Comparison of alternative versions of the job demand-control scales in 17 European cohort studies: the IPD-Work consortium
- Equal contributors
1 Institute of Environmental Medicine, Karolinska Institutet, Box 210, SE-171 77 Stockholm, Sweden
2 School of Health Sciences, Jönköping University, Jönköping, Sweden
3 Finnish Institute of Occupational Health, Helsinki, Finland
4 Department of Public Health, Ghent University, Ghent, Belgium
5 Department of Epidemiology and Public Health, University College London, London, UK
6 MRC Centre for Cognitive Ageing and Cognitive Epidemiology, University of Edinburgh, Edinburgh, UK
7 Versailles-Saint Quentin University, Versailles, France
8 School of Public Health, Université Libre de Bruxelles, Brussels, Belgium
9 Inserm U1018, Centre for Research in Epidemiology and Population Health, Villejuif, France
10 Department of Public Health, University of Helsinki, Helsinki, Finland
11 Department of Health Sciences, Mid Sweden University, Sundsvall, Sweden
12 Stress Research Institute, Stockholm University, Stockholm, Sweden
13 Department of Public Health and Clinical Medicine, Occupational and Environmental Medicine, Umeå University, Umeå, Sweden
14 Department of Public Health, University of Turku, Turku, Finland
15 Folkhälsan Research Center, Helsinki, Finland
16 Finnish Institute of Occupational Health, Turku, Finland
17 Occupational and Environmental Medicine, Uppsala University, Uppsala, Sweden
18 Department of Behavioral Sciences, University of Helsinki, Helsinki, Finland
BMC Public Health 2012, 12:62 doi:10.1186/1471-2458-12-62Published: 20 January 2012
Job strain (i.e., high job demands combined with low job control) is a frequently used indicator of harmful work stress, but studies have often used partial versions of the complete multi-item job demands and control scales. Understanding whether the different instruments assess the same underlying concepts has crucial implications for the interpretation of findings across studies, harmonisation of multi-cohort data for pooled analyses, and design of future studies. As part of the 'IPD-Work' (Individual-participant-data meta-analysis in working populations) consortium, we compared different versions of the demands and control scales available in 17 European cohort studies.
Six of the 17 studies had information on the complete scales and 11 on partial scales. Here, we analyse individual level data from 70 751 participants of the studies which had complete scales (5 demand items, 6 job control items).
We found high Pearson correlation coefficients between complete scales of job demands and control relative to scales with at least three items (r > 0.90) and for partial scales with two items only (r = 0.76-0.88). In comparison with scores from the complete scales, the agreement between job strain definitions was very good when only one item was missing in either the demands or the control scale (kappa > 0.80); good for job strain assessed with three demand items and all six control items (kappa > 0.68) and moderate to good when items were missing from both scales (kappa = 0.54-0.76). The sensitivity was > 0.80 when only one item was missing from either scale, decreasing when several items were missing in one or both job strain subscales.
Partial job demand and job control scales with at least half of the items of the complete scales, and job strain indices based on one complete and one partial scale, seemed to assess the same underlying concepts as the complete survey instruments.