Background: Maternal diabetes seems to increase the risk of childhood leukemia. While metformin may reduce the risk of cancer in adults, the impact of antidiabetic medication on childhood cancer risk remains unknown.
Objectives: To study whether in-utero exposure to maternal diabetes is associated with an increased risk of childhood leukemia and lymphoma and whether treatment with antidiabetics reduces the risk of these cancer types in the child.
Methods: Using nationwide registry data from Denmark, Finland, Norway and Sweden (1996-2018), a cohort of 4,278,623 liveborn singletons was followed from birth until first cancer, death, emigration, 20th birthday or December 31st, 2020 (or latest available), whichever occurred first. We used propensity score (PS) fine stratification with 50 strata to control confounding. Covariates included in the PS models were source country, birth year, maternal age, multiparity, chronic hypertension, cardiovascular disease, smoking, maternal cancer before pregnancy, fertility treatment, dispensing of glucocorticoids, antihypertensive, thyroid and lipid modifying drugs. We estimated PS-weighted hazard ratios (wHR) with 95% confidence intervals (CI) for leukemia or lymphoma in children exposed to diabetes and to use of antidiabetics in-utero using Cox proportional hazards regression.
Results: The study material included 2470 cases of leukemia and 811 lymphomas diagnosed during an average follow-up of 9.98 years. Compared to no diabetes, the wHR of any leukemia in 186 children born to a diabetic mother was 1.35 (95% CI 1.15-1.58) and the wHR of acute lymphoid leukemia was 1.34 (95% CI 1.12-1.60, 150 exposed cases). The risk of any leukemia was increased both in children exposed to chronic diabetes (wHR 1.69; 95% CI 1.27-2.26) and those exposed to gestational diabetes (wHR 1.26; 95% CI 1.04-1.51). When using children of unmedicated diabetic mothers as a reference, the wHRs were further increased for any leukemia in children exposed to any antidiabetics (wHR 1.45; 95% CI 1.01-2.08, 58 exposed cases) and for acute lymphoid leukemia in children exposed to insulin (wHR 2.04; 95% CI 1.35-3.07, 37 exposed cases). We did not find an association between maternal diabetes and lymphoma (wHR 0.96; 95% CI 0.67-1.34, 41 exposed cases) or its subtypes.
Conclusions: Maternal diabetes increased the risk of any childhood leukemia but not lymphoma. Treatment with antidiabetics tended further to increase the risk of leukemia and particularly that of acute lymphoid leukemia probably reflecting the severity of diabetes. Our findings support hyperinsulinemia, regardless of whether it is of endogenous or exogenous origin, as a mechanism by which maternal diabetes might promote childhood leukemia development.
