Effect of psychosocial interventions for people with dementia: an overview of systematic reviews
Systematic review
|Published
The objective of this review of systematic reviews was to assess the effects of psycho-social interventions for people with dementia.
Key message
Shortly after publication, we were made aware of Woods B, Rai HK, Elliott E, Aguirre E, Orrell M, Spector A. Cognitive stimulation to improve cognitive functioning in people with dementia. Cochrane Database of Systematic Reviews 2023. We have therefore chosen to make a separate assessment of this, see at the bottom of the page.
101 000 people in Norway were living with dementia in 2020, and the incidence increases with age. Our overview of systematic reviews identified 20 systematic reviews of high or moderate quality investigating the effect of 15 psychosocial interventions for people with dementia. We are uncertain of the effect of many outcomes, either because of missing data or very low certainty in the effect estimates.
Results summarized per intervention:
Cognitive therapy probably reduces symptoms of depression and improves activities of daily life and quality of life but may cause little or no change in cognition. Cognitive training probably reduces symptoms of depression and may improve cognition. Virtual cognitive interventions probably reduce symptoms of depression and improve cognition. Reminiscence therapy causes little or no change in cognition or symptoms of depression and probably little or no change in quality of life. Occupational therapy may cause little or no change in cognition and BPSD. Music therapy probably reduces symptoms of depression and BPSD, may improve quality of life and cause little or no change in cognition. Dance therapy may improve cognition and reduce BPSD. Animal assisted therapy probably reduces symptoms of depression and BPSD. Therapy using robots may reduce symptoms of depression but may cause little or no change in cognition. Multi component interventions probably cause little or no change in quality of life but may improve activities of daily life and may cause little or no change in cognition. Personally tailored activities may reduce BPSD and burden of care for next of kin but cause little or no difference in symptoms of depression and quality of life.
We do not know, or are very uncertain about, effects of cognitive stimulation, cognitive rehabilitation, de-escalation therapy and sensory stimulation for people with dementia.
Summary
Introduction
There were 101 000 people in Norway living with dementia in 2020. Approximately 15% of the Norwegian population above 70 years have dementia, and the incidence increases with age. Common symptoms of dementia are impaired memory, difficulty concentrating and communicating, reduced ability to orientate, and reduced function of daily living. Dementia may also lead to behavioural and personality changes as well as psychological symptoms. This review was made on commission from the Norwegian Directorate of Health.
Objective
The objective of this review of systematic reviews was to assess the effects of psychosocial interventions for people with dementia.
Methods
We conducted systematic literature searches in the Cochrane Database of Systematic Reviews, Epistemonikos, PsycInfo, Medline and Embase in June 2022. Two researchers independently assessed the relevance of references according to the inclusion criteria, first on title and abstract, and then in full text. We included systematic reviews on the effect of psychosocial interventions compared with no treatment, usual care or other treatment for people with dementia. Reviews that did not fulfil criteria for a systematic review (clear inclusion criteria, search in databases, and risk of bias assessment) were excluded. Two researchers independently assessed the quality of the most updated systematic reviews (with the newest literature searches) using the AMSTAR 2 checklist. One researcher extracted data on seven main outcomes: general cognition, depression, behavioural and psychological symptoms of dementia (BPSD), activities of daily life, quality of life, need for full time care, and burden of care for next of kin. Another researcher checked the data extraction. Confidence in the effect estimates were assessed using the GRADE approach. Where the systematic review authors had made their own GRADE assessments, we used these. For non-graded reviews, we made the judgements.
Results
We included 20 systematic reviews of high or moderate quality that collectively assessed the effects of 15 psychosocial interventions for people with dementia. In this executive summary we present the effect estimates with 95 % confidence interval (CI) for interventions and outcomes we have high or moderate confidence in, and only change and direction for results of low confidence.
Cognitive therapy probably causes a small reduction in symptoms of depression (SMD 0.23 SD lower, CI 0.37 lower to 0.10 lower), a small improvement in function of daily activities (SMD 0.25 SD lower, CI 0.40 lower to 0.09 lower) a small improvement in quality of life (SMD 0.31 SD higher, CI 0.13 higher to 0.50 higher) (moderate confidence), but small or no change in general cognition (low confidence).
Cognitive training probably causes a small to medium reduction in symptoms of depression (SMD 0.48 SD lower, CI 0.71 lower to 0.24 lower) (moderate confidence) and may improve cognition (low confidence).
