AIMS: To examine associations of life-course stress with cognition and diurnal cortisol patterns in older adulthood, as well as potential mediation effects of diurnal cortisol patterns and perceived stress on the association between life-course stress and cognition.

METHODS: 127 participants without dementia were selected from a cohort of Swedish memory clinic patients. Cross-sectional associations between scores on two chronic stress questionnaires (perceived stress, stressful life events (SLEs)), five cognitive domains (overall cognition, memory, working memory, processing speed, perceptual reasoning), and two measures of diurnal cortisol patterns (total daily output, diurnal cortisol slope), as well as potential mediation effects of diurnal cortisol patterns and perceived stress on associations between life-course stress and cognition, were assessed using linear regressions.

RESULTS: Greater lifetime exposure to SLEs was associated with worse memory, working memory, and processing speed performance, but not with diurnal cortisol patterns. A greater number of SLEs in late childhood was associated with worse working memory and processing speed, while a greater number of SLEs in non-recent adulthood were associated with better overall cognition and perceptual reasoning. Greater perceived stress was associated with a flattened diurnal cortisol slope, but not with cognition. No evidence for interplay between self-reported and physiological stress markers was found in relation to cognition, although there appeared to be a significant positive indirect association between economic/legal SLEs and the diurnal cortisol slope via perceived stress.

CONCLUSIONS: The associations between SLEs and cognition depend on the period during which SLEs occur, but seem independent of late-life cortisol dysregulation.