Explaining educational differences in return to work after early breast cancer: a mediation analysis.

BACKGROUND

The mechanisms behind the effect of education on the return to work (RTW) of survivors of breast cancer (BCS) are unknown.

METHODS

We used data from 2461 BCS from the prospective cohort CANTO. Using inverse odds weights we studied the mediating effect of pre-existing comorbidities, clinical factors, post-treatment health related quality of life, working conditions and household composition in the association between education and non-RTW. We used education in three categories. All models were adjusted for age at diagnosis.

RESULTS

Non-RTW increased at lower levels of education. Compared to high educated women, women with <high school education were more likely not to return to work (RR = 1.43, 95% CI 1.28-1.61). After accounting for all mediators, the direct effect of education on RTW remained elevated but was not statistically significant (RR = 1.18, 0.96-1.46). The association between <high school education and non-RTW was mediated by pre-existing comorbidities (RR = 1.06, 1.00-1.13), post-treatment health-related quality of life (RR = 1.09, 1.01-1.18) and working conditions (RR = 1.19, 1.09-1.29). Differences were observed by age group, with larger inequalities among younger women. Among women aged ≥50, education differences in non-return to work were totally accounted for by all mediators (RR = 0.99, 0.75-1.29) whereas they remained elevated among women aged below 50 (RR = 1.30, 0.99-1.72).

CONCLUSIONS

Individuals with lower education are less likely to return to work. Poorer working conditions account for a large part of this association.

CLINICAL TRIAL: Registration: NCT01993498.