Phthalates are a group of endocrine disrupting chemicals (EDCs) commonly used as plasticizers, in personal care products (PCPs), and building materials.(1) Due to their ubiquity, over 95% of the United States population has detectable urinary concentrations of phthalate metabolites.(2) Perinatal exposure to phthalates is associated with a variety of adverse health outcomes for child and parent.(1,3,4) Further, phthalates can cross the placenta, resulting in fetal exposure that is strongly correlated with parental phthalate exposure.(5) For the child, perinatal exposures have been associated with preterm birth, low birth weight, short gestational length, and adverse effects into childhood such as neurodevelopmental effects.(6,7) For the parent, exposure to phthalates during pregnancy has been associated with increased risk of gestational diabetes, as well as hypertensive disorders of pregnancy.(8–10) Still, studies have found that exposure to phthalates can be reduced via behavioral and product use change, which can have implications for health across the life course for the parent and child.(11–13) Thus, developing tools to reduce perinatal phthalate exposures have advantages for parent and child across the life course.

The emerging field of environmental health literacy (EHL) seeks to evaluate and characterize how environmental health information is conceptualized, attended, contextualized, and ultimately, used to inform behaviors. (14–16) Environmental health literacy is posited to be context and topic-specific, requiring development of environmental health literacy measurement tools that are population-specific.(14,16) To meet this goal, we previously developed and validated the Phthalate Environmental Reproductive Health Literacy (PERHL) Scale, which measures overall phthalate and reproductive health environmental health literacy.(17) Measurement via the PERHL Scale can assist in identifying individuals who may benefit from educational and behavioral environmental health literacy interventions.(17) Pregnancy is also a period of rapid behavior change in which there may be increased motivation to engage in behaviors that reduce phthalate exposures and their associated health effects.

To date, studies have not explored associations between environmental health literacy measurement tools and biomarkers of exposure, though there have been explorations of associations between environmental health literacy and exposures.(18,19) However, developing tools that demonstrate correlation with biomarkers of relevant environmental exposures will contribute to the developing environmental health literacy toolkit by providing non-invasive means of identifying individuals who may be at risk of higher environmental exposures.

The objective of this pilot study was to explore associations between scores from the PERHL Scale and subscales with urinary phthalate metabolite concentrations among participants of the Environmental Reproductive and Glucose Outcomes (ERGO) cohort.

The ERGO Study is a longitudinal pregnancy cohort focused on assessing the role of environmental exposures on reproductive and cardiometabolic health outcomes across pregnancy and the postpartum period. The details of ERGO study activities have been described elsewhere. (20) Briefly, ERGO recruited pregnant participants (n=653 unique pregnancies contributed by n=647 participants) without pre-existing diabetes from December 2016 to November 2020 from Boston, MA-area hospitals’ obstetrics clinics during routine prenatal visits.(20) Participant eligibility criteria included being at least 18 years of age, being less than 15 weeks of gestation at enrollment, and planning to receive prenatal care and deliver at a participating hospital. The ERGO cohort also includes participants enrolled in the Study of Pregnancy Regulation of INsulin and Glucose (SPRING) who contributed harmonized data during pregnancy and early postpartum but did not participate in the present study(21). For this pilot study, all ERGO participants who were at least one year postpartum (n=424) were invited to complete the PERHL survey. One hundred seventeen ERGO participants consented to complete the PERHL scale. Participants, all of whom had delivered within the last 5 years, completed the PERHL Scale remotely via REDCap between March 2021 and June 2022. Prior to completing the PERHL survey, ERGO participants did not engage in conversations about phthalates as part of their participation in the ERGO study. All participants were asked if they would like to opt-in to provide optional remote spot urine samples. Of the 117 participants who completed the PERHL Scale, 42 participants concurrently provided an optional remote spot urine sample. Participants collected remote urine samples using polypropylene sample containers shipped to them. Participants froze their samples until time of shipment. All samples were shipped in insulated containers with ice packs via overnight shipping. All urine samples were received in appropriate condition for analysis.(22)

