Mothers’ Autoimmune Diseases Linked to Higher Infant Mortality in Landmark Study of 1.4 Million Children

A sweeping analysis of more than 1.4 million Swedish births has found that children whose mothers live with autoimmune diseases face a modestly elevated risk of dying before reaching adulthood, with the excess danger concentrated almost entirely in the first year of life. The nationwide cohort study, published in eClinicalMedicine, tracked offspring born between 2001 and 2014 for up to two decades, and its authors say the pattern points to disturbances during fetal development and the perinatal period, rather than lingering risks that follow children into later childhood and beyond.

The research team, led by investigators at Karolinska Institutet, linked six national health and socioeconomic registers covering every live-born singleton in Sweden across a fourteen-year window. Of the 1,455,645 children included, 90,046, or 6.2 percent, were born to mothers diagnosed with an autoimmune disease before or during pregnancy. These conditions, which range from type 1 diabetes and autoimmune thyroiditis to rheumatoid arthritis, lupus and inflammatory bowel disease, arise when the immune system mistakenly attacks the body’s own tissues, sustaining chronic inflammation and producing autoantibodies that can cross the placenta.

Over a median follow-up of 15.3 years, yielding more than 22 million person-years of observation, 426 deaths occurred among children of mothers with autoimmune diseases compared with 5,728 deaths among the unexposed. After adjusting for birth year, maternal age, education, country of birth, parity, marital status, smoking in early pregnancy and pre-existing maternal cardiovascular disease and type 2 diabetes, children of exposed mothers had a 22 percent higher risk of death from any cause, corresponding to a hazard ratio of 1.22. The estimate proved stable across successive statistical models, suggesting that observable socioeconomic and health differences between the groups could not account for the association.

Beneath the overall figure lay striking heterogeneity. Maternal type 1 diabetes roughly doubled offspring mortality risk, with a hazard ratio of 1.99, and elevated risks also appeared for connective tissue diseases and autoimmune thyroiditis. In contrast, maternal skin, blood and gastrointestinal autoimmune conditions showed little or no association. The endocrine diseases carry particular biological weight because they combine immune dysregulation with metabolic disturbance, a combination known to impair placental function and fetal growth, and the study’s authors note that the type 1 diabetes findings align with earlier smaller studies that reported lasting consequences for children exposed to hyperglycemia in utero.

Perhaps the most consequential discovery concerned timing. When follow-up was divided by age, the association was strongest during infancy, with a hazard ratio of 1.21 in the first year of life, while estimates for childhood, adolescence and young adulthood were weaker and statistically imprecise. Cause-of-death analysis reinforced this developmental picture: the excess risk was driven overwhelmingly by natural causes, specifically conditions originating in the perinatal period, with a hazard ratio of 1.43, and congenital malformations and chromosomal abnormalities, at 1.29. There was no clear association with infections, cancers, circulatory or nervous system diseases, and no association whatsoever with unnatural deaths such as accidents, an outcome-based negative control that strengthens confidence in the specificity of the biological signal.

Determining whether maternal disease itself causes harm, rather than merely marking shared family vulnerabilities, is the central challenge of any observational study of this kind. The researchers deployed two complementary designs. First, they compared maternal autoimmune diseases with paternal autoimmune diseases during the same pregnancies. Because fathers share genes and household environments with their children but cannot influence fetal development through the intrauterine route, a paternal signal would suggest familial confounding. Instead, paternal autoimmune diseases showed essentially no association with offspring mortality, while the maternal association remained strong, and the maternal-to-paternal hazard ratio reached 1.40 for any autoimmune disease and 2.73 for type 1 diabetes.

Second, cousin comparisons, in which offspring of sisters discordant for autoimmune disease were contrasted while stratifying on the maternal grandmother, partially controlled for shared genetics and early-family environment. These analyses attenuated the overall association to a hazard ratio of 1.04, though the confidence interval widened considerably because far fewer informative comparisons were available. Maternal type 1 diabetes retained a substantially elevated estimate even in this restricted sample, at a hazard ratio of 2.63. Taken together, the authors argue, the family-based evidence is consistent with the hypothesis that shared familial factors alone do not fully explain the findings, leaving room for genuine intrauterine mechanisms, while acknowledging that extrauterine pathways tied to pregnancy and the postnatal period cannot be excluded.

To probe those mechanisms, the team turned to formal mediation analysis within a counterfactual framework, asking whether adverse birth outcomes carry the excess risk from maternal disease to offspring death. Preterm birth emerged as the dominant candidate, mediating an estimated 53 percent of the all-cause association and roughly three-quarters of the association with perinatal-cause deaths, though wide confidence intervals counsel caution about the precise magnitude. Small for gestational age birth mediated a smaller but notable share, about 17 percent of all-cause mortality and 24 percent of perinatal deaths, while congenital heart disease accounted for around 7.5 percent. Large for gestational age, intriguingly, mediated nothing, suggesting the effect runs through restricted growth and early delivery rather than through fetal overgrowth.

The biological logic supporting these findings is well established in developmental science. Chronic maternal inflammation, immune dysregulation and transplacental autoantibody transfer can damage placental function and disrupt fetal development, producing structural and physiological alterations that heighten vulnerability in the first weeks of life. In pregnancies complicated by type 1 diabetes, metabolic and inflammatory disturbances compound one another. Notably, the cohort was born from 2001 onward, an era in which clinical management of autoimmune disease in pregnancy improved substantially, yet excess early-life mortality persisted, indicating that suboptimal care alone is unlikely to explain the pattern. Sex-stratified analyses also hinted at greater vulnerability in boys, with a 28 percent increased risk in males versus 14 percent in females, echoing the well-documented male disadvantage in perinatal survival.

The study’s limitations temper but do not overturn its conclusions. Register data cannot capture diagnoses made exclusively in primary care, medication exposures during pregnancy or the full severity spectrum of maternal disease, and residual confounding by unmeasured behaviors or pregnancy management remains possible. Because only live births were included, any effect of maternal autoimmune disease on fetal loss would have been invisible, potentially underestimating the true impact. Some subgroup estimates rest on small numbers of deaths and must be treated as hypothesis-generating. Still, the authors emphasize that if confirmed, the findings identify pregnancy and the perinatal window as critical opportunities for intervention: optimizing maternal disease control before and during conception, with particular vigilance for endocrine conditions such as type 1 diabetes, alongside strengthened perinatal care, could reduce a measurable share of preventable early-life mortality.

Subject of Research: Association between maternal autoimmune diseases during pregnancy and offspring mortality from birth to young adulthood

Article Title: Maternal autoimmune diseases during pregnancy and mortality up to young adulthood in offspring: a register-based cohort study

Article References: Zhang, H., Janszky, I., Arkema, E. V., Möller, J., Liang, Y., Mo, X., & László, K. D. (2026). Maternal autoimmune diseases during pregnancy and mortality up to young adulthood in offspring: a register-based cohort study. eClinicalMedicine, Article 104237. https://doi.org/10.1016/j.eclinm.2026.104237

Image Credits: AI Generated

DOI: 10.1016/j.eclinm.2026.104237

Keywords: maternal autoimmune disease, offspring mortality, type 1 diabetes, perinatal outcomes, preterm birth, congenital anomalies, Swedish national registers, cohort study, intrauterine mechanisms, epidemiology, fetal development, autoimmune thyroiditis