A pilot biomonitoring study of cumulative phthalates exposure among Vietnamese American nail salon workers

Study population and sample collection

Our study population comprised a subset of participants sampled in a 2011 pilot intervention study designed to educate nail salon workers and owners about how to reduce occupational exposures to various hazardous chemicals.²⁶ That study included 26 Vietnamese American workers at eight San Francisco Bay Area nail salons and measured each worker’s personal breathing zone air for volatile organic compounds such as toluene and methyl methacrylate. During air monitoring sessions, each participant completed questionnaires on characteristics such as age, the number of other workers and customers on the day of sampling, types of nail care services performed, and individual protective behaviors. Prior to educating workers about how to reduce their exposures, post-shift urine samples were collected from 17 of these workers at six nail salons during busier days of the week (Thursday, Friday, and Saturday) and warmer months of the year (June and July). Written consent was obtained in Vietnamese, and human subjects institutional review and approval from the Cancer Prevention Institute of California was received on May 11, 2011 (Protocol # 2010-013). Urine samples were collected in four-ounce glass containers, shipped on ice, and stored at -20 °C prior to analysis. To reduce potential effects of freeze-thaw cycles, samples were thawed and aliquoted (2–4 mL) in pre-cleaned amber vials. Water laboratory blanks were also prepared in the same type vials and frozen.

Laboratory analysis

All urinary phthalate analyses were conducted at the Department of Toxic Substances Control (DTSC) in Berkeley, California. We measured four phthalate metabolites in this pilot study because our modest pilot study budget warranted a targeted approach to analyte selection, and our collaborators at DTSC sought to develop methods to measure urinary analytes for priority chemicals that would inform DTSC’s Safer Consumer Products program. We selected analytes based on earlier studies which suggested that mono-n-butyl-, mono-isobutyl-, mono(2-ethylhexyl)-, and monoethyl phthalates (MnBP, MiBP, MEHP, and MEP, respectively) may be associated with working in a nail salon.^24,25 To analyze urine samples in the lab, we adapted and validated the Silva et al. (2004) method used by the Centers for Disease Control (CDC) to measure urinary MnBP, MiBP, MEHP, and MEP concentrations using liquid chromatography-tandem mass spectrometry (LC-MS/MS).²⁷ Isotopically-labeled internal standards (IS) were used for quantitation (MnBP IS was used for MiBP) and Agilent Nexus ABS Bond Elut cartridges for solid-phase extraction (SPE) clean-up. Samples were analyzed using a Varian 320MS LC-MS/MS with a Phenomenex Gemini C6-Phenyl column (100mm x 2.00mm, 3µm). Baseline peak resolution was achieved for all metabolites, and no significant laboratory background was detected. Relative recoveries of quality control standards spiked into the urine before SPE ranged between 80–95%. Inter- and intra-day reproducibility and precision (coefficient of variation < 15%) were demonstrated in addition to accuracy (< 15% difference between expected and observed concentrations). The limit of detection (LOD) was defined as the lowest calibration standard providing a signal-to-noise ratio > 6. Once the method was validated, all 17 nail salon worker urine samples were analyzed and aliquots were submitted to the clinical laboratory at San Francisco General Hospital for creatinine measurements.

Statistical analysis

Nail salon worker data were compared to 2011–12 CDC-administered National Health and Nutritional Examination Survey (NHANES), a nationally representative questionnaire and physical survey of the U.S. civilian, non-institutionalized population. As part of the NHANES medical examination, spot urine samples are collected and shipped on dry ice to CDC’s National Center for Environmental Health for analysis.²⁸ Demographic and laboratory files were downloaded from the NHANES website in May 2015 (http://www.cdc.gov/nchs/nhanes.htm). Out of 9756 NHANES participants, 2489 were sampled for >10 urinary phthalate metabolites, quantified using high performance liquid chromatography coupled with tandem mass spectrometry (HPLC-MS/MS).²⁹ Primary comparison populations were restricted to pooled (men and women) and female-only NHANES Asian Americans 20–59 years old (n = 203 and n = 97, respectively). A secondary comparison population included all NHANES participants aged 20– 59 years (N = 1175).

