Obesity among school-aged youth is a growing public health problem in India [1]. This crisis appears to be rising rapidly, especially in cities and among affluent youth [2-4]. Compared to countries in the West, many negative sequelae of obesity, like diabetes, occur at earlier ages [4,5] and at lower BMI cut-points [6] in India, which makes obesity a particularly serious public health issue in this country.

It is critical to be able to accurately identify obesity among school-going youth, for clinical and epidemiologic purposes. Many growth references can be used [7-10]. Each uses age- and gender-specific body mass index (BMI) cut-points to classify young people according to their weight status (eg, overweight or obese). To date, all studies of obesity among school-going youth in India have employed either (a) a National reference [7]; (b) the older National Center for Health Statistics/ World Health Organization (NCHS/WHO) reference [8]; and/or (c) the International Obesity Task Force (IOTF) reference [9]. None have used the growth reference recently released by the WHO for school-aged children [5 to 19 years old) [10]. This growth reference was developed after the release of a new reference for preschool children (0 to 5 years old) in 2006 [11]. Data from a multi-national sample of preschool children, including children from India, were merged with the 1977 NCHS data (1-21 years old) to construct new growth curves for school-aged youth. This "new" WHO reference replaces the "old" WHO/NCHS reference.

The purpose of this paper is to compare this WHO reference [10] with the national [7] and IOTF [8] references to assess the extent to which they agree in regards to classifying weight status, with a special focus on identifying overweight or obese youth.

This cross-sectional study was conducted in 2006, when eight schools in New Delhi, India were recruited to participate in the study. Four were Private schools (middle-upper SES) and four were Government schools (low SES). Ethical clearances for the study were obtained from the appropriate boards in India and the US. Passive parental consent and active student assent were required to participate. The study included (a) anthropometric measures; (b) a survey of nutrition and physical activity behaviors; and (c) a survey specific to ‘westerni-zation’ or cultural adaptation within this setting.

Anthropometric data (i.e., height, weight) were collected from these students using standardized protocols adapted from Lohman and colleagues [13] to meet needs specific to this context. Weight was measured using a digital scale that was transported to each school. The scale was calibrated daily, and each student’s weight was recorded to the nearest 0.1 kg. Height was measured using a calibrated vertical bar and was recorded to the nearest 0.5 cm. Data were collected by two-person teams of trained research staff.

Body mass index (BMI) was used to classify the participants according to their weight status (i.e., underweight, normal weight, overweight, or obese) using age- and gender-specific cut-points recommended by three growth references: (a) the International Obesity Task Force (IOTF) [9]; (b) the WHO recommendations [10]; and (c) a National (i.e., Indian) study [7]. The classification of participants across these growth references was compared. Kappa statistics were calculated to measure agreement between each pair of growth references [14,15]. Kappa at 1.00 implies perfect agreement. Mixed-effects regression models were used to test for differences in the percentage of participants who could be classified as overweight or obese by grade level, gender, and school type. School was specified as a nested random effect in the regression models [16].

Using the IOTF, the WHO, and the Indian growth references, the prevalence of obesity was estimated at 3.25%, 4.95%, and 5.01%, respectively, while the prevalence of overweight was estimated to be 10.40%, 11.55%, and 12.87%, respectively. These estimates were combined to examine whether the prevalence of obesity/overweight varied by grade, gender, or school type (Table II). There were no differences by grade level. The prevalence of overweight/obesity was slightly higher among boys compared to girls, but this difference was only statistically significant when using the Indian reference (13.86% vs 9.80%, P=0.007). The prevalence of overweight/obesity, by comparison, was much higher among students enrolled in private schools compared to those in Government schools. The difference was statistically significant in each of the three growth references (WHO: 26.62% vs 3.98%, P<0.001; IOTF: 21.99% vs 3.13%, P<0.001; Indian: 28.61% vs 4.48%, P<0.001). The percentage of students who were either overweight or obese was 6 to 7 times higher among students in private schools compared to students in Government schools.

Table II



Differences in the Prevalence of Obesity/Overweight by Grade Level, Gender, and School Type (N=1818)

In studies of child and adolescent obesity, the International Obesity Task Force reference is typically recommended for use, especially for comparisons, worldwide [9]. In this study of school-going youth from India, the IOTF reference consistently classified participants in a lower weight status group, across all categories, compared to the new WHO references and Indian references. That is, compared to these references, the IOTF reference underestimated obesity. The WHO and Indian references, in contrast, were very similar.

Regardless of the growth reference used, it is clear that obesity is problematic among adolescents in large cities of India, like Delhi. In this study, the distribution of BMI was skewed, with a longer tail at higher BMI values, which is characteristic of this epidemic in other regions of the world [18]. Consistent with similar studies from other cities in India, the prevalence of obesity was somewhat higher among boys, as compared to girls [2-4]. In contrast to what is typically observed in the West, obesity is most problematic among more affluent youth in India, versus the less affluent. About 1 in every 4 (>25%) students enrolled in Private schools in this study, which are more affluent than Government schools, was overweight or obese. This estimate is greater than that reported in comparable studies of affluent, school-aged youth in other parts of India, like Hyderabad (<7%) [2] and Ludhiana, (<15%) [3], but consistent with study from Delhi (>25%) [19]. Epidemiologic studies of adults seem to suggest obesity is most problematic in north India, in cities like Delhi [20]. This may also be true for adolescents.

Schools were not randomly chosen to participate in the study, but were selected to be representative of the mix of types of schools in Delhi. The study, therefore, provides a sample of school-aged youth of different affluence, as well as both boys and girls from different grade levels [13].

Future research should be cognizant of the different growth references that can be used to define obesity among school-aged youth in India, including their strengths and weaknesses. An accurate understanding of which school-aged youth are affected by this emerging epidemic is critical to subsequent studies designed to understand the reasons for the same. These studies, in turn, must then be used to drive the development of effective intervention, which is urgently required in India.

Contributors: MHS is the Principal Investigator. She was involved in study design, implementation, development of data collection protocols and the conceptualization of the data analysis. MA was involved as Project Director of the study in India, and worked on all aspects of the design and implementation of the study, including development of data collection protocols. PD and VT assisted with all data analyses. RS oversaw data collection. KSR and CLP are Co-Investigators on the study and assisted with its design and the development of data collection protocols. MHS wrote the manuscript, with significant input from all other co-authors listed. The final manuscript was approved by all the authors.

Funding: University of Minnesota’s Obesity Prevention Center (Stigler, PI) and the University of Texas School of Public Health (Stigler, PI).

Competing interests: None stated.