Mental Health Literacy Among Future Health and Education Professionals: Validity Evidence of the Spanish Version of the Mental Health Literacy Questionnaire for Young Adults

Mental health literacy (MHL) plays a crucial role in the early identification, management, and prevention of mental health problems, especially among young adults. For future health and education professionals, robust MHL is a foundational component of clinical readiness, ensuring they are adequately prepared to recognize early signs of psychological distress, mitigate stigma, and facilitate timely referrals in healthcare and community settings. This study aimed to examine the internal structure and psychometric properties of the Mental Health Literacy Questionnaire for Young Adults (MHLq-YA) within a Spanish university population. An initial sample of 425 university students enrolled in Nursing and Education programs was recruited. Following data screening for invariant and aberrant responding, the final analytical sample comprised 416 participants. Confirmatory Factor Analysis (CFA) supported a four-factor structure (Knowledge, Stereotypes, Help-Seeking, and Self-Help) with an adequate and structural fit to the data, supporting its multidimensionality with an adequate model fit (CFI = 0.949, TLI = 0.944, RMSEA = 0.068). Evidence of convergent and discriminant validity was observed in three of the four dimensions, while the relationship between Knowledge and Self-Help warranted theoretical interpretation. Convergent validity was supported for three dimensions (Stereotypes, Help-Seeking, and Self-Help), whereas the Knowledge dimension showed comparatively weaker convergent validity (AVE = 0.436). Discriminant validity was supported for all factor pairs under the HTMT criterion, although the stricter Fornell–Larcker criterion was not met for the Knowledge–Self-Help pair, reflecting a conceptual and empirical overlap between these two dimensions that is discussed as a structural limitation of the model. Item Response Theory (IRT) analyses revealed high measurement precision in the lower and moderate ranges of the latent traits, particularly for the Help-Seeking and Self-Help subscales. Ceiling effects and cross-loadings in specific items were resolved through a stepwise model re-specification. The final 26-item version demonstrated robust psychometric performance. These findings support the MHLq-YA as a valid and reliable tool for assessing mental health literacy in academic settings. Crucially, it allows educators to identify gaps in clinical preparedness before students interact with actual patients or vulnerable populations. Ultimately, these results inform targeted interventions to improve mental health awareness, help-seeking behavior, and the overall quality of future clinical care provided by university students.