Psychiatric disorders and risk of subsequent dementia: Systematic review and meta‐analysis of longitudinal studies

OBJECTIVES: Although psychiatric disorders have been found to be associated with increased risk of dementia, previous findings are mixed, and the nature of these relationships remains poorly understood. We examined longitudinal associations between depression, anxiety, post‐traumatic stress disorders (PTSD), bipolar disorder (BPD), psychotic disorders and subsequent dementia. METHODS: We searched three databases for longitudinal, population‐based studies investigating associations between psychiatric disorders and dementia (PROSPERO registration: CRD42020209638). We conducted narrative synthesis, and random‐effects meta‐analyses to obtain pooled estimates. We used meta‐regression and stratified analyses to examine variation by sex, age‐at‐onset and follow‐up time. RESULTS: Fifty‐seven citations met eligibility criteria. Most studies focussed on depression (n = 33), which was associated with subsequent all‐cause dementia (pooled relative risk [RR]: 1.96, 95% confidence interval [CI]: 1.59–2.43; I (2) = 96.5%), Alzheimer's Disease (pooled RR: 1.9, 95% CI: 1.52–2.38; I (2) = 85.5%), and Vascular Dementia (pooled RR: 2.71, 95% CI: 2.48–2.97; I (2) = 0). Associations were stronger in studies with shorter follow‐up periods and for severe and late‐onset depression. Findings regarding anxiety were mixed, and we did not find evidence of an overall association (pooled RR: 1.18, 95% CI: 0.96–1.45; I (2) = 52.2%, n = 5). Despite sparse evidence, psychotic disorders (pooled RR: 2.19, 95% CI: 1.44–3.31; I (2) = 99%), PTSD and BPD were associated with subsequent dementia. CONCLUSIONS: People with psychiatric disorders represent high‐risk groups for dementia, highlighting the importance of ongoing symptom monitoring in these groups. Findings regarding temporality and age‐at‐onset indicate that depression symptoms could reflect prodromal dementia for some individuals. Further longitudinal research is required to determine whether psychiatric disorders represent causal risk factors or early markers of dementia neuropathology.