Respiratory Medicine (2013) 107, 673e683

Available online at www.sciencedirect.com

journal homepage: www.elsevier .com/locate/rmed

Functional and biological characteristics ofasthma in cleaning workers

David Vizcaya a,b,c,d,*, Maria C. Mirabelli e, Ramon Orriols f,g,Josep Maria Anto a,b,c,h, Esther Barreiro f,i, Felip Burgos f,j,Lourdes Arjona a,b,h, Federico Gomez f,j, Jan-Paul Zock a,b,h

aCentre for Research in Environmental Epidemiology (CREAL), Barcelona, Spainb IMIM (Hospital del Mar Research Institute), Barcelona, SpaincUniversity Pompeu Fabra (UPF), Barcelona, SpaindUniversity of Montreal Hospital Research Center (CRCHUM), Montreal, Quebec, CanadaeDepartment of Epidemiology and Prevention, Division of Public Health Sciences, Wake Forest Schoolof Medicine, Winston-Salem, North Carolina, USAfCentro de Investigacion en Red de Enfermedades Respiratorias (CIBERES), Instituto de Salud Carlos III(ISCIII), Bunyola, Majorca, Balearic Islands, SpaingDepartment of Pulmonology, Hospital Vall d’Hebron, Barcelona, Catalonia, SpainhCIBER - Epidemiologıa y Salud Publica, CIBERESP, Barcelona, Spaini Pulmonology Department-Muscle and Respiratory System Research Unit (URMAR), IMIM-Hospital delMar, Barcelona, Spainj Servei de Pneumologia, Hospital Clınic, IDIBAPS, Barcelona, Spain

Received 27 September 2012; accepted 15 January 2013Available online 20 February 2013

KEYWORDSAsthma;Cleaners;Inflammation;Occupational;Epidemiology;Case-control

* Corresponding author. UniversityCanada. Tel.: þ1 514 890 8000x15921

E-mail address: dvizcaya.epi@gma

0954-6111/$ - see front matter ª 201http://dx.doi.org/10.1016/j.rmed.201

Summary

Objectives: Cleaning workers have an increased risk of asthma but the underlying mechanismsare largely unknown. We studied functional and biological characteristics in asthmatic cleanersand compared these to healthy cleaners.Methods: Forty-two cleaners with a history of asthma and/or recent respiratory symptoms and53 symptom-free controls were identified. Fractional exhaled nitric oxide (FeNO) was mea-sured and forced spirometry with reversibility testing was performed. Total IgE, pulmonary sur-factant protein D and the 16 kDa Clara Cell secretory protein were measured in blood serum.Interleukins and other cytokines, growth factors, cys-leukotrienes and 8-isoprostane weremeasured in exhaled breath condensate. Information on occupational and domestic use ofcleaning products was obtained in an interview. Associations between asthma status, specific

of Montreal Hospital Research Centre (CRCHUM), 3875 Rue Saint-Urbain, Montreal, QC H2W 1V1,; fax: þ1 514 412 7106.il.com (D. Vizcaya).

3 Elsevier Ltd. All rights reserved.3.01.011

674 D. Vizcaya et al.

characteristics and the use of cleaning products were evaluated using multivariable linear andlogistic regression analyses.Results: Asthma was associated with an 8% (95% confidence interval (CI) 1e15%) lower post-bronchodilator FEV1, a higher prevalence of atopy (42% vs. 10%) and a 2.9 (CI 1.5e5.6) timeshigher level of total IgE. Asthma status was not associated with the other respiratory bio-markers. Most irritant products and sprays were more often used by asthmatic cleaners. Theuse of multiuse products, glass cleaners and polishes at work was associated with higher FeNO,particularly in controls.Conclusions: Asthma in cleaning workers is characterised by non-reversible lung function de-crement and increased total IgE. Oxidative stress, altered lung permeability and eosinophilicinflammation are unlikely to play an important underlying role, although the latter may be af-fected by certain irritant cleaning exposures.ª 2013 Elsevier Ltd. All rights reserved.

Introduction

A growing body of epidemiologic research now suggests thatcleaning workers are at increased risk of asthma and relatedrespiratory symptoms.1e5 In Catalonia, Spain, cleaning-related exposures have been identified as one of the maincauses of physician-diagnosed occupational asthma.6 Amongthe cleaning-related exposures investigated, occupationaluse of products containing respiratory irritants such as hy-pochlorite bleach and ammonia have been associated withasthma and asthma-related conditions.7e10 Domestic use ofcleaning products, in particular those in spray form, hasbeen also suggested as a risk factor for asthma.11,12

Despite the increasing epidemiological evidence sup-porting elevated risks of asthma among cleaners,1,5 limiteddata are available to describe the physiologic characteristicsof cleaning-related asthma or suggest underlying mecha-nisms by which cleaning-related exposures may impact res-piratory health. It has been proposed that inhalation ofcompounds with respiratory irritant properties may inducebronchial epithelial damage and facilitate allergic sensiti-zation by triggering a pro-inflammatory response, neuro-genic inflammation, increased lung permeability, andremodelling of the airways epithelium.13 Non-allergic-mediated bronchial inflammation has also been described;specifically, occupational and non-occupational exposuresalike may lead to non-eosinophilic bronchial inflammationand the onset or aggravation of asthma through non-allergicpathways.14 Specific sensitization to the individual compo-nents of cleaning product formulations may also play a rolein the observed risk of asthma among cleaners.15

Such inflammatory processes may be detected by char-acterization of biomarkers of respiratory health. Cytokinesand growth factors involved in the inflammatory response inasthma can be measured in exhaled breath condensate(EBC), a suitable non-invasive matrix for this purpose.16

Eosinophilic inflammation can be evaluated using thefraction of exhaled nitric oxide (FeNO).17 Finally, bronchialepithelium damage may alter lung permeability, and can beassessed by the blood serum levels of pulmonary proteinswhich move passively across the alveolar epithelial barrierinto the peripheral blood stream when the lung epitheliumpermeability is compromised.18

We conducted this study to describe the functional andbiological characteristics of asthma in cleaning workers and

to compare a group of cleaners with asthma with a group ofcleaning workers free of respiratory symptoms. Thesefindings are based on analysis of the second stage ofa multi-stage study of cleaning-related exposures andasthma. In the first stage of the project, we evaluated riskfactors for asthma and asthma-related symptoms amongprofessional cleaning workers.9 In this second stage, weassess the inflammatory profile, oxidative stress, sensiti-sation to aeroallergens, lung epithelium permeability,bronchial hyperresponsiveness and lung function in asth-matic and non-asthmatic cleaners identified in the firststage of the project. With this analysis, we providea description of the physiological and functional charac-teristics of asthma in cleaning workers compared toa healthy population with similar sociodemographic char-acteristics and evaluate associations of domestic andoccupational use of cleaning products with asthma andbiomarkers of respiratory health.

