 21, 290 women in the United

States, and more than 14, 000 died from this disease. Effective

early detection of ovarian cancer through regularly repeated

screening tests may have a real impact on survival and,

potentially, on mortality from the disease [3].

Type I ovarian cancers are generally large, unilateral, cystic

tumors at diagnosis with indolent behavior. They are thought

to usually develop from extraovarian benign lesions that

embedded in the ovary and subsequently undergo a series of

mutations resulting in malignant transformation. In this way,

low‐ grade serous carcinomas are thought to originate from

benign deposits of fallopian tube epithelium in the ovaries

(endosalpingiosis); endometrioid and clear cell carcinomas

from benign foci of endometrial tissue in the ovaries

(endometriosis); and most mucinous carcinomas from benign

foci of transitional epithelium from the tuboperitoneal

junction [4]. Type I ovarian cancers are considered low grade,

except for clear cell carcinomas, and account for only a small

fraction of ovarian cancer deaths [5].

Type II epithelial cancers are high grade and characterized by

involvement of both ovaries, aggressive behavior, late stage

at diagnosis, and low survival. They are thought to originate

as fallopian tube fimbriae carcinomas that spread to the

ovaries and/or peritoneum. Women with these cancers often

present with extensive extraovarian disease and ascites. Type

II cancers are primarily high‐ grade serous carcinomas, the

most common epithelial subtype, but also include

carcinosarcomas and undifferentiated carcinomas. It is

notable that although tumor grade is important in clinical

practice, it is not a robust independent prognostic indicator

[5].

Studies on miRNA in formalin fixed placenta tissues are

limited [2] especially in Malaysia thus this study aimed at

detecting the presence of mir-210 in formalin fixed normal

placenta tissues in order to provide essential information for

27 Health Scope, 2020, Vol. 3(3) Ainul Shafiqah et al.


analysis of the variation in miRNA levels in archived

placentas.

Nonepithelial cancers are typically less aggressive than

epithelial malignancies. Germ cell and sex cord‐stromal

tumors make up the majority of nonepithelial cancers but

account for only 3% and 2%, respectively, of all ovarian

cancers. Sex cord‐stromal tumors arise from various

connective tissue cell types, including granulosa, Sertoli,

and/or Leydig cells. Other nonepithelial ovarian cancers

include small cell carcinoma (hypercalcemic and non-

hypercalcemic types) and ovarian sarcoma [4].

Transvaginal ultrasound (TVUS) is the most practical

modality for assessment because it is widely available, well

accepted by patients, non-invasive, low cost and not use

ionizing radiation [6]. The pooled sensibility and specificity

were respectively 77.0% and 83.0% for TVUS [6].

DWI is a newly developed magnetic resonance functional

imaging technique based on water molecules movement

rather than structure. Malignant tumors are composed of

randomly organized tumor cells and the free movement of

water molecules inside malignant dense mass is hindered. The

inhibited diffusion of water is attributed to hypercellularity,

thus DWI could provide unique information of tissue structure

by tissue cellularity evaluation [7]. [7] The pooled SEN and

SPE were 0.86 (95% confidence interval [CI], 0.83–0.89) and

0.81 (95%CI, 0.77–0.84), respectively.


2.1 Protocol and registration

Overview of systematic reviews of diagnostic test accuracy

will be conducted. The systematic review will be performed

according to Preferred Reporting Items for Systematic

reviews and Meta-Analyses: the PRISMA statement [8]

http://www.primastatement.org/statement.htm. The search

was conducted in Google Scholar, Dimension and Wiley

Online Library databases that published in January 2011 until

April 2020. All methods for inclusion/ exclusion criteria, data

extraction and quality assessment will be specified in

advance. The protocol will be not registered.

2.2 Data sources and search

Studies were identified by searching electronic databases and

the limit were applied for language which is only English.

This search was applied to Google scholar (2011- Present),

Dimensions (2011- Present) and Wiley Online Library (2011-

Present). The search strategy will include database subject

headings and text words as follows: “ovarian cancer”,

“ultrasound”, “transvaginal ultrasonography”, “TVUS” or

“TVS”, “diffused weighted image” and “magnetic resonance

imaging”. All the identified studies will be retrieved, and their

references will also be checked for other relevant publications.

2.3 Study selection and data collection

Two assessors screened all abstracts and checked relevant full

texts independently. Studies were enrolled in the meta-

analysis if they satisfied the eligibility criteria, including types

of studies, types of participants, types of interventions, types

of outcomes, language and year publication. The studies will

first be screened by their titles and abstracts to exclude

evidently irrelevant article. After that, the full text retrieved

for further clarification.

In order to avoid inclusion of duplicate cohorts in the meta-

analysis in the case of 2 studies from the same authors, the

study period of each study was examined; if the data

overlapped, we chose the latest study according to the

publication date, considering that patients from the first study

was also included in the latest one. We used “snowball”

strategy to identify potential interesting papers by reading

reference list of those papers selected for full text reading. No

attempts were made to contact the authors.

• Type of studies: Prospective or retrospective articles

that will be chosen should have evaluated the

diagnostic accuracy of TVUS or DWI alone or

combination with other techniques in diagnosing

ovarian cancer, regardless of the applied protocol of

acquisition.

• Type of participants: Participant will be adult women

undergoing histopathology of biopsy or surgery and

no limitations to nationality. Patients with any stage

of the disease will be included.

• Type of intervention: TVUS and DWI will be

regarded as index tests because these tests are usually

used to detect ovarian cancer.

• Types of outcome: The primary outcomes will be

sensitivity (SEN), specificity (SPE), positive

predictive value, negative predictive value, area

under the curve, and their respective 95% confidence

intervals.

• Language and year publication: Articles were

published in English in January 2011 until present.

The Patients, Intervention, Comparator, Outcomes, Study

Design (PICOS) criteria were used for describing the studies

include. The diagnostic performance results as well as the

supplementary useful information on procedures and patients

were retrieved from selected main studies independently by

the same reviewers. Disagreements were resolved peacefully

by discussion between reviewers. The study period of each

study was screened to avoid inclusion of duplicate cohorts in

the meta-analysis in the case of two studies from the same

authors. The latest study according to the publication date was

taken if dates overlapped. No attempts were made to contact

the authors for further information regarding the studies.

2.4 Risk of bias in individual studies

Quality Assessment of Diagnostic Accuracy Studies-2

(QUADAS-2) is a tool that used to assess the quality of the

studies and this tool had been adapted in this study. There are

four domains in the QUADAS-2 format: 1) patient selection,

http://www.primastatement.org/statement.htm

Health Scope 28


2) index test, 3) reference standard, 4) flow and timing. The

possibility of bias and concerns about applicability were

evaluated in each domain (the latter not applying to the 4th

domain) and rated as low, high, or unclear risk. The outcomes

of quality assessment were used for descriptive purposes to

explore potential sources of heterogeneity and to provide an

assessment of the overall quality of the included studies.

The quality of methodology was evaluated independently by

the two reviewers, using a standard form with quality

assessment criteria and a flow diagram; discussion had been

done to resolve the disagreements.


We extracted or derived information on diagnostic

performance of TVS and MRI. A random effects model was

used to determine overall pooled sensitivity, specificity,

positive likelihood ratio (LR+) and negative likelihood ratio

(LR−). Positive and negative likelihood ratios (LRs) were

used to characterize the clinical utility of a test and to estimate

the post-test probability of disease. With TP, TN, FP, FN from

extracted 2 × 2 contingency tables, we quantified the pooled

SEN, SPE, LR, and DOR with 95% confidence intervals

(95%CI) to evaluate DWI diagnosis accuracy for ovarian

cancer. P value based on the likelihood ratio test were

provided (α=0.05, two-sided). LR+ and LR- were applied to

characterize the clinical utility of test and to assess the post-

test probability of disease. A LR of 0.2–5.0 provides weak

evidence for either ruling out or confirming the disease. A LR

of 5.0–10.0 and 0.1–0.2 provides moderate evidence to either

confirm or rule out the disease. A LR >10 or <0.1 provides

strong evidence to either confirm or rules out the disease.

Heterogeneity for sensitivity and specificity were explored

graphically thru constructing forest plots of sensitivity and

specificity of each primary study. Then, they were plotted in

the hierarchical summary receiver operating characteristic

(HSROC) space, the latter to identify whether any

heterogeneity could be attributable to an implicit threshold

effect. Besides, HSROC curves for each technique were

plotted to illustrate relationship between sensitivity and

specificity. Means of a test on the Q statistic and I2 index were

calculated to assess the presence of heterogeneity. A p-value

<0.1 points to heterogeneity. The I2 index was measured to

define the percentage of overall variation across studies that

are due to heterogeneity rather than chance. The I2 value of

25%, 50% and 75% would be considered to specify low,

moderate, and high heterogeneity, respectively.

All analysis was performed using Meta-analytical Integration

of Diagnostic Accuracy Studies (MIDAS) and (METANDI)

commands in STATA version 13.0.


3.1 Search result

The search of Google Scholar, Dimension and Wiley Online

Library databases provided a total 355 citations. After

removal of duplicate records, 307 citations were remained.

From these, 296 were discarded because it was clear from the

title and abstract that they did not meet the criteria. The

remaining 11 papers were examined. Finally, 5

[9][10][11][12][13] studies were excluded because these

articles did not meet inclusion criteria and the remaining 6

[14][15][16][17][18][19] were included in the review and

meta-analysis. There were no additional related studies were

found from references cited in the articles included in the

review.

3.2 Characteristics of included studies

A total of 6 studies published between January 2012 to April

2020 reporting on 1627 patients were included in the final

analyses. Among these 1627 women, 974 had ovarian cancer.

Two of the studies were retrospective studies and the

remaining were prospective studies. Three studies used TVUS

to assess the ovarian cancer and another two studies used

DWI. Only one study assesses ovarian cancer using both

TVUS and DWI on the same patients. Figure 1 below is a

flowchart summarizing literature identification and selection.

Records identified through database searching

(Google Scholar, Dimension, Wiley Online Library)

(n= 355)

Records after duplicates removed

(n= 307)

Records screened

(n= 307)

Full-text articles

assessed for eligibility

(n= 11)

Studies included in

qualitative synthesis

(n= 6)

Studies included in quantitative

synthesis (meta-analysis)

(n = 6)

Records excluded

(n= 296)

Full-text articles

excluded

(n= 5):

- Studies not using

TVUS [9]

- Patient did not

have ovarian cancer

[10]

- There is not

enough data [11]

- Studies not using

DWI [12]

- Studies about

adnexal masses [13]

Figure 1 Flow chart showing studies selection

process



Table 1 Characteristic of included studies in this systematic

review according to PICOS criteria

Study Study

design

No. of

patient

Age (years) Obser-

vers

TVUS

Obser-

vers

DWI

Mean Min-

max

(Semelka et

al., 2014)

Pro- 100 44 16-65 NA Multip

le

(Zikan et al.,

2016)

Pro- 191 59 39-79 Multip

le

NA

(Fischerova

et al., 2017)

Pro- 393 59 40-76 Multip

le

NA

(Testa et al.,

2012)

Pro- 115 59 31-85 Multip

le

NA

(Mohammed

et al., 2020)

Retro- 44 46 21-78 NA Single

(Li et al.,

2012)

Retro- 131 NA NA NA Single

Pro-, Prospective; Retro-, Retrospective

Table 2 The technical aspects of MR protocols Study Tesla Mark b

value

Thickness

(mm)

FOV

(mm)

Matrix T2-

fused

(Semelka et

al., 2014)

1.5 1000 - - - Yes

(Li et al.,

2012)

1.5 1000 6 320-

420

256 x 256 Yes

(Mohammed

et al., 2020)

1.5 1000 - - - Yes

Table 1 shows PICOS features of studies included. All studies

were based on the radiologist’s or physician’s impression. The

technical aspects of MR protocols are described in Table 2.

All studies used 1.5T and were used surface coil. For TVUS,

all studies followed the standard protocol.

3.3 Methodological quality of included studies

Regarding risk of bias and the domain patient selection, one

study was considered high risk and the other five studies were

considered low. Concerning the domain index test regarding

DWI, three studies adequately describe the method of index

test as well as how it was performed and interpreted. The other

three studies not applicable because it is for TVUS. Regarding

TVUS, four studies adequately described the method of index

test as well as how it was performed and interpreted. For the

domain reference standard, all studies were likely interpreted

the reference standard results with knowledge of the results of

the index test. Regarding the domain flow and timing, the time

elapsed between the index test and reference standard were

unclear in three studies.

Regarding applicability, for the domain patient selection, five

studies were deemed to include patients that matched the

review question. For the domain index test, regarding DWI,

all studies were considered as having low concerns for

applicability as the index test was described well enough for

study replication as was the reference standard domain.

However, for TVUS, the domain index test was high.

0% 50% 100%

Patient selection

Index test DWI

Index test TVUS

Reference standard

Flow and timing

Proportion of studies with low,

high or unclear risk of bias

Low High Unclear

0% 50% 100%

Patient selection

Index test DWI

Index test TVUS

Reference standard

Proportion of studies with low,

high or unclear concerns

regarding applicability

Low High Unclear

Figure 2 Histogram plot showing quality

assessment (risk of bias and concerns about

applicability) for all studies included in the meta-

analysis

Health Scope 30


3.4 Diagnostic performance of TVUS and DWI in

detecting ovarian cancer

Individual results of TVUS are shown on the forest plots in

Figure 3. Overall, pooled sensitivity, specificity, LR+, and

LR− of TVUS for detecting ovarian cancer were 75% (95%

confidence interval [CI]=57%–87%), 92% (95% CI=80%–

97%), 8.97 (95% CI=3.21–25.08), and 0.28 (95% CI=0.14–

0.52), respectively. High heterogeneity was established for

sensitivity (I²=92.28%; Cochran Q=38.85; p=0.00).

Furthermore, high heterogeneity was found for specificity

(I²=88.54%; Cochran Q=26.17; p=0.00).

Figure 3 Forest plot for sensitivity and specificity for TVUS

On the other hand, pooled sensitivity, specificity, LR+, and

LR− of DW-MRI for detecting ovarian cancer not applicable.

This is because for STATA version 13.0. to be working, all

the value for true-positive, false-positive, false-negative and

true-negative must not have value 0.

HSROC curves are shown in Figure 4. The figure shows the

summary point with a 95% prediction region and 95%

confidence region. It can be observed that TVUS techniques

has wider prediction contour than the confidence contour,

respectively. The region of confidence is plotted from the CI

around the summary point and shows that, the ‘real value’

would be estimated to be inside the region 95% of the time.

The region of prediction around the summary point specifies

the region where the results from a new research in the future

are expected to lie. Furthermore, the prediction region is wider

compared to the region of confidence because it goes more

than the improbability in the presented data.

Figure 4 HSROC curve for TVUS

Fagan nomogram (Figure 5). show that a positive test for

TVUS significantly increases the pretest probability ovarian

cancer, from 12% to 55% in case of TVUS, while a negative

test significantly decreases the pretest probability, from 12%

to 4%. There is no apparent sign of publication bias in the

funnel plots presented in Figure 6 and the Deeks adaptation

for funnel plot asymmetry was significant for both TVUS

(p=0.20).

Figure 5 Fagan nomogram showing how pre-test probability

change when the test is performed (post-test probability)

depending on a positive or negative result for TVUS



Figure 6 Funnel plot according to Deeks for graphical

exploration of publication bias

Table 3 Pooled results of meta-analysis for TVUS and DWI

Analysis

Sensitivity

(95% CI)

Specificity

(95% CI)

LR+

(95% CI)

LR-

(95% CI)

DOR

(95% CI)

DWI NA NA NA NA NA

TVUS 0.75 (0.57,

0.87)

0.92 (0.80,

0.97)

8.97

(3.21,

25.08)

0.28

(0.14,

0.52)

32.59

(6.90,

153.9)

4. CONCLUSION

Overall, the evidence is not sufficiently robust to determine

the diagnostic accuracy between TVUS and DWI to detect the

ovarian cancer. This is because for STATA version 13.0. to

be working, all the value for true-positive, false-positive,

false-negative and true-negative must not have value 0. Due

to that, the diagnostic accuracy for DWI cannot be determine.

Sensitivity, specificity, LR+, and LR− of TVUS for detecting

ovarian cancer were 75% (95% confidence interval

[CI]=57%–87%), 92% (95% CI=80%–97%), 8.97 (95%

CI=3.21–25.08), and 0.28 (95% CI=0.14–0.52), respectively.

High heterogeneity was established for sensitivity

(I²=92.28%; Cochran Q=38.85; p=0.00). Furthermore, high

heterogeneity was found for specificity (I²=88.54%; Cochran

Q=26.17; p=0.00).

Notwithstanding, some limitations of the meta-analysis also

should be acknowledged. First, only a small number of studies

were included in the final meta-analysis because many studies

were excluded based on eligibility criteria and may not be

qualified to evaluate the diagnostic accuracy. All included

studies were published in English which may have negated

some of the gray literature.

In conclusion, TVUS has high specificity for detection

ovarian cancer. TVUS should be considered as good enough

for being used in clinical settings with limited resources.

REFERENCES

[1] F. Bray, J. Ferlay, I. Soerjomataram, R. L. Siegel, L.

A. Torre, and A. Jemal, “Global cancer statistics 2018:

GLOBOCAN estimates of incidence and mortality worldwide

for 36 cancers in 185 countries,” CA. Cancer J. Clin., vol. 68,

no. 6, pp. 394–424, 2018.

[2] S. Coburn, F. Bray, M. Sherman, and B. Trabert,

“International patterns and trends in ovarian cancer incidence,

overall and by histologic subtype,” Int J Cancer, vol. 140, no.

11, pp. 2451–2460, 2017.

[3] A. Coovadia and J. A. Eggert, “Future Perspectives

in Cancer Screening and Early Detection,” Semin. Oncol.

Nurs., vol. 33, no. 2, pp. 219–222, 2017.

[4] L. A. Torre et al., “Ovarian cancer statistics, 2018,”

CA. Cancer J. Clin., vol. 68, no. 4, pp. 284–296, 2018.

[5] R. J. Kurman and I. M. Shih, “The dualistic model of

ovarian carcinogenesis revisited, revised, and expanded,” Am.

J. Pathol., vol. 186, no. 4, pp. 733–747, 2016.

[6] G. M. Borrelli, L. A. de Mattos, M. de P. Andres, M.

O. Gonçalves, R. M. Kho, and M. S. Abrão, “Role of Imaging

Tools for the Diagnosis of Borderline Ovarian Tumors: A

Systematic Review and Meta-Analysis,” J. Minim. Invasive

Gynecol., vol. 24, no. 3, pp. 353–363, 2017.

[7] X. Yuan, L. Guo, W. Du, F. Mo, and M. Liu,

“Diagnostic accuracy of DWI in patients with ovarian

cancer,” Med. (United States), vol. 96, no. 19, pp. 1–7, 2017.

[8] D. Moher, A. Liberati, J. Tetzlaff, D. G. Altman, and

T. P. Group, “Preferred Reporting Items for Systematic

Reviews and Meta-Analyses : The PRISMA Statement,” vol.

6, no. 7, 2009.

[9] Z. Liang, J. S. B, D. Zhang, and L. Gao, Small Object

Detection Using Deep, vol. 1, no. September 2018. Springer

International Publishing, 2018.

[10] J. Kaijser et al., “Imaging techniques for the pre-

surgical diagnosis of adnexal tumours,” Best Pract. Res. Clin.

Obstet. Gynaecol., vol. 28, no. 5, pp. 683–695, 2014.

[11] P. Yadav, “Magnetic resonance imaging in the

evaluation of female pelvis,” Med. J. Dr. D.Y. Patil Univ., vol.

9, no. 5, p. 627, 2016.

[12] S. M. Mansour, S. Saraya, and Y. El-Faissal, “Semi-

quantitative contrast-enhanced MR analysis of indeterminate

ovarian tumours: When to say malignancy?,” Br. J. Radiol.,

vol. 88, no. 1053, 2015.

[13] K. Shimada et al., “Ultrasound-based logistic

regression model LR2 versus magnetic resonance imaging for

discriminating between benign and malignant adnexal

masses: a prospective study,” Int. J. Clin. Oncol., vol. 23, no.

3, pp. 514–521, 2018.

Health Scope 32


[14] D. Fischerova et al., “Ultrasound in preoperative

assessment of pelvic and abdominal spread in patients with

ovarian cancer: a prospective study,” Ultrasound Obstet.

Gynecol., vol. 49, no. 2, pp. 263–274, 2017.

[15] W. Li, C. Chu, Y. Cui, P. Zhang, and M. Zhu,

“Diffusion-weighted MRI : a useful technique to discriminate

benign versus malignant ovarian surface epithelial tumors

with solid and cystic components,” no. October 2011, pp.

897–903, 2012.

[16] G. Mohammed, A. El, and M. A. M. Ahmed, “The

Usefulness of Diffusion Weighted and Contrast Enhanced

Magnetic Resonance Imaging in Characterization of

Inconclusive Ovarian Mass,” pp. 24–33, 2020.

[17] R. Semelka, K. Shawky, S. AbdelMoniem, and H.

Tantawy, “Role of Mri and Diffusion Mri in Evaluation of

Pelvic Masses of Gynaecological Origin,” Zagazig Univ.

Med. J., vol. 20, no. 5, pp. 1–14, 2014.

[18] A. C. Testa et al., “Ultrasound evaluation of intra-

abdominal sites of disease to predict likelihood of suboptimal

cytoreduction in advanced ovarian cancer: A prospective

study,” Ultrasound Obstet. Gynecol., vol. 39, no. 1, pp. 99–

105, 2012.

[19] M. Zikan et al., “A prospective evaluation of

ultrasound accuracy in the prediction of rectosigmoid

infiltration in patients with epithelial ovarian cancer,” p. 368,

2016.

Health Scope 33


RESEARCH ARTICLE

Association of 18F-FDG PET/CT imaging metabolic parameters with

tumour size in breast carcinoma

Ain Nabila Zahari, Mohd Hafizi Mahmud*



Abstract:

Tumour size is a well-known independent prognostic factor that plays an important role in the

management and diagnosis of cancer. 18F-Fluorodeoxyglucose Positron Emission

Tomography/Computed Tomography (18F-FDG PET/CT) has been widely used in oncology which

their metabolic parameters including maximum standardized uptake value (SUVmax) and metabolic

tumour volume (MTV) are important PET indexes to signify the degree of tumour aggressiveness.

The correlation of SUVmax with tumour size has been widely reported in the literatures, however

such correlation with MTV was not well reported. This study aims to evaluate the association

between tumour size and PET/CT metabolic parameters in breast carcinoma. PET/CT images of

breast cancer patients (n = 10) who underwent 18F-FDG PET/CT examination for staging were

reviewed. The patients were histopathologically confirmed with breast carcinoma. PET/CT derived

metabolic parameters (SUVmax and MTV) of the primary breast carcinoma lesions (n=15) were

quantified. The parameters were associated with the tumour size of the primary lesions using

Spearman Rho Correlation. All the metabolic parameters have positive statistically significant

association with tumour size where SUVmax (p < 0.001, r = 0.817) and MTV (p < 0.001, r = 0.954). 18F-FDG PET/CT-derived metabolic parameters (SUVmax and MTV) are valuable imaging markers

associating with tumour size of the breast lesions in which better association is demonstrated in

volume based metabolic parameter (MTV) as compared to SUVmax. Therefore, both parameters have

prognostic value for the evaluation of breast carcinoma.


Mohd Hafizi Mahmud


Keywords: 18F-FDG PET/CT, breast carcinoma, metabolic tumour volume, standardized uptake

value

1. INTRODUCTION

World Health Organization (WHO) has revealed that

breast cancer is the most prevalent cancer among females in

the world, and it is the second major cause of death from

cancer among women [1]. Breast cancer has been reported as

the major cancer among females in Malaysia [2]. Its high

incidence and mortality make breast cancer is a major health

problem [3] as it affects more than one million women all over

the world [4]. Thus, accurate staging of breast cancer is vital

for clinical management decisions to prevent delayed urgent

treatment [5]. As a result, non-invasive diagnostic tools for

staging and tumour behaviour prediction are becoming

increasingly crucial for breast cancer management [6].

An advanced hybrid molecular imaging, 18F-

fluorodeoxyglucosepositron-emission tomography/computed

tomography (18F-FDG PET/CT) has been widely used in

clinical practice to characterize and stage tumours non-

invasively. It can identify breast cancer in its early stage,

reflect the glycolytic changes of tumours, and simultaneously

evaluate the whole body’s responses to the tumours in vivo

[1]. Besides, it has improved the effectiveness of imaging in

staging patients with breast cancer [7]. PET/CT can provide

quantitative biomarkers or known as metabolic parameters

such as standardized uptake value (SUV) and metabolic

tumour volume (MTV), which reflect tumour receptor status,

the degree of tumour heterogeneity, and treatment response

[8]. SUV has been widely used as PET parameter for

estimating the metabolic activity of tumours. However, SUV

has been reported to cause overestimation in obese patients as

this parameter is calculated based on the whole-body weight

metric including fat contribution of the patients [1]. Thus,

accurate staging of the cancer will be affected. To compensate

this limitation, the role of volumetric parameter derived from

FDG PET/CT such as MTV has been explored as this

parameter measures the metabolic burden of the whole tumour

[9]. This parameter has been shown to be independent

prognostic factors in several oncological studies [10].

34 Health Scope, 2020, Vol. 3(3) Ain Nabila et al.

Generally, the diagnosis of breast cancer is based on clinical

examination in combination with imaging and is confirmed by

pathological assessment [11]. The clinicopathological

features such as tumour size are one of the important factors

in making the clinical and pathological assessment of breast

cancer. Tumour size is one of the staging criteria for various

types of cancer and has a well-known prognostic role [12, 13].

Tumour size may influence patients’ staging (T) status, thus

having an impact on subsequent surgical and oncological

management, including the type of treatment. The accuracy of

preoperative tumour measurement is great importance in

deciding patients’ eligibility for conserving their breasts [14],

as more accurate tumour staging can be achieved, and a

correct treatment protocol can be followed.

Combining clinicopathological data with metabolic

parameters may further improve diagnostic efficiency and

evaluation prediction of the prognosis of the patient with

cancer [15]. Significant correlation of SUVmax and tumour

size have been reported in the previous literatures [6, 10, 16,

17]. However, such impact on MTV parameter is not well

reported [18, 19]. Therefore, this study aims to evaluate the

association between tumour size and PET/CT metabolic

parameters in breast carcinoma.


2.1 Study Population

This retrospective study was approved by the Research

Ethics Committee of Faculty of Health Sciences Universiti

Teknologi MARA (UiTM REC/03/2 UG/MR/93) and

informed consent was waived due to the retrospective design.

Breast cancer patients who underwent whole-body 18F-FDG

PET/CT examination between April 2015 to March 2017 has

been reviewed. Only PET/CT examinations for staging were

included in the study. All the cases have been confirmed with

histopathological examination. The PET/CT images were

selected using purposive sampling method based on the

inclusion and exclusion criteria.

2.2 18F-FDG PET/CT Image Acquisition

PET/CT images were acquired using an integrated

PET/CT system (GE HEALTHCARE DISCOVER). None of

the patients had a blood glucose level >130 mg/dL before 18F-

FDG injection. The scanning was acquired 60 min following

intravenous 18F-FDG administration. All patients were placed

in the supine position and the scans were acquired with the

patients immobilized. Six-to-eight-bed positions were used,

and the acquisition time was 2–2.5 min per position. Non

contrast-enhanced CT imaging was started at the vertex and

extended to the upper thigh, subsequently PET scanning was

performed over the same body region. CT data were used for

attenuation correction, and images were reconstructed using a

standard ordered-subset expectation-maximization algorithm.

2.3 Imaging and Data Analysis

The PET/CT images were reviewed using a workstation

(Syngo. Via, Siemens Medical Solutions, Chicago, IL) that

provided multiplanar reformatted images in transverse,

coronal, and sagittal planes. The images of the same patient

were analysed frame by frame. Tumour mapping was

performed by manual contouring of the lesions guided by a

nuclear medicine physician. Areas of abnormally intense

tracer uptake were recorded. Standard spherical regions of

interest (ROI) were placed over the increased pathological

uptakes of the primary lesions on PET/CT images to obtain

the SUVavg and SUVmax. These parameters were determined

automatically by the Syngo software following the delineation

of the ROI on the selected lesions in the PET/CT images. A

volumetric ROI around the outline of primary tumour in the

breast was placed on the axial PET/CT images and the borders

of the ROI were adjusted manually by visual inspection of the

primary lesions outline to avoid overlapping on adjacent

FDG-avid structures or lesions (Figure 1). Tumour size was

expressed by the maximum diameter of the MTV measured

by contouring margins defined with threshold of SUV of 2.5.

Figure 1. Tumour mapping of breast lesion in 18F-FDG PET/CT

image. Note the breast lesion margin was delineated by the

red line.

2.4 Statistical Analysis

SPSS 21.0 software for Windows (IBM Corp., Armonk,

NY, USA) was used for the statistical analysis. Continuous

variables were expressed as median (interquartile range)

(IQR). Two-tailed Spearman’s rho correlation test was

performed to evaluate association between tumour size and

metabolic parameters with p-value < 0.05 was considered as

statistically significant.


A total of 15 lesions from 10 patients with the median

(IQR) age of 51 (10) years old were quantified in this study.

Tumour and metabolic parameters characteristics are

presented in Table 1.

Health Scope 35

Table 1. Patients, tumour characteristics and metabolic

parameters

Association of metabolic parameters and tumour size of the

breast lesion is demonstrated in Table 2. Both parameters

show strong association with tumour size (SUVmax, r = 0.817;

MTV, r = 0.954) and are significantly associated (p < 0.001).

These associations are expressed in the scatter plot graphs, as

shown in Figure 2.

Table 2. Spearman rho correlation analysis between FDG

PET/CT metabolic parameters and tumour size of breast

carcinoma

(a)

(b)

Figure 2. Association between tumour size and metabolic

parameters (a) SUVmax and (b) MTV, respectively

(p < 0.001).

In this study, significant associations have been observed

between tumour size of breast carcinoma and metabolic

parameters (SUVmax and MTV). The findings of the present

study are in accordance with the previous literatures [6, 10,

13, 24]. Kitajima et al. [6] and Chang et al. [10] reported a

significant association between SUVmax and MTV with

tumour size and grade. Furthermore, several studies have

reported association in increased metabolic activities with an

increase of tumour size [6, 20]. Higher glucose uptake usually

reflects larger tumour sizes or worse malignancy, so that the

higher FDG signals around the advanced stage of tumour [21,

22]. Ayaz et al. [16] and Jain et al. [23] claimed SUV was

dependent on tumour size with increased uptake seen in larger

tumour size, tumour grade and stage of breast lesions. This is

probably because of the increased aggressivity of breast

cancer and unfavourable changes in tumour behaviour as the

time proceeds and tumour grows further [16].

In the present study, higher correlation magnitude (r) is

observed in MTV and TLG as compared to SUVmax in relation

to tumour size. This finding is supported by several previous

literatures. Önner et al. [13] and Vatankulu et al. [24] reported

the tumour size showed stronger positive association with

MTV than SUVmax. MTV can reflect the metabolic volume,

which is the FDG-avid volume in the tumour, rather than the

size of the mass. MTV associates with the size of the mass and

provide more accurate measurement than the maximum or

minimum diameters, especially for lesions with non-FDG

uptake necrosis inside [1]. Thus, MTV may more accurately

reflect the tumour activity and grade of malignancy as

compared to SUVmax [1, 25]. SUVmax does not represent the

whole tumour metabolic burden because the value is

generated from only one voxel [26]. SUVmax is limited to

provide area information from the high metabolic uptake of a

particular lesions area, which can make it difficult to identify

the expansion of active lesions within malignant tumours

accurately based on this parameter alone and difficult to assess

how much measured volume reflects the viable tumour region

[27]. MTV has been proven to be more accurate because they

highly correlate with the tumour size of the breast lesion.

Therefore, MTV is a superior metabolic parameter associating

with the tumour size in relative to SUVmax. Hence this

parameter is useful PET index to signify the aggressiveness of

the lesions.

4. CONCLUSION

18F-FDG PET/CT-derived metabolic parameters (SUVmax,

and MTV) are valuable imaging markers associating with the

tumour size of the breast lesions. MTV as a volume based

metabolic parameter demonstrates better association with

tumour size in breast lesions as compared to SUVmax.

Therefore, both parameters have prognostic value for the

evaluation of breast carcinoma.

ACKNOWLEDGEMENTS

The authors would like to thank the Research Ethics

Committee of Faculty of Health Sciences Universiti

Teknologi MARA for the ethics approval of the study.

Variable Median (IQR)

Age 51.00 (10.00)

Tumour size, cm2 2.58 (5.17)

Metabolic parameters

SUVmax

MTV

8.90 (5.87)

5.96 (22.62)

Variable Tumour Size

r p value

SUVmax 0.817 < 0.001

MTV 0.954 < 0.001

36 Health Scope, 2020, Vol. 3(3) Ain Nabila et al.

REFERENCES

[1] W. Chen, L. Zhu, X. Yu, Q. Fu, W. Xu, and P. Wang,

“Quantitative assessment of metabolic tumor burden

in molecular subtypes of primary breast cancer with

FDG PET/CT,” Diagnostic Interv. Radiol., vol. 24,

no. 6, pp. 336–341, 2018.

[2] I. P. Gopal Ram, “Vital Role of Water Treatment in

Energy Management.,” Energy Manag. (New Delhi),

vol. 7, no. 1, pp. 45–51.

[3] A. Jiménez-Ballvé et al., “Prognostic value of

metabolic tumor volume and total lesion glycolysis

in 18F-FDG PET/CT scans in locally advanced

breast cancer staging,” Rev. Española Med. Nucl. e

Imagen Mol. (English Ed., vol. 35, no. 6, pp. 365–

372, 2016.

[4] M. Taghipour, R. Wray, S. Sheikhbahaei, J. L.

Wright, and R. M. Subramaniam, “FDG avidity and

tumor burden: Survival outcomes for patients with

recurrent breast cancer,” Am. J. Roentgenol., vol.

206, no. 4, pp. 846–855, 2016.

[5] D. Groheux, A. Cochet, O. Humbert, J. L. Alberini,

E. Hindié, and D. Mankoff, “18F-FDG PET/CT for

staging and restaging of breast cancer,” J. Nucl.

Med., vol. 57, p. 17S–26S, 2016.

[6] K. Kitajima, Y. Miyoshi, T. Yamano, S. Odawara, T.

Higuchi, and K. Yamakado, “Prognostic value of

FDG-PET and DWI in breast cancer,” Ann. Nucl.

Med., vol. 32, no. 1, pp. 44–53, 2018.

[7] D. M. Schuster, “Clinical Utility of PET Scanning in

Breast Cancer Management,” Am. J. Hematol.

Oncol., vol. 11, no. June, pp. 20–25, 2015.

[8] C. Bailly et al., “Exploring tumor heterogeneity

using PET imaging: The big picture,” Cancers

(Basel)., vol. 11, no. 9, pp. 1–17, 2019.

[9] S. Y. Park et al., “Prognostic value of total lesion

glycolysis by 18F-FDG PET/CT in surgically

resected stage IA non-small cell lung cancer,” J.

Nucl. Med., vol. 56, no. 1, pp. 45–49, 2015.

[10] C. C. Chang, C. J. Chen, W. L. Hsu, S. M. Chang, Y.

F. Huang, and Y. C. Tyan, “Prognostic Significance

of Metabolic Parameters and Textural Features on

18F-FDG PET/CT in Invasive Ductal Carcinoma of

Breast,” Sci. Rep., vol. 9, no. 1, pp. 1–11, 2019.

