ORIGINAL COMMUNICATION Prevalence of overweight and obesity ...

European Journal of Clinical Nutrition (2003) 57, 471–474 ß 2003 Nature Publishing Group All rights reserved 0954–3007/03 $25.00 www.nature.com/ejcn

ORIGINAL COMMUNICATION Prevalence of overweight and obesity among Bahraini adolescents: a comparison between three different sets of criteria AM Al-Sendi2, P Shetty3 and AO Musaiger3* 1 College of Health Science, Ministry of Health, Bahrain; 2Food and Nutrition Division, Food and Agriculture Organisation, Rome, Italy; and 3Bahrain Center for Studies and Research, Bahrain

OBJECTIVE: The aim of this study is to determine the prevalence of overweight and obesity among Bahraini adolescents using three different sets of criteria=standards. DESIGN: Cross-sectional prevalence study. SETTING: Intermediate and secondary schools in Bahrain. SUBJECTS: The study included a population-representative sample of 506 Bahraini students (249 males and 257 females) between 12 and 17 y of age. The sample was selected using multistage stratified random sampling technique. MEASUREMENT: Anthropometric measurements including weight, height and triceps and subscapular skinfolds were taken on the adolescents. Age was verified against school records. To minimize inter-observer error, weight and height were taken by one person while skinfold was taken by two trained persons (one for each sex). RESULTS: The overall prevalence of obesity among Bahraini boys and girls was high, especially in girls. Obesity was highest (21% in males and 35% in females) when the WHO recommended criteria of BMI for age and skinfolds for age percentiles were applied and lowest (15% in boys and 18% in girls) when the age and sex specific BMI cut-off values of Cole et al were used. Compared with those of WHO criteria, estimates of overweight and obesity prevalence obtained with Must et al and Cole et al were generally close. CONCLUSIONS: Our data revealed a much higher prevalence rate of obesity in the Bahraini adolescent population than was previously reported, especially among girls. The BMI reference values of Must et al and that of Cole et al gave relatively similar estimates and appear to be more practical for use in surveys aimed at estimating the prevalence of overweight and obesity among adolescents than the WHO recommended composite criteria. SPONSORSHIP: Ministry of Health, Bahrain. European Journal of Clinical Nutrition (2003) 57, 471 – 474. doi:10.1038=sj=ejcn=1601560 Keywords: body mass index; adolescents; overweight; obesity

Introduction Childhood and adolescent obesity is a public health problem of increasing concern both in the developed countries and in countries undergoing cultural and economic transition. Data from the Arab Gulf Countries, including Bahrain, indicate an

*Correspondence: AO Musaiger, Director, Environmental and Biological Researches Programme, Bahrain Center for Studies and Research, PO Box 496, Manama, Bahrain. E-mail: [email protected] Guarantor: AM Al-Sendi Contributors: AM Al-Sendi, P Shetty, AO Musiager Received 29 November 2001; revised 11 June 2002; accepted 18 June 2002

increase in the prevalence of overweight and obesity among the adolescent population (Eid et al, 1986; Musaiger et al, 1993; Al-Nuaim et al, 1996). In Bahrain, several studies were carried out to estimate obesity in school children. These studies have relied on the weight for height tables, skinfold measurement or body mass index (BMI) criteria to determine overweight and obesity (Amine, 1980; Blair & Gregory, 1985; Musaiger et al, 1993). However, none of these studies used the recently proposed measurement of obesity by WHO (1995) and Cole et al (2000). Several criteria and various standards have been proposed to assess obesity in children and adolescents. One of the commonly used standard is that developed by Must et al

Prevalence of overweight in Bahraini adolescents AM Al-Sendi et al

472 (1991) which are based on data from the US (NHANES 1) surveys. The 85th and 95th percentiles of BMI have been suggested to define overweight and obesity in the adolescents (Himes & Dietz, 1994). The World Health Organization Expert Committee has recommended the use of both the BMI-for-age and the skinfold-for-age as the best indicators for the assessment for obesity in adolescents (WHO, 1995). The WHO has also recommended the use of Must et al reference data when the local reference is not available. More recently, Cole et al (2000) have proposed age- and sex-specific BMI cut-off points for overweight and obesity in children which are linked to the adult obesity cut-offs of 25 and 30 kg=m2 and based on pooled international data. Different criteria were used by many European researchers, as 85th and 97th percentiles were used to define overweight and obese individuals based on gender-age-specific BMI (Rolland-Cachera et al, 2001). Criteria for overweight and obesity vary in their applicability between different countries and across age groups and there is a need to determine which is the most appropriate standard or cut-off to use in assessing the degree of obesity in any population. The purpose of this study, therefore, is to determine the prevalence of obesity and overweight in Bahraini adolescents using three different sets of criteria= standards, namely those developed by Must et al (1991), the composite criteria of BMI for age and skinfold for age recommended by the WHO Expert Committee (WHO, 1995) and the recently published standards of Cole et al (2000).

