Clinical effectiveness of music interventions for dementia and depression in older people (MIDDEL): a multinational, cluster-randomised controlled trial

BACKGROUND: Dementia and depression are among the leading causes of global disease burden. Effective and scalable interventions are needed to address the effect of these conditions, and music interventions are a promising non-pharmacological approach. The aim of this study was to determine the effectiveness of music interventions on depressive symptoms among care home residents with dementia in Australia, Germany, the Netherlands, Norway, Türkiye, and the UK. METHODS: Music Interventions for Dementia and Depression in Elderly care was a large, multinational, cluster-randomised controlled trial with a 2 × 2 factorial design to examine the effects of group music therapy, recreational choir singing, or both compared with standard care. The trial was done in 86 care home units across Australia, Germany, the Netherlands, Norway, Türkiye, and the UK. Care home units were required to host at least ten residents who met the inclusion criteria. Participants were required to be aged 65 years or older; a full-time resident in a participating care home unit; have dementia as indicated by a Clinical Dementia Rating score of 0·5-3 and a Mini-Mental State Examination score of 26 or less; have mild depressive symptoms as indicated by a Montgomery-Åsberg Depression Rating Scale (MADRS) score of at least 8; and a clinical diagnosis of dementia. Care home units with residents with dementia and depressive symptoms were randomly assigned (1:1:1:1; block randomisation stratified by site, using a computer-generated list) to group music therapy, recreational choir singing, a combination of these strategies, or standard care. The primary outcome was MADRS assessed at 6 months in the intention-to-treat population, which included all participants with available data. Assessors were masked but care staff, intervention providers, and residents were not masked due to the nature of the intervention. Intervention effects were analysed with ANCOVA for the total sample and per country. The trial was registered at ClinicalTrials.gov, NCT03496675, and is completed. FINDINGS: Between July 18, 2018, and Feb 1, 2023, 86 care home units with 1021 residents were enrolled and randomly assigned to one of the four groups. 22 care home units with 258 residents were randomly assigned to group music therapy, 22 care home units with 281 residents were allocated to recreational choir singing, 21 care home units with 244 residents were assigned to a combination of both group music therapy and recreational choir singing, and 21 care home units with 238 residents were assigned to standard care. The mean age of residents was 85·6 years (SD 7·4); most residents (747 73·2%) were female and 274 (26·8%) were male. Intention-to-treat analysis of 751 residents with data at 6 months showed no significant effect on MADRS scores of either recreational choir singing versus no recreational choir singing (β 0·4 95% CI -1·3 to 2·1; p=0·68), group music therapy versus no group music therapy (β 0·8 -1·0 to 2·6; p=0·37), or the interaction between recreational choir singing and group music therapy (β -0·6 -3·1 to 1·9, p=0·63; β represents mean difference estimated from ANCOVA). Effects varied between countries. No related adverse events were reported and acute medical hospital admission rates were similar across groups at 3 months and 6 months. INTERPRETATION: Internationally, active group music interventions as conducted in this study do not reduce depressive symptoms more than standard care in the long term. Country was the strongest predictor for differences in effects, underlining the importance of cultural and systemic differences. Intervention guidelines and health-care policies need to be carefully tailored to the specific contexts of care home populations and levels of care. Future multisite trials should focus on more narrowly defined target groups or contexts to reduce the risk of heterogeneity overshadowing potential effects of interventions. Although music interventions might be beneficial for people with dementia, there is a need to harmonise their implementation and investigate the mechanisms through which they work. FUNDING: EU Joint Programme - Neurodegenerative Disease Research, and National Health and Medical Research Council, Australia.