Objective: 1) To estimate the heritability of body mass index (BMI) in twins aged 16y and 17y, with a special emphasis on gender-specific genetic effects and 2) to compare heights, weights, BMIs, and prevalences of 'overweight' (BMI > or = 25 kg/m2) in these twins and in singletons aged 16.5y.
Design: Cross-sectional and longitudinal epidemiological questionnaire study of twins at ages 16y and 17 y, and cross-sectional study of singletons at age 16.5y.
Measurements: BMI (kg/m2) was calculated from self-reported heights (m) and weights (kg).
Subjects: 4884 twins (2299 boys, 2585 girls) at baseline (age 16 y), 4401 twins (2002 boys, 2399 girls) at age 17 y, and 2509 singletons (1147 boys, 1362 girls) at age 16.5 y. Both twin and singleton samples are nationally representative.
Results: At the ages of 16y and 17y, genetic effects accounted for over 80% of the interindividual variation of BMI. The correlations for male-female pairs were smaller than for either male-male or female-female dizygotic pairs. The singletons, especially the boys, had a higher BMI than the twins. Nine percent of singleton boys, but only 4-6% of twin boys and twin and singleton girls were 'overweight' (BMI > or = 25 kg/m2).
Conclusions: Among adolescents, genetic factors play a significant role in the causes of variation in BMI. The genetic modelling suggested that the sets of genes explaining the variation of BMI may differ in males and females. At this age, the twin boys, but not girls, seem to be leaner than singletons. Further follow-up will indicate whether these small differences disappear, and if not, what implications it might have to the generalizability of twin studies.