Norwegian Registry Study Links Maternal Diabetes to Doubled Risk of Multiple Sclerosis in Offspring

A recently published retrospective study from Norway including more than 1 million individuals found that maternal diabetes was associated with a doubling of the child’s future risk of multiple sclerosis (MS), with additional risk increases seen in those born large for gestational age (LGA) and reduced risk observed among those born small for gestational age (SGA). Published in JAMA Neurology, the findings suggest that susceptibility to MS may begin as early as the prenatal period.¹

This closed cohort study comprised data from Norweigian national registers, including the Medical Birth Registry of Norway from 1967 to 1989. From January 2009 to 2019, live births (n = 1,303,802) were followed until the first event of MS diagnosis, death, emigration, or end of follow-up. In the analysis, Cox proportional hazards models were used to assess associations between adverse pregnancy outcomes and MS among participants aged at least 18 years at follow-up initiation who were MS-free in the prior year, with analyses conducted from February through October 2025.

Study authors, which included senior investigator Sarah E. Tom, PhD, MPH, assistant professor of epidemiology at Columbia University, identified 4295 MS cases among the 1,166,731 infants from the study. Using that original cohort, 114,864 (9.8%) participants were born SGA; 114,025 (9.8%) were born LGA; and 49,598 (4.2%) participants were exposed to maternal HDPs, 4778 (0.4%) to maternal placental abruption, and 2662 (0.2%) to maternal diabetes.

After adjusting for confounders, the hazard ratio (HR) for maternal diabetes was 2.15 (95% CI, 1.37-3.37) in the incidence sample and the odds of MS were higher (OR, 1.99; 95% CI, 1.32-3.01) in the prevalence sample. In the sibling analysis, maternal diabetes was not associated with MS, although estimates were imprecise due to few cases (OR, 1.17; 95% CI, 0.33-4.19). When gestational diabetes was modeled as the exposure, the point estimate was higher but still uncertain given only 5 exposed MS cases (OR, 1.86; 95% CI, 0.77-4.50), and no associations were observed for preterm birth, placental abruption, or hypertensive disorders of pregnancy.

“While it is well established that children with high BMI and diabetes are more likely to develop MS in adulthood, our findings suggest that the roots may lie in the perinatal period,” the study authors concluded. “Early metabolic exposures may influence immune system programming and future growth trajectories. Future epidemiologic studies should examine markers of neonatal adiposity and growth to better understand how early-life factors shape MS risk.”

After adjusting for confounders, the HR for being LGA vs AGA was 1.13 (95% CI, 1.03-1.25), while for being SGA it was less pronounced, at 0.88 (95% CI, 0.78-0.98) in the incidence sample. In the prevalence sample, the association between LGA and MS was attenuated (OR, 1.09; 95% CI, 1.00-1.19), whereas the association between SGA and MS remained similar (OR, 0.87; 95% CI, 0.79-0.96). Notably, an adjustment for additional confounders did not meaningfully change these findings.

READ MORE: Neuroprotective Agent Privosegtor Gains FDA Breakthrough Designation for Optic Neuritis

In the study, birth weight showed a linear relationship with both MS incidence and prevalence in the study population. Investigators reported that higher birth weight was associated with an increased risk of incident MS (HR, 1.07; 95% CI, 1.04-1.11) as well as a greater likelihood of prevalent MS (OR, 1.08; 95% CI, 1.05-1.11 per standard deviation increase in birth weight). In contrast, gestational age was not independently associated with MS outcomes, with no meaningful associations observed for either MS incidence (HR, 1.00; 95% CI, 0.98-1.01) or MS prevalence (OR, 1.00; 95% CI, 0.99-1.01). Notably, adjustment for diabetes did not alter the observed relationships.

Additional modeling analyses also supported the stability of the findings. When birth weight and gestational age were fitted using cubic splines with 4 knots, investigators found no improvement in model fit (log likelihood P > .05). High latitude was associated with MS (OR, 1.16; 95% CI, 1.10-1.23); however, including latitude as a confounder had minimal impact, shifting effect estimates by 2% or less. Across exposures, E-values ranged from 1.40 to 3.72, suggesting mild to moderate robustness to potential unmeasured confounding.

The study authors noted that the findings do not support previous observations of null effects of either birth weight, or birth weight for gestational age, on the risk of MS. “Individuals who are SGA and LGA may experience differences in BMI trajectories during childhood, which may then influence the MS risk,” they wrote. “While LGA is associated with early weight gain and childhood obesity, which is a well-established risk factor for MS, evidence on SGA is less consistent, with several studies reporting lower childhood BMI in this group, which could contribute to lower MS risk. Alternatively, there may be a critical window in early life during which adiposity has a stronger influence on MS risk—effects that later changes in BMI may not reverse.”

Although the study leveraged registry data from more than 1 million individuals, limitations include potential misclassification or underreporting of adverse pregnancy outcomes (APOs), particularly maternal diabetes, given lower-than-expected exposure prevalence. Residual confounding cannot be ruled out because key maternal factors (eg, obesity, smoking, vitamin D status) were unavailable, and evolving diagnostic practices over time for both APOs and MS may have introduced nondifferential misclassification that could bias results toward the null. Temporal shifts in obstetric/neonatal care and maternal risk profiles may also have influenced APO rates, and generalizability may be limited.

REFERENCE
1. Wolfova K, Engdahl BL, Horn J, et al. Maternal Pregnancy Outcomes and Offspring Risk of Adult-Onset Multiple Sclerosis. JAMA Neurol. Published online January 12, 2026. doi:10.1001/jamaneurol.2025.5255