SFEBES2023 Poster Presentations Thyroid (63 abstracts)
Cardiff University, Cardiff, United Kingdom
Introduction: The Controlled Antenatal Thyroid Screening II (CATS) study, a large randomised trial of thyroxine supplementation for suboptimal gestational thyroid function (SGTF), reported higher attention deficit hyperactivity disorder (ADHD) scores in 9 year-old children exposed to higher thyroid hormone (TH) in utero. We investigated if this was accompanied by altered myelination.
Methods: Eighty-five children aged 11-16 years (untreated SGTF (n=21), normal GTF (n=24), or treated SGTF (optimally treated (n=21), over-treated (n=19)) recruited from the CATS cohort underwent quantitative characterisation of white matter microstructure and myelination using 3.0T MRI. Myelination was assessed by quantitative magnetization transfer (qMT) and compared to diffusion-weighted MRI (dMRI). qMT parameters included macromolecular proton (bound pool; BPF) fraction. White matter (WM) bundles were reconstructed from dMRI data in tracts known to be affected by TH exposure and/or implicated in ADHD risk (corpus callosum, cingulum bundles, inferior and superior longitudinal fasciculi). The median BPF was calculated for each bundle. Relationships between thyroid hormone, treatment group and BPF for each WM bundle were assessed using linear mixed effects models. Child age, sex and Tanner scores were included as fixed effects.
Results: No associations were found between free T4 or TSH at 12 weeks and median BPF along studied tracts. No effect of treatment group on median BPF was identified, despite significant differences in fractional anisotropy (FA). Linear regression revealed significant effects of treatment on FA in the Corpus Callosum, with posthoc tests revealing greater FA in over-treated compared with the untreated group (P=0.033). Correlations between FA and BPF were weak and not statistically significant.
Conclusion: This is the first study to explore the effects of in utero TH exposure on myelination in humans. Lack of correlation between FA and BPF suggests that differences in FA observed between over-treated and untreated individuals were not driven by myelin.