INTERNATIONAL SCENE

Prevalence of Glucose Intolerance inUrban and Rural Communities in SaudiArabiaA.R. Al-Nuaim

King Khalid University Hospital, King Saud University, Riyadh, SaudiArabia

The prevalence and associated factors for glucose intolerance among Saudi populationsin urban and rural communities were investigated among 13 177 subjects, 15 years andover, from different regions of Saudi Arabia. The data were standardized using the knownage structure of the Saudi population. The sample was randomly selected from subjectswho participated in the National Epidemiological Household Study for Chronic MetabolicDiseases. Medical and social history was ascertained from all the study population duringhouse visits. All subjects were then requested to attend a local primary care centre forphysical examination and phlebotomy for measurement of random plasma glucose (RPG).A 75 g oral glucose tolerance test was employed for subjects with borderline values.WHO criteria for diagnosis of diabetes mellitus (DM) and impaired glucose tolerance(IGT) were applied. Mean RPG from the urban population was significantly higher thanin the rural population. Age adjusted prevalence of DM was significantly higher in theurban population (males 12 %, 95 % CI 11–13 and females 14 %, 95 % CI 13–15) thanin the rural population (males 7 %, 95 % CI 7–8 and females 7.7 %, 95 % CI 7–9) andis among the highest in the world. The prevalence of DM increased with age. The lowestand highest prevalences of DM in the urban population were 2 % for subjects aged 15–20 years and 49 % for female subjects aged 51–60 years. The lowest and highestprevalences of DM among rural population were 1 % for subjects aged 15–20 years and29 % for female subjects over the age of 60 years. Fifty-six per cent of diabetic patientswere newly diagnosed at the time of the study. Age adjusted prevalence of IGT was notsignificantly higher in the urban population. The highest prevalence of obesity, BMI.30,was among urban female subjects. Age, obesity, and family history of DM were associatedwith DM. Considering the young nature of Saudi population, the prevalence is expectedto increase in the near future. There is a need to develop a multi-disciplinary approachfor the general population with special attention to female subjects for prevention throughcontrolling modifiable risk factors such as obesity and sedentary life style, improvingglycemic control of the diabetic population, and early identification and treatment ofdiabetic complications. 1997 by John Wiley & Sons, Ltd.

Diabet. Med. 14: 595–602

No of Figures: 2. No of Tables: 5. No of Refs: 20

KEY WORDS Prevalence Diabetes mellitus Urban and rural communities Saudi Arabia

Received 26 July 1996; revised 22 January 1997; accepted 25 January 1997

Saudi Arabia is located in the Middle East betweenIntroductionlongitudes 35°–56° East and latitudes 16°–32° North.Most of the surface area (which exceeds 2 240 000 km2)Diabetes mellitus (DM) is a chronic metabolic disease

and a major risk factor for coronary heart disease (CHD).1 is desert. According to the Saudi National PopulationCensus (SNPC) held in 1992,7 the population was 13.2It is the leading cause of blindness and renal failure in

the developed countries. Establishing the prevalence of million, of which 70 % is under 30 years of age. SaudiArabia is divided into five administrative regions: Central,DM in a given community is essential for estimating the

financial requirements of implementing prevention and East, West, North, and South. There is a regional variationwith respect to the extent of urbanization and life style,care programmes as well as being necessary for monitor-

ing the effectiveness of such programmes. Over the last for example, the Eastern, oil producing region, was thefirst to adopt the western life style, compared with the20 years, DM has become more prevalent in developing

countries and certain ethnic minorities.2–6 Southern region which is the least urbanized and stilldepends largely on agriculture. Saudi Arabia has gonethrough significant economic development over the last

* Correspondence to: Professor Abdul Rahman Al-Nuaim, King Khalid30 years, which has resulted in financial, demographic,University Hospital, King Saud University, P.O. Box 2925 (38), Riyadh

11461, Saudi Arabia and epidemiological changes. Society has become more

595CCC 0742–3071/97/070595–08$17.50 1997 by John Wiley & Sons, Ltd. DIABETIC MEDICINE, 1997; 14: 595–602

INTERNATIONAL SCENEaffluent, with new nutritional habits and westernized life and every third house then visited. All subjects over the

age of 15 years in these houses were asked to participatestyle. Significant internal migration from rural to urbancommunities occurred in the early years of economic in this study.

