The short version
- People with moderate or severe Alzheimer’s enjoyed old songs about as much as healthy people their age, even when the songs felt less familiar.
- As the disease advanced, the songs that felt most familiar were ones that had been popular when the listener was a younger child, around age 4 rather than 10.
- Songs with a softer beat and simpler harmony tended to feel more familiar and more personal, though the effect was small.
As Alzheimer’s disease advances, the songs a person still finds familiar come from earlier and earlier in their childhood, and their pleasure in old music does not fade. That is the main finding of a University of Helsinki study published on October 20 in Scientific Reports.
The design
The team, led by Ilja Salakka with Teppo Särkämö and Anni Pitkäniemi, asked 101 Finnish adults aged 60 and over to rate 70 short song excerpts. The songs were 10 traditional folk songs and 60 pop songs, 20 each from the 1950s, 1960s and 1970s. For every song, each person answered three questions on a 1-to-5 scale: how pleasant it felt, how familiar it was, and how strongly it was tied to their own life.
There were three groups. Forty were healthy, 29 had mild Alzheimer’s dementia, and 32 had moderate or severe dementia, split by a standard cognitive screening score. People did the task where they liked, often at home, on a computer or tablet. Some people with dementia had a relative or staff member help with the device, told to give only technical help.
The researchers also measured each song’s sound with software, scoring things like beat strength, brightness and how often the harmony changes. They then checked which features went with stronger memories.
This was not a treatment trial. Nobody received music therapy. It is a snapshot of how people at different stages respond to the same songs.
What they found
Pleasure held steady. Average pleasantness ratings were 3.79 for the healthy group, 3.71 for mild dementia and 3.76 for moderate or severe dementia, with no real difference.
Memory faded, but less than you might expect. Familiarity fell from 4.42 in the healthy group to 4.02 in the most affected group. The sense that a song was part of one’s own life fell from 2.66 to about 2.4 in both dementia groups. The authors call these gaps “rather small,” which fits earlier work showing musical memory holds up better than most other memory in Alzheimer’s.
The clearest result was about timing. For each song, the team worked out how old each listener was when it was popular. In the healthy group, familiarity peaked for songs from around age 10. In mild dementia the peak was age 8, and in moderate or severe dementia it was age 4. The authors describe this as the first experimental evidence that musical memory follows the same backward drift seen in other memories in Alzheimer’s.
The sound of the songs mattered a little. Songs with a weaker beat tended to feel more familiar, and songs with simpler, slower-moving harmony felt more personal. This matches the team’s 2021 study of 113 healthy older adults, where softer, less driving songs also went with stronger memories. But the musical features explained only about 1 percent of the variation in ratings.
The limits
The song list was short and Finnish, and the ratings were the listeners’ own impressions, not tested recall. The groups were small, and they differed in age and in the share of women, which the analysis adjusted for. Help from relatives at home may have nudged some answers, especially in severe dementia, as the authors note. The age-at-song calculation assumes people heard each song the year it was on the charts. And because each person was tested once, the study shows differences between groups, not change within one person over time.
Why it matters
Music is widely used in dementia care, and earlier research put the peak of music-evoked memories in older adults between about ages 10 and 18. This study suggests that for people further along, the songs that still feel familiar may be ones from much earlier, from when they were small children. The authors say the findings could help in “selecting the music material” for future programs. Our recent look at a singing-versus-conversation trial is a reminder that whether music helps is a separate question from which music to choose.
What to watch next: the larger Helsinki study this data came from is testing music interventions in dementia, and the team says the memories participants recorded for each song will be reported separately.
