Age, period, and cohort effects for future employment, sickness absence, and disability pension by occupational gender segregation: a population-based study of all employed people in a country (> 3 million).

Objectives: The occupational gender segregation of the labour market is very strong, both in Sweden and in North America. Nevertheless, there is little knowledge on how this is associated with employees' future employment or morbidity. The objectives of this study were to explore age, period, and cohort effects on future employment and morbidity in terms of sickness absence (SA) or disability pension (DP) among women and men employed in numerically gender-segregated or gender-integrated occupations.

Methods: Based on Swedish nationwide register data, three population-based cohorts of all people living in Sweden, with a registered occupation, and aged 20-56 years at inclusion in 1985 (N = 3,183,549), 1990 (N = 3,372,152), or 2003 (N = 3,565,579), respectively, were followed prospectively for 8 years each. First, descriptive statistics of employment and SA/DP at follow-up were calculated, related to level of gender segregation/integration of occupation at inclusion. Second, differences between birth cohorts (those born in 1929-1983, respectively) were estimated within each of the periods 1985-1993, 1990-1998, and 2003-2011, using mean polish analyses.

Results: Women and men in gender-segregated occupations differed in relation to future employment rates and SA/DP. However, these differences decreased over time. Furthermore, the results show a birth cohort effect; those born in 1943-1956 remained in employment to a higher extent and also had lower rates of SA/DP than all other birth cohorts.

Conclusion: Differences between people in the five categories of gender-segregated occupations decreased over time. Although age and period are important when explaining the outcome, also birth cohort effects have to be considered, both from a public and an occupational health perspective.