In this study of PCP use among adults affiliated with a large U.S. university, we observed that for certain types of PCP, specifically hair products, beauty products, skincare products, and perfumes and colognes, the average hazard scores of products used differed by sociodemographic factors including age, gender, income, and marital status. Additionally, we observed higher risk of recent use of hair products with high hazard ratings among non-Hispanic Black women, middle aged participants, and females. Furthermore, we observed that participants’ perceptions around the health impacts of PCPs and the regulation of these products by government agencies were associated with the hazards scores of the products they used, with participants who expressed concern about the safety and regulation of PCPs generally using products with lower average hazard scores. Importantly, our results also suggested that behaviors such as use of healthy product apps and websites and reading ingredient labels were associated with use of products with lower hazard scores. By linking the specific products used by participants to data on the hazard profiles of those products, these results represent an important extension of the prior research which has more generally characterized PCP use by type or category of product.

Prior literature has documented sociodemographic differences in the number and type of PCPs used as well as the frequency of use. For example, in a prior analysis of this same university-based sample, we previously showed that Asian and Pacific Islander females used fewer hair products and more intimate care and menstrual products than non-Hispanic White females, while non-Hispanic Black females used more hair products, perfumes, oral care, and intimate care and menstrual products than non-Hispanic White females [17]. In a community-based study in California, Latina women tended to be heavier makeup users (in terms of product number and frequency) than other groups, while use of fragrances and certain types of hair products (including those linked to reproductive cancers and preterm birth) was most common among Black women [29]. Other research in women from New York City reported that African-American and African-Caribbean women were more likely to use a variety of hair products compared to White women, including products known to contain extracts of placenta (a hormonally active tissue) or other endocrine disruptors [30]. While the average hazard scores of the specific PCPs used by participants in the last 24 h did not differ by race and ethnicity, when we looked at the use of products with high hazard scores specifically, it emerged that non-Hispanic Black females were significantly more likely than non-Hispanic White females to have recently used a PCP with a high hazard score, a result driven by the use of hair products with high hazard scores. A number of prior studies have demonstrated persistent racial disparities in exposure to chemicals commonly found in PCPs (such as phthalates) and differences in PCP use have been proposed as a primary contributor to the unequal exposure burden observed among some race and ethnicity groups [30,31,32,33]. Our prior work in this cohort further showed that Black females used more haircare products and perfumes than other racial and ethnic groups [17]. The use of more hair products as well as hair products with higher hazard ratings among Black females may be an important contributor to health disparities and warrants further investigation including evaluation of specific patterns of use of products such as hair relaxers and oils as well as drawing distinctions between rinse-off and leave-on products which we were not able to do here.

We also observed that a disproportionately higher number of hair products used by diverse participants were not found in the Skin Deep® Database and thus their hazard profiles are unknown. This raises concerns about equitable public access to safety information, which is particularly important considering previous research linking hair products containing EDCs to breast cancer risk in African American women [34]. Of note, EWG also recognizes similar concerns and has prioritized the addition of products marketed to Black females in recent Skin Deep® updates with 1600 products being added in early 2024 alone (personal communication). Additionally, our findings highlight the need for further research in this area with more diverse and inclusive samples. Prior research focusing on males found significant correlations between PCP use and increased urinary concentrations of EDCs [35], as well as an increased risk of prostate cancer associated with hair dye use [9]. In our prior research in this sample, we observed that males used fewer products than females [17] and we add to those results here by showing that average hazard scores of products used by male respondents were similar to those used by females, suggesting potential risks to males as well. At the same time, for every product category we studied, males had a lower risk of having used a product with a high hazard score in the last 24–48 h compared to females, again showing that the greatest risks associated with PCP use may be to females. The existing literature and our current findings emphasize the importance of studies with diverse samples of both males and females to better understand the differences in PCPs used and their associated hazards. Further, inclusion of individuals belonging to sexual and gender minority populations including non-binary and transgender females, who are also frequent users of PCPs, must also be prioritized.

