The Holland Sleep Disorders Questionnaire: Factorial structure and measurement invariance in a psychiatric sample relative to the general population

Teus Mijnster,Klaas J. Wardenaar,Gretha J. Boersma,Maaike M. van Veen,Daniëlle Cath,Gerard A. Kerkhof,Marike Lancel
DOI: https://doi.org/10.1002/mpr.70004
2024-10-04
International Journal of Methods in Psychiatric Research
Abstract:Abstract Objectives Although common, sleep disorders often remain undiagnosed in psychiatric patients. A screening instrument, like the Holland Sleep Disorders Questionnaire (HSDQ) could improve this. Previous work indicated a 6‐factor structure for the HSDQ, but this hasn't been investigated in psychiatric patients. Methods HSDQ data was collected in a psychiatric‐outpatient sample ( n = 1082) and general‐population sample ( n = 2089). Internal reliability of the HSDQ was investigated and Confirmatory Factor Analyses (CFA) were used to compare 1‐, 6‐, and second‐order 6‐factor models in both samples. Next, multigroup‐CFA was used to investigate measurement invariance. Results Except for one subscale, internal reliability was acceptable in both samples. The 6‐factor structure model fitted best in both samples and investigation of measurement invariance showed evidence for equality of the overall factor structure (configural invariance). Addition of equality constraints on factor loadings (metric invariance) and item thresholds (scalar invariance) showed good fit for all fit statistics, except for one. Exploratory analyses identified three items for metric and three different items for scalar invariance explaining this non‐invariance. Conclusion The HSDQ has a 6‐factor structure in psychiatric patients, which is comparable to the general population. However, due to the observed non‐invariance, users should be cautious with comparing HSDQ scores between psychiatric and general populations.
psychiatry
What problem does this paper attempt to address?