Heart HealthResearch PaperOpen Access

Maternal Diabetes Raises Offspring Cardiovascular Disease Risk by Up to 29%

A Swedish cohort of 4.3 million found maternal diabetes during pregnancy significantly increases offspring CVD risk, with pregestational diabetes driving the highest hazard.

Monday, May 18, 2026 0 views
Published in JAMA Cardiol
A pregnant woman in a clinical setting having blood glucose measured by a nurse, with a newborn in a hospital bassinet visible in the background

Summary

A nationwide Swedish study of over 4.27 million people born between 1973 and 2014 found that children of mothers with diabetes during pregnancy face a 16% higher overall risk of cardiovascular disease — rising to 29% for pregestational diabetes. Specific CVDs elevated include heart failure (65% higher risk), cerebrovascular disease (31%), atrial fibrillation (27%), and venous thromboembolism (20%). Congenital heart disease, preterm birth, and being large for gestational age partly explained these links, mediating approximately 32%, 16%, and 14% of the association respectively. Sibling analyses confirmed the findings persist after accounting for shared family factors, pointing to prenatal exposure as a genuine independent contributor to long-term cardiovascular risk.

Detailed Summary

Maternal diabetes during pregnancy — including both gestational diabetes mellitus (GDM) and pregestational diabetes (type 1 and type 2) — is already known to raise risks of complications at birth, but its long-term cardiovascular consequences for offspring have remained poorly characterized. This large Swedish registry study addresses that gap with unprecedented statistical power, following more than 4.27 million individuals for a median of 27.6 years (IQR 17.2–37.4 years), making it one of the most comprehensive investigations of intergenerational cardiovascular risk ever published.

The cohort included everyone born in Sweden between January 1, 1973, and December 31, 2014, with follow-up through December 31, 2023. Of the total, 61,336 individuals (1.46%) were exposed to maternal diabetes in utero. Cardiovascular disease diagnoses were captured from national inpatient and outpatient registers. Exposures were categorized as any maternal diabetes, gestational diabetes only, and pregestational diabetes (type 1 or type 2). Cox proportional hazards regression models estimated hazard ratios after adjusting for relevant covariates. Crucially, sibling-matched analyses were conducted to control for shared familial genetic and environmental confounders.

The primary finding was that any maternal diabetes was associated with a 16% increased risk of overall CVD in offspring (HR 1.16; 95% CI, 1.12–1.20). The risk gradient was clearly related to diabetes severity: gestational diabetes conferred an HR of 1.11 (95% CI, 1.05–1.17), while pregestational diabetes carried an HR of 1.29 (95% CI, 1.21–1.38). Among specific CVD subtypes, the most striking association was with heart failure (HR 1.65; 95% CI, 1.37–2.00), followed by cerebrovascular diseases (HR 1.31; 95% CI, 1.12–1.52), atrial fibrillation (HR 1.27; 95% CI, 1.05–1.54), and venous thromboembolism (HR 1.20; 95% CI, 1.07–1.34). These findings held in sibling analyses, strengthening the causal interpretation.

Mediation analyses identified three key perinatal and early-life pathways. Congenital heart disease (CHD) was the largest mediator, directly and/or indirectly explaining 31.87% of the association between maternal diabetes and offspring CVD risk. Preterm birth mediated 16.06% of the association, and being large for gestational age (LGA) mediated 14.18%. These results suggest that a substantial proportion of risk operates through identifiable, clinically actionable perinatal events, while meaningful residual risk remains unexplained.

The clinical and public health implications are substantial. With gestational diabetes prevalence rising globally alongside obesity rates, the downstream CVD burden in offspring populations could grow considerably. Identifying offspring of diabetic mothers — especially those also born preterm, with CHD, or LGA — as a high-risk group from birth could enable earlier cardiovascular surveillance, lifestyle intervention, and preventive care. For clinicians, these findings argue for flagging maternal diabetes in pediatric and adult medical records as a cardiovascular risk factor in its own right.

Key Findings

  • Maternal diabetes associated with 16% higher overall CVD risk in offspring (HR 1.16; 95% CI, 1.12–1.20) across 4.27 million individuals
  • Pregestational diabetes carried a 29% higher offspring CVD risk (HR 1.29; 95% CI, 1.21–1.38) vs. 11% for gestational diabetes (HR 1.11; 95% CI, 1.05–1.17)
  • Heart failure risk in offspring was 65% higher in those exposed to maternal diabetes (HR 1.65; 95% CI, 1.37–2.00)
  • Cerebrovascular disease risk was 31% elevated (HR 1.31; 95% CI, 1.12–1.52) and atrial fibrillation 27% elevated (HR 1.27; 95% CI, 1.05–1.54)
  • Venous thromboembolism risk was 20% higher (HR 1.20; 95% CI, 1.07–1.34) in offspring of mothers with diabetes
  • Congenital heart disease mediated 31.87%, preterm birth 16.06%, and large for gestational age 14.18% of the maternal diabetes–offspring CVD association
  • Associations persisted in sibling-matched analyses, supporting independence from shared familial/genetic confounders

Methodology

Nationwide population-based cohort using linked Swedish national registers; 4,274,414 individuals born 1973–2014 followed through December 31, 2023 (median 27.6 years). Exposures were gestational and pregestational diabetes identified from health registers; CVD outcomes captured from inpatient and outpatient national registers. Cox proportional hazards regression estimated adjusted HRs; sibling-matched analyses controlled for shared familial factors; counterfactual mediation analyses quantified contributions of CHD, preterm birth, and LGA.

Study Limitations

The study relied on register-based diagnoses, which may undercount milder or undiagnosed cases of both maternal diabetes and offspring CVD, potentially underestimating true associations. Detailed clinical data on diabetes severity, glycemic control, and medication use during pregnancy were not available, limiting mechanistic inference. Residual confounding from unmeasured socioeconomic, lifestyle, and genetic factors cannot be fully excluded despite sibling analyses; no conflicts of interest were reported.

Enjoyed this summary?

Get the latest longevity research delivered to your inbox every week.