Timing and duration of work-family experiences and mental health in young adulthood: Applying feature selection

Background

Little is known about whether and how the timing and duration of work-family experiences are associated with subsequent mental health outcomes in the current generation of young adults. Such information may inform the development of timely interventions to promote better mental health. In this study, we use two complementary approaches to examine the complex relationship between work, family and mental health in young Dutch adults born around 1990. First, we examine the association between work-family trajectories and mental health. Second, we examine which timing and duration features of work-family trajectories are related to subsequent mental health.

Methods

Sequences of monthly work-family states from ages 18 to 28 years were constructed in a sample of Dutch young adults participating in the TRAILS cohort study (N = 992). Mental health was operationalised as experiencing internalising or externalising problems at age 29. We created a typology of work-family trajectories by clustering individual sequences and used regression analysis to examine the association between work-family trajectories and mental health. Next, we used the Boruta feature selection algorithm to identify timing and duration features of work-family trajectories related to mental health at age 29 and conducted a regression analysis to examine the associations between the identified features and mental health. All regression analyses were controlled for mental health problems between ages 11 and 16.

Results

Women in the work-family trajectories long education and inactive were most likely to experience internalising problems at age 29. Women in the work-family trajectory inactive were also most likely to experience externalising problems. In line, the feature selection analysis showed that both being inactive and being in education in the first stage of young adulthood, being a parent and in work in the last stage of young adulthood and the total duration of inactivity were associated with internalising problems. Being inactive in the first and the last stages of young adulthood and being inactive and a parent in the third stage of young adulthood were associated with externalising problems. Men in the work-family trajectory inactive were most likely to experience both internalising and externalising problems. Additionally, men in the work-family trajectory continuous education and work were most likely to report externalising problems. The feature selection analysis showed that being in education in the last stage of young adulthood was associated with internalising problems. Similar to women, inactivity in the first stage of young adulthood was associated with internalising and externalising problems at age 29.

Conclusion

Work-family trajectories in young adulthood are associated with mental health problems at age 29, even when controlling for early-life mental health problems. In particular, the timing and duration of inactivity and education were found to be associated with subsequent mental health problems. Combining sequence analysis and feature selection provides a more nuanced understanding of the dynamics between work-family experiences and mental health.