adolescents and according to the method described by Fidanza (1991). Weight and height measurements were taken by a single person, while the skinfold thickness measurements were carried out by two skilled persons, one for each sex. Data were analysed to determine overweight and obesity rates, using the NHANES 1 reference data for BMI for age (Must et al, 1991), the age- and sex-specific BMI cut-off points reported by Cole et al (2000), and World Health Organization criteria of BMI-for-age and skinfold-for-age (WHO, 1995). Overweight and obesity status were determined by using the following criteria: 1. Using the age and sex specific BMI reference data of Must et al (1991) — overweight, 85th to < 95th percentile; obese, 95th percentile. 2. Using the BMI cut-off values of Cole et al (2000) — these are sex- and age-specific BMI cut-off points consistent with the internationally accepted cut-off points for adults overweight and obesity (ie BMI of 25 and 30, respectively). 3. Using the WHO expert committee composite criteria of BMI-for-age and skinfold-for-age (WHO, 1995) (based on Must et al, 1991) — overweight, BMI 85th to 94th percentile plus < 90th percentile of both the triceps and subscapular skinfold for age; obese, BMI 85th percentile plus 90th percentile of both the triceps and sub-scpaular skinfold for age.

Subjects and method The study population was Bahraini male and female intermediate and secondary school students aged 12 – 17 y. A population-representative sample of 506 students (249 males and 257 females) was selected using multi-stage stratified random sampling technique. The study was conducted between September and November 2000. The sample was selected from different educational grades within the elementary and secondary school levels and different geographic areas of Bahrain to ensure that the appropriate age groups and different strata of the target population are represented. Bahrain was first divided into its four governorates, and all the intermediate and secondary schools in each governorate were numbered and then a proportional sample of four or eight schools were randomly chosen from each governorate. After multistage sampling 25 schools, a systematic random sample of students was drawn from each school using the student name lists. Only Bahraini students were included in the study and a replacement sample was selected to substitute the non-Bahraini in each educational level. Students’ reported age and date of birth were verified against the school’s records, which in turn were based on the students’ birth certificates. Of the total sample, 506 students were included, and 34 students (6.7%) were excluded because of missing one or more data. Anthropometric measurements including weight, height, triceps and subscapular skinfold thickness were taken on the European Journal of Clinical Nutrition

Results Mean standard deviation and median of BMI for male and female Bahraini adolescents are presented in Table 1. As age increased the mean BMI increased. However at all ages females have greater BMI than males. The prevalences of overweight and obesity estimated using three different sets of criteria for the Bahraini boys and girls are shown in Table 2. The overall prevalence of obesity is relatively high, especially in females. Based on the WHO recommended criteria (WHO, 1995), about 21% of the males and 35% of the females were obese. The highest percentage of obesity in boys occurred at age 14 y (29%) and in girls occurred at age 16 y (42.5%). Overweight was more than twice (8%) as prevalent in boys as in girls (3.5%). Rates of overweight and obesity showed no specific pattern of variations with age, in both sexes. When Must et al (1991) reference values for overweight were applied, an overall prevalence of overweight, which was higher than that obtained with the WHO criteria, ranging from 13% in boys to 18.7% in girls. Relative to the Must et al (1991) indicator, Cole et al (2000) cut-offs gave slightly lower estimates of obesity prevalence, suggesting that age-specific BMI cut-off points were more stringent in identifying obese cases than the BMI centile-based standard. Nearly 15% of the boys and 18% of the girls were obese. In contrast, overall rates of overweight


473 Table 1 Mean, standard deviation and median of BMI for male and female Bahraini adolescents Age (y)

n

Mean BMI

s.d.

