Hair Relaxer Use and Risk of Uterine Leiomyomata in African-American Women

Lauren A. Wise; Julie R. Palmer; David Reich; Yvette C. Cozier; Lynn Rosenberg

Disclosures

Am J Epidemiol. 2012;175(5):432-440. 

In This Article

Results

Nearly 94% of women in our cohort reported using hair relaxers for at least 1 year (Table 1). The frequency of hair relaxer use ranged from as low as 8.6% for 1–2 times per year to as high as 32.6% for ≥7 times per year. Frequency of use was highest among participants who were current users of oral contraceptives, had higher income, had a white collar occupation, had higher educational attainment, resided in the South or Midwest, or were foreign born; it was lowest among older women and women who resided in the West or Northeast. Among the 2,259 cases and 2,154 noncases for whom we had genotyping data, the frequency of hair relaxer use increased with decreasing percentage of European ancestry. Weak correlations were observed between frequency and duration (Spearman's correlation coefficient (r) = 0.22, P < 0.001), frequency and number of burns (r = 0.23, P < 0.001), and duration and number of burns (r = 0.17, P < 0.001) (data not shown).

During 199,991 person-years of observation, 7,146 incident cases of uterine leiomyomata diagnosed by ultrasound (n = 4,630) or surgery (n = 2,516) were reported (Table 2). The multivariable incidence rate ratio for hair relaxer use relative to nonuse was 1.17 (95% confidence interval (CI): 1.06, 1.30). There was no clear pattern of risk for age at first use of hair relaxers. Multivariable incidence rate ratios increased monotonically with increasing frequency of use (P trend < 0.001), and the incidence rate ratio comparing frequency of use ≥7 times per year versus never use was 1.23 (95% CI: 1.10, 1.37). Duration of use and number of burns were also positively associated with risk of uterine leiomyomata, but there was no clear dose-response relation. There was no association with type of formulation used most often (lye vs. no-lye), although nearly 10% of women reported "don't know" to this question. The score variable, which incorporated data on frequency and duration of hair relaxer use, showed evidence of a dose-response relation. Relative to quintile 1, the multivariable incidence rate ratio increased from 1.09 (95% CI: 0.97, 1.22) for quintile 2 to 1.25 (95% CI: 1.12, 1.40) for quintile 5 (P trend < 0.001).

Among the long-term hair relaxer users (i.e., ≥10 years of use), positive trends for frequency of use were found overall (Table 3). None of the incidence rate ratios differed statistically across strata of body mass index, education, or region (Table 3), but associations appeared to be stronger among leaner women and women living in the South. Notably, there was virtually no association observed among obese women (body mass index ≥ 30). Results were also similar across age (Table 3) and different time periods of study observation (data not shown), indicating that the proportional hazards assumption was met. Results for ever use and duration of use of hair relaxers were somewhat stronger among surgical cases, but results for frequency of use did not differ according to method of diagnosis (data not shown). Finally, when the sample was confined to the 7,741 women who reported a pelvic ultrasound examination within 5 years of the 2007 questionnaire (4,134 cases and 61,948 person-years), a subgroup of women for whom undiagnosed uterine leiomyomata are less likely, results were slightly stronger than those found in the whole sample: Multivariable incidence rate ratios comparing frequency of use 3–4, 5–6, and ≥7 times/year with 1–2 times/year were 1.18 (95% CI: 0.99, 1.42), 1.19 (95% CI: 0.99, 1.41), and 1.25 (95% CI: 1.05, 1.49) (P trend = 0.02).

Table 4 presents results for selected hair relaxer variables in relation to uterine leiomyomata risk among women for whom genotyping data were available (n = 4,413). Weakly positive findings were still found for ever use, frequency of use, duration of use, number of burns, and hair relaxer score. Effect estimates were attenuated when we controlled for percentage of European ancestry.

Comments

3090D553-9492-4563-8681-AD288FA52ACE
Comments on Medscape are moderated and should be professional in tone and on topic. You must declare any conflicts of interest related to your comments and responses. Please see our Commenting Guide for further information. We reserve the right to remove posts at our sole discretion.
Post as:

processing....