Methods

Study design and population

We conducted a case-control study nested within a largecross-sectional study of asthma among cleaning companyemployees in Barcelona, Spain. The study design andmethods of the questionnaire survey have been describedpreviously.9 Briefly, in 2008 we obtained self-administeredquestionnaires including information on respiratory symp-toms and asthma from 761 cleaning workers currentlyemployed at 37 cleaning companies in Barcelona. No mea-surements of lung function or any biomarker were con-ducted for the first stage of the project. Based on suchquestionnaires, we identified 70 prevalent cases withasthma symptoms (wheeze, chest tightness, breathlessnessat rest, breathlessness after exercise and nocturnalbreathlessness attack) in the last year and/or with a historyof asthma and 121 controls without any lower tract respi-ratory symptom and without a history of asthma. A similarsymptom-based definition of asthma has been used previ-ously.7 Between December 2008 and September 2009,selected cases and controls were interviewed by tele-phone. Those who were still employed as cleaning workersand who still met the case and control inclusion criteriawere invited to participate in a detailed clinic visit. Forty-

Figure 1 Selection flowchart of the study population from a previous cross-sectional study in cleaning workers.

Asthma in cleaning workers 675

676 D. Vizcaya et al.

two cases (60%) and fifty-three controls (44%) were finallyenrolled in the study (Fig. 1). Eligible participants and non-participants did not differ in age, educational level, sex,smoking status and symptoms either among cases or con-trols. A higher prevalence of adult-onset asthma was foundamong non-participant cases compared to participants(Supplement Table 1).

The present study was approved by the ethics commit-tee of Parc de Salut Mar, Barcelona, and the participantsprovided written informed consent.

Face-to-face interview

Information on respiratory symptoms, job history, domesticand occupational cleaning-related exposures, smokinghabits and demographic characteristics was obtained duringa computer-assisted face-to-face interview. Respiratoryhealth questions were taken from the Spanish version of theEuropean Community Respiratory Health Survey ques-tionnaire.19 Data on the use of cleaning products in theprevious year and the average number of hours per week ofproduct use was obtained separately for domestic andoccupational cleaning activities. The selected list of clean-ing products was based on findings from previous studies.7,9

Lung function testing

Forced expiratory volume in the first second (FEV1), forcedvital capacity (FVC) and forced expiratory flow between25% and 75% of FVC (FEF25e75%) were measured with anEasyOne portable spirometer (ndd Medical Technologies,Zurich, Switzerland) before and 15 min after the admin-istration of 400 mg salbutamol via metered dose inhaler,following standard recommendations.20,21 FEV1 and FVCwere expressed as percentages of the age-, sex- and height-specific predicted values. All models were adjusted forcigarette pack-years.

All participants who were eligible for methacholinechallenge testing were invited to a second clinic visit wherea bronchial hyperresponsiveness (BHR) test was conductedfollowing the ECRHS protocol.19

Fraction of exhaled nitric oxide (FeNO)

FeNO was measured using an electrochemical portabledevice (NIOX-MINO; Aerocrine, Solna, Sweden) with a con-stant airflow rate of 50 mL/s and following internationalrecommendations.22 Levels were expressed as parts perbillion (ppb).

Determination of biological markers in exhaledbreath condensate (EBC)

EBC was collected using an EcoScreen� condenser (JaegerGmbH, Wurzburg, Germany) following ATS/ERS Task Forcerecommendations.23 Treatment of samples has beendescribed previously.24,25 8-Isoprostane was analysed usingan enzyme-linked immunosorbent assay (ELISA; CaymanChemical, Ann Arbor, MI, USA). BD Cytometric Bead Array(CBA; BD Biosciences, Erembodegem, Belgium) and the BDFACSCalibur Flow Cytometer (Becton Dickinson, San Jose,

CA, USA), a particle-based immunoassay, were used tomeasure the following 10 cytokines and 2 growth factors:vascular endothelial growth factor (VEGF), basic fibroblastgrowth factor (FGF), tumour necrosis factor (TNF), inter-leukin (IL) 2, IL-4, IL-5, IL-8, IL-10, IL-12, IL-13, interferon-gamma (IFN-g), and IFN-geinduced protein 10 (Ip10). Weused the manufacturer recommendations of the corre-sponding lower limit of detection for each biomarker.

Determination of biological markers in blood serum

Blood serum samples were collected by venipuncture. CC16and SP-D were analysed using commercial kits (BiovendorLaboratornı medicına a.s., Modrice, Czech Republic).24 Theconcentration of total IgE, and specific IgE against Dustmite (Dermatophagoides pteronyssinus) and latex in serumwas determined using Chemoluminescent immunoanalysis(IMMULITE 2000. Siemens). We evaluated the levels ofspecific IgE against common aeroallergens using the Pha-diatop test (Pharmacia ImmunoCAP; Pharmacia, Uppsala,Sweden) as a proxy for atopy.

Data analysis

Levels of all cytokines and growth factors measured in EBCwere dichotomised as detectable or non-detectable toincrease the statistical power. Analysis was conducted ifmore than 5% of cases and controls were detectable. Thedistributions of the concentration of all biomarkers followeda log-normal shape. The associations between asthma anddichotomous outcomes were evaluated using multivariablelogistic regression, while associations with (log-trans-formed) continuous outcomes were evaluated using multi-variable linear regression. Associations with biomarkerswere expressed as Geometric Mean Ratios, computed as theexponential of the beta coefficient obtained from multi-variable linear regression models. All models were adjustedfor age, sex and smoking status (never, former and currentsmoker). The association between asthma and the use ofcleaning products was evaluated with multivariable logisticregression models for each cleaning product analysed. Allmodels were adjusted by domestic use of the studiedcleaning product, occupational use of the studied cleaningproduct, age, sex and smoking status (never, former andcurrent smoker). Statistical analyses were performed usingSAS version 9.1 (SAS Institute Inc., Cary, NC, USA).

Results

Cases and controls were both predominantly women(Table 1). Cases were on average six years younger thancontrols and more likely to smoke. A relatively high pro-portion of both groups was born outside Spain (30%) andreported educational attainment of primary education orless (cases: 62%, controls: 70%). Most cases and controlswere overweighted (60% and 64% with BMI � 25, respec-tively). The score of asthma26 among cases was on average2.2, 24% of them reported having had asthma confirmed bya physician, and 17% had their first asthma attack after theage of 16. Nineteen cases (45%) presented current asthmaas defined in previous studies.9 Atopy and sensitisation to

Asthma in cleaning workers 677

latex and dust mite was more prevalent among cases thancontrols (Table 1).

Lung function testing

Forced spirometry tests of 28 (67%) cases and 44 (83%)controls met the ATS/ERS quality criteria (Table 2) andwere, therefore, considered for analyses. Measurements ofFEV1, the FEV1/FVC ratio and FEF25%e75% were significantlylower in cases than in controls. No significant differences inFVC were observed. A higher proportion of cases as com-pared to controls presented a percentage of predicted FEV1

below 85% both before and after the inhalation of bron-chodilator. On the other hand, we found no differences in

Table 1 Demographic and respiratory health character-istics of the studied population.

