Of all the claims made for music, the ones about depression are the most likely to be oversold. They are also, it turns out, among the better evidenced — provided you are precise about what was tested, by whom, and for how long. The difference between “music helps depression” and what the research actually shows is large enough to be worth spelling out.

Music therapy is not the same as listening to music

The first distinction matters more than anything else in this article. The trials discussed below did not test people putting on a sad album at home. They tested music therapy: a structured clinical process delivered by a trained and accredited music therapist, typically involving improvisation, playing instruments together, songwriting or guided listening, within a therapeutic relationship and alongside the person’s usual treatment.

That is a different intervention from private listening, with different training requirements and different aims. Conflating the two is the most common error in popular coverage of this research, and it makes claims sound far broader than the evidence supports.

What the Cochrane review found

The reference point is the Cochrane review by Aalbers and colleagues, published in 2017. It included nine studies with 421 participants — a small evidence base by the standards of this kind of review, and the authors say so.

When music therapy was added to treatment as usual and compared with treatment as usual alone, the review reported:

A reduction in clinician-rated depressive symptoms, standardised mean difference −0.98 (95% CI −1.69 to −0.27), based on four studies with 219 participants, at moderate-quality evidence. A similar reduction in patient-reported depressive symptoms, SMD −0.85 (95% CI −1.37 to −0.34), four studies, 142 participants, also moderate quality. Improvements in anxiety and in functioning compared with treatment as usual alone. No significant difference in quality of life (SMD 0.32, confidence interval crossing zero, 67 participants, low-quality evidence). No significant difference in the number of people leaving the study early, which matters because dropout is a rough proxy for how acceptable an intervention is.

The review also found no evidence that music therapy was associated with more adverse events than usual treatment alone.

Three qualifications that change the picture

The first is duration. The authors describe the benefits as short-term. The included trials measured outcomes at the end of treatment or shortly after; none established that gains persist over months. Anyone citing this review as evidence of lasting change is going further than the data.

The second is scale. Four studies and a couple of hundred participants is a thin base on which to rest a confidence interval, and the intervals here are wide. A standardised mean difference of −0.98 sounds large, but the range runs from −1.69 to −0.27 — that is, from a substantial effect to a small one.

The third is the comparison being made. Music therapy was added to usual treatment and compared against usual treatment alone. When the review looked at music therapy against psychological therapy, the results were uncertain, and it found no clear answer on whether one form of music therapy outperforms another. So the finding is about addition, not substitution.

A note on what “added to usual treatment” means

That design detail is easy to skim past and is in fact the heart of the matter. Everyone in these trials was receiving their ordinary care. The question was whether adding structured music therapy sessions on top of it produced measurable change compared with ordinary care alone.

It is a useful question precisely because it mirrors how music therapy is used in practice — as one element within a treatment plan, arranged by clinicians, not as an alternative to anything. Nothing in the review suggests it should replace established treatments, and the authors make no such suggestion.

Where the field is going

The authors’ own recommendations are instructive. They call for larger, well-designed trials, particularly in children and adolescents, where the evidence is thinnest. They also ask for clearer reporting: what exactly the intervention involved, what the comparison group received, and who delivered the sessions. That request is a quiet acknowledgement that “music therapy” in the published literature has covered a wide range of practices, which makes pooling results across studies difficult and the pooled numbers harder to interpret.

Some of this has moved on since 2017, and anyone with a professional interest should look at the trials published since. But the shape of the conclusion has not changed much: a moderate-quality signal, from a small number of studies, for short-term improvement in depressive symptoms when music therapy is added to usual care.

That is a real finding, and it is also a modest one. It deserves neither the dismissal it sometimes gets nor the enthusiasm that turns it into a claim about playlists. Anyone dealing with persistent low mood has questions that belong with a doctor or a qualified mental health professional, who can say what is appropriate in their situation — including, where it is available, a referral to an accredited music therapist.

For related reading on everyday listening and mood, see our piece on why sad music appeals.


Sources

  • Cochrane Database of Systematic Reviews — Aalbers, Fusar-Poli, Freeman, Spreen, Ket, Vink, Maratos, Crawford, Chen & Gold, Music therapy for depression (2017), Issue 11, Art. No. CD004517. 9 studies, 421 participants. DOI: 10.1002/14651858.CD004517.pub3.
  • Cochrane plain-language summary — Music therapy for depression.

This article summarises published research and is not medical advice. Decisions about treatment belong with a qualified healthcare professional.

Created with the help of artificial intelligence.