Physical activity and change in body mass index from adolescence to mid-adulthood in the 1958 British cohort

Background Prevention of obesity has focused on childhood as a target period. Our aim was to assess whether frequency of adolescent physical activity affected subsequent body mass index (BMI) gain through to mid-adulthood. Methods The British birth cohort of all births in 1 week in March 1958, includes information on physical activity frequency and BMI for several ages, 11-45 years. We examined relationships between activity in adolescence and trajectories of BMI between 16 years (or 23 years) and 45 years using multi-level models. Effects of change in activity on BMI and on change in BMI were tested using ANOVA. Results Physical activity at 11 years had no effect on the BMI trajectories, in males or females. More active females at 16 years gained BMI more slowly than others, by 0.007 kg/m(2)/year per activity category over the period 16-45 years, whereas the most active males gained BMI faster than others, by 0.005 kg/m(2)/year per activity category. This effect in males was not evident on the BMI trajectory from 23 to 45 years. Consistent with these analyses, change in activity was associated with change in BMI in females, e.g. females active at 16 and 42 years gained less BMI than inactive females (2.1 vs 2.5 kg/m(2)/10 years). Results for males were inconsistent over the time periods examined. Conclusions Physical activity may lessen the gains in BMI from adolescence onwards, but relationships vary with age, and in later adolescence show opposite effects for males and females. Decreasing activity between adolescence and mid-adulthood in males, and inactivity in both life stages in females may increase BMI gain.