News · Heart & Metabolic
Children exposed to diabetes in the womb developed it five times as often
Gestational diabetes is described as ending at delivery. Following 2.85 million children to a median age of 16, exposure before birth tracked far higher diabetes rates, and control around conception shifted the odds.
- Children of mothers with diabetes before pregnancy developed it about five times as often.
- Gestational diabetes, which resolves at birth, still tracked 75% higher risk in the child.
- Better blood sugar control around conception tracked lower risk in the child.
- Rates were higher again among First Nations families, at nearly eight times.
- 2.85 million children followed to a median age of 16.
Gestational diabetes is explained to patients as temporary. It appears in pregnancy, it is managed through pregnancy, and it goes away at delivery. The follow-up advice concerns the mother’s own risk of type 2 diabetes later.
The child is usually left out of that conversation, which is odd, because the child is the one who spent nine months in the altered environment.
Writing in Diabetes Care, researchers followed 2,851,583 children to a median age of 16 years and counted how many developed diabetes themselves.
The gradient
Children whose mothers had diabetes before becoming pregnant developed diabetes at around five times the rate of children whose mothers had none.
Children exposed to gestational diabetes, the kind that resolves at delivery, developed it at around 1.75 times the rate. Lower than the preexisting group and a long way above the unexposed.
Among children of First Nations mothers both figures were higher: roughly 7.7 times and 2.6 times respectively.
The number that moves this past genetics
The obvious objection writes itself. Type 2 diabetes is strongly heritable, mothers with diabetes pass on genes as well as a uterus, and a fivefold gap is what family history looks like.
The study has an answer, and it is the most important line in the paper. Among children of mothers with preexisting diabetes, each 1% lower blood sugar measure around conception carried about 11% lower risk in the child.
That is a gradient among mothers who all had diabetes. Genes do not vary with how well a mother’s blood sugar happened to be controlled in the months around conception. Something about the pregnancy environment itself is doing work.
It is not proof. Mothers with better control differ from mothers with worse control in income, health literacy, access to care and much else, all of which travel to the child by other routes. But it is the kind of within-group gradient that pure inheritance does not predict.
What the exposure might be doing
The leading account is developmental programming: a fetus developing in a high-glucose environment adapts to it, and those adaptations persist into a life that no longer matches.
Higher maternal glucose crosses the placenta, the fetal pancreas responds by producing more insulin, and growth and fat deposition shift accordingly. Whether that early tuning is what shows up as diabetes twenty years later is a hypothesis this study is consistent with rather than one it tests.
Reading the First Nations figures properly
The higher figures among First Nations families are real and they are not an explanation of anything on their own.
Gestational diabetes is caused by the hormonal changes of pregnancy, along with genetic and lifestyle factors, and it is more common where diabetes is more common generally. The conditions that make it so are material: food cost and availability, access to care, and the long consequences of dispossession. A number like this measures an outcome, not a cause.
That the study was First Nations-engaged and co-led matters here. Research on Indigenous health has a poor history of reporting differences without the communities concerned having any say in how they are framed.
The proportions worth keeping straight
Five times a small number is still a small number. Diabetes before the mid-twenties remains uncommon, and the large majority of exposed children in this cohort did not develop it.
Where a fivefold figure bites is at scale. Gestational diabetes affects a substantial and rising share of pregnancies, and a modest per-child risk multiplied across that many pregnancies produces a lot of diabetes a generation later.
What it changes
Blood sugar control before and during pregnancy is already the goal, for reasons concerning the mother and the delivery. This adds a reason with a twenty-year horizon.
It also argues for something the current framing discourages: that gestational diabetes is worth a note in the child’s record, not just the mother’s. A child whose exposure is known can be watched. A child whose exposure ended, on paper, at delivery cannot.
People also ask
What did the study find?
Among children of non-First Nations mothers, adjusted hazard ratios for incident diabetes were 5.18 (95% CI 4.82-5.57) with maternal preexisting diabetes and 1.75 (1.65-1.84) with gestational diabetes. Among children of First Nations mothers the figures were 7.68 (5.91-9.98) and 2.59 (2.07-3.24). Each 1% lower periconception HbA1c carried a hazard ratio of 0.89 (0.80-0.995) in children of mothers with preexisting diabetes.
What is periconception HbA1c?
A measure of average blood sugar over the preceding two to three months, taken around the time of conception. It reflects how well diabetes was controlled in the window when the pregnancy was starting.
Is this just inherited genes?
Partly, certainly. Type 2 diabetes runs strongly in families. What the HbA1c finding adds is a gradient within mothers who all had diabetes, which genes alone would not produce, and which points at the intrauterine environment as well.
Why does gestational diabetes matter if it resolves?
Because the exposure the child had does not resolve. The pregnancy happened at higher blood sugar whatever occurred afterwards, and this cohort shows the child's risk raised long after the mother's condition ended.
Why were rates higher in First Nations families?
The study was First Nations-engaged and co-led, and reports the difference rather than explaining it away. Higher background diabetes prevalence, and the material and historical conditions that produce it, sit behind figures like these.
How big is the absolute risk?
Diabetes in children and young adults remains uncommon, so a fivefold relative increase still leaves most exposed children unaffected. The figures matter at population scale, across millions of pregnancies.
What could a family do with this?
Blood sugar control before and during pregnancy is already a goal of care, and this adds a reason that extends to the child. Any change to diabetes management in pregnancy belongs with a clinical team. This is general information rather than medical advice.