| Summary: | Background: Ethnicity is associated with childhood obesity, with Black African origin girls in particular being more vulnerable to overweight and obesity than their White European peers. In the UK, ethnic minorities often live and attend school in poor urban areas which may influence their opportunity for physical activity and a healthy diet. Aim: To examine neighbourhood and school effects on ethnic differences in Body Mass Index (BMI) and waist circumference trends in adolescence. Methods: Multilevel analysis of longitudinal data on BMI and waist circumference [standard deviation scores (SDS)] from 3401 adolescents in the Determinants of Adolescent Social well-being and Health (DASH) Study (870 White UK, 778 Black Caribbean, 504 Nigerian/Ghanaian, 386 Other African, 418 Indian, and 445 Pakistani/Bangladeshi). Forty-nine London schools participated in the study and the same pupils were surveyed at 11-13yrs and 14-16yrs. Neighbourhood measures included deprivation, crime, and ethnic density; school measures included ethnic density, school socioeconomic status (SES) (academic performance, free school meals, unauthorised absence) and ethos. Individual and family characteristics were also examined (including dietary and physical activity measures, family SES, and parental overweight). Results: Between 11 and 16yrs ethnic differences in BMI emerged in boys and persisted in girls; compared to their White UK peers Black Caribbean and Nigerian/Ghanaian boys and girls, and Other African girls, had a greater mean BMI SDS. These patterns were not observed for waist circumference, signalling ethnic differences in fat distribution or body composition. The DASH pupils overall had large waists compared to the 1990 Growth Reference population. The ethnic minority pupils, with the exception of the Indians, were more likely to live in more deprived, higher crime, less green areas than their White UK peers. However the ethnic minority pupils often attended better performing schools than the White UK pupils. The overall variance in body size at neighbourhood level or school level was small (<4%), and area or school context measures had little or no effect on ethnic differences in body size. Individual characteristics (such as age, pubertal status, and skipping breakfast) and maternal overweight were strong correlates of body size but did not explain the ethnic differences observed. Conclusions: There were significant ethnic differences in BMI in adolescence, emergent in late adolescence for boys. Neighbourhood and school contexts did not explain the ethnic differences in BMI age trends.
|
|---|
