Adaptive Music
Can Music Help People With Dementia?
By the Starchild Music team · November 22, 2026 · 6 min read

The short answer
Music can help some people with dementia, but the effects are modest. A 2025 Cochrane review of 30 trials found that a course of music-based therapy probably improves depressive symptoms slightly and may reduce overall behavioral problems by the end of treatment, with no evidence of long-term effects. Smaller studies point to music a person already likes, and to active involvement, as worth trying.
Music can help some people with dementia, mostly by easing low mood and some behavior problems for a while. The effects are modest and are strongest during and shortly after a course of sessions. It is not a cure and it does not slow the disease.
This is general information, not medical advice. Anyone caring for a person with dementia should involve the person's doctor or care team.
What the best summary of the evidence says
The Cochrane Collaboration, which reviews health evidence, updated its review of music-based therapeutic interventions for people with dementia in 2025. It included 30 randomized trials with 1,720 participants in 15 countries. Most participants lived in institutions. Most interventions mixed active parts, such as singing or playing, with receptive parts, such as listening. Every trial included at least five sessions.
Compared with usual care, at the end of treatment:
- Depressive symptoms: music-based interventions "probably improved" them slightly, with moderate-certainty evidence.
- Overall behavioral problems: low-certainty evidence that they "may have improved" them.
- Agitation or aggression: moderate-certainty evidence of "likely" no improvement.
- Anxiety, emotional well-being, social behavior and cognition: low or very low certainty, with no clear improvement.
Compared with other activities, music may have been better for social behavior at the end of treatment, with low-certainty evidence. The reviewers found no evidence of effects four or more weeks after treatment ended. Adverse effects were measured inconsistently, but "no serious adverse events were reported." They also say the studies were at high risk of performance bias, which means people in the trials usually knew which group they were in.
Does it matter which music?
The Cochrane abstract does not say which music works best, so the next evidence is smaller and should be read with care.
A 2010 study by Sung, Chang and Lee gave 29 nursing home residents with dementia a 30-minute session of music chosen to fit their personal preferences, twice a week for six weeks, delivered by trained nursing staff. A comparison group of 23 received usual care. At week six, the music group had significantly lower anxiety scores. It was a quasi-experimental design, not a randomized trial, so it cannot prove cause.
A 2013 trial by Sakamoto, Ando and Tsutou randomly assigned 39 people with severe Alzheimer's disease to passive music, interactive music or no music, using individualized music. Both kinds of music caused short-term signs of relaxation. Interactive music brought the greatest improvement in emotional state and a greater long-term reduction in behavioral symptoms than passive music or none. It is a small trial, and the abstract does not spell out what the interactive sessions involved.
Together they suggest two ideas to try, not rules: choose music the person already knows and likes, and look for ways to take part, not just listen.
Why this connects to adaptive music
Adaptive music means music that changes to fit the listener. In dementia care, the listener changes day to day. A calm song might soothe in the morning and feel dull in the afternoon. The studies above are not about adaptive software. But their message is the same as the idea behind adaptive music: the right music depends on the person and the moment. For a related idea, see the iso principle, which starts with music that matches a mood before moving it.
Gentle ways to try it
These are practical ideas drawn from the studies, not medical instructions.
- Start with the person's own favorites. Ask family, or look at what they listened to when they were young.
- Keep sessions regular and short. The trials in the Cochrane review used at least five sessions, and the nursing home study used 30 minutes twice a week.
- Invite participation, such as humming, singing along, clapping or tapping, if the person enjoys it.
- Watch the reaction. If the person seems upset, stop and try something else. The reviewers say adverse effects may be rare, but studies recorded them inconsistently.
- Keep volume modest, and reduce competing noise such as a television.
- Tell the care team what you try and what you see, so it can be part of the care plan.
What we do not know
- Whether any gains last. The Cochrane review found no evidence of long-term effects.
- Which kind of music works best. The review's abstract does not rank them.
- How much of the benefit comes from the music and how much from company and attention. Comparing music with usual care cannot separate the music from the extra company and attention that come with a session.
Sources
- Europe PMC: Music-based therapeutic interventions for people with dementia (Cochrane Database of Systematic Reviews, 2025)
- Europe PMC: A preferred music listening intervention to reduce anxiety in older adults with dementia in nursing homes (Sung, Chang and Lee, 2010)
- Europe PMC: Comparing the effects of different individualized music interventions for elderly individuals with severe dementia (Sakamoto, Ando and Tsutou, 2013)
Find a song to license
Every song on Starchild can be heard in different styles and in your key before you license it.