Associations of Circadian Rhythms with Cognitive Performance in Patients with Amnestic Mild Cognitive Impairment (aMCI)

Background/Objectives: Circadian rhythm disruption is linked to cognitive decline, yet it remains unclear how behavioral and physiological rhythm markers are differently associated with cognition in amnestic mild cognitive impairment (aMCI). The primary aim of this study was to compare sleep–wake timing, rest–activity rhythm (RAR), and dim light melatonin onset (DLMO) between patients with aMCI and cognitively normal controls. Exploratory analyses further examined their associations with domain-specific cognitive performance. Methods: Eighteen aMCI patients and 21 cognitively normal controls (NC) enrolled. Cognitive function was assessed using the Korean version of the Consortium to Establish a Registry for Alzheimer’s Disease Neuropsychological Battery (CERAD-K). Participants underwent 5-day actigraphy to assess sleep–wake timing and non-parametric RAR variables, including interdaily stability (IS), intradaily variability (IV), and relative amplitude (RA). DLMO was determined from hourly salivary melatonin samples collected over five hours before sleep onset under dim-light conditions. Group comparisons of circadian markers were conducted as the primary analyses, and generalized linear models were used for exploratory analyses of associations between circadian markers and cognitive outcomes. Results: Groups did not significantly differ in sleep–wake timing, RAR parameters and DLMO. Sleep–wake timing variables and DLMO were not significantly associated with cognitive performance. Higher IS was associated with better visuospatial memory and executive function, whereas higher RA was associated with poorer verbal memory among aMCI patients. Conclusions: Although sleep–wake timing and melatonin phase did not differ between groups nor predict cognitive performance, higher daily rhythm stability was linked to better non-verbal memory and executive functioning. In contrast, high RA may relate to poorer verbal memory in aMCI, suggesting that elevated RA may not reflect true circadian robustness required for optimal cognition.