The PERHL Scale is a validated tool to assess environmental health literacy related to phthalate exposures during pregnancy. Overall scores range from 6 to 30, representing low to high environmental health literacy, respectively. In addition to the overall PERHL Scale score, there are 6 subscale scores capturing latent factors related to environmental health literacy: 1) ‘Awareness of Phthalate Reproductive Health Impacts’ (8 questions that address the extent to which respondents believe that exposure to phthalates may be associated with reproductive health impacts), 2) ‘Scientific Uncertainty’ (4 questions that address the extent to which respondents feel uncertain about the adverse effects of phthalates on environmental and human health due to various possible sources of uncertainty), 3) ‘Protective Behavior/Risk Control’ (5 questions addressing whether respondents try to minimize their use of certain personal care and household products due to concern of phthalate exposures, and to what degree they believe reducing their phthalate exposures is beneficial to their health), 4) ‘Regulatory Interest’ (3 questions addressing respondents’ belief that phthalates should be regulated based on their potential impacts to the environment and human health), 5) ‘Awareness of Phthalate Exposure Routes’ (3 questions addressing respondents’ understanding of sources of phthalates into the environment, and possible exposure pathways that may affect reproductive health outcomes), and 6) ‘General Phthalate Knowledge’ (3 questions very broadly addressing respondents’ belief that phthalates are in personal care products, and that people can be exposed to them by using personal care products). PERHL questions and the subscales they belong to are provided in Table 1.

Urine samples were analyzed for concentrations of 14 phthalate and 2 replacement phthalate metabolites, described in Table 2(23) Samples were analyzed using previously described methods.(24) Urinary concentrations were specific-gravity (SG)-corrected to account for urinary dilution.(25) We created three phthalate summary measures for our model outcomes by calculating molar sums of SG-corrected metabolite concentrations representing exposure to: 1) parent phthalate, DEHP: ∑DEHP=mEHP278+mEHHP294+mEOHP292+mECPP308, 2) personal care product (PCP)-associated phthalates: ∑PCP=mBP222+mEP194+miBP222, and 3) parent butyl-phthalates: ∑butyl=mBP222+mBzP190+miBP222. Summary measure concentrations were log₂-transformed due to skewed distributions and modeled as continuous outcomes.

We used linear regression models to estimate the associations of continuous PERHL Scale and subscale scores with concentrations of each phthalate summary measure, controlling for educational attainment (college degree or higher vs. no college degree), age (years), and race and ethnicity (non-Hispanic White vs. non-White) of the participant. The percent change in molar sums of phthalate metabolite concentrations were calculated from model estimates using the following formula: %change=(2β-1)*100. We additionally explored whether there were associations between education with the three summary phthalate measures.

Due to the small sample size in this pilot study, an alpha value of 0.1 was used as the significance threshold to account for the exploratory nature of the analysis and to minimize risk of type II errors. All analyses were performed in R Software version 4.2.1.

Participant demographics are displayed in Table 3 for the present study population. Participants were mainly non-Hispanic White (71.8%), had a mean age of 33.1 years, and had a college degree or beyond (86.3%). This differed from the sociodemographic characteristics of participants in the full ERGO cohort, which had a lower percentage of non-Hispanic White participants (58%), and lower percentage obtaining a college degree or beyond (79%). The mean age of the ERGO cohort was 32.9 years. Participants who opted in to provide urine samples (n=42) were mainly non-Hispanic White (73.8%), had a mean age of 34.0 years, and had a college degree or beyond (88.1%).

Table 4 displays the estimated percent difference in urinary concentrations of phthalate summary measures per 1-point increase in PERHL Scale & subscale scores. Higher scores on the PEHRL Scale and subscales were generally associated with lower ΣDEHP, Σbutyl, and ΣPCP metabolite concentrations. Higher PERHL Scale scores were significantly associated with lower ΣDEHP metabolite concentrations. Specifically, a one-point increase on the overall PERHL Scale was associated with a −7.2% (95% CI: −15.0 1.3) decrease in ΣDEHP metabolite concentrations. Higher ‘Protective Behavior/Risk Control’ and ‘Regulatory Interest’ subscale scores were significantly associated with lower ΣDEHP and Σbutyl phthalate metabolite concentrations after adjustment for potential confounders. A one-point increase in the ‘Protective Behavior/Risk Control’ subscale score was significantly associated with a −30.3% (95% CI: −50.1, −2.6) decrease in ΣDEHP, and a −30.6% (95% CI: −51.5, −0.63) decrease in Σbutyl metabolite concentrations. Similarly, a one-point increase in the ‘Regulatory Interest’ subscale score was significantly associated with a −33.9% (95% CI: −53.5, −6.0) decrease in ΣDEHP, and a −28.4% (95% CI: −51.45, 5.4) decrease in Σbutyl metabolite concentrations. We found no evidence of an association between educational attainment and PERHL score or summary phthalate metabolite measures.

