Prevalence of obesity and its associations among adolescents: a community-based, birth cohort study from Sri Lanka
 
Niriella MA, De Silva ST, Ediriweera DS, Kasturiratne A, Jayasinghe C, Rajindrajith S, De Silva AP, Pathmeswaran A, de Silva HJ 
Faculty of Medicine, University of Kelaniya, Ragama, Sri Lanka
  
Background 
Childhood and adolescent obesity (AO) is an emerging problem in many developing countries. AO predisposes to insulin resistance, type 2 diabetes, hypertension, hyperlipidemia and liver disease, as well as progression to obesity and cardiovascular disease in adulthood. AO is a complex disorder, with genetics, metabolism, lifestyle, and eating habits likely to play a role in its development. Community based data on AO in Sri Lanka is lacking. 
Aims  
We investigated community prevalence and associations for AO among a selected birth cohort in urban Sri Lanka.  
Method
14-year-olds (Year 2000 birth cohort) from Ragama Medical Officer of Health area, Gampaha District, Sri Lanka, were included. Demographic and lifestyle data was collected. Standard blood biochemistry was performed. Physical activity (PA) and sedentary time (ST) were assessed using the short version of the International Physical Activity Questionnaire (IPAQ). Body mass index (BMI), waist circumference (WC) and blood pressure were measured. Total body fat (TBF%) was measured using impedance method. Homeostatic Model Assessment of Insulin Resistance (HOMA-IR) was calculated. Multiple logistic regression was used to assess relationships between demography, life style habits, anthropometry, TBF% and biochemistry with obesity.  
 
Results  
508 adolescents participated in the study. 261 (51.6%) were girls, median birth weight was 2.9 (IQR 2.6-3.2) kg, 9 (1.78%) had maternal gestational diabetes (GDM), and median age of menarche was 12 (IQR 11-13) years.  61 (12.0%) had general obesity (GO) (BMI >age-sex equivalent of 23kg/m2) [33 (11.4%) were girls and 28 (12.7%) were boys; p=0.76]. 136 (27.0%) had central obesity (CO) (WC >cut-off for age and sex) [97 (37.3%) were girls and 39 (16.0%) were boys; p<0.01]. Median TBF% was 19.4 (IQR 14.2-23.5). 10 (2.0%) had abnormal TBF% [3 (1.1%) were girls and 7 (2.9%) were boys; p=0.21]. Those engaging in some-PA [below recommendation (duration <150 min/week)] and recommended-PA [at or above recommendation (duration ≥150 min/week)] among girls were 129 (49.4%) and 132 (50.6%), while among boys were 116 (48.1%) and 125 (51.9%), respectively. Those with ST >=4 hour,  <4 hours for girls and boys were 63 (24.4%), 195 (75.6%) and 31 (12.7%), 213 (87.3%) respectively. HOMA-IR median was 120.3 (IQR 86.2-173.8). On multiple logistic regression, birth weight (OR-1.01;p<0.01), age at menarche (OR-1.48;p<0.05), TBF% (OR-2.27;p<0.001) and HOMA-IR (OR-1.01;p<0.001) were significantly associated with CO, while PA or ST were not.  

Discussion
CO among adolescent girls was significantly more than boys. Adolescent CO was associated with birth weight, age at menarche (among girls), abnormal TBF% and insulin resistance.