Family history of first and second degree relatives wasdevelopment and has slowed down more recently. Manyendemic diseases have been eradicated but have been recorded. Education level was classified as illiterate,

elementary, junior high school, high school and universityreplaced by chronic non-communicable diseases suchas hypertension, cardiovascular diseases, and diabetes education. Personal income was sought and recorded.

All the interviewed subjects were requested to attendmellitus.8

The prevalence and associated factors of glucose the Primary Care Centre for weight, height, and bloodmeasurements. Random blood samples were drawn inintolerance in Saudi Arabia among adult subjects living

in urban and rural communities is the subject of this an EDTA tube, centrifuged and the plasma stored frozentill completion of the target sample. The samples werestudy. Cross-sectional household epidemiological survey

was conducted through different regions of Saudi Arabia. sent frozen to the central laboratory at the King SaudUniversity, Riyadh; they were stored at −20 °C untilWHO criteria for diagnosis of DM have been used9 to

allow comparison with other populations. assayed. Upon completion of the target number (100and 50 subjects per physician in each district in city orvillage), records were sent to central office in Riyadh forPatients and Methodsdata entry. A computer program was designed, using D-Base IV software package for data entry. After theStudy Populationcompletion of data entry, the final sample, adjustedfor gender, age, regional, urban vs rural populationThe National Epidemiological Household Survey for

chronic metabolic diseases, including DM was conducted distribution in accordance with the SNPC data, wasrandomly selected using Statpac Gold software packageamong Saudi subjects over the age of 15 years old in

regions of Saudi Arabia between 1990 and 1993. This from records of different region and age groups.9

national study had several objectives, among themestimation of the prevalence of DM, hypercholesterolae- Methodmia, and obesity at national, regional, and tribal levels.Based on this, a sample size of 37 000 was calculated Plasma glucose was analysed by a glucose oxidase

method (Beckman Paragon, Fullerton, CA, USA). Theto fulfil all the objectives. The objective of this paperwas to study the prevalence and risk factors of glucose intra- and inter-assay coefficients of variation were 1.7 %

and 2.6 %, respectively. Random plasma glucose (RPG)intolerance in urban and rural communities. A multistagestratified cluster random sampling technique was used was interpreted as per WHO criteria.9 Group 1:

,5.5 mmol l−1, no diabetes; Group 2: .11.1 mmol l−1,for selection of the study population. The assignedpopulation sample of the study was distributed between diagnosis of DM; Group 3: 5.5–11.1 mmol l−1, Borderline,

requiring oral glucose tolerance test (OGTT). OGTT wasthe different regions in accordance with the regionalpopulation distribution as per the SNPC. There was an performed using 75 g anhydrous glucose, diluted in

400 ml water and consumed over 4 min. Subjects forinitial adjustment for area type, urban vs rural, populationdistribution in each region as per the SNPC. Cities and OGTT were advised to fast overnight and 2 h post-

glucose load blood was assayed as described. OGTTvillages of each region were listed and then randomselection of cities and villages was conducted in was interpreted as follows: plasma glucose at 2 h .11.1

would qualify for diagnosis of DM, plasma glucose .7.8accordacne with the allocated share of each region inthe national sample. The administrative maps of the and ,11.1 at 2 h would qualify for diagnosis of impaired

glucose tolerance test (IGT).selected cities and villages were reviewed, then therewas random selection of number of districts in these In order to calculate the estimate of error among

subjects with plasma glucose below 5.5 mmol l−1, acities and villages.Primary care physicians working in these localities random selection of 250 subjects with plasma glucose

below 5.5 mmol l−1 from different regions, gender, andwere assembled and given orientation lectures andworkshops in different aspects of the study, including age groups was subjected to OGTT.collecting personal, demographic, social and medical-history data, specifically relating to DM. Medical records Overweight and Obesityof patients self-reporting DM were reviewed to confirmthe diagnosis. The primary care physicians were also WHO criteria were adopted, and overweight was defined

as BMI 25–30, while obesity was defined as BMI.30.10trained on proper methods of blood handling, andmeasurement of height and weight which was done atthe local primary care centre usually 2–3 days after the Statistical Analysisinitial home visit.

