Diurnal Cortisol Slope Mediates the Association Between Affect and Memory Retrieval in Older Adults With Mild Cognitive Impairment: A Path-Analytical Study

Abstract

Background Memory deficits are linked to dysfunctional HPA axis activity and negative affect in older adults. This study evaluated the mediating effect of the diurnal cortisol pattern on the relationship between affect and memory in older people with mild cognitive impairment (MCI). Methods This longitudinal study recruited 189 Chinese older adults with MCI from elderly centers in Hong Kong. The participants completed assessments of affect, salivary cortisol, and digit spans at baseline; neurocognitive assessments on verbal fluency, memory retrieval, and digit spans at 6-month follow-up; and instrumental activities of daily living (IADL) at 1-year follow-up. Structural equation modeling examined the direct and indirect effects of negative affect on memory and IADL via diurnal cortisol pattern. Results Controlling for covariates, negative affect significantly predicted flattened diurnal cortisol slopes (beta = 0.17, p < 0.05) but not memory or IADL (p = 0.23 - 0.91) directly. Diurnal cortisol slopes negatively predicted memory retrieval (beta = -0.20, p < 0.05), which in turn positively predicted IADL (beta = 0.22, p < 0.01). The indirect effect from negative affect to IADL via cortisol slope and memory retrieval was significant and negative (alpha beta gamma = -0.05, 95% bootstrapped CI = -0.248 to -0.001). Discussion The present study established certain temporal linkages among affect and cortisol slopes at baseline, memory retrieval at 6 months, and functional decline 1 year later in older adults with MCI. Flattened diurnal cortisol slopes might mediate the detrimental effects of negative affect on memory retrieval and functioning across 1 year.