[11] J. Huszno and Z. Kolosza, “Molecular characteristics

of breast cancer according to clinicopathological

factors,” Mol. Clin. Oncol., vol. 11, no. 2, pp. 192–

200, 2019.

[12] A. A. Kasangian et al., “The prognostic role of tumor

size in early breast cancer in the era of molecular

biology,” PLoS One, vol. 12, no. 12, pp. 1–12, 2017.

[13] H. Önner et al., “Which of the fluorine-18

fluorodeoxyglucose positron emission

tomography/computerized tomography parameters

are better associated with prognostic factors in breast

cancer?,” Medicine (Baltimore)., vol. 98, no. 22, p.

e15925, 2019.

[14] Y. Z. Jiang, C. Xia, W. T. Peng, K. Da Yu, Z. G.

Zhuang, and Z. M. Shao, “Preoperative measurement

of breast cancer overestimates tumor size compared

to pathological measurement,” PLoS One, vol. 9, no.

1, pp. 1–5, 2014.

[15] B. Yang et al., “Correlations Study Between 18F-

FDG PET/CT Metabolic Parameters Predicting

Epidermal Growth Factor Receptor Mutation Status

and Prognosis in Lung Adenocarcinoma,” Front.

Oncol., vol. 9, no. July, pp. 1–9, 2019.

[16] S. Ayaz, S. S. Gültekin, Ü. Y. Ayaz, and A. Dilli,

“Initial fludeoxyglucose (18F) positron emission

tomography-computed tomography (FDG-PET/CT)

imaging of breast cancer - Correlations with the

primary tumour and locoregional metastases,” Polish

J. Radiol., vol. 82, pp. 9–16, 2017.

[17] O. C. Guler, N. Torun, B. A. YılDıRım, and C. Onal,

“Pretreatment metabolic tumour volume and total

lesion glycolysis are not independent prognosticators

for locally advanced cervical cancer patients treated

with chemoradiotherapy,” Br. J. Radiol., vol. 91, no.

1084, 2018.

[18] A. Myssayev et al., “Usefulness of FDG PET/CT

derived parameters in prediction of histopathological

finding during the surgery in patients with pancreatic

adenocarcinoma,” PLoS One, vol. 14, no. 1, pp. 1–

15, 2019.

[19] K. Pak et al., “Original article Prognostic value of

metabolic tumor volume and total lesion glycolysis

in breast cancer : a meta-analysis,” no. June, 2020.

[20] Z. Huang, S. Chen, and J. Li, “Correlation between

18F-FDG PET/CT Metabolic Parameters and

Lymph Node Metastasis of Esopageal Cancer,” Int.

J. Clin. Exp. Med., vol. 12, no. 3, pp. 2690–2696,

2019.

[21] E. A. Wellberg et al., “The glucose transporter

GLUT1 is required for ErbB2-induced mammary

tumorigenesis,” Breast Cancer Res., vol. 18, no. 1,

pp. 1–15, 2016.

[22] Y. Zhang and J. Wang, “Targeting uptake

transporters for cancer imaging and treatment,” Acta

Pharm. Sin. B, vol. 10, no. 1, pp. 79–90, 2020.

[23] A. Jain et al., "Tumor characteristics and

metabolic quantification in carcinoma breast: An

institutional experience", Indian Journal of Cancer,

vol. 54, no. 1, p. 333, 2017.

[24] B. Vatankulu et al., “Do 18F-FDG PET/CT findings

have a relationship with histopathological and

immunohistochemical factors of breast cancer in

men?,” Nucl. Med. Commun., vol. 37, no. 12, pp.

1273–1281, 2016.

[25] S. H. Son et al., “Prognostic implication of

intratumoral metabolic heterogeneity in invasive

ductal carcinoma of the breast,” BMC Cancer, vol.

14, no. 1, p. 585, 2014.

[26] H. J. Im, T. Bradshaw, M. Solaiyappan, and S. Y.

Cho, “Current Methods to Define Metabolic Tumor

Volume in Positron Emission Tomography: Which

One is Better?,” Nucl. Med. Mol. Imaging (2010).,

vol. 52, no. 1, pp. 5–15, 2018.

[27] S. Ogawa, M. Itabashi, C. Kondo, M. Momose, S.

Sakai, and S. Kameoka, “Prognostic value of total

lesion glycolysis measured by 18F-FDG-PET/CT in

Health Scope 37

patients with colorectal cancer,” Anticancer Res.,

vol. 35, no. 6, pp. 3495–3500, 2015.

Health Scope 38


RESEARCH ARTICLE

Environmentally-extended input-output analysis for Malaysia

Farah Ayuni Shafie1*, Nursyazwani Suparman1, Umi Aida Adlina Mesarandi1, Dasimah Omar2,

Subramaniam Karuppannan1, Oliver Ling Hoon Leh2

1Centre of Environmental Health and Safety, Faculty of Health Sciences, Universiti Teknologi MARA Cawangan Selangor,

Kampus Puncak Alam, 42300 Bandar Puncak Alam, Selangor, Malaysia,2Faculty of Architecture, Surveying and Planning,

40000 Shah Alam, Selangor, Malaysia

Abstract:

Sustainability has become the key goal in every aspect of the environment especially in the diverse

urban systems. Sustainable practices require a controlled setting within an urban system where their

practicality and efficacy could be assessed. The aim of the study is to establish an environmental

assessment tool based on urban metabolism approach to assist decision-making during

environmental and economic assessment in an urban planning process. The outcome provides

understanding on the means of integrating carbon footprint and monetary factor to oversee the

expenditure of a nation in general, or a household or an individual in particular, in relation to global

warming potential. This study applied a retrospective cross sectional study to evaluate sustainability

in economic-environmental input-output model for greenhouse gases emission monitoring and

climate change adaptation. The potential impact from economic sectors’ contribution on greenhouse

gas emissions which eventually leads to global warming and climate change was quantitatively

assessed by examining the contribution of greenhouse gas emission from each economic activity.

Malaysia greenhouse gas emission of all 120 economic activities is 4.87 kg CO2-eq./cap/day. The

use of publicly available data to assemble the matrix representation enables comprehensive

assessment of the environmental impacts of a country in a manner which is not only of high

technology but also relatively fast and cost effective.


Farah Ayuni Shafie


Keywords: carbon dioxide emission, economic input-output table

1. INTRODUCTION

Development is often carried out without fully considering the future consequences and creates numerous environmental impacts including increased air and water pollution [1-2], climate alteration [3-4], emergence and re-emergence of disease [5-7], insufficient housing and sanitation facilities [8-9], as well as overcrowding [9-10] and traffic congestion [11-12], which would result in negative environmental ramifications to the cities.

Growth in terms of economy of a country is generally aided by industrial activities which further drive the process of urbanisation. The urban metabolism approach is seen to provide a structured format in specially-designed tools and could identify potential environmental impacts that can be avoided as well as society and environmental benefits that can be enhanced. This study is anticipated to provide a novel environmental assessment by using input-output analysis for national and regional assessment while serving as a platform for enhanced strategic environmental assessment where a suitable and consistent method is applicable across countries and economic sectors. It enables comparison at the local and international level in the evaluation of the environmental

impact in line with the mission to improve climate change urban policies and regulations.


The basic understanding in economic input-output table

was to see the product branches by lines and columns. The

input-output table shows the amount of money paid for each

economic sector (lines) to manufacture the products of the

other sectors (columns). The most significant part of the Input

Output table is to derive the “inverse Leontief” matrix or the

matrix (I-A)-1. The table can be generated by a spreadsheet

software function (Microsoft Excel). The flow of computation

along the economic data to environmental outcomes and the

step-by-step computation are described in Figure 1. Economic

input-input analysis takes into account the monetary fluxes it

generates in the total national economy and the average

environmental impact of each economic sector in carbon

dioxide emissions per currency.

39 Health Scope, 2020, Vol. 3(3) Shafie et al.

Figure 1. The Flow of Computation for Environmental-

Economic Input-Output Model

This exercise provides an understanding on environmental

impact assessment because large-scale environmental impacts

in a particular area can be studied. Calculation and analysis

on greenhouse gas emissions, carbon and water footprints,

pollution, and embodied energy can be produced as well.


3.1 Individual Consumption

The first co-efficient matrix generated was the individual consumption. The total Malaysian individual consumption via 120 economic activities was RM 33, 021.98 per year. Hence, the crude Malaysian individual consumption per month was RM 2751.83. This finding concurs with the data provided by the Department of Statistics, Malaysia where the average monthly individual income in 2004 was RM 3,250 [13]. This suggested that an average Malaysian purchased higher number of items or services with higher income, indicating that one would have a greater purchasing power as their income increases.

Based on the 120 economic activities, the “semi-conductor devices, tubes and circuit boards” was the highest economic activity with a total individual consumption of RM 3708.18/cap/year. The second highest individual consumption was from the “office, accounting and computing machinery” that contributed RM 3039.69/cap/year followed by “TV, radio receivers and transmitters and associate goods” with RM 2219.48/cap/year. The summary of the top five highest economic activities for Malaysian individual consumption is illustrated in Table 1.

Table 1. The Highest Economic Activities for Malaysian Individual Consumption

Malaysia's electronics industry is leading the manufacturing sector. The production is in high capacity with a variety of semiconductor gadgets, high-end consumer electronic and information and communication technology (ICT) products being manufactured [14]. Semiconductor devices, tubes and circuit boards are produced using silicon and different materials that encourage the controlled conduction of power. The products incorporate memory chips, picture sensors, diodes, transistors, and sun-powered cells. These items are frequently utilised within machines, telecom items, and different electronic gadgets and commonly being used in Malaysia as important tools for communication. The results show that host-site institutional support is more important than regional integration in influencing firms’ capacity to upgrade their technological capabilities [15].

Alongside other developing nations, Malaysia is moving towards becoming one of the developed countries in the world where technology enhances profitable revenue to the country. Hence, the use of computers and machines facilitates manpower and provides technical support to the industry. Manufacturing of office, accounting and computing machinery has been the second highest economic activity in Malaysian individual consumption. It is important as it is an ICT-producing activity that provides a large number of investment goods to other sectors [16], [17].

3.2 Individual Consumption Based on Key Categories

An analysis was also performed in this study to determine the individual consumption by economic activities based on the ten key categories and the results are presented in Table 2. The highest Malaysian individual consumption was in “manufacturing industry and construction” with RM 13,542.90 per capita per year. Malaysia has been developing over the years, enhancing the economic base for the country.

No. Economic activities Individual

consumption

(RM/cap/year)

Percentage

1. Semi-conductor devices,

tubes and circuit boards

3708.18 11.23

2. Office, accounting and

computing machinery

3039.69 9.21

3. TV, radio receivers &

transmitters &associate

goods

2219.48 6.72

4. Crude oil and natural gas 1977.66 5.99

5. Motor Vehicles 1049.41 3.18

Health Scope 40

Table 2. Malaysia’s Individual Consumption by Ten Key

Categories (RM/capita/year).

The manufacturing industry is seen as the leading boundary of modernisation and skilled job establishment as well as a fundamental source of a mixture of spill over effect [18]. Thus, this industry has been seen as a crucial factor in developing the country. The distribution of individual consumption over economic activities is illustrated in Figure 2.

Figure 2. Distribution of Individual Consumption by

Economic Activities Based on Ten Categories

This can be supported with extensive urbanisation, where the increment of demands and supplies from the industries are indispensable. According to [19] there were relatively young and large working-age populations in the demographic characteristics that could provide support for income growth. It is anticipated that young people tend to spend more on non-essential goods and services, making them a generally high marginal propensity consumers. The least individual consumption by sector was “forest conversion” with RM 71.60 per capita per year.

3.3 Greenhouse Gas Emission Intensity

The findings presented in Table 3 are the factors of greenhouse gas emissions intensity (kg CO2-eq./RM) for each economic sector. Greenhouse gas emission intensity value is very significant in the case where there is a need to increase production while minimising its influence on climate change. The particular economic sector will need to reduce greenhouse gas emissions per unit of product [20].

Table 3. Summary of Malaysian GHG Emission Intensity in 2005 by Ten Key Categories of Economic Activities

Key category GHG

emission

(kt CO2-eq. / year)

Gross

output

(RM)

GHG

emission

intensity (kg CO2-eq.

/RM)

Transport 35587 265520.44 0.134

Residential 1812 15960.69 0.114

Energy industry 58486 1677288 0.035

Forest conversion 24111 1244083 0.019

Rice production 1861 225368.33 0.008

Mineral products 9776 2077312.1 0.005

Commercial 2122 833229.1 0.003

Construction &

manufacturing industry

26104 11977942 0.002

Emissions & removals from soil

4638 5302135 0.001

Emission & removals

from oil and gas system

4638 2414051 0.001

The computation of greenhouse gases emission intensity is necessary for this study to determine the Global Warming Potential by economic sectors and also to measure the total greenhouse gases emission for a region or a city.

3.4 Impact Assessment

The highest contributor for greenhouse gases emission (GHG) comes from the activities of “crude oil and natural gas sector”. “Petroleum refinery” activities come in second followed by “motor vehicle”, “office, accounting and computing machinery” as well as “semi-conductor devices, tubes and circuit boards”. The summary of greenhouse gases emission for the above activities is shown in Table 4.

The “crude oil and natural gas” category contributes the highest in terms of to Malaysia greenhouse gas emission with 0.732 kg CO₂-eq./day. This is mainly because Malaysia is one of the crude oil producers in the world and ranked 29th in the world concerning crude oil production. Energy is known to be a driver to development. The main energy supply shall be secured together with green energy initiatives [21].

No. Economic activities Individual consumption

(RM/cap/year)

1 Forest conversion 71.60

2 Rice production 85.40

3 Residential 731.40

4 Emissions & removals from soil 871.90

5 Emission from oil and gas system 926.90

6 Mineral products 1,438.00

7 Energy industry 2,305.00

8 Transport 2,645.90

9 Commercial 7,588.80

10 Manufacturing industry and

construction

13,542.90


Table 4. Highest Economic Activities in terms of Greenhouse Gas Emission

The second highest GHG contributor is the “petroleum refinery” activities with 0.597 kg CO₂-eq./day. Its processes comprising heavy to light hydrocarbon distillation and crude oil separation generate substantial amount of GHG. GHG emission from petroleum refinery directly contributes towards the amount of Global Warming Potential. The third highest contributor to Malaysia GHG emission is from “motor vehicles” activities with 0.501 kg CO₂-eq./day. In 2005, a total of 14,816,407 motor vehicles were registered in Malaysia.

The electrical and electronics industry registered massive growth in exports, thus maintaining Malaysia’s position as a net exporter of electrical and electronic products. Malaysia’s production and export of semiconductors ranked among the top five in the world during the period between 2000 to 2004. Exports of electronic products by sub-sectors: (i) office, accounting and company machinery; RM 115.4 billion and (ii) semi-conductor devices, tubes and circuit boards; RM 100.4 billion; make “office, accounting and computing machinery” activities the fourth highest GHG emission contributor in Malaysia with 0.443 kg CO₂-eq./cap/day followed by “semi-conductor devices, tubes and circuit boards” with 0.299 kg CO₂-eq./cap/day.

The economic activities linked with their demand are shown in Figure 3.

Figure 3. Economic Activities in terms Of Carbon Dioxide Emission for Malaysia

From the analysis, it is shown that the demand for transport and communication creates the highest greenhouse gases emissions (2.536 kg CO₂-eq./cap/day) followed by the manufacturing industries with 1.554 kg CO₂-eq./cap/day. The third highest greenhouse gases emitter is from the energy industry sector with 0.867 kg CO2-eq./cap/day. In the United States of America, the largest source of its greenhouse gas emissions is “electricity production” which accounted for 31 %

of U.S. greenhouse gas emissions in 2013, followed by “transportation” at 27 % [22].

Though the sequence is different, the dominating economic sectors seem to reach an outcome. Energy-related carbon dioxide production accounts for 80 percent of these emissions; mostly from burning coal, oil, and natural gas [22].

3.5 Sensitivity Analysis

Often, household expenditure data refer to the national scale and if the data need to be downscaled to the smaller region such as municipality or city level, some uncertainty may be introduced [23]. Therefore, a sensitivity analysis was undertaken to evaluate the influence of using different estimates of household expenditure. Sensitivity analysis is a methodology in uncertainty analysis and is thus important to both calibration and validation. It investigates if the variation in the output of a model is attributed to the variation of its output [24].

Assumption A: Due to lack of data available from the municipalities, the total amount of taxes collected in each municipality is assumed as the total consumption of goods and services of an individual living in each municipality.

Assumption B: A factor was calculated to provide an assumption in order to determine the GHG emission in each municipality. The GHG emission for each municipality is obtained by multiplying each factors with respect to its municipality with Malaysia GHG emission.

The greenhouse gas emissions at four different municipalities (Putrajaya, Kajang, Shah Alam and Klang) were calculated to provide an overview of the environmental impacts related with the consumption of goods and services in each municipality by using a set of data obtained at national level. These data were downscaled to municipality level, which had introduced some uncertainty. Therefore, a sensitivity analysis was undertaken to determine how different values of an independent variable will impact a particular dependent variable under a given set of assumptions [25].

The sensitivity analysis provided the weightage to each of the municipalities against the national data. With the designated weightage or factor, the municipality greenhouse gases emissions could be computed and generated. The Malaysia’s greenhouse gas emission was then multiplied with these factors to compute the GHG emission in all the four municipalities (Table 5).

Economic activities GHG emission

(kg CO2-eq./day)

Crude oil and natural gas 0.732

Petroleum refinery 0.597

Motor Vehicles 0.501

Office, accounting and computing machinery 0.443

Semi-conductor devices, tubes and circuit boards 0.299

0 1 2 3

Transport…

Emissions…

Rice…

Forest…

Commercial

Manufacturi…

Carbon dioxide emission

Eco

no

mic

act

ivit

ies

Motor Vehicles

Land Transport

Water Transport

Air Transport

http://www.epa.gov/climatechange/ghgemissions/sources.html

Health Scope 42

Table Error! No text of specified style in document.. Sensitivity Analysis of Greenhouse Gas Emission

From the expenditure data at municipal level, Kajang lead the

individual expenditure with RM 411.49, followed by

Putrajaya (RM 339.85), Shah Alam (RM 293.20) and Klang

(RM 195.65). According to [26] Putrajaya recorded the

highest mean monthly household consumption expenditure

(RM 5,627), followed by Kuala Lumpur (RM 5,559) and

Selangor (RM 4,646).

The mean monthly household consumption expenditure for

Malaysia increased from RM 2,190 in 2009 to RM 3,578 in

2014 which was 9.8% per annum at the nominal value. While

at a real value, the annual growth rate is 7.5% for the same

period. The mean monthly

household consumption expenditure in urban area showed an

increment at a rate of 9.3% annually from RM 2,465 to RM

3,921 while in rural also increased at a rate of 8.4% annually

from RM 1,599 to RM 2,431 for the period of 2009 to 2014.

4. CONCLUSION

This study has explained the formulation of the environmentally extended input-output analysis. The purely economic input-output table has given out some significant variables and outputs when appropriately expanded with environmental datasets. The initial economic data in terms of input-output transactions coupled with greenhouse gases emission inventory could derive greenhouse gases emission for each economic activities. It also allows analysis particularly in supply chain, improving on the results or assisting to identify errors. The use of publicly available data to assemble the matrix representation enables comprehensive assessment of the environmental impacts of a product or service with effective cost, time and manner.

ACKNOWLEDGEMENTS

Authors would like to thank Erasmus Mundus Man, Health, Environment and Biodiversity in Asia (MAHEVA) research grant from the European Commission for the methodological development in Universitat Autonoma de Barcelona, Spain.

REFERENCES

[1] Gao, J., Liu, Y., "Deforestation in Heilongjiang Province of

China, 1896–2000: Severity, spatiotemporal patterns and

causes", Applied Geography, 35(1-2) 345-352, 2012.

[2] Salahodajev, R., “Over-urbanization and air pollution: a cross-

country analysis,” Int Eco Letters, 69–75, October, 2014.

[3] Argüeso, D., et al., "Temperature response to future

urbanization and climate change", Climate Dynamics, 42, (7-

8), 2183-2199, 2013.

[4] Grimm, N., et al., "The impacts of climate change on

ecosystem structure and function", Frontiers in Ecology and

the Environment, 11(9), 474-482, 2013.

[5] Zell, R., "Global climate change and the emergence/re-

emergence of infectious diseases", International Journal of

Medical Microbiology Supplements, 293, 16-26, 2004.

[6] Adhikari, S., et al., "A study of the relationship between

infectious diseases and health economics: some evidences

from Nepal", Asian Pacific Journal of Tropical Disease, 6(6),

437-442, 2016.

[7] Tong, M., et al., "Perceptions of capacity for infectious disease

control and prevention to meet the challenges of dengue fever

in the face of climate change: A survey among CDC staff in

Guangdong Province, China", Environmental Research, 148,

295-302, 2016.

[8] Petsch, S., et al., "Modeling, Monitoring, and Visualizing

Carbon Footprints at the Urban Neighborhood Scale", Journal

of Urban Technology, 18(4), 81-96, 2011.

[9] McGranahan, G., "Realizing the Right to Sanitation in

Deprived Urban Communities: Meeting the Challenges of

Collective Action, Coproduction, Affordability, and Housing

Tenure", World Development, 68, 242-253, 2015.

[10] World Health Organization (WHO), “What are the health risks

related to overcrowding?”, 2014.

[11] Khreis, H., et al., "The health impacts of traffic-related

exposures in urban areas: Understanding real effects,

underlying driving forces and co-producing future

directions", Journal of Transport & Health, 3(3), 249-267,

2016.

[12] Shahrokni, H., “Smart Urban Metabolism. KTH Royal

Instritute of Technology”, 2015.

[13] Malaysia Department of Statistics Malaysia, Chapter 5: Urban

Environment/Human Settlements, 2013.

[14] Malaysian Investment Development Authority, Electrical and

Electronic. [Online]. Available:

http://www.mida.gov.my/home/electrical-and-

electronic/posts/. [Accessed: Jan. 6, 2016].

[15] Rasiah, R., Shan, Y., "Institutional support, regional trade

linkages and technological capabilities in the semiconductor

industry in Malaysia", Asia Pacific Business Review, 22(1),

165-179, 2014.

[16] Ghosal B., “An Impact of ICT on the Growth of Capital

Market-Empirical Evidence from Indian Stock Exchange”

Information and Knowledge Management, 2(11), 1, 2011.

[17] Schreyer, P., "The Contribution of Information and

Communication Technology to Output Growth", OECD

Science, Technology and Industry Working Papers, 2000.

[18] Shahbaz, M., et al., "The effect of urbanization, affluence and

trade openness on energy consumption: A time series analysis

Key category

GHG emission (kg CO₂-eq./day)

Malaysia Kajang Putrajaya Shah Alam

Klang

Weightage

factor

1.8 0.5 3.8 3.1

Energy

industry

0.823 1.481 0.412 3.1274 2.5513

Transport 1.305 2.349 0.653 4.959 4.0455

Constructio

n and

1.228

2.210

0.614

4.6664

3.8068

manufacturi

ng industry

Emission from oil and

gas system

0.597

1.075

0.299

2.2686

1.8507

Commercial 0.535 0.963 0.268 2.033 1.6585

Forest

conversion

0.005 0.009 0.003 0.019 0.0155

Emissions & removals

from soil

0.025

0.045

0.013

0.095

0.0775

Mineral products

0.112 0.2017 0.056 0.4256 0.3472

Rice

production

0.002 0.0034 0.001 0.0076 0.0062

Residential 0.238 0.429 0.119 0.9044 0.7378

Total 4.870 8.766 2.435 18.506 15.097

https://www.statistics.gov.my/index.php?r=column/cglossary2&menu_id=eWd2VFdIZ2xpdzBmT2Y0a0pweDcwQT09&keyword=UmVqMGkxQ29QbE9DejZLRnEzTlRlUT09&release=1



http://www.mida.gov.my/home/electrical-and-electronic/posts/

http://www.mida.gov.my/home/electrical-and-electronic/posts/


in Malaysia", Renewable and Sustainable Energy Reviews, 47,

683-693, 2015.

[19] Central Bank of Malaysia. “Economic Developments 2010”

Kuala Lumpur, 2010. [Online].

Available: http://www.bnm.gov.my/files/publication/ar/en/20

10/cp01.pdf.

[20] Bonesmo, H., et al., "Greenhouse gas emission intensities of

grass silage based dairy and beef production: A systems

analysis of Norwegian farms", Livestock Science, 152, (2-3),

239-252, 2013.

[21] Foo, K., "A vision on the opportunities, policies and coping

strategies for the energy security and green energy

development in Malaysia", Renewable and Sustainable Energy

Reviews, 51, 1477-1498, 2015.

[22] United States Environmental Protection Agency. Sources of

Greenhouse Gas Emissions. 2015. [Online]. Available:

http://www3.epa.gov/climatechange/ghgemissions/sources.ht

ml.

[23] Dias, A., et al., "Environmentally extended input–output

analysis on a city scale – application to Aveiro

(Portugal)", Journal of Cleaner Production, 75, 118-129,

2014.

[24] Pianosi, F., et al., "Sensitivity analysis of environmental

models: A systematic review with practical

workflow", Environmental Modelling & Software, 79, 214-

232, 2016.

[25] Mattila, T., “Input-output analysis of the networks of

production, consumption and environmental destruction in

Finland”, Aalto University, Finland, 2013.

http://www.bnm.gov.my/files/publication/ar/en/2010/cp01.pdf

http://www.bnm.gov.my/files/publication/ar/en/2010/cp01.pdf

http://www3.epa.gov/climatechange/ghgemissions/sources.html

http://www3.epa.gov/climatechange/ghgemissions/sources.html

Health Scope 44


RESEARCH ARTICLE

Caregivers’ burdens and its’ association with activities of daily living performance of individuals with stroke: A cross-sectional study

Nur Syafrawa Nabihah Mohamad Razali1, Siti Salwa Talib2, Nor Faridah Ahmad Roslan3,

Ahmad Zamir Che Daud1*

1Centre of Occupational Therapy, Faculty of Health Sciences, Universiti Teknologi MARA Cawangan Selangor Kampus

Puncak Alam, 42300 Bandar Puncak Alam, Selangor, Malaysia; 2Centre of Occupational Therapy, Faculty of Health

Sciences, Universiti Teknologi MARA Cawangan Pulau Pinang, 13200 Kepala Batas, Pulau Pinang, Malaysia; 3Department of Rehabilitation Medicine, Faculty of Medicine, Universiti Teknologi MARA Cawangan Selangor, 47000

Sungai Buloh, Selangor, Malaysia

Abstract:

Stroke leads to disability and has become the leading cause of dependence in daily life activities.

Caring of individuals with stroke requires a lot of energy, time and resources to fulfil the

individuals’ needs. This situation leads to burdens and strains to the caregivers. There is a limited

study on the relationship between Activities of Daily Living (ADL) performance of stroke’s

individuals with caregivers’ burdens in literature. Thus, the aim of this study was to examine the

association between ADL performance of individuals with stroke and the caregivers’ burdens. A

cross-sectional study was conducted with 59 individuals with stroke and their caregivers aged

ranged from 20 to 86 years old. It was found that there was no significant correlation between the

ADL performance of the individuals with stroke and the caregivers’ burdens. However, there was

a significant association found between caregivers’ burdens and the duration of having the stroke

(p=0.030). This study suggests that caregivers’ burden is not associated with dependency level in

ADL of individuals with stroke, but the duration of having the stroke does.


Ahmad Zamir Che Daud


Keywords: Activities of daily living, caregivers’ burden, stroke

1. INTRODUCTION

There was a rise in the incidence and prevalence of stroke in Malaysia between 2010 to 2014 [1]. More than half of stroke survivors have initial mobility deficits and more than 30% of the survivors are unable to ambulate by themselves six months after the stroke [2]. This situation limits the individuals’ ability to perform Activity of Daily Living (ADL). Some individuals with stroke require care from family members or professional caregivers [3]. A huge physical, mental, and financial burden for the individuals, their families, and society might be contributed by the long duration of disabilities [4]. A study reported individuals’ functional status (ADL) was found to be one the strongest predictors of total burden among Polish populations [5].

Malaysian has a unique culture in terms of value, norms, and customs, which would result in a different finding on caregivers’ burdens. The diversity in ethnic and races, which have different religion and beliefs in Malaysia will have an influence on the level of burden among caregivers. A study suggested that non-white caregivers have lower levels of caregiver stress, burden and depression as compared to the white caregiver [6]. Therefore, findings from previous

studies lead to further exploration given the multi-ethnic and culture of the Malaysian population.

Sociodemographic data of caregivers and individuals with stroke also influence the caregivers’ burdens. Previous studies indicates that female gender, low educational attainment, residence with the care recipient, higher number of hours spent caregiving, depression, social isolation, financial stress, and lack of choice in being a caregiver are risk factors for caregivers’ burdens [7–10]. Malaysian caregivers would have different results pertaining to caregivers’ burdens, especially when culture and racial factor is considered.

Performances in ADL may influence the level of caregivers' burdens. Caregivers who experience high levels of burden were reported to have deteriorations in psychological well-being and overall quality of life [11]. Therefore, it is important to enhance ADL performance of individual with the stroke, which directly can reduce the caregivers' burdens. Occupational therapist needs to play a significant role not only in improving the ADL performance among individuals with stroke, but also to decrease the caregivers' burdens and also improve their quality of life.

45 Health Scope, 2020, Vol. 3(3) Nur Syafrawa Nabihah et al.

A significant amount of time spent, financial resources and energy are the characteristics of caregiving; to extend in which the stroke’ individuals or care receiver being dependent on the effort of caregivers to the point where it exceeds the normal care. Due to lack of other available alternatives or due to the strong wish of the caregivers to stays in a home setting, close relatives and spouses were often involved in caregiving over longer periods and they performed it voluntarily. Displeasing activities such as helping an incontinent individuals with bladder management or helping a physically impaired in transferring can be physically exhausting as it was part of caregiving tasks [12]. However, most studies conducted in Malaysia were on the burden of the caregivers and the underlying factors for the burden, and the coping strategies for the caregivers, but there is limited study on the association of dependency level in ADL and caregivers’ burdens in taking care individual with stroke.

Therefore, the aim of this study was to identify the relationship between the level of ADL performances of individuals with stroke and the caregivers’ burdens in Malaysia. This study also explored the association between sociodemographic characteristics of caregivers with stroke’s individual ADL performance and burden.


2.1 Design

A cross-sectional survey was conducted to 59 individuals with stroke and their primary caregivers from January to May 2019. This design was chosen because it allows researchers to investigate associations between contributing factors and the outcome of interest, which is caregivers’ burdens [13]. Besides, a cross-sectional study is generally quick, easy, and cheap to perform and it is often based on a questionnaire survey [14], [15]. This study was conducted at few branches of the National Stroke Association of Malaysia (NASAM); Petaling Jaya, Ampang, Melaka, Johor Bahru and Ipoh. Ethical approval from Ethic Committee Universiti Teknologi MARA (UiTM) has been obtained to conduct this study (Approval no: 600-IRMI 5/1/6). Approval to collect data in the NASAM settings has been granted by the NASAM headquarters. Confidentiality and anonymity of the participants were guaranteed to ensure the privacy of their information

2.2 Sampling

The participants involved in this study were individuals with stroke and their caregivers. Purposive sampling was used to recruit them in this study. Individuals with stroke who fulfilled the following criteria; (i) a confirmed diagnosis of stroke from medical professionals; (ii) absence of other disabling and/or psychiatric conditions; (iii) and independent in activities of daily living (ADL) before stroke were included in this study. They were excluded if; (i) have other associated neurological problems; (ii) stay in the hospital or long-term care facilities; (iii) have a severe communication impairment; (iv) have an unstable condition; (v) and refuse to participate in the study. As for caregivers, the inclusion criteria were; (i) the main person engaged in the caregiving (ii) does not receive any payment for providing care; (iii) and does not provide long-term care for another person at the

same time, The caregivers were excluded if; (i) they refused to participate, (ii) have severe cognitive and/or communication impairment; and (iii) could not read and understand Malay to complete Zarit Burden Interview.

The sample size was calculated using G-power software. The number of sample was based on empirical guidelines for Interpreting the Magnitude of Correlation Coefficients from Hemphill[16]. By using distribution of correlation for assessment review in the upper third; the coefficients slope was chosen is 0.35, which indicates medium effect size that produces power (1-β err prob) of 0.80 resulting a total sample size of 59.

2.3 Data collection procedure

Data was collected using demographic questionnaire, Modified Barthel Index (MBI) and Zarit Burden Interview. Level of ADL performance of the stroke individuals was measured using the Modified Barthel Index (MBI) and caregivers’ burden was measured by using Malay Zarit Burden Interview (ZBI).

The Modified Barthel Index (MBI) was developed by Shah et al., [17] to evaluate the physical functional ability in individuals with disabilities. The index of 10 items includes personal hygiene, bathing, feeding, toilet use, stair climbing, dressing, bowel control, bladder control, ambulation, and transfer. Each item is scored from 0 (unable to perform the task) to 15 (fully independent). A score of 0-20 indicates total dependence, 21-60 indicates severe dependence, 61-90 indicates moderate dependence, 91-99 indicates slight dependence, and 100 indicates independence. The Cronbach's alpha coefficient and KMO statistical magnitude of MBI were 0.916 and 0.854, indicating the high reliability and validity of this index [18]. This observational assessment was conducted by an occupational therapist to examine the performance of individuals with stroke in ADL.

Malay Zarit Burden Interview (MZBI) is a self-report measure was used to examine (i) the caregiver subjective burden with a 22-item questionnaire on a 5-point scale 0 (never) to 4 (nearly always); and (ii) two subscales measuring personal strain and role strain [19]. Item ratings are summed to achieve a total score that can range from 0 to 88, with higher scores represented greater burden. The score of 0-21 indicates little or no burden, 21-40 indicates mild to moderate burden, 41-60 moderate to severe burden and 61-88 indicates severe burden. The MZBI was demonstrated fairly good psychometric properties in assessing the caregivers’ burdens in local Malaysian population with high internal consistency (α = 0.898) [20].

2.4 Data analysis

Statistical Package for Social Science (SPSS) version 25.0 software was used to analyse all the data collected. The demographic characteristics of the participants were calculated using the descriptive statistic. Normality testing was done using the Shapiro-Wilk test and the assumptions of normal distribution were violated. The association between ADL performance and the caregivers’ burdens was analysed using Spearman’s correlation test. The associations of caregivers’ burden and ADL performance of individuals with demographic characteristics were analysed using spearman correlation, Mann-Whitney U and Kruskal Wallis test.

Health Scope 46


3.1 Description of participants

A total of N=59 participants from National Stroke Association of Malaysia (NASAM) were involved in this study. Table 1 displays descriptive statistics of participants’ demographic data. The age distribution of the participants ranged between 20 to 86 years old with a median age of 68.00 (12.00) for the individuals with stroke and 60.00 (23.00) for caregiver’s age. The involvement of male participants with stroke were n=29 (49.2%) and n=23 (39%) for caregivers while the involvement of female participants with stroke were n=30 (50.8%) and n=36 (61%) for caregivers. The median duration of having stroke was 24.00 (52.00).

