In this large cohort of reproductive-aged Black individuals, we used LCA to identify subgroups of participants with distinct PCP use patterns that differed by SES. When compared to each other, the latent classes ranged in frequency of PCP use which included a class of lower overall PCP use, a class with lower overall PCP use except for higher use of nailcare products, a class with lower overall PCP use except for higher use of skin creams, a class with overall moderate PCP use, a class with higher use of makeup, haircare, and skin creams, and a class with higher overall PCP use. The starting point of this analysis was to establish groupings of participants based on PCP use and then examine the characteristics of each group. Participants who reported using PCPs more frequently were more likely to have higher SES, including higher educational attainment, higher income, and a higher likelihood of working full-time. The results of this study support and build on previous findings that Black women with higher SES were more likely to use multiple hair products [16]. This work demonstrates the importance of considering PCP exposures concurrently with other socio-demographic characteristics, lifestyle factors, and health behaviors when modeling health risks.

Prior studies have evaluated racial/ethnic differences in PCP use and provide important evidence that Black women and children are more likely to use PCPs that may contain more harmful ingredients than products used by White women and children [1, 5,6,7, 17, 24, 25]. For example, Black women have reported more frequent use of nailcare products compared to White women [7, 25,26,27,–28], and use of these products has been linked to higher urinary concentrations of mono-n-butyl phthalate and mono-ethyl phthalate [29,30,31,32,33]. Some studies have also reported more frequent use of skin creams among Black women compared to White women [25, 27]; use of these products has been associated with higher urinary concentrations of parabens, phthalates, and phenols [32,33,34,35,36,37]. Parabens, phthalates, and phenols have also been linked to makeup use [32,33,34,35, 37, 38], although studies examining racial/ethnic differences in the use of makeup have reported mixed results [7, 17, 25, 27]. Black women are more likely than White women to use hair products that contain placenta (a potential source of estrogen hormones), parabens, and phthalates [3,4,5, 9, 13, 14]. Black women are also more likely than White women to use scented vaginal products (e.g., douches and sprays), and perfume [1, 6,7,8, 17, 25]. Some of these products have been found to have higher concentrations of hormonally active chemicals (e.g., parabens, phthalates, per- and polyfluoroalkyl substances (PFAS)) than products more commonly used by White women [3,4,5, 9]. As described in Zota et al. [9], increased use of scented vaginal douches and other fragranced intimate care products may be driven by odor discrimination--racial discrimination based on a long-standing societal myth of odors among Black women. This has been perpetuated by targeted marketing of vaginal and intimate care products towards Black women [9]. We note that only 25–30% of the SELF cohort reported regular use of vaginal products such as powder, douche, or lubricant over the previous 12 months, and use of fragrance-free panty liners, tampons, and pads was common. However, use of vaginal products was more common in this cohort when participants were younger [39]. Additional research that examines the use of both scented and fragrance-free products in conjunction with different types of discrimination (e.g., based on hairstyles or odor) is warranted.

Motivations driving PCP use decisions were not assessed in SELF, and evidence supporting SES-related differences in the frequency of PCP use, especially among Black women, is limited. More frequent use of a combination of PCPs among Black women with higher SES (e.g., use of makeup, hair products, and skin creams) may be related to lifestyle differences and/or long-standing pressures on Black women to maintain high perceived beauty standards when in professional and public settings [9, 16, 40,41,42,43]. These perceptions stem from institutionalized racism that historically embraces European beauty standards [9, 10]. For example, until 2014 the US Army banned certain hairstyles worn primarily by Black women [41]. This type of racial discrimination, reinforced by targeted marketing to Black women that promotes the use of products to lighten skin, straighten hair, or use scented vaginal products, can lead to internalized racism that influences an individual’s PCP use [9, 10, 25].

Most participants in the SELF cohort did not report avoiding products with parabens, Bisphenol-A, and triclosan. Despite data indicating that Black women are aware of toxic chemicals in PCPs, other factors such as higher cost of “clean” products, neighborhood availability, and lack of adequate labeling can preclude cleaner choices [3, 16, 24, 25, 43,44,45]. Despite the evidence of adverse health effects, PCPs remain poorly regulated with fragmented government oversight. Federal law currently does not require the disclosure of proprietary ingredients, such as fragrance chemicals, to consumers or regulatory agencies. However, some states, such as California, have introduced laws that remove trade secret protections and require companies to disclose chemicals in personal care and beauty products [46]. A federal bill called the Cosmetic Fragrance and Flavor Ingredient Right to Know Act of 2023–2024 [47] has been introduced to Congress and, if passed, would require companies to publicly disclose a full list of fragrance and flavor ingredients in their products on product labels and websites.