Controlsn Z 53

Casesn Z 42

Female 47 (89%) 39 (93%)Age, mean � SD 48 � 8 42 � 10Smoking

Never smoker 30 (57%) 14 (33%)Former smoker 15 (28%) 7 (17%)Packs-year, mean � SD 5 � 6 28 � 31

Current smoker 8 (15%) 21 (50%)Packs-year, mean � SD 15 � 8 29 � 24

Country of birthSpain 37 (70%) 30 (71%)Other 16 (30%) 12 (29%)

Educational levelLess than primary school 7 (13%) 5 (12%)Primary school 30 (57%) 21 (50%)Secondary school or higher 16 (30%) 16 (38%)

Body mass index (kg/m2)<20 2 (4%) 1 (2%)20 to 24.9 17 (32%) 16 (38%)25 to 29.9 25 (47%) 12 (29%)� 30 9 (17%) 13 (31%)

Years employed as a cleaningworker, mean � SD

11.6 � 8.0 12.0 � 8.2

Doctor diagnosed asthma e 10 (24%)Adult onset asthma e 7 (17%)Current asthma a e 19 (45%)Asthma score, mean (SD) e 2,2 (1,4)Chronic cough e 16 (38%)Chronic phlegm e 10 (24%)Upper respiratory tractsymptoms

19 (36%) 27 (64%)

Atopyb 5 (10%) 17 (42%)Sensitisation to latexc 1 (2%) 3 (7%)Sensitisation toD. pteronyssinus c

2 (4%) 13 (31%)

n (%) unless otherwise indicated.a Wheeze with breathlessness and/or attack of asthma in the

last year and/or currently taking medication for asthma.b Phadiatop test positive. [specific IgE]>0.35 kU/L at least for

1 of 10 common aeroallergens.c Concentration of specific IgE in blood serum higher than

0.35 kU/L.

the proportion of cases and controls that showed a per-centage of predicted FVC below 85%. Reversibility ofbronchial obstruction was similar in cases and controls bothas a percentage of change and as a difference of FEV1 orFVC before and after the inhalation of bronchodilator. Asubgroup of 11 cases underwent methacholine challengetesting and showed higher prevalence of BHR compared to11 controls.

Biological markers

Levels of FeNO, SP-D and CC16 were similar in both groups(Table 3). Cases and controls did not differ in the per-centage of detectable levels of ILs and growth factors. Cys-leukotrienes and 8-isoprostane were detectable in almostall the analysed samples of EBC and no differences betweencases and controls were found in the average level of bothmarkers. Cases had significantly higher levels of total serumIgE than controls after adjusting by sex, age and smokingstatus.

Domestic and occupational use of cleaning productsand asthma

Cases and controls reported frequent domestic and occu-pational use of hypochlorite bleach, soaps or detergentsand degreaser (Table 4). The association with asthmasymptoms varied depending on the cleaning product andwhether the exposure occurred at home or at work. Casesreported a more frequent use of multi-use products com-pared to controls. The highest risk of asthma was observedamong those who used multi-use products at both settingsin the previous year (Supplement Table 2). Additionally,asthma was associated with occupational use of multi-useproducts in spray form (Table 4). Occupational use ofsoaps or detergents showed a statistically significant in-verse association with asthma. In general, ORs appearedhigher when evaluating domestic exposure compared tooccupational exposure.

Regarding the association between biomarkers and theuse of cleaning products both at home and at work, theinitial analysis strategy was stratifying for cases and con-trols showing highly imprecise results. Therefore, we con-ducted a cross-sectional analysis including both cases andcontrols adjusted for age, sex and smoking status. Occu-pational use of multi-use products during the previous year,adjusted for domestic use of the same product, was asso-ciated with increased levels of FeNO, serum IgE and FGF inEBC among cases and controls regardless of their symp-tomatic status (Table 5). Cleaning workers who used glasscleaners and polishes or waxes at work showed higher levelsof FeNO. Occupational use of glass cleaners was also asso-ciated with higher levels of TNF-a. The use of hydrochloricacid and degreasers at work during the previous year wasassociated with increased IgE levels. Regarding exposuresat home, domestic use of ammonia during the previous yearwas associated with higher levels of 8-isoprostane in EBCand the use of hydrochloric acid with higher levels of TNF-a. Geometric means of biomarkers concentrations in EBCare provided in the online supplement (Supplement Tables3 and 4).

Table 2 Functional characteristics of cases and controls.

Controls (n Z 44) Cases (n Z 28) Adj. Coeff. (95%CI) b

PrebronchodilatorFEV1/FVC (%) 81.5 (4.4) 76.7 (6.5) �4.4 (�7.4 to �1.5)FVC (% Predicted), mean (SD) 96.7 (12.9) 97.8 (12.7) 3.2 (�3.7e10.1)FEV1 (% Predicted), mean (SD) 99.3 (13.1) 92.8 (11.2) �6.8 (�14.0 to 0.3)FVC (% Predicted) < 85, n (%) 10 (21) 7 (22) 0.5 (0.1e2.2)FEV1 (% Predicted) < 85, n (%) 5 (11) 9 (28) 1.5 (0.3e6.5)FEF25e75% (L/s) 2.9 (0.9) 2.4 (0.9) �0.5 (�1.0 to �0.1)

Postbronchodilatorc

FEV1/FVC (%) 83.4 (4.9) 77.9 (7.1) �5.2 (�8.8 to �1.6)FVC (% Predicted), mean (SD) 95.5 (12.7) 97.3 (13.2) �1.2 (�9.6 to 7.2)FEV1 (% Predicted), mean (SD) 100.2 (12.4) 93.7 (10.7) �7.8 (�14.9 to �0.7)FVC (% Predicted) < 85, n (%) 11 (23) 9 (28) 1.4 (0.4e4.5)FEV1 (% Predicted) < 85, n (%) 4 (8) 8 (25) 2.0 (0.4e9.3)FEF25e75% (L/s) 2.9 (1.2) 2.2 (1.5) �1.0 (�2.3 to 0.3)

FEV1/FVC < 0.7, n (%) 0 (0) 3 (11) e

FVC postBD - FVC preBD (mL), mean (SD) �25.2 (17.6) 5.6 (20.5) 9.8 (�127e146)FEV1 postBD - FEV1 preBD (mL), mean (SD) 36.2 (12.6) 62.8 (14.7) 1.5 (�83e86)FVC postBD/FVC preBD (%), mean (SD) 99.1 (5.2) 100.4 (6.0) 1.1 (�3.0e5.1)FEV1 postBD/FEV1 preBD (%), mean (SD) 101.4 (4.6) 102.7 (5.7) 0.2 (�2.9e3.3)Bronchodilator challenge test cutpoints, n (%)C: 10% change in FEV1 or FVC 2 (5) 3 (11) 2.7 (0.5e14.1)D: 150 ml change in FEV1 or FVC 8 (19) 6 (24) 1.2 (0.3e4.2)

Bronchial hyperesponsiveness (PD20 < 1 mg), n (%)a 0 (0) 6 (55) e

Bronchial hyperesponsiveness (PD20 <2 mg), n (%)a 3 (27) 9 (82) 8.2 (0.7e97)a Metacholine challenge test. N Z 11 controls and 11 cases.b Coefficient and 95%CI from linear regression models adjusted for age, height, sex, and packs-year smoked.c Spirometry 15 min after the inhalation of 400 mg of salbutamol: n (controls, cases) Z 44, 26.

