BackgroundLittle is known about the interrelationships among adverse childhood experiences (ACEs), post-traumatic stress disorder (PTSD), dissociative disorders, depression, and dementia risk. We sought to investigate associations of ACEs, PTSD, dissociative disorders, and depression with incident dementia and explore whether these associations may be interrelated through mediation.
MethodsThis prospective cohort study used population-based UK Biobank data, including 502 355 participants recruited at 22 assessment centres who completed questionnaires, an interview, and physical assessments at baseline (2006–2010). Data are linked to participants’ electronic health records from primary care, hospital admissions, and death registers through November 30, 2022, and to the results of the UK Biobank online mental health survey (2016–2017). Cox regression and g-formula-based mediation analyses were used to examine associations between self-reported ACEs, self-reported PTSD symptoms, diagnosed PTSD, dissociative disorders, depression, and dementia.
ResultsIn the final sample (n = 434 215, mean (SD) age 56.58 (8.07) years), ACEs (hazard ratio (HR)1point: 1.10; 95% CI 1.02–1.20), diagnosed PTSD (HR: 2.09; 95% CI 1.38–3.18), dissociative disorders (HR: 3.96; 95% CI 2.55–6.15), depression (HR: 2.17; 95% CI 2.05–2.30), and self-reported PTSD symptoms (HR1point: 1.09; 95% CI 1.06–1.11) were associated with increased dementia risk, after adjusting for sociodemographic characteristics. Self-reported PTSD symptoms explained 75.26% (P < .001) of the excess dementia risk associated with ACEs. Depression explained 4.51% (P = .02) of the ACEs-dementia link, 8.42% (P < .001) of the diagnosed PTSD-dementia link, and 10.29% (P < .001) of the dissociative disorders-dementia link.
ConclusionsIndividuals with ACEs, PTSD, dissociative disorders, or depression appear to be at increased risk of dementia, potentially through both shared and unique associations. However, these findings should be interpreted with caution due to potential limitations in statistical power.