Interactions between health and work: effect of occupational psychosocial conditions on changes in mental health

Outline of reasons and objectives
Current knowledge makes it possible to assert the existence of links between exposure to certain psychosocial risk factors (PRFs) and short-term and long-term damage to health. Exposure to PRFs is generally defined using the Karasek and Siegrist models. Other concepts have emerged (e.g. violence at work, and job insecurity), but, so far, they remain little explored, in particular through longitudinal studies.
The objective of this study was to examine the effect of changes in exposures to a wide variety of PRFs on changes in mental health (depression or anxiety) from 2006 to 2010, as well as on changes in the perceived state of health, and on changes in the limitations on daily activity due to health problems.

Approach
The “Sante et Itineraire Professionnel” (“Health and Career Path”) survey is a longitudinal study on the general population organised jointly by the French Ministries for Health and for Labour, and concerning 14,000 subjects in 2006, 11,000 of whom were able to be questioned again in 2010. This study concerned the 5 684 subjects in jobs both in 2006 and in 2010.
Four exposure situations (subjects exposed both in 2006 and in 2010; subjects exposed in 2006 only; subjects exposed in 2010 only; subjects not exposed either in 2006 or in 2010) for 17 PRFs were analysed to see how they linked up respectively with degradation in mental health, occurrence of limitations on activity, and degradation in perceived state of health in 2010 relative to 2006. These changes were analysed separately in the men and in the women, while taking account of the past career path, of major events (in their personal or professional lives), of the job characteristics, and of the individual characteristics, collected in 2006 or in 2010.

Main results
For the men and the women alike, high complexity of the work and lack of recognition were the exposures that were declared most in 2006 and in 2010. The changes in exposure between 2006 and 2010, although moderate, tended to be deteriorations, in particular in terms of pressure at work and of lack of recognition.
The associations observed with the PRFs were more numerous for mental health and perceived state of health than for limitations on activity. Degradation in health is, overall, more frequent in cases of exposure to PRFs in 2010 only or in cases of exposure both in 2010 and in 2006, compared with exposure in 2006 only. In particular, perception of job insecurity is associated with degradation in all three health indicators both in the men and in the women. The exposures to a feeling of fear during work (own safety/security or safety/security of others) or to a poor work-life balance were significant in the women, and exposures to high quantities of work or to low use of current skills were significant factors in men. Having more “autonomy“ in 2010 in the men and more recognition or resources for doing quality work in 2010 in the women reduces the risk of deterioration of health in 2010.

Discussion
The results observed argue in favour of PRFs having an immediate effect on deterioration of health and the question of whether these effects are reversible remains to be answered. The results also highlight that it is necessary to have a standardised and validated measurement tool that covers all of the PRFs, and such a tool is currently lacking.
Studies on changes in the health of ill subjects or intervention studies focused specifically on the changes in working conditions could usefully feed into addressing the issue of characterising the effects of PRFs. An article for prevention specialists and scientific presentations and articles are planned.