678 D. Vizcaya et al.

Discussion

This study suggests that cleaning workers with asthma orasthma symptoms are characterised by non-reversible air-way obstruction and non-eosinophilic inflammation. Airwayobstruction is indicated by the results of basal spirometryand bronchodilator challenge testing; non-eosinophilicinflammation is indicated by the results of FeNO testing.We found no differences between cases and controls inlevels of biomarkers of oxidative stress or remodelling of theairways. Asthmatic cleaners are more often atopic and showhigher serum total IgE levels than asymptomatic cleaners.

This is the first study that evaluated thoroughly thebiological characteristics of asthma in cleaning workers.The lower values of post-bronchodilator FEV1, FEF25e75%

and the ratio FEV1/FVC of cases compared to controlssuggest a functional phenotype with airflow limitation. Thereversibility of bronchial obstruction was not different be-tween cases and controls. It has been previously reportedthat non-eosinophilic asthmatics with low FEV1, are lesscommonly bronchodilator test positive compared to eosi-nophilic asthmatics.27 Thickening of the reticular basalmembrane, considered a histopathologic feature of occu-pational asthma, may be an alternative explanation for thenon-reversibility of airways obstruction.28 However, thenon-reversibility obstruction of the airways may be due tothe presence of other related diseases such as chronicobstructive pulmonary disease (COPD) among symptomaticcases. We evaluated the proportion of cases and controls

with FEV1/FVC lower than 0.70, a diagnostic criterion forCOPD accepted by the international respiratory societies,29

and found that only three out of 42 cases met this criteria.Therefore, although it cannot be excluded that cases haveother diseases than asthma, our results are not incompat-ible with known asthma phenotypes. Cases also showeda higher prevalence of unspecific bronchial hyper-responsiveness, which was measured in a subsample of 11cases and 11 controls using the metacholine challenge test.No differences in age, basal lung function, sex and smokinghabit were found between eligible participants and non-participants of methacholine challenge testing.

Cases and controls showed, on average, very similarlevels of FeNO, even after adjusting for age, sex and smokingstatus. Additional adjustment for other known determinantssuch as atopy and body mass index yielded very similar re-sults, suggesting that these factors were not confounding therelationship between asthma and FeNO in our population.Limiting the analysis to cases with current asthma, or to in-dividuals with adult-onset asthma, did not change the re-sults. This lack of difference in FeNO suggests thateosinophilic inflammation does not play a predominant rolein asthma in cleaning workers. This result is compatible withprevious studies reporting that the non-eosinophilic asthmaphenotype is characteristic of irritant-induced asthma,which has been proposed as themainmechanistic hypothesisfor cleaning-related asthma and asthma-like disorders.14,30

The increased levels of IgE observed among cases com-pared to controls and the increased proportion of Phadiatop

Table 3 Biological characteristics of cases and controls.

Controls(n Z 51)

Cases(n Z 41)

GM ratioa

(95%CI)

GM GM

FeNO (ppb) 19.9 17.9 1.1 (0.8e1.3)Serum total [IgE] (IU/mL) 14.9 39.7 2.9 (1.5e5.6)Serum [SP-D] (ng/mL) 39.2 31.2 0.8 (0.5e1.1)Serum [CC16] (ng/mL) 6.8 5.9 0.9 (0.8e1.2)Exhaled breath condensate (pg/mL)

[8-isoprostane] 1.9 1.7 1.2 (0.7e1.8)[Cys-leukotrienes] 55.5 52.0 1.6 (0.7e3.9)

Exhaled breathcondensate

Lower limit ofdetection (pg/mL)

% Detectable % Detectable OR* (95%CI)

FGF 3.4 35% 41% 0.6 (0.2e1.7)IL-13 0.6 20% 32% 1.2 (0.4e3.8)TNF-a 0.7 18% 29% 1.3 (0.4e4.1)IFN-g 1.8 10% 5% 0.3 (0.0e2.1)VEGF 4.5 6% 5% 0.8 (0.1e6.6)IL-4 1.4 4% 5% n.a.IL-8 1.2 4% 5% n.a.Ip10 0.5 2% 2% n.a.IL-5 1.1 2% 5% n.a.IL-10 0.1 2% 5% n.a.IL-12 0.6 2% 5% n.a.IL-2 11.2 0% 2% n.a.

FeNO: Fraction of exhaled nitric oxide. SP-D: Surfactant Protein D. CC16: 16 kDa Clara Cell Protein. EBC: Exhaled Breath Condensate.FGF: basic Fibroblast Growth Factor. VEGF: Vascular Endothelial Growth Factor. TNF-a: Tumour Necrosis Factor alpha. IFN-g: Interferongamma. Ip10: IFN-geinduced protein. IL: Interleukin.GM: geometric mean. n.a.: not analysed (<5% of detectables).a Exponential of the coefficient from linear regression models of the log-transformed variable adjusted for age, sex and smoking

status.

Asthma in cleaning workers 679

test positives may be related to lung damage due to theinhalation of irritants, which facilitates an immunologicalresponse to sensitizers.13,14 In addition, a previous studyshowed an association between atopic sensitisation and theexposure to non-allergenic disinfectants.31 Alternatively,some cleaning products contain sensitizers, which mayexplain the increased levels of IgE.15 In a previous study wefound no association between atopy and asthma and/orchronic bronchitis in domestic cleaning workers, but highertotal serum IgE level.7 It is important to highlight thatphadiatop test is a method that assesses the levels ofspecific IgE levels for ten common aeroallergens, and mayhave a component of false positives.32 On the other hand,an international study with a caseecase design found thatasthmatic cleaners had less atopy than asthmatic officeworkers.33 In addition to the above mentioned causes forincreased IgE and atopy, it is important to take into accountthat a pre-existing immunological asthma may be aggra-vated after the exposure to irritants at work.34

Up to date there are no reference values for the studiedEBC biomarkers. For this reason, rather than descriptive weperformed a classic case-control analysis to evaluate theassociation of these biomarkers and cleaning-relatedasthma. Cases and controls showed similar levels of8-isoprostane in EBC, which according to the literature is inline with our results of FeNO.35 Cys-leukotrienes, growthfactors and cytokines levels were also similar in cases and

controls. The replication of the analyses restricted to caseswith either adult-onset asthma or current asthma did notshow any further association. It has been suggested thatirritant-induced asthma is usually less severe in terms ofcontrol of the disease than immunological asthma, and maybe an alternative explanation for the lack of associationsbetween respiratory biomarkers and asthma among clean-ing workers in our study.36

We evaluated both occupational and domestic exposureto cleaning products, and its association with asthmasymptoms and respiratory biomarkers. Occupational anddomestic use of multi-use products appeared as a con-sistent risk factor. In addition, multi-use products’ use inspray form, likely facilitates inhalatory exposure,11 wasstrongly associated with asthma. Multi-use product isa generic term that refers to complex formulae for cleaningproducts containing chemicals such as benzenesulfonic acidand aliphatic alcohols. We also found indications that theuse of irritants including ammonia and hypochloritebleach,7e9 polishes or waxes,37 glass cleaners, and dustmop products was associated with asthma.