A cluster sampling procedure was adopted, where Data were analysed using Statpac Gold software pack-age.11 Results were presented as means and percentages,every third parallel street in each locality was selected

596 A.R. AL-NUAIM

INTERNATIONAL SCENEin tables and graphs. Mann–Whitney U test was used Resultsfor comparison between medians of continuous variablesand a p-value less than 0.05 was considered statisti- Study Populationcally significant.

Not all subjects with borderline value of RPG attended Among subjects who were invited initially to participatein the national epidemiological study, 92 % respondedfor OGTT. To study whether or not there was a difference

between the OGTT responders and the OGTT non- for home interview at their houses and 69 % attendedfor the physical examination and blood investigation.responders, we randomly selected 500 subjects from

each group, matched for gender, age and region. The final adjusted sample for this study was 13 177, andas this sample was randomly selected after completionVariables such as income, family or personal history of

DM, body mass index, and DM diagnosed by RPG were of data entry, the response rate for attending the PrimaryCare Clinic and for initial blood investigation was 91 %.compared between the two groups.

To calculate the age adjusted prevalence of DM or The demographic characteristic of the adjusted sampleis shown in Table 1. Totals of 1600 (25 %) male subjectsIGT, a direct standardization adjustment for age was

employed in accordance with the following national and 1652 (28 %) female subjects had RPG in the rangeof 5.5–11.1 mmol l−1. Of these, only 57 % urban andpopulation age distribution7 of Saudi subjects: 15–20

years (23 %), 21–30 years (34 %), 31–40 years (19 %), 69 % rural residents attended for OGTT, with 60 % ofthe target group overall having OGTT. The comparison41–50 years (10 %), 51–60 years (8 %) and .60 years

(6 %). The use of the known age structure of Saudi between the subgroup of OGTT responders and OGTTnon-responders showed no significant difference withpopulation permits direct adjustment of age-related

mortality and morbidity, rather than using indirect respect to age, gender, income, family or present historyof DM, body mass index or DM diagnosed by RPG. Theadjustments based on the world standard population.

Direct standardized adjustment was employed when randomly selected subjects with RPG below 5.5 mmol l−1

who underwent OGTT showed diabetic response (RPGcalculating the prevalence of DM and IGT, specificallywhen adding up the estimated number of subjects with at 2 h .11.1 mmol l−1) among 5 % and 8 % of male and

female subjects respectively, and IGT response (RPG atRPG below 5.5 mmol l−1 who would exhibit diabetic orIGT responses should they be challenged with OGTT. 2 h 7.8–11.1 mmol l−1) among 8 % and 13 % of male

and female subjects respectively. This estimate of errorMultiple regression analysis using the BMPD statisticalsoftware12 was conducted to assess the relative likelihood increased with age.that subjects with certain socio-demographic character-istics will have diabetes mellitus. The variables in the

Table 1. Demographic characteristics of the national diabetesregression model were examined in a stepwise forwardmellitus study in Saudi Arabia

selection procedure. The interaction between explanatoryvariables was assessed and was synergistically significant Variable Category Number (%)between family history and increasing age with calculatedsynergistic index of Rothman .1. Comparison was made Gender Male 6873 (52)

Female 6304 (48)between urban and rural population after being analysedAge groups 15–20 3012 (23)separately. The independent effects of the subjects’

21–30 6304 (34)characteristics were adjusted for other variables and all31–40 3012 (19)

the variables were treated as categorical. 41–50 4431 (10)51–60 999 (8)60. 774 (6)Overall Prevalence of DM

Regions West 3792 (29)Central 3293 (25)The prevalence of DM was calculated based on addingSouth 3179 (24)

up the number of subjects with confirmed self-reporting East 1942 (15)DM to the number of subjects with random plasma North 971 (7)

Area typeglucose .11.1 mmol l−1. The per cent of subjects whoA. National Urban 7495 (57)exhibited diabetic response on OGTT was used to

Rural 5682 (43)calculate the number of subjects who would exhibitB. Regional West Urban 1861 (49)

diabetic response among subjects with borderline serum Rural 1931 (51)glucose, who did not show up for OGTT. This total Central Urban 2353 (71)

Rural 940 (29)number of borderline subjects with diabetic responseSouth Urban 1506 (47)who either showed up for OGTT or did not show up

Rural 1673 (53)for OGTT was added to the first two groups of diabeticEast Urban 1183 (61)

subjects, either by history or by random serum plasma Rural 759 (39)glucose. The overall prevalence of DM was calculated North Urban 592 (61)

Rural 379 (39)after taking into consideration the estimate of erroramong subjects with plasma glucose ,5.5 mmol l−1.

