Effects of combined multimodal exercise and cognitive training on cognitive, psychological, functional and health-related quality of life outcomes in older adults living with mild dementia

Abstract

Persons with dementia (PwD) experience cognitive decline and neuropsychiatric symptoms (NPS), resulting in impaired activities of daily living (ADLs) and reduced health-related quality of life (HRQoL). Given the limitations of drug therapies, non-pharmacological interventions remain the treatment of choice. Previous research has shown the benefits of exercise and cognitive interventions for PwD. However, whether there are synergistic effects for combined exercise and cognitive interventions in this population are unclear.

A systematic review and network meta-analysis was conducted to examine the effects of combined exercise and cognitive interventions for PwD, and to identify the most effective designs. The review included 29 randomized controlled trials (RCTs) involving persons with mild cognitive impairment (MCI) and dementia to broaden its scope. The results showed that combined interventions for MCI and dementia were more effective than exercise alone, but were comparable with cognitive interventions alone. Multimodal exercise combined with cognitive training was most effective in improving global cognition. This review further revealed limited studies have focused on persons with mild dementia, who still retain the ability to engage meaningfully in interventions. Based on these findings, a 12-week caregiver-assisted combined multimodal exercise and cognitive training (CA-MECT) intervention was developed. This intervention comprised online training implemented through videoconferencing and supplemented with home visits and caregiver-assisted home-based training. A pilot study demonstrated its feasibility, acceptability, and preliminary effects in older adults living with mild dementia.

The main study adopted a sequential mixed-methods design, including a single-blind, parallel-group RCT and a qualitative study. Seventy-two participants were recruited and randomly assigned to either the CA-MECT intervention (n = 36) or control group (n = 36). Validated instruments were used to evaluate global cognition, cognitive domains, NPS, instrumental ADLs (IADLs), and HRQoL at baseline (T0), immediately post-intervention (T1), and 3 months post-intervention (T2). Semi-structured qualitative interviews were conducted to explore participant-caregiver engagement experience and perceived benefits.

Overall, the adherence rate for the CA-MECT intervention was 78.71%. The dropout rates were 15.28% at T1 and 22.22% at T2. The generalized estimating equation analysis showed that, compared to the control group, the intervention group exhibited significantly greater improvements in global cognition and various subdomains of episodic memory, including immediate recall, short-term and long-term delayed recall, cued recall, and word recognition at T1. These benefits were maintained at T2. The intervention group also showed significantly greater improvements in short-term memory at T2. Additionally, the intervention group exhibited a significant reduction in the number and severity of NPS and caregiver distress related to NPS at T1. However, the effects on executive function, processing speed, attention, IADLs, and HRQoL were not significant. Qualitative findings showed that PwD experienced improved physical, cognitive, and social engagement, along with greater self-efficacy. Caregivers also reported positive changes in their attitudes and coping strategies regarding dementia. However, PwDs’ limited receptiveness to caregivers and mismatches between tasks and their interests and abilities hindered training.

The CA-MECT intervention effectively improved cognitive function and reduced NPS among older adults living with mild dementia. It can be integrated as a non-pharmacological treatment for managing dementia.