Blood pressure among overweight adolescents from urban school children in Pune, India

Transcription

1 (07) 61, & 07 Nature Publishing Group All rights reserved /07 $.00 ORIGINAL ARTICLE Blood pressure among overweight adolescents from urban school children in Pune, India S Rao, A Kanade and R Kelkar Biometry and Nutrition Group, Agharkar Research Institute, G G Agarkar Road, Pune , India Objectives: To examine the magnitude of overweight and its association with blood pressure (BP) among adolescents. Design: Cross-sectional study with all children in age range 9 16 years (n ¼ 1146 boys and 77 girls) from two schools catering to urban affluent high socio-economic class (HSE), for anthropometric measurements by trained investigators and BP measurement by a pediatrician using sphygmomanometer. Results: The prevalence of overweight based on conventional body mass index (BMI) cutoff was 27.5% for boys and.9% for girls but varied for different indicators. Prevalence of high systolic blood pressure (HSBP) was 12.0% in boys and 9.7% in girls and increased with increasing levels of BMI, weight, triceps skin fold thickness (TSFT) and percent body fat. Mean level of SBP among overweight children was significantly (Po0.001) higher by about 12 mm Hg, whereas that for diastolic blood pressure was higher by 8 mm Hg (Po0.001) as compared to their non-overweight (age, sex-matched) counterparts. This was true in both sexes and for all indicators used for assessing overweight. Prevalence of HSBP increased suddenly beyond BMI value of kg/m 2 in boys and 21.5 kg/m 2 in girls, beyond TSFT value of 12 mm for boys and 14 mm for girls whereas such cutoffs for body fat were above 25% in both sexes. These cutoffs appear much lower than the conventional ones and therefore indicate the need for validation of conventional cutoffs in different populations. Conclusions: Our findings highlight that BP measurement needs to be a routine part of physical examination in school children, and the use of cutoffs anchored to metabolic risks may be essential for assessment of obesity. EuropeanJournalofClinicalNutrition(07) 61, doi:.38/sj.ejcn ; published online 29 November 06 Keywords: overweight adolescents; body mass index; body fat; blood pressure; India Introduction Obesity in adolescents is a major public health problem in developed countries and in some parts of developing countries, too. It is a growing nutritional concern in countries like India, which are witnessing nutritional transition (Popkin et al., 01). It is being realized that with the growing popularity of fast foods, the transition is greatly affecting the food basket of the people. In addition, decreasing physical activity leading to sedentary life style is likely to promote obesity and related health problems. Obesity in adults is known to be a major risk factor for many non-communicable diseases and, especially, for cardiovascular disorders. Childhood obesity, too, has important Correspondence: Dr S Rao, Animal Science Division, In-charge, Biometry & Nutrition Unit, Agharkar Research Institute, G G Agarkar Road, Pune , India. s: raoari@yahoo.com, srao@aripune.org Received 19 December 05; revised 19 September 06; accepted 27 September 06; published online 29 November 06 health consequences for children and is a major antecedent of adult obesity (Popkin and Udry, 1998). Persistence of childhood obesity into adulthood has been shown by several studies (Rolland Cachera et al., 1987; Siervogel et al., 1991; Serdula et al., 1993). Guo et al. (1994) have shown that overweight during childhood, especially beyond 8 years, is an important risk factor for overweight at the age of 35 years. Obesity in adolescents is not only associated with hypertension and abnormal lipid profile, but also shows clustering of risk factors for cardiovascular disease (Chu et al., 1998). Reported studies on adolescent obesity in the urban Indian population are unfortunately scarce, and have not studied its health consequences. Thus, Kapil et al. (02) report the prevalence of obesity in school children from the affluent class of Delhi to be only 6 8%, whereas Ramachandran et al. (02) report it to be as high as 16 18% in school children from Chennai and, further, they associate it with lower physical activity. As obesity is an important risk factor for hypertension, we studied adolescent obesity and blood pressure (BP) levels in school children from the urban population in Pune.

