Maternal autoimmune disease in pregnancy tied to higher type 1 diabetes risk in children

A mother with lupus, rheumatoid arthritis or an autoimmune thyroid condition can spend a pregnancy wondering whether her own immune system will leave a mark on her child. A large Swedish study published on 26 October in JAMA Pediatrics gives a partial answer for one immune-mediated disease — type 1 diabetes — and the result is more reassuring than the headline figure first suggests [s1].

What the study looked at

The researchers used Sweden's national health registers to follow 1,455,645 children born between 2001 and 2014, tracking each one from birth until a type 1 diabetes diagnosis, death, emigration or the end of 2023 [s1]. Over a median of 15.3 years, 10,273 children — 0.71% of the group — developed the condition [s1]. The cohort was split almost evenly between boys and girls, at 51.4% male [s1].

Because the registers record diagnoses for both parents, the team could do something most studies cannot: compare the effect of a mother's autoimmune disease against a father's. That comparison is the heart of the paper. If pregnancy itself — the intrauterine environment a mother provides — were driving the child's risk, mothers should matter far more than fathers [s1].

What it found

Children whose mothers had an autoimmune disease during pregnancy were more likely to develop type 1 diabetes, with an adjusted hazard ratio of 1.57 (95% CI, 1.46 to 1.68) [s1]. In plain terms, their risk was about 57% higher than that of children whose mothers had no such diagnosis.

The striking part is what happened on the paternal side. Fathers with an autoimmune disease before their child's birth carried almost the same elevated risk, with an adjusted hazard ratio of 1.51 (95% CI, 1.37 to 1.67) [s1]. When the researchers compared the two directly, there was no overall difference between mothers and fathers (ratio of hazard ratios 1.04; 95% CI, 0.92 to 1.17) [s1]. A gradient ran through the data by parental diabetes status: risk was highest when both parents were affected, then paternal type 1 diabetes, then maternal [s1].

That symmetry matters. If something unique to pregnancy were the main mechanism, fathers would not track so closely with mothers. A father does not carry a pregnancy, so his immune system cannot shape the intrauterine environment his child develops in; whatever risk he transmits must travel through genes or a shared home. The near-equal maternal and paternal estimates therefore point to shared family susceptibility — the genes and environment relatives hold in common — rather than the womb itself.

To probe that further, the team ran a cousin analysis, which compares children who share grandparents and therefore some inherited risk. Here the maternal association shrank but did not vanish, falling to a hazard ratio of 1.27 (95% CI, 1.08 to 1.50) [s1]. The persistence of a signal after accounting for shared family factors leaves room for some genuinely prenatal contribution, even if it is smaller than the raw number implies [s1].

How to read it

Several things temper alarm. The absolute risk stays low: fewer than one child in a hundred developed type 1 diabetes across the whole cohort, and a 57% relative increase on a small baseline is still a small number in real terms [s1]. The study is observational, so it can show association but not prove that a mother's condition causes her child's diabetes. And the headline maternal figure blends inherited risk with anything specific to pregnancy — which is exactly why the paternal comparison and cousin analysis were done [s1].

The authors also note heterogeneity across specific conditions: some autoimmune diseases showed parent-specific patterns, suggesting a handful may act through routes unique to the mother or father rather than through shared genetics alone [s1]. Those disease-by-disease differences are where any truly intrauterine mechanism is most likely to hide, and they will need dedicated study rather than a single pooled estimate. A condition in which maternal antibodies cross the placenta, for instance, could plausibly leave a prenatal mark that a paternal diagnosis would not — but establishing that for any individual disease requires far more cases than a broad grouping provides.

It is also worth being clear about what "autoimmune disease during pregnancy" covers. The registers capture a broad spectrum of conditions diagnosed before or during pregnancy, from thyroid disease to inflammatory arthritis to lupus, lumped together for the headline estimate [s1]. These differ enormously in how they behave, how they are treated and whether they involve antibodies that can reach a fetus, so the single pooled hazard ratio is best read as an average across a varied set rather than a figure that applies evenly to any one diagnosis.

The study was published alongside an invited commentary in the same issue of the journal [s2]. For families, the practical takeaway is narrow. A parent's autoimmune disease is one of many factors that nudge a child's type 1 diabetes risk, most of it inherited and none of it currently preventable. There is no screening or intervention this study calls for, and nothing a mother did or did not do in pregnancy emerges as the culprit. What the work offers clinicians is a clearer map of where the risk comes from — and a reminder that a mother's diagnosis, viewed in isolation, overstates her particular role.