Median

Male 12 13 14 15 16 17

41 46 48 37 43 34

18.5 21.2 22.3 21.0 22.5 24.4

4.0 5.6 5.8 4.9 5.3 7.1

17.0 19.8 21.0 19.2 21.2 22.1

Female 12 13 14 15 16 17

50 43 40 41 40 43

21.5 21.8 22.0 24.7 25.3 25.0

6.0 4.4 5.0 7.0 7.7 7.0

20.5 21.3 20.9 23.3 23.7 23.4

were higher than those obtained by Must et al (1991) standard and ranged from 15% in males to 24.5% in females. In general, estimates of the prevalence of overweight and obesity obtained with Must et al (1991) and Cole et al (2000) were generally close. The proportions of obese categorized on the basis of these two criteria were identical at age 12, 15, 16 and 17 y in boys and at age 14 and 17 y in girls. Both criteria showed that obesity was most frequent at age 14 y in boys and at age 15 y in girls.

Discussion Despite the different estimates obtained by the three sets of criteria used, our findings indicate a high prevalence of

overweight and obesity among adolescents in Bahrain. The trend in obesity among the females studied was very similar to the data reported by Musaiger et al (2000), which were collected in 1992. However, our findings showed a remarkable decline in overweight and an increase in obesity using the same reference data (Must et al, 1991). The prevalence of overweight for the same age group (12 – 17 y) was 36.5% in 1992 and declined to 18.7% in our study, whereas that of obesity was 7% and increased to 20.2%. These observations suggest a trend of greater fat accumulation in adolescent girls during the 1992 – 2000 period. In boys, obesity level was also higher than that reported previously (Amine, 1980; Musaiger et al, 1993a,b) and appeared to be strongly influenced by the higher percentage of obesity among the 14-y-old males (29.2%). The observation that the highest proportion of obesity among girls occurred at age 16 y was in agreement with previous report (Musaiger et al, 2000). In both sexes and at almost all ages, the WHO composite criteria gave lower estimates of overweight but higher estimates of obesity prevalence than the indicators of either Must et al (1991) or Cole et al (2000). This appears to be related to the fact that, according to the WHO recommended criteria, overweight is defined on the basis of the BMI ( 85th BMI percentile) while obesity is defined on the basis of both BMI ( 85th BMI percentile) and skinfold ( 90th percentile of triceps and 90th percentile of subscapular skinfold). As a result, when these cut-offs were applied to the data set, participants with BMI values at or above the 85th percentile and with high subcutaneous fat were shifted from the overweight category to the obese category, resulting in lower estimates of overweight and higher estimates of obesity than those determined by other standards such as those of Must et al (1991) and Cole et al

Table 2 Prevalence (%) of overweight and obesity among male and female Bahraini adolescents using three reference data Age (y) Weight status

Reference data

Males

12

13

14

15

16

17

Total

(n ¼ 41)

(n ¼ 46)

(n ¼ 48)

(n ¼ 37)

(n ¼ 43)

(n ¼ 34)

(n ¼ 249)

Overweight

WHO (1995) Must et al (1991) Cole et al (2000)

7.3 19.5 19.5

8.7 10.9 15.2

8.3 12.5 16.7

5.4 8.1 8.1

9.3 9.3 9.3

11.8 20.6 23.5

8.4 13.3 15.3

Obese


17.1 4.9 4.9

23.9 21.7 17.4

29.2 25.0 20.8

13.5 10.8 10.8

18.6 18.6 18.6

23.5 14.7 14.7

21.3 16.5 14.9

(n ¼ 50)

(n ¼ 43)

(n ¼ 40)

(n ¼ 41)

(n ¼ 40)

(n ¼ 43)

(n ¼ 257)