2 634 Methods Subjects Two convent public schools, one catering to boys and another to girls, from the urban affluent high socioeconomic class (HSE) were considered for the study. These two schools actually represented the elite class of the society, as the school fees were enormously high compared to other schools in the city, and as such were affordable to parents from the high-income group. A majority of the parents were either businessman or professionals. A cross-sectional study was conducted by complete enumeration method. Thus, all the children from the 5th to the th standard, covering the age range 9 16 years (1146 boys and 77 girls) from these schools, were included in the study. Absentees (5.7% boys and 8.6% girls) on the days of the actual survey were the only exclusions. The study was approved by Institutional Research Advisory Committee. Measurements Anthropometry Measurements were recorded in duplicate by trained investigators using standard procedures. Nevertheless, an interobserver variability (IOV) study was carried out before starting the study. Between-investigator variation was negligible for the measurements, weight and body fat, taken with the help of digital equipments. However it was seen in case of skin fold measurements (CV% for triceps-5.5). Investigator with the lowest standard deviation in case of skin folds was kept constant for a particular measurement during the study. Weight was recorded (up to g) using the electronic weighing balance (Suysan, India), standing and sitting heights were measured by stadiometer (upto 0.1 cm), skin fold thickness at triceps was measured (upto 0.1 mm) using the Harpenden s Calipers (CMS Instruments, London, UK), whereas body fat was measured using the Omran (HBF 0, Japan) equipment that works on the principle of bioelectrical impedance analysis. Age assessment was carried out using birth date records from the school. Blood pressure A pediatrician who accompanied the team measured the BP using sphygmomanometer (mercury). It was measured in a sitting position, on the left hand, using an appropriate size cuff, after a child had taken rest for at least min. As the measurements of the students were taken during school timings, we were limited to a single measurement of BP. We defined high systolic blood pressure (HSBP) and high diastolic blood pressure (HDBP) independently when measured SBP or DBP of a child was above the 95th percentile of value of BP corresponding to his/her age, sex and height percentile given by the Task Force Recommendations (National High Blood Pressure Education Program, 1996). An index for BP was defined as a ratio of observed BP divided by this reference value of 95th percentile of BP, and was expressed as percent. Thus, index for SBP (SBPI) as well as for DBP (DBPI) was computed for the individual child. This corrects the effect of height on BP and allows for severity of BP elevation for comparison between groups of children of different heights. Assessment of overweight Body mass index For assessing the prevalence of overweight children, the conventional cutoff (Must et al., 1991) of body mass index (BMI) above the 85th percentile was used. Similarly, age sex specific cutoffs for BMI developed by Cole et al. (00) were also used to define overweight children. These cutoffs are linked with adult cutoff and extrapolated to childhood as proposed at the International Obesity Task Force (IOTF) meeting in 1997 (Dietz and Bellizzi, 1999). Body fat (%) Another indicator based on body fat percent was also considered for estimating the prevalence of overweight. Conventional cutoff (X25% for boys andx% for girls), defined by Williams et al. (1992), was used for estimating prevalence of overweight in adolescents. Skin fold at triceps A relatively simple method of assessing the percentage of body fat in children would be the measurement of the subcutaneous fat layer, namely skin fold thickness. The cutoff of 85th percentile given by Must et al. (1991) was used for skin fold at triceps for defining overweight. Weight for age This indicator is used to grade obesity on the basis of a simple basic measurement of the body weight expressed as percent of expected weight for age (WHO, 1983). The cutoff used was 1% or above for defining overweight. Statistical methods Age adjustment was required in view of the significant (Po0.000) correlation of age with body weight (r ¼ 0.22 and 0.48) and BMI (r ¼ 0.22 and 0.33 for boys and girls, respectively). In case of girls, such adjustment was required additionally for body fat (%) and triceps skin fold thickness (TSFT) being significantly (Po0.000) correlated with age (r ¼ 0.36 and 0.24, respectively). Mean values of various anthropometric measurements for the two sexes for a given age group and the mean values of BP measurements among overweight and non-overweight group of children were