The nail salon urine sample and NHANES data were analyzed using SAS software, Version 9.4 (SAS Institute Inc., Cary, NC, USA). Phthalate metabolite concentrations below the LOD were imputed with LOD divided by √2. To correct for urine dilution, metabolite concentration (μg/L) was divided by urinary creatinine concentration (g/L), which is often used as a surrogate for urine dilution.^30,31 A potency weighted cumulative sum of phthalates daily intake (Σandrogen-disruptor, μg/kg/day) was constructed as described elsewhere.^32,33 Briefly, daily intakes of DnBP, DiBP, DEHP, and DEP were weighted by relative potency factors and summed using the following equation: Σandrogen-disruptor = (DnBP x 1.0) + (DiBP x 0.24) + (DEHP x 0.61) + (DEP x 0.024). Daily intake for each phthalate was back-calculated from creatinine-corrected urinary metabolite concentrations (μg/g) using creatinine excretion rates and fractional excretion values of metabolites relative to parent compounds. Because MEHP was the only DEHP metabolite included in this analysis, an F_UE value of 0.062 was used for DEHP estimation.³⁴ Average creatinine excretion rates of 23 mg/kg/day and 18 mg/kg/day were applied to men and women, respectively.³⁵ Concentrations of Σandrogen-disruptor and urinary metabolites were log-transformed prior to statistical testing to account for non-normal distributions, and statistical and marginal significance were defined at p < 0.05 and p < 0.10 for two-sided tests, respectively.

Descriptive statistics were calculated for worker and salon characteristics, such as average hours worked per week, number of services performed on the day of sampling, and salon volume in cubic meters of air (m³). We calculated detection frequencies (% > LOD) of measured urinary metabolites and performed univariate statistics on creatinine-corrected concentrations, non-creatinine-corrected concentrations, and our cumulative exposure metric (Σandrogen-disruptor). Univariate statistics included the arithmetic mean, geometric mean (GM), geometric standard deviation (GSD) or geometric standard error (GSE), range, and 25^th, 50^th, 75^th, and 95^th percentiles, and were calculated for all 17 nail salon workers and female-only workers (n = 15). These were also calculated for NHANES comparison populations, accounting for sample population weights and the complex NHANES survey design according to analytical guidelines.²⁸

Phthalate concentrations were compared between nail salon workers and NHANES 2011–12 participants using a parametric t-test for MnBP, MEHP, MEP, and the cumulative exposure metric (Σandrogen-disruptor). For statistical hypothesis testing, NHANES values were assumed to be the true population values. The Wilcoxon signed rank non-parametric t-test was used for the MiBP comparison, since several MiBP outliers were observed among nail salon workers. This analysis was performed separately for pooled (men and women) and female-only Asian Americans aged 20–59 years old, in addition to all race/ethnicities of the same age range.

Because urine dilution correction and daily intake estimation may both introduce uncertainty and variability into back-calculated biological exposure metrics,^30,31,36,37 we conducted a supplementary analysis to evaluate whether alternate approaches impacted results. Two additional cumulative exposure metrics were calculated by first applying RPF weights directly to non-creatinine-corrected concentrations measured in urine (μg/L), and then by applying RPF weights to creatinine-corrected urinary metabolite concentrations (μg/g).³² Compared to the original metric (Σandrogen-disruptor), which both corrected for urine dilution and estimated daily intake, the first supplementary metric provided a comparison to an absence of both potential sources of uncertainty/variability, while the second provided a comparison to the absence of daily intake estimation only.

To facilitate interpretation of results, we additionally compiled a summary of urinary phthalate metabolite concentrations measured in relevant biomonitoring and health impact studies. These included two previous biomonitoring studies that reported creatinine-corrected values, one that collected 25 post-shift urine samples from predominantly Asian nail salon workers in Maryland, USA,²⁴ and another that collected 30 urine samples exclusively from the general population in Vietnam.³⁸ One other known study measured urinary MnBP and MiBP in U.S. nail salon workers, but the study population was predominantly white and specific gravity-corrected concentrations were the only values reported, which limited comparability with our results.²⁵ Lastly, we included five epidemiology studies that reported significant associations between non-corrected concentrations of urinary metabolites measured in this study and adverse male developmental outcomes,^39–43 excluding studies that reported specific-gravity corrected values only, as they were not directly comparable.

