When Knowledge Falls Short: A Systematic Review on the Correlation of Mental Health Knowledge With Stigma and Help-Seeking

Improving Mental Health Literacy (MHL) is considered important to facilitate help-seeking for mental health problems. However, it is crucial to ascertain the true significance of increasing mental health knowledge (the central component of MHL) for prevention efforts. This systematic review analyses the direction and effect size of the correlations between knowledge and different facets of stigma (personal, perceived, self) and help-seeking (attitudes, intention, behavior), highlighting current research gaps. We conducted a PRISMA-based systematic analysis of quantitative cross-sectional and longitudinal data on the associations between mental health knowledge and mental health related stigma or help-seeking in Western populations. Peer-reviewed articles published since 1997 were retrieved via PubMed, PsycINFO and Web of Science, theses via DART-EUROPE, EBSCO-OpenDissertation and Google. The number of studies, types of outcomes, country, population, effect direction and size (median, range) are reported. We identified 47 studies on the association between knowledge and stigma, and 38 on its association with help-seeking. Knowledge showed a medium-sized association with personal stigma (Mdn r = −0.28; range = −0.54; −0.06) and attitudes towards help-seeking (Mdn r = 0.29; range = 0.04; 0.58) and a small association with self-stigma (Mdn r = −0.16; range = −0.39; −0.02), help-seeking intention (Mdn r = 0.15; range = −0.07; 0.40) and help-seeking behavior (Mdn r = 0.15; range = −0.04; 0.68). Knowledge was not consistently related to perceived stigma (Mdn r = −0.01; range = −0.29; 0.16). While the direction of the correlations meets our expectations, the only small to moderate effect sizes indicate that knowledge is relevant but not sufficient for reducing self-stigma and improving help-seeking. We identified several research gaps, such as a scarcity of studies with rigorous methodological standards and with older or clinical populations. We also propose including a broader range of potentially empowering variables in addition to knowledge in future studies.