3 tested using Student s t-test. The proportion of children with HSBP in overweight and non-overweight groups was tested using Z-test. Linear trends in various anthropometric measurements by age were tested using analysis of variance (ANOVA) whereas those in proportions of overweight children by weight, BMI, body fat and TSFT were tested using Z-test for a linear trend in proportion (Snedecor and Cochran, 1968). All the analyses were carried out using SPSS/ PC þ 11.0 version for windows. Results The mean values of various anthropometric measurements are plotted in Figure 1.The weight, height, sitting height and BMI generally increased with age in both sexes. On the other hand, body fat percent and TSFT were not significantly different among various age groups in boys, but showed a significantly (Po0.000) increasing trend in the case of girls. Mean height and sitting height of boys was significantly (Po0.01) higher than that of girls above 13 years of age. In contrast, body fat (%) and BMI were significantly (Po0.01) higher in girls than in boys above 13 years. TSFT in case of girls was higher than in boys all through adolescence, but the differences widened significantly (Po0.01) above 12 years. Thus, above the age of 13 years, girls were fatter than boys. It is known that during the span of adolescent growth, significant changes take place in weight, height and body composition at varying chronological ages. Therefore, the prevalence of overweight was computed for children according to three broad categories of age. These categories differed for the two sexes owing to the obvious differences in age of maturity. Among girls, the mean recalled age for onset of Weight (kg) Height (cm) Body fat (%) Sitthing height (cm) BMI (kg/m 2 ) Triceps (mm) AGE (yr) AGE (yr) Figure 1 Mean values of various measurements by age for adolescent boys and girls.

4 636 Table 1 Prevalence of overweight among adolescent boys and girls according to age by various indicators Age (years) N BMI485th percentile BMI (IOTF) TSFT 485th percentile Body fat a (%) Weight for agex1 (%) HSBP (%) HDBP (%) o X Total o X Total Abreviations: BMI, body mass index; HDBP, high diastolic blood pressure; HSBP, high systolic blood pressure; IOTF, International Obesity Task Force. a X25% for boys and X% for girls. menarche was years in our subjects. Hence, the categories for girls were defined as pre-pubertal (o11 years), pubertal (11 13 years), and post-pubertal (X13 years) whereas the respective categories were taken 1 year later in case of boys in view of the fact that they mature later compared to girls. The prevalence of overweight by five different indicators defined earlier is shown in Table 1. The estimate of overall prevalence of overweight obtained by TSFT was lowest (11.7% in boys and 7.4% in girls), whereas that based on percent body fat was highest (35.8% for boys and 24.1% for girls) compared to other indicators. Among boys, the prevalence of overweight for any indicator was higher in pre-pubertal age, but decreased as growth advanced. In contrast, among girls it was lower in prepubertal stage, but increased in pubertal and post-pubertal period when indicators based on body fat are considered, that is, TSFT and percent body fat, probably owing to fat deposition associated with onset of menarche. However, the prevalence of overweight decreased with age when weight for age was considered but did not vary when BMI was considered. The overall prevalence of HSBP was higher than that for HDBP in both sexes (.5 and 7.0% in boys; 9.7 and 2.6% in girls, respectively). As obesity is a major risk factor for hypertension, for each overweight child age sex-matched normal, non-overweight child was randomly selected from the same school. The prevalence of HSBP and HDBP among non-overweight and overweight children, identified by various indicators, was examined (Table 2). The mean SBP and DBP