Comment

This pilot study of 17 nail salon workers provides evidence of elevated individual metabolite and cumulative phthalates exposure in comparison to a national sample of the U.S. Asian American population. Specifically, nail salon workers had significantly higher levels of MnBP, MiBP, and MEHP, as well as higher levels of exposure to multiple phthalates that were combined in a summary metric based on their relative toxicities (Σandrogen-disruptor). To our knowledge, this is the first biomonitoring assessment of phthalates in an exclusive sample of Vietnamese nail salon workers in California, who comprise a uniquely vulnerable occupational group. This pilot study suggests that nail salon workers are exposed to higher levels of biologically relevant phthalate combinations than the general population.

Individual metabolite concentrations exceeded the NHANES 95^th percentile in several nail salon workers, and for more than one metabolite in two participants, indicating that some workers may be simultaneously exposed to phthalate combinations at very high levels relative to the general Asian American population. Most nail salon workers had higher cumulative phthalates daily intake than the NHANES 75^th but not 95^th percentiles. However, the NHANES Asian American comparison population conceivably might include Vietnamese nail salon workers, which could have skewed the distribution (and 95^th percentile) towards the upper end of extreme values. Due to existing evidence that combined phthalate exposures may have greater adverse health effects than individual phthalate exposures, future studies in a larger sample of nail salon workers could examine whether these pilot findings can be replicated.^23,44–46

While this analysis did not assess direct connections between phthalates and health impacts, MnBP, MiBP, and MEHP concentrations were similar to or above those measured in studies reporting adverse developmental endpoints in male offspring.^39–42,44 For example, in a large prospective cohort study across four major U.S. cities (N = 753) that found a significant association between MEHP and reduced anogenital distance among baby boys, Swan et al. (2015) reported a non-creatinine corrected GM of 1.9 ng/mL (95% CI: 1.8, 2.1), which was >1.5 times lower than the GM of 3.5 ng/mL (95% CI: 2.4, 5.3) observed among female nail salon workers in this study.⁴⁰ These exposure levels are of potential concern for women in the nail salon workforce who are of reproductive age and may become pregnant.

MnBP, MEHP, and MEP concentrations measured in this study were lower than those collected seven years prior, reflecting U.S. biomonitoring trends for these metabolites between 2001 and 2012.²⁴ Similarly, MiBP concentrations were two times higher in the current nail salon worker population, which is consistent with a reported >100% increase in urinary MiBP levels over the last decade.⁴⁷ These temporal changes may reflect product reformulations in response to increased phthalates scrutiny and consumer preferences for safer alternatives over the last decade.^47–49 Due to the U.S. Food and Drug Administration’s (FDA) limited oversight over cosmetic ingredients, large public campaigns in recent years have pressured beauty companies, such as OPI, Sally Hansen, and Revlon, to remove phthalates from nail polish and other personal care products.⁵⁰ Subsequent data suggest that DiBP may be replacing the better-known DnBP, which has long been used as a hardener in nail polish to preserve shape and color.^{47,48,51–53} However, DnBP is still detected in many nail care products, including those claiming to be phthalate free,^15,17 and both compounds contribute to androgen-mediated risks associated with phthalate co-exposures.^23,54 Continued product testing and biomonitoring research will be required to evaluate these trends going forward.

