Working longer hours and body weight: An Australian study using household panel data (with measures of paid and unpaid time) to provide gender-specific estimates.

Time is a resource for health, and when time is constrained, people have less opportunity to maintain good health. This study focuses on the relationship between paid work hours (with a focus on long hours) and body weight for Australian men and women. Time is conceptualised as a 24-hour system, including time in paid work, time in unpaid work, and discretionary time (available for health promoting activities). We propose that to accurately estimate the relationship between long paid work hours and body weight, analyses need to take account of unpaid work hours, and that this is particularly important for women. Analyses utilised 16 waves of data from the Household, Income and Labour Dynamics in Australia (HILDA) panel study, with data on hours in paid and unpaid work and BMI at all waves (n = 113,084 observations, 54,664 from women and 58,424 from men). We used Mixed Effects models with a Two Stage Residual Inclusion (2SRI) approach to account for time unobserved heterogeneity and the reciprocity between time uses (paid and unpaid) and body weight. The results showed that for every 10 hours worked above the gender-specific average, women weighed 762 grms more and men weighed 1.34 kgs more. When the analyses were restricted to sedentary jobs this increased to 938 grms for women and 1.68 kgs for men. We contribute new evidence highlighting the importance of including unpaid work hours in research that aims to understand working time and health.