Smoothened percentiles for height, weight and body mass index of urban school going adolescents aged 11-17 years in national capital territory region of Delhi, India

PURPOSE: The aim of the study was to estimate centiles by using improved statistical smoothing procedure, the Box–Cox power-exponential (BCPE) method, in urban northern Indian adolescents aged 11–17 years. MATERIALS AND METHODS: Data were collected cross-sectionally by measuring specific anthropometric features such as height, weight, and mid-upper arm circumference in school-based adolescents aged 11–17 years including both boys (n = 838) and girls (n = 788) enrolled in government educational institutions in urban Delhi. We used a state-of-the-art statistical methodology (BCPE method) to establish centile curves. RESULTS: The model fitted before smoothing revealed that weight, height, and BMI did not follow a normal distribution; both skewness and kurtosis were observed in all three variables. After correcting both skewness and kurtosis, estimated empirical percentile values showed a gradual increase in weight, height, and BMI in both boys and girls. Girls had higher weight and height than boys in initial ages and observed a steep increase in boys in both weight and height in later ages. BMI was higher in girls than boys and visibly higher during 14–16 years of age. The 50(th) percentile value of BMI was smaller in all the ages in our study than that in other studies. CONCLUSIONS: Smoothened percentile values derived for BMI by using the state-of-the-art statistical methodology may help policymakers to promote better growth in urban adolescents.