Music therapy for substance misuse: sixteen people, one significant EEG difference

Sixteen people from a community substance-use treatment service took part in this pilot. Some had six weekly music therapy sessions added to their standard treatment; the rest had standard treatment alone. The team, at Anglia Ruskin University’s Cambridge Institute for Music Therapy Research, collected resting-state and event-related EEG, craving and mood questionnaires, and semi-structured interviews. Their stated aim was to find out whether all of those measures can be collected together in a real clinical service, and the answer is that they can.

Read result by result, the study found one significant group difference and several non-significant ones.

Craving Thermometer ratings dropped immediately after music therapy sessions, and the authors note that the perceived improvement “may emerge primarily in the immediate post-session window”. The Brief Substance Craving Scale, which is the standardised instrument and the one a clinician would act on, showed no significant group difference after baseline adjustment. Resting-state beta power, often described as elevated during craving, also showed no significant group difference, though the reduction in the music therapy group was descriptively larger. Frontal alpha asymmetry did differ significantly between groups after baseline adjustment, and the authors put it forward as a candidate domain for future work rather than a result. P3d amplitude on a Go/NoGo task, the inhibition measure, rose non-significantly.

The interviews add what the numbers cannot. Participants described craving reduction, emotional reflection and a sense of therapeutic connection. They also described a reluctance to discuss craving, which tells you something about asking people in treatment to rate their cravings on a scale.

We grade it mixed. The rubric score of 2.9 puts it in weak territory on method alone, and the partial-findings override is the more informative label: one outcome moved, the primary craving measure did not, and the authors are explicit that larger trials with better allocation, clearer comparators and longer follow-up are needed.

Limitations. Sixteen participants split across two groups cannot detect a modest effect, and the authors do not claim otherwise. Allocation was not robust, by their own description, so the groups may have differed before anything started. The music therapy group also got six extra hours of human attention, which standard treatment alone does not provide, so the comparison is not of music against nothing-else-changed. There is no follow-up after the six weeks. We could not obtain the full text; Elsevier refused the automated request despite the paper being open access, so this grade rests on a detailed abstract and the request is logged in docs/reports/fulltext-requests.md.

This paper is the subject of our news check on the coverage that reported its non-significant beta result as a marked reduction.