Work hours and cortisol variation from non-working to working days

This study aims at modelling the relationship between the number of work hours per week and cortisol variation across 3 days by comparing non-working day to working day in a population of day-shift workers. Questionnaire data and saliva samples for cortisol analysis were collected on 132 day-shift workers employed in 13 workplaces in Canada. Consenting workers provided 5 saliva samples a day (awaking, 30 min after awaking, 2 p.m., 4 p.m., bedtime) repeated 3 times (Saturday, Tuesday, Thursday) over a week, to provide measures at work and non-work times and settings. Multilevel regression models were estimated from cortisol measurements at each occasion within a day at level-1, days at level-2 and workers at level-3. Controlling for gender and age, results revealed significant variations in salivary cortisol concentration between sample, day and worker levels. Cortisol increases linearly from non-working day to work days and work hours per week interacted with days, revealing a positive association between the number of work hours per week and cortisol concentrations during these days. Work hours per week did not interact with awaking, 30 min after awaking, 2 p.m., 4 p.m., and bedtime cortisol concentrations. Gender and age had neither main effects nor interaction effects. No significant cortisol variations were found between workplaces. These results suggest that work hours act as a stressor that is associated with significant variations in cortisol concentrations over working days. Non-working days may contribute to stress reduction in workers who experience longer work hours.