Aims
Observational studies suggest delirium increases risk of incident dementia. However, studies have primarily been conducted in small samples of surgical patients with limited adjustment for potential confounders. We aimed to address these limitations using largescale hospital data.
Methods
Data from 650,773 patients aged >75 years were considered (>4.7M episodes; 2009–2014). Diagnoses of dementia and delirium were extracted from ICD-10 codes. Patients with dementia at or prior to the index episode were excluded. Delirium/no delirium pairs were identified by matching patient age, sex, frailty score, episode principal diagnosis, length of stay and intensive care length of stay, and were followed for 5 years. A Fine-Gray hazard model was used to estimate the association of delirium with incident dementia while accounting for the competing risk of death.
Results
Data from 33,670 matched pairs were analysed (48% male; mean (SD) age = 83.9 (6.4) years). Collectively, 76% (N=51,101) of patients died and 20% (N=13,665) developed dementia. Patients with delirium had 35% higher risk of death (HR = 1.35 [95% CI = 1.32–1.39]) and more than double the risk of incident dementia (sHR = 2.29 [95% CI = 2.21–2.38]) than patients without delirium.
Conclusions
Delirium was a strong risk factor for incident dementia in this large cohort of older patients. Accounting for key demographic and clinical factors and the competing risk of death strengthens the evidence for delirium’s independent effect on dementia pathophysiology. The clinical implications of delirium as a potentially modifiable risk factor for dementia are substantial.