Abstract A036: Hair relaxers use and breast cancer risk by tumor estrogen receptor status: Results from the Ghana Breast Health study
Kaitlin E. White,Quiera Booker,Nicholas Titiloye,Lawrence Edusei,Louise Brinton,Beatrice Wiafe,Ernest Adjei,Jonine Figueroa,On behalf of the Ghana Breast Health Study Team
DOI: https://doi.org/10.1158/1538-7755.disp24-a036
2024-09-23
Cancer Epidemiology Biomarkers & Prevention
Abstract:Background: Hair relaxer use is highly prevalent among women of African descent world-wide and in previous analyses within the Ghana Breast Health Study (GBHS) ever use was associated with a 60% elevated risk of breast cancer. Subsequent studies in the GBHS showed selected alterations in estrogen metabolites with hair relaxer use among controls, suggesting a possible hormonal mechanism may underlie the relation of hair relaxers to breast cancer risk. We investigated if the association between relaxer use and breast cancer risk differed by tumor estrogen-receptor (ER) status and hypothesized ever-users would be at a higher risk of ER+ tumors. Methods: The GBHS is a population-based case-control study conducted between 2013– 2015 in Accra and Kumasi of 1,071 cases with pathologically confirmed invasive breast cancer and 2,106 controls, aged 18-74 years. A standardized interview-based questionnaire collected information on currency and duration of use of hair relaxers and whether products used were lye or non-lye formulations. ER status was available for 915 cases, determined using either mRNA expression data from nCounter® Breast Cancer 360TM Panel (65% of 915 cases) or immunohistochemistry. We used multivariable multinomial logistic regression models to estimate odds ratios (OR) with 95% confidence intervals (CI) for ER+ vs. ER- breast cancers. Models were adjusted for age, hospital site, education, body size, age at menarche, parity, age at first birth, median breastfeeding months/child, menopausal status. To determine if risk related to exposure to relaxers differed by ER status, we used logistic regression analyses with ER+ tumors as the reference and calculated heterogeneity p-values (p-het). Results: The majority of women reported ever use of relaxers (96% of cases and 94% of controls). Of the 1,071 cases included in analyses, 422 (44%) were ER+, 438 (41%) ER-, and 156 (15%) unknown for ER. Elevated risk did not differ by ER status for ever compared to never use (OR ER+ =1.67, 95% CI [0.95–2.93]; OR ER - =1.59, 95% CI [0.92–2.76], p-het=0.90). Risks were highest among former users (OR ER+ =2.19, 95% CI [1.21–3.95]; OR ER- =2.21, 95% CI [1.24–3.93], p-het=0.98. Additional analyses showed elevated risk for those with long durations of use, and those who predominantly used non-lye products, but associations did not differ according to ER tumor status. Conclusion: Our analysis of relaxer use in Ghanaian women showed elevated breast cancer risk, particularly for former, long-term users of non-lye formulations, irrespective of tumor ER status. Additional analyses by tumor and other characteristics related to biologically aggressive tumor types needed to determine if associations differ beyond ER consistent with disparities seen in African ancestry women. Further, future research is warranted regarding specific products used that may provide insights into hormonal and non-hormonal mechanisms underlying observed associations. Citation Format: Kaitlin E. White, Quiera Booker, Nicholas Titiloye, Lawrence Edusei, Louise Brinton, Beatrice Wiafe, Ernest Adjei, Jonine Figueroa, On behalf of the Ghana Breast Health Study Team. Hair relaxers use and breast cancer risk by tumor estrogen receptor status: Results from the Ghana Breast Health study [abstract]. In: Proceedings of the 17th AACR Conference on the Science of Cancer Health Disparities in Racial/Ethnic Minorities and the Medically Underserved; 2024 Sep 21-24; Los Angeles, CA. Philadelphia (PA): AACR; Cancer Epidemiol Biomarkers Prev 2024;33(9 Suppl) nr A036.
oncology,public, environmental & occupational health