The cross-sectional association between the use of spe-cific products and biomarkers were heterogeneous but,interestingly, occupational use of multi-use products wasassociated with increased levels of FeNO, total IgE and FGF.This is suggestive of an inflammatory profile more related toimmunological asthma.13,14 We assessed the associations of

Table 4 Occupational and domestic use of cleaning products in the previous year and asthma.

Occupational use in the last year Domestic use in the last year

Controls(n Z 53)

Cases(n Z 42)

ORa (95%CI) Controls(n Z 53)

Cases(n Z 42)

ORb (95%CI)

n (%) n (%) n (%) n (%)

Ammonia 10 (19) 12 (29) 2.7 (0.9e8.2) 14 (26) 13 (31) 1.1 (0.4e3.2)Bleach 47 (89) 39 (93) 1.1 (0.1e11) 41 (77) 37 (90) 4.0 (0.8e21)Degreasers 26 (49) 25 (59) 1.2 (0.5e3.0) 33 (62) 31 (76) 1.5 (0.5e4.6)Drain products 3 (6) 1 (2) 0.2 (0.0e2.9) 5 (9) 6 (15) 1.3 (0.3e6.0)Dust mop products 18 (34) 19 (45) 1.9 (0.7e5.2) 8 (15) 8 (20) 0.8 (0.2e3.1)Glass cleaners 19 (36) 17 (40) 1.0 (0.3e2.7) 28 (53) 31 (76) 3.3 (1.1e9.9)Hydrochloric acid 4 (8) 5 (12) 1.5 (0.3e7.7) 6 (11) 4 (10) 0.6 (0.1e3.0)Limescale removers 23 (43) 20 (48) 0.2 (0.1e0.7) 17 (32) 25 (61) 3.8 (0.8e19)Multi-use products 12 (23) 15 (33) 2.3 (0.7e7.0) 21 (40) 27 (66) 2.1 (0.8e5.8)Polishes and waxes 7 (13) 6 (14) 1.1 (0.2e5.2) 3 (6) 7 (17) 3.9 (0.8e19)Soaps or detergents 37 (70) 21 (50) 0.2 (0.1e0.7) 34 (64) 28 (68) 1.5 (0.5e4.5)Stain removers 5 (9) 4 (9) 0.8 (0.1e4.1) 5 (9) 11 (27) 2.7 (0.7e9.8)Spray or aerosol formMulti-use products 5 (9) 7 (17) 4.1 (1.0e18) n.a. n.a. n.a.Degreasers 8 (15) 9 (21) 1.1 (0.4e3.1) n.a. n.a. n.a.Dust mop products 16 (30) 17 (40) 1.5 (0.6e3.9) n.a. n.a. n.a.Limescale removers 4 (8) 4 (10) 1.5 (0.5e5.0) n.a. n.a. n.a.Glass cleaners 16 (31) 14 (33) 1.2 (0.3e5.9) n.a. n.a. n.a.Num different sprays

0 22 (42) 14 (33) 1 n.a. n.a. n.a.1e2 22 (42) 17 (40) 0.8 (0.3e2.4) n.a. n.a. n.a.3e5 9 (17) 11 (26) 2.1 (0.6e7.4) n.a. n.a. n.a.

n.a.: information not available.a Odd ratios for asthma and occupational use of cleaning products from logistic regression models adjusted for age, domestic use of the

product, sex and smoking habit. Reference category for each model: no occupational use of the product ever in the last year. Onedifferent model for each product.b Association between asthma and domestic use of cleaning products using logistic regression models adjusted for age, occupational

use of the product, sex and smoking habit. Reference category for each model: no domestic use of the product ever in the last year. Onedifferent model for each product.

680 D. Vizcaya et al.

biomarkers and products without avoiding any possiblemechanistic pathway, thuswedid not adjust for case-controlstatus. We initially performed a separate analysis for casesand controls, although it was limited by statistical power(Supplement Tables 5 and 6). Nevertheless, the associationsbetween multi-use products and FeNO, total IgE and FGFpointed in the same direction for cases and controls, sug-gesting a sub-clinical effect independent of asthma.

A particular strength of this study was the detailedassessment of functional and biological characteristics ofasthma. We evaluated biomarkers indicative of a variety ofpathophysiological processes including inflammation, oxi-dative stress, airways damage and affected permeability.Additional strengths included a confirmation of the case orcontrol status in an intermediate step between the initialquestionnaire and the clinic visit. The inclusion and exclu-sion criteria were largely based on reported respiratorysymptoms, which typically show a considerable variabilityover time when repeatedly assessed. This is related to theintermittent nature of the underlying respiratory condition(true variation) as well as to measurement error.38 Thus,potential misclassification of asthma status was reduced byfollowing the conservative approach in which both casesand controls met the inclusion criteria for case or control

status twice, approximately one year apart. As a result, ourfinal study population included cases with persistentasthma and/or asthma symptoms and controls withouttemporary respiratory symptoms.

There are a number of potential limitations that need tobe considered. First, the study population was relativelysmall. This affected the statistical power for detectingsmall differences between cases and controls and limitedthe assertiveness of the conclusions. This is due in part tothe cross-sectional-nested and workforce-based nature ofthe study, what made difficult approaching participants dueto an already limited population and a very restrictiveSpanish legislation on personal data protection that madeimpossible to access any workers’ personal data from thecompanies.39 However, our industry-based approach pro-vided us a less biased study population, with a wide rangeof occupational exposures. In our opinion, the large amountof biological and questionnaire data obtained from a highlyinaccessible population and the reduced misclassificationof asthma compensate the low sample size. Furthermore,our results are consistent with previous reports on asthmaand cleaning workers.1,5 Second, several cleaning productshave strong odours that may have been overreported byasthmatic cases and potentially lead to a bias of our results

Table 5 Associations between occupational and domestic use of cleaning products and biomarker levels among all cases and controls.

Continuous variables Categorical variables (detectable vs. non-detectable)

FeNO CC16 SP-D IgE 8-Isoprostane Cys-leukotrienes FGF TNF-a IL-13 IFN-g

GMRa (95%CI) GMRa (95%CI) GMRa (95%CI) GMRa (95%CI) GMRa (95%CI) GMRa (95%CI) ORb (95%CI) ORb (95%CI) ORb (95%CI) ORb (95%CI)

AmmoniaOccupational 0.9 (0.7e1.1) 1.0 (0.8e1.3) 1.0 (0.7e1.5) 1.4 (0.7e3.0) 0.9 (0.5e1.4) 1.2 (0.5e3.1) 0.7 (0.2e2.3) 1.4 (0.4e4.8) 0.7 (0.2e2.6) 0.9 (0.1e6.6)Domestic 1.1 (0.8e1.4) 0.9 (0.7e1.2) 1.0 (0.7e1.5) 0.6 (0.3e1.2) 1.7 (1.1e2.7) 0.4 (0.2e1.1) 1.4 (0.5e4.0) 1.5 (0.5e5.1) 2.5 (0.8e8.6) 1.6 (0.2e12.5)

BleachOccupational 1.2 (0.8e1.9) 1.4 (0.9e2.2) 1.6 (0.7e3.5) 2.2 (0.5e9.3) 0.8 (0.3e1.9) 1.2 (0.2e7.5) n.a. n.a. n.a. n.a.Domestic 0.9 (0.7e1.3) 0.9 (0.6e1.2) 0.9 (0.5e1.5) 1.0 (0.4e2.6) 0.8 (0.4e1.6) 0.9 (0.3e2.9) n.a. n.a. n.a. n.a.