Table 1. Descriptions of participants’ demographic statistics

Demographic characteristics

N % Median (IQR)

Age of stroke’ Individuals

- - 68.00 (12.00)

Duration of stroke - - 24.00 (52.00)

Caregivers’ age - - 60.00 (23.00)

Gender of

Male

Female

29

49.2

-

30 50.8 -

Caregivers’ gender

Male

Female

23

39

-

36 61 -

Caregivers’ marital status

Single

Married

13

22

-

46 78.0 -

Caregivers’ employment status

Yes

No

12

20.3

-

47 79.7 -

Caregivers’ monthly income

<RM4,000

>RM4,000

46

78.0

-

13 22.0 -

Caregivers’ race

Malay

Chinese

Indian

9

15.3

-

43 72.9 -

6 10.2 -

Relationship with

stroke’s individuals

Spouse

Family members

34

25

57.6

42.4

-

-

Receiving help from others

Yes

No

33

55.9

-

26 44.1 -

* Interquartile Ranger (IQR)

Majority of the participants were Chinese (n=43, 73.9%), while the minority of the participants were Malay (n=9, 15.3%) and Indian (n=6, 10.2%). More than half of the caregivers were married (n=46, 78%), while (n=13, 22%) were single. 47 caregivers (79.7%) were not working and the rest (n=12, 20.3%) were still working. Most of the caregiver has monthly income of <RM4,000 (n=46, 78%), while the other (n=13, 22%) has a monthly salary of >RM4,000. The number of caregivers receiving help from others in caregiving was n=33 (55.9%) while caregivers that does not receive help was n=26 (44.1%). Most caregivers were the spouse for the individuals with stroke (n=34, 57.6%), and the rest were family members.

3.2 Level of performance in ADL and caregivers’ burdens

Table 2 shows the median and IQR of MBI and MZBI were 74.00 (47.00) and 25.00 (22.00) respectively. The minimum score was 5.00 while the maximum score was 100.00 for ADL performance of individuals with stroke. Meanwhile, the minimum score for MZBI was 1.00 and the maximum is 66.00 for the caregivers’ burdens. Out of the 59 participants with stroke, there were five participants with total dependency level, 18 participants with severe dependency level, 20 participants with moderate dependency level, 15 participants with slight dependency level and one was independent.

Table 2. Level of ADL activities and Level of Caregivers’ burdens

(N=59) Minimum Maximum Median (IQR)

MBI 5.00 100.00 74.00 (47.00)

MZBI 1.00 66.00 25.00 (22.00)

3.3 Relationship between performances in ADL and caregivers’ burdens

A Spearman correlation coefficient was conducted to examine the correlation of ADL performance of individuals with stroke and the caregivers’ burdens. The results showed the ADL performance was not statistically correlated with the caregivers’ burdens (r= -0.086, p=0.52). Table 3 showed the result of relationship between level of ADL performance and caregivers’ burdens.

47 Health Scope, 2020, Vol. 3(3) Nur Syafrawa Nabihah et al.

Table 3. Relationship between level of ADL performance and level of caregivers’ burden

Correlation Zarit Caregivers’ burdens

Modified Barthel Index r = -0.086

p-value = 0.52

3.4 Association between socio-demographic data with caregivers’ burdens and ADL performance of individual with stroke

As for the association between sociodemographic characteristics and caregivers’ burden, the duration of having stroke and caregivers’ burdens was found to have significant negative weak correlation through Spearman correlation coefficient (r= -0.281, p= 0.03). Surprisingly, the result suggests the higher the duration of stroke, the lower the caregivers’ burden. Other demographic characteristics of individuals with stroke and their caregivers were found not to be associated with burdens.

Results of the association between demographic characteristics and individuals’ with stoke ADL performance indicate there was a significant negative moderate correlation between age of individuals with stroke and ADL performance (r= -0.414, p<0.01). It can be concluded that the older the age of the individual with stroke, the poorer the performance in ADL. Additionally, there was significant association between the relationship of caregiver and the ADL performance of individuals with stroke (z=-2.54, p=0.01). This result suggests that individuals with stroke who was taken care by spouse had greater ADL performance (M=84, IQR=28.25) than those who was taken care by other family members (M=50.00, IQR=48.00). Other demographic characteristics of individuals with stroke and their caregivers were found not to be associated with ADL performance.

3.5 Discussion

This study examined the caregivers’ burdens and its’ association with ADL performance of individuals with stroke. The main finding emphasized in the present study was that there is no significant correlation between caregivers’ burdens and ADL performance of the individuals with stroke. The finding contradict with a study that reported there is statistically significant difference between impairment in activity of daily living of older people with the level of caregivers’ burdens in India [21]. This contradiction is probably due to the caregivers in this present study felt it is their responsibility to provide care to the needy family members[22]. A study reported that personal fulfilment is one of the several benefits when taking care of their loved ones[23].

Findings of the study show that 20 out of the 59 (33.9%) participants has moderate dependency level, followed by 18 (30.5%) participants having severe dependency, 15 (25.4%) participants with slight dependency, 5 (8.5%) participants with total dependency level and 1 (1.7%) is totally independent. Most of the stroke survivors are able to perform ADL with maximum to moderate help from caregivers. The level of independency in ADL performance are diverse across cultures and diagnosis. For instance, a study reported

that there is high dependency level in ADL among elderly in India[21].

As for the caregivers’ burdens, 25 out of 59 (42.4%) participant felt little to no burden, 21 (35.6%) participants felt mild to moderate burden, 9 (15.3%) participants felt moderate to severe burden and only 4 (6.8%) of the participants felt severe burden when caring for individuals with stroke. This finding contradicts with a study that found the caregivers of people with dementia have high and persistent rates of burden[24].

Previous studies reported that the relationship with stroke survivor, receiving help in caregiving of the individuals with stroke, duration of having stroke, income level of the family, caregivers’ health status and hours of caregiving per day were the factors that significantly increase the level of burden and strains experienced by the informal caregivers[7], [25], [26]. However, in this study the duration of having stroke was the only variable that has significant difference with level of burden among caregivers. The other demographic characteristics such as individuals’ with stroke and caregivers’ age, gender of individuals with stroke and their caregivers, caregivers’ marital status, caregivers’ working status, caregivers’ monthly income, race of individuals with stroke and their caregivers, caregivers’ relationship with individuals with stroke, and receiving help in caregiving seemed not to have any association with caregivers’ burdens. A study discovered that some of the factors that could contribute caregivers’ burdens among Korean population includes caregiver’s mental health states, caregiver’s anxiety, caregivers being daughter in law, individual’s with stroke mental health states, and their physical deficits [27], which were not addressed in this present study. Thus, a future study needs to consider these factors when trying to untangle the caregivers’ burdens of the individuals with stroke.

In this study, it was found that the age of individuals with stroke is associated with ADL performance. The younger the stroke individuals, the better the performance in ADL. The relationship between stroke individuals and their caregiver was also found to be associated with ADL performance. Those who were taken care by their spouse were found to be having a higher ADL performance as compared to those who were taken care by other family members. It can be said that the ability to perform ADL tasks were influenced by vast variation of physical and emotional problems that the individuals with stroke had to deal with[28]. The environmental and sociodemographic characteristics also plays a significant role in ADL performance of individuals with stroke.

The major limitation of this study is the samples were relatively small and might not represent the number of individuals with stroke who attended NASAM for rehabilitation. As this is a cross-sectional study, it could not describe the experience of caregivers’ burdens in taking care of individual with stroke more details. Thus, a future study with a larger sample is required and it will be good to incorporate qualitative data to describe the experience of caregivers in taking care of a family member with stroke.

Health Scope 48 4. CONCLUSION

This study suggests that there is no significant association between ADL performance of stroke individuals and the caregivers’ burdens. The caregivers can experience little to no burden, despite the stroke individuals have total dependency in ADL performance. This study found that some of the sociodemographic characteristics of the stroke individuals and the caregivers have significant association with burdens of caregivers and ADL performances. It is found that the duration of having stroke significantly associated with caregivers’ burdens. The longer the duration of having stroke, the less burden felt by the caregivers. The age of individuals with stroke and the individual’s relationship with caregivers were found to be associated with performance in ADL.

ACKNOWLEDGEMENTS

We would like to thank all participants involved in the study and NASAM for their great cooperation. This study was supported by LESTARI Grant (Ref no: 600-IRMI/Dana KCM5/3/LESTARI(223/2017), Universiti Teknologi MARA.

REFERENCES

[1] Z. A. Aziz et al., “Acute Stroke Registry Malaysia, 2010-

2014: Results from the National Neurology Registry,” J.

Stroke Cerebrovasc. Dis., vol. 24, no. 12, pp. 2701–2709,

2015.

[2] T. Umehara et al., “Can the Amount of Interventions

during the Convalescent Phase Predict the Achievement of

Independence in Activities of Daily Living in Patients with

Stroke? A Retrospective Cohort Study,” Journal of Stroke

and Cerebrovascular Diseases, 2018.

[3] K. Jaracz, B. Grabowska-Fudala, K. Górna, J. Jaracz, J.

Moczko, and W. Kozubski, “Burden in caregivers of long-

term stroke survivors: Prevalence and determinants at 6

months and 5 years after stroke,” Patient Educ. Couns.,

vol. 98, no. 8, pp. 1011–1016, 2015.

[4] W. Y. Lin et al., “Predicting post-stroke activities of daily

living through a machine learning-based approach on

initiating rehabilitation,” Int. J. Med. Inform., vol. 111, no.

November 2017, pp. 159–164, 2018.

[5] K. Jaracz, B. Grabowska-Fudala, K. Górna, and W.

Kozubski, “Caregiving burden and its determinants in

Polish caregivers of stroke survivors,” Arch. Med. Sci.,

vol. 5, pp. 941–950, 2014.

[6] C. M. Connell and G. D. Gibson, “Racial, ethnic, and

cultural differences in dementia caregiving: review and

analysis.,” Gerontologist, 1997.

[7] R. D. Adelman, L. L. Tmanova, D. Delgado, S. Dion, and

M. S. Lachs, “Caregiver Burden,” JAMA, vol. 311, no. 10,

p. 1052, Mar. 2014.

[8] H. Rigby, G. Gubitz, and S. Phillips, “A systematic review

of caregiver burden following stroke,” Int. J. Stroke, vol. 4,

no. 4, pp. 285–292, 2009.

[9] H. Rigby et al., “Caring for stroke survivors: baseline and

1-year determinants of caregiver burden,” Int. J. Stroke,

vol. 4, no. 3, pp. 152–158, 2009.

[10] D. J. Camak, “Addressing the burden of stroke caregivers:

a literature review,” J. Clin. Nurs., vol. 24, no. 17–18, pp.

2376–2382, 2015.

[11] M. Grant et al., “Family Caregiver Burden, Skills

Preparedness, and Quality of Life in Non-Small Cell Lung

Cancer,” Oncol. Nurs. Forum, 2013.

[12] C. Fekete, “Caregiver Burden,” in International

Encyclopedia of the Social & Behavioral Sciences: Second

Edition, 2015.

[13] K. A. Levin, “Study design III: Cross-sectional studies,”

Evid. Based. Dent., vol. 7, no. 1, pp. 24–25, 2006.

[14] P. Sedgwick, “Case-control studies: advantages and

disadvantages,” Bmj, vol. 348, p. f7707, 2014.

[15] C. J. Mann, “Observational research methods. Research

design II: cohort, cross sectional, and case-control

studies,” Emerg. Med. J., vol. 20, no. 1, pp. 54–60, 2003.

[16] J. F. Hemphill, “Interpreting the Magnitudes of Correlation

Coefficients,” Am. Psychol., vol. 58, no. 1, pp. 78–79,

2003.

[17] S. Shah, F. Vanclay, and B. Cooper, “Improving the

sensitivity of the Barthel Index for stroke rehabilitation,” J.

Clin. Epidemiol., vol. 42, no. 8, pp. 703–709, 1989.

[18] L. Pei et al., “Factors associated with activities of daily

living among the disabled elders with stroke,” Int. J. Nurs.

Sci., vol. 3, no. 1, pp. 29–34, 2016.

[19] A. Taub, S. B. Andreoli, and P. H. Bertolucci, “Dementia

caregiver burden: reliability of the Brazilian version of the

Zarit caregiver burden interview,” Cad. Saude Publica,

vol. 20, pp. 372–376, 2004.

[20] V. K. Shim, C. G. Ng, and I. Drahman, “Validation of the

Malay Version of Zarit Burden Interview (MZBI),”

Malaysian J. Psychiatry, vol. 26, no. 1, 2018.

[21] S. Ajay, A. Kasthuri, P. Kiran, and R. Malhotra,

“Association of impairments of older persons with

caregiver burden among family caregivers: Findings from

rural South India,” Arch. Gerontol. Geriatr., vol. 68, pp.

143–148, 2017.

[22] A. Z. C. Daud, J. Judd, M. Yau, and F. Barnett, “Barriers

of occupation-based intervention,” Asian J. Qual. Life, vol.

1, no. 4, pp. 1–10, 2016.

[23] L. G. Collins and K. Swartz, “Caregiver care.,” Am. Fam.

Physician, vol. 83, no. 11, pp. 1309–17, Jun. 2011.

[24] H. Brodaty, M. Woodward, K. Boundy, D. Ames, and R.

Balshaw, “Prevalence and predictors of burden in

caregivers of people with dementia,” Am. J. Geriatr.

Psychiatry, 2014.

[25] C. A. Gbiri, O. A. Olawale, and S. O. Isaac, “Stroke

management: Informal caregivers’ burdens and strians of

caring for stroke survivors,” Ann. Phys. Rehabil. Med., vol.

58, no. 2, pp. 98–103, 2015.

[26] Z. Kurtulus Tosun and M. Temel, “Burden of Caregiving

for Stroke Patients and The Role of Social Support Among

Family Members: An Assessment Through Home Visits,”

2017.

[27] S. Choi-Kwon, H.-S. Kim, S. U. Kwon, and J. S. Kim,

“Factors affecting the burden on caregivers of stroke

survivors in South Korea,” Arch. Phys. Med. Rehabil., vol.

86, no. 5, pp. 1043–1048, 2005.

[28] H. A. Haghgoo, E. S. Pazuki, A. S. Hosseini, and M.

Rassafiani, “Depression, activities of daily living and

quality of life in patients with stroke,” J. Neurol. Sci., vol.

328, no. 1–2, pp. 87–91, May 2013.

https://orchid.uitm.edu.my/irmis/view/boundary/application/myresearch/view_details.php?file_no=PPh5ZEhxqSM+LDVu5ZnWlJwx3uqqk4XowuB/n6gax3CLgBO8CbWqSYr/UYcFgQhNK+LUZZFYJSM+qIW95cjckw==

https://orchid.uitm.edu.my/irmis/view/boundary/application/myresearch/view_details.php?file_no=PPh5ZEhxqSM+LDVu5ZnWlJwx3uqqk4XowuB/n6gax3CLgBO8CbWqSYr/UYcFgQhNK+LUZZFYJSM+qIW95cjckw==

Health Scope 49


RESEARCH ARTICLE

Evaluating microbiological quality of kitchen equipment in school canteens

Rosnaini P. Ramli, Hanis Syakira Salman, Lea Nursuria Sariza Sazari, Alia Azmi*

Centre of Environmental Health and Safety, Faculty of Health Sciences, Universiti Teknologi MARA Cawangan

Selangor Kampus Puncak Alam, 42300 Bandar Puncak Alam, Selangor, Malaysia

Abstract:

Food preparation requires constant safety and health supervision and careful handling of raw

materials. Due to the nature of processes occurring during food preparation, equipment within any

food establishment will come into contact with various types of microbiological organism. If

cleanliness is not ensured, equipment such as dishwashing sponges and cutleries can be a source of

contamination, leading to foodborne diseases. This is a serious health hazard, especially if it

occurs within a setting that has vulnerable populations, such as a school canteen. As such, this

study aims to evaluate the microbiological quality of kitchen equipment in school canteens, as well

as the effectiveness of the cleaning agents used in the kitchen. Carried out in three schools in

Puncak Alam, microbiological swabs were obtained from sponges, cutleries and plates used in

each of the school canteen. Based on the microbiological analysis, Salmonella was detected the

highest (mean; 4 log10 cfu/ml), compared to E. coli (mean; 3.4 log10 cfu/ml) and S.aureus, which

showed the least growth (mean; 2.05 log10 cfu/ml).


Alia Azmi


Keywords: Food hygiene, kitchen equipment, microbiological quality

1. INTRODUCTION

Every school, regardless of primary, secondary or kindergarten, has its own canteen. It is commonly known that during cleaning process for kitchen utensils and equipment, sponges are used to eliminate food residues. However, food residues will stick to the sponge surfaces and if stored in favorable environment for the microbial growth, kitchen sponge will turn into a medium that promotes bacterial growth.

Food hygiene is an important factor in food preparation. This is because biological food contamination can easily occur if proper hygiene is not practiced. Biological contamination such as bacteria, viruses, fungi, protozoa and helminthes are major causes of foodborne diseases, all of which with severities that range from mild to chronic, or life threatening, or both [1]. In this study, school canteen is selected because it generally produces a large quantity of meal for the students. The contamination of food by pathogen at school canteen will result in the occurrence of food borne disease amongst a large number of people.

The incidence of food borne disease such as diarrhea and abdominal pains, which are mostly reported by the students after eating at the school are presumed as microbial food poisoning [1]. Incidence of food borne disease highlights the issue of food hygiene preparation, which includes factors such as food handler hygiene, equipment, utensils, storage and raw material preparation. Even preparing meals too

early and stored at incorrect temperature will allow microbial growth.

This study evaluates the microbiological quality of different kitchen equipment at school canteens. The laboratory work includes determining microbial count in used kitchen sponge, identifying presence of bacteria in kitchen cutleries washed by used kitchen sponge and investigating the effectiveness of detergent used in different school canteens.

2. METHODOLOGY

2.1 Determination of Microbial Count in Used Kitchen Sponges

Samples were taken from three different school canteens located in Puncak Alam, Selangor. The schools were labelled with SC1, SC2 and SC3 onsite to differentiate the samples. The samples were taken after receiving permission from the headmaster of each school. The sampling activities took place in the month of October 2019.

During the sampling, aseptic techniques were implemented to avoid cross contamination of the samples. The sample, which is kitchen sponge, is not based on any specific types of sponge but rather depending on whichever type of kitchen sponge used in the school canteen. For SC1 and SC3, polyethylene sponges were used while SC2 used a stainless

50 Health Scope, 2020, Vol 3(3) Rosnaini et al.

steel sponge. Once collected, the samples were stored in an icebox with a temperature of approximately 0 - 4°C and were taken back to the laboratory for culture process.

In the laboratory, the sponge samples were aseptically cut in 8cm³ and placed into Vortex Shaker approximately for 30 seconds. After dilution, the sample was pipetted to make 3 dilutions: 1] 10 times, 2] 100 times and 3] 1000 times. The sample of 100 µl from each serial dilution is pipetted and plated on MacConkey agar for subculture of E. coli and Salmonella spp. and on Mannitol Salt agar for S. aureus. After 72 hours, total colony count for Mannitol Salt Agar are counted [2].

The subculture procedure is conducted for E. coli and Salmonella spp. by referring to the method from [3]. The sub culturing process was done by isolating the target colony streaked onto other agar which is Eosin Methylene Blue and Xylose Lysine Deoxycholate.

2.2 Identification of Bacterial found on Kitchen Cutleries Washed by Used Kitchen Sponges

The contamination of food contact surfaces along with poor handling method by food handlers may increase the risk of foodborne disease through cross-contamination. Contamination of food contact surfaces such as kitchen counter, kitchen cutlery, and equipment is mainly caused by ineffective cleaning procedure of the food contact surface [4]. Dry environmental swab test was used to identify the presence of pathogenic bacteria on kitchen utensils, which is plates and spoons [5].

2.3 Determination of Efficiency of Detergent Used in

School Canteen

Microorganism test - The effectiveness of detergents used was measured by performing the disc agar diffusion method. The samples were tested for three types of bacteria, which were E.coli, Salmonella and S.aureus, similar with the bacteria tested for kitchen sponge. The bacteria were provided by University Teknologi Mara (UiTM) Puncak Alam’s microbiology laboratory. The bacteria was inoculated in peptone broth for 24 hours before pipetted into Muller-Hilton agar.

Detergent selection and sensitivity screening - The detergent selected was based on detergent used by the school canteen which was Glo for SC1, Axion Lemon for SC2 and Kuat Harimau for SC3. The samples were labeled as DSC1, DSC2 and DSC3 representing each school canteen respectively. Since DSC2 and DSC3 is a paste type of detergent, both were reconstituted in sterile water to obtain a stock solution prior to testing. For DSC2 and DSC3, 0.17 g/ml and five-dilution stocks were used for the test. The disk agar diffusion method procedure was done in biosafety cabinet. The petri dish were divided into three sections for each detergent tested. Each section represent one sample that was taken and was labelled (DSC1, DSC2, DSC3). Whatman filter paper sized about 6mm was placed into each section. A drop of sample was dropped on the filter paper. The plates were incubated for 24hours at 37°C. The diameter of inhabitation zones was measured and recorded


3.1 Determination of Microbial Count in Used Kitchen Sponges in School Canteen

This study was carried out in a sterile laboratory in order to ensure no cross contamination. The sample agar plates were incubated for 24 hours at a temperature of 35°C - 37°C for E. coli and Salmonella; and for 72 hours at the same temperature for S.aureus. After the spread and incubation, the bacteria colony formed was counted. The total colony count (Log10 cfu/ml) was then used to determine the level of bacterial growth. The data are presented in the following Figure 1 and 2

E. coli Growth – As presented in Figure 1, the growth of E. coli, which is 4.27 log10 cfu/ml from SC2, was the highest total colony count. This is followed by 4.23 log10 cfu/ml from SC1. The lowest total number of total colony count was for SC3, which is 2.70 log10 cfu/ml. The mean of the total colony count of E.coli for all school canteens was 3.4 log10 cfu/ml.

Figure 1. E.Coli growth in kitchen sponges.

Values are in Log10 (10 in multiplication), Cfu/ml (colony

forming unit per milliliter) according to amount of dilution

respectively

In a study by [6], sponges in households are used daily and is in contact with dishwashing liquid at least twice a day. The study also showed an increasing growth of E. coli in used kitchen sponge, which supports the findings of our study. The study also noted that, although raw food is probably the main source of contamination, other kitchen surfaces such as sink, waste trap, cleaning cloth, face cloths and surrounding area could also act as semi-permanent source or reservoir which harbor and encourage the establishment of bacterial growth. [7] stated that E. coli 0157 serotypes may attach to the sponge more firmly and therefore transfer less frequently than E.coli ATCC11229 serotypes. E.coli 0157 serotypes, unfortunately are associated with foodborne disease such as abdominal cramps, bloody diarrhea and vomiting.

The occurrence of E. coli on kitchen sponge may potentially come from food handlers with poor hand washing hygiene. Even though most of healthy adult will recover from E. coli illness within a week, some particularly young children and older adult may develop life threatening form of kidney failure called hemolytic uremic syndrome [8]. This situation might increase the occurrence of foodborne illness.

Salmonella Growth - Figure 2 presents the number of total colony count for salmonella from three different school canteens at three levels of dilution. Based on the calculated

Health Scope 51

result, the highest total colony count is from SC2, at 5.13 log10 cfu/ml at 10̄³ dilution. This is followed by 4.96 log10 cfu/ml from SC1 at 10̄³ dilution. The lowest total colony count was from SC1, which was 2.51 log10 cfu/ml at 10̄¹ dilution. The mean of the total colony count from Salmonella spp. was 4 log10 cfu/ml

Figure 2. Salmonella growth in kitchen sponges.

Values are in Log10 (10 in multiplication), Cfu/ml (colony

forming unit per milliliter) according to amount of dilution

respectively

Salmonella is a bacterium that is often associated with eggs and egg-containing food. However, according to a study by [9], the highest salmonella detected are from ground chicken. Besides that, Salmonella has also been found on broiler chicken, pasteurized egg products and ground beef. In a study carried out by [10], 15% of sponges and dishcloths contained Salmonella and 20% contained S. aureus. The study also stated that, Salmonella can also be found in 60% of cellulose sponges and 86% of loofahs. The use of same cutting board while preparing raw food and uncooked food such as fruit or salad without disinfection can also transmit the pathogens to the food.

In general, results from this study indicate that the relationship between the types of food prepared by school canteen and the presence of Salmonella on kitchen sponge. This is because common food prepared by the school canteen contains poultry such as chicken, egg and beef, all of which contain Salmonella. The kitchen contact surface in food preparation such as chopping board, kitchen counter and equipment’s can become the medium for the bacterial growth. Hence, the transfer of bacteria to kitchen sponge while cleaning the chopping board, knives and other kitchen equipment can become the cause of bacterial growth

S. aureus Growth – The highest total colony count was 3.95 log10 cfu/ml at dilution 10̄³ from SC3, while the lowest was from SC1 which was not detected on either dilution 10 ̄ ² and 10 ̄ ³ respectively. The mean for all total colony count was 2.05 log10 cfu/ml.

S. aureus is carried in the nose of 25-35% of humans and it is a common cause of serious and life-threatening infection [11]. [12] has stated that S. aureus can survive for long periods once it is shed into the environment and its presence on environmental surfaces in the home has been reported in numerous studies. 30% of S. aureus are found on sponge and cloth. S. aureus have also been identified in various types of food such as meat, poultry, salads and bakery products. S. aureus infection will resolve within 24 to 48

hours of the onset of infection, however the effect of S.aureus can be severe for infants, elderly and immune-compromised patients. The concern is for the children at school that may have low immune system and digestive disorder [13].

Poor hand hygiene may contribute to high levels of S.aureus on the hands of food handlers [14]. The human hands have the ability to transmit pathogens from any contact surfaces or human body, to the food, especially if hands are not properly washed. Not only that, wet equipment such as sponges, dishcloths and sink drain will constantly act as reservoir that harbors and encourage the growth of potential pathogens.

3.2 Identification of Microbiological Found on Kitchen Cutleries

Most of the inspections carried out by health inspectors visually examined kitchen cutleries for issues such as dirt. For our study, kitchen cutleries from the school canteen are taken for microbiological evaluation. The kitchen cutleries chosen were the ones that been washed by used kitchen sponge taken for microbial testing. The dry cutleries were picked for dry swab procedure. The limitation for the test might come from uncontrolled environment since the sampling procedures were done at the school canteen itself and not in sterilized environment. However, the equipments used for the sampling were sterilized with alcohol swab.

Table 1 indicates the presence of bacteria on kitchen

cutleries at the school canteen. The result shows that bacteria

were present in all samples. A study on drinking water and

utensils showed that E. coli could commonly be found on

utensils [15]. The study also stated that the quality of water

and cleanliness of food preparing utensil did not improved

even with the usage of pipe water. The authors also noted

that E. coli count in water and food utensil found in the

treated household are not significantly lower than those

found in the control household

52 Health Scope, 2020, Vol 3(3) Rosnaini et al.

Table 1 Presence of Bacteria in Used Kitchen Cutleries

Furthermore, a study done by [16] stated that spoon and forks had the highest rate of heterotrophic contamination. The factor that causes the high rate of heterotrophic contamination in spoon and forks is due to poor washing technique.

The presence of E. coli , Salmonella spp. and Clostridium spp. has become the special concern and possibly the greatest danger associated with water food processing and drinking purpose [17-19]. The presence of bacteria on kitchen utensil is an indicator of poor sanitary qualities of food utensils, ineffective washing technique as well as poor handling and storage of cleaned utensil. The presence of bacteria on kitchen utensil can be a source of foodborne disease.

3.3 Efficiency of Detergents Used In School Canteen

Based on the lab observations, DSC1 was more efficient at reducing the contamination of Salmonella with 1.9 cm diameter of inhabitation zones. E.coli and S.aureus on the other hand, recorded 1.1cm and 1.3 cm respectively. For DSC2, disc agar containing Salmonella showed 1.5cm diameter of inhibition while E.coli and S.aureus did not show any inhabitation. DSC3 did not show any inhabitation for all bacteria tested.

Detergent is a surfactant or a mixture of surfactants with cleaning properties in dilute solution [20]. As awareness on food hygiene preparation increases, the effectiveness of detergent used has also become a concern. The aim of detergent is to totally eradicate food poisoning microorganism in any food surfaces such a utensils, facilities or equipment’s in the food processing line [21, 22]. However, based on our tests, the detergents have little only on certain types of the bacteria tested. However, based on the studies by [23, 24] detergents, if used correctly, can be highly effective against E.coli and S.aureus

4. CONCLUSION

Kitchen equipment such as used kitchen sponges and cutleries have become one of the important medium for the growth of pathogenic bacteria [25]. This is worrying because food handlers are not able to see the growth of bacteria on their kitchen equipment. Based on the results of this study, pathogenic bacteria are easily found on used kitchen sponges at the school canteen. Hygiene inspections by the District Health Office only covers issues such as proper storage, floor and cooked food. However, the hygiene of food contact surfaces such as kitchen sponges, dishcloths and kitchen cutleries is not included in the inspection. Students are quite susceptible to food borne illness because of their age and immune system that are still developing. Some of the students may also have an allergic reaction to certain products or food. As students also did not have many options in choosing their food at the canteen, it is important to take a proper steps in ensuring all food are safe to consume.

ACKNOWLEDGEMENTS

The authors would like to acknowledge the contribution of Tuan Haji Mohd Pozi as a former supervisor. Gratitude is also extended to all academic and laboratory staff of the Centre of Environmental Health & Safety, UiTM Puncak Alam for their help and assistance.

REFERENCES

[1] Ajao, A. T., & Atere, T. G., “Bacteriological Assesment and

Hygienic Standard of Food Canteens In Kwara State

Polytechnic , Ilorin , Nigeria”. African Scientist, 10(3), 173–

180, 2009.

[2] Jovanovska, S, et al., "Invisible Cohabitants: Investigating the

Microbial Presence in the Kitchen Sponges of Maastricht."

The Maastricht Journal of Liberal Arts 10: 69-83, 2018.

[3] Zulfakar, S, et al., “Microbiological Assessment of Food

Contact Surfaces in Residential College Cafeterias at a Local

University in Malaysia”. Jurnal Sains Kesihatan Malaysia,

16(02), 33–38, 2018.

[4] Hussaini, J., “Effect of Scourers on Utensils and Bacterial

Survival”. Indian Journal of Science, 1(1), 54–57, 2012.

[5] Environmental Swabbing. Australia: NSW Food Authority.

Retrieved from Food Authority NSW Government, 2013.

[6] Erdogrul, Ö., & Erbilir, F., “Microorganisms in Kitchen

Sponges”. Internet Journal of Food Safety. 2000.

[7] Rossi, E. M., Scapin, D., & Tondo, E. C., “Survival and

transfer of microorganisms from kitchen sponges to surfaces

of stainless steel and polyethylene”. Journal of Infection in

Developing Countries, 7(3), 229–234, 2013.

[8] Mayo Clinic., “E. Coli: Symptoms and Causes”, Mayo Clinic,

2019

[9] White, P.L., et al., "Salmonella Enteritidis in meat, poultry,

and pasteurized egg products regulated by the US Food

Safety and Inspection Service, 1998 through 2003." Journal

of food protection 70(3) 582-591, 2007.

[10] Britton, L. A., “Microbiological threats to health in the

home”. Clinical Laboratory Science : Journal of the

American Society for Medical Technology, 16(1), 10–15,

2003.

[11] Roberts, M. C., et al., "Characterization of Methicillin resistant Staphylococcus aureus isolated from public surfaces

on a University Campus, Student Homes and Local

Community." Journal of applied microbiology 110(6)1531-

Food Premises Kitchen Cutlery Presence Of Bacteria

(Yes/ No)

School canteen 1

Spoon Yes

Plate Yes

School canteen 2

Spoon Yes

Plate Yes

School canteen 3

Spoon Yes

Plate Yes

Health Scope 53

1537, 2011.

[12] Scott, E, Susan D, & Maureen C., "A pilot study to isolate

Staphylococcus aureus and methicillin-resistant S aureus from

environmental surfaces in the home." American journal of

infection control 36(6) 458-460.

[13] Lin, J, et al., "Non-hospital environment contamination with

Staphylococcus aureus and methicillin-resistant

Staphylococcus aureus: proportion meta-analysis and features

of antibiotic resistance and molecular genetics."

Environmental Research 150, 528-540, 2016.

[14] Tan, S. L., et al., "Microbiological quality on food handlers'

hands at primary schools in Hulu Langat District, Malaysia."

International Food Research Journal 20(5) 2973, 2013.

[15] Hasan, M. M., Nicolas G., "Bacterial contamination of

drinking water and food utensils: impacts of piped water on

child health in North-Western Bangladesh." Water resources

and rural development 10, 33-44, 2017.

[16] Rakhshkhorshid, M. et al., "Survey of cooking utensils and

dishes microbial contamination rate in the cafeteria of

Zahedan University of medical sciences, 2015." International

Journal of Biomedical and Healthcare Science 6(2) 187-193,

2016.

[17] Nyenje, M. E., et al., "Foodborne pathogens recovered from

ready-to-eat foods from roadside cafeterias and retail outlets

in Alice, Eastern Cape Province, South Africa: public health

implications." International Journal of Environmental

Research and Public Health 9(8) 2608-2619, 2012.

[18] Momba, M.N.B et al., "Abundance of pathogenic

Escherichia coli, Salmonella typhimurium and Vibrio

cholerae in Nkonkobe drinking water sources." Journal

of water and health 4(3) 289-296, 2006.

[19] Momtaz, H, et al., "Detection of Escherichia coli,

Salmonella species, and Vibrio cholerae in tap water

and bottled drinking water in Isfahan, Iran." BMC

public health 13(1) 556, 2013. [20] Ikegbunam, M. N., et al., "Antimicrobial activity of some

cleaning products against selected bacteria." International

Research Journal of Pharmaceutical and Applied Sciences 3,

133-135, 2013.

[21] Nillian, Elexson, et al., "Efficiency of Detergents against

Microbial Biofilm Growth in Kuching, Sarawak." Clinical

Microbiology: Open Access, 2016.

[22] Ramm, L, et al., "Pathogen transfer and high variability in

pathogen removal by detergent wipes." American journal of

infection control 43(7) 724-728, 2015.

[23] Dai, M., et al., "Preparation and Investigation of High-

Efficiency Antibacterial Liquid Dishwashing Detergent."

Transactions of Tianjin University 25(4) 322-329, 2019.

[24] Kusumaningrum, H. D., et al., "Effects of antibacterial

dishwashing liquid on foodborne pathogens and competitive

microorganisms in kitchen sponges." Journal of Food

Protection 65 (1)61-65, 2002.

[25] Sharma, M. et al., "Effective household disinfection methods

of kitchen sponges." Food Control 20(3) 310-313, 2009.

Health Scope 54


RESEARCH ARTICLE

Removing copper, chromium and nickel in industrial effluent using hydroxide precipitation versus sulphide precipitation

Siti Rohana Mohd Yatim1, Siti Nor Hazliyana Kasmuri1, Hanin Nazhifah Syahjidan1, Nurul

Shahirazni Mokhtar1, Nur Ain Zainuddin2*

1Centre of Environmental Health and Safety, Faculty of Health Sciences, Universiti Teknologi MARA Cawangan

Selangor Kampus Puncak Alam, 42300 Bandar Puncak Alam, Selangor Malaysia; 2Faculty of Chemical Engineering,

Universiti Teknologi MARA 40450 Shah Alam, Selangor Malaysia

Abstract:

Rapid growth of industries, such as the electroplating industry has led to large scale water

pollution, due to the release of wastewater containing high concentration of heavy metals such

as chromium (Cr), copper (Cu) and nickel (Ni) that does not comply with the Industrial Effluent

Regulation (IER) 2009. This experiment was conducted to determine the efficiency of removal

heavy metals by using hydroxide precipitation and sulphide precipitation. The optimum

condition, which is pH and coagulant dosage were determined by jar test. Jar test was conducted

on two different types of chemical precipitation, which is hydroxide precipitation using sodium

hydroxide (NaOH) and sulphide precipitation using sodium sulfide (Na2S). The result shows the

optimum condition of removal Cr, Cu and Ni in hydroxide precipitation was at pH 8.5 while for

sulphide precipitation was at pH 10. The optimum coagulant dosage (ferric chloride = FeCl3) for

hydroxide precipitation and sulphide precipitation was 0.8 mL. For hydroxide precipitation, Cr

= 98.65%, Cu = 2.81% and Ni = 99.90% were removed while for sulphide precipitation Cr =

91.29%, Cu = 99.99% and Ni = 99.97%. From the results, it showed that sulphide precipitation

was more effective to remove Cr, Cu and Ni compared to hydroxide precipitation as it can

remove heavy metals at broad range of pH.


