Is assessment of depression equivalent for migrants of different cultural backgrounds? Results from the German population-based Gutenberg Health Study (GHS)

Abstract

BACKGROUND: Bearing in mind the multicultural background of a national population, little is known about the measurement invariance across different cultures or ethnicities of frequently used screeners for depression. For this reason, the main objective of the current study is to assess the measurement invariance of the Patient Health Questionnaire (PHQ-9) across groups with different migration backgrounds. METHODS: We provided psychometric analyses (descriptive statistics at item and scale level, reliability analysis, exploratory [EFA] and confirmatory factor analyses [CFA]) comparing a native population with first- and second-generation migrants of the German population-based Gutenberg Health Study with N = 13,973 participants completing the PHQ-9. Furthermore, we conducted measurement invariance analyses among different groups of first-generation migrants. RESULTS: Comparing the native population with first- and second-generation migrants, a higher prevalence for mental distress was found for first generation. Although mean score patterns were similar for all groups, analyses of item loadings among first-generation migrants yielded some variance in patterns pointing out that certain items have a distinct impact on depression for specific groups. With regard to the factorial validity for all groups, EFA and CFA provided evidence for the proposed one latent factor structure of the PHQ-9. Depression assessed by the PHQ-9 turned out to be equivalent from a psychometric perspective across different groups stratified by their migration background. CONCLUSIONS: Overall, results of thorough scale and item analyses, especially multigroup confirmatory analyses, provided support that depression, assessed by the PHQ-9, can be considered as psychometrically equivalent across all analyzed groups.

Cite this work

No public copy is stored for this record, so only the metadata is cited.

Stored only when both papers are in the corpus — 27 citations exist upstream.

Papers that cite the same references — the relational fallback while embeddings are still backfilling.