DegreasersOccupational 1.1 (0.9e1.4) 0.9 (0.7e1.1) 0.8 (0.6e1.2) 1.8 (1.0e3.3) 0.5 (0.3e0.7) 0.7 (0.3e1.6) 0.8 (0.3e2.1) 0.6 (0.2e1.6) 0.7 (0.2e1.9) 0.2 (0.0e1.7)Domestic 1.1 (0.9e1.4) 1.0 (0.8e1.2) 1.1 (0.7e1.6) 0.9 (0.4e1.7) 0.7 (0.5e1.1) 0.8 (0.3e2.0) 1.1 (0.4e3.3) 1.7 (0.5e6.4) 1.1 (0.3e3.8) 0.3 (0.0e2.1)

Dust mop productsOccupational 1.0 (0.8e1.2) 0.9 (0.8e1.2) 1.0 (0.7e1.5) 1.8 (0.9e3.4) 0.9 (0.6e1.4) 1.5 (0.6e3.3) 1.1 (0.4e2.9) 1.8 (0.6e5.3) 0.7 (0.2e2.1) 3.2 (0.5e20.7)Domestic 1.1 (0.8e1.4) 1.0 (0.8e1.4) 0.8 (0.5e1.2) 1.0 (0.4e2.5) 1.5 (0.8e2.6) 0.6 (0.2e1.8) 0.8 (0.2e2.9) 0.6 (0.1e2.9) 0.5 (0.1e2.8) 0.5 (0.0e6.2)

Glass cleanersOccupational 1.3 (1.1e1.7) 0.9 (0.7e1.1) 1.0 (0.7e1.5) 1.4 (0.7e2.7) 1.1 (0.7e1.7) 1.7 (0.7e4.2) 1.4 (0.5e3.9) 3.1 (1.0e9.8) 1.2 (0.4e3.8) 1.0 (0.2e5.8)Domestic 0.9 (0.7e1.1) 1.0 (0.8e1.3) 0.9 (0.6e1.2) 1.3 (0.6e2.5) 0.9 (0.6e1.5) 0.6 (0.2e1.4) 1.3 (0.5e3.5) 1.7 (0.5e5.9) 1.4 (0.4e4.3) 0.7 (0.1e4.0)

Hydrochloric acidOccupational 1.2 (0.9e1.7) 0.9 (0.6e1.2) 1.0 (0.5e1.7) 3.5 (1.2e9.8) 1.0 (0.5e2.1) 1.1 (0.3e4.0) 0.8 (0.2e3.9) 0.4 (0.1e3.1) 0.4 (0.1e2.4) 0.8 (0.1e10.9)Domestic 0.7 (0.5e1.0) 0.8 (0.6e1.2) 1.1 (0.6e2.0) 0.8 (0.3e2.1) 0.9 (0.4e1.9) 0.5 (0.1e1.6) 2.0 (0.5e8.8) 7.3 (1.4e39.1) 4.0 (0.8e20.3) 2.6 (0.2e33.4)

Limescale removersOccupational 1.1 (0.9e1.4) 1.0 (0.8e1.3) 0.8 (0.6e1.2) 1.6 (0.8e3.0) 0.7 (0.4e1.1) 1.8 (0.8e4.2) 1.0 (0.4e2.6) 0.3 (0.1e1.1) 0.2 (0.1e0.8) 0.1 (0.0e1.5)Domestic 1.0 (0.8e1.3) 0.9 (0.7e1.1) 1.2 (0.8e1.7) 1.1 (0.6e2.2) 1.1 (0.7e1.7) 1.1 (0.5e2.6) 1.1 (0.4e2.9) 3.0 (0.9e9.9) 1.8 (0.6e5.9) 2.0 (0.3e14.0)

Multi-use productsOccupational 1.4 (1.1e1.7) 1.0 (0.8e1.2) 0.9 (0.6e1.4) 2.1 (1.0e4.2) 1.3 (0.8e2.1) 1.9 (0.8e4.9) 2.9 (1.0e8.7) 2.5 (0.8e7.7) 1.1 (0.4e3.4) 3.1 (0.5e17.4)Domestic 1.0 (0.8e1.2) 1.0 (0.8e1.3) 1.0 (0.7e1.5) 1.0 (0.5e2.0) 1.0 (0.6e1.5) 1.0 (0.4e2.2) 1.0 (0.4e2.8) 0.9 (0.3e2.6) 1.0 (0.3e2.9) 2.5 (0.4e17.6)

Polishes and waxesOccupational 1.6 (1.2e2.1) 1.2 (0.8e1.6) 1.4 (0.8e2.4) 1.5 (0.6e4.3) 1.4 (0.7e2.7) 1.0 (0.3e4.0) 0.7 (0.2e3.2) 1.6 (0.4e7.6) 1.8 (0.4e8.1) 1.0 (0.1e10.8)Domestic 1.0 (0.7e1.3) 0.8 (0.6e1.2) 1.2 (0.7e2.2) 0.7 (0.2e2.0) 1.5 (0.7e3.1) 1.2 (0.3e4.9) 1.1 (0.2e5.3) 1.3 (0.2e7.2) 2.0 (0.4e10.5) 2.9 (0.2e54.5)

Soaps or detergentsOccupational 0.9 (0.8e1.1) 1.0 (0.8e1.3) 0.8 (0.6e1.2) 0.6 (0.3e1.1) 0.8 (0.5e1.2) 0.7 (0.3e1.7) 1.1 (0.4e2.7) 0.7 (0.2e1.9) 0.6 (0.2e1.6) 0.2 (0.0e1.5)Domestic 1.2 (0.9e1.4) 1.0 (0.8e1.3) 1.2 (0.8e1.7) 0.9 (0.4e1.7) 1.3 (0.8e2.0) 0.9 (0.4e2.1) 1.3 (0.5e3.7) 0.6 (0.2e1.8) 0.9 (0.3e2.6) 0.2 (0.0e1.2)

n.a.: not analysed. FeNO: Fraction of exhaled NO. SP-D: Serum surfactant Protein D. CC16: Serum 16 kDa Clara Cell Protein. EBC: Exhaled Breath Condensate. FGF: EBC basic FibroblastGrowth Factor. VEGF: EBC Vascular Endothelial Growth Factor. TNF-a: EBC Tumour Necrosis Factor alpha. IFN-g: EBC Interferon gamma. Ip10: EBC IFN-geinduced protein. IL: EBCInterleukin. Bold indicates statistically significant association at the 95% confidence level.a Geometric means ratio and 95% confidence intervals from multivariable linear regression models of the log-transformed variables

including all cases and controls. Independent variables included in the models: domestic use of cleaning product, occupational use ofcleaning product, age, sex and smoking habit.b Odds ratio and 95% confidence intervals from multivariable logistic regression models including all cases and controls. Independent

variables included in the models: domestic use of cleaning product, occupational use of cleaning product, age, sex and smoking habit.