Nur Ain Zainuddin

Email:[email protected]

Keywords: hydroxide precipitation, sulphide precipitation, turbidity removal

1. INTRODUCTION

Wastewater discharged by industrial activities is usually

contaminated with a variety of foreign substances that can be

toxic and harmful to environment and public health. One of

the common substances that is used in industrial activities is

heavy metals, especially in electroplating industry.

Electroplating is a branch of the metalworking industry in

which metal surfaces are treated by means of chemical and

electrochemical processes involving a sequence of pre-

treatment, coating and surface conversion baths to provide the

manufactured products with various properties and

characteristics such as protection from corrosion, increases

aesthetics values and surface hardness enhancement. [1]. The

process of plating involves various types of heavy metals.

This leads to the discharge of wastewater containing heavy

metals such as chromium (Cr), zinc (Zn), cyanide (Cn),

copper (Cu) and nickel (Ni) [1,2] that can cause pollution to

water bodies, resulting in water that is no longer consumable

and safe for drinking and daily use [3]. Due to the excessive

amount of heavy metals that have been discharge from the

electroplating activities, there are a few methods that have

been applied for heavy metal removal. The most common

method used is chemical precipitation as it is the simplest

method in removing heavy metals from aqueous solutions (4-

6]. However, there are a few limitations in treating mixed

metal effluents, as the process is affected by the optimum pH

and coagulant dosage.

The most common method used in chemical precipitation is

hydroxide precipitation and sulphide precipitation [7,8].

According to Lewis [9], hydroxide precipitation is widely

used in industry for removing heavy metals including Cr, Cu

and Ni. However, through a research by Nur et al., [8], it is

stated that sulphide precipitation gives a better performance in

removal of heavy metals as it can remove mixed heavy metals

due to the lower solubility of metal sulphide over a broad pH

range. In general, the treatment of wastewaters containing

heavy metals for Cr, Cu and Ni involve coagulation process

that uses chemical to enhance the process and as it is most

effective [10]. Chemical precipitation is the process of adding

other chemicals or coagulant to enhance the removal of heavy

55 Health Scope, 2020, Vol. 3(3) Yatim et al.

metals by significant increment in particles size of precipitates

[11]. Coagulant is mainly used for separation of small

molecules or suspended solid contents from solution by

destabilizing the particles, which inhibit the bond between

water particles [10,12]. Additionally, the process of adding

precipitant influence the formation of bond. The reaction of

hydroxide precipitation is shown in Equation 1 while reaction

for sulphide precipitation in Equation 2:

M2+ + 2(OH)- → M(OH)2

(1)

M2+ + S2- → MS

(2)

Basically, the rinsing process of plating activities contains a

high concentration of hexavalent chromium, Cr (IV). Cr (IV)

is more toxic than trivalent chromium, Cr (III) as it has been

defined as very toxic and harmful based on its characteristics

of mutagenicity, teratology and carcinogenicity [13,14]. Due

to the higher toxicity level, Cr (VI) need to undergo reduction

process in order to removed it and meet the standard. The

reduction of Cr (IV) to Cr (III) must be conducted in low pH

by using chemical reducing agents such as sulfur dioxide

(SO2) or sodium metabisulfite (Na2S2O5) [13-15]. Based on

research by Volesky [16] and Mo et. al [17] the hydroxide

precipitation is done under alkaline condition. According to

Nur et. al., [8], sulphide precipitation process should be

conduct in an alkaline condition to prevent H2S gas formation

and it stated that the performance of sulphide precipitation

treatment was better than hydroxide precipitation as it

removes a high amount of nickel and copper. Therefore, pH

selection plays important role in making improvement on

removal rate.

In this research, the pH range was selected based on the

Theoretical Metals Solubility Curve in order to identify at

which pH that the all heavy metals listed achieved maximum

removal rate. Figure 1(a) showed the metals hydroxide

solubility curve graph. From the graph, it indicated that Cr, Cu

and Ni can be removed between pH 6.5 to 10.5 as pH 8.5 –

9.5 for copper [8], for chromium is pH 7.5 – 8.5 [18] and for

nickel pH 9.0 - 10.0 [5] while for sulphide precipitation as

stated above, it can remove heavy metals at broad pH but to

avoid the production of H2S gas, sulphide precipitation was

conducted in alkaline condition. The pH between 8.0 until

10.0 is chosen. Daud et al., [12] indicated that there are several

factors affecting the effectiveness of the coagulation-

flocculation process in chemical precipitation. This includes

types of coagulants, chemical dosage, initial pH, mixing

speed, settling time and coagulant aid addition. Therefore, this

study aimed to determine the best way in removing the mixed

heavy metals in plating industrial effluent.

Figure 1: Metals hydroxide solubility graph

(Source: United States Environmental Protection Agency,

1983)


In this experiment, there were several steps that needed to

be carried out before running the experiment, which includes

calibration of apparatus for initial reading. All the data were

recorded from raw effluent preparation until the final result.

The materials and apparatus used are also stated below. The

initial parameters were taken and recorded to compare with

the final result.

2.1 Raw effluent preparation and materials

The raw effluent samples were obtained from plating

company in Klang, Selangor that contains nickel, chromium

and copper. Approximately 15 liters of sample were needed

for the experiment.

2.2 Heavy metals characterization

The effluent samples underwent several measurements,

including measurement for heavy metals concentration, pH

value, chemical oxygen demand (COD), turbidity value and

total suspended solid. These measurements were taken before

and after experiments.

2.2.1 Heavy metals measurement by using Inductively

Coupled Plasma- Optical Emission Spectrometers (ICP-

OES)

Inductively Coupled Plasma-Optical Emission

Spectrometers (ICP-OES) (Thermo Scientific iCAP 6000

series) was used. 10 ml of wastewater was utilized to achieve

the first objective, which is characterization of element

content in the wastewater.

Health Scope 56

2.2.2 Chemical oxygen demand measurement

To determine the chemical oxygen demand (COD) for the

sample, DRB200 reactor and HACH DR2800 was used.

Before the usage of DRB200 Reactor, it needed to be heated

from ambient temperature to 150oC. The sample vial was

filled with 2 mL of sample while for blank sample, 2 mL of

deionized was filled by using pipet. Both vials were put in the

DRB200 Reactor for preheating for 2 hours. After the heating

process was completed, the vials were left in the reactor that

had been switched off. Both vials were allowed to cool for

about 20 minutes until the temperature reaches 120oC or less.

Both vials were inverted few times in warm condition, and

were then placed in tube rack to cool until they reach room

temperature. The blank sample was cleaned and inserted into

cell folder. Button ZERO was pushed to ensure screen

displayed the value 0.0 mg/L COD. This step was repeated as

the sample was inserted into the cell holder. Both results for

initial and final readings were recorded.

2.2.3 Total suspended solid measurement

The suspended solid was measured by using Whatmann 1

filter paper. The weight of filter paper was measured using

electronic analytical balance (ADAM Nimbus Analytical

Balance) and recorded before usage (initial weight). After the

jar test processes were carried out, the sample was filtered and

dried in the oven (MMM Medcenter Venticell Drying Oven)

at 350C – 400C to evaporate the moisture. When the sample

was totally dried and the weight become constants, the oven

was switched off and let the sample cool down before weighed

(final weight). The weight of sample was obtained from the

subtraction of filter paper final weight minus initial weight

divided by total volume of the sample. It is shown in Equation

3 below (APHA 2540 D Method).

Total suspended solid (mg/L) = (A – B) (3)

V

Where: A = mass of filter paper + dried residue (mg)

B = mass of filter paper (mg)

V = total volume of the sample (L)

2.3 Jar test

To evaluate the optimum pH and coagulant dosage for the

precipitation process, a few sequences of jar test was carried

out using flocculator (Stuart SW6 Flocculator). The

adjustment of pH was varied to five different pH which is 6.5,

7.5, 8.5, 9.5 and 10.5 in each beaker respectively by adding

sulphuric acid (H2SO4) or sodium hydroxide NaOH. In this

experiment, ferric chloride (FeCl3) was used as the coagulant.

Firstly, the samples of 300 mL of raw effluents were poured

into five different unit of 500 mL beakers by using measuring

cylinder. To get optimum pH, each beaker was set to five

different pH levels as stated above and the coagulant dosage

(FeCl3) was set at a constant 0.5 mL. Meanwhile, for the

optimum coagulant dosage, the five beakers were set to a same

optimum pH but different coagulant dosages which 0.2 mL,

0.4 mL, 0.6 mL, 0.8 mL and 1.0 mL for each beaker. The

process involved rapid mixing, 150 rpm by using flocculator.

At the end of the experiments, the turbidity was measured and

recorded. Besides, the mass of the suspended solid produced

by each test also was measured by the gravimetric method and

the result was recorded.

2.4 Reduction of hexavalent chromium (IV) to trivalent

chromium (III)

For converting Cr (IV) to Cr (III), the sample of effluent

was filled in a beaker where the pH was determined and

sulfuric acid was added until a pH value of 2 was reached. The

justification for the pH setpoint of 2 has been selected as the

most cost-effective. At the same time, the oxidation reduction

potential (ORP) of the solution was measured, and sodium

metabisulfite [13-15] was added until an ORP value of

approximately 280 mV was achieved. Reactions that occurred

reduced hexavalent chromium to trivalent chromium (Cr+6 to

Cr+3).

2.5 Comparison hydroxide and sulphide precipitation

Before the completion of the jar test, the pH was set to

optimum level by adding H2SO4 or NaOH. Then the optimum

dosage of coagulant, FeCl3 was added into the sample. After

dosing the sample, the speed of stirrer was set to 150 rpm for

3 minutes. Following that, the precipitant, which is hydroxide

or sulphide, was added into the sample and while stirring

continues. This is then followed by the addition of coagulant

booster. Once mixed well, the mixer was turned off and the

beaker was left for 30 minutes to allow the flocs to settle.

Finally, the concentration of Cr, Cu and Ni were measured

using ICP-OES and the sludge produced was observed. The

steps were repeated using different precipitant, NaOH and

Na2S. The result from both precipitants was compared

2.6 The analysis of variance (ANOVA)

The results of the hydroxide and sulphide precipitation

were analyzed using two-factor with replication ANOVA in

Excel Microsoft Software.


Wastewater effluent from any industry activities included

electroplating industry need to be comply with the Industrial

Effluents Regulations 2009 (IER 2009) which included the

parameters such as pH value, COD, and heavy metal

concentration. Table 1 showed the standard parameter that

needed to be comply with IER 2009. The initial result of the

wastewater indicates that a few parameters does not comply

with the standard as the result obtain for pH was acidic (4.46),

COD = 128 mg/L and the heavy metals (Cr = 15.1509 mg/L,

Cu = 6.8128 mg/L, Ni = 63.8948 mg/L) were exceeded the

standard. The initial turbidity value was 89.7 NTU. This study

seeks to analyze the comparison between hydroxide

precipitation and sulphide precipitation.


Table 1: The parameter that needed to be comply based on

Fifth Schedule and Eight Schedule in IER 2009.

Parameters Standard B

pH value 5.5 – 9.0

COD, mg/L 200

Chromium trivalent (Cr3+), mg/L 1.0

Copper (Cu), mg/L 1.0

Nickel (Ni), mg/L 1.0

3.1 Effect of optimum pH

The efficiency of heavy metals removal was significantly

affected by pH alteration. From the metal solubility graph, it

is indicated that Cr, Cu and Ni can be optimally removed at a

pH range of 6.5 to 10.5, with a pH range of 8.5 – 9.5 for copper

[8], pH 7.5 – 8.5 for chromium [18] and pH 9.0 - 10.0 for

nickel. Meanwhile, sulphide precipitation can remove heavy

metals at broad pH but to avoid the production of H2S gas,

sulphide precipitation needed to be conducted in alkaline

condition [8]. Therefore, a pH range between 8.0 until 10.0

was selected. The pH of wastewater varied between pH 6.5

until 10.5 for hydroxide precipitation while for sulphide

precipitation, it was between pH 8.0 until 10.0. Both were

conducted at constant FeCL3 dosage of 0.5 mL (coagulant)

and PAC dosage of 0.5 mL (coagulant booster).

Figure 2 shows the result of pH value vs percentage of

turbidity removal, %. For the hydroxide precipitation, the bar

graph showed a decrease from pH 6.5, 7.5 and 8.5. There was

a slight increase at pH 9.5 but it decreases again at pH 10.5.

The lowest turbidity removal was 49.05% (45.7 NTU) which

the pH 6.5 while the highest turbidity removal was 75.14%

(22.3 NTU) which at pH 8.5. As stated by Volesky [16] and

Mo et. al [17], the hydroxide precipitation should be carried

out in alkaline condition. The optimum pH for hydroxide

precipitation was pH 8.5 as it showed the highest percentage

of turbidity removal, out of all five different pH. Therefore,

this supports the evidence presented by Volesky [16] and Mo

et. al [17].

The line graph showed a slightly increase from pH 8 to pH 8.5

and decline from pH 9 and 9.5 for sulphide precipitation.

Observation also showed a slight decrease for pH 10. The

lowest turbidity removal was 52.40% (42.7 NTU), in which

the pH was 8.5 while the highest was 88.41% (10.4 NTU) at

pH 10. Based on the research by Nur et. al., [8], sulphide

precipitation should be done under alkaline condition to avoid

the production of hydrogen sulphide gas. From the

experiment, it showed that the optimum pH for sulphide

precipitation was at pH 10. It indicated that efficiency of

sulphide precipitation increases as the pH increased. This

showed, that sulphide precipitation can remove heavy metals

at alkaline condition. A research by Shahad [19] also stated

that the condition of alkaline sample which is the higher pH

can reduce the turbidity value. This is due the opposite charge

of an ion of the colloid that affect the sedimentation process.

Figure 2: The result of pH value vs percentage of turbidity

removal, % for hydroxide and sulphide precipitation.

3.2 Effect of coagulant (FeCl3) and coagulant booster

(PAC) dosage

The efficiency of turbidity removal from wastewater is

significantly affected by the dosage of coagulant added to the

wastewater. By varying the coagulant dose between 0.2 mL

until 1.0 mL with interval of 0.2mL and at constant optimum

pH of 8.5 for hydroxide precipitation, pH 10.0 for sulphide

precipitation and constant flocculant dosage of 0.5 mL, the

graph of optimum coagulant dosage, mL vs percentage of

turbidity removal, % was plotted. Observations revelead the

optimum coagulant dose for maximum removal of turbidity

was obtained for hydroxide and sulphide precipitation.

Figure 3 shows the result of the optimum coagulant dosage,

mL vs percentage of turbidity removal, %. In this experiment,

ferric chloride (FeCl3) was used as the coagulant. For the

hydroxide precipitation, the bar graph showed an increase of

turbidity removal from 0.2 mL, 0.4 mL, 0.6 mL and 0.8 mL.

However, there was slightly decreased at 1.0 mL. The result

for the 0.2 mL of FeCl3 showed the lowest turbidity removal

which 23.86% (65.3 NTU) while the highest turbidity removal

was 75.92% (21.6 NTU), which the FeCl3 dosage was 0.8 mL.

From the result it showed that, as the coagulant increased, the

percentage of turbidity removal also increased. However,

when the amount of coagulant increased to 1.0 mL, the

percentage of turbidity removal was decreased. The alteration

of pH by NaOH can cause the metals inside the water to

ionized into metals ions which were considered stable in

water. These metals ions considered hard to eliminate by a

little amount of FeCl3 dose. Therefore, the solubility of the

metals inside the water increases at this point. From the trend,

the result can indicate that the increment of coagulant dosage

at certain amount can enhance the removal of turbidity but as

the amount exceed, the efficiency of coagulant was decreased.

The graph showed an unstable result for sulphide

precipitation. The FeCl3 dosage of 0.4 mL showed the lowest

turbidity removal of 99.29% (0.63 NTU) and the highest

turbidity removal at 99.73% (0.24 NTU), in which the FeCl3

dosage was set at 0.8 mL. Therefore, 0.8 mL of FeCl3 dosage

was chosen as the optimum coagulant dosage for hydroxide

and sulphide precipitation. From the trend, the addition of

dosage of FeCl3 in sulphide precipitation showed varied result.

This may happen due to the ionization of metals. Na2S was

Health Scope 58

used as precipitant. Therefore, there was no addition of

hydroxide ions that can affected the turbidity removal and

easily to remove. Overall sulphide precipitation showed better

results compared to hydroxide precipitation by adding varied

amount of FeCl3 dosage.

Figure 3: The result of coagulant, FeCl3 dosage, mL vs

percentage of turbidity removal, % for hydroxide and

sulphide precipitation.

Figure 4 shows the result of coagulant booster (PAC) dosage,

mL vs percentage of turbidity removal, %. In this experiment,

the pH and FeCl3 dosage was set to the optimum value. The

optimum pH for hydroxide and sulphide precipitation was 8.5

and 10.0 respectively while the optimum FeCl3 dosage was

0.8 mL for both. The results are presented in the bar chart

below. The trend shows an increased percentage of turbidity

removal, % from 0.2 mL, 0.4 mL and 0.6 mL. However, the

results decreased for 0.8 mL and 1.0 mL. The lowest turbidity

removal was 70.12% (26.8 NTU) for 0.2 mL of PAC and the

highest turbidity removal was 89.52% (9.4 NTU) with 0.6 mL

of PAC. For the bar graph of sulphide precipitation, the result

was increased for 0.2 and 0.4 mL of PAC and decreased when

more PAC was used. However, the result shows the lowest

turbidity removal was 97.97% (1.82 NTU) for 0.8 mL of PAC

dosage and the highest was 99.45% (0.49 NTU) for 1.0 mL of

PAC dosage. Thus, for hydroxide and sulphide precipitation

the optimum coagulant booster (PAC) dosage was 0.6 mL and

1.0 mL respectively.

Addition of PAC may affect the change of pH, which may also

affect turbidity removal. PAC was added to improve

coagulation process and reduces the amount of FeCl3 used by

increasing the amount of positive charges that help to create

larger, heavier and easily settled particles [20]. From Figure

4, we can observe an improvement of performance of turbidity

removal for hydroxide precipitation when PAC was added as

it increased the basicity condition. However, in this

experiment, sulphide precipitation are observed to be better at

removing heavy metals by reduction of turbidity.

Figure 4: The result of coagulant booster, PAC vs

percentage of turbidity removal, % for hydroxide and


3.3 Effect of precipitant

The efficiency of turbidity removal can be affected by

precipitant dosage. To conduct the hydroxide and sulphide

precipitation, the pH was set to the optimum pH, which for

hydroxide precipitation, was pH 8.5. Meanwhile, for sulphide

precipitation the pH was 10.0. The coagulant dosage also was

adjusted to the optimum dosage for each treatment. The

precipitant dosage varied from 0.2 mL to 1.0 mL with an

interval of 0.2. The graph of effect precipitant dosage to the

turbidity removal was plotted. The precipitant used for

hydroxide precipitant is NaOH, while sulphide precipitation

used Na2S.Thus, the optimum precipitant dosage for each

treatment was obtained.

Figure 5 showed the result of precipitant dosage, mL vs

percentage of turbidity removal, % for hydroxide and sulphide

precipitation. The result observed from the bar chart for

hydroxide precipitation showed an increased from 0.2 mL of

NaOH dosage until 0.8 mL. at 1.0 mL of NaOH the result

showed a slightly decreased. The lowest turbidity removal

was at 0.2 mL of NaOH, which is 79.14% (18.71 NTU) while

the highest was at 0.8 mL which 98.42% (1.42 NTU). Thus,

0.8 mL of NaOH was the optimum precipitant dosage as it

showed the highest percentage of turbidity removal. For

sulphide precipitation, the bar chart showed a significant

increase from 0.2 mL of Na2S until 1.0 mL. The lowest

percentage turbidity removal was 94.24% (5.17 NTU) while

the highest was at 1.0 mL which 99.28% (0.65 NTU). Thus,

1.0mL of Na2S was the optimum precipitant dosage for


Overall, the result indicated from the hydroxide precipitation

showed that the increment of precipitant dosage can enhance

the turbidity removal but at certain point when the precipitant

was exceed, there was no reaction occur between the ions as

the higher hydroxide group in the water made the ions become

stable and not easily to remove [12]. This is due to the reaction

between the ions in the water that was affected by factors such

pH and coagulant dosage. As stated by Nur et al., [8] sulphide

precipitation can remove heavy metals at broad pH. The result

of the experiment aligned with the statement as the sulphide

precipitation showed a better result for turbidity removal

compared to hydroxide precipitation.


Figure 5: The result of precipitant dosage, mL vs percentage

of turbidity removal, % for hydroxide and sulphide

precipitation.

3.4 The removal concentration of chromium (Cr), copper

(Cu) and nickel (Ni) of the effluent wastewater

Figure 5 shows the result for the percentage removal of

heavy metals concentration of chromium (Cr), copper (Cu)

and nickel (Ni) for hydroxide precipitation and sulphide

precipitation. The line graph indicated the hydroxide

precipitation while the bar graph indicated the sulphide

precipitation. For the hydroxide precipitation, the result of

removal was Cr = 98.65%, Cu = 2.81% and Ni = 99.90%

while for sulphide precipitation Cr = 91.29%, Cu = 99.99%

and Ni = 99.97%. From the result, it showed that Cr and Ni

can be removed at the same pH (8.5) value as it was in the

same range of solubility. However, at pH 8.5, it only removed

Cu at minimum value while for the sulphide precipitation, the

result showed more than 90% of heavy metal concentration

were removed at same pH, which is pH 10.0. This showed that

sulphide precipitation can remove heavy metals at broad range

of pH.

Figure 6: The result for percentage removal of Cr, Cu

and Ni concentration, %.

Additionally, the COD result also decreased from 128 mg/L

to 59 mg/L for hydroxide precipitation and 12.9 mg/L for

sulphide precipitation. Apart from that, the sludge from

sulphide precipitation (0.41 g) was more than hydroxide

precipitation (0.02 g). Thus, from the results as shown in the

Table 2, it can state that sulphide precipitation was better than

hydroxide precipitation for removing heavy metals in the

electroplating wastewater.

Table 2: The result of chemical oxygen demand (COD) and

total suspended solid after hydroxide and sulphide

precipitation treatment.

Parameter Hydroxide

precipitation

Sulphide

precipitation

Chemical

Oxygen Demand

(COD), mg/L

59 12.9

Total suspended

solid, g

0.41 0.02

3.5 Two-factor ANOVA analysis of the removal of Cr, Cu

and Ni via hydroxide precipitation and sulphide

precipitation

The comparison of performance on Cr, Cu and Ni

removal for hydroxide precipitation and sulphide precipitation

was analyzed using a two-factor ANOVA analysis. The two-

factor with replication ANOVA analysis was conducted based

on the experiment of chemical precipitation by using Excel

Software. The ANOVA analysis for each chemical

precipitation was presented in Table 3.

Table 3: The result of two-factor ANOVA analysis.

Source of

variation

dF F p-value F crit

Sample 2.80845 1 2.80845 1.530212

Columns 8.529636 2 4.264818 2.323729

Interaction 16.82971 2 8.414857 4.58492

Table 3 indicates the result of two-factor ANOVA analysis for

the experiment. The sample row showed the F statistics and

p-value for differences between hydroxide and sulphide

precipitation. F critical value (1.530212) is bigger than F value

of 1. This indicates the acceptance of null hypothesis, which

means the observation for the hydroxide and sulphide

precipitation are not the same in favor of the alternative. Since

the p-value is greater than 0.05, it showed another evidence

that the null hypothesis is accepted. The column row showed

that F critical value, 2.323729 is bigger that F value, 2 which

means of the observation by the columns called hydroxide and

sulphide precipitation are not same in favor of alternative is

that there is no difference between at two of them and the p-

value is bigger than 0.05 which means the null hypothesis is

accepted. The interaction row showed the F critical value

(4.58492) is bigger than F value (2), which also means the null

hypothesis is accepted. The p-value, which is also greater than

0.05, also indicates that the null hypothesis is accepted. The

result from the interaction showed the differences of heavy

metals (Cr, Cu, Ni) is independent of the chemical

precipitation. However, it also indicates the reverse, in which

the effect of hydroxide and sulphide precipitation on the

turbidity removal is independent of the types of heavy metals

(Cr, Cu, Ni).

Health Scope 60

4. CONCLUSION

In this experiment, the optimum pH, coagulant (FeCl3) dosage

and coagulant booster (PAC) dosage were determined through

several sequences of jar test. The performance of hydroxide

precipitation and sulphide precipitation was compared to

evaluate which is the more effective method to remove Cr, Cu

and Ni in the effluent of industrial wastewater. The result

shows that hydroxide precipitation, Cr = 98.65%, Cu = 2.81%

and Ni = 99.90% were removed while for sulphide

precipitation Cr = 91.29%, Cu = 99.99% and Ni = 99.97%.

Meanwhile, sulphide precipitation was more effective in

removing Cr, Cu and Ni compared to hydroxide precipitation,

as it can removed heavy metals at broad range of pH. As a

conclusion, the performance of sulphide precipitation

treatment was better than hydroxide precipitation. In future,

the experiment can be improved for getting the better result to

analyze the effectiveness of the performance. Not only that, it

is also recommended to determine the heavy metals

concentration based on the percentage of turbidity removal as

it can show the exact interaction of the heavy metals reduction

from the ions reaction with the pH alteration and coagulant

dosage.

ACKNOWLEDGEMENTS

The authors would like to thanks to the Faculty of Chemical

engineering and Faculty of Health Sciences for the facilities

provided.

REFERENCES

[1] Hackbarth, F.V., et al., “Removal of hexavalent chromium

from electroplating wastewaters using marine macroalga

Pelvetia canaliculata as natural electron donor.” Chemical

Engineering Journal, 290, 477–489, 2016.

[2] Mansur et al., “Selective extraction of zinc (II) over iron (II)

from spent hydrochloric acid pickling effluents by liquid-

liquid extraction.” Journal of Hazardous Materials, 150(3),

669–678, 2008.

[3] Bhatnagar, Amit, & Sillanpää, Mika, “Utilization of agro-

industrial and municipal waste materials as potential

adsorbents for water treatment-A review.” Chemical

Engineering Journal, 157(2–3), 277–296, 2010.

[4] Grimshaw, P., et al., “Cyclic electrowinning/precipitation

(CEP) system for the removal of heavy metal mixtures from

aqueous solutions.” Chemical Engineering Journal, 175(1),

103–109, 2011.

[5] Coman, V., Robotin, B., & Ilea, P., “Nickel

recovery/removal from industrial wastes: A review.”

Resources, Conservation and Recycling, 73, 229–238, 2013.

[6] Shuangchen, M., et al., “Research on desulfurization

wastewater evaporation: Present and future perspectives.”

Renewable and Sustainable Energy Reviews, 58, 1143–1151,

2016.

[7] Fu, Fenglian, & Wang, Q., “Removal of heavy metal ions

from wastewaters: A review.” Journal of Environmental

Management, 92(3), 407–418, 2011.

[8] Nur Ain, Z., et al., “Removal of Nickel, Zinc and Copper

from Plating Process Industrial Raw Effluent Via Hydroxide

Precipitation Versus Sulphide Precipitation” OP Conf.

Series: Materials Science and Engineering, 1–6, 2019.

[9] Lewis, Alison Emslie, Review of metal sulphide

precipitation. Hydrometallurgy, 104(2), 222–234, 2010.

[10] Carolin, C., et al., “Efficient techniques for the removal of

toxic heavy metals from aquatic environment: A review.”

Journal of Environmental Chemical Engineering, 5(3),

2782–2799, 2017.

[11] Chen, Q., et al., “Precipitation of heavy metals from

wastewater using simulated flue gas: Sequent additions of

fly ash, lime and carbon dioxide.” Water Research, 43(10),

2605–2614, 2009.

[12] Daud, N. M., et al., “Response surface methodological

analysis for the optimization of acid-catalyzed

transesterification biodiesel wastewater pre-treatment using

coagulation–flocculation process.” Process Safety and

Environmental Protection, 113(2014), 184–192, 2018.

[13] Gheju, M., & Balcu, I., “Removal of chromium from Cr (

VI ) polluted wastewaters by reduction with scrap iron and

subsequent precipitation of resulted cations.” Journal of

Hazardous Materials, 196, 131–138, 2011.

[14] Espada, J. J., Gallardo, M., & Molina, R., “Life cycle

assessment and techno-economic evaluation of alternatives

for the treatment of wastewater in a chrome-plating

industry.” Journal of Cleaner Production.172, 2351–2362,

2018.

[15] Liu, C., et al., “A new technology for the treatment of

chromium electroplating wastewater based on biosorption.”

Journal of Water Process Engineering, 11, 143–151, 2016.

[16] Volesky, B., “Detoxification of metal-bearing effluents:

Biosorption for the next century.” Hydrometallurgy, 59(2–

3), 203–216, 2001.

[17] Mo, J., et al., “A review on agro-industrial waste (AIW)

derived adsorbents for water and wastewater treatment.”

Journal of Environmental Management, 227(July), 395–

405, 2018.

[18] Brbooti, M. M., et al., “Removal of Heavy Metals Using

Chemicals Precipitation.” Eng & Tech. Journal, 29, 595–

612.

[19] Shahad Salim Mohammed. “Effect of pH on the Turbidity

Removal of Wastewater.” Open Access Library Journal,

02(12), 1–9, 2011, 2005.

[20] Al-Asheh, Sameer, & Aidan, Ahmed., “Operating

Conditions of Coagulation-Flocculation Process for High

Turbidity Ceramic Wastewater.” Journal Water

Environmental Nanotechnology, 2(22), 80–87, 2017.

[21] Environmental Quality Act 1974 (Act 127), Environmental

Quality (Industrial Effluent) Regulations 2009. Petaling

Jaya: International law Book Services, 2017.

[22] United States Environmental Protection Agency, Methods

for Chemical Analysis of Water and Wastes (Vol. 2), 1983.

Health Scope 61

RESEARCH ARTICLE

School functional assessment: performance in functional communication among special education at school

Nursyahirah Anuar, Rosilah Wahab*

Centre of Occupational Therapy, Faculty of Health Sciences, Universiti Teknologi MARA Cawangan Selangor Kampus


Abstract:

Functional communication is a critical skill used in school and students with special needs often have

difficulties performing compared to their peers. The School Function Assessment (SFA) is able to

contextually measure the performance of functional skill in special needs students and compare them

between their peers within the Special Education Integration Program (SEIP) classes using the norm-

referenced criterion scores. 63 special needs students of eight different disabilities between ages seven

to nine years old were assessed by their teachers in this research using the SFA (Part III) to investigate

the relationship between age, gender and diagnosis with overall activity performance and performance

in functional communication. The findings discovered that special needs students perform generally

poorly across all components of SFA (Part III) and specifically in functional communication. There are

no significant differences between gender (p=1.000) and ages (p = 0.095) with functional

communication. There is also no significant differences were found between activity performance in

functional communication and diagnosis (p=0.075). This research will increase understanding about the

functional skill skills of special needs students of different age, gender and diagnosis and enable

educators and therapists alike to plan effective interventions for the students.


Rosilah Wahab


Keywords: School functional assessment, special education, functional communication

1. INTRODUCTION

In this new era of science and technology, education is important for all people including those with disabilities. Education plays a huge role in a person’s life such as individual development and nation building [1]. Education involves academic, non-academic and educational activities. School is one of the essential places where the educational process can be accomplished. In schools, students’ abilities are measured through their school performance which an issue that deeply concerns students, parents, teachers and authorities in Malaysia and other countries. School performance is described as the results shown by the students of what they have learned and usually measured through school grades [2]. However, children with disability are acknowledged to encounter hurdles in school compared to peers despite their phenotype disability [3]. On the other hand, school performance also can be measured through activity performance at school. Activity performance can be described as the student’s ability to initiate and achieve specific functional school-related activities that are classified by the general task area [4].

School functions will give a significant impact on students’ successes and difficulties [5]. School activity limitations and less participation by students with special needs than by their typically developing peers is commonly reported [6]. Similarly, students with disabilities often have difficulty meeting performance expectations in school because physical, cognitive, or social impairments affect their ability to participate actively in classroom learning activities,

express their knowledge clearly, and to interact with their peers [7].

From a human right perspective, education is an intrinsic right which possessed by every individual regardless of their status and background [1]. Children with special needs have the right for quality education which involves not only academic subjects but also skills development that can help them survive in the challenging ‘real world’ in the future [8]. In fact, special education concern with any necessary services or approach for children with various forms of disabilities and also concern with the learning of gifted children and other marginalized children [1]. Besides, the main aim of special education integrated program is to ensure that students with special needs can learn in the least restrictive environment and also be able to develop their social and communication skills effectively [8].

The effective communication increases the effectiveness of students’ performance [9]. This stressed the importance of communication in the educational process. Communication education helps to develop the ability to communicate with an array of different types of people in different situations including peers, parents and teachers [10]. Next, communication as a key to successful collaboration in all educational environments. In addition, according to [11], a robust and comprehensive communication system provides students with the ability to convey a rich range of communicative functions, promoting academic and social inclusion across their school day. Moreover, a transactional process of ongoing verbal and nonverbal behaviors from a


minimum of two people is known as communication [12], where one person responding accordingly after another person initiating the communication [13]. Communication enhances the quality of classroom instruction and is key to successful collaboration in educational environments [14]. The main focus of this study is to determine the performance of functional commination among special education at school.


2.1 Participant sampling

A mixed-method sampling was conducted where

random sampling was used for the schools within Petaling

Perdana, Selangor and a convenience sampling was done

within those schools to choose appropriate and suitable

students for the study. The criteria to be included in the

study were primary school students with special needs aged

between 7 to 9 years old and attend school at least 3 times a

week. Teachers (who will be assessing the students) must be

able to understand English and have at least three years’

experience in teaching special needs students. Participants

who were excluded were special needs students with

multiple disabilities other than learning disabilities (deaf and

blind).

2.2 Instruments

2.2.1 Demographic questionnaire

Information on the name, age, gender and specific disability

of the sample being assessed is recorded. The assessor may

provide extra information such as secondary health

conditions, personal observations or notes from visiting

community Occupational Therapists if deemed appropriate

and reasonable. This questionnaire is in simple English to

ease the process of answering.

2.2.2 School Function Assessment (SFA) (Part III)

The School Function Assessment (SFA) is a criterion-based

assessment that can be administered through observation and

without active participation of the student. It is given in

English and fully explained to the assessors prior to data

collection to ensure the measure is understood. The

measurement was chosen due to its ability to measure the

school function of a special needs’ child in their school and

have the results be comparable to a set norm provided [7].

SFA comprises three parts which are Part I: Participation,

Part II: Task Support and Part III: Activity Performance.

Only Part III of the assessment is used in this research as it

focuses on rating the activity performance of the special

needs’ child with or without assistance.