Virtual cognitive interventions probably lead to small improvements in general cognition (SMD 0.36 SD higher, CI 0.16 higher to 0.57 higher) and small to medium reductions in symptoms of depression (SMD 0.45 SD lower, CI 0.79 lower to 0.12 lower) (moderate confidence).
Reminiscence therapy causes little or no change in cognition (SMD 0.11 SD higher, CI 0 to 0.23 higher) or symptoms of depression (SMD 0.03 SD lower, CI 0.15 lower to 0.1 higher) (high confidence), and probably causes little or no change in quality of life (SMD 0.11 SD higher, CI 0.12 lower to 0.33 higher) (moderate confidence). The effect of reminiscence therapy in activities of daily living is uncertain (low confidence).
Occupational therapy may cause little or no change in cognition and BPSD (low confidence).
Music therapy, measured at the end of the intervention, causes a small reduction in symptoms of depression (SMD 0.27 SD lower, CI 0.45 lower to 0.09 lower) and BPSD (SMD 0.23 SD lower, CI 0.46 lower to 0.01 lower) (moderate confidence), and may cause a small improvement in quality of life and little or no change in cognition (low confidence). Four weeks or more after the end of music therapy, there may be little or no change in general cognition, symptoms of depression and BPSD or quality of life (low confidence).
Dance therapy may improve general cognition and reduce BPSD (low confidence).
Animal assisted therapy probably leads to a small to moderate reduction in symptoms of depression (SMD 0.45 SD lower, CI 0.63 lower to 0.27 lower) and BPSD (SMD 0.43 SD lower, CI 0.62 lower to 0.23 lower) (moderate confidence).
Therapy using robots (human like and animal like) may lead to a small reduction in symptoms of depression but may cause little or no change in cognition (low confidence).
Multi component interventions probably cause little or no change in quality of life (authors did not present effect estimates, but reported moderate confidence in the results), but may improve activities of daily life and may cause little or no change in cognition (low confidence).
Personally tailored activities may lead to a small to moderate reduction in BPSD and reduced burden of care for next of kin, but little or no change in symptoms of depression and quality of life (low confidence).
All the main outcomes were not reported for all the therapies, and for several of the main outcomes that were reported, the confidence in the effect estimate was very low. We do not know, or are very uncertain about, effects of cognitive stimulation, cognitive rehabilitation, de-escalation therapy and sensory stimulation for people with dementia.
Effects were generally assessed immediately or shortly after the therapy ended, therefore there are very little information available regarding long term effects of these psychosocial therapies.
Discussion
People with all degrees and many types of dementia were included in the systematic reviews. The available knowledge thus includes a wide population, but we have little direct information on effects for specific subgroups. Furthermore, the effect of some of the therapies are uncertain because of missing data or because available data is highly uncertain (low certainty of evidence). One important, and recurrent reason for downgrading our confidence in the effect estimates was wide confidence intervals, so wide as to include both no effect, small effect or moderate effect. There is a need for new randomized trials with larger numbers of participants to achieve more precise effect estimates and sound conclusions. Studies with a more homogenous population may provide more precise effect estimates, but may be less relevant for clinical practice, because people with dementia is a highly heterogeneous group with multiple and overlapping challenges. There were few reviews on the burden of next of kin, and none of the included reviews reported the need for full time care. New studies should include these outcomes and should have a sufficiently long follow up to inform on long term effects of psychosocial therapies. Within the Norwegian health care system, we expect that psychosocial therapies will be tailored to the level of daily activities and preferences of the person with dementia, and that different therapies may be combined. However, we know little about the effect of personally tailored therapies and only a few combinations of therapies were identified in our review. There is thus a need for more research about these types of psychosocial therapies.
Conclusion
The following psychosocial therapies are probably beneficial for people with dementia: cognitive therapy, cognitive training, virtual cognitive therapy, reminiscence therapy, music therapy, and animal assisted therapy. Additionally, dance therapy, therapy with robots, multi component interventions, and personally tailored activities may be beneficial, but occupational therapy may have little or no effect. We do not know, or are very uncertain about, effects of cognitive stimulation, cognitive rehabilitation, de-escalation therapy, and sensory stimulation for people with dementia.
Effects were assessed for general cognition, depression, BPSD, activity of daily living, quality of life, or burden on next of kin. We have not found information about all the main outcomes for any of the psychosocial therapies. Nor do we know if these psychosocial therapies affect the need for full time care for people with dementia. Outcomes were generally measured immediately, or shortly after end of therapy, and, hence, there is a lack of information on long term effects of these psychosocial therapies.