Asth

main

cleaningworke

rs681

682 D. Vizcaya et al.

away from the null.40 However, the differences in the risksfor domestic and professional use of some odorous products(e.g., ammonia) suggest that overreporting was unlikely tohave introduced a major bias. Third, in our study popula-tion, adult-onset asthma cases may be underrepresented ascompared to the previous study9 (Supplement Table 1).Underrepresented diagnosed asthma cases may have ledour results to an underestimation of the true differences inbiological and functional markers between cases and con-trols. Finally, we controlled potential confounding byadjustment for age, sex and smoking status in all analyses.Current smoking was strongly related to asthma in ourstudy, and rather than being a strong risk factor for asthmathis was likely driven by the selection criteria for controls,excluding those with chronic bronchitis symptoms (chroniccough and chronic phlegm). These symptoms were not partof the inclusion or exclusion criteria for cases and, asa result cases had more chronic bronchitis symptoms thancontrols. Therefore, the association between currentsmoking and asthma was determined by the cases withchronic bronchitis symptoms. Indeed, when excluding caseswith chronic bronchitis symptoms, the association betweencurrent smoking and case/control status attenuated. Whenthe main analyses were repeated without adjustment forsmoking status, or when excluding current and formersmokers, no major differences in the associations could befound. Thus, in spite of a potential overadjustment forsmoking, this is unlikely to have had a major influence onour findings.

In conclusion, this work contributes to disentangling thephysiological characteristics of the respiratory disordersassociated to cleaning-related exposures. Our results sug-gest that a non-reversible bronchial obstruction componentis present in cleaners with asthma and/or asthma symptomsand that eosinophilic inflammation and oxidative stress areunlikely to play a key role in the increased risk of asthmasymptoms. Our findings about occupational and domesticexposure to cleaning products are complementary to thosepublished previously in this study population9,41 and high-light the importance both for general public health andoccupational safety. Further studies on the mechanisms ofasthma in individuals exposed to cleaning agents arerecommended.

Funding

This study was funded by Instituto de Salud Carlos III/Eu-ropean Regional Development Fund, grant number PI 06/1378. The authors acknowledge partial funding to thisresearch from the CIBER Epidemiologıa y Salud Publica(CIBERESP), Spain.

Conflict of interests

The authors declare that they have no conflict of interest.

Acknowledgements

The authors thank Maitane Perez-Olabarrıa and MercedesDiaz Castillo for technical support and Manolis Kogevinas

and Jordi Sunyer for critical review of the manuscript. Theauthors also thank the participant companies and workers.This study was funded by Instituto de Salud Carlos III/Eu-ropean Regional Development Fund, grant number PI 06/1378. The authors acknowledge partial funding to thisresearch from the CIBER Epidemiologıa y Salud Publica(CIBERESP), Spain.

Appendix A. Supplementary data

Supplementary data related to this article can be found athttp://dx.doi.org/10.1016/j.rmed.2013.01.011.

References

1. Jaakkola JJ, Jaakkola MS. Professional cleaning and asthma.Current Opinion in Allergy and Clinical Immunology 2006;6(2):85e90. Epub 2006/03/08.

2. Kogevinas M, Anto JM, Sunyer J, Tobias A, Kromhout H,Burney P. Occupational asthma in Europe and other industri-alised areas: a population-based study. European CommunityRespiratory Health Survey Study Group. Lancet 1999;353(9166):1750e4. Epub 1999/05/29.

3. Kogevinas M, Zock J-P, Jarvis D, Kromhout H, Lillienberg L,Plana E, et al. Exposure to substances in the workplace andnew-onset asthma: an international prospective population-based study (ECRHS-II). The Lancet 2007;370(9584):336e41.

4. Medina-Ramon M, Zock JP, Kogevinas M, Sunyer J, Anto JM.Asthma symptoms in women employed in domestic cleaning:a community based study. Thorax 2003;58:950e4.

5. Zock JP, Vizcaya D, Le Moual N. Update on asthma andcleaners. Current Opinion in Allergy and Clinical Immunology2010;10(2):114e20. Epub 2010/01/23.

6. Orriols R, Costa R, Albanell M, Alberti C, Castejon J, Monso E,et al. Reported occupational respiratory diseases in Catalonia.Occupational and Environmental Medicine 2006;63(4):255e60.Epub 2006/03/25.

7. Medina-Ramon M, Zock JP, Kogevinas M, Sunyer J, Torralba Y,Borrell A, et al. Asthma, chronic bronchitis, and exposure toirritant agents in occupational domestic cleaning: a nestedcase-control study. Occupational and Environmental Medicine2005;62(9):598e606. Epub 2005/08/20.

8. Mirabelli MC, Zock JP, Plana E, Anto JM, Benke G, Blanc PD,et al. Occupational risk factors for asthma among nurses andrelated healthcare professionals in an international study.Occupational and Environmental Medicine 2007;64(7):474e9.Epub 2007/03/03.

9. Vizcaya D, Mirabelli MC, Anto JM, Orriols R, Burgos F, Arjona L,et al. A workforce-based study of occupational exposures andasthma symptoms in cleaning workers. Occupational and Envi-ronmental Medicine 2011;68(12):914e9. Epub 2011/05/12.

10. Dumas O, Donnay C, Heederik DJ, Hery M, Choudat D,Kauffmann F, et al. Occupational exposure to cleaning prod-ucts and asthma in hospital workers. Occupational and Envi-ronmental Medicine 2012;69(12):883e9. Epub 2012/10/04.

11. Zock JP, Plana E, Jarvis D, Anto JM, Kromhout H, Kennedy SM,et al. The use of household cleaning sprays and adult asthma:an international longitudinal study. American Journal of Res-piratory and Critical Care Medicine 2007;176(8):735e41. Epub2007/06/23.

12. Le Moual N, Varraso R, Siroux V, Dumas O, Nadif R, Pin I, et al.Domestic use of cleaning sprays and asthma activity in fe-males. The European Respiratory Journal: Official Journal ofthe European Society for Clinical Respiratory Physiology 2012;40(6):1381e9. Epub 2012/04/13.

Asthma in cleaning workers 683

13. Maestrelli P, Boschetto P, Fabbri LM, Mapp CE. Mechanisms ofoccupational asthma. The Journal of Allergy and ClinicalImmunology 2009;123(3):531e42. quiz 43-4. Epub 2009/03/14.

14. Douwes J, Gibson P, Pekkanen J, Pearce N. Non-eosinophilicasthma: importance and possible mechanisms. Thorax 2002;57(7):643e8. Epub 2002/07/04.

15. Quirce S, Barranco P. Cleaning agents and asthma. Journal ofInvestigational Allergology and Clinical Immunology 2010;20(7):542e50. quiz p. 2, following 50. [Epub 2011/02/15].

