Aims/hypothesis The aim of this study was to examine the associations of birthweight, gestational age and childhood BMI (assessed at a mean age of 5 years) with a self-report of a doctor diagnosis of diabetes in middle age. Methods We studied a birth cohort of 5,793 individuals who were born between 1950 and 1956 in Aberdeen, Scotland, and who responded to a questionnaire administered in 2000. Results Birthweight and gestational age were inversely associated with diabetes. These associations remained with additional adjustment for indicators of childhood and adult socioeconomic position, maternal complications of pregnancy, adult smoking, adult BMI and simultaneous adjustment for each other: the adjusted odds ratio per unit increase in birthweight z score was 0.73 (95% CI 0.60-0.88), the odds ratio per week increase in gestational age was 0.91 (95% CI 0.82-1.00) and odds ratio for preterm birth was 2.04 (94% CI 1.18-3.53). The positive association of childhood BMI with diabetes was attenuated on adjustment for adult BMI.
Conclusions/interpretationIn this population, who were born in the 1950s, poor intrauterine growth and preterm birth are associated with an increased risk of diabetes.