levels as well as prevalence of HSBP and HDBP were significantly (Po0.001) higher among overweight children compared to their counterpart, by any indicator. Thus high BMI, high TSFT, high body fat percent and high weight for age were associated with higher prevalence of HSBP and HDBP. In particular, when children were identified as overweight based on TSFT, the prevalence of HSBP as well as HDBP was highest among boys (27.6 and 17.9%, respectively) as well as girls (.5 and 8.9%, respectively). On the other hand, it was lowest when they were identified as overweight based on percent body fat. Sex-related differences were observed in the prevalence of high BP. Thus, among overweight girls, the prevalence of HSBP was higher, whereas in overweight boys that of HDBP was higher for any indicator. The mean values of SBPI and DBPI (i.e. after correcting the effects of age, gender and height on BP) are compared between overweight and non-overweight children by various indicators in Figure 2. It can be observed that mean SBPI and mean DBPI were significantly (Po0.000) higher for overweight children than for their counterparts, in both sexes, by any indicator. Further, among overweight children, the sex differences seen in mean SBP values disappeared when their mean SBPI were compared, whereas they remained significant in case of DBPI. As overall prevalence of HDBP was relatively smaller than that for HSBP, we examined further the prevalence of HSBP for different levels of various indicators adjusted for age and is shown in Figure 3. Prevalence of HSBP showed positive association with various indicators of overweight. It increased significantly (Po0.000) with BMI in both sexes, but increase was sharp beyond the value of BMI.0 kg/m 2 in boys and beyond 21.5 kg/m 2 in girls. Similar significant trend (Po0.000) was seen with age-adjusted body weight, and the increase in prevalence was higher beyond 48 kg. It also increased as fatness increased and the significant (Po0.000) increase in the prevalence was observed beyond 25% in both sexes. SBP also showed positive association with TSFT, a measure of subcutaneous fat and the significant (Po0.000) increase in prevalence of HSBP was seen beyond the value of 12 mm in boys and 14 mm in girls. Discussion Obesity s persistence into adulthood, its resistance to treatment and its health consequences make it critical to understand adiposity in children. A review of childhood

5 Table 2 Prevalence (%) and mean (7s.d.) BP values for overweight and non-overweight adolescent boys and girls, according to different indicators Indicator Overweight a,b Nonoverweight Overweight b Nonoverweight c BMI n % HSBP % HDBP SBP (mm Hg) DBP (mm Hg) BMI(IOTF) n % HSBP % HDBP SBP (mm Hg) DBP (mm Hg) TSFT n % HSBP % HDBP SBP (mm Hg) DBP (mm Hg) Body fat n % HSBP % HDBP SBP (mm Hg) DBP (mm Hg) Weight for Age n % HSBP % HDBP SBP (mm Hg) DBP (mm Hg) Abreviations: BMI, body mass index; DBP, diastolic blood pressure; HDBP, high diastolic blood pressure; HSBP, high systolic blood pressure; IOTF, International Obesity Task Force; SBP, systolic blood pressure; TSFT, triceps skin fold thickness. a BMI above 85th percentile for given age, sex as well as above IOTF cutoff for given age, sex body fat X25% for boys and X% for girls, TSFT above 85th percentile for given age, sex; weight for age X1%. b Differences in all the measures significant (Po0.01) compared to nonoverweight. c Randomly selected non-overweight subjects matched for age and sex. obesity indicates that most studies reported are from Europe and North America (Dietz and Bellizzi, 1999) and validation studies in other populations are needed. Moreover, in the Indian context, as an epidemic of non-communicable diseases like hypertension, diabetes and cardiovascular diseases is on the rise, investigations on adolescent obesity assume special significance. In view of the rising affluence in India leading to a sedentary life style and mushrooming of food stalls for fast foods especially in urban areas, it was thought that emerging trends in adolescent obesity in urban populations needed to be studied. We examined the magnitude of obesity and its association with high BP among urban adolescents from the affluent population in Pune, India. Comparison of mean weights and heights to NCHS median showed that boys had comparable values up to 14 years and were heavier and shorter