The Kessler 6 psychological distress scale (K6) in swedish: a psychometric study of reliability, factor structure and measurement invariance in a large population-based sample.

@article{ehnsio2026kessler,
  title = {{The Kessler 6 psychological distress scale (K6) in swedish: a psychometric study of reliability, factor structure and measurement invariance in a large population-based sample.}},
  author = {Karolina Ehnsiö and Sanna Tiikkaja and Johnny Pellas},
  journal = {BMC psychology},
  year = {2026},
  month = oct,
  doi = {10.1186/s40359-026-05857-9},
  url = {https://europepmc.org/article/MED/42920194},
  note = {CC BY},
}

TY  - JOUR
TI  - The Kessler 6 psychological distress scale (K6) in swedish: a psychometric study of reliability, factor structure and measurement invariance in a large population-based sample.
AU  - Karolina Ehnsiö
AU  - Sanna Tiikkaja
AU  - Johnny Pellas
PY  - 2026
DA  - 2026/10/23
JO  - BMC psychology
DO  - 10.1186/s40359-026-05857-9
UR  - https://europepmc.org/article/MED/42920194
AB  - Background  Psychological distress is a major public health concern, and brief, reliable screening instruments are essential for early detection and monitoring. The six-item Kessler Psychological Distress Scale (K6) is widely used internationally; however, evidence regarding its psychometric properties in Sweden remains limited. This study examined the reliability, factor structure, and measurement invariance of the Swedish version of the K6 in a large population-based sample.Methods  Data were obtained from the Life and Health Survey (2022), conducted in five regions in Mid-Sweden. The study included 31,759 adults aged 18-102 years. Internal consistency was assessed using McDonald's omega total. Structural validity was evaluated using confirmatory factor analysis, comparing a one-factor model, a two-factor model representing depression and anxiety, and a bifactor model comprising a general psychological distress factor and two specific factors representing depression and anxiety. Measurement invariance of one-factor model was examined across age groups and by gender using multigroup CFA, testing configural and metric invariance.Results  Internal consistency was high across all demographic groups (McDonald's omega total 0.883-0.899). The one-factor model demonstrated best fit, with acceptable values according to CFI, TLI, and SRMR, but unacceptable according to RMSEA (CFI = 0.996, TLI=0.993, RMSEA=0.069, SRMR=0.020), with standardized loadings ranging from 0.70 to 0.91. The two-factor model showed acceptable fit according to CFI, TLI and SRMR, but unacceptable fit according to RMSEA (CFI=0.989, TLI=0.982, RMSEA=0.109, SRMR=0.040). Standardized factor loadings were high for both factors (Factor 1: 0.78; Factor 2: 0.78-0.91). The correlation between the two latent factors was extremely strong (r=.99). The bifactor model did not converge and therefore could not be evaluated or interpreted further. Configural and metric invariance were supported across the age groups and between men and women.Conclusions  The Swedish version of the K6 demonstrates strong reliability and supports a unidimensional structure. Measurement invariance testing supported comparable measurement of psychological distress across age and gender. The scale is suitable for monitoring psychological distress in public health surveys and may also be useful as a brief screening tool in clinical practice. Further research should establish clinically relevant threshold values in a Swedish context.Trial registration  Not applicable.
SN  - 2050-7283
ER  -