Understanding Public Perceptions of Mental Illness in Ghana: Implications for Community-Based Mental Health Care

Background: Despite mental health needs affecting an estimated 3 million Ghanaians, few access formal treatment due to infrastructure gaps, limited education, and sociocultural worldviews shaped by spirituality and religious life. While Ghana’s 2012 Mental Health Act marked a shift toward decentralized, community-based care, public stigma remains a barrier to help-seeking. Mental health stigma in Ghana is complex and multidimensional, encompassing perceptions of dangerousness, social distance, disclosure concerns, and treatability. Understanding stigma within sociocultural contexts is essential for informing mental health policy and practice. To explore how stigma shapes public perceptions of mental illness, this study used a vignette-based design to examine whether Ghanaians recognize symptoms of mental distress as a “problem” and how that recognition relates to different domains of stigma. Specifically, we examined: (1) whether levels of stigma varied based on the vignette subject’s sociodemographic characteristics or symptom type, and (2) how stigma dimensions were associated with whether respondents believed the vignette subject had a mental health problem.

Methods: A convenience sample of 658 Ghanaian adults was recruited from Accra and Tamale through in-person outreach in public spaces. Participants completed a 45-minute tablet-based survey in English, Twi, or Dagbani. The survey included a randomly assigned vignette describing one of five mental health conditions (depression, anxiety, psychosis, suicidal ideation, or gambling disorder), followed by standardized stigma measures. Additional variables included religiosity, mental health knowledge, help-seeking attitudes, self-stigma, and prior exposure to mental illness. Data were analyzed using MANOVA to assess stigma variation by vignette characteristics and hierarchical logistic regression to examine predictors of perceiving the vignette character as having a problem.

Results: Participants were predominantly under 40 (78.7%), urban-residing (80.5%), and identified as Christian (57%) or Muslim (42.2%). MANOVA showed that gender and symptom type did not significantly affect stigma levels. Logistic regression revealed that married participants had lower odds of recognizing a problem than single participants (AOR = 0.51, p < .05). Higher mental health knowledge was associated with decreased odds of problem recognition (AOR = 0.98, p < .05). Relational stigma (AOR = 0.94, p < .001) and treatability stigma (AOR = 0.97, p < .01) were inversely associated with problem recognition, while public stigma (AOR = 1.04, p < .01) and disclosure stigma (AOR = 1.03, p < .05) were positively associated. The final model explained 31% of the variance and showed good fit.

Conclusions: Stigma in Ghana operates across multiple dimensions and differentially influences whether individuals recognize symptoms as a mental health “problem.” Higher knowledge may normalize symptoms or reflect more nuanced interpretations. Positive associations between disclosure/public stigma and problem recognition suggest ambivalence—individuals may recognize distress but anticipate discrimination. Findings support stigma reduction efforts prioritizing mental health literacy, public education, and engagement with traditional and religious leaders. Policy reforms should expand community-based services and integrate mental health into broader care systems. Rather than reinforcing a singular narrative of stigma as uniformly harmful, these findings point to more complex interpretations of mental health that can inform contextually grounded approaches to care.