Background: A narrow time window in infancy may be relevant for the aetiology of immune-mediated type 1 diabetes. We investigated whether a non-specific immune stimulation in the first year of life, as resulting from Bacillus Calmette-Guérin (BCG) vaccination, was associated with childhood diabetes.
Methods: Using data from a birth cohort assembled through linkage of administrative databases, 78,492 subjects born in 1974 were the object of the present analysis. Information was extracted from the birth, death, and BCG vaccination registries. Diabetes-related health services were obtained from administrative health databases (physician billing claims and hospitalisation data) until 1994. Subjects were classified as having diabetes according to two validated definitions: (1) ≥2 diabetes-related medical visits within 2 years or ≥1 hospitalisation for diabetes; and 2) ≥4 diabetes-related medical visits within 2 years. Cox proportional hazards regression was used to estimate adjusted hazard ratios (HR) and 95% confidence interval (CI), adjusted for potential confounders.
Results: Forty-four per cent of subjects were BCG vaccinated in the first year of life. According to the first and second definition, respectively, 293 (0.37%) and 230 (0.29%) subjects were classified as having diabetes. There was no association between BCG vaccination in the first year of life and risk of diabetes with either definition (HR(def1) = 0.92, 95% CI 0.73, 1.17; HR(def2) = 1.04, 95% CI 0.80, 1.37), and results did not differ by sex.
Conclusions: Given the potentially critical importance of the exposure window and paucity of studies addressing BCG vaccination timing in relation to diabetes risk, this question deserves further investigation.
Keywords: BCG; Bacillus Calmette-Guérin; administrative health databases; birth cohort; diabetes.
© 2015 John Wiley & Sons Ltd.