There are two types of activity to observe and measure

which are then further broken into specific tasks such as

Travel, Maintaining and Changing Positions, Recreational

Movement, Manipulation with Movement, Using Materials,

Setup and Cleanup, Hygiene, Eating and Drinking, Clothing

Management, Up/Down Stairs, Written Work, Computer and

Equipment Use, Functional Communication, Memory and

Understanding, Following Social Conventions, Compliance

with Adult Directives and School Rules, Task

Behavior/Completion, Positive Interaction, Behavior

Regulation, Personal Care Awareness and Safety.

The tasks Up/Down Stairs and Computer and Equipment

Use were not included in the overall assessment for this

research due to limitations and physical structures of the

school buildings in which this research was carried out.

Furthermore, this research will focus more specifically on

the functional communication aspect of the assessment.

2.3 Assessment of participants

Special education teachers are to assess students

within their class that fits the criteria of the study. The

teachers only need to report on normal behavior of their

students and rate them accordingly on the School Function

Assessment Part III. As this does not require the students to

follow any extra directions other than their usual in-class

directives, it does not interrupt the school schedule in any

way. Each teacher is required to sign a consent form prior to

beginning the assessment to signify willingness to

participate.

2.6 Data analysis

The questionnaire will be analysed using Statistical Package for the Social Sciences (SPSS) version 25 in two stages. The first stage is using descriptive analysis for data such as demographic and scores from the questionnaire. The descriptive analysis of the demographic information and the total scores in the barriers questionnaires will be analysed by using mean (SD), or median (IQR: 25% - 75%) depends on the distribution of the data collection.

As for the second stage, an inferential analysis that aimed to identify relationship and to determine whether to accept or reject the hypothesis of the study. The inferential analysis which is types of hypothesis testing depends on the distribution data and types of data obtained. If the data is normally distributed, parametric testing will be conducted such as the t-test or one-way ANOVA will be used to accept or reject the hypothesis. To determine associations between variables, a Pearson correlation coefficient if data is normally distributed or Spearman rho if data is abnormally distributed will be conducted.


3.1 Demographic variables

The table 1 below shows the demographic data of the students such as gender, age and diagnosis. There were 39 male and 24 female students. Next, the age ranged from 7 to 9 years old. The mean age is 8.02 years old. Moreover, there are 27 students with autism spectrum disorder (ASD), 3 students with attention deficit/hyperactivity disorder (ADHD), 8 students with Down syndrome (DS), 12 students with slow learner (SL), 3 students with cerebral palsy (CP), 5 students with global developmental delay (GDD), 3 students with learning disabilities (LD) and 2 students with mental retardation (MR).

Health Scope 63

Table 1: The demographic data of the students.

Gender

Frequency Percentage (%)

Male 39 61.9

Female 24

38.1

Age Frequency Percentage (%)

7 21 33.3

8 20

31.7

9 22

34.9

Diagnosis Frequency Percentage (%)

ASD

27 42.9

ADHD 3

4.8

DS 8

12.7

SL 12 19.0

CP 3 4.8

GDD 5 7.9

LD 3 4.8

MR 2 3.2

3.2 The level of activity performance among special education students

SFA divided activity performance into physical tasks and cognitive/behavioural tasks. The mean criterion scores of the students will be compared to the criterion cut-off scores for each specific functional task. The physical tasks including travel; maintaining and changing position; recreational movement; manipulation with movement; using materials; set up and clean up; eating and drinking; hygiene; clothing management; and written work. Moreover, the optional tasks such as up/downstairs; and computer and equipment use are will not be scored since the students not routinely required to perform the task. Besides, the cognitive/behavioural tasks including functional communication; memory and understanding; following social conventions; compliance with adult directives and school rules; task behaviour/completion; positive interaction; behaviour regulation; personal care awareness; and safety.

Figure 1 shows the level of physical tasks performance of the students. Generally, the students achieved scores significantly below that of children their age in all tasks.

This study shows the special education students achieved scores significantly below that of children their age in all physical tasks’ performance. The two highest mean criterion scores are maintained and changing position; and travel which more than 75.00. However, both tasks still below each criterion cut-off score which means the students’ performance are below of their typically developing peers. While, the four lowest mean criterion scores are written work, hygiene, using materials, and recreational movement which below 55.00. Other than that, the lower the gap different between mean criterion score and the criterion cut-off score of a task shows better ability of students in

performing that task and vice versa. The three lowest gaps different between mean criterion score and the criterion cut-off score in physical tasks are travel; maintain and changing position; and manipulation with movement. Most students have better ability to perform those tasks. While, the three biggest gaps different between mean criterion score and the criterion cut-off score in physical tasks are eating and drinking; hygiene; and clothing management. Most students have lower ability to perform those tasks.

Figure 1: The level of physical tasks performance.

Figure 2 shows the level of cognitive/behavioural tasks performance of the students. Generally, the students also achieved scores significantly below that of children their age in all tasks.

Figure 2: The level of cognitive/behavioural tasks performance.

This study shows the special education students achieved scores significantly below that of children their age in all cognitive/behavioural tasks performance. The three highest mean criterion scores are memory and understanding; compliance with adult directives and school rules; and behaviour regulation which more than 50.00. However, both tasks still below each criterion cut-off score which means the students’ performance are below of their typically developing peers. While, the five lowest mean criterion scores are functional communication, following social conventions, task behaviour/completion, personal care awareness, and safety which below 47.00. Other than that, the lower the gap different between mean criterion score and the criterion cut-off score of a task shows better ability of students in performing that task and vice versa. The three lowest gaps different between mean criterion score and the criterion cut-off score in cognitive/behavioural tasks are memory and understanding; compliance with adult directives


and school rules; and behaviour regulation. Most students have better ability to perform those tasks. While, the three biggest gaps different between mean criterion score and the criterion cut-off score in cognitive/behavioural tasks are positive interaction; personal care awareness; and safety. Most students have lower ability to perform those tasks.

Students with disabilities often have difficulty meeting performance expectations in school because physical, cognitive or social impairments affect their ability to participate actively in classroom learning activities, express their knowledge clearly, and to interact with their peers [7]. School activity limitations and less participation by students with special needs than by their typically developing peers are commonly reported [6]. However, there is evidence by study which pointed out that children with struggle with aspects of adaptive functioning in the school setting [3]. As a result, students with DS were reported to demonstrate the greatest challenges with the following cognitive-behavioural activities such as following social convention, functional communication, compliance with adult directives and school rules, personal care awareness, task behaviour-completion, positive interaction, and safety [3]. Other than that, there is no collaboration network with the occupational therapist in hospital and lack of law enforcement in Malaysia resulted to indirect and formal occupational therapy in special education classes [15]. As one of occupational therapy domains is education and implements to develop functional tasks among special needs students.

3.3 The level of functional communication performance among special education students

Functional communication is one of a task in cognitive/behavioural tasks. The mean criterion scores of the students will be compared to the criterion cut-off scores for this specific functional task.

Figure 3 shows the level of functional communication performance of the students. Generally, the students achieved scores significantly below that of children their age in this task.

. The functional communication domain measures a student’s ability to communicate all types of information to peers and adults. Items include communicates yes/no, acceptance/refusal, or choice between two or more items; “hungry” or “thirsty”; “sick”, “hurt” or “help”; need for help with a functional (non-academic) task; first and last name; where something is located in classroom or school; short messages to another person; inquiries/requests for information; short messages from one person to another; basic safety information; describe an object well enough to enable correct identification; where he/she would go or what he/she would do if lost; and complex (3 steps) directions to others. However, this study shows the special education students achieved scores significantly below that of children their age in functional communication performance. The mean criterion scores for this task is 38.97. However, these tasks still below its criterion cut-off score of 91.00 which means the students’ performance are below of their typically developing peers. Other than that, the lower the gap different between mean criterion score and the criterion cut-off score of a task shows better ability of students in performing that

task and vice versa. The gap different for this task is 52.03 which more than half of the criterion cut-off score. This shows most students has lower ability in performing this task. Next, this study also shows 92.1% number of students failed in this task. There are 5 students passed with 58 students failed in functional communication task.

This finding shown there is a significant lower of functional communication performance among special education students [16]. Communication enhances the quality of classroom instruction and is key to successful collaboration in educational environments [14]. This statement is supported which improving communication skills allow children to take an active part in social settings and provide them with the tools necessary to engage in meaningful conversations [17]. Communication is often impaired in children with special needs [18]. It is supported that the child’s ability to interact and engage with their environment related to these communication difficulties [19]. Moreover, students having problems in verbal communication including in classroom instruction [20]. In the study stressed that symptoms of inattention and hyperactivity typically co-occur with poor communication skills and low level of literacy in children with ADHD [21]. Other than that, reinforces the importance of communication to enhance the organizational life [22]. In addition, classroom teachers can maintain communication skills on these children in natural and unstructured contexts as there is strong correlation exists between communication skills of children and the frequency of communication opportunities [23]. The teachers have to provide frequent communication opportunities for them in the classrooms in order to maintain their communication skills [13].

Figure 3: The level of functional communication performance.

3.4 The Association between functional communication performance with age

The association between functional communication performance with age is analysed using Fisher’s Exact Test since three of the cell’s frequency are small. Those cells with an expected count of <5 is 50%. Moreover, the Fisher’s exact test used to determine if there are non-random associations between two categorical variables. The variables are criterion cut-off scores of the functional communication performance will be associated with the age of the students.

Table 2 shows there was no significant association between the student’s functional communication performance and the

Health Scope 65

age (p=0.095). The proportion of pass student was highest among those with aged 9 years old, while most of the children assessed as being of fail in functional communication performance were from student aged 7 years old.

The age range of the students in this study are 7 (n=21), 8 (n=20), and 9 (n=22) years old. The mean age is 8.02 years old. This study shows there was no significant association between the student’s functional communication performance and the age (p=0.095). Next, the proportion of pass student was highest among those with aged 9 years old, while most of the children assessed as being of fail in functional communication performance were from student aged 7 years old. Moreover, this study shows only one student aged 8 years old (5.0%) and four students aged 9 years old (18.2%) passed the functional communication task with criterion score above 91. Furthermore, all students aged 7 years old (100.0%) and nineteen students aged 8 years old (95.0%) together with eighteen students aged 9 years old (81.8%) failed in this task. There was no significant association shown since the differences between functional communication performance and age are small.

Table 2: Assessed functional communication performance by age.

Age

Functional communication performance

Pass Fail Total

7 0 (0.0%) 21 (100.0%) 21 (100.0%)

8 1 (5.0%) 19 (95.0%) 20 (100.0%)

9 4 (18.2%) 18 (92.1%) 22 (100.0%)

Total 5 (7.9%) 58 (92.1%) 63 (100.0%)

x2 = 4.313; p > 0.05

The hypothesis stated there is an association between functional communication performance with age is rejected. Children learn best when communication is tailored to their specific developmental age, needs and interests [24]. It is not only developmental trends in children’s understanding of conversation as a means of communicating and legitimising knowledge but also as the progressive age groups able to understand, the level of conceptual difficulty is vital to each group which requiring more complex contents, contexts and forms [25]. During age of 7 through 10 years, children gradually develop into more independent and capable of exploring the world around them [24]. They are able using more sophisticated language, learn huge amount of new information and acquire new skills including literacy. Furthermore, children have different in abilities and needs at different ages are things to be considered when communicating with them. Thus, there are several principles that can be applied when communicating with children which are communication should be age-appropriate and child-friendly; address the child holistically; and be positive and strengths-based [24].

3.5 The difference between functional communication

performance with gender

The difference between functional communication performance with gender is analysed using Fisher’s Exact Test since two of the cell’s frequency are small. Those cells with an expected count of <5 is 50%. Moreover, the Fisher’s exact test used to determine if there are non-random associations between two categorical variables. The variables are criterion cut-off scores of the functional communication performance will be associated with the gender of the students.

Table 3 shows there was no significant difference between the student’s functional communication performance and the gender (p=1.000). The proportion of pass student was highest among female students (8.3%), while most of the children assessed as being of fail in functional communication performance were male students (92.3%).

Table 3: Assessed functional communication performance by gender.

Gender

Functional communication performance

Pass Fail Total

Male 3 (7.7%) 36 (92.3%) 39 (100.0%)

Female 2 (8.3%) 22 (91.7%) 24 (100.0%)

Total 5 (7.9%) 58 (92.1%) 63 (100.0%)

x2 = 0.008a; p > 0.05b

a Pearson Chi Square Test

b Fisher’s Exact Test

The students in this study were male (n=39) and female (n=24). This study shows there was no significant difference between the student’s functional communication performance and the gender (p=1.000). Next, the proportion of pass student was highest among female students, while most of the children assessed as being of fail in functional communication performance were male students. Moreover, this study shows only two female students (8.3%) and three male students (7.7%) passed the functional communication task with criterion score above 91. Furthermore, twenty-two female students (91.7%) and thirty-six male students (8.3%) failed in this task. There was no significant difference shown since the differences between functional communication performance and gender are small.

The hypothesis stated there is a significant difference between functional communication performance with gender is rejected. This finding conflicts with the previous research o which found that gender differences between the dyads emerged and it was shown that the different kinds of communication had different effects for the different genders [26]. The study done shows that girls generally use more positive forms of communications while boys use more controlling acts and engaged in more domineering exchanges than girls [27]. Moreover, boys’ conversations are characterised by greater independence, competitiveness and dominance whereas girls’ conversations are characterised by closeness, cooperation and interpersonal exchange [28]. In addition, mentioned that boys delay behind girls in the development of many communication features [29]. Other


than that, it is emphasize that the effect size of gender on children’s language largely depends on their age and the language aspects [29]. In contrary, a study mentioned that generally differences decrease with age [30]. However, “these differences are seldom statistically significant, but the careful observer cannot ignore the amazing consistency with which theses small differences appear in one investigation after another, each being conducted by the different experimenter, employing different techniques, different subjects, and sampling different geographical populations” [31]. Furthermore, a study confirmed the existence of gender differences but it is limited and often interact with a variety of factors such as age [32].

3.6 The difference between functional communication performance with diagnosis

The difference between functional communication performance with diagnosis is analysed using one-way ANOVA. Moreover, one-way ANOVA used to determine whether there are any statistically significant differences between the means of two or more independent groups. The criterion scores of the functional communication performance will be associated with the diagnosis of the students. The diagnosis including ASD, ADHD, DS, GDD, MR, SL, CP and LD.

Table 4 shows there was no significant difference between the student’s functional communication performance and the diagnosis (p=0.075). Generally, all the diagnosis achieved scores significantly below that of children their age in this task. However, the three highest mean criterion scores are among SL, CP and ADHD, while the three lowest mean criterion scores are among GDD, DS and LD. Furthermore, ASD and MR in the middle between the highest and the lowest mean criterion score.

Table 4: Assessed functional communication performance by diagnosis.

Diagnosis n Mean (SD) F-stats

(df)

p-value

ASD 27 35.2 (20.1) 1.98 (7) 0.075

ADHD 3 42.0 (18.0)

DS 8 26.1 (19.0)

GDD 5 23.6 (26.5)

MR 2 39.5 (21.9)

SL 12 59.5 (25.9)

CP 3 53.0 (41.1)

LD 3 33.3 (57.7)

The diagnosis of students in this study are ASD (n=27), ADHD (n=3), DS (n=8), GDD (n=5), MR (n=2), SL (n=12), CP (n=3) and LD (n=3). This study shows there was no significant difference between the student’s functional communication performance and the diagnosis (p=0.075). Generally, all the diagnosis achieved scores significantly below that of children their age in this task. Moreover, the

two highest mean criterion scores are SL and CP which more than 52.0. However, both diagnoses still below criterion cut-off score of 91 which means the students’ performance are below of their typically developing peers. Furthermore, the four lowest mean criterion scores are ASD, LD, DS and GDD which below 36.0. There was no significant difference shown since the differences between functional communication performance and diagnosis are small.

The hypothesis stated there is a significant difference between functional communication performances with diagnosis is rejected. In this study, mean criterion scores of students with ASD is 35.2 (SD 20.1) which performance below of their typically developing peers. This finding was supported by previous which mentioned that children with ASD have limited communication skills as most of them having difficulty with communication and language as well as social interactions [33]. ASD is characterised by deficits in two core domains which are social communication and social interaction, restricted repetitive patterns of behaviour, interests and activities [34]. Next, mean criterion scores of students with LD is 33.3 (SD 75.7) which performance below of their typically developing peers. This finding is supported by a pointed out that children with SLD pose a real challenge when it comes to communication and have little or no speech [35]. Moreover, mean criterion scores of students with DS is 26.1 (SD 19.0) which performance below of their typically developing peers. This finding is supported mentioned that DS display a relative weakness in communication skills specifically in expressive language and grammar [36]. It is explained in their study that individuals with DS have a characteristic profile of language and communication strengths and difficulties which receptive language is typically stronger than expressive language [37]. Furthermore, a significant delay in two or more domains including gross or fine motor skills, cognition, speech or language, personal or social skills, or activities in daily living [38].

4. CONCLUSION

In a nutshell, education students showed poor performance in all activities assessed in SFA, which includes hygiene performance. No significant differences in functional communication performance were obtained among special education students with different age and gender. For diagnosis, result of significant difference was obtained between diagnosis pair of Down syndrome and slow learner, while no significant difference was obtained between other pair of diagnosis.

Student’s poor performance in school functional task, including hygiene highlighted the possibility of students to have issue with their development. It also highlights the possibility of the children to have underlying problems with their body functions that limit their engagement in functional activity. The inability to do functional task might be related to physical problems and cognitive problems caused by the disorder/ impairment. Specific evaluations and interventions targeted specifically at the root cause of the problem should be implemented to help special education students to perform better in academic and function activities at school.. This highlights the importance of establishing a specific

Health Scope 67

individualized educational programme (IEP) that involves multi-disciplinary team at school which includes other profession such as occupational therapy, physiotherapy and speech therapy in order to cater the needs of special education students holistically.

ACKNOWLEDGEMENTS

The authors would like to thank the Ministry of Education (MoE), Jabatan Pelajaran Negeri Selangor and Faculty of Health Sciences, Universiti Teknologi MARA, Puncak Alam for the permission and supports given.

REFERENCES

[1] M. N. Abdul Nasir and A. N. A. Erman Efendi,

“Special education for children with disabilities in

Malaysia: progress and obstacles,” Geografia:

Malaysian Journal of Society & Space, 2016.

[2] H. A. Lamas, “School Performance Sobre el

rendimiento escolar Citar Como,” Propósitos y

Representaciones, 2015, doi:

10.20511/pyr2015.v3n1.74.

[3] L. A. Daunhauer, D. J. Fidler, and E. Will, “School

function in students with down syndrome,”

American Journal of Occupational Therapy, 2014,

doi: 10.5014/ajot.2014.009274.

[4] W. Coster, “Occupation-Centered Assessment of

Children,” American Journal of Occupational

Therapy, 1998, doi: 10.5014/ajot.52.5.337.

[5] W. Strein and M. Kuhn-McKearin, “School Function

Assessment,” in Encyclopedia of Autism Spectrum

Disorders, 2013.

[6] R. J. Simeonsson, D. Carlson, G. S. Huntington, J. S.

McMillen, and J. L. Brent, “Students with

disabilities: A national survey of participation in

school activities,” Disability and Rehabilitation,

2001, doi: 10.1080/096382801750058134.

[7] W. J. Coster, M. C. Mancini, and L. H. Ludlow,

“Factor structure of the school function assessment,”

Educational and Psychological Measurement, 1999,

doi: 10.1177/00131649921970099.

[8] M. N. Mohamad Taib, “‘The Role of Special

Classes, Resource Rooms and Other Arrangements

in Regular Schools - Meeting Individual Educational

Needs while Forming a Cohesive Society,’” Journal

of Special Education in the Asia Pacific, vol. 5, pp.

64–68, 2009.

[9] K. Zuoud and R. Shboul, “The effects of

communication skills in developing preparatory tear

students’ performance,” Academics World 82nd

International Conference, Langkawi, Malaysia 18-

19 June, 2018.

[10] V. A. Reed, K. McLeod, and L. McAllister,

“Importance of selected communication skills for

talking with peers and teachers: Adolescents’

opinions,” Language, Speech, and Hearing Services

in Schools, 1999, doi: 10.1044/0161-1461.3001.32.

[11] P. Mirenda, “Supporting individuals with

challenging behavior through functional

communication training and AAC: Research

review,” AAC: Augmentative and Alternative

Communication, 1997, doi:

10.1080/07434619712331278048.

[12] L. B. Olswang, L. Svensson, and S. Astley,

“Observation of classroom social communication:

Do children with fetal alcohol spectrum disorders

spend their time differently than their typically

developing peers?,” Journal of Speech, Language,

and Hearing Research, 2010, doi: 10.1044/1092-

4388(2010/09-0092).

[13] M. J. Tsai, “Communication interaction in special

education preschool classrooms,” International

Journal of Developmental Disabilities, 2016, doi:

10.1179/2047387715Y.0000000013.

[14] S. P. Morreale and J. C. Pearson, “Why

communication education is important: The

centrality of the discipline in the 21st century,”

Communication Education, 2008, doi:

10.1080/03634520701861713.

[15] M. Zainol and R. A. Majid, “Pelaksanaan Terapi

Carakerja Demi Penguasaan Kemahiran Motor Halus

Murid Bekeperluan Khas di Sekolah,” Journal of

Special Education, vol. 3, no. 1, pp. 81–91, 2013.

[16] S. T. Egilson and W. J. Coster, “School Function

Assessment: Performance of Icelandic students with

special needs,” Scandinavian Journal of

Occupational Therapy, 2004, doi:

10.1080/11038120410020737.

[17] D. R. Beukelman and P. Mirenda, Augmentative &

alternative communication: Supporting children and

adults with complex communication needs. Paul H.

Brookes Publishing, 2013.

[18] K. Himmelmann, K. Lindh, and M. J. C. Hidecker,

“Communication ability in cerebral palsy: A study

from the CP register of western Sweden,” European

Journal of Paediatric Neurology, 2013, doi:

10.1016/j.ejpn.2013.04.005.

[19] J. H. Bensted-Smith, “Participation, Activity

Performance and Task Support of Cerebral Palsied

Children Aged Between 4 and 12 Years in

Mainstream Schooling Environments in Gauteng,”

Master Degree, University of the Witwatersrand,

Johannesburg, 2014.

[20] L. Abbeduto, “Development of Verbal

Communication in Persons with Moderate to Mild

Mental Retardation,” International Review of

Research in Mental Retardation, 1991, doi:

10.1016/S0074-7750(08)60104-4.

[21] E. Hawkins, S. Gathercole, D. Astle, and J. Holmes,

“Language Problems and ADHD Symptoms: How

Specific Are the Links?,” Brain Sciences, vol. 6, no.

4, p. 50, 2016, doi: doi: 10.3390/brainsci6040050.

[22] S. P. Morreale, J. M. Valenzano, and J. A. Bauer,

“Why communication education is important: a third

study on the centrality of the discipline’s content and

pedagogy,” Communication Education, 2017, doi:

10.1080/03634523.2016.1265136.

[23] J. Sigafoos, D. Roberts, M. Kerr, D. Couzens, and A.

J. Baglioni, “Opportunities for communication in

classrooms serving children with developmental


disabilities,” Journal of Autism and Developmental

Disorders, 1994, doi: 10.1007/BF02172226.

[24] B. Kolucki and D. Lemish, “Communicating with

Children,” 2011.

https://www.unicef.org/cwc/files/CwC_Final_Nov-

2011.pdf.

[25] P. J. Leman, “Argument structure, argument content,

and cognitive change in children’s peer interaction,”

Journal of Genetic Psychology, 2002, doi:

10.1080/00221320209597967.

[26] S. M. Murphy and D. Faulkner, “Gender differences

in verbal communication between popular and

unpopular children during an interactive task,”

Social Development, 2006, doi: 10.1111/j.1467-

9507.2006.00331.x.

[27] P. J. Leman, S. Ahmed, and L. Ozarow, “Gender,

gender relations, and the social dynamics of

children’s conversations,” Developmental

Psychology, 2005, doi: 10.1037/0012-1649.41.1.64.

[28] C. Leaper, “Influence and Involvement in Children’s

Discourse: Age, Gender, and Partner Effects,” Child

Development, 1991, doi: 10.2307/1131178.

[29] S. Adani and M. Cepanec, “Sex differences in early

communication development: Behavioral and

neurobiological indicators of more vulnerable

communication system development in boys,” 2019,

doi: 10.3325/cmj.2019.60.141.

[30] B. P. Lange, H. A. Euler, and E. Zaretsky, “Sex

differences in language competence of 3- to 6-year-

old children,” Applied Psycholinguistics, 2016, doi:

10.1017/S0142716415000624.

[31] D. Mccarthy, “Some Possible Explanations of Sex

Differences in Language Development and

Disorders,” Journal of Psychology: Interdisciplinary

and Applied, 1953, doi:

10.1080/00223980.1953.9712848.

[32] A. Etchell et al., “A systematic literature review of

sex differences in childhood language and brain

development,” Neuropsychologia. 2018, doi:

10.1016/j.neuropsychologia.2018.04.011.

[33] J. N. Cadette, C. L. Wilson, M. P. Brady, C. Dukes,

and K. D. Bennett, “The Effectiveness of Direct

Instruction in Teaching Students with Autism

Spectrum Disorder to Answer ‘Wh-’ Questions,”

Journal of Autism and Developmental Disorders,

2016, doi: 10.1007/s10803-016-2825-2.

[34] K. Y. Teng, “Contributions of Inclusive Education

Programme, Teachers’ Attitudes and Efficacy on

Social Skills Among Students with Autism Spectrum

Disorder,” Universiti Teknologi Malaysia, 2016.

[35] P. Mutumburanzou, “Communication Skills for

Children with Severe Learning Difficulties,”

European Journal of Educational Sciences, vol. 5,

no. 2, pp. 75–83, 2018.

[36] E. Berglund and M. Eriksson, “Communicative

development in Swedish children 16-28 months old:

The Swedish early communicative development

inventory - Words and sentences,” Scandinavian

Journal of Psychology, 2000, doi: 10.1111/1467-

9450.00181.

[37] G. E. Martin, J. Klusek, B. Estigarribia, and J. E.

Roberts, “Language characteristics of individuals

with down syndrome,” Topics in Language

Disorders, 2009, doi:

10.1097/TLD.0b013e3181a71fe1.

[38] M. Shevell et al., “Practice parameter: Evaluation of

the child with global developmental delay: Report of

the quality standards subcommittee of the American

Academy of Neurology and The Practice Committee

of the Child Neurology Society,” Neurology, 2003,

doi: 10.1212/01.WNL.0000031431.81555.16.

Health Scope 69


RESEARCH ARTICLE

Nutritional status and feeding practices among toddlers in selected nurseries in Puncak Alam, Selangor, Malaysia

Naleena Devi Muniandy*, Nik Aishah Aliffah Mahdy, Ifrah Ruwaida Mat Saad

Center of Nutrition and Dietetics, Faculty of Health Sciences, Universiti Teknologi MARA Cawangan Selangor Kampus


Abstract:

Nutrition intake and growth during the first five years of life is vital to ensure optimum growth

and cognitive development of a child. The National Health and Morbidity Survey in this

country indicated that stunting was highest among children 24 to 35 months while overweight

was highest among 36 to 47 months, indicating the importance of nutrition assessment in these

groups of children. Hence, this study was designed to assess the current growth and nutrition

intake and history of feeding practices of the participants. Sixty-seven participants from nine

selected nurseries were included in this cross-sectional study. The anthropometry measurements

were taken and compared with the WHO growth charts. A three days food diary was used to

assess the nutrition intake of the participants. The study revealed that the participants consumed

more than twice than the national protein recommendation (boys: 30.7 ± 4.4, and girls 29.7±

5.0) and low calcium intake (boys: 369.7±108.5, girls: 444±51.9). The results also showed a

poor history of feeding practices, in which only 25% of the participants were exclusively

breastfed until six months and early initiation of solids and semi-solids, as around 25% of the

participants were introduced to solids or semi-solids before six months. The results strongly

suggest that more initiative needs to be taken to educate mothers on infant and young child

feeding practices as nutritional intake and growth during this period are associated with many

medical conditions in the future.


Naleena Devi Muniandy


Keywords: Early feeding practices, nutritional status, nurseries toddlers

1. INTRODUCTION

Nutritional status is important in assessing the state of

nourishment in a child. Growth and dietary intake of toddlers

are essential as these parameters reflect feeding practices

during infancy and current nutrition intake, which are known

risk factors to develop obesity or other non-communicable

diseases in the future [1,2]. Factors that influence the

nutritional intake of children include inadequate food intake,

inappropriate feeding practices, perinatal care practices,

socio-economic status, undesirable socio-cultural practices,

maternal literacy and employment, etc. [3-6].

Traditionally, feeding practices of infants and young children

were reported as breastfeeding and complementary feeding

practices. However, in recent years complementary feeding

has been replaced with the term solid, semi solid or soft food

as it is easier to measure food intake using these terms [7]. In

this study, feeding practices are reported as breastfeeding

and solid and semi solids feeding practices.

Early feeding practices is one of the factors that have a

significant influence on growth and cognitive performance

among children globally. Studies show that inappropriate

feeding practices during the first year of life is the main

factor for malnutrition and poor cognitive performance in

children under five years old [8,9]. Early feeding practices

include breastfeeding practices and solid and semi solid

feeding practices during the first two years of life.

Early feeding practices are associated with growth and

development of toddlers. Studies show that time of

introduction, taste and texture of solids during the first year

of life has a strong effect on moulding the eating preferences

of a child in later life [4]. Studies indicate that the main

factors associated with feeding practices of a child is

maternal literacy and employment [5,10].

The report by National Health and Morbidity Survey

(NHMS) indicated that there was an increase in the

percentage of children grouped under the normal weight

category from 2006 to 2015 [11,12]. Those studies too

70 Health Scope, 2020, Vol. 3(3) Naleena et al.

indicated that the percentage of children grouped under the

normal weight category had reduced from 83.7% in 2006 to

80.3% in 2015. The data also reported an increase in the

prevalence of both overweight and underweight, suggesting

that the country still suffers the issue of dual burden of

malnutrition.

The increase in the prevalence of underweight and

overweight in children under five in this country calls for

intervention much earlier in the life cycle. There is fair

amount of studies reported on nutrition intake during the first

year of life; however, data on nutritional status during first to

the third year of life is scarce. This may be due to the

technical effort involved in assessing the nutritional intake of

the children during this stage. Referring to the existing

studies that link the importance of nutritional intake during

early stages of life with growth and nutritional practices in

early childhood, hence it is worth to investigate the nutrition

status of children aged 1 to 3 years old in this country.

Hence, this study was conducted in selected childcare

nurseries in Puncak Alam, Selangor to assess the growth and

nutrition intake of the toddlers and evaluate the history of the

feeding practices of the toddlers.

The study was conducted in nurseries as it is convenient to

approach study participants and their respective mothers

within the selected age in these settings. The study will give

an idea of the baseline characteristics of anthropometry,

current nutrition intake and early feeding practices data of

children in the selected age in Puncak Alam, Selangor.


2.1 Study Design and Location

This was a cross-sectional study conducted in nine

selected nurseries in Puncak Alam, Selangor. A simple

random sampling was used to choose the nine selected

nurseries which were located in Puncak Alam Selangor.

Participants from all the nine nurseries were recruited

according to convenient sampling. Participants whose

parents were willing to participate in this study only were

enrolled in the study.

2.2 Participants

Children aged one to three years old were recruited in the

study based on the inclusion criterias of (1) aged 1 to 3 years

old, (2) were born between 37 to 42 gestation weeks and (3)

have births weight of more than 2.5kg and less than 4.5kg.

The exclusion criterias are (1) the participant was not

suffering from any illnesses or diseases that can influence

food intake and nutritional status, such as gastrointestinal

problems, allergy, inborn error, asthma, diabetes, congenital

heart disease, kidney disease, liver failure, etc and (2)

participants were not admitted to the hospital anytime during

the past six months.

2.3 Sample Size

96 participants were needed after calculating for sample

size using Krejcie & Morgan (1979) formula including a

20% dropout.

2.4 Instruments and methods

Table 1. Objectives, study variables and instruments of the

study

Objective Study Variables Instruments

To assess the

anthropometry status of the children

Weight for age

Height for age Weight for height

Head circumference

MUAC

WHO Growth charts

Weighing Scale SECA 813

Body Meter SECA

206 Measuring tape

To evaluate the

nutritional intake of the children

Total macronutrient

and micronutrients intake

Food Diary (3days)

RNI Nutritionist Pro

2.5 Data Collection

Data collection was conducted after seeking permission

from the respected kindergartens and after obtaining

permission from the parents through a consent form. A

questionnaire that consists of 4 parts which were

demographic details, prenatal characteristics, child

characteristics and feeding history was used to collect

information needed. The anthropometric measurements such

as weight, height, head circumference and mid upper

armcircumference of the participants were taken during

breaktime by a trained research assistant (RA). All

measurements were taken according to the NHANES

Anthropometry Procedures Manual [8]. Type of

measurements and instruments used is given in Table 1. The

RA later helped in filling in the questionnaire by

interviewing the parents of the participants. At the end of the

interview, the RA educated the parent or caregiver that will

be feeding the child on how to fill in the Three Days Food

Diary.

Participants' parents or caegivers were expected to fill in all

the solids/ semi solid and liquid taken for the whole day on

two weekdays and one weekend. The portion and quantity of

food were recorded according to the given handouts that

consisted a list and picture of household measurements.

Parents/ caregivers were given two weeks to complete the

given food diary. Returned food diaries were examined by

the RA and attached to the questionnaire before data

analysis.

2.6 Data Analysis

Anthropometry data was interpreted using WHO Growth

Charts. Three days food record was interpreted using diet 4

software. Then the average values were compared to

Recommended Nutrient Intake (RNI) of Malaysia. Data was

entered and analyzed using Statistical Package for the Social

Sciences (SPSS) version 17. Discriptive statistics were

presented in tables and graphs.


3.1 Demographic information

67 participants from a total of 80 participants completed

and returned the questionnaire. All the participants were

Muslim and lived with their parents who were both working.

The distributions between the gender of the participants were

nearly equal in which 50.7% (n=34) of them were boys and

Health Scope 71

49.3% (n=33) were girls. The distribution of participants

according to their age is given in table 3.1. There were

43.4% who were one year old, 31.3% aged two years old and

25.5% aged three years old. All the mothers had a minimum

of secondary school education in which 46.3% of them had a

SPM/STPM/Diploma education while the remaining 53.7%

had a tertiary education, as shown in Table 3.1.

Table 2. Demographic information of the subjects (n=67)

Demographic characteristics n Percentage (%)

Gender

Boy 34 50.7

Girl 33 49.3 Age

1 years old 29 43.4

2 years old 21 31.3 3 years old 17 25.4

Maternal education level

Secondary SPM/ SLE 14 20.9

STPM/ Diploma/ A –

Level/ Sijil 17 25.4

Tertiary

Degree 33 49.2 Master/ PhD 3 4.5

3.2 Anthropometry characteristics of the participants.

The anthropometry characteristics of the participants

according to the WHO Growth Charts are tabulated in table

3.2. None of the participants were severely underweight or

overweight; nearly 90% of the participants fell in the

category of normal weight. However, 15% of the participants

were stunted and 1.5% were severely stunted. Nearly 96% of

the participants had normal BMI and MUAC while 97% of

them had normal head circumference for age.

The percentage of underweight was higher in boys (14.7%)

than girls (6.1%). Same goes to the prevalence of stunting

which was higher in boys (17.6%) than girls (12.1%).

Table 4. Proportion of participants according to WHO

growth chart.

