Thyroid hormone (TH) is crucial during neurodevelopment, but high levels of TH have been linked to neurodegenerative disorders. No data on the association of thyroid function with brain imaging in the general population are available. We therefore investigated the association of thyroid-stimulating hormone and free thyroxine (FT4) with magnetic resonance imaging (MRI)-derived total intracranial volume, brain tissue volumes, and diffusion tensor imaging measures of white matter microstructure in 4683 dementia- and stroke-free participants (mean age 60.2, range 45.6–89.9 years). Higher FT4 levels were associated with larger total intracranial volumes (β = 6.73 mL, 95% confidence interval = 2.94–9.80). Higher FT4 levels were also associated with larger total brain and white matter volumes in younger individuals, but with smaller total brain and white matter volume in older individuals (p-interaction 0.02). There was a similar interaction by age for the association of FT4 with mean diffusivity on diffusion tensor imaging (p-interaction 0.026). These results are in line with differential effects of TH during neurodevelopmental and neurodegenerative processes and can improve the understanding of the role of thyroid function in neurodegenerative disorders.

Objective: To investigate the association of cerebral white matter microstructural integrity with mortality. Methods: We included 4,294 individuals, free from stroke and dementia (mean age 63.6 years, 44% male) from the population-based Rotterdam Study (2006-2011). Diffusion-MRI was used to assess the microstructural integrity of the white matter, both globally and for specific white matter tracts. Fractional anisotropy and mean diffusivity were the measures used to quantify white matter integrity. All-cause mortality and cause-specific mortality was recorded with a median follow-up time of 5.4 and 4.6 years, respectively. Cox regression models, adjusted for age, sex, APOE ϵ4 allele carriership, cardiovascular risk factors, and macrostructural MRI changes, were used to estimate hazard ratios. Results: During the follow-up time, 216 participants (5.0%) died of all causes, 31 (0.7%) of cardiovascular causes, and 102 individuals (2.4%) died of noncardiovascular causes. Each SD decrease in fractional anisotropy and each SD increase in mean diffusivity was associated with a 1.37-fold (95% confidence interval: 1.20-1.57) and a 1.49-fold (95% confidence interval: 1.28-1.75) higher hazard of all-cause mortality, respectively. The associations were more prominent with cardiovascular mortality than with noncardiovascular mortality. In tract-specific analyses, we observed that association tracts were more prominently related to mortality. Conclusions: Our findings suggest that impairments in cerebral white matter, even at early stages, are not limited to adverse brain outcomes and they are related to mortality, especially from cardiovascular causes.