Music Engagement and Personal Recovery: Validation of the Healthy-Unhealthy Music Scale (HUMS) Among Marginalized Young Adults Diagnosed with Serious Mental Illness

Background and Purpose: Young adulthood is a critical developmental period when serious mental illnesses (SMI) often emerge, yet treatment engagement remains low. Personal recovery involves developing self-management strategies to manage symptoms and improve quality of life. Although music engagement is an accessible and commonly used self-management strategy, it can serve both adaptive (e.g., boosting mood) and maladaptive (e.g., rumination) functions. The Healthy-Unhealthy Music Scale (HUMS) is a novel measure that was developed to assess these functions, yet psychometric validation has been limited to non-clinical adolescent samples. This study evaluates the psychometric properties of the HUMS among marginalized YA-SMI.

Methods: Participants were 107 YA-SMI (schizophrenia-spectrum, bipolar, and major depressive disorders); 61.7% men, 27.1% women, and 11.2% non-binary or transgender; mean age 26.7 (SD=4.7), with 89.6% identifying as Black, Latinx, or Multiracial. The 13-item HUMS captures healthy and unhealthy music engagement on a 5-point scale across two hypothesized factors. Structural equation modeling (SEM) was used to conduct a confirmatory factor analysis (CFA) to compare the two-factor model (healthy, unhealthy) to an alternative three-factor model (healthy, rumination, avoidance). Analyses were conducted in Mplus using robust maximum likelihood estimation. Global fit was evaluated using Swain-corrected indices (RMSEA, CFI, p-value close fit) for smaller samples, with localized ill-fit examined via modification indices. Reliability was assessed using hierarchical omega. Validity analyses included bivariate correlations (construct validity) and multiple regression (predictive validity).

Results: The initial two-factor model demonstrated poor fit (χ²[df=64]=92.06, p=0.01; RMSEA=0.07; CFI=0.90; SRMR=0.08; p-close fit=0.15). Inclusion of one theoretically relevant correlated error improved fit (χ²[df=63]=74.37, p=0.15; RMSEA=0.04; CFI=0.96; SRMR=0.08; p-close fit=0.62). All items loaded significantly onto their respective factors (range: 0.44–0.79), and the factor correlation was non-significant. Although the three-factor model yielded good fit (χ²[df=61]=71.89, p=0.16; RMSEA=0.04; CFI=0.96; SRMR=0.07; p-close fit=0.62), the solution was inadmissible due to a factor correlation of r=1.10 between rumination and avoidance, suggesting statistical indistinguishability. The two-factor model was therefore most parsimonious. Subscales demonstrated adequate composite reliability (healthy: ω=0.78; unhealthy: ω=0.79). The unhealthy subscale was associated with greater emotion regulation difficulties (r=0.46, p<0.001), less social connectedness (r=-0.24, p<0.01), and less personal recovery (r=-0.24, p<0.05). The healthy subscale was associated with greater self-efficacy (r=0.30, p<0.001) and personal recovery (r=0.30, p<0.01). More healthy music engagement was associated with greater personal recovery (b=0.82, SE=0.22, p<0.001), over and above sociodemographic characteristics and social determinants of health (childhood adversities, discrimination); unhealthy music use was non-significant. The model was significant (F(9,97)=4.44, p<0.001, R²=.29).

Conclusions and Implications: Findings support the two-factor structure of the HUMS among marginalized YA-SMI, extending validation beyond non-clinical adolescent samples. Adaptive and maladaptive patterns of music engagement were evident, suggesting music's dual role as a self-management strategy. Research should examine whether HUMS items require adaptation, test measurement invariance across subgroups, and explore relationships between music engagement and clinical recovery (psychiatric symptomatology). Findings can inform measurement of healthy and unhealthy music engagement, along with novel psychosocial interventions that promote intentional, healthy music use while minimizing potential risks–advancing inclusive inquiry for the common good that centers the voices and experiences of marginalized YA-SMI.