Some differences in PCP hazard scores were noted by participant age as well. Of particular note, the perfumes and fragrances used by younger participants have higher hazard scores than those used by the older groups. This is of concern given that of all product types, perfumes and fragrances tend to have the highest hazard scores due to the widespread inclusion of known endocrine disruptors like phthalates in fragrance formulations. These concerns are further heightened by recent media reports suggesting rising interest in and use of fragrances and other scented PCPs among even tweens and teenagers [36, 37]. At the same time, in our study, overall, average hazard PCP scores were highest among participants aged 40–59, a result that appeared to be mostly driven by use of beauty products. This may reflect conscious or unconscious societal pressures to adhere to beauty standards that favor youth leading to the use of products containing ingredients that include anti-aging or other “active” ingredients that may potentially have higher hazard scores. Because prior literature has largely focused on race and ethnicity-based disparities in PCP use or on limited age-bands (e.g., college students or reproductive-age women), little is known about how age influences product choice and use. While more work is needed to understand these results, it is important to note that like gestation, puberty, and pregnancy, the perimenopause is increasingly viewed as another life stage during which hormonal axes are in flux and may be more vulnerable to disruption by chemical exposures [38, 39].

Participants’ perceptions regarding the safety of PCPs and the effectiveness of government regulations were associated with the hazard scores of the products they used. Generally, participants who expressed greater concern about product safety or the effectiveness of regulations tended to use products with lower average hazard scores. In a prior analysis in this sample, we demonstrated that participants who agreed that PCPs could have harmful health impacts were more likely to use healthy product smart phone apps and websites or to read ingredient labels [28]. Collectively, these results suggest that educating consumers around potential hazardous exposures in PCPs and current U.S. policies around their regulation could help to inform consumer choices.

These observations raise important issues regarding environmental health literacy- namely how people conceptualize and contextualize environmental health information to shape their behaviors [40, 41]. Specific to PCPs, for example, some consumers may assume that the U.S. government closely regulates the products on the market, testing them for their safety, when in reality, there is very little safety testing beyond color additives [14, 42]. Choosing safer products therefore requires some environmental health literacy to navigate the vast PCP market. At present, research on environmental health literacy specific to PCPs is limited, although a recent study focused on phthalates, a common ingredient in PCPs, indicated race- and ethnicity-based disparities in knowledge around sources of exposure and health impacts [43]. In that study, participants who demonstrated greater knowledge of phthalates were also more likely to self-report avoiding PCPs containing phthalates. Together, results of that research and the current study indicate that improving environmental health literacy around chemicals in PCPs, particularly in minoritized communities, may shift PCP use behaviors and in so doing, be an effective approach to reducing known disparities in chemical exposures.

Additional research is needed to determine the extent to which use of “safer” products (i.e., those with lower hazard scores) translates to body burden of chemicals commonly found in PCPs, such as phthalates and phenols. In one study of teenage girls, replacing the usual PCPs they used for “cleaner” alternatives for 3 days resulted in a roughly 25–50% reduction in certain urinary phthalate and paraben metabolites [44]. Importantly, after that intervention, 71% of participants said they would subsequently buy products without phthalates, parabens, triclosan, and oxybenzone. Similar studies are needed to understand whether such reductions in body burden of EDCs can be replicated in real-life settings when consumers are empowered with knowledge and tools to independently choose safer products rather than being provided with pre-selected PCPs.

To that end, we observed differences in PCP hazard scores of products used by participants in relation to their self-reported purchasing behaviors. Participants who reported using healthy or “clean” product apps, regularly reading ingredient labels, and looking for labels indicating natural, non-toxic, or eco-friendly ingredients used PCPs with lower average hazard scores, suggesting that apps and labels may be effective in helping consumers to select safer products. At the same time, this is an imperfect solution as some prior work has noted that certain ingredients are often undisclosed on product labels (e.g., cyclosiloxanes) or are “hidden” under proprietary terms like “fragrance” [12, 45]. It also puts the burden on the consumer to research products and navigate complicated ingredient lists, which may be increasingly difficult with the recent rise in “clean washing”, or the false advertisement of products to appear safer to consumers [46, 47].

Many of the associations observed were small in magnitude, for instance differences of less than 1 in average hazard scores between those who use healthy product apps and websites versus those who do not. At present, the clinical relevance of these differences in terms of individual participants’ health is not clear, however, on a population level, even small differences may increase the burden of disease. In addition, the consistent daily use of PCPs over years or even decades is such that even minor differences in chemical exposures at one time point may add up to considerably higher cumulative lifetime exposure compared to those who use products with slightly lower hazard scores. Lastly, our observation that some demographic groups had higher risk of recent use of a PCP with high hazard ratings (e.g. Black women and hair products) suggests priorities for future work linking product use to health outcomes.