Our analysis addresses previous gaps in this literature. First, although it is important to understand racial/ethnic differences in PCP use, our study examined differences in PCP use among a cohort of Black individuals and identified related socio-economic characteristics, health behaviors, clinical characteristics, and behaviors related to product use. This information helps to build a more comprehensive understanding of how social factors may influence PCP use. This could be used to inform future research that examines how environmental factors may contribute to commonly observed health disparities and how these factors influence product availability, accessibility, and patterns of use. Second, the literature on PCPs has generally focused on single categories of products. To the best of our knowledge, this is the first study to utilize a mixture approach that captures real-world usage patterns across several categories of PCPs. Third, we used LCA to capture complex patterns of PCP use and identify distinct groups of participants with similar product use profiles. Examining the product use probabilities across groups revealed substantial differences in the use of certain products (e.g., nail products and skin cream products) that have been identified as being used more by Black women compared to White women. For example, when compared to the Lower Overall class, the Higher Nailcare class is distinguished by higher/more frequent use of nail products, and the Higher Skincare class is distinguished by higher/more frequent use of skin creams. LCA is a mixture model that accounts for correlations between PCPs such that the PCPs within classes are related but classes are independent of each other. Therefore, with control for potential confounders, these latent classes can be used as exposure variables to investigate associations between PCP use and other outcomes without the multiple-testing problems that arise when associations between individual products and other outcomes are examined. Understanding different patterns of PCP use across multiple PCP categories provides insight into whether certain patterns are associated with other risk factors for hormone-mediated health outcomes such as earlier age of menarche, breast and uterine cancer, uterine fibroids, and cardiometabolic health. Finally, no previous studies have examined how PCP patterns across different product categories differ by SES and other lifestyle factors among Black women.

Except for Gaston et al. [16], it is difficult to compare our results showing SES-related differences in PCP use among Black women to other study populations. Only a couple of studies have examined both socio-economic and racial/ethnic differences in PCP use [17, 18, 24]. However, likely due to small sample sizes, these studies did not report SES differences in PCP use by race/ethnicity or among Black women. Among studies that have evaluated socio-economic differences in PCP use and PCP-related EDC concentrations [16,17,18, 48, 49], there have been conflicting results in patterns of use. The findings in the current study are consistent with previous work of the same cohort that found women with higher SES were more likely to use multiple hair products [16]. Also consistent with the current study, a study of pregnant women in Ottawa, Canada reported that, compared to women with lower incomes, women with higher incomes were more likely to use more PCPs [49]. However, information on the race/ethnicity of the participants was not provided. A study of usage patterns of PCPs in California households found that women with a college education were more likely to use sunscreen, insect repellent, facial cleanser, and professional application of nail products and hair dye [18]. This study, which was majority White and <3% African American, also reported that compared to White women, African American women were more likely to have their nails professionally treated, use leave-in hair treatments, deodorant, facial cleanser, and bath gel. In several studies, compared to women with lower SES, women with higher SES had higher urinary concentrations of benzophenone-3 and triclosan [2, 48, 50, 51], chemicals often found in sunscreen, antibacterial soaps, body washes, deodorants, skin cleansers, and fluoride toothpaste (FDA.gov). In contrast to the findings in our study, a study of pregnant women (<10% non-Hispanic Black) living in Boston, Massachusetts reported that women with lower SES reported significantly higher product use, including bar soap, perfume, and nail polish [17].

This study also has several limitations. Study criteria required that participants have an intact uterus at the time of enrollment. However, participants were not queried about their gender, a socially constructed term that encompasses identity, expression, and social position with many categories beyond the binary of female and male [52, 53]. In contrast to examining the use of individual products, LCA creates manageable categorical data elements that summarize complex patterns of PCP use. However, classes can be difficult to interpret. Labels were assigned to different classes based on our observation and interpretation of the probability-based weights for class membership, and there is some subjectivity in choosing the shorthand label descriptors for different classes. Also, the categories identified with LCA are specific to the SELF-study population and may not be generalizable to other populations. Future studies examining PCP use patterns in other cohorts will help determine how PCP patterns vary across other study populations. Due to the nature of self-reported data, it is possible that PCP use in this study was misclassified. The LCA approach assigns individuals to classes based on their probability of class membership which may result in non-differential misclassification. Finally, we were unable to capture the actual products used, the chemical composition of the products, or variability in the intensity of use (e.g., heaviness of application). Although future studies may want to capture this level of information, specific product formulations change frequently likely due to changes in the availability and cost of ingredients.