beyond 14 years. In girls, these values were comparable up to 13 years, and they had lower weights as well as heights beyond 13 years. However, mean values for TSFT were on the higher side compared to NCHS standard for boys across all ages, but it was not true in case of girls. A similar comparison with studies reported from India (Vijayraghavan et al., 1971; Agarwal et al., 1992) for children from well to do families showed that our subjects were taller and heavier. This could be attributed to the secular changes in growth. However, the fact that mean TSFT values observed in this study are comparable to those reported two decades back (Vijayraghavan et al., 1974; Kapoor et al., 1991) for children from well to do families, shows that there is hardly any secular change with respect to triceps skin folds in Indian adolescents. BMI, defined as kg/m 2, offers a reasonable measure of fatness in children and adolescents. Although its validity across diverse samples of youth from different age, sex and ethnic groups has not been evaluated, it is used widely because of the relative ease and accuracy of the basic measurements (Malina and Katzmarzyk, 1999). Several studies have focused on validation of various anthropometric measures for assessing fatness and seem to recommend mostly BMI (Dietz and Bellizzi, 1999) and TSFT (Deurenberg et al., 19; Sarria et al., 1998). Our observations show that prevalence of overweight children differs greatly with different indicators and it was lowest using TSFT cutoff but highest using percent body fat cutoff, in both sexes. Although BMI is widely used as a surrogate measure of adiposity, it is a measure of excess weight relative to height rather than excess body fat. In view of the fact that Asians are believed to have higher body fat for the same BMI compared to Westerners, assessment of body fat is essential. Although methods of body fat assessment like densitometry or DEXA are more accurate but are complex, costly and impractical for field studies, alternative simple methods like BIA or surrogate measures like skin folds have become popular. Our observation that mature boys had less body fat than girls is similar to that reported by Dietz and Bellizzi (1999) based on estimates of body fat by DEXA. Higher body fat in girls was probably responsible for higher BMI, especially beyond 14 þ years. A similar observation has been reported by Freedman et al. (05) who have estimated body fat using DEXA. The fact that these trends in body fat are observed in adolescents from White population and also confirmed in Indian 637

6 638 BMI BMI(IOTF) Tricep skin fold thickness Body Fat Weight for Age Overweight Non-overweight Figure 2 Mean values of SBP and DBP indices in overweight and non-overweight children by various indicators.

7 BMI Weight 639 % prevalence % prevalence 0 < > 23 (kg/m 2 ) 0 < (kg) Body Fat TSFT % prevalence % prevalence 0 < >=35+ (%) fat 0 < (mm) Figure 3 Prevalence of HSBP according to various indicators of overweight in adolescent children. (All parameters are age adjusted in case of girls and only weight, BMI in case of boys). population, gives us confidence in the estimates of body fat based on BIA. We observed that the prevalence of overweight children was higher in younger boys and decreased as age advanced, but the trend was reversed in girls. Mean skin folds at triceps were also higher in girls than boys throughout adolescence, but the differences were striking beyond 12 years. Must et al. (1991), too, have reported a similar observation using NHANES-I data. This is consistent with the greater gains in muscle and bone experienced by boys during adolescence and greater gains in body fat experienced by girls (Sardinha et al., 1999). Therefore, it cannot be denied that these variations in age, sex and maturation impose more limitations on BMI as a measure of adiposity in the pediatric population than in adult population (Neovius et al., 04). In view of this, examining the relation of obesity with morbidity appears to be best available alternative. In fact, one of the criteria suggested for choosing a measure of adiposity is the clinical validity, that is, ability to predict morbidity (Kraemer et al., 19). A single observation available on BP was the limitation of the study and may perhaps be the reason for higher prevalence of HSBP and HDBP observed in this population. However, it