Association of Music Interventions With Health-Related Quality of Life

Abstract

Increasing evidence supports the ability of music to broadly promote well-being and health-related quality of life (HRQOL). However, the magnitude of music's positive association with HRQOL is still unclear, particularly relative to established interventions, limiting inclusion of music interventions in health policy and care. To synthesize results of studies investigating outcomes of music interventions in terms of HRQOL, as assessed by the 36- and 12-Item Health Survey Short Forms (SF-36 and SF-12). MEDLINE, Embase, Web of Science, PsycINFO, ClinicalTrials.gov, and International Clinical Trials Registry Platform (searched July 30, 2021, with no restrictions). Inclusion criteria were randomized and single-group studies of music interventions reporting SF-36 data at time points before and after the intervention. Observational studies were excluded. Studies were reviewed independently by 2 authors. Data were independently extracted and appraised using GRADE criteria (Grading of Recommendations, Assessment, Development, and Evaluations) by multiple authors. Inverse-variance random-effects meta-analyses quantified changes in SF-36 mental and physical component summary (respectively, MCS and PCS) scores from preintervention to postintervention and vs common control groups. SF-36 or SF-12 MCS and PCS scores, defined a priori. Analyses included 779 participants from 26 studies (mean [SD] age, 60 [11] years). Music interventions (music listening, 10 studies; music therapy, 7 studies; singing, 8 studies; gospel music, 1 study) were associated with significant improvements in MCS scores (total mean difference, 2.95 points; 95% CI, 1.39-4.51 points; P < .001) and PCS scores (total mean difference, 1.09 points; 95% CI, 0.15-2.03 points; P = .02). In subgroup analysis (8 studies), the addition of music to standard treatment for a range of conditions was associated with significant improvements in MCS scores vs standard treatment alone (mean difference, 3.72 points; 95% CI, 0.40-7.05 points; P = .03). Effect sizes did not vary between music intervention types or doses; no evidence of small study or publication biases was present in any analysis. Mean difference in MCS scores met SF-36 minimum important difference thresholds (mean difference 3 or greater). In this systematic review and meta-analysis, music interventions were associated with clinically meaningful improvements in HRQOL; however, substantial individual variation in intervention outcomes precluded conclusions regarding optimal music interventions and doses for distinct clinical and public health scenarios.