Infant’s

characteristics

Boys, n = 35 Girls, n = 32 Total, n=67

(%) (%) (%)

Weight-for-age

Severe Underweight

0 0 0

Underweight 14.7 6.1 10.4 Normal 85.3 93.9 89.6

Overweight 0 0 0

Height-for-age

Severe stunting 2.9 0 1.5

Stunting 17.6 12.1 14.9

Normal 79.4 87.9 83.6 BMI-for-age

Severe wasting 0 0 0

Wasting 0 6.1 3 Normal 97.1 93.3 95.5

Overweight 2.9 0 1.5

Obese 0 0 0 MUAC-for-age

Severe wasting 0 0 0

Wasting 2.9 6.1 4.5 Normal 97.1 93.9 95.5

HC-for-age

Microcephaly 2.9 3 3 Normal 97.1 97 97

Macrocephaly 0 0 0

3.3 History of breastfeeding

Breastfeeding practices were reported as exclusively

breastfed and any kind of breastfeeding which may include

exclusive breastfeeding, breastfed with solid intakes or

mixed feeding (formula milk and breast milk). This study

found that 53.7% of the participants were exclusively

breastfed for the first six months. Most participants were

provided any kind of breastfeeding for up to 3 months

(40.3%). Only 3% of the participants received any kind of

breastfeeding after six months.

Table 3. History of breastfeeding

n %

Infant feeding practices

exclusively breastfed for the first six months 36 53.7

Duration of any kind of breastfeeding

< 1 month 5 7.5

1st month 9 13.4

3rd month 27 40.3

5th month 7 10.4

6th month 17 25.4

> 6months 2 3

3.4 Time of introduction to solid or semi-solid food

The finding indicated that most of the participants (68.7%)

were introduced to solid and semi-solid food at the age of 6

months, followed by 19.4% at 3 to 4 months 7.5% below 3

months old and 4.5 % after six months.

Figure 1: Time of initiation to solid or semi-solid food

3.5 Solid and Semi solids feeding practices

Most mothers practiced appropriate feeding practices during

the introduction to solid and semi-solid as shown in figure

3.2. Most of the mothers (68.7%) introduced solid and semi

solids to their infants at six months. However, less than 50%

of the mothers did not practice providing the same kind of

new food for three days consecutively before trying out

something new. 62.7% did offer a particular food that was

rejected by the child a few times and 73.1% practiced

introducing one new food at a time. There were around

72 Health Scope, 2020, Vol. 3(3) Naleena et al.

37.3% of the mothers who did not practice placing the child

in a proper position before feeding.

Figure 2. Percentage of mothers that practiced appropriate

solid and semi-solid feeding practices

3.6 Comparison between the current nutrient intake of

the participants with the Recommended Nutrient Intake

(RNI) of Malaysia.

Both male and female participants consumed protein

higher than the national recommendations. The mean protein

intake of the male participants was 18.7 g higher than the

recommendation while the mean protein intake of the girl

participants was 17.7g higher than the recommendations.

The mean fat intake of the participants was also higher than

the recommendation.

However, the calcium intake of both the male and female

participants were lower than the recommendation. The mean

calcium intake of the male and female participants was

330mg and 256mg lesser than the recommendation. The iron

intake of the participants from the gender were around 62%

higher than the recommendation. The comparison of the

mean of selected nutrient intake of the participants for 3 days

with the RNI of Malaysia is shown in Table 3.4.

Table 6. Comparing nutrient intake of subjects with

Recommended Nutrient Intake (RNI)

Boys, n=35 Girls, n=32

Intake

(Mean±SD) RNI (%)

Intake

(Mean±SD) RNI (%)

Energy (Kcal) 984.89±81.8 980 101 939.1±33.1 900 103 Protein (g) 30.7±4.4 12 181 29.7±5 12 175

Carbohydrate

(g) 134.4±25 - - 125.3±16.1 - -

Fat (g) 39.8±6.7

27-

38 - 36±4.9

25-

35 -

Calcium (mg) 369.7±108.5 700 74 444±51.9 700 89 Iron (mg) 9.7±1.2 6 162 9.9±0.5 6 165

3.7 Comparison of selected macronutrients and

micronutrients intake of the participants according to

commercial and homemade food intake

The distribution of selected macro and micronutrient

intake among the participants were grouped according to

those who had 100% of homemade food in the selected 3

days and those who had at least one commercial food in the

3 days. Table 3.4 shows that there was a significant

difference in the calorie intake of the participants from both

the groups.

Table 5. Comparison between the types of complementary

food with the nutrient intake of the subjects

Nutrient

Commercial food,

n=11

Homemade food,

n=56 p-value

Mean ± SD Mean ± SD

Energy 926±62.93 969.48±65.24 0.046* Carbohydrate 126.21±18.74 130.62±21.96 0.536

Protein 30.56±5.77 30.08±4.48 0.754

Fat 35.65±6.94 38.39±5.92 0.177 Calcium 433.67±99.22 401.04±91.43 0.290

Iron 9.64±0.91 9.81±0.9 0.578

*level of significance at p<0.05

3.8 Discussion

Most of the subjects have normal anthropometry

measurements. The results of this study showed a higher

prevalence of stunting among male children compared to

females. These results correspond well with their calcium

intake in which the female children seem to consume higher

calcium intake compared to the males. Studies in a few third

world county also indicated a higher prevalence of stunting

in males compared to female’s children [1, 2]. However, the

higher prevalence of stunting among the males may be a

result of poor calcium and protein intake in male infants

during the first year of life in third world countries. Studies

also have indicated that early weaning of breastfeeding as a

source of stunting among the children in third world

countries hence it may be worth understanding the

differences in breastfeeding practices among the male and

female infants in this country to draw a better conclusion [2].

The percentage of exclusively breastfed participants in this

study were 53.7% which is higher than a previous study in

Klang valley that reported the prevalence of breastfeeding as

43% [3]. The results of any kind of breastfeeding also

declined notably from 3 months on words suggesting that

mothers begin to quit breastfeeding after resuming work. As

reported in Table 1, all the mothers were employed. A

qualitative study conducted in Malaysia also reported that

lack of awareness on methods of breastmilk expression and

storage and even lack of facilities to express and store the

breastmilk at work were essential factors to early cessation

of breastfeeding [4].

The results from this study show nearly 20% of the

participants were introduced to solids or semi solids at 4 to5

months, while around 7.5% of the participants were

introduced to solids and semi-solids within the third month

of their life. This might be a reason for nearly 15% of the

participants to be stunted as studies have shown that early

introduction to solids and semi-solids will compromise on

the milk intake of the infants hence result in a reduced height

for age [2].

Education on early feeding practices needs to be enforced

among mothers as the results indicated that around 30 to

40% of the mothers were not practicing appropriate solid or

semi solid feeding practices. It is important to establish good

feeding practices as it molds lifelong eating habits [8].

Education can be carried out both at mother and child clinics

or home visits as evidence suggest that early intervention

through these platform has helped to reduce the intake of

formula milk and increase duration of breastfeeding, reduce

Health Scope 73

the intake of sugar sweetened beverages and improved

child’s BMI [9].

The results of this study also showed that the participants

consumed more than twice higher protein intake than the

recommendations. Studies show that high protein intake will

result in accelerated growth and will increase the risk of

being overweight or obese later in childhood [5, 6]. However

there are convincing studies that suggest the type of protein

rather than the amount itself is associated with the risk of

developing obesity in the future [5-7].

The study also reported that those who consumed at least one

commercial food during the three days logged in their food

diary, indicated a higher mean energy intake compared to

those who consumed food for all three days from home or

from the nursery. This is a good indicator to show that

commercial food may contain higher energy intake

compared to home-cooked food as studies have reported a

high level of sugar, salt and fat in commercially prepared

food [10]. Hence a high intake of this commercially

available food may result in future obesity.

4. CONCLUSIONS

The findings from this study are in line with the national

statistics of this country that indicated a rapid increment in

childhood obesity and stunting in this country in recent years

[11-13]. However, more cross-sectional data on growth and

nutritional intake in participants of the same group should be

gathered to draw a conclusive hypothesis that may be used to

implement early intervention on feeding practices in

children. Given the broad literature on the effect of type of

protein consumed in children during early life may have a

stronger impact on the future development of obesity; hence

a more comprehensive dietary assessment should be done

mainly on the types of protein intake during infancy and the

association with overweight and obesity in early childhood

and adulthood. In addition, it is worth to follow up a cohort

from birth to the first 5 years to study the effect of early

nutritional intake on growth and eating practices throughout

childhood.

ACKNOWLEDGEMENTS

The authors would like to thank the Faculty of Health

Sciences, Universiti Teknologi MARA, Puncak Alam

campus and to all the owners of the involved nurseries,

caregivers, the subjects and their parents who participated in

the study.

REFERENCES

[1] Abrams, E. M., & Becker, A. B.. Food introduction and

allergy prevention in infants. CMAJ, 2015

[2] Adnan, N., & Muniandy, N. D.The relationship between

mothers’ educational level and feeding practices among

children in selected kindergartens in Selangor, Malaysia: A

crosssectional study. Asian Journal of Clinical Nutrition. Asian

Journal of Clinical Nutrition, 2012.

[3] Agedew, E., & Chane, T.. Prevalence of Stunting among

Children Aged 6–23 Months in Kemba Woreda, Southern

Ethiopia: A Community Based Cross-Sectional Study..

Advance in Public Health, 2015

[4] Bork, K. A., & Diallo, A.. Boys Are More Stunted than Girls

from Early Infancy to 3 Years of Age in Rural Senegal. The

Journal of Nutrition.,2017.

[5] C. Agostoni et al., “Complementary feeding: a commentary by

the ESPGHAN Committee on Nutrition.,” Journal of

Pediatric Gastroenterology and Nutrition, 2008

[6] Campoy, C., Campos, D., Cerdó, T., Diéguez, E., & Garciá-

Santos, J. A. (2018). Complementary feeding in developed

countries: The 3 Ws (When, what, and why?). Annals of

Nutrition and Metabolism, 73(suppl 1), 27–36.

[7] Cattaneo, A., Williams, C., Pallás-Alonso, C. R., Hernández-

Aguilar, M. T., Lasarte-Velillas, J. J., Landa-Rivera, L., …

Oudesluys-Murphy, A. M.. ESPGHAN’s 2008

recommendation for early introduction of complementary

foods: How good is the evidence? Maternal and Child

Nutrition. (2011)

[8] Centres for Disease Control and Prevention. Anthropometry

procedures manual. National Health and Nutrition

Examinatory Survey (NHANES). (2007).

[9] Chekol, D. A., Biks, G. A., Gelaw, Y. A., & Melsew, Y. A..

Exclusive breastfeeding and mothers’ employment status in

Gondar town, Northwest Ethiopia: A comparative

crosssectional study. International Breastfeeding Journal,

12(1), 1–9. (2017)

[10] Findings, C. H.. NHMS2016ReportVolumeII-

MaternalChildHealthFindingsv2. Retrieved from

http://iku.moh.gov.my/images/IKU/Document/REPORT/2016/

NHMS2016ReportVolumeIIMaternalChildHealthFindingsv2.p

df(2008)

[11] Gandhi, S., Godara, N., Modi, A., & Kantharia, S. (2014).

Impact of feeding practices on nutritional status of children in

rural area of Navsari district. International Journal of Medical

Science and Public Health, 3(11), 1338.

[12] Hausner, H., Hartvig, D. L., Reinbach, H. C., Wendin, K., &

Bredie, W. L. P.. Effects of repeated exposure on acceptance

of initially disliked and liked Nordic snack bars in 9-11 year-

old children. Clinical Nutrition.(2012)

[13] Health Canada, Canadian Paediatric Society, Dietitians of

Canada, & Breastfeeding Committee for Canada. (2012).

Nutrition for healthy term infants: recommendations from birth

to six months. Canadian Journal of Dietetic Practice and

Research : A Publication of Dietitians of Canada

Health Scope 74


RESEARCH ARTICLE

Development of a healthy cookie to promote toddlers’ brain development:

Brainy Bites

Airin Fatehah, Naleena Devi*

1Centre of Nutrition and Dietetics, Faculty of Health Sciences, Universiti Teknologi MARA Cawangan Selangor Kampus

Puncak Alam, 42300 Bandar Puncak Alam, Selangor, Malaysia.

Abstract:

This study aims to develop a healthy cookie designed for toddlers. “Brainy Bites” was developed

by incorporating ingredients that are linked with healthy brain development and improved

cognitive function in early childhood. Proximate analysis revealed that Brainy Bites contain

67.71 ± 0.23 carbohydrate, 3.87± 0.07 crude protein, 18.51 ± 0.24 fat, 1.80 ± 0.01crude fiber,

1.63 ± 0.01 ash and 8.29 ± 0.10 moisture. Physical analysis done on Brainy Bites indicated that

the developed cookie had 0.5 cm thickness, 3.7 cm width, 7.4 spread ratio, 18.18% bake loss and

1392.4 g hardness, respectively. There were significant difference (p < 0.05) in the nutrients

content of Brainy Bites and the two cookies that were compared in which Brainy Bites contains

significantly higher amount of crude protein, crude fiber and ash content . Brainy Bites will serve

as a great snack for toddlers at home, nurseries or for travelling. The cookie will be accepted in

the market and would be a better choice of snack for toddlers as it contains no added sugar and

salt, and contains ingredients such as raisins, almonds and chia seeds that may help in the brain

development among toddlers.


Naleena Devi Muniandy


Keywords: Brain development, cookie, early nutrition, healthy snack, toddler

1. INTRODUCTION

Toddlers are children aged one to three years old who are

highly curious and have a high sense of independence [1].

Toddlers should have a balanced, nutritious, and high energy

meal as this is the age of rapid growth and development [2].

One of the most common nutrition challenges among toddlers

is “picky eating” [3].

Picky eaters are prone to become under-nourished as they

refuse to eat many types of food and tend to have meals that

are unhealthy [4]. Picky eaters often have a poor intake of

fruits and vegetables which lead to poor consumption of

dietary fiber [5].

In this 21st century, most parents are concerned about their

children’s’ cognitive and physical development [6]. One of

the most important factors contributing to cognitive

development in toddlers is nutritional intake [7]. Hence,

parents nowadays are giving priority to provide their children

with food that is important for their brain development [8].

Studies show that the important nutrients for brain

development include iron, choline, vitamin B12, antioxidant

such as resveratrol, Docosahexaenoic acid (DHA), α-linolenic

acid (ALA), and Arachidonic acid (AA) [9-12].

Iron is an important micronutrient needed for healthy brain

development as iron is essential for normal neurotransmitter

development and production [13]. Insufficient intake of iron

in the diet will affect the normal development of the central

nervous system [14]. Choline is an essential nutrient for

memory and brain development [12]. Choline is needed for

the synthesis of neurotransmitter namely acetylcholine,

phospholipids, and lipoprotein plus methyl group metabolism

[12]. Providing choline during critical period of development

will improve the child’s memory.

Vitamin B12 is necessary to be taken during early childhood

because of the importance in brain development of children

[11]. Providing adequate amount of vitamin B12 in children

would improve or enhance their brain performance as well as

performance during school [11]. One of the antioxidants that

have functioned with brain health is resveratrol which is a

polyphenolic phytoalexin found abundantly from fruits such

as grapes, berries and peanuts [15]. Resveratrol improves

cognitive function, enhance psychomotor function, and reduce

oxidative stress [10].

The major Long Chain Fatty Acids that are often linked with

brain development are docosahexaenoic acid (DHA) which is

an Omega 3 Fatty acid and the Arachidonic Acid (AA) which

75 Health Scope, 2020, Vol. 3(3) Airin Fatehah et al.

is an Omega 6 fatty acid [9]. Research has also proven that

DHA is also linked with visual functionality in healthy

children [16].

Hence this study aimed to develop a nutritious cookie for

toddlers using ingredients such as raisins, almonds, and chia

seeds that are often linked with healthy brain development

and cognitive performance in children. Brainy Bites will serve

as a great snack for toddlers as it is convenient, nutritious and

contains many ingredients that help in promoting a healthy

brain development for the targeted age group.

2. METHODOLOGY

2.1. Product Development

Table 1 shows the cookie formulation of the product. The

standard dough mixture of 462.18 g yielded approximately 80

cookies. There were four types of flour used: whole meal

flour, white flour, cornstarch, and rice flour with the

proportion of 42 g, 70 g, 15 g, and 9.88 g respectively. Wheat

germ (10g), chia seed (2g), almond (55g), raisins (55g),

banana (85g), canola oil (25g and Omega 3 fortified eggs

(35g) were added in the mixture.

Table 1. Cookie formulation in this study.

Ingredients Amounts (g) / 462.18 g Percentage (%)

Whole Meal Flour 42 9.09

White Flour 70 15.15

Cornstarch 15 3.25

Rice Flour 9.88 2.13

Wheat Germ 10 2.16

Chia Seed 2 0.43

Raisins 55 11.9

Banana 85 18.39

Almond 55 11.9

Canola Oil 25 5.41

Omega-3 Fortified Egg 35 7.57

Cinnamon 1.3 0.28

Vanilla Extract 2 0.43

Raisins, almonds, wheat germ and whole meal flour were first

pre-dried in an oven at 170̊ C for three minutes. Next, raisins

were cut into small bits using a knife and almond, wheat germ

and whole grain flour were finely grounded using a dry

blender. After this step, the 3 different type of flours (wheat

flour, corn starch and rice flour) were mixed in a bowl

together with all the other dry ingredients such as the chopped

raisins, grounded ingredients and chia seeds. The whole

banana was mashed with a spoon. All the wet ingredients such

as the canola oil, egg, mashed banana was combined in a

different bowl. Cinnamon and vanilla extract were added into

the wet ingredients. The wet ingredient mixture was then

combined with the dry ingredients and were kneaded into

cookie dough. Once homogenized, the cookie dough was

rolled to a thin layer and was cut using a cookie cutter to form

a doughnut shape. The cookie was then baked at a

temperature of 170̊ C for 10 minutes. Figure 1 shows the flow

chart of the cookie development.

Figure 1. Flowchart of the development of Brainy Bites

2.2 Proximate Analysis

Proximate analysis was done on the developed cookie and

the nutrient content was compared with two existing cookies

in the market. Crude fiber, moisture and ash were determined

by following the method from Association of Official

Analytical Chemists, AOAC [17]. Crude protein was

determined using Kjeldahl method while fat was determined

using Soxhlet method. Carbohydrate content of the developed

cookie was determined using calculation. The proximate

analysis was done at the Food Analysis Laboratory, Centre of

Nutrition and Dietetics Puncak Alam Campus, UiTM

Selangor,

2.3 Physical Analysis

The physical analysis of the cookie was determined on the

day of baking. The physical analysis done include thickness,

width, spread ratio, bake loss and hardness. The thickness of

the developed cookie was measured by stacking six cookies

and measured using a measuring scale [18]. The cookies were

restacked to obtain the average reading. The width of six

cookies was also measured by laying them edge to edge [18].

They were rotated 90 ̊ and re-measured to obtain the average

reading. The spread ratio of cookie was calculated by dividing

the width and thickness of the cookie [19]. The bake loss of

five cookies was measured for their weight before and after

baking [20]. The hardness of cookie was measured by using

texture analyzer (model TA.XT plus Stable Micro Systems,

England) using a blade probe with a load cell of 5 kg applied

at a crosshead speed of 60 mm/min [21]. The hardness of

cookie was expressed in gram (g). A higher reading indicated

greater force needed and therefore tougher cookie.

Health Scope 76


2.4 Comparison of Nutrient Content of The Developed

Cookie with Two Selected Cookies Available in the

Country for The Same Age Group.

The nutritional content of the newly developed cookie was

compared with two selected cookies for the same age group.

The two cookies that were chosen for comparison were

chosen based on availability, age group and nutritional

content.

2.5 Statistical Analysis

The data were analyzed using Statistical Package for

Social Sciences (SPSS) version 21.0. Descriptive statistics

such as ANOVA was performed and reported.

3. RESULT AND DISCUSSION

3.1 Product Development

Figure 2 shows the final product of Brainy Bites. The

selection of ingredients used to formulate this cookie took into

consideration several factors such as its nutrient content,

availability of the raw ingredients, good flavor, aroma, and

baking quality as well. The ingredients were chosen based on

their nutrient content related to nutrition that are associated

with toddlers' brain development. Literature review was first

done to identify which nutrients best help in enhancing brain

development and which food ingredients contain these

nutrients. In this recipe, there were no addition of sugar and

sodium as sugar to promote health food choices for this age

group.

Figure 2. Brainy Bites

3.2 Proximate Analysis

Proximate analysis was done on Brainy Bites and two

other compared cookies for the same age group to analyze

their nutrient content. Table 2 summarizes the results of the

proximate analysis on the three cookies.

Table 2 shows the result of proximate analysis done on the

three compared cookies. The results represent the mean ±

standard deviation (SD) from duplicate1, triplicate2 and

quadruplicate3 samples. The results for every nutrient in every

cookie depend on the ingredients incorporated in each cookie.

The results indicate that ‘Brainy Bites” has a significantly

higher protein and fibre content compared to the other two

control cookies.

Table 2. Nutrient content of Brainy Bites, Cookie B and

Cookie C.

% Nutrients

(per 100g)

Sample p-

value* Brainy Bites Cookie B Cookie C

Carbohydrate1 67.71 ± 0.23 77.14 ± 1.19 57.77 ± 1.81 0.001

Crude

Protein3 3.87± 0.07 2.92 ± 0.04 2.23 ± 0.06 0.000

Fat1 18.51 ± 0.24 13.19 ± 1.17 36.38 ± 1.84 0.001

Crude Fiber1 1.80 ± 0.01 1.70 ± 0.04 1.50 ± 0.06 0.012

Moisture2 8.29 ± 0.10 5.15 ± 0.03 2.34 ± 0.05 0.000

Ash1 1.63 ± 0.01 1.63 ± 0.01 1.31 ± 0.00 0.000

ANOVA significant at p<0.05

3.3 Physical Analysis

Physical analysis on Brainy Bites which included

thickness, width, spread ratio, bake loss and hardness was

conducted. Table 3 summarizes the physical characteristics

for Brainy Bites.

Table 3. Physical analysis of Brainy Bites.

Physical Analysis Brainy Bites

Thickness (cm) 0.5

Width (cm) 3.7

Spread ratio 7.4

Bake loss (%) 18.18

Hardness (g) 1392.4

The thickness of the cookie was 0.5cm each with the width of

3.7 cm measured using a measuring scale. This was done to

ensure toddlers were able to hold the cookie and eat

independently as they prefer to eat independently at this age.

Therefore, giving autonomy to feed themselves could help

them improve their skills especially eating skills [22].

A cookie with a high spread ratio (higher width and lower

thickness) is considered a positive cookie characteristic [23].

Brainy Bites has an average spread ratio of 7.4. Due to the

incorporation of high fiber ingredients, the amount of

hydrophilic spot which binds water is high in the cookie

dough making the dough's viscosity high and thus resulting in

a reduced spread ratio [23]. The low spread ratio also might

be due to the absence of sugar in the dough as sugar helps in

increasing spread ratio by absorbing the moisture content of

the cookie [24].

Texture is important to analyze a cookie's quality. The higher

the hardness reading of a cookie, the more force is needed to

break it. indicates that greater force thus tougher cookie

texture. The average hardness of this developed cookie was

1392.4 g. A study reported that the hardness of a cookie

77 Health Scope, 2020, Vol. 3(3) Airin Fatehah et al.

increased with the incorporation of whole meal wheat in

which the cookie developed had a reading of 3670 g [23]. The

result was higher than the result for Brainy Bites. The

hardness of a cookie might be affected by the baking time too

[21] and moisture content of a cookie [24]. Sugar addition in

most cookie recipes will help to give the cookie its crispiness

aside from adding sweetness [24]. However, In Brainy Bites,

there was no addition of sugar. Therefore, this might be the

reason why Brainy Bites was slightly hard.

3.3 Comparison of Nutrient Content in Brainy Bites with

Selected Two Cookies for The Same Age Group.

Brainy Bites, Cookie B and Cookie C were compared for

their nutrient content. As summarized in Table 2, one-way

ANOVA* was done and it was found that there was

significant difference (p < 0.05) for all the nutrients in the

three cookies. Cookie B had the highest carbohydrate content

followed by Brainy Bites and Cookie C. The values were

77.14 ± 1.19, 67.71 ± 0.23 and 57.77 ± 1.81, respectively. The

high carbohydrate content of Cookie B might be due to the

ingredients such wheat flour, sugar, banana, skim milk

powder, apple puree, and barley malt extract. The lower

carbohydrate results in Brainy Bites was due to the

incorporation of ingredients such as a mixture of whole meal

flour, white flour, cornstarch, rice flour and the usage of

banana and raisins as the fruit and no additional sugar was

added.

Brainy Bites contained the highest amount of crude protein

followed by Cookie B and Cookie C. The mean percentage

were 3.87 ± 0.07, 2.92 ± 0.04, and 2.23 ± 0.06, respectively.

Brainy Bites contained the highest amount of crude protein

because there was an incorporation of high protein ingredients

such as egg, almond, whole meal flour, and wheat germ.

Cookie B in cooperated skimmed milk powder while Cookie

C used full cream milk powder as the protein source.

Cookie C contained the highest amount of fat followed by

Brainy Bites and Cookie B. The mean percentage of fat were

36.38 ± 1.84, 18.51 ± 0.24, and 13.19 ± 1.17, respectively.

Cookie C incorporated palm oil and full cream milk powder

which added up to the high-fat content in the cookie. Brainy

Bites fat content was mainly from canola oil and almond

which were also good sources of monounsaturated fatty acid

and polyunsaturated fatty acid. Cookie B fat content was

mainly from palm oil.

Brainy Bites had the highest percentage of crude fiber

followed by Cookie B and Cookie C which were 1.8 ± 0.01,

1.7 ± 0.04, and 1.5 ± 0.06, respectively. Brainy Bites was

purposely made with ingredients high in fiber such as whole

meal flour, wheat germ, chia seed, almond, raisins, and

banana. This was to help toddlers receive some fiber as

toddlers are mostly picky eaters and have a poor intake of

vegetables and fruits that are main source of fiber [25].

Cookie B had added inulin while Cookie C might have

obtained its fiber value from the wheat flour.

The percentage of moisture content in Brainy Bites, Cookie B

and Cookie C were 8.29 ± 0.10, 5.15 ± 0.03, and 2.34 ± 0.05,

respectively. The developed cookie contained the highest

percentage of moisture compared to the other two. The high

moisture content might be due to the high fiber content in the

developed cookie because of the high-water binding capacity

[26].

Brainy Bites had the highest percentage of ash followed by

Cookie B and Cookie C. Brainy Bites was designed to be a

functional cookie that helps in toddlers’ brain development by

incorporating various ingredients such as raisins, almonds and

chia seeds that are linked with healthy brain development and

enhanced cognitive function . The high ash content proved the

availability of minerals in the developed product.

4. CONCLUSION

The development of a healthy cookie designed for toddlers

that contains ingredient that are associated with good brain

development and improved cognitive performance was

successfully developed. Brainny Bites will serve as a great

snack for toddlers at home, nurseries or for travelling as it is a

healthy choice of snack for toddlers. The cookie will be

accepted in the market and would be a better choice of snack

for toddlers as it contains no added sugar and salt, a higher

amount of crude protein, crude fiber and ash compared to the

other two cookies that were designed for the same group

which are widely available in the market.

ACKNOWLEDGEMENTS

The authors would like to express their gratitude to the

Faculty of Health Sciences and Faculty of Hotel and Tourism

Management, Universiti Teknologi MARA Selangor Puncak

Alam Campus for granting permission to use the facilities

available in completing this study.

REFERENCES

[1] "CDC's Infant and Toddler Nutrition website", Centers for

Disease Control and Prevention, 2020. [Online]. Available:

https://www.cdc.gov/nutrition/infantandtoddlernutrition/index.html.

[2] M.C. Bartolo. "Nutrition in Childhood", Journal of Malta

College of Family Doctors, vol. 3, no. 1, pp. 12-20, 2014.

[3] J. Lumeng, A. Miller, D. Appugliese, K. Rosenblum and N.

Kaciroti, "Picky eating, pressuring feeding, and growth in

toddlers", Appetite, vol. 123, pp. 299-305, 2018.

[4] K. van der Horst, D. Deming, R. Lesniauskas, B. Carr and K.

Reidy, "Picky eating: Associations with child eating characteristics

and food intake", Appetite, vol. 103, pp. 286-293, 2016.

[5] C. M. Taylor, K. Northstone, S. M. Wernimont, and P. M.

Emmett, “Picky eating in preschool children: Associations with

dietary fibre intakes and stool hardness,” Appetite, vol. 100, pp. 263–

271, May 2016.

[6] A. Vijaindren, “The cost of children’s education,” New Straits

Times, 05-Nov-2017.

[7] A.7 Nyaradi, J. Li, S. Hickling, J. Foster, and W. H. Oddy, “The

role of nutrition in childrens neurocognitive development, from

pregnancy through childhood,” Frontiers in Human Neuroscience,

vol. 7, no. 7, Mar. 2013.

[8] L. Hart, S. Damiano, C. Cornell and S. Paxton, "What parents

know and want to learn about healthy eating and body image in

preschool children: a triangulated qualitative study with parents and

Early Childhood Professionals", BMC Public Health, vol. 15, no. 1,

2015.

[9] A. Eilander, D. Hundscheid, S. Osendarp, C. Transler and P.

Zock, "Effects of n-3 long chain polyunsaturated fatty acid

Health Scope 78


supplementation on visual and cognitive development throughout

childhood: A review of human studies", Prostaglandins,

Leukotrienes and Essential Fatty Acids, vol. 76, no. 4, pp. 189-203,

2007.

[10] S. M. Poulose, N. Thangthaeng, M. G. Miller, and B. Shukitt-

Hale, “Effects of pterostilbene and resveratrol on brain and

behavior,” Neurochemistry International, vol. 89, pp. 227–233, Jul.

2015.

[11] S. Venkatramanan, I. E. Armata, B. J. Strupp, and J. L.

Finkelstein, “Vitamin B-12 and Cognition in Children,” Advances in

Nutrition, vol. 7, no. 5, pp. 879–888, Sep. 2016.

[12] S. H. Zeisel and K.-A. D. Costa, “Choline: an essential nutrient

for public health,” Nutrition Reviews, vol. 67, no. 11, pp. 615–623,

Nov. 2009.

[13] M. Georgieff, "Long-term brain and behavioral consequences of

early iron deficiency", Nutrition Reviews, vol. 69, pp. S43-S48,

2011.

[14] J. Beard, "Why Iron Deficiency Is Important in Infant

Development", The Journal of Nutrition, vol. 138, no. 12, pp. 2534-

2536, 2008.

[15] F. Jardim et al., "Resveratrol and Brain Mitochondria: a

Review", Molecular Neurobiology, vol. 55, no. 3, pp. 2085-2101,

2017.

[16] A. S. Ryan, J. D. Astwood, S. Gautier, C. N. Kuratko, E. B.

Nelson, and N. Salem, “Effects of long-chain polyunsaturated fatty

acid supplementation on neurodevelopment in childhood: A review

of human studies,” Prostaglandins, Leukotrienes and Essential Fatty

Acids (PLEFA), vol. 82, no. 4-6, pp. 305–314, Feb. 2010.

[17] "Official methods of analysis of AOAC International (Book,

2000) [WorldCat.org]", Worldcat.org, 2020. [Online]. Available:

https://www.worldcat.org/title/official-methods-of-analysis-of-aoac-

international/oclc/44761301. [Accessed: 17- Jul- 2020].

[18] "AACC Approved Methods of Analysis, 11th Edition - AACC

Method 44-15.02. Moisture -- Air-Oven

Methods", Methods.aaccnet.org, 2020. [Online]. Available:

http://methods.aaccnet.org/summaries/44-15-02.aspx. [Accessed: 17-

Jul- 2020].

[19] E. I. Zoulias, S. Piknis, and V. Oreopoulou, “Effect of sugar

replacement by polyols and acesulfame-K on properties of low-fat

cookies,” Journal of the Science of Food and Agriculture, vol. 80,

no. 14, pp. 2049–2056, 2000.

[20] L. Ho and N. Abdul Latif, "Nutritional composition, physical

properties, and sensory evaluation of cookies prepared from wheat

flour and pitaya (Hylocereus undatus) peel flour blends", Cogent

Food & Agriculture, vol. 2, no. 1, 2016.

[21] D. Mudgil and S. Barak, “Composition, properties and health

benefits of indigestible carbohydrate polymers as dietary fiber: A

review,” International Journal of Biological Macromolecules, vol.

61, pp. 1–6, 2013.

[22] D. Gaul and J. Issartel, "Fine motor skill proficiency in typically

developing children: On or off the maturation track?", Human

Movement Science, vol. 46, pp. 78-85, 2016.

[23] N. Čukelj, D. Novotni, H. Sarajlija, S. Drakula, B. Voučko and

D. Ćurić, "Flaxseed and multigrain mixtures in the development of

functional biscuits", LWT, vol. 86, pp. 85-92, 2017. Available:

10.1016/j.lwt.2017.07.048

[24] Joachim, B. D. (2018). Fine Cooking Issue 126, (126), 2–5.

[25] S. Kranz, Y. Marshall, A. Wight, P. Bordi and P. Kris-Etherton,

"Liking and consumption of high-fiber snacks in preschool-age

children", Food Quality and Preference, vol. 22, no. 5, pp. 486-489,

2011.

[26] P. Kaur, P. Sharma, V. Kumar, A. Panghal, J. Kaur and Y. Gat,

"Effect of addition of flaxseed flour on phytochemical,

physicochemical, nutritional, and textural properties of

cookies", Journal of the Saudi Society of Agricultural Sciences, vol.

18, no. 4, pp. 372-377, 2019.

Health Scope 79


RESEARCH ARTICLE

The relationship between physical activity level with depression, anxiety,

and stress among full-time housewives in Selangor, Malaysia

Rafidah Abdullah Hassan, Nur Islami Fahmi Mohd Teng*

Centre of Nutrition and Dietetics, Faculty of Health Sciences, Universiti Teknologi MARA Cawangan Selangor Kampus


Abstract:

Physical activity was suggested to improve mental health. However, limited study was done to examine the

relationship, especially among housewives. The aim of this study was to determine the relationship between

physical activity level with the depression, anxiety, and stress among full-time housewives. A cross-sectional

study was conducted among 297 respondents. The respondents were obtained by purposive sampling method.

Socio-demographic, DASS-21 and SQUASH questionnaires were administered to the respondents. It was

found that 50.5% of respondents were overweight. Level of MET/min of all of the respondents were reported

low. The study indicates a high prevalence of housewives who have anxiety (45.1%) followed by stress

(39.4%) and depression (34.7%). There is no relationship found between physical activity levels with mental

health. As conclusion, effort to reduce sedentary behavior together with improving mental health among

housewives is crucial..


Nur Islami Fahmi Mohd Teng


Keywords: Anxiety, depression, full-time housewife, physical activity, stress.

1. INTRODUCTION

Physical activity can prevent mental disorders [1].

Participating in physical activity was likely to improve positive

mood, enhance satisfaction, increase confidence and

encouraging good health [2]. Nowadays, people around the

world had to deal with common mental illnesses such as

depression, anxiety, and stress. Additionally, psychological

health can change the way people think, feel and act [3]. Mental

health problems have occurred in people with intense emotions

associated with other problems such as personal issues, job, or

family issues [4]. Moreover, long-term mental disorder can

lead to depression, anxiety, and other health problems such as

obesity, cholesterol, and hypertension [5].

1.1 Definition

Physical activity was referred to either fitness or exercise

[6]. Physical activity, as well as sports and other tasks were

included playing, working, rest, and performed household

chores [7]. Physical activity was necessary to improve and

maintain health [8].

