Hair Relaxers Not Associated with Breast Cancer Risk: Evidence from the Black Women's Health Study

African American women, especially younger women, commonly use hair relaxers to straighten their hair, often over periods of many years. “Lye” relaxers contain sodium hydroxide; no-lye relaxers contain calcium hydroxide and guanidine carbonate; and “thio” relaxers contain thioglycolic acid salts (1). There is no evidence that any of these substances is carcinogenic. However, hair relaxer ingredients are only partly monitored by the Food and Drug Administration: manufacturers need not list flavorings, fragrances, and ingredients that are trade secrets (2). California recently enacted the first state cosmetic regulatory law, the California Safe Cosmetics Act, which will require reporting of potentially harmful ingredients to the state, which will then inform consumers (3). Thus, the carcinogenic potential of hair relaxers is currently unknown. Hair relaxers can cause burns and lesions in the scalp, facilitating entry of hair relaxer constituents into the body (1, 4-7). The higher incidence of breast cancer in U.S. Black women than in U.S. White women before age 45 years is only partly explained by known risk factors (8, 9). In light of all these considerations, we assessed the relation of hair relaxer use to the incidence of breast cancer in African American women.

The Black Women's Health Study began in 1995 when Black women, ages 21 to 69 years, from across the United States completed postal health questionnaires (9, 10). Fifty-nine thousand women, whose names and addresses were deemed to be valid, have been followed every 2 years through postal questionnaires; ≥80% of the cohort has completed each follow-up questionnaire through 2003. Institutional review boards of Boston University and Howard University approved the study protocol. The participants provided consent by filling out questionnaires and giving written consent for medical records.

The baseline 1995 questionnaire obtained information on many factors including risk factors for breast cancer and medical history. Biennial follow-up questionnaires updated information on risk factors and ascertained incident breast cancer. The 1997 follow-up questionnaire included questions about use of “chemical hair straighteners”: age at first use (<10, 10-19, 20-29, ≥30 years), frequency of use per year (about 1 time, 2 times, 3-4 times, 5-6 times, ≥7 times), total number of years of use (<1, 1-4, 5-9, 10-14, 15-19, ≥20), number of burns (a break in the skin, not just tingling) during the application of chemical straighteners (never, 1-2 times, 3-4 times, 5-9 times, ≥10 times), and type of formulation used most often (lye, no-lye, don't know).

During each follow-up cycle, we mail multiple waves of questionnaires to women who have not yet responded. During the 1997 cycle, 1,060 women returned duplicate questionnaires. Weighted κ values (11) for agreement between responses on the hair relaxer use questions were 0.79 for age at first use, 0.79 for frequency of use, 0.80 for duration of use, 0.71 for number of burns, and 0.54 for type of relaxer (all P < 0.05) indicating good to excellent reproducibility.

The validation of self-reported cases of incident breast cancer in the Black Women's Health Study is an ongoing process. To date, 927 of 1,069 women who reported incident breast cancer since baseline in 1995 through 2005 have been approached, and medical record data have been obtained for 589 (64%) women. Among the latter, 586 (99.5%) women were confirmed to have had breast cancer, of which 88% was invasive cancer. An additional 135 women who were not willing to release their medical records because of concerns about confidentiality confirmed that their diagnosis was indeed breast cancer. Because of the high degree of accuracy of reporting of breast cancer by Black Women's Health Study participants, we have included all self-reported cases of breast cancer in the analysis, with the exception of the cases disconfirmed by medical record data.

We began follow-up for the present analysis in 1997, the year in which the information on hair relaxers was collected, and we assessed that use in relation to the subsequent occurrence of breast cancer reported on the 1999, 2001, and 2003 questionnaires. Among 53,177 women who completed the 1997 questionnaire, 5,010 were excluded because they had prevalent breast cancer (826) or another cancer (662), did not complete the information on hair relaxer use (793), or did not complete a follow-up questionnaire after 1997 (2,729), which left 48,167 women for the present analysis.

Age-stratified Cox regression was used to estimate incidence rate ratios (IRR) for hair relaxer use in relation to the incidence of breast cancer (12). The Anderson-Gill data structure was used to update time-varying covariates (13). Women who never used hair relaxers (2,255) or used them for less than a year (1,394), referred to hereafter as “nonusers,” composed the reference category for the estimation of IRRs. Women contributed person-years from 1997 until the occurrence of breast cancer, loss to follow-up, death, or the end of follow-up in 2003, whichever came first. Control for age at menarche, age at first birth, parity, family history of breast cancer, benign breast disease, alcohol use, physical activity, educational attainment, and geographic region of residence did not alter the IRRs by >2%. We therefore controlled only for age (1-year intervals) and questionnaire cycle in the Cox models.

Women <45 years of age were more likely to use hair relaxers than older women (94% versus 89%). Use before age 20 years (73% versus 30%) and use with a frequency of at least 7 times a year (33% versus 22%) were also greater among the younger women. There was little variation in use across categories of the breast cancer risk factors considered as potential confounders in the present analyses.

During 266,298 person-years of follow-up, 574 cases of incident breast cancer were ascertained. The IRR for hair relaxer use relative to nonuse was 1.04 [95% confidence interval (95% CI), 0.78-1.39]. IRRs were close to 1.0 for all categories of duration of use, frequency of use, age at first use, years of use, and type of relaxer used (Table 1). Among the women with more intensive use, those who used hair relaxers 7 or more times a year, the IRR for ≥20 years of use compared with no use was 0.98 (95% CI, 0.78-1.30). Results were unchanged when in situ cases were excluded.

Within strata of age (<45 and ≥45 years), none of the IRRs differed materially from 1.0 (Table 2).

For use at least 7 times a year for ≥20 years relative to nonuse (Table 3), the IRR was 0.84 (95% CI, 0.42-1.71) among women <45 years of age and 1.03 (95% CI, 0.68-1.56) among women ≥45 years of age.

In the present study of African American women, increases in breast cancer risk were not associated with any categories of duration of hair relaxer use, frequency of use, age at first use, number of burns experienced during use, or type of relaxer used. Of particular interest, null associations were observed among younger women who used relaxers at earlier ages and more frequently than older women. Thus, the findings provide empirical evidence that hair relaxers are not carcinogenic to the breast and do not contribute to the higher incidence of breast cancer in African American women than in White women before age 45 years.

Hair dye use has been associated with an increased risk of various cancers in some studies, but these results have generally not been confirmed (14). The present study is the first to assess hair relaxers in relation to a cancer.

Because our study was prospective, biased reporting of hair relaxer use is not a concern, although random misclassification of use would have tended to dilute associations. Follow-up rates were sufficiently high to allay concerns about selective losses. We found that self-reports of breast cancer were accurate and that potential confounding factors did not alter the risk estimates. The study was large and had excellent statistical power to detect small increases in risk. Because the study had no information on individual brands, however, a harmful effect of particular products cannot be ruled out.

We thank the participants in the Black Women's Health Study for their valuable contributions.