16. Corradi M, Gergelova P, Mutti A. Use of exhaled breath con-densate to investigate occupational lung diseases. CurrentOpinion in Allergy and Clinical Immunology 2010;10(2):93e8.Epub 2009/12/10.

17. Stewart L, Katial R. Exhaled nitric oxide. Immunology and Al-lergy Clinics of North America 2007;27(4):571e86. v. Epub2007/11/13.

18. Broeckaert F, Bernard A. Clara cell secretory protein (CC16):characteristics and perspectives as lung peripheral biomarker.Clinical and Experimental Allergy: Journal of the British So-ciety for Allergy and Clinical Immunology 2000;30(4):469e75.Epub 2000/03/16.

19. Burney PGJ, Luczynska C, Chinn S, Jarvis D. The Europeancommunity respiratory health survey. European RespiratoryJournal 1994;7(5):954e60.

20. Miller MR, Hankinson J, Brusasco V, Burgos F, Casaburi R,Coates A, et al. Standardisation of spirometry. The EuropeanRespiratory Journal: Official Journal of the European Societyfor Clinical Respiratory Physiology 2005;26(2):319e38. Epub2005/08/02.

21. Pellegrino R, Viegi G, Brusasco V, Crapo RO, Burgos F,Casaburi R, et al. Interpretative strategies for lung functiontests. The European Respiratory Journal: Official Journal ofthe European Society for Clinical Respiratory Physiology 2005;26(5):948e68. Epub 2005/11/03.

22. American Thoracic S, European Respiratory S. ATS/ERS rec-ommendations for standardized procedures for the online andoffline measurement of exhaled lower respiratory nitric oxideand nasal nitric oxide. American Journal of Respiratory andCritical Care Medicine 2005;171(8):912e30. Epub 2005/04/09.

23. Horvath I, Hunt J, Barnes PJ, Alving K, Antczak A, Baraldi E,et al. Exhaled breath condensate: methodological recom-mendations and unresolved questions. The European Respira-tory Journal: Official Journal of the European Society forClinical Respiratory Physiology 2005;26(3):523e48. Epub2005/09/02.

24. Font-Ribera L, Kogevinas M, Zock JP, Gomez FP, Barreiro E,Nieuwenhuijsen MJ, et al. Short-term changes in respiratorybiomarkers after swimming in a chlorinated pool. Environ-mental Health Perspectives 2010;118(11):1538e44. Epub2010/09/14.

25. Rodriguez-Trigo G, Zock JP, Pozo-Rodriguez F, Gomez FP,Monyarch G, Bouso L, et al. Health changes in fishermen 2years after clean-up of the Prestige oil spill. Annals of InternalMedicine 2010;153(8):489e98. Epub 2010/08/25.

26. Sunyer J, Pekkanen J, Garcia-Esteban R, Svanes C, Kunzli N,Janson C, et al. Asthma score: predictive ability and risk fac-tors. Allergy 2007;62(2):142e8. Epub 2007/02/15.

27. Wenzel SE, Schwartz LB, Langmack EL, Halliday JL,Trudeau JB, Gibbs RL, et al. Evidence that severe asthma canbe divided pathologically into two inflammatory subtypes with

distinct physiologic and clinical characteristics. AmericanJournal of Respiratory and Critical Care Medicine 1999;160(3):1001e8. Epub 1999/09/03.

28. Mapp CE, Boschetto P, Maestrelli P, Fabbri LM. Occupationalasthma. American Journal of Respiratory and Critical CareMedicine 2005;172(3):280e305. Epub 2005/04/30.

29. Rabe KF, Hurd S, Anzueto A, Barnes PJ, Buist SA, Calverley P,et al. Global strategy for the diagnosis, management, andprevention of chronic obstructive pulmonary disease: GOLDexecutive summary. American Journal of Respiratory andCritical Care Medicine 2007;176(6):532e55. Epub 2007/05/18.

30. Quirce S, Gala G, Perez-Camo I, Sanchez-Fernandez C,Pacheco A, Losada E. Irritant-induced asthma: clinical andfunctional aspects. The Journal of Asthma: Official Journal ofthe Association for the Care of Asthma 2000;37(3):267e74.Epub 2000/06/01.

31. Preller L, Doekes G, Heederik D, Vermeulen R, Vogelzang PFJ,Boleij JSM. Disinfectant use as a risk factor for atopic sensiti-zation and symptoms consistent with asthma: an epidemio-logical study. European Respiratory Journal 1996;9(7):1407e13.

32. Vidal C, Gude F, Boquete O, Fernandez-Merino MC, Meijide LM,Rey J, et al. Evaluation of the phadiatop test in the diagnosis ofallergic sensitization in a general adult population. J InvestigAllergol Clin Immunol 2005;15(2):124e30. Epub 2005/07/29.

33. Zock JP, Kogevinas M, Sunyer J, Jarvis D, Toren K, Anto JM,et al. Asthma characteristics in cleaning workers, workers inother risk jobs and office workers. The European RespiratoryJournal: Official Journal of the European Society for ClinicalRespiratory Physiology 2002;20(3):679e85. Epub 2002/10/03.

34. Tarlo SM, Liss G, Corey P, Broder I. A workers’ compensationclaim population for occupational asthma: comparison ofsubgroups. Chest 1995;107(3):634e41.

35. Montuschi P, Corradi M, Ciabattoni G, Nightingale J,Kharitonov SA, Barnes PJ. Increased 8-isoprostane, a marker ofoxidative stress, in exhaled condensate of asthma patients.American Journal of Respiratory and Critical Care Medicine1999;160(1):216e20. Epub 1999/07/03.

36. Wenzel SE. Asthma: defining of the persistent adult pheno-types. Lancet 2006;368(9537):804e13. Epub 2006/08/29.

37. Obadia M, Liss GM, Lou W, Purdham J, Tarlo SM. Relationshipsbetween asthma and work exposures among non-domesticcleaners in Ontario. American Journal of Industrial Medicine2009;52(9):716e23. Epub 2009/07/18.

38. Chinn S, Jarvis D, Burney P, Luczynska C, Ackermann-Liebrich U, Anto JM, et al. Increase in diagnosed asthma butnot in symptoms in the European Community RespiratoryHealth Survey. Thorax 2004;59(8):646e51. Epub 2004/07/30.

39. Spain Ley Organica 15/1999, de 13 de diciembre, de Proteccionde Datos de Caracter Personal. Boletın Oficial del Estado 14-12-1999:43088e99.

40. Delclos GL, Gimeno D, Arif AA, Benavides FG, Zock JP. Occu-pational exposures and asthma in health-care workers: com-parison of self-reports with a workplace-specific job exposurematrix. American Journal of Epidemiology 2009;169(5):581e7.Epub 2009/01/08.

41. Mirabelli MC, Vizcaya D, Marti Margarit A, Anto JM, Arjona L,Barreiro E, et al. Occupational risk factors for hand dermatitisamong professional cleaners in Spain. Contact Dermatitis 2012;66(4):188e96. Epub 2012/01/25.