An Exploration of Job Strain with Australian Workers' BMI: Is there an Association?

Abstract

Work stress is prevalent globally and within Australia, with the potential to affect individuals, organisations, and economies adversely. Similarly, the increase in population BMI is an ongoing trend globally and within Australia that is associated with negative outcomes for individuals' and population health outcomes, as well as economic loss for individuals, organisations, and economies. The simultaneous rise in each, in combination with other evidence to hypothesise a connection, provides reasons to suspect there may be an association between work stress and increased BMI. Job strain is a widely used and validated measure of work stress (Karasek, 1979; Karasek & Theorell, 1990). Existing research investigating an association between job strain and BMI is mixed, and there is a paucity of research within Australia. The main aim of this thesis is to explore evidence for an association of job strain with BMI using data from the Personality and Total Health Through Life (PATH) survey - a planned longitudinal survey, sampling three age cohorts (20s, 40s, 60s) from the Canberra region (ACT, Australia). Four investigations building on from each other were conducted: The first analyses chapter used cross-sectional data to examine whether there was evidence for an association of job strain with BMI. Logistic regression methods were used to examine whether job strain predicted overweight or obesity (n= 1947 males and 1981 females, from the 20s and 40s cohorts sampled at wave 2). Simultaneously, linear regressions modelled BMI continuously as a function of job strain using the same sample. No evidence for an association of job strain with overweight, obesity, or continuous BMI was found. However, the possibility that job strain may be associated with BMI over time remained. The second analyses chapter investigated the possibility that job strain may be associated with BMI over time. Four waves of data measured from the 40s cohort (n= 1130 males and 1141 females) across 12-years were used to explore this possibility in four varied multi-Level models (MLMs). No evidence was found that job strain was associated with BMI over time. The third analyses chapter considered that individuals' higher or lower than average BMI at baseline might moderate an effect of job strain on BMI over time. To examine this, the four MLMs used in the prior chapter were re-employed to analyse the same 40s sample, with an interaction term for baseline BMI and job strain included. Baseline BMI was modelled continuously and categorically in interaction with job strain (in separate analyses), and job strain was also tested as a predictor of BMI over time in analyses stratified by baseline BMI category. Whilst there were some inconsistent results between analyses with BMI modelled differently, particularly for females, there was minimal evidence that a relationship between job strain and BMI was dependent upon BMI at baseline. The fourth analyses chapter examined whether counts of job strain might accumulate over time to affect BMI at follow-up, and whether this may be dependent upon baseline BMI. Participants from the 40s cohort with data on job strain at every wave and BMI at baseline and follow-up were analysed (n= 525 males and 464 females). Linear regression was used to examine evidence for whether BMI at follow-up was associated with 1, 2, or 3 and 4 (combined) counts of job strain, compared to individuals who did not experience job strain (0 counts). There was no evidence of a pattern of increasing association between more counts of job strain and BMI outcomes, both in direct analyses and in those dependent upon BMI at baseline. Overall, the findings from this thesis suggest no evidence that job strain is meaningfully associated with BMI in this Australian sample. However, this research does point towards the need to explore job strain in connection with BMI in further large, Australian population-based longitudinal samples. Show more