Background Childhood obesity affects one in ten children in England by age 5, and one in five by age 11. Existing evidence suggests a possible relationship between caesarean section (CS) birth and higher risk of overweight/obesity in childhood, however maternal obesity is a strong confounder in this relationship. With CS rates rising by 4% globally per year, we aimed to examine the relationship between mode of delivery and overweight/obesity in childhood.

Methods SLOPE is a linked population-based cohort of anonymised routine antenatal, birth and child healthcare records in Hampshire, UK (2003–2018). Delivery method was categorised into unassisted vaginal delivery, assisted vaginal delivery and CS (including elective and emergency). Child body mass index (BMI) was measured as part of the National Child Measurement Programme in England. Children were identified as overweight/obese if their age- and sex-adjusted BMI was above the 85th percentile. Generalised linear modelling for outcome at two time points; 4–5 years (n=30,229) and 10–11 years (n=14,305) was conducted, adjusting for clustering within families. Modelling was introduced in stages with the choice of covariates informed by a Directed Acyclic Graph, first adjusting for maternal BMI, then adding in confounders including maternal age, ethnicity, educational attainment, parity, smoking status at booking appointment, pre-eclampsia, and previous CS (model C) and then birthweight and gestational age at birth as potential mediators (model M). Analyses were also stratified by maternal BMI category (underweight: <18.5, normal weight: 18.5 to <25, overweight: 25 to <30, obese: ≥30 kg/m²) at booking.

Results Of children delivered by CS, 25.0% and 33.7% were overweight/obese by 4–5 years and 10–11 years respectively, compared to 21.9% and 31.0% respectively with vaginal births. In unadjusted analysis, CS was associated with increased risk of overweight/obesity at 4–5 years (relative risk (RR) 1.13, 95% Confidence Interval (95% CI) 1.08–1.19), and at 10–11 years (RR 1.08, 95% CI 1.02–1.14), however both were attenuated by adjusting for maternal BMI. In stratified analyses, CS delivery was associated with increased risk of childhood overweight/obesity at 4–5 years only in normal weight women (model C: RR 1.15, 95% CI 1.04–1.27, model M: RR 1.14, 95% CI 1.02–1.26), but not in 10–11 year models.

Conclusion Maternal weight status at the start of pregnancy is a strong confounder in the relationship between mode of delivery and childhood overweight/obesity. In stratified analyses, this association was evident only for children of normal weight women. If this relationship is causal, the potential mechanisms need to be explored.