Background: Approximately 670,000 people in the UK have dementia. Previous literature suggests that physical exercise could slow dementia symptom progression.
Objectives: To estimate the clinical effectiveness and cost-effectiveness of a bespoke exercise programme, in addition to usual care, on the cognitive impairment (primary outcome), function and health-related quality of life (HRQoL) of people with mild to moderate dementia (MMD) and carer burden and HRQoL.
Design: Intervention development, systematic review, multicentred, randomised controlled trial (RCT) with a parallel economic evaluation and qualitative study.
Setting: 15 English regions.
Participants: People with MMD living in the community.
Intervention: A 4-month moderate- to high-intensity, structured exercise programme designed specifically for people with MMD, with support to continue unsupervised physical activity thereafter. Exercises were individually prescribed and progressed, and participants were supervised in groups. The comparator was usual practice.
Main outcome measures: The primary outcome was the Alzheimer's Disease Assessment Scale - Cognitive Subscale (ADAS-Cog). The secondary outcomes were function [as measured using the Bristol Activities of Daily Living Scale (BADLS)], generic HRQoL [as measured using the EuroQol-5 Dimensions, three-level version (EQ-5D-3L)], dementia-related QoL [as measured using the Quality of Life in Alzheimer's Disease (QoL-AD) scale], behavioural symptoms [as measured using the Neuropsychiatric Inventory (NPI)], falls and fractures, physical fitness (as measured using the 6-minute walk test) and muscle strength. Carer outcomes were HRQoL (Quality of Life in Alzheimer's Disease) (as measured using the EQ-5D-3L) and carer burden (as measured using the Zarit Burden Interview). The economic evaluation was expressed in terms of incremental cost per quality-adjusted life-year (QALY) gained from a NHS and Personal Social Services perspective. We measured health and social care use with the Client Services Receipt Inventory. Participants were followed up for 12 months.
Results: Between February 2013 and June 2015, 494 participants were randomised with an intentional unequal allocation ratio: 165 to usual care and 329 to the intervention. The mean age of participants was 77 years [standard deviation (SD) 7.9 years], 39% (193/494) were female and the mean baseline ADAS-Cog score was 21.5 (SD 9.0). Participants in the intervention arm achieved high compliance rates, with 65% (214/329) attending between 75% and 100% of sessions. Outcome data were obtained for 85% (418/494) of participants at 12 months, at which point a small, statistically significant negative treatment effect was found in the primary outcome, ADAS-Cog (patient reported), with a mean difference of -1.4 [95% confidence interval (CI) -2.62 to -0.17]. There were no treatment effects for any of the other secondary outcome measures for participants or carers: for the BADLS there was a mean difference of -0.6 (95% CI -2.05 to 0.78), for the EQ-5D-3L a mean difference of -0.002 (95% CI -0.04 to 0.04), for the QoL-AD scale a mean difference of 0.7 (95% CI -0.21 to 1.65) and for the NPI a mean difference of -2.1 (95% CI -4.83 to 0.65). Four serious adverse events were reported. The exercise intervention was dominated in health economic terms.
Limitations: In the absence of definitive guidance and rationale, we used a mixed exercise programme. Neither intervention providers nor participants could be masked to treatment allocation.
Conclusions: This is a large well-conducted RCT, with good compliance to exercise and research procedures. A structured exercise programme did not produce any clinically meaningful benefit in function or HRQoL in people with dementia or on carer burden.
Future work: Future work should concentrate on approaches other than exercise to influence cognitive impairment in dementia.
Trial registration: Current Controlled Trials ISRCTN32612072.
Funding: This project was funded by the National Institute for Health Research (NIHR) Health Technology Assessment programme and will be published in full programme and will be published in full in Health Technology Assessment Vol. 22, No. 28. See the NIHR Journals Library website for further project information. Additional funding was provided by the Oxford NIHR Biomedical Research Centre and the Oxford NIHR Collaboration for Leadership in Applied Health Research and Care.