Prenatal iron exposure and childhood type 1 diabetes

Population-based prospective cohort (Norwegian MoBa). Participants: pregnant women in Norway recruited mid-pregnancy (1999-2008); 94,209 pregnancies with complete exposure data; offspring followed for type 1 diabetes to May 1, 2017; 373 T1D cases (188 girls, 50.4%). Nested sub-studies included cord plasma iron biomarkers (ferritin and soluble transferrin receptor) and HFE genotyping; cord blood DNA methylation (1,062 participants) with replication in ALSPAC (≈1,000 newborns); additional microbiota analyses in NoMIC (552 infants).

Iron-containing dietary supplements taken during pregnancy. Median daily dose among users: 14 mg/day (mean 26.9 mg/day, SD 39.4 mg/day). Duration across pregnancy varied; timing categorized as use ≤17 weeks, >17 weeks, or both; cumulative dose estimated from daily dose and duration.

Maternal iron supplement use during pregnancy was associated with higher risk of childhood type 1 diabetes in offspring: HR 1.33 (95% CI 1.06-1.67). Early exposure (≤17 weeks) associated with HR 1.75 (1.08-2.86). Iron-only supplements: HR 1.71 (0.83-3.53); iron with other nutrients: HR 1.60 (0.95-2.69). Iron from foods and maternal anaemia were not risk factors. Maternal HFE genotype associated with offspring T1D; fetal HFE genotype linked to differential methylation near HFE in cord blood. Mid-pregnancy inflammatory cytokine patterns differed by iron supplementation. Excess prenatal iron exposure may increase T1D risk, but causality cannot be established; replication and mechanistic studies needed; findings suggest cautious consideration of universal prenatal iron supplementation guidelines.

Observational design; potential residual confounding; limited power for rarer genotypes; reliance on self-reported supplement use; generalizability to non-Norwegian populations.