Structural barriers may also reduce access to safer products for some populations. For example, in a study of eight Boston neighborhoods, retail stores in lower income neighborhoods (particularly those in communities of color) carried a higher percentage of hair products that were rated as highly hazardous in the Skin Deep® database compared to retail stores in wealthier neighborhoods [48]. That study also noted that higher priced hair products were less likely to have average to high hazard scores (compared to lower priced products), however stores in lower SES neighborhoods were more likely to carry lower priced hair products. This suggests that in addition to potential knowledge barriers, accessing safer PCPs can be a challenge and ultimately, perpetuate disparities in EDC exposures. Notably, while we observed some racial and ethnic differences in the hazards of the particular PCPs used by participants, it is also possible that disparities are less pronounced among communities of higher average socioeconomic status (such as in an academic setting like the one our sample represents) who have access to internet retailers and reliable transportation to access multiple brick and mortar retailers. In a general population, even stronger racial and ethnic differences in PCP hazards might be observed.

We note multiple strengths of the current work. In contrast to most prior research in this area which queried number and types of products used, we collected comprehensive data on the specific products used including brand, formulation, and color. This granularity allowed us to link the specific products used to the Skin Deep® database, which provided a quantitative measure of product hazards. Additionally, our sample was more diverse in age and gender than prior studies, most of which have focused on reproductive age females. While that group is clearly one of interest given concerns about the reproductive and developmental toxicity of chemicals found in PCPs (e.g., phthalates, phenols), increasingly, evidence suggests that health risks associated with chemical exposures in PCPs may be relevant to all gender identities, with multiple critical windows of susceptibility across the lifespan. Additional work is needed to understand PCP use in males, older females (who tended to use some types of products with higher average hazard scores (e.g., beauty products)), and non-binary and transgender individuals. Finally, we included items on perceptions of harm associated with PCP use, an area that has received relatively little attention to date, but is important in terms of understanding consumer behaviors and informing future intervention strategies.

At the same time, there are several limitations of this research that should be considered. We did not collect details on the use of the PCPs such as the amount of product applied, how many times it was used during the prior 24–48 h, and whether the product was left on or rinsed off, all of which may impact chemical absorption and potential toxicological effects [49]. Leave on products, in particular, may have higher levels of certain phthalates and parabens [3]. Importantly, many products did not match to the Skin Deep® database resulting in missing data on hazard scores. Because products enter the database through reporting by brands, retailers, or consumers, smaller brands that are carried by fewer retailers and used by fewer consumers may be less likely to be added to the database. Among the products that commonly did not match to the Skin Deep® database were intimate care products (e.g., maxi pads, tampons, wipes, douches), oral care products (e.g. flosses, mouthwashes), direct to consumer brands (e.g., Avon, Mary Kay), and certain store brands (e.g. Bath and Body Works, generic brands like CVS Health and Up & Up). By necessity, we pursued a complete case approach and results could have been biased in either direction as a result. With the ongoing expansion of coverage of products in Skin Deep®, particularly those used by people of color, missingness may be less of an issue in future studies. This expansion is essential not just for future research, but to help all consumers make more informed decisions about their PCPs. In addition, PCP formulations change over time and we are unable to verify that the PCPs used by participants are the exact same formulations (e.g. same ingredient profile) as when those products were entered into Skin Deep®. While other databases exist to evaluate risks associated with certain PCPs, we opted to use data from Skin Deep® as it is particularly widely-used and well-known, and its hazard scores were developed based on 17 evidence-based criteria. We additionally note that our sample, based in an academic community, was relatively homogeneous and tended to be well educated with higher income, which may limit generalizability.

In summary, in this study of university-affiliated U.S. adults, we observed that the chemical hazards encountered through PCPs varied by consumers’ sociodemographic characteristics. We additionally observed that people’s perceptions around PCP safety and regulation were associated with the hazards of the products they used, suggesting the importance of education and environmental health literacy in shaping consumer choices. Finally, our results suggested that use of smartphone apps and websites promoting cleaner, safer beauty choices may be an effective way to help consumers choose PCPs that may pose fewer health hazards. These results suggest that individual-level behaviors may help to protect consumers from potentially harmful chemical exposures in PCPs in the absence of much-needed updates to U.S. regulations on the chemical ingredients found in PCPs.