Psychometric properties and population norms of the Brief Symptom Inventory-18 (BSI-18) based on a representative...

The Brief Symptom Inventory-18 (BSI-18) is a widely used self-report measure of psychological distress assessing somatization, depression, and anxiety. Although German population norms were published in 2017 based on data from 2009, updated psychometric evidence using modern statistical methods is lacking.

The BSI-18 was administered to a random probability sample representative of the German residential population aged 16 years and older (N = 2,519; 52.5% women; age range 16-96 years), collected between December 2020 and March 2021. Psychometric evaluation included confirmatory factor analysis (CFA) with the weighted least squares mean and variance adjusted (WLSMV) estimator, internal consistency (McDonald's omega and Cronbach's alpha with bootstrap confidence intervals), measurement invariance across gender, age, and clinical screening status, construct validity via latent correlations with the PHQ-9, GAD-7, and CFQ-11, and age- and gender-stratified population norms. Cross-cohort comparisons were conducted with the normative sample from 2009.

The correlated three-factor model showed good fit (χ²(132) = 644, CFI = 0.986, RMSEA = 0.039, 90% CI [0.036, 0.042]). Internal consistency was good to very good (Global Severity Index [GSI]: ω = 0.966; subscales: ω = 0.855 to 0.893). Full measurement invariance was established across gender, age, and clinical screening status. Latent correlations confirmed the hypothesized convergent validity pattern for depression (PHQ-9: r = 0.912), anxiety (GAD-7: r = 0.889) and somatization (CFQ-11: r = 0.580). Women scored higher on all scales (d = -0.11 to -0.14). Cross-cohort comparisons revealed no significant change in the GSI (d = -0.05, p = 0.105) or depression (d = 0.04, p = 0.200), whereas somatization (d = -0.11, p < 0.001) and anxiety (d = -0.07, p = 0.006) scores were significantly lower in 2020/21.

The German BSI-18 demonstrated sound psychometric properties, including full measurement invariance. The updated population norms provide a useful reference for clinical practice. Cross-cohort comparisons revealed no general increase in distress during the COVID-19 pandemic.