While absolute levels have changed over time, the difference between nail salon workers and the general population remained relatively constant between the previous study (which compared samples collected from workers in 2003–4 to NHANES 2001–2)²⁴ and our study (which compared samples collected from workers in 2011 to NHANES 2011–12). Both studies found higher MnBP and MiBP concentrations among nail salon workers, which is consistent with common use of their parent compounds in nail and other beauty products.^25,47 MEP levels were substantially but not significantly higher among nail salon workers in both studies. Although typically associated with personal care product use in non-occupational populations,^18,19,25,55 MEP exposures at the nail salon may be negligible relative to high background concentrations measured in the general population. Indeed, MEP levels were the highest observed in our study as well as previous studies assessing this compound in the U.S. population.^32,47

Our MEHP results are also consistent with prior research which found the highest MEHP concentrations among nail salon workers,²⁴ despite DEHP’s prevalent association with dietary intake from food packaging and processing equipment, rather than from cosmetic or beauty product use.⁴⁴ Levels were nearly five times that of U.S. adults and higher than every other occupational group evaluated (including phthalate and PVC manufacturing workers).²⁴ Oxidative (secondary) metabolites of DEHP were not significantly elevated in nail salon workers, which suggested the possibility of an unknown DEHP source that may be unique to nail salons. We speculate that acrylic nail powder may be a possible unique MEHP exposure source in this workforce, as it contains a catalyst that metabolizes to benzoic acid.^56,57 Given that MEHP is an ester derivative of benzoic acid, acrylic nail powder may conceivably be a direct source of MEHP exposure in nail salons, which would explain why higher primary but not secondary metabolites (which can only be formed in the liver) were observed previously. However, a major limitation of our study is that we did not measure oxidative metabolites for DEHP, which constitute a larger fraction of absorbed parent compound dose than MEHP. Consequently, measuring MEHP alone might underestimate DEHP daily intake. Additionally, oxidative metabolites are less sensitive than primary monoesters to laboratory contamination because they are formed solely from phthalate metabolism in the liver.⁵⁸ Future biomonitoring studies that evaluate DEHP should include at least one oxidized metabolite.

Other study limitations should be addressed in future work. First, our sample size was too small to predict phthalate metabolite and cumulative exposure concentrations from questionnaire data on worker and salon characteristics, thus limiting statistical power to identify specific occupational sources of phthalates exposure. Forthcoming research on nail salon exposures should also include the collection of pre-shift urine samples in addition to air samples collected during the shift, which would help differentiate between occupational and non-occupational (environmental) exposures. The addition of pre-shift urine samples may be the best way to assess phthalate exposures in nail salon workers, given the short metabolic half-lives of phthalates (less than 24 hours).²⁶ Including questionnaire data about non-occupational exposure sources, such as at-home personal care product use and dietary behaviors, would also facilitate delineation between occupational and environmental exposures. Additionally, while prior research has demonstrated moderate long-term reliability of urine spot samples in adults for some metabolites, including MEP and MnBP, studies evaluating reproducibility in other populations are somewhat equivocal.^59–61 Furthermore, intra-person variability may be higher in the workplace due to short-term exposure spikes from specific job tasks.²⁶ Longitudinal data collection could help resolve this issue in future studies.

Lastly, although nail salon workers in this study were compared to NHANES Asian Americans of the same age range sampled within a year of workers in this study, some non-occupational differences between the groups could potentially explain exposure disparities. For example, workers in this study were Vietnamese-born, while the NHANES Asian American population may include second or third generation immigrants. Vietnamese immigrants might use different personal care products or eat different foods than those born in the United States, which could result in observed exposure disparities between workers and the NHANES sample that are not occupationally based. However, Guo et al. (2011) biomonitored 30 people in Vietnam and reported equal or lower creatinine-corrected phthalate metabolite levels than workers in this study,³⁸ suggesting our findings are not likely to due to cultural differences in personal care product use or diet originating from the country of origin. Other differences between the groups, such as location, sampling strategy (convenience rather than multistage probability), data collection time of day (post-shift compared to random), and month or season of sampling (summer rather than throughout the year), might also partially explain exposure differences.

Despite limitations, findings from this pilot study suggest that nail salon workers are disproportionately exposed to multiple phthalates daily at levels associated with adverse effects on reproductive health and development. Comparing Vietnamese nail salon workers to a national sample of Asian Americans provides useful insight and supportive evidence that higher concentrations of individual metabolites and potency weighted cumulative phthalates levels may be due to workplace exposures. Further research should more broadly characterize phthalate exposures and their health implications in this workforce. Continued efforts should be made to identify workplace practices that minimize exposure (i.e., ventilation and glove use) as well as upstream opportunities for promoting safer alternatives.