Aims To perform a longitudinal analysis of the association between childhood body mass index (BMI) and later risk of Type 1 diabetes, controlling for socio-economic status, birthweight, height in early and late childhood, breastfeeding history and pubertal status.
Methods Analysis of the 1970 British Birth Cohort, followed up at age 5, 10 and 30 years (n = 11 261). Data were available on birthweight, breastfeeding; height, weight, pubertal status, socio-economic status at age 10 years; self-report data on history of diabetes (type, age at onset) at age 30 years. Cox proportional hazards models were used to examine relations of childhood growth, socio-economic status and breastfeeding history to the incidence of Type 1 diabetes between 10 and 30 years of age.
Results Sixty-one subjects (0.5%) reported Type 1 diabetes at 30 years of age; 47 (77%) reported onset ≥ age 10 years. Higher BMI z-score at 10 years predicted higher risk of subsequent Type 1 diabetes (hazard ratio 1.8, 95% confidence interval 1.2 to 2.8, P = 0.01) when adjusted for birthweight, pubertal status, breastfeeding history and socio-economic status. Repeating the model for childhood obesity, the hazard ratio was 3.1 (1.0, 9.3; P = 0.05). Birthweight, breastfeeding, height growth and pubertal timing were not associated with incidence of Type 1 diabetes.
Conclusions Higher BMI in childhood independently increased the risk of later Type 1 diabetes, supporting suggestions that obesity may provide a link between Type 1 and Type 2 diabetes. This supports observations of a rise in Type 1 diabetes prevalence. Reduction in childhood obesity may reduce the incidence of Type 1 as well as Type 2 diabetes.