Females Overweight


4.0 22.0 28.0

2.3 20.9 27.9

5.0 15.0 22.5

7.3 17.1 22.0

0.0 20.0 25.0

2.3 16.3 20.9

3.5 18.7 24.5

Obese


38.0 20.0 16.0

32.6 14.0 11.6

22.5 12.5 12.5

39.0 29.3 24.4

42.5 22.5 20.0

37.2 23.3 23.3

35.4 20.2 17.9

European Journal of Clinical Nutrition


474 (2000). Recently, Flegal et al (2001) compared the prevalence of overweight in US children calculated with Cole et al and Must et al standards and found that Cole et al’s standards produced higher estimates of overweight and obesity in adolescents than the standard of Must et al. In this study we found a similar trend for the estimate of overweight prevalence but not for that of obesity. Sexual maturation is an important factor influencing BMI and body composition (Daniels et al, 1997; Bini et al, 2000). In the present study however, pubertal staging was not performed due to some cultural constraints. A previous work in which puberty was determined indirectly (by questioning girls about age of menartche and attainment of adult voice in boys) have found that the mean age of puberty in Bahraini boys was 14.6 y and that of menarche in girls was 13.0 y (Musaiger et al, 1993). The high prevalence of obesity documented in this study has important public health implications given the recent evidence linking childhood and adolescent obesity and increased risk of obesity and morbidity in adulthood (Must & Strauss, 1999). The health statistics in Bahrain indicate that obesity-related chronic diseases such as heart disease, diabetes and hypertension are the main public health problems and represented more than 35% of total deaths (Ministry of Health, 2000). Therefore any programme to reduce the occurrence of obesity among adolescents will have a great impact. Although BMI is not an exact measure of body fat, BMI for age-based standards such as that of Must et al or Cole et al appear to be more preferable to use than the WHO criteria (which are based on both BMI and skinfold). Skinfold thickness measurements are subject to considerable inter- and intra-user error, whereas measures based on height and weight are simple to obtain in wide variety of settings and reliable (Himes, 1989). This is particularly relevant to developing countries, where people highly skilled in skinfold measurement are scarce. The high subcutaneous fat in adolescents suggested by the high percentage of obese boys and girls identified using the composite BMI and skinfold criteria (ie WHO recommended criteria) is a cause of concern since it is known that the reliability of skinfold measurement decreases as body fat increases. Furthermore, the use of a BMI reference such as that of Must et al or Cole et al would allow ease of comparison with corresponding studies from other countries in the world.

European Journal of Clinical Nutrition

References Al-Nuaim AR, Bamgboye EA & Al-Herbish A (1996): The pattern of growth in Saudi Arabian male school children. Int. J. Obes. Relat. Metab. Disord. 20,1000 – 1005. Amine EK (1980): Bahrain Nutrition Status Survey. Abu-Dubi: UNICEF, Gulf Area Office. Bini V, Celi F, Berioli MG, Bacosi ML, Stella P, Giglio P, TosTi L & Falorni A (2000): Body mass index in children and adolescents according to age and pubertal stage. Eur. J. Clin. Nutr. 54, 214 – 218. Blair D & Gregory WB (1985): The Nutrition Status of Bahraini Girls aged 7 – 18 years. Bahrain: Bahrain Sport Institute. Cole TJ, Bellizzi MC & Flegal KM (2000): Establishing standard definition for child overweight and obesity worldwide: International survey. Br. Med. J. 320, 1240 – 1243. Daniels SR, Khoury PR & Morrison JA (1997): The utility of body mass index as a measure of body fatness in children and adolescents: differences by race and gender. Pediatric 99, 804 – 807. Eid N, Al-Hotti S, Bourishly N & Khalafawi M (1986): Nutritional anthropometry of school children in Kuwait. Nutr. Report. Int. 33, 253 – 260. Fidanza F (1991): Nutritional Assessment: a Manual for Population Studies, ed. F Fidanza. London: Chapman & Hall. Flegal KM, Ogden CL, Wei R, Kuczmarski RL & Johnson CL (2001): Prevalence of overweight in US children: comparison of US growth charts from the Centers for Disease Control and Prevention with other reference values for body mass index. Am. J. Clin. Nutr. 73, 1086 – 1093. Himes JH (1989): Reliability of anthropometric methods and replicate measurements. Am. J. Anthropol. 79, 77 – 80. Himes JH & Dietz WH (1994): Guidelines for overweight in adolescent preventative services: recommendations from and expert committee. Am. J. Clin. Nutr. 59, 307 – 316. Ministry of Health (2000): Health Statistics. Bahrain, Manama. Musaiger AO, Matter AM, Alekri SA & Mahdi A (1993): Obesity among secondary school students in Bahrain. Nutr. Health 9, 25 – 32. Musaiger AO, Al-Ansari M & Al-Mannai M (2000): Anthropometry of adolescent girls in Bahrain, including body fat distribution. Ann. Hum. Biol. 27, 507 – 515. Must A & Strauss RS (1999): Risks and consequences of childhood and adolescent obesity. Int. J. Obes. Relat. Metab. Disord. 23(Suppl 2), S2 – S11. Must A, Dallal GE & Dietz WH (1991): Reference data for obesity: 85th and 95th percentile body mass index (wt=ht2) and triceps skinfold thickness. Am. J. Clin. Nutr. 53, 839 – 846. Rolland-Cachera MF, Deheger M & Bellisle F (2001): Definition actuelle et evolution de la frequence de l’obesite’ chez l’enfant. Cah. Nutr. Diet. 36, 108 – 112. WHO (1995): Physical status: the use and interpretation of anthropometry. WHO Technical Report Series no. 854, p 271. Geneva: World Health Organization.