Existing longitudinal studies on the relationship between working time arrangements (WTA) and work-family conflict have mainly focused on the normal causal relationship, that is, the impact of WTA on work-family conflict over time. So far, however, the reversed relationship, that is, the effect of work-family conflict on adjustments in WTA over time, has hardly been studied. Because work-family conflict is highly prevalent in the working population, further insight in this reverse relationship is invaluable to gain insight into secondary selection processes. The aim of this study is to investigate whether work-family conflict is prospectively related to adjustments in work schedules, working hours, and overtime work, and to explore sex differences and different time lags in this relation. Data of the prospective Maastricht Cohort Study were used. To study the effect of work-family conflict on a change from shift- to day work over 32 months of follow-up, male three-shift (n = 727), five-shift (n = 932), and irregular-shift (n = 451) workers were selected. To study effects of work-family conflict on reduction of working hours over 12 and 24 months of follow-up, respectively, only day workers (males and females) were selected, capturing 5809 full-time workers (> or =36 h/wk) and 1387 part-time workers (<36 h/wk) at baseline. To examine effects of work-family conflict on refraining from overtime work over 12 months of follow-up, only day workers reporting frequent overtime work at baseline were selected (3145 full-time and 492 part-time workers). Cox regression analyses were performed with adjustments for age, educational level, and presence of a long-term illness. Work-family conflict was associated with a significantly increased risk of changing from shift- to day work over 32 months of follow-up in three-shift workers (relative risk [RR] = 1.77, 95% confidence interval [CI] 1.19-2.63) but not in five-shift workers (RR = 1.32, 95% CI 0.78-2.24) and irregular-shift workers (RR = 0.81, 95% CI 0.50-1.31). Within day workers, work-family conflict among full-time workers was associated with a significantly increased risk of reducing working hours during 1 yr of follow-up in women (RR = 2.80, 95% CI 1.42-5.54) but not men (RR = 1.34, 95% CI 0.81-2.22). In part-time workers, work-family conflict was associated with a significantly increased risk of reducing working hours during 1 yr of follow-up both in women (RR = 1.99, 95% CI 1.04-3.82) and men (RR = 4.03, 95% CI 1.28-12.68). Whereas the effects of work-family conflict on a reduction of working hours somewhat decreased among female full-time workers after 2 yr of follow-up (RR = 2.13, 95% CI 1.24-3.66), among male full-time workers the effects increased and reached statistical significance (RR = 1.53, 95% CI 1.05-2.21). Work-family conflict was not significantly associated with refraining from overtime work over 1 yr of follow-up. This study shows that work-family conflict has important consequences in terms of adjustments in work schedules and working hours over time, with considerable sex differences. The study thereby clearly illustrates secondary selection processes both in shift- and day workers, with significant implications for labor force participation, emphasizing the need for prevention of work-family conflict.