Poor mental health has been described as low self-esteem,

mood disturbance, and lack of self-confidence, sleep problem

and social isolation [9]. Depression was a mood disorder in

which it can affect influence the thought, emotion, and social

lives of people [10]. Meanwhile, anxiety was an emotional state

associated with temporary changes of mental state that causes

people to experience unease, and unwanted physical symptoms

such as increased in adrenaline hormone when there was a

threat [11]. Stress was a body's nonspecific response towards

any stimulation from the stressor and not associated with nerve

damage and anxiety, as most people believed [12].

1.2 The prevalence of physical activity and mental illness

Malaysia was marked with least physically active with

more than 60 percent of the adult population engaged in

sedentary lifestyles [13]. In facts, the prevalence of adult

sedentary behaviour was high in Kedah (66.4%), followed by

Negeri Sembilan (66.3%), Kuala Lumpur (63.6%), Malacca

(62.2%), Selangor (60.1%), Sarawak (59.2%) compared to

other Malaysian states (National Health and Morbidity Survey

[14]. Additionally, physical inactivity among the Malaysian

adult population has led to global mortality by 6 percent [14].

Woman, particularly housewives, has poor physical activity

[13]. Furthermore, long-term sedentary lifestyle practices can

contribute to non-communicable diseases (NCDs), including

high blood pressure, heart disease, stroke, type 2 diabetes

mellitus, and cancer [14].

Malaysia's mental illness is rising by years. The statistics

regarding the prevalence of the mental health illness in

Malaysia were scarce [15]. Recent data presented by the

NHMS in 2015 showed that 3 out of 10 Malaysian’s adult

population above 16 years had mental problems. The

prevalence of mental illness among Malaysian adults was

increased three-fold from the year 1996 to the year 2015.

80 Health Scope, 2020, Vol. 3(3) Aireen Fatihah et al.

Furthermore, the prevalence of mental illness among

Malaysian adults was higher in women (30.8%) compared to

men (27.6%) [14].

1.3 Previous studies conducted on the relationship between

physical activity with the depression, anxiety and stress

Previous studies show that physical activity was associated

with lower psychological distress [16]. In addition, higher

levels of physical activity were associated with higher

psychological well-being [17]. Also, recent studies suggested

that sitting time has directly associated with high level of

psychological distress [18]. Other studies also found that

physical activity has positive effects on biological, as well as

mental [19]. Moreover, other recent studies stated that there

was strong and positive relationship between physical activities

with mental health [20]. Besides that, according the researcher

have found that mental disorders have been associated with

sedentary activity and increased physical activity facilitates

better mood disturbances [9]. Furthermore, mental problems

have significant associated with sedentary behaviour and light

physical activity [21]. However, physical activity has been

reported negative association with mental illness [22-25].

Therefore, there is a need to study the relationship between

mental health and physical activity among women, as the

prevalence of mental illness was reported higher among them.

As there is limited study being done, particularly among

Malaysian, this study will provide insight on the

implementation for future study.


2.1 Study design

A cross-sectional study was carried out among the 297

respondents which focused on full-time housewives in

Selangor, Malaysia. This study was a quantitative study and

carried out around certain residential areas, recreational parks,

and shopping malls in Selangor State.

The respondents were obtained by purposive sampling method.

The respondents were assessed using a validated questionnaire

in order to measure the level of depression, anxiety, stress and

also physical activity. The questionnaires that have been used

included Socio-demographic, Depression, Anxiety and Stress

Scale-21 Items (DASS-21) and Short Questionnaire to Assess

Health Enhancing Physical Activity (SQUASH).

2.2 Inclusion and exclusion criteria

The respondents were selected based on inclusion criteria,

which were includes married women, full-time housewife,

aged 18 to 65 years, residents of Selangor, and able to read and

write in either Malay or English. This study excluded

respondents who have a mental health problem and physical

disability.

2.3 Procedure

The data collection was begun once the ethical application

has been approved (June 2019). The data was collected around

the residential area, recreational parks and shopping malls.

There were three self-administered questionnaires that

included socio-demographic, DASS-21 and SQUASH. All

three questionnaires have been distributed to respondents

through direct and indirect by posting in social media platforms.

The questionnaires were distributed directly to those

respondents who fulfill the inclusion criteria and agree to

participate in this study. Then, the respondents were given a

consent form and a set of questionnaires (socio-demographic,

DASS-21 and SQUASH) after they agreed to participate in this

study. Furthermore, the respondents were informed about the

intent of this study as well as instruction of questionnaires. It

took ten to fifteen minutes for respondents to complete all these

questionnaires.

Besides that, indirect distribution of questionnaires was done through the social networking such as Facebook, Telegram, WhatsApp, and Twitter. The statement and criteria of inclusion and exclusion were clearly stated. The websites link (questionnaires) was given to the respondents who met the inclusion criteria and willing to participate in this study. The directions for each questionnaire that were presented in layman languages, clearly and easily understood. In addition, the respondents were given the contact number of the researcher to ask any queries of the questionnaires.

2.4 Ethics

The legal approval for this study has been applied from the Research Ethics Committee of Universiti Teknologi MARA (UiTM).

2.5 Analysis

The data was evaluated using Statistical Package for Social Sciences (SPSS) versions 21.0. The descriptive statistic was used to summarize the data by finding the measures of the central tendency (median) and finding the measures of spread. The median and IQR was used instead of the mean and standard deviation since the data were not normally distributed. Meanwhile, the Spearman Correlation Test was used to find the relationship between the physical activity level with depression, anxiety and stress among full-time housewives. The Spearman Correlation Test was used instead of Pearson Correlation Test because of the data was not normally distributed. Furthermore, the Pearson Chi-square Test was also used in this study to observe the association between factors from socio-demographic and health with depression, anxiety and stress among full-time housewives.


Table 1. Socio-demographic and health background of the

respondents (n=297)

Characteristic n (%)

Health Scope 81

SOCIO-DEMOGRAPHIC BACKGROUND

Age, years (Mean ±SD) 39.6 ± 10.3

Age (Min – Max) 20-65

BMI classification

Underweight 21 (7.1) Normal 126 (42.4)

Overweight 99 (33.3)

Obese class I 44 (14.8)

Obese class II 7 (2.4) Race

Malay 297 (100)

Educational level

No formal education 9 (3.0)

Primary education 5 (1.7)

Secondary education 92 (31.0)

STPM/Matriculation/Diploma 98 (33.0)

Bachelor 93 (31.3)

Side job

Has side job 83 (27.9)

No side job 214 (72.1)

Estimated monthly income

No income 197 (66.3)

Income less than RM 3900 81 (27.3)

Income between RM 3900 to RM8299 19 (6.4)

Marriage status

Married

Single mother

286 (96.3)

11 (3.7) Marriage age, years (Mean ±SD) 14.84 ± 10.315

Husband’s age, years (Mean ±SD) 42.44 ± 11.527

Husband’s educational level

No formal education 7 (2.3)

Primary education 13 (4.4)

Secondary education 92 (31.0)

STPM/Matriculation/Diploma 85 (28.6)

Bachelor 100 (33.7)

Husband’s working status

Retire 26 (8.7)

Not working 5 (1.7)

Working 266 (89.6)

Husband’s working sector

None/retired 26 (8.8)

Private 182 (61.2)

Government 89 (30.0)

Husband’s side job

Has side job 45 (15.2)

No side job 252 (84.8)

Estimated husband’s monthly income

No income 4 (1.3)



Income more than RM8300 52 (17.5)

Estimated household income



Income more than RM8300 53 (17.8)

HEALTH BACKGROUND Pregnancy status

Pregnant 28 (9.4)

Not pregnant 269 (90.6)

Breastfeeding status

Currently breastfeed 63 (21.2)

Not breastfeed 234 (78.8)

Menopause

Menopause 41 (13.8)

Not menopause 256 (86.2)

Chronic disease state

Yes 29 (9.8)

No 268 (90.2)

Musculoskeletal disorder

Yes 28 (9.4)

No 269 (90.6)

Prolonged depression

Experience depression 6 (2.0)

Not experience depression 291 (98.0)

Incident of concussion

Experience concussion 7 (2.4)

Not experience concussion 290 (97.6)

Smoking status

Smoking 3 (1.0)

Not smoking 294 (99.0)

Alcoholic status

Drinking alcohol 14 (4.7)

Not drinking alcohol 283 (95.3)

Data were analyzed using descriptive statistics

Table 2. The prevalence of physical activity among

housewives (n=297)

Median (IQR)a n (%)

Commuting Activities

(Total min/week) 120.00 (270)

Short duration 149 (50.2)

Long duration 148 (49.8) (MET-minutes) 240.00 (472)

High intensity 148 (49.8) Low intensity 149 (50.2)

Leisure Activities

(Total min/week) 120.00 (330) Short duration 160 (53.9)

Long duration 137 (46.1)

(MET-minutes) 380.00 (1050)

High intensity 150 (50.5)

Low intensity 147 (49.5)

Household Activities

(Total min/week) 1200.00 (1380)

Short duration 149 (50.2

Long duration 148 (49.8)

(MET-minutes) 2940.00 (3000)

High intensity 154 (51.9)

Low intensity 143 (48.1) a Data was not normally distributed


Table 3. The prevalence of depression, anxiety and stress

among housewives (n=297) Variables Median (IQR)a n (%)

Depression 6 (9) 103 (34.7)

Anxiety 6 (11) 134 (45.1)

Stress 8 (12) 117 (39.4) a Data was not normally distributed



Figure 1. Prevalence distribution of depression, anxiety and

stress

Table 4. Correlation between Total MET-Minutes (total

intensity of physical activity per week) with depression,

anxiety, and stress among housewives (n=297)

*Data were analyzed using Spearman’s Correlation test

*Correlation significant at (p <0.05) and (r >0.26)

Table 5. Association between socio-demographic data with

depression among housewives (n=297) Normal Depression x2

(df)

p value

n (%) n (%)

Side job

Side job (n=83) 43 (22.2)

40 (38.8)

1

*0.002

No side job (n=214) 151

(77.8)

63

(61.2) Husband working sector

Private sector (n=208) 144

(74.2)

64

(62.1)

1

*0.030

Government sector (n=89) 50 (25.8)

39 (37.9)

a Fisher’s Exact Test was used when the expected count of <5 is more than 20%

*p<0.05 indicates significant difference

*Data was analyzed by using Pearson Chi-square Test

Table 6. Association between health background with

depression among housewives (n=297)

Normal Depression x2

(df)

p

value n (%) n (%)

Menopause status Has menopause (n=41) 33 (17.0) 8 (7.8)

1

*0.028 Not menopause (n=256) 161 (83.0) 95 (92.2) a Fisher’s Exact Test was used when the expected count of <5 is more than 20%

*p<0.05 indicate significant difference

Data was analyzed by using Pearson Chi-square Test

Table 7. Association between health background with

anxiety among housewives (n=297)

Normal Anxiety x2

(df) p value

n (%) n (%)

Experienced prolonged depression

Experiences (n=6) 0 (0) 6 (4.5)

1

0.008a Not experiences

(n=291)

163 (100) 128 (95.5)

aFisher’s Exact Test was used when the expected count of <5 is more than 20%

*p<0.05 indicate significant difference

Data was analyzed by using Pearson Chi-square Test

3.1 The demographic and health background

Table 1 shows the demographic and health background of

housewives (n=297) who participated in this study. The

demographic background has been assessed via socio-

demographic questionnaires. The results show the mean age for

housewives was 39.6 years old. Nearly half of the respondents

were also listed as normal BMI (42.4%). Meanwhile, about

33.3 percent of respondents were classified as overweight,

followed by 14.8 percent of obese class I, 2.4 percent of obese

class II and 7.1 percent of underweight BMI.

All the respondents in this study were Malay and Islam.

Approximately 31.3 percent of respondents have a bachelor’s

degree and 33 percent have a tertiary education (STPM /

Matriculation / Diploma). A total of 31 percent of respondents

have secondary education (PMR and SPM). Meanwhile, 1.7

percent of respondents have primary education, and 3.0 percent

have no formal education.

This study shows that most respondents have no side job (not

worked) (72.1%) compared to have a side job (worked)

(27.9%). Furthermore, more than half of the respondents have

no income (66.3%) as many of them did not work.

Approximately 27.3 percent of respondents who worked have

an income less than RM 3900. However, 6.4 percent of

respondents who worked have an income between RM 3900

and RM 8299.

Next, about 96.3 percent of respondents who were participated

in this study were marriage. The mean of the respondents’

marriage age was 14.8. The mean of the husbands’ age was

42.4. About 33.7 percent of the respondents’ husbands have a

bachelor's degree followed by have a secondary education

(31.0%), tertiary education either STPM or Matriculation or

Diploma (28.6%), primary education (4.4%) and no formal

education (2.3%).

Based on the results, most of the respondent’s husbands have

currently been working (89.6%), followed by retired (8.7%)

and not working (1.7%). Also, more than half of the

respondent's husbands were working for the private sector

(61.2%) followed by the government sector (30.0%) and none

or retired (8.8%). Moreover, the majority of the respondents’

husbands have no side job (84.8%); meanwhile only 15.2

percent have a side job. In addition, almost half of the

respondents' husbands have an estimated monthly income

between RM 3900 to RM 8299 (42.8%) followed by an income

less than RM 3900 (38.4%), an income above RM 8300 (17.5%)

and no an income (1.3%). Lastly, almost half of the

respondents' whose husbands have an estimated household

income between RM 3900 to RM 8299 (42.3%) followed by

RM 3900 (40.0 %) and an income above RM 8300 (17.3 %).

Health information was obtained via socio-demographic

questionnaires. The study shows that most of the respondents

were currently not pregnant (90.6%), currently not

breastfeeding their children (78.8%), and most of them still not

have menopause (86.2%). In addition, most of the respondents

DASS-21 Item Coefficient (r) P value

Depression -0.067 0.252 Anxiety -0.073 0.208

Stress 0.040 0.491

Health Scope 83

have a good health condition which have no chronic disease

(90.2%) and musculoskeletal disorders (90.6%). However,

some of the respondents had chronic diseases such as

hypertension and diabetes (9.8%). Also, only 9.4 percent of

respondents have experienced musculoskeletal disorders such

as back pain. The study also shows that the majority of

respondents had no experienced of prolonged depression

(98.0%), and no experienced an incident of concussion (97.6%).

Nevertheless, only 2.0 percent have experienced of prolonged

depression and 2.4 percent of the respondents have a history of

concussion. Most of the respondents were not smoking (99.0%)

and not taking alcohol (95.3%). However, only 1.0 percent of

the respondents were smoking and 4.7 percent of them were

drinking alcohol.

3.2 The prevalence of physical activity among housewives

Table 2 indicates the prevalence of physical activity among

housewives (n=297). The information on physical activity was

obtained via SQUASH questionnaires. The physical activity

was divided into three domains which were commuting

activities, leisure activities, and household activities. The

length of time spent on activities (total minutes/week) and

physical intensity (MET-minutes) for commuting leisure and

household activities was assessed. The median and IQR were

used instead of the mean and standard deviation (SD) due to

the data were not normally distributed. The median and IQR of

the total minutes/week for commuting, leisure, and household

activities were 120.00 (270), 120.00 (330), and 1200.00 (1380)

respectively. Meanwhile, the median and IQR of the MET-

minutes for commuting, leisure, and household activities were

240.00 (472), 380.00 (1050), and 2940.00 (3000) respectively.

The results show that half of the respondents were engaged in

short duration of the commuting (50.2%), leisure (53.9%) and

household activities (50.2%). In addition, nearly half of the

respondents were engaged in the high intensity of MET-

minutes for commuting (49.8%), half of them were engaged in

the high intensity of MET-minutes of leisure (50.5%) and

household task (51.9%). However, the respondents were

considered to have low MET-minutes in commuting, leisure,

and household activities, although half of the respondents have

high intensity of MET-minutes. It because of the average of

short duration of total min/week was reported (housewives

spent more time on a short duration compared to long duration

of total minutes/week).

The present study found that full-time housewives have low

physical activity in which half of the respondents have engaged

in a short duration of commuting, leisure, and household

activities. This finding suggests that the respondents spent less

time on commuting, leisure and household activities such as

walking, cycling, gardening, jogging, sports, and house chores.

This finding was consistent with previous studies whereby

housewives were associated with less physical activity [13].

Furthermore, other studies found that housewives tended to be

involved in light-to-moderate intensity of physical activity such

as housekeeping, home grocery, household items, and taking

care of children [26]. There were several factors could

influence the involvement of housewives in physical activity

which were included cultural barriers, limited transportation,

and lack of time due to the household workload and duties to

care of other family member [7].

3.3 The prevalence of depression, anxiety and stress among

housewives

The information on depression, anxiety, and stress was

obtained via DASS-21 questionnaires. The level of depression,

anxiety and stress was measured. Table 3 represents the

prevalence of depression, anxiety and stress among housewives

(n=297). The level of depression, anxiety and stress was

measured using the median and IQR due to the data were not

normally distributed. This study shows that the median score

and IQR of depression was 6 (9), anxiety; 6 (11) and stress; 8

(12), respectively. The results show a high percentage of

housewives who have anxiety (45.1%) followed by stress

(39.4%) and depression (34.7%).

Figure 1 shows the prevalence of depression, anxiety and stress

among housewives (n=297) which categorized based on the

severity. The severities were further divided into normal, mild,

moderate, severe and extremely severe. Based on this study,

more than half of the respondents were classified as normal.

Nonetheless, some of the respondents reported either have mild,

moderate, severe and extremely severe. About 25.9 percent of

the respondents have mild depression, moderate depression

(9.4%), severe depression (1.7%) and extremely severe (2.4%).

Besides that, about 7.4 percent of the respondents have mild

depression, moderate depression (18.9%), severe depression

(10.1%) and extremely severe (8.8%). Meanwhile, 14.5 percent

of the respondents have mild depression, moderate depression

(12.8%), severe depression (4.0%) and extremely severe

(3.4%).

The results show a high percentage of housewives who have

anxiety (45.1%) followed by stress (39.4%) and depression

(34.7%) among full-time housewives. This finding was

supported previous studies whereby full-time housewives

reported more anxiety than working housewives [27]. Also,

other studies have found that full-time housewives were highly

anxious and poorly satisfied with their lives [28]. Furthermore,

the full-time housewives who lived within the nuclear family

were more anxious than lived in the joint family [29]. Besides

that, this study also shows the prevalence of stress was quite

high among full-time housewives which indicate their life were

stressful. This result was consistent with previous studies

whereby full-time housewives have perceived more stress than

working housewives [6]. Moreover, this finding indicates that

full-time housewives perceived less depression because of

have ample time to enjoy their life and have a good relationship

with family compared to working housewives which consistent

with previous studies [30].

3.4 The relationship between Total MET-Minutes with

depression, anxiety, and stress among housewives

Table 4 demonstrates correlation between Total MET-

Minutes (total intensity of physical activity per week) with

depression, anxiety, and stress among housewives (n=297).

Spearman’s Correlation Test was used instead of the Pearson

Correlation Test due to the data was not normally distributed.

The results show that there was no significant (p>0.05)

between total MET-minutes of commuting, leisure and

household activities with depression, anxiety and stress. This

study also shows there was no correlation between all variables

(r < 0.25).


Based on the results, there was no significant correlation

between MET-minutes of commuting, leisure and household

activities with depression, anxiety and stress. This study was

particularly in some studies whereby a negative association of

physical activity with depression, anxiety disorders has been

found [22-25]. The effect of physical intensity toward the level

of depression, anxiety, and stress was not presented in this

study because of poor measurements of psychological stress,

and physical activity (duration and intensity). Also, inaccurate

information on physical activity either over-reported or under-

reported could happen due to housewives tend to self-estimate

the duration and intensity in the SQUASH questionnaires. Thus,

these factors were leads to inconsistent findings to recent

studies which found that engaged more in physical activity

drives to a better physical and psychological status in which

contributed to the positive association between physical and

mental health [31].

3.5 Association between socio-demographic data and

health background with depression, anxiety, and stress

among housewives

Pearson Chi-square Test was used since the data was not

normally distributed while Fisher’s Exact Test was used when

the expected count of <5 and more than 20%. Total depression

was further categorized into normal (0 to 9 scores of depression)

and has depression (10 to 42 scores of depression). In addition,

total anxiety also further categorized into normal (0 to 6 scores

of anxiety) and has anxiety (7 to 42 scores of anxiety).

Meanwhile, total stress was categorized into normal (0 to10

scores of stress) and has stress (11 to 42 scores of stress).

Table 5 shows the association between socio-demographic data

with depression among housewives (n=297). The results show

that there was a significant association between have a side job

(p=0.002), and the husband’s working sector (p=0.030). Table

6 shows the association between health background with

depression among housewives (n=297). The results show that

there was a significant association between menopause status

(p=0.028) with depression. Table 7 shows the association

between health background with anxiety among housewives

(n=297). The results show there was a significant association

between experienced prolonged depression (p=0.008) with

anxiety. Meanwhile, other factors in the socio-demographic

and health background that have not been mentioned had no

significant association with depression, anxiety, and stress.

Almost half of the respondents who worked (have a side job)

were reported experiencing depression. This finding was

supported by other researchers whereby working housewives

have experienced greater stress and depression than full-time

housewives due to face greater challenges in their lives [32]. In

addition, working housewives were reported may have

mentally exhausted throughout the day due to the challenges

during additional tasks and house chores effort [33]. Moreover,

there was a significant association between the husband’s

working sectors with depression. This study indicates that

housewives whose husbands worked with the government

sector have more depression compared to the private sector.

This finding can be linked with financial standing [34]. Hence,

this study indicates that housewives whose husbands worked

with the private sector has less depression may due to good

financial standing compared to the government sector.

However, the associations between government and the private

working sector with mental problems were not yet to be

addressed by scholars in Malaysia.

The study shows that there was a significant association

between menopauses with depression. Less than half of the

respondents who non-menopausal but reported depressed and

also some respondents who have menopausal were reported

depressed. The evidence has shown that the incidence of

depression in women usually rose due to hormone fluctuation

for instances estrogen and progesterone [35]. This finding was

supported by other researcher an increased prevalence of

depression correlates with hormonal changes in women,

especially during puberty, before menstruation, after

pregnancy, and during pre-menopause [36]. The researcher

further explained that changes in ovarian hormones and decline

in estrogen level cause women to experience specific forms of

depression-related disease, including pre-menstrual dysphoric

disorder, post-partum depression and post-menopausal

depression and anxiety [36].

In addition, there was a significant association between

experienced prolonged depressions with anxiety. Previous

studies suggest that anxiety disorders have been found more

prevalent among individuals with a family history of major

depressive disorder, disturbed family environment, childhood

sexual abuse, low self-esteem, and low-educated [37]. This

finding was also supported by other researchers whereby past

diagnoses of psychological disorder, previously used

psychotherapy as well as excessive alcohol, smoking, drug

abuse, lower socioeconomic status, and gastrointestinal

disorders were associated with anxiety [38].

The present finding shows different results compared to other

studies whereby depression, anxiety and stress were not

affected by the housewife’s education level, marriage status,

estimated monthly income, husband education level, husband

working status, husband's side job, husband estimated monthly

income. In contrast, previous studies in Malaysia have found

the socio-demographic differences in depression [39-42]. The

findings from these studies were difficult to interpret due to the

lack of different samples in terms of age groups and

geographical areas which consistent from previous study

conducted [34]. However, this finding was vary with previous

studies whereby most of the respondents have a good health

background which many of the respondents were have not

chronic disease, no musculoskeletal disorder, no history of

concussion, and also not taking alcohol and cigarette.

3.6 Limitation & suggestion for future research

This study was lacking the respondents from other Malay

ethnics since all the respondents were Malay. Besides that, the

Health Scope 85

challenge throughout the study was place barrier. It is due to

the full-time housewives were quite hard to find other than at

recreational places and shopping malls. Lastly, the respondents

may over-report the duration of physical activity which affects

the findings of this study. It may due to the respondents to

misunderstand questionnaire (SQUASH), and there is no

opportunity to get clarification especially for indirect

distribution through social networking.

It is recommended to recruit the respondents from urban and

rural areas to observe the difference between physical activity

and the level of depression anxiety and stress. Moreover, it is

suggested to conduct a longitudinal study instead of cross-

sectional to gain a better sense of physical activity with the

level of depression, anxiety and stress among housewife by

observing the same respondents over a time. Lastly, it also

recommended to future researcher to conduct an intervention

study regarding the effect of physical intensity on depression,

anxiety and stress among full-time housewives.

4. CONCLUSION

Physical activity benefit both physical and mental health.

The present study suggests that that full-time housewives have

low MET-minutes of commuting, leisure and household

activities. This study also found that higher prevalence of the

housewives who have anxiety followed by stress and

depression. Additionally, this study suggests that there was no

significant correlation between MET-minutes of commuting,

leisure and household activities with depression, anxiety and

stress among full-time housewives. Besides that, this study

found some factors of socio-demographic and health

background has been associated with mental health such as side

job, husband working sector, menopause and experienced

prolonged depression. It is best for future research to develop

intervention of specific physical activity that can help to

improve mental health among this population.

ACKNOWLEDGEMENTS

I was appreciated and wish to thank all of my

respondents for their contribution and willingness to

be part of my subject for this study.

REFERENCES

[1] P. Ekkekakis, “Routledge Handbook of Physical Activity and

Mental Health.,” 1st ed., D. B. Cook, L. L. C. S. N. Culos-

Reed, P. E. J. L. E. M. Hamer, K. A. M. G. J. Reed, and J.

A. J. S. M. Ussher, Eds. New York: Routledge, 2013.

[2] S. Biddle, “Physical Activity and Mental Health: Evidence is

Growing,” pp. 176–177, 2016.

[3] Mental Illness Awareness & Support Association (MIASA),

“Frequent ask question (FAQ). What is Mental Health?,”

2018. [Online]. Available: https://miasa.org.my/faq.html.

[4] B. H. Peterson, “What is Mental Health?,” J. Christ. Educ.,

2015.

[5] N. Richardson, “The Effects of Physical Activity On Perceived

Stress, Anxiety And Life Satisfaction,” 2014.

[6] K. Singh, “a Study of Physicalactivity, Exercise, and Physical

Fitness: Definitions and Bifurcation for Physical Related

Research.,” Acad. Sport. Sch., vol. II, no. III, pp. 1–5, 2013.

[7] World Health Organization (WHO), “Physical Activity,” 2018.

[8] E. Al-Eisa, S. Buragadda, and G. R. Melam, “Association

between physical activity and psychological status among

Saudi female students,” BMC Psychiatry, 2014.

[9] National Institute of Mental Health (NIMH), “Transforming the

Understanding and Treatment of Mental Illnesses.,”

2019.[Online].Available:

https://www.nimh.nih.gov/index.shtml.

[10] P. Swift, E. Cyhlarova, I. Goldie, and C. O’Sullivan, “Living

with Anxiety: Understanding the role and impact of anxiety

in our lives,” Ment. Heal. Aware. Week, pp. 1–48, 2014.

[11] H. Selye, “Stress without Distress,” Psychosom. Med., vol. 38,

no. 1, pp. 72–73, 1976.

[12] T. C. Lian, G. Bonn, Y. S. Han, Y. C. Choo, and W. C. Piau,

“Physical activity and its correlates among adults in

Malaysia: A cross-sectional descriptive study,” PLoS One,

vol. 11, no. 6, pp. 1–14, 2016.

[13] National Health and Morbidity Survey (NHMS), “Non-

Communicable Disease, Risk Factors & Other Health

Problems Volume II: Mental Health Problem In Adults.,” N.

A. Tahir Aris, Muhammad Fadhli Mohd Yusoff, Abdul

Aiman Abd Ghani, N. Ahmad, Mohd Azahadi Omar, Tee

Guat Hiong, Nur Hazwani Mohd Hasri, and N. A. K.

Fadzilla Mohd Radzi, Nur Syazwani Manan, Eds. 2015.

[14] Abdul Kadir Abu Bakar, “Malaysian Mental Healthcare

Performance,” Malaysian J. Psychiatry, vol. 20, no.

Editorial, pp. 1–3, 2016.

[15] C. Gm et al., “Exercise for depression ( Review ) Summary of

findings for the main comparison,” no. 9, 2013.

[16] C. Kingsbury, P. Bernard, I. Doré, G. Hains-Monfette, C.

Sabiston, and A.-J. Romain, “Dose response association of

objective physical activity with mental health in a

representative national sample of adults: A cross-sectional

study,” PLoS One, vol. 13, no. 10, p. e0204682, 2018.

[17] A. L. Rebar, C. Vandelanotte, J. Van Uffelen, C. Short, and M.

J. Duncan, “Associations of overall sitting time and sitting

time in different contexts with depression, anxiety, and

stress symptoms,” Ment. Health Phys. Act., vol. 7, no. 2, pp.

105–110, 2014.

[18] E. Anderson and G. Shivakumar, “Effects of Exercise and

Physical Activity on Anxiety,” Front. Psychiatry, vol. 4, no.

April, pp. 10–13, 2013.

[19] M. A. Harris, “The relationship between physical inactivity and

mental wellbeing: Findings from a gamification-based

community-wide physical activity intervention,” Heal.

Psychol. Open, vol. 5, no. 1, 2018.

[20] P. Bernard et al., “Effects of a six-month walking intervention

on depression in inactive post-menopausal women: A

randomized controlled trial,” Aging Ment. Heal., vol. 19,

no. 6, pp. 485–492, 2015.

[21] M. Gerber and U. Pühse, “Review Article: Do exercise and

fitness protect against stress-induced health complaints? A

review of the literature,” Scand. J. Public Health, vol. 37,

no. 8, pp. 801–819, 2009.

[22] S. Tufik, L. Bittencourt, H. K. M. Antunes, M. T. De Mello, V.

de A. Lemos, and R. Santos-Silva, “Relationship between

physical activity and depression and anxiety symptoms: A

population study,” J. Affect. Disord., vol. 149, no. 1–3, pp.

241–246, 2013.

[23] J. L. Durstine, B. Gordon, Z. Wang, and X. Luo, “Chronic

disease and the link to physical activity,” J. Sport Heal. Sci.,

vol. 2, no. 1, pp. 3–11, 2013.

[24] N. J. Hegberg and E. B. Tone, “Physical activity and stress

resilience: Considering those at-risk for developing mental

health problems,” Ment. Health Phys. Act., vol. 8, pp. 1–7,

2015.

[25] Y. Y. Chan et al., “Lifestyle, chronic diseases and self-rated

health among Malaysian adults: results from the 2011

https://miasa.org.my/faq.html

https://www.nimh.nih.gov/index.shtml


National Health and Morbidity Survey (NHMS),” BMC

Public Health, 2015.

[26] A. Rahman, et al., “Anxiety and coping strategies of married

women”. Biological Science, 2-10, 2017.

[27] N. Kaur, N. Panwar, H. . Thind, and M. . Farooqi, “A

comparative study of working and non-working married

women: Effect of Anxiety level on life satisfaction.,” Indian

J. Psychol. Ment. Heal., 2012.

[28] A. Iqbal, R. Nadeem, and N. Fatima, “Anxiety in non-working

women with reference to their education, family system and

number of children,” Pakistan J. Med. Sci., 2004.

[29] J. S. Hyde, “Women, Men, Work, and Family: Expansionist

Theory Updated,” 2016.

[30] J. Ohrnberger, E. Fichera, and M. Sutton, “The relationship

between physical and mental health: A mediation analysis,”

Soc. Sci. Med., vol. 195, no. October, pp. 42–49, 2017.

[31] H. A. Hashmi, M. Khurshid, and I. Hassan, “Marital

Adjustment, Stress and Depression among Working and

Non-Working Married Women,” Internet J. Med. Updat. -

EJOURNAL, 2007.

[32] S. M. Dibaji, S. H. Reza Oreyzi, and M. R. Abedi, “Occupation

or Home: Comparison Housewives and Working Women in

the Variables of Stress, Depression and Perception of

Quantitative, Mental and Emotional Home Demands,” Rev.

Eur. Stud., 2017.

[33] S. H. Yeoh, C. L. Tam, C. P. Wong, and G. Bonn, “Examining

depressive symptoms and their predictors in Malaysia:

Stress, locus of control, and occupation,” Front. Psychol.,

vol. 8, no. AUG, pp. 1–10, 2017.

[34] M. Smith and J. Jaffe, Depression in Women, vol. 15, no. 10.

Elsevier Inc., 2011.

[35] P. R. Albert, “Why is depression more prevalent in women?,”

Journal of Psychiatry and Neuroscience. 2015.

[36] C. Blanco, J. Rubio, M. Wall, S. Wang, C. J. Jiu, and K. S.

Kendler, “Risk factors for anxiety disorders: Common and

specific effects in a national sample,” Depress. Anxiety,

2014.

[37] E. Martín-Merino, A. Ruigómez, S. Johansson, M. A.

Wallander, and L. A. García-Rodriguez, “Study of a cohort

of patients newly diagnosed with depression in general

practice: Prevalence, incidence, comorbidity, and treatment

patterns,” Prim. Care Companion J. Clin. Psychiatry, 2010.

[38] S. F. Kader Maideen, S. Mohd Sidik, L. Rampal, and F.

Mukhtar, “Prevalence, associated factors and predictors of

depression among adults in the community of Selangor,

Malaysia,” PLoS One, 2014.

[39] J. Kaur et al., “Prevalence and correlates of depression among

adolescents in Malaysia.,” Asia. Pac. J. Public Health, 2014.

[40] A. Rashid and I. Tahir, “The Prevalence and Predictors of

Severe Depression Among the Elderly in Malaysia,” J.

Cross. Cult. Gerontol., 2014.

[41] K. L. Tan and H. Yadav, “Depression among the urban poor in

Peninsular Malaysia: A community based cross-sectional

study,” J. Health Psychol., 2013.

[42] P. Ekkekakis. Routledge Handbook of Physical Activity and

Mental Health. Pp. 2-26. ISBN: 978-0-415-78299-9.

Health Scope 87


RESEARCH ARTICLE

Evaluation on the knowledge, attitude, and practices toward evidence-

based practice in computed tomography among radiographers

Nur Ezzatie Aqillah Mohd Sauti , Lyana Shahirah Mohamad Yamin*

Centre of Medical Imaging, Faculty of Health Sciences, Universiti Teknologi MARA Cawangan Selangor Kampus


Abstract:

The study aims to identify the level of knowledge, attitude, and practices toward evidence-based practice

(EBP) in computed tomography (CT) among radiographers and factors that influence the use of EBP in CT

among the radiographers. This is a cross-sectional study to evaluate the knowledge, attitude, and practice

toward evidence-based practice in computed tomography among radiographers. The research was done by

using a self-administered questionnaire covering socio-demographic features, knowledge on EBP,

understanding of specific terms related to EBP, access and availability toward literature, attitudes and

practices, and also barriers perceived toward EBP in CT that had been distributed to 30 radiographers in

Hospital Muar. All the data obtained were analyzed using SPSS software version 21. The level of

knowledge toward EBP in CT among radiographers was high, where 63.3% of the respondents had high

knowledge, and only 36.7% of the respondents had a low level of knowledge toward EBP in CT. The

radiographer also had a high level of attitude and practices by 56.7%, and 43.3% of the respondents had a

low level of attitude and practices toward EBP in CT. The barriers that affect the implementation of EBP

among radiographers involve in CT was the inability to critically appraise literature. The lack of information

resources and insufficient time also become the most factors that influence the use of EBP in CT. In

conclusion, from the findings, most of the radiographers had high a level of knowledge, attitude, and

practices toward evidence-based practice in CT and understood most of the specific terms related to EBP.