597GLUCOSE INTOLERANCE IN URBAN AND RURAL SAUDI ARABIA

INTERNATIONAL SCENEcategory of subjects was calculated with respect to ageRural vs Urban Populations(Figure 2). Means of RPG for urban subjects were higherthan for rural subjects, whether for male or femaleThe rural population was older than the urban population.

The difference, however, was not significant. The urban subjects, in all age groups. This difference, however,reached a significant level only in the age group of 31–population was significantly taller for both male and

female subjects (p = 0.0001), and significantly heavier 40 years among female subjects (p = 0.02). There was aprogressive increase in means of RPG with age for allfor both male and female subjects (p = 0.0001). Mean

BMI for the urban population was significantly higher the subject categories till the 6th decade, followed by adecrease in means of RPG. There was a differencethan for the rural population for male and femal subjects

(p = 0.0001) and the prevalence of overweight and obesity between means of RPG for each category of subject,when compared with the mean of RPG of the consecutivewas significantly higher among the urban population for

male and female subjects (p = 0.0001; Table 2). age group. However, this difference was not significantwithin all the age groups and was more likely to besignificant in the early rather than late age groups.Plasma Glucose Level Distribution

Median random plasma blood glucose concentration Prevalence of Diabetes Mellitus and(RPG) for the urban population was significantly higher Impaired Glucose Tolerancethan for rural population for male and female subjects(p = 0.0001). The median 2 h post-glucose load plasma Age-adjusted prevalence of diabetes mellitus (DM) was

significantly higher for the urban population, whenglucose during OGTT was also significantly higher forthe urban population than for the rural population for compared with the rural population, whether for male

or female subjects. The prevalence of impaired glucoseboth male and female subjects (p = 0.0001; Table 2).The shape of distribution of means of RPG for all the tolerance (IGT) was higher, however, not significantly

so for the urban population, either male or femalefour categories is shown in Figure 1 with right skewnessshift of the curve for urban population. (Table 3).

A significant proportion of diabetic patients in theMeans (SD) of plasma RPG concentration for each

Table 2. Comparison of the urban and rural populations

Variable Male Female

Urban Rural p-value Urban Rural p-value

Number of records 3952 2921 3543 2761Age (yr)

Range 15–100 15–90 15–95 15–10025th 21 21 22 22Median 29 30 ,0.0001 28 29 ,0.000175th 40 42 40 40

Height (cm)Range 106–197 100–196 100–187 103–184Mean (SD) 165 (8) 164 (9) ,0.0001 154 (7) 153 (7) ,0.0001

Weight (kg)Range 28–140 21–135 28–155 21–136Mean (SD) 69 (16) 64 (15) ,0.0001 64 (15) 58 (14) ,0.0001

BMI (kg m−2)Range 11–68 11–54 11–70 11–72Mean (SD) 25 (5) 24 (5) ,0.0001 27 (6) 25 (6) ,0.0001

OverweightPrevalence (%) 31 26 ,0.0001 29 24 ,0.0001(BMI 25–30)

ObesityPrevalence 18 12 ,0.001 28 18 ,0.0001(%) (BMI.30)

Number of records 3508 2752 3087 2602Random blood glucose mmol l−1

Range 2–34 2–40 2–40 2–38Median 5.1 4.5 ,0.0001 4.9 4.5 ,0.0001

Number of records 532 427 550 4412-h Post-OGTT glucose (mmol l−1)

Range 2–38 2–26 2–30 2–30Median 5.5 5.0 ,0.0001 5.6 5.4 ,0.0001

598 A.R. AL-NUAIM

INTERNATIONAL SCENE

Figure 1. Frequency distribution of random plasma glucose (RPG) among the urban and rural populations in Saudi Arabia. —■—Male urban, —❒— male rural, —●— female urban, —❍— female rural

Figure 2. Mean (SD) of random plasma glucose (RPG) among Saudi subjects living in urban and rural communities with respectto age groups. —■— Male urban, —❒— male rural, —●— female urban, —❍— female rural

Table 3. Age-adjusted prevalence (%) and 95 % confidence interval of total, self-reported DM, DM (RPG), DM(OGTT), and IGT in urban and rural populations

