4,068 posters, 219 videos, 101 topics, 69 sessions, 14,365 authors, 6,128 institutions

Alzheimer's Association International Conference 2026

July 12-15, 2026 | London, United Kingdom, and Online

13539

Investigating Temporal Relationships Between Psychiatric Disorders And Subsequent Dementia: A Swedish Population-based Cohort Study

Preview of Investigating Temporal Relationships Between Psychiatric Disorders And Subsequent Dementia: A Swedish Population-based Cohort Study

View Poster Contact Presenter

Poster Presenter

Jean Stafford

Authors

Jean Stafford, Praveetha Patalay, James B. Kirkbride, Marcus Richards, Robert Howard

Affiliations

UCL, University of Edinburgh

Part of Topic

Public Health -> Epidemiology -> Risk, Reserve and Resilience

Temporal relationships between psychiatric disorders and dementia

Key takeaways: Psychiatric disorders were associated with higher dementia rates across diagnostic groups. Associations were strongest for severe mental illness and later-life psychiatric diagnoses. Patterns suggest both possible prodromal dementia and longer-term vulnerability.

INTRODUCTION

•Mental health conditions associated with later dementia.

•Evidence beyond depression in limited, and causality and underlying mechanisms remain unclear.

AIM

•Investigate associations with dementia by age and time since psychiatric diagnosis.

•Examine variation by sex and baseline comorbidities.

STUDY DESIGN

•Born 1932–1985.

•Psychiatric disorder cases matched by birth year to comparisons without psychiatric disorder on same date.

•Followed until all-cause dementia, death, emigration or 31st Dec 2021.

•Stratified Cox models; Fine and Gray models - competing mortality.

COHORT & EXPOSURES
Total cohort n=3,345,970; psychiatric disorder n=816,065; matched comparison n=2,529,905

Psychiatric conditions examined

•Depression, anxiety and PTSD

•Psychotic disorders and bipolar disorder

•Alcohol and substance use disorder

Most common psychiatric diagnoses

•Anxiety, somatoform and dissociative disorders: 29.8%

•Alcohol and substance use disorder: 29.5%

•Unipolar depression: 28.9%

Median follow-up: 13.2 years (IQR: 6.8–20.1) Median age-at-entry: 39.6 years

RESULTS

Main finding 1: psychiatric disorders associated with higher dementia rates

Overall dementia rates were elevated across all psychiatric disorder groups. Strongest overall associations were found for severe mental illness, particularly psychotic disorders.

Main finding 2: associations varied by age at diagnosis

Associations were present across age-at-diagnosis groups and were generally strongest for later-life psychiatric diagnoses, particularly for severe mental illness.

Subgroup and sensitivity analyses

•Associations found in men and women; slightly stronger in men.

•Stronger associations with baseline cardiovascular disease and diabetes.

•Stronger associations with vascular dementia diagnosis.

•Associations remained in competing risk models.

TEMPORALITY & INTERPRETATION

Short-term pattern

Associations strongest close to diagnosis for mid- and late-life groups — may indicate prodromal dementia, diagnostic overlap or increased clinical contact.

Long-term pattern

Elevated rates persisted over longer follow-up, suggesting reverse causality alone is unlikely to fully explain the associations.

MEANING OF FINDINGS

Psychiatric disorders may be both early markers of neurodegenerative disease and indicators of longer-term vulnerability to dementia.

•Findings highlight brain health inequalities across psychiatric diagnoses.

•New psychiatric symptoms in later life, particularly severe mental illness, may warrant cognitive and functional monitoring.

•Further research should examine possible cardiometabolic, inflammatory, stress-related, social and genetic pathways.

STRENGTHS, LIMITATIONS & CONTACT

Strengths

•Population-based cohort

•Temporal analyses and long follow-up

•Sensitivity analyses and competing risks models

Limitations

•Register-based ascertainment

•Changes in recording over time

•Possible residual confounding

•Potential detection bias

Comments

Email is required but is not shown publicly.

Sort comments

Loading comments...