Response to invited commentary: Joensuu et al. respond to "Structure and context matters"

Unique associations of the Job Demand-Control-Support model subscales with leisure-time physical activity and dietary energy intake

Leisure-time physical activity (LTPA) and dietary energy intake are two important health behaviours, which at too low or high levels respectively, are associated with overweight and obesity. This study explores associations between subscales of the Job Demand-Control-Support (JDCS) model, LTPA and dietary energy intake. A cross-sectional design sampled current employees (N=433) from a South Australian cohort using a computer-assisted telephone interview and a self-completed food frequency questionnaire. In analyses adjusted for sex, age, and sociodemographic variables, higher levels of skill discretion were associated with increased odds for attaining sufficient physical activity (OR=2.45; 95% CI=1.10-5.47). Higher levels of decision authority were associated with reduced odds (OR=0.43; 95% CI=0.20-0.93) for being in the highest tertile of daily energy intake. Higher scores for coworker support were associated with increased odds (OR=2.20; 95% CI=1.15-4.23) for being in the highest tertile of daily energy intake. These findings support the consideration of the individual JDCS subscales, since this practice may reveal novel associations with health behaviour outcomes, thereby presenting new opportunities to improve employee health and wellbeing.

Differential associations of job control components with both waist circumference and body mass index

INTRODUCTION

The Job Demand-Control-Support (JDCS) model is commonly used to investigate associations between psychosocial work factors and employee health, yet research considering obesity using the JDCS model remains inconclusive.

OBJECTIVE

This study investigates which parts of the JDCS model are associated with measures of obesity and provides a comparison between waist circumference (higher values imply central obesity) and body mass index (BMI, higher values imply overall obesity).

METHODS

Contrary to common practice, in this study the JDCS components are not reduced into composite or global scores. In light of emerging evidence that the two components of job control (skill discretion and decision authority) could have differential associations with related health outcomes, components of the JDCS model were analysed at the subscale level. A cross-sectional design with a South Australian cohort (N = 450) combined computer-assisted telephone interview data and clinic-measured height, weight and waist circumference.

RESULTS

After controlling for sex, age, household income, work hours and job nature (blue vs. white-collar), the two components of job control were the only parts of the JDCS model to hold significant associations with measures of obesity. Notably, the associations between skill discretion and waist circumference (b = -.502, p = .001), and skill discretion and BMI (b = -.163, p = .005) were negative. Conversely, the association between decision authority and waist circumference (b = .282, p = .022) was positive.

CONCLUSION

These findings are significant since skill discretion and decision authority are typically combined into a composite measure of job control or decision latitude. Our findings suggest skill discretion and decision authority should be treated separately since combining these theoretically distinct components may conceal their differential associations with measures of obesity, masking their individual importance. Psychosocial work factors displayed stronger associations and explained greater variance in waist circumference compared with BMI, and possible reasons for this are discussed.

A Longitudinal Analysis of Changes in Job Control and Mental Health

Deteriorating job control has been previously shown to predict poor mental health. The impact of improvement in job control on mental health is less well understood, yet it is of policy significance. We used fixed-effects longitudinal regression models to analyze 10 annual waves of data from a large Australian panel survey (2001-2010) to test within-person associations between change in self-reported job control and corresponding change in mental health as measured by the Mental Component Summary score of Short Form 36. We found evidence of a graded relationship; with each quintile increase in job control experienced by an individual, the person's mental health increased. The biggest improvement was a 1.55-point increase in mental health (95% confidence interval: 1.25, 1.84) for people moving from the lowest (worst) quintile of job control to the highest. Separate analyses of each of the component subscales of job control-decision authority and skill discretion-showed results consistent with those of the main analysis; both were significantly associated with mental health in the same direction, with a stronger association for decision authority. We conclude that as people's level of job control increased, so did their mental health, supporting the value of targeting improvements in job control through policy and practice interventions.