Variable Male Female

Urban Rural p-value Urban Rural p-value% (95 % CI) % (95 % CI) % (95 % CI) % (95 % CI)

Number of records 3952 2921 3543 2761Total DM 12 (11–13) 7 (6.3–8.1) 0.0001 14 (13–15) 8 (7–9) 0.0001Self-report DM 7 (5–7) 4 (3.3–4.7) 0.0001 6 (5–7) 4 (3–5) 0.0001DM (RPG) 3 (3–4) 2 (2–3) 0.002 5 (5–7) 3 (2–3) 0.0001DM (OGTT) 2 (1–3) 1 (0.2–2) 0.33 2 (1–3) 1 (0.2–2) 0.25IGT 10 (9–11) 8 (7–9) 0.28 11 (10–12) 8 (7–9) 0.095Ratio of newly diagnosed 0.45 0.44 0.78 0.56 0.49 0.08DM/total DMRatio of IGT/total DM 0.8 1.1 0.056 0.8 1.04 0.056

different categories, 44–56 %, were newly diagnosed at reach significance. The highest ratio was found amongthe female urban population.the time of study. The ratio of newly diagnosed DM to

total DM patients was higher among female subjects for Prevalence of DM increased with age (Table 4). Thelowest and highest prevalences of DM among the urbanurban and rural populations, but the difference did not

599GLUCOSE INTOLERANCE IN URBAN AND RURAL SAUDI ARABIA

INTERNATIONAL SCENETable 4. Age-specific prevalence (%) and 95 % confidence interval of diabetes mellitus in urbanand rural population in Saudi Arabia

Age Gender Urban Rural(yr)

Number of % (95 % CI) Number of % (95 % CI)records records

15–20 Male 960 2 (1–3) 728 1 (0.3–2)Female 737 2 (1–3) 587 1 (0.2–2)

21–30 Male 1215 3 (2–4) 770 2 (1–3)Female 1438 5 (4–6) 1008 3 (2–4)

31–40 Male 834 9 (7–11) 652 4 (3–6)Female 586 15 (12–18) 517 8 (6–10)

41–50 Male 394 28 (24–32) 320 17 (13–21)Female 366 36 (31–41) 292 11 (7–15)

51–60 Male 288 39 (33–45) 255 22 (17–27)Female 247 49 (43–55) 209 26 (20–32)

.60 Male 261 35 (29–41) 196 19 (14–25)Female 169 42 (35–49) 148 29 (22–36)

population were 2 % and 49 % for subjects aged 15– Discussion20 years and for female subjects aged 51–60 years,respectively. The lowest and highest prevalences of DM There was a clear distinction among Saudi subjects living

in urban and rural communities with respect to glucoseamong the rural population were 1 % and 29 % forsubjects aged 15–20 years and for female subjects over levels and the prevalence of DM. Mean RPG concen-

trations whether for male or female subjects living inthe age of 60 years, respectively.urban communities were significantly higher than theircounterparts living in rural communities and overallFactors Associated with Diabetes Mellitusglucose intolerance was higher among the urban popu-lation, especially among female subjects. Right skewnessAnalysis of socio-demographic and clinical characteristics

for the study population have shown that family history of RPG distribution among the urban population indicatesthe larger segment of this group with higher values of RPG.of DM, age, BMI, and region of residence were important

and statistically significant variables associated with DM Family history of DM, age, overweight, and obesityare recognized risk factors for DM among different(p,0.05). Gender, income, and education did not attain

statistical significance at the 10 % level for inclusion in populations and our data show that populations inSaudi Arabia are no exception. This study showedthe model. The regression coefficients, adjusted odds

ratios, and 95 % confidence intervals are presented in intercorrelation between body mass index (BMI), familyhistory, and geographical regions, which precludes citingTable 5. The risk of diabetes mellitus significantly

increased with increasing age in the urban and rural obesity as the exclusive cause for the regional variationin the prevalence of DM. This was more visible in thepopulations. Subjects less than 20 years of age were

defined as the reference category in this analysis. Subjects urban population. However, the relative low and highrisk of incidence of the Southern and Eastern regions40 years and above had more than 10 times the risk of

diabetes mellitus (p , 0.01). Family history of diabetes can partially be explained by the fact that these tworegions have the lowest and highest prevalence of obesitymellitus was associated with a risk almost 3 fold higher

than in subjects with no family history of diabetes. Using in the country, 12 % and 20 % for male and femalesubjects, respectively, in the Southern region, and 23 %data from the West as a reference category, urban

subjects in the North, South, and Central regions and and 30 % for male and female subjects, respectively, inthe Eastern region.14those from the rural South had lower risks than the West.

