Sleep disturbance and adverse outcomes in dementia: A retrospective cohort study using electronic health records data.

Objectives: Dementia and sleep disturbance co-occur frequently. Previous research has focused on sleep disturbance and dementia risk, but its impact in established dementia is less clear. The aim of this study was to investigate the frequency of recorded sleep disturbance at the time of dementia diagnosis and its association with adverse outcomes.

Design & setting: Using data from a large South London healthcare provider, a retrospective cohort study was assembled.

Participants: Patients aged 55 years and older with a dementia diagnosis received between 2007 and 2018 (n = 17,235).

Measurements: Recorded sleep disturbance at the time of dementia diagnosis was ascertained using natural language processing algorithms. Multivariate Cox regression models were applied to identify associations with the following adverse outcomes: unplanned hospitalisation, emergency department attendance, mental health crisis care, mortality.

Results: The prevalence of recorded sleep disturbance at the time of dementia diagnosis was 29.0 %. After adjustment for twenty-two different potential confounders sleep disturbance was associated with significantly increased risk of all adverse outcomes (unplanned hospitalisation: adjusted hazard ratio (aHR): 1.06, 95 % confidence interval (CI): 1.01-1.10; emergency department attendance: aHR: 1.09, 95 % CI: 1.05-1.14; mental health crisis care: aHR: 1.58, CI: 1.36-1.84) apart from mortality.

Conclusion: Sleep disturbance at the time of dementia diagnosis is an indicator for a worse prognosis. Preventing and treating sleep disturbance in people with dementia could pose an opportunity to reduce adverse outcomes.