Associations Between Self-Rated Health and Mortality in the Norwegian Women and Cancer (NOWAC) Study.

Abstract

We investigated the association between self-rated health (SRH) and cancer incidence and SRH and all-cause mortality among Norwegian women. We used data from 110,104 women in the Norwegian Women and Cancer (NOWAC) cohort aged 41-70 years at baseline. We used flexible parametric survival analysis with restricted cubic splines to calculate hazard ratios (HRs) and 95% confidence intervals (CIs) for the association between SRH and mortality in the entire cohort. We employed the same method in a multistate design to assess associations between baseline SRH and 1) cancer incidence, and 2) all-cause mortality in subgroups of women who did and did not receive a cancer diagnosis during follow-up. With very good SRH as reference category for all associations and median age at end of follow-up, lower SRH was associated with increased mortality (HRgood SRH 1.19, 95% CI 1.12-1.26) and HRpoor SRH 1.81, 95% CI 1.66-1.97). Lower SRH at baseline was associated with cancer incidence (HRgood SRH 1.14, 95% CI 1.08-1.20 and HRpoor SRH 1.44, 95% CI: 1.32-1.58). Poor baseline SRH was associated with increased mortality for women who received a cancer diagnosis (HRpoor SRH 1.20, 95% CI 1.04-1.39), and SRH showed a strong association with increased mortality for women who stayed cancer free (HRgood SRH 1.59, 95% CI 1.44-1.77 and HRpoor SRH 3.34, 95% CI 2.91-3.84). Lower SRH at baseline predicted increased cancer risk and all-cause mortality in middle-aged to older women. Poor SRH at baseline predicted all-cause mortality in women who later received a cancer diagnosis. Both good and poor SRH at baseline predicted all-cause mortality in women who stayed cancer-free, and the association was stronger for these women compared to both the entire cohort and to women who were subsequently diagnosed with cancer.