Alteration in intrinsic and extrinsic functional connectivity of resting state networks associated with subclinical hypothyroidism

Abstract Subclinical hypothyroidism (SCH) is characterised by a mild elevation of thyroid‐stimulating hormone (range 5‐10 μIU mL ‐1 ) and normal free triiodothyronine and free thyroxine. Cognitive function impairment is well known in thyroid disorders such as hypothyroidism and hyperthyroidism, although little is known about deficits in brain functions in SCH subjects. Also, whether hormone‐replacement treatment is necessary or not in SCH subjects is still debatable. To obtain an insight into the cognition of SCH subjects, intrinsic and extrinsic functional connectivity (FC) of the resting state networks (RSNs) was investigated. For resting state data analysis, we used an unbiased, data‐driven approach based on independent component analysis and dual‐regression, which can emphasise widespread changes in FC without restricting to a set of predefined seeds. In total, 28 SCH subjects and 28 matched healthy controls (HCs) participated in the study. RSN analysis showed significantly decreased intrinsic FC in somato‐motor network (SMN) and right frontoparietal attention network (RAN) and increased intrinsic FC in default mode network (DMN) in SCH subjects compared to control subjects. The reduced intrinsic FC in the SMN and RAN suggests neurocognitive alterations in SCH subjects in the corresponding functions, as also evident from the deficit in the neuropsychological performance of the SCH subjects on behavioural tests, such as digit span, delayed recall, visual retention, recognition, Bender Gestalt and Mini‐Mental State Examination. We also found a significant reduction in extrinsic network FC between DMN and RAN; SMN and posterior default mode network; and increased extrinsic FC between SMN and anterior default mode network in SCH subjects compared to controls. An altered extrinsic FC in SCH suggests functional reorganisation in response to neurological disruption. The partial correlation analysis between intrinsic and extrinsic RSN FC and neuropsychological performances, as well as clinical indices, provides interesting insights into the brain‐behaviour relationship in SCH subjects.