Incidence and Prediction of Falls in Dementia: A Prospective Study in Older People

BACKGROUND Falls are a major cause of morbidity and mortality in dementia, but there have been no prospective studies of risk factors for falling specific to this patient population, and no successful falls intervention/prevention trials. This prospective study aimed to identify modifiable risk factors for falling in older people with mild to moderate dementia. METHODS AND FINDINGS 179 participants aged over 65 years were recruited from outpatient clinics in the UK (38 Alzheimer's disease (AD), 32 Vascular dementia (VAD), 30 Dementia with Lewy bodies (DLB), 40 Parkinson's disease with dementia (PDD), 39 healthy controls). A multifactorial assessment of baseline risk factors was performed and fall diaries were completed prospectively for 12 months. Dementia participants experienced nearly 8 times more incident falls (9118/1000 person-years) than controls (1023/1000 person-years; incidence density ratio: 7.58, 3.11-18.5). In dementia, significant univariate predictors of sustaining at least one fall included diagnosis of Lewy body disorder (proportional hazard ratio (HR) adjusted for age and sex: 3.33, 2.11-5.26), and history of falls in the preceding 12 months (HR: 2.52, 1.52-4.17). In multivariate analyses, significant potentially modifiable predictors were symptomatic orthostatic hypotension (HR: 2.13, 1.19-3.80), autonomic symptom score (HR per point 0-36: 1.055, 1.012-1.099), and Cornell depression score (HR per point 0-40: 1.053, 1.01-1.099). Higher levels of physical activity were protective (HR per point 0-9: 0.827, 0.716-0.956). CONCLUSIONS The management of symptomatic orthostatic hypotension, autonomic symptoms and depression, and the encouragement of physical activity may provide the core elements for the most fruitful strategy to reduce falls in people with dementia. Randomised controlled trials to assess such a strategy are a priority.