We found that higher scores on the validated PERHL Scale and subscales were overall associated with decreased phthalate metabolite concentrations. Scant literature has explored the relationship between environmental health literacy and biomarkers of exposure. One randomized controlled trial (n = 63) found that an environmental health literacy-based intervention was successful in improving indoor air quality, significantly reducing indoor PM2.5 (z = 56.39, p<0.1) and volatile organic compounds (z=10.11, p<0.01) in the treatment group and resulted in significant differences between the treatment and control groups in the level of cotinine (z=−1.07, p = 0.15) and trans,trans-muconic acid (tt-MA) (z = −1.44, p=0.31), bioindicators of exposure to the volatile exposures of interest.(26) Environmental health literacy was assessed via an adapted version of Lichtveld et al.’s environmental health literacy scale, which consists of a series of questions on air, water, and food as well as general environmental health literacy. All questions are related to environmental health knowledge, attitudes, or behaviors via Likert 5-point scale questions and summed, with higher scores indicating higher environmental health literacy. However, although environmental health literacy was assessed as a part of this study, the direct associations between environmental health literacy score and biomarkers of exposure were not investigated. The results of this study suggest that educational interventions can be effective at reducing indoor air pollutants and, further, reducing biomarkers of exposure.

Overall, the associations seen between higher PERHL scores and lower biomarkers of phthalate exposure support the hypothesis that environmental health literacy may be a protective factor against harmful environmental exposures that are amenable to behavioral interventions. Previous conceptualizations of environmental health literacy have hypothesized that increasing awareness or knowledge of a given environmental exposure leads to protective behavior, and subsequently, reduced exposures. However, the associations between higher scores on the ‘Awareness of Phthalate Exposure Routes’ and ‘General Phthalate Knowledge’ subscales suggest that assessment of environmental health literacy via measurement of knowledge on a topic, alone, may not be sufficient to capture the full scope of environmental health literacy dimensions. Future efforts to develop environmental health literacy measurement tools should explore latent factors beyond topical knowledge that contribute to environmental health literacy.

Additionally, we found higher ‘Protective Behavior/Risk Control’ scores were associated with significant decreases in ΣDEHP and in Σbutyl metabolite concentrations, but not ΣPCP metabolite concentrations. We hypothesize that this may be due to broad awareness and avoidance of phthalates in certain plastic, food packaging, and household products, but an unawareness of phthalates in personal care products. Additional research should explore the driver of the difference in associations across phthalate exposures sources.

Although the theory of environmental health literacy continues to evolve, it has been suggested that environmental health literacy of an individual is likely to be topic and context-specific. Levels of environmental health literacy will vary based on personal experiences and background relevant to a particular environmental exposure in various contexts.(27) Building tools to assess individual environmental health literacy related to specific contexts and environmental exposures is critical to ensure that environmental health literacy can be measured to identify opportunities for intervention to build environmental health literacy and facilitate actions aimed at the reduction of exposures, where possible. This work seeks to fill the gap by assessing the association between a validated environmental health literacy scale and biomarkers of phthalate exposure. Previous work has found that increasing environmental health literacy related to endocrine disrupting chemicals can facilitate exposure-reduction behaviors.(28)Thus, developing non-invasive tools to assess environmental health literacy that are associated with physical indicators of exposures to the chemical of concern can provide a non-invasive means of identifying individuals who may benefit from educational interventions.(28)

This study has several limitations. This pilot study has a small sample size, with only 42 participants opting to provide remote urine samples. This limits the generalizability of these findings to broader populations. Additional exploration of the associations between PERHL scale scores and biomarkers of phthalate exposures should be pursued in larger samples with more diverse participant characteristics.

Overall, associating environmental health literacy scale scores with relevant biomarkers of exposure provides a means of validation of the environmental health literacy tool, and can contribute to future environmental health interventions by providing a quick, non-invasive tool to identify individuals who may be most at risk for lower environmental health literacy and higher phthalate exposure.