Rural subjects in the North and Eastern regions had Several studies have been conducted, either amongpopulations of Arab countries, who share the Saudimarginally significant increased risk compared with the

Western region. Arabian, racial characteristics, nutritional and exercisehabits or among populations who have undergoneUsing the Western region as a reference category, the

Southern region had a low risk compared to the West, significant change in life style. We have limited thecomparison to the studies using WHO criteria forwhile the risk was higher in the East. Subjects who were

either midly or moderately obese had thrice the risk of diagnosis, to enable proper comparison between differentstudy populations. The prevalence rate of DM amongdiabetes in the rural areas. But the urban population

showed a marginally increased risk in the subjects with the Saudi population over 30 years of age in this currentstudy was the highest among all the cited studies, e.g.moderate obesity.

600 A.R. AL-NUAIM

INTERNATIONAL SCENETable 5. Comparison of multiple logistic regression coefficients, odds ratios, and 95 % confidenceintervals of potential factors associated with diabetes mellitus in urban and rural areas in Saudi Arabia

Variable Coefficient Odds-ratio 95 % confidence interval forodds ratio

Urban Rural Urban Rural Urban Rural

Constant −5.0 −5.4Age (yr)

,20 – – 1.0 1.0 – –20–29 0.5 0.4 1.7 1.4 1.1–2.7 0.7–2.830–39 1.7 1.3 5.2 3.8 3.3–8.2 2.0–7.240–49 2.8 2.2 17.2 8.7 11.0–27.1 4.6–16.550–59 3.4 2.7 29.3 14.6 18.5–46.4 7.7–27.560+ 3.3 2.9 27.0 19.0 16.8–43.5 10.0–36.2

Family history of DMNo – – 1.0 1.0 – –2.1–3.6Yes 0.9 1.0 2.4 2.7 2.0–3.0

RegionWest – – 1.0 1.0 – –North −0.1 0.4 0.9 1.6 0.6–1.4 1.0–2.5South −0.5 −0.3 0.6 0.7 0.5–0.8 0.5–1.0Central −0.3 0.1 0.8 1.1 0.6–1.0 0.7–1.6East 0.1 0.3 1.1 1.4 0.8–1.4 1.0–2.0

BMI (kg m−2),20 – – – – – –20–24 −0.1 0.2 0.9 1.3 0.6–4.0 0.7–2.225–30 0.2 1.1 1.3 2.9 0.8–1.9 1.7–5.0.30 0.4 1.1 1.5 3.0 1.0–2.3 1.7–5.3

9.2 % and 9.1 % among Tunisian male and female Acknowledgementssubjects over 30 years of age,15 8.4 % and 2.6 % among

The author acknowledges the contribution of primaryurban and rural male subjects, and 7.9 % and 1.6 %care physicians who were involved in this study inamong urban and rural female subjects among Indiandifferent regions of Saudi Arabia: the critical review bysubjects over 30 years of age.16 The prevalence of DMK. Al-Rubeaan, King Khalid University Hospital, Y. Al-among adult subjects of Pima Indians and AboriginesMazrou, Ministry of Health, and O. Al-Attas, College ofhas exceeded 40 %.17,18

Science, who were partners throughout this study, andThe rate of newly diagnosed DM was similar amongby M. Hammami, King Faisal Specialist Hospital anddifferent populations, 40 % and 54 % of urban and ruralResearch Center. The statistical support and advice of E.diabetic subjects, respectively, from Tunisia,15 35 % andA. Bamgboye, S. Mirdad, and F. Kwski, King Khalid76 % of urban and rural diabetic subjects from India,16

University Hospital is gratefully acknowledged, and the40 % and 30 % of male and female subjects fromprofessional excellence of M. E. Ali with respect toColumbia19 and 46 % of urban Brazil20 making the Saudicomputer programming, analysis, graphics and preparingArabian rates 50 % and 46% among the urban and ruralthe manuscript is highly appreciated.populations. Such rates are alarming, and probably reflect

the magnitude of the disease in the community and thelack of insight of the Saudi subjects on the importance

Referencesof control of risk factors such as obesity and sedentarylife style.

