Characteristics of the study sample.

Objective
The Swedish Demand-Control-Support Questionnaire (DCSQ) is used to measure psychosocial work environment. Nine of the original 17 DCSQ items were used in the Trøndelag Health Study (HUNT) in 2017–2019 (HUNT4); three items from each of the three dimensions of demand, control, and support. The goal of this paper was to assess the internal reliability and dimensional structure of the nine DCSQ items used in HUNT4.
Method
HUNT4 participants registered with an occupation, and who had responded to all DCSQ items were included in the sample. Internal reliability and consistency of the nine DCSQ items were tested using composite reliability and item total correlation. A confirmatory factor analysis (CFA) was performed to determine whether the nine DCSQ items used in HUNT4 had a similar factor structure as the original 17 items. CFA was performed on the sample as a whole, before testing the models on different occupational groups to check if the dimensional structure remained the same.
Results
Of 56,041 HUNT4 participants (participation rate: 54%), 21,187 had a known occupation and responded to all nine DCSQ items. The mean age of the sample was 50.6 years (SD = 10.2), and the sample consisted of 57.3% females and 42.7% males. The best model fit was achieved by removing the item “Work creativity” (λ = 0.398, item total correlation 0.334) from the model. The same three-factorial structure as in the original DCSQ was seen with the remaining eight DCSQ items, with good internal consistency of all three dimensions (composite reliability ranged from 0.709 to 0.851). This dimensional structure remained the same when tested on all occupational groups.
Conclusion
The results indicate that the shortened version of the Swedish DCSQ used in HUNT4 can be used to assess aspects of demand, control, and social support at work.