Physical activity interventions for hospitalised people living with dementia: systematic review and meta-analysis.

Abstract

Purpose: People living with dementia are vulnerable to adverse outcomes of hospitalisation, such as functional decline and deconditioning. The aim of this review was to examine the effectiveness of hospital-based interventions involving physical activity on deconditioning outcomes in people living with dementia.

Methods: Five databases (MEDLINE, CENTRAL, Embase, PsycINFO, and CINAHL) were searched to January 2025 to identify eligible studies. Two researchers independently carried out screening, data extraction and quality assessment.

Inclusion criteria: adults with a diagnosis of dementia, acute hospital admission, and interventions involving physical activity during hospitalisation. Data were pooled for strength, balance, functional outcomes, readmission, falls, length of stay, and delirium. We used GRADE to assess the certainty of evidence.

Results: 2179 records were identified and six studies (n = 576 with dementia) included: one randomised-controlled trial, five non-randomised or quasi-randomised. Studies were conducted in an "acute care for elders unit" (n = 2), acute psychogeriatric ward (n = 2), general medical unit (n = 1), and geriatric rehabilitation ward (n = 1) across six countries. In meta-analyses, there was less decline in basic activities of daily living at 3 months in the physical activity group (change from pre-admission status); two studies, mean difference 1.27 (95% CI 0.36-2.18) but there were no differences found in all other outcomes at discharge or 3 months, compared to usual care. Certainty of the evidence is low/very low.

Conclusion: Due to the limited available evidence, there is uncertainty about the effect of physical activity interventions above usual care. More high-quality research is needed to improve outcomes for people living with dementia in hospital.