Aim — To examine the association between childhood food insecurity and incident diabetes.
Methods — Using health administrative databases linked to the Canadian Community Health Survey, we conducted a population‐based cohort study of children aged <18 years from Ontario, Canada. Children without diabetes who had a household response to the Canadian Community Health Survey Household Food Security Survey Module were followed for a median of 9.5 years for incident diabetes. Multivariable Cox proportional hazards models were used to examine the association between childhood food insecurity and incident diabetes, adjusted sequentially for important clinical and socio‐economic risk factors.
Results — We included 34 042 children, of whom 5.3% lived in food‐insecure households. There were 184 new cases of diabetes, diagnosed at a median age of 16 and 18.5 years in food‐secure and food‐insecure children, respectively. In unadjusted analysis, childhood food insecurity was associated with an increased risk of incident diabetes (hazard ratio 1.69, 95% CI 1.01 to 2.81). When adjusted for clinical and socio‐economic confounders, the relationship was no longer statistically significant (hazard ratio 1.55, 95% CI 0.91 to 2.66, adjusted for clinical confounders; hazard ratio 1.30, 95% CI 0.72 to 2.37, adjusted for clinical/socio‐economic confounders). Our results remained robust in sensitivity analyses.
Conclusions — Although food‐insecure children are a medically and socially vulnerable population, they do not appear to be at increased risk of incident diabetes over a median of 9.5 years.