Lyana Shahirah Mohamad Yamin


Keywords: Attitude, computed tomography, evidence-based practice, knowledge, practices

1. INTRODUCTION

Evidence-based practice (EBP) is an approach that

organizes, which can be used in clinical decision-making

that involves the best current evidence accessible from

research and clinical skill [16]. It becomes a wide-reaching

concern for numerous professional bodies and is considered

as a basis for health care quality. As stated by Eid AbuRuz et

al. [5], EBP serves as a method to solve the problems based

on research evidence while considering the patient’s values

and clinical experience. In radiology, there are rapidly

increasing technological advancements that need the

radiographer to provide service using currently available

research evidence to get the best decision that benefits the

patients and prevent harm [14].

Computed Tomography (CT) has been known as the

revolutionary diagnostic imaging tool since its introduction.

It produces sectional digital imaging techniques that produce

images different from the conventional radiographic

technique in several aspects. The images produce referred to

as transaxial images because it refers to planar sections that

perpendicular to the long axis of the patient. CT uses a

computer to collect, process, and reconstruct the data

transmission from the patients’ body. It helps to produce the

images by eliminating the superimposition structures,

improves image contrast, and detects difference tissue

contrast less than 10% [15].

Computed Tomography is widely used and become a

reference standard for some emergency diagnosis. The rise in

the use of CT has its effect as the ionizing radiation from CT

scans is associated with an increased risk of subsequent

malignancies. This is important to appropriately identified

injured patient who really need the CT scan and who does

not need. A study conducted by Halaweish et al. [10] on the

patient with head and abdominal trauma. The result shows

that there is a decreased number of patients need to undergo

CT after the evidence-based guideline is used. This help to

reduce the number of patients being exposed to unnecessary

radiation. With respect to population health, a sensitive

clinical decision is used that can greatly reduce unnecessary

risks and costs. Evidence-based practice is constructed to

reduce those potential harms that occurred in CT. It helps to

prevent overdiagnosis and overtreatment as identifying

incidental findings or other findings of unknown clinical

88 Health Scope, 2020, Vol. 3(3) Nur Ezzatie Aqillah et al.

significance [13]. This will result in costly follow-up

procedures that lead to additional radiation and potential

morbidity from invasive testing.

Computed Tomography has shown its progressive year by year technological advances, thus increase the widespread uptake of CT in clinical practice. Evidence-based practice understanding, interpretation, and application of the best current evidence into radiology practice that optimizes patient care. Lavelle et al. [12] also stated that the evidence-based practice (EBP) is widely accepted as a standard that supports robust decision making. Radiographers indeed generate and consequently use evidence in their daily service delivery, but most of them are not routinely using EBP. In radiography, actively pursuing new evidence within the fields can decrease the possibility of clinical practice become out of date and be able to provide high-quality services for health care users [9].


2.1 Study Design

This is a cross-sectional study design to evaluate the

knowledge, attitude, and practices of radiographers toward

EBP in CT. A survey questionnaire will be given to 30

diagnostic radiographers working in Hospital Pakar Sultanah

Fatimah, Muar Johor, who agree to take part in this study.

The questionnaire was adapted from the article by Antwi

WK [1] and C. Costa et al. [3]. A pilot was conducted

among 15 radiographers that involve in CT to gather

information about the study’s validity and reliability of the

questionnaire. Half of the sample size was chosen to undergo

the pilot study. A set of the questionnaire has been given to

the radiographer working there by google form.

2.2 Data Collection

The questionnaire designed was divided into two parts.

The first part consists of a demographic variable on

radiographers’ age, gender, qualification, and years of

working and involve in CT. The second part includes 29

items that can be divided into four sections, which is to

access the knowledge level of a radiographer, their

understanding of specific terms related to EBP, attitudes and

practices, and also barriers perceived toward EBP in CT.

In this study, 5 Likert-scale will be used with 1 = strongly

disagree, 2 = disagree, 3 = neutral, 4 = agree and 5 =

strongly agree for answers. For each of the response items,

the response of “3 and above” is considered to be high

knowledge, attitude and practice toward EBP while “less

than 3” were considered to have low knowledge, attitude,

and practice. Several items related to access and availability

toward literature and to understand specific terms related to

EBP required categorical response (‘YES/NO’). A score of 1

was given for each yes that indicate correct or ‘I know’

answer and for no that indicates incorrect answer or ‘I don’t

know’ answer was scored with zero.

2.3 Data Analysis

All data were analyzed using the Statistical Package for

the Social Sciences (SPSS) software version 21. Descriptive

analysis for all data such as demographics, knowledge,

attitude and practices, and scores from the barrier’s

questionnaires. The data were then analyzed using

descriptive statistics. For evaluation of the differences

between radiographer’s level of knowledge, attitude and

practices toward EBP in CT based on their years working in

CT and qualification of study, the ANOVA test was used.

Post-hoc Anova test was carried out to determine the

differences between radiographer’s number of years working

in CT and the level of their knowledge toward EBP in CT if

there is a differences.


3.1 Demographic Data Analysis

Table 1: Data of Socio-demographic of participants Demographic Data Frequency Percentage (%)

Gender Male

Female

12

17

40.0

56.7

Age 20-29

30-39

40-49 >50

20

7

3 0

66.7

23.3

10.0 0

Years of Working <5 years 5-10 years

11-15 years >15 years

13 12

2 3

43.3 40.0

6.7 10.0

Duration with CT Scan

<6 months 1-5 years

6-10 years

>10 years

12 11

3

4

40.0 36.7

10.0

13.3

Entry Level

Qualification

Diploma

First Degree Master

PhD

22

8 0

0

73.3

26.7 0

0

Table 1 shows the demographic data of the respondents. The

total number of female radiographers (56.7%) was more than

the number of male radiographers (40.0%). Most of the

respondents, 20 (66.7%) were aged 20-29 years old, and only

3 (10%) were aged 40-49 years old. Most of the

radiographers have been working as radiographers at the

hospital for less than five years and between 5-10 years with

43.3% and 40.0%, respectively. Most of the respondents, 22

(73.3%) were from diploma level qualification, with most of

them involved in CT scan for less than six months (40%) and

between 1-5 years (36.7%).

Health Scope 89

3.2 Knowledge toward EBP in CT

Figure 1 shows the respondent’s knowledge of EBP in

CT. Most of the respondents agree (43.3%) and strongly

agree (43.3%) that they learned the foundation of EBP, and

only 1 (3.3%) stated for strongly disagree. This can be

considered most of the participants have at least some

knowledge regarding the EBP. Half (50%) of the

respondents agree that they involved in training in search

strategies, with only 1 (3.3%) strongly disagree in stated that

they familiar with medical search engines. This shows that

most of the respondents have knowledge of search strategies

and familiar with medical search engines. The other studies

conducted by Antwi WK et al. [1] among radiographers in

Ghana also agreed learned foundation of EBP, but only a

few of them had knowledge using medical search engines

such as MEDLINE. Most of the respondents strongly agree

(36.7%) and agree (33.3%) that they had formal training in

critical appraisal, and most of them agree (46.7%) that they

were confident in appraisal skills and search skills. The mean

for radiographer’s knowledge toward EBP in CT is 4.0667.

This shows that radiographers have high knowledge of EBP

in CT, which is 63.3%. This was support by the previous

study done by Nalweyiso et al. [14] toward radiographer, in

which the majority of them had high knowledge of EBP.

Figure 1: Knowledge toward EBP in CT

Figure 2: The specific terms regarding EBP in CT

Figure 2 shows that majority of the respondents answer

‘YES’ to the terms such as relative risk, systematic review,

meta-analysis, confidence interval, heterogeneity, and

publication bias, which have large differences with

respondents that answer ‘NO’. This shows that most of the

respondents understand most of these terms used. There is

only slightly different between the percentage of ‘YES’ and

‘NO’ answer for terms such as absolute risk, which 60%

knows and 40% didn’t know and also odds ratio terms with

56.7% answer ‘yes’ and 43.3% answer ‘no. The study

conducted by Guidetti et al. [8] also stated that the research

terms with the lowest understanding were odds ratio.

Table 2: Knowledge Level toward EBP in CT N Frequency (%)

Valid Low 11 36.7

High 19 63.3

Total 30 100

Table 2 shows the percentage of the respondent’s knowledge

level toward EBP in CT. It shows that the percentage of the

respondents (63.3%) who have high knowledge toward EBP

in CT is greater than the respondents who have low

knowledge (36.7%). Based on table 7, the differences

between the knowledge level with the duration years

working in CT and their entry-level qualification was

calculated. From the results, the respondents involved in CT

scan were <6 months (40%), 1-5 years (36.7%), 6-10 years

(10%), and >10 years (13.3%). The mean score for

respondents involves less than six months was 4.4167, 1-5

years were 3.6667, 6-10 years were 3.5000, and more than

ten years were 4.5417. Based on ANOVA, there is a

difference between the knowledge level and their years

working in CT scan as the p values are 0.006, which is less

than (p value> 0.05). After the post-hoc test was done, there

are only differences between radiographers working less than

six months and 1-5 years as the p-value less than 0.05 while

the other combination has no differences. There are also no

differences between entry-level qualification and their

knowledge level as the p-value is 0.369.

90 Health Scope, 2020, Vol. 3(3) Nur Ezzatie Aqillah et al.

3.3 Attitude and Practices toward EBP in CT

The findings of this study were that the respondents have

a good attitude and practices toward EBP in CT. Figure 3

shows the attitude and practice of respondents toward EBP

in CT. Most of the respondents agreed (40%) and strongly

agreed (40%) that EBP is essential for improving the quality

of services and practices in CT. This also supported by the

previous study by Elshami et al. [6], in which 99% of

respondents agreed that EBP important as it improves

practices and ensures better patient care. Only 1 (3.3%)

disagrees and strongly disagrees that EBP contributes to the

improvement of the decision making about technical and

other practices in CT. This shows that most of the

respondents accept the use of EBP in clinical practice CT.

The respondents agree (n=16) that EBP promotes the

participation of the patient, and only 2 (6.7%) strongly

disagree that EBP reinforces the guarantee of patient safety.

Most of the respondents agree that EBP promotes the

confidence of the users and scientific reading and research

results contribute to enhancing the effectiveness and

efficiency of the procedures with 63.3% and 56.7%,

respectively. This can be concluded that radiographers used

the EBP in their clinical practice as it improves their

procedures involving CT.

0

10

20

30

40

50

60

70

The

assumptions

of EBP are

essential forimproving the

quality ofservice and

practices inCT

The EBP

contributes to

the

improvementof the

decisionmaking about

the technicalproceduresand other

practices inCT

EBP

promotes the

participations

of patient,taking into

account theirinformed

consent ontechnical

procedures

and otherpractices in

CT

The EBP

reinforces the

guarantee of

patient safety

The EBP

promotes the

confidence of

the usersabout myself,

my technicalprocedures

and otherpractices

Scientific

reading and

research

results,specially

directed toCT,

contribute toenhancing theeffectiveness

andefficiency of

theprocedures

Percentages (%)

Figure 3: Attitudes and Practices toward EBP in CT

Table 3 shows the attitudes and practices level of the

respondents toward EBP in CT. It shows that most of the

respondents (56.7%) have high attitudes and practices

toward EBP in CT, and 13 (43.3%) of them have low

attitudes and practices. One-way ANOVA test had done to

determine the differences between the duration working in

CT and qualification level with their attitudes and practices

toward EBP in CT and the results show that there are no

differences between the duration working in CT scan and

qualification level with their attitudes and practices as the p-

value are 0.691 and 0.230 respectively. Therefore, they were

excluded for the next step of analysis (post-hoc). This is

supported by the study done by Eid AbuRuz et al. [5] that

years of experience do not have a correlation with their

attitudes and practices toward EBP.

Table 3: Attitudes and Practices Level toward EBP in CT N Frequency (%)

Valid Low 13 43.3

High 17 56.7

Total 30 100

3.2 Barriers toward implementation of EBP in CT

Table 4: Mean Barriers of EBP in CT Barriers Mean Std.

Deviation

I don’t have insufficient time

3.47 0.819

Lack of information resources

3.40 0.968

Lack of research skills

3.33 0.959

Inability to critically appraise literature

3.60 0.855

Lack of generalization of research findings

3.33 0.758

Inapplicability of research findings to

individual patients

3.30 0.750

Lack of understanding of statistical analysis

3.20 1.031

Lack of collective support among colleagues

3.07 0.944

Lack of interest 3.13 0.973

Table 4 shows the mean of the barriers perceived toward

implementation of EBP in clinical practice of CT. The result

shows that the three highest mean values, which are 3.60 for

inability to critically appraise literature, 3.47 for insufficient

time and 3.40 for the lack of information resources. This also

support by the previous study by C. Costa and A. F.

Abrantes [3], which majority of the radiographer agreed that

knowledge to transfer the result of literature to their daily

practice is one of the barriers in implementing the EBP in

CT. Another study done by Antwi WK et al. [1] and

Guidetti et al. [8] stated that lack of time had been shown to

be one of the major barriers to the use of evidence in practice

while lacking information resources as the second major

barriers. As stated by Antwi WK et al. [1], the time to

answer the clinical questions, selects an effective search

strategy, finding a source cover the topic under questions and

synthesizing multiple information to formulate an

answerable question was the steps involve when using

evidence-based as practice.

Then, the other barriers that affect the use of EBP in CT

among radiographers are lack of research skills and lack of

generalization of research findings with mean 3.33. This

shows that radiographers are lack skills in finding the

information related to research. The same findings show in

the previous study conducted by Chukwuani et al. [4], which

stated that access to research data is perceived as barriers by

nearly half of the respondents.

Health Scope 91

Table 5: Differences between demographic variable and

knowledge, attitude and practice level toward EBP in CT

ANOVA (p-value.)

Knowledge Level Attitudes and Practices Level

Duration

with CT Scan

0.006

0.691

Entry Level

Qualification

0.369

0.230

4. CONCLUSION

In conclusion, from the results of the study, most of the

radiographers agreed and have high knowledge regarding the

EBP in CT by 63.3% (n=19), and most of them understand

most of the specific terms related to EBP. Then, 56.7% of

the radiographers have a high level and positive attitudes and

practices toward EBP in CT. Inability to critically appraisal

literature, insufficient time and lack of information resources

are identified to be the main barriers that affect the use of

EBP in CT. Therefore, effort from everyone to advance the

use of EBP in CT should focus on decreasing these barriers.

Lastly, there is a difference between the duration working

with CT and their knowledge level toward EBP in CT but no

differences with their attitudes and practices. There are also

no differences between the entry-level qualification and both

their knowledge, attitudes, and practices toward EBP in CT

as there is no significant.

ACKNOWLEDGMENTS

The authors would like to thank the Centre of Medical Imaging, Faculty of Health Sciences, Universiti Teknologi MARA, Puncak Alam campus for the chance and support. Big thanks also to everyone that involve in completing this research.

REFERENCES

[1] Antwi WK, K. K. (2015). Evidence-Based Practice in Radiography: Attitudes, Beliefs, Knowledge and Practices of Radiographers in

Ghana. OMICS Journal of Radiology, 04(01).

https://doi.org/10.4172/2167-7964.1000176 [2] Anuradha, C., Jacob, K. S., Shyamkumar, N. K., & Sridhar, G.

(2013). Evidence-based practice in radiology: Knowledge, attitude,

and perceived barriers to practice among residents in radiology. European Journal of Radiology, 82(5), 894–897.

https://doi.org/10.1016/j.ejrad.2013.01.026

[3] C. Costa, A. F. Abrantes, L. P. R. et al. (2017). Evidence-based practice in radiology : the radiographer perspective. 1–9.

[4] Chukwuani, A. E., Osanaiye, A., & Obinna, F. E. (2017). Attitude

Towards Research Evidence Utilization in Radiography Practice. Http://Www.Sciencepublishinggroup.Com, 2(4), 162.

https://doi.org/10.11648/J.BSI.20170204.15

[5] Eid AbuRuz, M., Abu Hayeah, H., Al Dweik, G., & Yousef Al Akash, H. (2017). Knowledge, Attitudes, and Practice about Evidence-Based

Practice: A Jordanian Study. Health Science Journal, 11(2), 1–8.

https://doi.org/10.21767/1791-809X.1000489 [6] Elshami, W., Elamrdi, A., Alyafie, S., & Abuzaid, M. (2016).

Continuing professional development in radiography: practice, attitude

and barriers. International Journal of Medical Research & Health Sciences, 5(1), 68. https://doi.org/10.5958/2319-5886.2016.00015.1

[7] García Villar, C. (2011). Evidence-based radiology for diagnostic

imaging: What it is and how to practice it. Radiología (English

Edition), 53(4), 326–334.

https://doi.org/10.1016/j.rxeng.2011.02.001

[8] Guidetti, S., Heiwe, S., Kajermo, K. N., Samuelsson, M., Andersson,

I., & Wengstro, Y. (2011). Evidence-based practice : attitudes ,

knowledge and behaviour among allied health care professionals.

23(2), 198–209. [9] Hafslund, B., Clare, J., Graverholt, B., & Wammen Nortvedt, M.

(2008). Evidence-based radiography. Radiography, 14(4), 343–

348. https://doi.org/10.1016/j.radi.2008.01.003

[10] Halaweish, I., Riebe-Rodgers, J., Randall, A., & Ehrlich, P. F. (2018).

Compliance with evidence-based guidelines for computed tomography

of children with head and abdominal trauma. Journal of Pediatric Surgery, 53(4), 748–751.

https://doi.org/10.1016/j.jpedsurg.2017.07.008

[11] Joseph, D. Z., Laushugno, S. S., Mathew, A., Musa, G., Sani, M. U., Specialist, S., … State, B. (2019). KNOWLEDGE AND

PERCEIVED BARRIERS IN THE APPLICATION OF

EVIDENCE-BASED MEDICALRADIOGRAPHY IN NORTH, 4(2), 111–119.

[12] Lavelle, L. P., Rcsi, F., & Dunne, R. M. (2015). Evidence-based

Practice of Radiology 1 EBP = evidence-based practice, EORTC

= European Organisation for Research and Treatment of Cancer,

GCP = graph of conditional probability, LI-RADS = Liver

Imaging Reporting and Data System, PBL = practice-based lea. RadioGraphics, 35, 1802–1813.

https://doi.org/10.1148/rg.2015150027

[13] Miglioretti, D. L., & Smith-Bindman, R. (2011). Overuse of computed

tomography and associated risks. American Family Physician, 83(11),

1252–1254. [14] Nalweyiso, D. I., Kabanda, J., Mubuuke, A. G., Sanderson, K., &

Nnyanzi, L. A. (2019). Knowledge, attitudes and practices towards

evidence based practice: A survey amongst radiographers. Radiography, xxxx. https://doi.org/10.1016/j.radi.2019.03.004

[15] Seeram, E. (2010). Computed tomography: Physical principles and

recent technical advances. Journal of Medical Imaging and Radiation Sciences, 41(2), 87–109. https://doi.org/10.1016/j.jmir.2010.04.001

[16] Snibsøer, A. K., Graverholt, B., Nortvedt, M. W., Riise, T., &

Espehaug, B. (2018). Evidence-based practice profiles among bachelor students in four health disciplines: A cross-sectional study. BMC

Medical Education, 18(1), 1–10. https://doi.org/10.1186/s12909-018-

1319-7

Health Scope 92

RESEARCH ARTICLE

Effect of low dose ionizing radiation-induced oxidative stress on sperm count and sperm motility in male mice Nur Adira Mohamad Yusuf, Mohd Hafizi Mahmud*

Centre of Medical Imaging, Faculty of Health Sciences, Universiti Teknologi MARA Cawangan Selangor Kampus Puncak Alam, 42300 Bandar Puncak Alam, Selangor, Malaysia

Abstract:

Male infertility is contributed to approximately 50% of infertility worldwide. Reactive oxygen species (ROS) is known as one of the major factors contributing to male infertility. This animal modeling study aimed to evaluate the effects of oxidative stress induced by low dose ionizing radiation exposure on sperm count and sperm motility. Twelve adult male ICR mice were randomly divided into two groups (control group, n=6 and irradiated group, n=6). Both groups received filtered tap water and a normal pellet diet for five days of the acclimatization period prior to total body exposure of 200 µGy x-ray on the 6th day. All the mice were sacrificed in cervical dislocation. The epididymis was evaluated for sperm analysis using a computer-assisted sperm analyzer (CASA) system. The sperm count and sperm motility decreased significantly in the irradiated group as compared to the control group (95.8 ± 40.2 million vs 49.3 ± 19.3 million, p < 0.05 and 40.8 ± 13.3 % vs 23.2 ± 10.3 %, p < 0.05), respectively. Low dose ionizing radiation-induced oxidative stress would reduce the sperm count and sperm motility as a consequence of the increased level of ROS in the testis.


Mohd Hafizi Mahmud


Keywords: Low dose ionizing radiation; oxidative stress; male infertility; sperm count; sperm motility

1. INTRODUCTION

Ionizing radiation has become a necessary part in modern medicine for medical diagnostic and therapeutic. Ionizing radiation breaks the chemical bonds of biological macromolecules following its interaction with living cells. Ionizing radiation can affect proteins, nucleic acids and complex lipids as a result of the generation of reactive oxygen species (ROS) via radiolysis of water or alteration of mitochondrial functions [1]. ROS is highly chemical entities performing vital signaling functions and occurs in nature in the cells [2, 3]. ROS activity is regulated by an amount of enzymatic and non-enzymatic antioxidants. The imbalance of increased generation of ROS and antioxidant mechanisms leads to oxidative stress as manifested by increased oxidative molecular damage to biomolecules [4]. As a consequence, the oxidative damage leads to reduce growth, fertility and survival [5]. Low dose ionizing radiation is defined as ionizing radiation with dose < 100 mGy [6]. The key sources of low-dose radiation exposure to the general population are diagnostic medical exposures and natural background radiation [7]. Low dose ionizing radiation has been significantly associated with the generation of ROS [2, 8]. Previous reports have shown significant correlation between genetic damage and oxidative damage following the effect of ROS generated from ionizing radiation [5, 9].

Male infertility as a result of ROS generation occurs via two mechanisms. Initially, ROS damages the sperm

membrane and decreases sperm mobility thus affecting their fusion capability with the oocyte. Furthermore, transmission of sperm information to the embryo is interrupted as a result of DNA damage of the sperm [10]. Increased levels of ROS have been reported in the semen of approximately 25 - 40% of infertile men [11].

Many studies have addressed the effect of several inducers of oxidative stress on the sperm quality in male mice such as radiofrequency electromagnetic radiation [12], radioactive cesium [13] and gamma rays [14, 15]. Nevertheless, limited studies have futher investigated such effect from acute low dose x-ray irradiation. Hence, the present study was designed to investigate in vivo effects of low dose ionizing radiation induced oxidative stress on sperm count and sperm motility in mice.


2.1 Animal handling

The animal study was conducted in accordance with the criteria set by Universiti Teknologi MARA Committee of Animal Research and Ethics (UiTM CARE) concerning the use of experimental animals. Twelve (n=12) healthy six weeks old male Mus musculus mice species weight of 28 - 30 grams were obtained from Laboratory Animal Facility and

93 Health Scope, 2020, Vol. 3(3) Nur Adira et al.

Management (LAFAM), UiTM Selangor Branch, Puncak Alam Campus. Mice were individually housed under standard conditions (controlled atmosphere with a 12 hrs light/dark cycle at 20 – 25 oC). The mice were acclimatized for five days and normal pellet diet along with filtered tap water was given Ad libitium.

2.2 Animal irradiation

The mice were divided randomly into irradiated and control group. The irradiated group (n=6) were exposed with single fractionated of 200 µGy x-ray for total body irradiation on the day sixth. The low dose ionizing radiation was irradiated by using x-ray (Philips Bucky Diagnost) performed under the guidelines provided by the Centre of Medical Imaging, Faculty of Health Sciences, Universiti Teknologi MARA Selangor Branch, Puncak Alam Campus. The irradiated and control mice were sacrificed by cervical dislocation and their epididymis tissues were immediately removed for further analysis following the treatment procedure.

2.3 Semen collection and sperm analysis

For preparation of epididymal sperm, the mice were sacrificed by cervical dislocation. About 3 mm cauda epididymis was dissected and placed in 1 ml of medium M16 [14]. The semen samples were incubated at 37oC for 15 min. Subsequently, the sperm count and motility were analyzed by a Hamilton-Thorne computer assisted sperm analyzer (CASA) (Hamilton Thorne Research, Beverly, MA, USA) (Figure 1). Figure 1: Sperm analysis using Hamilton-Thorne computer assisted sperm

analyzer (CASA)


All data were analyzed using SPSS software statistical package version 21 (SPSS Inc. Chicago, IL, USA). The data were tested for normality and analyzed with the Student’s Independent T-test. The results were expressed as the mean ± standard deviation (SD) with a p -value < 0.05 was deemed statistically significant.


Significant reduction of sperm count and sperm motility were observed in the irradiated group as compared to the control group (p < 0.05) as shown in Table 1. Upon irradiation, sperm count and sperm motility were reduced by 48.5% and 43.1%, respectively.

Table 1: Mean sperm count and sperm motility of the mice in control and irradiated groups following low dose ionizing

radiation

Sperm analysis

Control Irradiated

group group

Mean Difference (95% CI)

p-value

Sperm count (106)

95.8 ± 40.2 49.3 ± 19.3 46.5 (5.81 - 87.19)

0.029*

Sperm motility (%)

40.8 ± 13.3 23.2 ± 10.3 17.7 (3.36 - 32.97)

0.028*

The results are expressed as mean±standard deviation. *Denotes significant result with p < 0.05 In the present study, we investigated the effects of acute low dose ionizing radiation (200 µGy) on the sperm quality to understand the underlying toxicity of low dose irradiation as a result of oxidative stress. Testicular tissue becomes vulnerable to oxidative stress-induced pathologies due to the inherent abundance of highly unsaturated fatty acids, high metabolic activity, high mitotic activity and the presence of potential ROS- generating systems [17, 18]. Testicular tissue is one of the most radiosensitive tissues, which can be significantly functionally impaired by low doses of radiation as little as 0.1 Gy [19]. Evidence suggest that ionizing radiation affects the testicular tissue by disturbing the normal spermatogenic metabolism, proliferation and differentiation, which results in mutagenesis or apoptosis of radiosensitive cells, low sperm counts and defective sperm function [20]. Oxidative stress is known as a contributing factor in the development of male infertility and defecting sperm function [21]. Various factors inducing oxidative stress in animal models including infections, tetracycline, streptozotocin (STZ), chromic acid, bisphenol and tertiary-butyl hdyroperoxide (TBHP) have been reported [22–27]. As x-ray exposure leads to ROS formation, hence x-ray irradiation is a reliable method for induction of ROS in animal and human studies.

Epididymal sperm counts and motility are a widely used, simple and sensitive method of assessing the effects of male reproductive toxicants [28] and these parameters are useful in diagnosing male infertility [12]. The present study is in accordance with the previous studies investigating the effect of x-ray irradiation on human sperm quality. Ionizing radiation has been shown to reduce the total number of spermatozoa and the number of motile spermatozoa [14]. Exposure to ionizing radiation leads to abnormalities in spermatogenesis including low sperm count, produce abnormal spermatozoa, and damaged sperm function [29]. Study by Kumar et al. [30] found that low dose ionizing radiation from occupational exposure changes the motility pattern of sperm without affecting the fertility. On the

Health Scope 94

contrary, higher dose (> 2Gy) of ionizing radiation was denoted from those studies as compared to the present study. A mice model study by Mohamed et al. [31] reported that acute low dose ionizing radiation (100 µGy) would induce oxidative stress. Hence our findings postulate that acute low dose ionizing radiation could significantly affect the epididymal sperm counts and sperm motility.

4. CONCLUSION

Low dose ionizing radiation-induced oxidative stress would lead to decrease the sperm count and sperm motility as a result of increased level of ROS in the mice testis. This finding may serve as important implication for estimating risks of male infertility related to low dose ionizing radiation among radiation workers. This study model may be proposed as a valuable tool for further investigation on the effects of various antioxidants activities on germinal tissues. The sperm analysis on sperm count and sperm motility without histological analysis was insufficient to evaluate the effect of ionizing radiation-induced oxidative stress on the sperm quality. Hence, further investigation on other kinematic parameters of spermatozoa including progressive motility, linearity and curvilinear velocity integrated with histological analysis of sperm would be valuable in the future investigation of the effect of low dose ionizing radiation-induced oxidative stress on the sperm quality.

ACKNOWLEDGEMENTS

The authors would like to express their appreciation to the radiographers of the Medical Imaging Laboratory, Centre of Medical Imaging, Faculty of Health Sciences and staffs of LAFAM, Universiti Teknologi MARA for their invaluable cooperation and assistance during conducting the study.

REFERENCES

[1] Kam, W.W., Banati, R.B. Effects of ionizing radiation on mitochondria. Free Radical Biology & Medicine, 2014. 65: p. 607- 619

[2] Azzam, E., Jay-Gerin, J.P. and Pain, D. Ionizing radiation-induced metabolic oxidative stress and prolonged cell injury. Cancer Letters, 2011. 327 (1-2): p. 48 - 60

[3] Murphy, M., et al., Unraveling the biological roles of reactive oxygen species. Cell Metabolism, 2011. 13 (4): p. 361- 366

[4] Jones, D.P. Redefining oxidative stress. Antioxidant & Redox Signaling, 2006. 8 (9-10): p. 1865 -1879

[5] Costantini, D. Variation in oxidative stress threats and hormesis acrcoss environments: In Oxidative Stress and Hormesis in Evolutionary Ecology and Physiology. 2014. Springer-Verlag: p. 75-109

[6] United Nations Scientific Committee on the Effects of Atomic Radiation (UNSCEAR). Summary of Low-Dose Radiation Effects on Health. New York: United Nations; 2011

[7] United Nations Scientific Committee on the Effects of Atomic Radiation. Sources and effects of ionizing radiation. New York: United Nations; 2008

[8] Smith, J.T., Willey, N.J. and Hancock, J.T. Low dose ionizing radiation produces too few reactive oxygen species to directly affect antioxidant concentrations in cells. Biology Letter. 2012. 8 (4) : p. 594-597

[9] Galván, I. et al., Chronic exposure to low-dose radiation at Chernobyl favours adaptation to oxidative stress in birds. Functional Ecology. 2014. 28 (6): p. 1387-1403

[10] Tremellen, K. Oxidative stress and male infertility - a clinical perspective. Human Reproduction Update. 2008. 14 (3): p. 243 -258.

[11] Padron, O.F. et al., Seminal reactive oxygen species and sperm motility and morphology in men with spinal cord injury. Fertility & Sterility. 1997. 67 (6): p. 1115-1120

[12] Kumaria, K. et al., Effects of intermediate frequency magnetic fields on male fertility indicators in mice. Environmental Research. 2017. 157: p. 64 – 70

[13] Okano, T. et al., Effects of environmental radiation on testes and spermatogenesis in wild large Japanese field mice (Apodemus speciosus) from Fukushima. Scientific Reports. 2016. 6: p. 23601

[14] Gong, E.J., et al., Low-dose-rate radiation exposure leads to testicular damage with decreases in DNMT1 and HDAC1 in the murine testis. Journal of Radiation Research . 2014. 55 (1): p. 54-60

[15] Khan, S., et al., Radioprotective potential of melatonin against 60Co γ-ray induced testicular injury in male C57BL/6 mice. Journal of Biomedical Science. 2015. 22 (1): p. 1-15

[16] Nabavi, N., Todehdehghan, F and Shiravi, A. Effect of caffeine on motility and vitality of sperm and in vitro fertilization of outbreed mouse in T6 and M16 media. Iranian Journal of Reproductive Medicine . 2013. 11 (9) : p. 741-746

[17] Sheweita, S. A., Abdulkarim, M. and Al-Sawaf, H. Mechanisms of male infertility: Role of antioxidants. Current Drug Metabolism. 2005. 6 (5) :p 495-501

[18] Sisodia, R. et al., Spinacia oleracea modulates radiation- induced biochemical changes in mice testes. Indian Journal of Pharmaceutical Sciences. 2008. 70 (3): p. 320-326

[19] Howell, S.J. and Shalet, S.M. Spermatogenesis after cancer treatment: damage and recovery. Journal of the National Cancer Institute. 2005. 34: p. 12 - 17

[20] Ibrahim, R. Y. M. and Ghoneim, M. A. M. Study of some biochemical and molecular changes induced by radiation hormesis in testicular tissues of male rats. International Journal of Advanced Research. 2014 .2 (7): p. 397- 407

[21] Kaur, P., Kaur, G. and Bansal, M. Upregulation of AP1 by tertiary butyl hydroperoxide induced oxidative stress and subsequent effect on spermatogenesis in mice testis. Molecular and Cellular Biochemistry. 2008. 308 (1-2): p.177-181

95 Health Scope, 2020, Vol. 3(3) Nur Adira et al.

[22] Wang, A. et al., Generation of reactive oxygen species by leukocytes and sperm following exposure to urogenital tract infection. Achieves of Andrology. 1997. 39 (1): p. 11-17

[23] Acharya, U.R. et al., Testicular dysfunction and antioxidative defense system of Swiss mice after chromic acid exposure. Reproductive Toxicology. 2006. 22 (1): p. 87 - 91

[24] Shrilatha, B. Occurrence of oxidative impairments, response of antioxidant defenses and associated biochemical perturbations in male reproductive milieu in the Streptozotocin-diabetic rat. International Journal of Andrology. 2007. 30 (6): p. 508 - 518

[25] Shrilatha, B. and Muralidhara. Early oxidative stress in testis and epididymal sperm in streptozotocin-induced diabetic mice: its progression and genotoxic consequences. Reproductive Toxicology. 2007. 23 (4): p. 578 - 587

[26] Farombi, E.O. et al., Tetracycline-induced reproductive toxicity in male rats: effects of vitamin C and Nacetylcysteine. Experimental and Toxicological Pathology. 2008. 60 (1): p. 77- 85

[27] Fatemi, N.et al. Effect of tertiary-butyl hydroperoxide (TBHP)-induced oxidative stress on mice sperm quality and testis histopathology. Andrologia. 2013. 45 (4): p. 232 - 239.

[28] Migliore, L. et al., Assessment of sperm DNA integrity in workers exposed to styrene. Human Reproduction. 2002. 17 (11): p. 2912 - 2918

[29] Sun Hee, L. and Ho Shin, C. Reduced male fertility in childhood cancer survivors. Annals of Pediatric Endocrinology & Metabolism. 2013. 18 (4): p. 168 -72.

[30] Kumar, D. et al., Semen abnormalities, sperm DNA damage and global hypermethylation in health workers occupationally exposed to ionizing radiation. PLoS ONE. 2013. 8 (7): e69927

[31] Mohamed, M.I. et al., Induction of oxidative stress following low dose ionizing radiation in ICR mice. World Journal of Medical Sciences. 2014. 10 (2): p. 198 - 203

Preface

Health Scope (e-ISSN: 2735-0649) is a peer-reviewed and evidence-based scien�fic research book published by Faculty of Health Sciences, Universi� Teknologi MARA, Puncak Alam Campus, Malaysia. The research book's mission is to promote excellence in health sciences and a range of disciplines and special�es of allied health professions. It welcomes submissions from academic and health professionals’ community. The research book publishes evidence-based ar�cles with solid and sound methodology, clinical applica�on, descrip�on of best clinical prac�ces, and discussion of relevant professional issues or perspec�ves. Ar�cles can be submi�ed in the form of research ar�cles, reviews, case reports, and le�ers to the editor or short communica�ons. The journal's priori�es are papers in the fields of Physiotherapy, Occupa�onal Therapy, Optometry, Medical Laboratory Technology, Environmental Health & Safety, Nursing, Nutri�on & Diete�cs, Medical Imaging and Basic Sciences. Relevant ar�cles from other disciplines of allied health professions may be considered for publica�on.

Preface

bookcover white edit

Documents