Skip navigation

No evidence for music therapy for dementia

"..the published literature demonstrates that music/music therapy is an effective method overall for symptoms of dementia". This is the conclusion of a meta-analysis [1]. But is it right?


The search was extensive and looked at published studies in peer-reviewed journals that empirically addressed the use of music therapy for dementia. Twenty-one studies were included with study numbers from 3 to 40 patients. Twelve studies had fewer than 15 subjects, and only one had more than 30.


Extensive and impenetrable statistical analysis concluded that music therapy was beneficial. What it was beneficial for was not stated specifically, but could have been as varied as agitation during bathing to sleeping.


There is no mention in the paper nor in the titles of the included papers, of the key words randomisation or blinding. Although we don't know it for a fact, it is highly likely that most of these studies were neither randomised nor blind.

In the absence of information on two of the most important sources of study bias, with more than half the studies with trivial size, with no stated outcome and no clear indication of the intervention, what can we conclude? Only that the conclusion of the meta-analysis is dangerous and probably wrong. At best we can conclude that there is no hard evidence that music therapy is of any value in dementia.


  1. SM Koger et al. Is music therapy an effective intervention for dementia? A meta-analytic review of the literature. Journal of Music Therapy 1999 XXXVI: 2-15.
previous or next story in this issue