Answers are grounded in this paper and cite their sources.
Adequate gestational thyroid function is important for offspring neurodevelopment, yet behavioural implications of variation within the normal range are unclear. We examined associations with externalising and internalising problems across repeated assessments in childhood and adolescence. 2,606 mother–child dyads from the Amsterdam Born Children and their Development (ABCD) cohort with maternal free thyroxine (FT4) and thyroid-stimulating hormone (TSH) measurements in the first half of pregnancy and at least one child SDQ assessment. Primary analysis excluded women with overt thyroid disease or thyroid medication use and children born < 32 weeks or < 1500 g. Linear mixed-effects models were applied to Strengths and Difficulties Questionnaire (SDQ) scores at 5–6, 11–12 and 15–16 years, via mother-, teacher-, and self-report. Secondary analyses included logistic regression for borderline SDQ cut-offs, sex interactions, and effect modification by TSH. FT4 showed a U-shaped association with teacher-reported externalising problems (quadratic β = 9.49, FDR-adjusted p = 0.03), with higher scores at both low and high FT4. Higher FT4 was associated with lower child-reported internalising problems (β = -0.17, FDR-adjusted p = 0.04). For both primary associations, there was no evidence of FT4-by-age interaction, suggesting differences in problem levels rather than thyroid-related differences in symptom change over time. At age 5–6, FT4 showed a non-linear association with teacher-reported hyperactivity/inattention above the threshold, with a negative linear term (OR = 0.83, 95% CI: 0.72 to 0.97) and a positive quadratic term (OR = 1.11, 95% CI: 1.02 to 1.20). At age 11–12, higher FT4 was associated with fewer self-reported peer problems (OR = 0.84, 95% CI:0.71–0.99). In exploratory analyses, curvature of the FT4-externalising association varied by maternal TSH; no sex modification was detected. Within a cohort free of overt thyroid disease, maternal thyroid function in the first half of pregnancy was associated with domain- and rater-specific differences in child internalising and externalising problems. We found no evidence that these associations changed across the available repeated assessments, suggesting differences in average problem scores rather than differences in how symptoms changed with age.
No takes yet. Share an insight, caveat, or question.
Keestra et al. (2026) studied this question.
Synapse has enriched 10 related papers. Consider them for comparative context: