Abstract
BACKGROUND: Psychological distress has been linked with later cognitive impairment and dementia, although the nature of the association remains unclear. Using a multi-cohort approach, we examined longitudinal associations of psychological distress with subsequent cognition and dementia, testing whether findings varied by age of assessment, severity, and persistence of psychological distress. METHODS: We used five longitudinal studies: Caerphilly Prospective Study, English Longitudinal Study of Ageing, National Child Development Study, National Survey of Health and Development, and Whitehall II. We examined associations with changes in cognition using linear and mixed effects models, and dementia using logistic regression. Results were pooled using two-stage individual participant data meta-analysis. FINDINGS: Pooled analyses (total N=24,564) showed that greater psychological distress was associated with lower subsequent cognitive level (β=-0.03 [95% CI: -0.06; -0.01]; I2=69.7%). Associations were present for clinically-significant (β=-0.06 [-0.12; -0.00]; I2=62.2%), persistent (β=-0.12 [-0.23; -0.02]; I2=82.5%) and intermittent distress (β=-0.09 [-0.12; -0.05]; I2=0%). Baseline distress was not associated with rates of subsequent cognitive decline. Psychological distress was associated with subsequent dementia (OR=1.12 [1.04; 1.20]; I2=0%), including for clinically-significant (OR=1.28 [1.07; 1.52; I2=0%]), persistent (OR=1.43 [1.04; 1.98]; I2=44.0%) and intermittent symptoms (OR=1.32 [1.02; 1.71]; I2=40.3%). Dementia was associated with psychological distress assessed in later life (age 65-75 OR: 1.29 [1.18; 1.40]; I2=13.9%), but not mid-life (age 45-54 OR: 1.09 [0.93; 1.28]; I2=34.0%). INTERPRETATION: In this multi-cohort study, psychological distress was associated with subsequent dementia and lower subsequent cognitive level, including for both persistent and intermittent distress. Associations with dementia were present when distress was assessed at older ages but not at ages 45-54 years, suggesting that associations might partly represent early preclinical markers of dementia neuropathology. Findings highlight the potential relevance of psychological distress in informing dementia prevention and supporting identification of high-risk groups, both of which are major global public health priorities.