Background
There is conflicting evidence that caesarean births are linked to type 1 diabetes (T1D). It is hypothesized that lack of exposure to the birth-canal microbiome may influence the risk of T1D. Children born by prelabor caesarean are not exposed to the birth-canal microbiome, and births by caesarean during labor (intrapartum caesarean) may be exposed.
The study aims to estimate risk of T1D among children delivered by caesarean compared with normal vaginal delivery (NVD), and whether the risk of T1D differs by caesarean type (prelabor caesarean, intrapartum caesarean).
Materials and Methods
This population-based study linked routinely-collected, de-identified perinatal, birth, and hospitalization data from the South Australian Early Childhood Data Project for all births from 1999 to 2013 (n=286,054). T1D cases were identified using inpatient hospitalizations ICD-10-AM diagnosis codes (E10-E109). Risks of T1D according to 1) all caesareans, and 2) stratified into prelabor or intrapartum caesarean, compared to NVD, were assessed by Cox proportional hazard regression. Analyses were adjusted for confounders identified a priori and involved multiple imputation to address missing information.
Results
There were 541 children with T1D diagnosed from 2001-2014. In this cohort, 31.7% children were delivered by caesarean and 68.3% had vaginal birth. Compared to NVD, the hazard ratio for all caesareans was 1.03 (95% CI 0.79, 1.34), for prelabor caesarean 1.00 (0.73, 1.37), and for intrapartum caesarean 1.06 (0.78, 1.43).
Conclusion
In this large whole-of-population study, no association was found between caesarean section and T1D, regardless of whether the neonate/fetus was exposed to the birth-canal microbiome.