isiXhosa translation of the Patient Health Questionnaire (PHQ-9) shows satisfactory psychometric properties for the measurement of depressive symptoms [Stage 2].

Abstract

Depression is a debilitating illness, and stigma associated with it often prevents people from seeking support. Easy-to-administer and culturally- inclusive tools can allow for early screening for depressive symptoms in primary care clinics, especially in resource-limited settings. In this pre-registered pilot study (Stage 1 Report available at DOI: 10.3389/fpsyt.2022.840912), we produced an open-access isiXhosa-language version of the nine-item Patient Health Questionnaire (PHQ-9), a well-validated measure of depression incidence and severity, using a transcultural translation framework. We validated this isiXhosa PHQ-9 in a sample of N = 47 adolescents living with and without HIV in Cape Town, South Africa who speak isiXhosa at home. Reliability, convergent validity, and criterion validity were assessed, with T scores on the Achenbach System of Empirically Based Assessment Youth Self Report (YSR) form completed previously as reference standard. Our isiXhosa PHQ-9 exhibited satisfactory reliability, with Cronbach's $\alpha =0.866$, inter-item correlations ranging from 0.229 to 0.730, and mean item-total correlation of 0.69. PHQ-9 score and Withdrawn/Depressed component T scores on the Youth Self Report were moderately associated (Spearman's $\rho =0.40,p=0.011)$, indicating acceptable convergent validity. The isiXhosa PHQ-9 showed satisfactory criterion validity (area under the receiver operating characteristic curve, AUC = 0.706), but these analyses were under-powered. Principal component analysis revealed a one-factor solution, with 45.8% of variance explained by the first principal component and all factor loadings above conventional thresholds. Our isiXhosa translation of the PHQ-9 thus exhibited satisfactory psychometric properties in this pilot validation study and performed comparably to other PHQ-9 versions validated in different languages in African and global contexts. This questionnaire may serve as an invaluable culturally-inclusive screening tool for measuring depressive symptoms among isiXhosa speakers. Caution must be exercised as screening tools including the PHQ-9 may over- or under-estimate prevalence of depression. Further validation in larger, independent cohorts may enable wider use of our isiXhosa PHQ-9 as a screening tool in clinics, research studies, and mental health non-profits who serve amaXhosa.