Higher circulating cortisol is associated with amyloidogenesis in post‐menopausal women: The Framingham Heart Study

Abstract

AbstractBackgroundReports of a correlation between higher baseline cortisol and faster progression of Alzheimer’s disease (AD) in longitudinal cohorts and more advanced AD pathology in cross sectional studies raise the possibility that cortisol dysregulation contributes to AD pathogenesis. The current data, however, is less strong with regards to possible mechanisms underlying this relationship or the direction of influence between higher cortisol and AD severity. The goal of this study was to examine the associations between cortisol levels and incident cerebral amyloid and tau PET deposition.MethodsWe present a prospective cohort study of cognitively normal individuals derived from Framingham Heart Study (generation 3‐mean age 39.6±8.1 years, 48.5% women) who had measures of fasting morning plasma cortisol at baseline and PET measures of brain amyloid and tau 15±1.6 years later. Fasting cortisol levels were examined as tertitles (Cortisol level (min,max)‐ T1:3.92‐11.12; T2:11.13‐15.50; T3:15.76‐50.07). Linear regression was used to evaluate associations baseline plasma cortisol (in tertiles) with amyloid and tau deposition, adjusted for age, sex, and the duration between cortisol and PET imaging assessments. We assessed interactions by ApoE4 positivity, sex and menopausal status (women only).ResultsWe found no association between serum cortisol and amyloid/tau burden, and no significant interaction with age or ApoE4 status. However, the association between cortisol and amyloid retention interacted significantly with sex in our primary regions of interest (posterior cingulate cortex [PCC] β:0.059±0.027, p = 0.031; precuneus [PC] 0.071±0.025, 0.005, and whole brain [represented by FLR] 0.039±0.014, 0.006). Comparing the highest relative to the middle cortisol (the referent) tertiles among women, amyloid deposition was higher in PCC (β±SE: 0.059±0.027, p = 0.031), PC (0.071±0.025, 0.005), and FLR ROIs (0.039±0.014, p = 0.006). The same relationships did not hold in men or in relations to tau. Furthermore, the effect was only present in postmenopausal women, with the higher amyloid deposition in PC (β:0.103±0.040, p = 0.012) and FLR ROIs (0.071±0.020, 0.001) as well as superior temporal sulcus (secondary outcome). We did not observe any statistically significant association comparing bottom to the middle cortisol tertile.ConclusionAlthough future research is needed, the results suggest that serum cortisol may contribute to Alzheimer’s pathogenesis by driving amyloidogenesis in postmenopausal women.