Circulating 27-hydroxycholesterol, sex steroid hormones, and risk of mortality among postmenopausal females with breast cancer: The Multiethnic Cohort Study.
Abstract
Background
Biomarkers in the hormonal and metabolic pathways linking obesity and mortality can influence survival after breast cancer diagnosis.
Methods
We prospectively examined associations with mortality for pre-diagnostic plasma levels of circulating 27-hydroxycholesterol (27HC), estrone, testosterone, and sex hormone binding globulin (SHBG) measured on average 6.7 years before breast cancer diagnosis in 1,009 African American, Japanese American, Latino, Native Hawaiian, and White postmenopausal females not using menopausal hormone therapy at blood draw from the Multiethnic Cohort (MEC) study (1993-2017). Hazard ratios (HRs) and 95% confidence intervals (CI) were estimated using multivariable Cox proportional hazards regression adjusting for selected factors related to demographics, prognosis, lifestyle, and medical history.
Results
Over an average 7.8 follow-up years (standard deviation=4.2), 300 total deaths occurred, including 83 breast cancer-specific deaths. Inverse associations were found for 27HC (highest versus lowest tertile: HR=0.73, 95% CI=0.54-0.98) with all-cause but not breast-cancer specific mortality, and estrone (highest versus lowest tertile: HR=0.37, 95% CI=0.20-0.70) with breast cancer-specific but not all-cause mortality. SHBG was positively associated with all-cause (highest versus lowest tertile: HR=1.76, 95% CI=1.28-2.41) but not breast cancer-specific mortality. No associations were found for testosterone. There was no statistically significant heterogeneity in associations by body mass index, estrogen receptor status, cancer stage, or race and ethnicity.
Conclusions
This seminal study of racially and ethnically diverse female breast cancer survivors found associations for pre-diagnostic 27HC, estrone, and SHBG with mortality. IMPACT Future research studies should prioritize larger breast cancer cohorts to confirm our findings and evaluate associations across important subgroups.