cannot be overlooked that the difference in the mean level of SBP among overweight and non-overweight children was significantly (Po0.001) high and was about 12 mm Hg, whereas that for DBP was about 8 mm Hg (Po0.001). This was true in both sexes and for all indicators used for assessing overweight. An association of overweight and fatness with high BP has also been observed by Iman et al. (03) among children from Kuwait and in CATCH study by Dwyer et al. (1998) among an ethnically and geographically diversified group of students. In younger children (1 7 years) from 11 major cities in China, He et al. (00) observed that the differences in mean BP levels were around 5 mm Hg in SBP and 4 mm Hg in DBP. The larger differences in mean BP values observed in this study probably indicate that overweight during adolescence further amplifies the differences in BP levels. Prevalence of HSBP was higher in overweight girls than overweight boys, whereas the reverse was true in case of prevalence of HDBP. However, even in non-overweight children, the prevalence of HSBP or HDBP was observed to be of the order of 5 %, and indicates that the cutoffs for assessing overweight need to be lower than conventional cutoffs in case of the Indian population. Increase in the prevalence of HSBP was sharp and linear with increase in body fat percent, BMI or TSFT. It increased

8 6 suddenly beyond the value of 25% for body fat in both sexes. In case of girls, this is much lower than the conventional cutoff of % body fat used for assessing overweight. In case of BMI, prevalence increased beyond.0 kg/m 2 in boys (and beyond 21.5 kg/m 2 in girls) and is again lower than the respective IOTF cutoff. For example, in the age range 16 years IOTF cutoffs for boys vary between and 23.3 kg/m 2 and for the observed sample in our study, the estimated weighed average would be kg/m 2 (a similar estimate for girls is 22.6 kg/m 2 ). Similar is the case with TSFT (12 vs 16.4 mm for boys; and 14 vs 22.3 mm for girls), indicating that most cutoffs appear much lower than the conventional ones. Recently, Sardinha et al. (1999) too have reported that the recommended cutoffs for Portugese boys and girls also correspond to lower percentiles than those recommended for US boys and girls, whereas Malina and Katzmarzyk (1999) show cutoffs for body fat to be more sensitive than BMI. Our observations, therefore, indicate the need for validation of conventional cutoffs in different populations. In conclusion, the prevalence of adolescent overweight/ obesity is indeed increasing in urban well-off populations in India. Our observations also highlight the fact that obesity is a major factor associated with increased BP levels and physical examination of school children could be a routine part of school curriculum. However, owing to probable population differences in relative risks, customized classification systems derived from national data are required. The findings in this study underscore the importance of examining childhood obesity in relation to the functional outcome. All the more so because South-east-Asian populations are believed to have relatively higher percent of body fat and that these populations are passing through nutritional transition. Thus, the results of our study highlight the fact that using cutoffs anchored to metabolic risks may be essential in assessing obesity. Acknowledgements We are grateful to Dr VS Rao, Director, ARI, for encouraging and providing the facilities to carry out this work in the institute. We greatly appreciate the help of other staff members in the Department during data collection. References Agarwal DK, Agarwal KN, Upadhyay SK, Mittal R, Prakash R, Rai S (1992). Physical and sexual growth patter of affluent children from 5 to 18 years of age. Indian Pediatr 29, Chu NF, Rimm EB, Wang DJ, Liou HS, Shieh SM (1998). Clustering of cardiovascular disease risk factors among obese school children: The Taipei Children Heart Study. Am J Clin Nutr 67, Cole TJ, Bellizzi MC, Flegal KM, Dietz WH (00). Establishing a standard definition for child overweight and obesity worldwide: International survey. Br Med J 3, Deurenberg P, Pieters JJL, Hautvast JGAJ (19). The assessment of body fat percentage by skinfold thickness measurements in childhood & young adolescents. Br J Nutr 63, Dietz WH, Bellizzi MC (1999). Introduction. The use of body mass index to