1. Reaven GM. Non-insulin-dependent diabetes mellitus,Finally, as 70 % of the Saudi population is ,30 years abnormal lipoprotein metabolism and atherosclerosis.of age, the prevalence of DM and IGT must be expected Metabolism 1987; 36: 1–8.

2. Dowse GK, Gareeboo H, Zimmet PZ, Alberti KGMM etto increase in the future in both urban and ruralal. High prevalence of NIDDM and impaired glucosecommunities. There is a need to define the risk factorstolerance in Indian, Creole, and Chinese Mauritians.for DM among Saudi subjects and develop a multi- Diabetes 1990; 39: 390–396.

disciplinary approach targeted to the general population, 3. King H, Zimmet P. Trends in the prevalence and incidencewith special attention to female subjects, for prevention. of diabetes: non insulin dependent diabetes mellitus.

World Health Stat 1988; 41: 190–196.

601GLUCOSE INTOLERANCE IN URBAN AND RURAL SAUDI ARABIA

INTERNATIONAL SCENE4. Zimmet P, Kelly west lecture 1991. Challenge in diabetes 14. Al-Nuaim AR, Al-Rubeaan K, Al-Mazrou Y, et al. High

epidemiology—from West to the rest. Diabetes Care 1992; prevalence of overweight and obesity in Saudi Arabia.15: 232–252. Int J Obes 1996; 20: 547–552.

5. McLarty DG, Swai ABM, Kitange HM, Masuki G, Mtinangi 15. Alwan A, King H. Diabetes in the eastern mediterraneanBL, Kilima DM et al. Prealence of diabetes and impaired region. Wld Hlth Statist Quart 1992; 45: 355–359.glucose tolerance in rural Tanzania. Lancet 1989; 1: 16. Ramachandran A, Dharmaraj D, Snehalatha C, Viswana-871–875. than M. Prevalence of glucose intolerance in Asian

6. King H, Rowers M. Diabetes in adults is now a third Indians. Diabetes Care 1992; 15: 1348–1355.world problem. Bull WHO 1991; 69: 643–648. 17. Ali O, Khalid BAK, Tan TT, Wu LL, Satinah O, et al.

7. Vital Statistics, Saudi Central Department for Statistics. Prevalence of NIDDM and impaired glucose tolerance inNational Population Census. Ministry of Planning: aborigines and Malaysia and their relationship to socio-Riyadh, 1992. demographic, health, and nutritional factors. Diabetes

8. Al-Balla SR, Bamgboye EA, Al-Balla SR, Al-Sekait M, Al- Care 1993; 16: 68–75.Rasheed R. Pattern of adult admission into medical wards 18. Saad MF, Knowler WC, Pettitt DJ, Nelson RG, Mott DM,of King Khalid University Hospital, Riyadh (1985–1990). Bennett PH. The natural history of impaired glucoseSaudi Med J 1993; 13: 8–13. tolerance in the Pima Indians. N Engl J Med 1988; 319:

9. WHO Study Group. Diabetes Mellitus. Technical Report 1500–1506.Series 727. Geneva: World Health Organization, 1985. 19. Aschner P, King H, Torrado MTP, Rodriguez BM. Glucose10. Bray GA. Definition, measurement and classification of

intolerance in Colombia. Diabetes Care 1993; 16: 90–93.the syndrome of obesity. Int J Obes 1978; 2: 99–112.20. Mallerbi DA, Franco LJ, The Brazillian Cooperative Group11. David S. Walonick Associates Inc. Statpac Gold Statistical

on the study of Diabetes Prevalence. Multicenter studyAnalysis Package—Version 3.2. 1988.of the prevalence of diabetes mellitus and impaired12. Dixon WJ (ed). BMDP Statistical Software, Inc. Berkeley,glucose tolerance in the urban Brazilian population ageUSA: University of California Press, 1990.30–69 years. Diabetes Care, 1992; 15: 1509–1516.13. Dowse GK, Zimmet PZ, King H. Relationship between

prevalence of impaired glucose tolerance and NIDDM ina population. Diabetes Care 1991; 14: 968–974.

602 A.R. AL-NUAIM