Mainstreaming mental health in occupational health and safety : an exploration of benchmark dose modelling to regulate job demands and job control

Abstract

The move to service and knowledge industries in industrialised countries means that workers are now less exposed to physical hazards. However, exposure to psychosocial hazards remains widespread. These psychosocial hazards are related to the way work is designed and managed (for example, heavy workloads) and are associated with an increased risk of common mental disorders. Psychosocial hazards are usually regulated as part of the general duty requirements of occupational health and safety (OHS) legislation. To date, there are no specific regulations for psychosocial hazards that delineate the acceptable level of exposure to these hazards. Nevertheless, acceptable exposure levels, based on critical exposure levels, are a part of OHS regulatory framework for many occupational exposures. The lack of such regulatory guidelines for psychosocial hazards makes it difficult for risk assessment and risk management of these hazards. This research sought to adapt the benchmark dose (BMD) method to identify critical exposure levels for psychosocial hazards and mental health. It included a series of studies that addressed three key challenges: 1) the multifactorial nature of mental health, 2) self-report assessment of exposure and mental health and 3) curvilinear dose{u00AD}response relationships. The analyses were based on two waves of the Personality and Total Health (PATH) Through Life study (N = 4004), which collects data every four years from age cohorts living in south-eastern Australia. I chose job demands and job control as exposure variables because they are employer modifiable and have a well{u00AD}established association with poor mental health. Poor mental health was measured as depression and anxiety symptoms, the most common mental disorders in the workforce. A sequenced approach to BMD analysis was undertaken by first conducting unadjusted BMD analysis, which is the conventional BMD analysis. This was then extended by conducting adjusted analysis, taking into account other risk factors for poor mental health (previous mental health status, personality and socio-demographic factors). Based on the 15-item measure of job control, adjusted analyses showed that workers needed to report they 'often' or 'sometimes' have control over many (nine out of 15) aspects of their job to reduce the risk of poor mental health. For job demands, keeping heavy workload demands to a medium level is best for reducing the risk of poor mental health. This translates to workers saying they 'often' or 'sometimes' have heavy workload demands to two out of four different aspects of job demands measured in this study. Adjusted analyses with a second depression and anxiety scale produced similar results, suggesting that critical exposure levels are comparable between different measurements of mental health. The current study contributes to the field in three ways. First, this study extends the concept of the critical exposure level approach to psychosocial hazards. Second, it provides an adapted BMD method to enable identification of critical exposure levels for psychosocial hazards, which can be used in future BMD studies. Third, critical exposure levels, such as those identified in this thesis, can be used to develop acceptable exposure levels and health-based risk assessment tools for psychosocial hazards.