assess obesity in children. Am J Clin Nutr (Suppl), 123s 125s. Dwyer TJ, Stone EJ, Yang M, Feldman H, Webber LS, Must A et al. (1998). Predictors of overweight & over fatness in a multiethnic pediatric population. Am J Clin Nutr 67, Freedman DS, Wang J, Maynard LM, Thornton JC, Mei Z, Pierson Jr RN et al. (05). Relation of BMI to fat and fat-free mass among children and adolescents. Int J Obese Relat Metab Disord 29, 1 8. Guo SS, Roche AF, Chumlea WC, Gardner JD, Siervogel RM (1994). The predictive values of childhood body mass index values for overweight at age 35 y. Am J Clin Nutr 59, He Q, Ding ZY, Fong DYT, Karlberg J (00). Blood pressure is associated with body mass index in both normal & obese children. Hypertension 36, Iman S, Hind AN, Ali A, Farida M, Bayoomi A, Mohamed AB (03). Juvenile hypertension in Kuwait: prevalence & influence of obesity. Int Pediatr 18, Kapil U, Singh P, Pathak P, Dwivedi SV, Bhasin S (02). Prevalence of obesity amongst affluent adolescent school children in Delhi. Ind Pediatr 39, Kapoor G, Aneja S, Kumari S, Metha SC (1991). Triceps skinfold thickness in adolescents. Indian J Med Res 94, Kraemer HC, Berkowitz RI, Hammer LD (19). Methodological difficulties in studies of obesity I. Measurement issues. Ann Behav Med 12, Malina RM, Katzmarzyk PT (1999). Validity of Body mass index as indicator of the risk & presence of overweight in adolescence. Am J Clin Nutr, 131s 136s. Must A, Dallal GE, Dietz WH (1991). Reference data for obesity. 85th & 95th percentiles of body mass index (wt/ht 2 ) & triceps skinfold thickness. Am J Clin Nutr 53, National High Blood Pressure Education Program (1996). Update on the Task Force Report (1987) of high blood pressure in children and adolescents (September 1996): Washington, DC: US Government Printing office. National Institute of Health Publication No , pp Neovius MG, Linne YM, Barkeling BS, Rossner SO (04). Sensitivity & specificity of classification system for fatness in adolescents. Am J Clin Nutr, Popkin BM, Udry JR (1998). Adolescent obesity increases significantly in second and third generation US migrants: The national longitudinal study of adolescent health. J Nutr 128, 1 6. Popkin BM, Hortan S, Kim S, Mahal A, Jin S (01). Trends in diet nutritional status & diet related non-communicable diseases in China & India: The economic costs of the nutrition transition. Nutr Rev 59, Ramachandran A, Snehalatha C, Vinitha R, Thayyil M, Satishkumar CK, Sheeba L, Joseph S, Vijay V (02). Prevalence of overweight in urban Indian adolescent school children. Diab Res Clin Pract 57, Rolland Cachera MF, Deheeger M, Guillood-Bataille M, Avons P, Patois E, Sempe M (1987). Tracking the development of adiposity from one month of age to adulthood. Ann Human Biol 14, Sardinha LB, Going SB, Teixeira PJ, Lohman TG (1999). Receiver operating characteristic analysis of body mass index, triceps skinfold thickness & arm girth for obesity screening in children and adolescents. Am J Clin Nutr, 95. Sarria A, Garcia-Llop LA, Moreno LA, Fleta J, Morellon MP, Bueno M (1998). Skinfold thickness measurements are better predictor of body fat percentage than body mass index in male Spanish children and adolescence. Eur J Clin Nutr 52, Serdula MK, Ivery D, Coates RJ, Freedman DS, Williamson DF, Byers T (1993). Do obese children become obese adults? A review of the literature. Prev Med 22,

9 Siervogel RM, Roche AF, Guo S, Mukherjee D, Chumlea WC (1991). Patterns of change in weight/stature 2 from 2 to 18 yrs: finding from long term serial data for children in the Fels longitudinal growth study. Int J Obes 15, Snedecor GW, Cochran WG (1968). Statistical Methods. Indian Edition, Oxford and IBH publishing Co. pp Vijayraghavan K, Singh D, Swaminathan MC (1971). Heights and weights of well nourished Indian school children. Ind J Med Res 59, Vijayraghavan K, Singh D, Swaminathan MC (1974). Arm circumference and fat fold at triceps in well nourished Indian school children. Ind J Med Res 62, Williams DP, Going SB, Lohman TG, Harsha DW, Srinivasan SR, Webber LS et al. (1992). Body fatness and risk for elevated blood pressure, total cholesterol and serum lipoprotein ratios in children and adolescents. Am J Public Health 82, World Health Organisation (1983). Measuring changes in nutritional status. Guidelines for assessing the nutritional impact of feeding programmes for vulnerable groups. WHO: Geneva,