Background: The objectives of this study were to examine the factors modifying the relationship between cortisol level and prevalent/incident cognitive impairment in older adults and to verify whether these relationships were non-linear.
Methods: Data were collected from 1,226 individuals aged 65 and older by two in-home interviews separated by 12 months. Cortisol level was measured using saliva samples taken at the beginning of the baseline interview before cognitive, mental, and physical health evaluations. Prevalent and incident cognitive impairment were defined using the Mini-Mental State Examination scores according to normative data for age, education level, and sex.
Results: High morning cortisol level increased the risk of incident cognitive impairment in participants with anxiety or depressive episode while low cortisol level increased the risk in participants without anxiety or depressive episode. In high educated participants, but not in low educated participants, high morning cortisol level was associated with prevalent cognitive impairment and high afternoon cortisol level increased the risk of incident cognitive impairment. The results also suggested that lower morning cortisol values could increase the risk of incident cognitive impairment in individuals with few chronic diseases. A curvilinear relationship was observed between morning cortisol and the probability of incident cognitive impairment, but further analyses suggested that it was likely explained by anxiety and depressive episode.
Conclusions: These results suggest that cognitive impairment in older adults is linked to higher or lower cortisol level depending on characteristics such as anxiety, depressive episode, education level, and physical health.