https://orcid.org/0009-0001-6069-6340
Figures
Abstract
Styrene remains a major hazard in fiberglass-reinforced plastic (FRP) manufacturing. The current 10 ppm 8-h TLV-TWA is half the former limit, and the movement by several European states toward comparable or lower OELs highlights the need for fresh exposure–response data. In this study, eighty-five Korean FRP workers were monitored cross-sectionally. Full-shift breathing-zone styrene was measured by GC-FID; post-shift urine was analysed for mandelic acid (MA) and phenyl-glyoxylic acid (PGA), and dermal uptake was estimated with fluorescent tracers. Neuro-irritative symptoms were assessed by questionnaire and clinically verified in a subset. The results showed that median styrene levels were 18.65 ppm (spray-up), 12.42 ppm (hand lay-up) and 6.37 ppm (closed-mold). Urinary MA and PGA correlated with air levels (r = 0.78, 0.77). Dermal styrene load showed a moderate correlation with urinary MA (r = 0.42, p < 0.001). Symptom prevalence rose from 19% to 71% across exposure quartiles (adjusted OR = 5.6). A biomarker-based model using urinary mandelic acid (MA) with covariates (age, ventilation) showed strong apparent discrimination (AUC = 0.93). We propose 0.38 mg/g creatinine (MA) as a candidate operational (“early-warning”) threshold, pending external validation. In conclusion, integrated air, biological and dermal metrics reveal dose-dependent acute effects at or below 10 ppm. In this cross-sectional analysis, higher styrene exposure was associated with increased acute symptoms at or below ~10 ppm; these associations warrant confirmation in longitudinal studies with repeated biomonitoring. We present 0.38 mg/g creatinine (MA) as a candidate operational (“early-warning”)threshold to flag workers for closer evaluation; external validation is needed, and engineering controls remain the primary means of risk reduction.
Citation: Kwon O-H, Kim K-Y (2025) Occupational exposure to styrene and acute health effects among fiberglass-reinforced plastic workers: An integrated environmental and biological monitoring study. PLoS One 20(11): e0334962. https://doi.org/10.1371/journal.pone.0334962
Editor: Jin Su Kim, Korea Institute of Radiological and Medical Sciences, KOREA, REPUBLIC OF
Received: September 5, 2025; Accepted: October 3, 2025; Published: November 26, 2025
Copyright: © 2025 Kwon, Kim. This is an open access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.
Data Availability: All relevant data files are available from the Zenodo repository (DOI: 10.5281/zenodo.17248606).
Funding: This study was supported by the Research Program funded by SeoulTech (Seoul National University of Science and Technology).
Competing interests: No authors have competing interests.
1. Introduction
Styrene (C₈H₈) is a pivotal industrial monomer predominantly used in the synthesis of polystyrene plastics and unsaturated polyester resins [1,2]. The global styrene market, estimated at approximately $35 billion in 2023, is experiencing growth driven by the increasing demand in the fiberglass-reinforced plastic (FRP) sector, notably in boat manufacturing, automotive components, and construction materials [3,4]. In South Korea, the FRP industry contributes over $2 billion annually to the national economy [5]. The production of FRP components typically employs open-mold processes, such as hand lay-up and spray-up techniques, wherein workers manually apply styrene-containing resins to the molds [6,7]. Although these processes are cost-effective and versatile, they pose a significant potential for occupational exposure due to the high vapor pressure of styrene (0.67 kPa at 25°C) and the open nature of the work environment [8]. In contrast to closed-mold processes that contain emissions, open-mold operations can result in airborne styrene concentrations that frequently exceed the established occupational exposure limits [9,10].
Despite extensive research on styrene toxicity over the decades, substantial knowledge gaps persist regarding exposure-response relationships for acute health effects in real-world occupational settings. Most prior studies have focused on chronic health outcomes or have been conducted in controlled laboratory environments that may not accurately reflect actual workplace conditions [11,12]. Moreover, few studies have integrated multiple exposure assessment methodologies, including environmental monitoring, biological markers, and quantified dermal exposure, to provide a comprehensive understanding of the total internal dose of these chemicals [13,14].
In 2018 the American Conference of Governmental Industrial Hygienists (ACGIH) proposed lowering the styrene 8‑h Threshold‑Limit‑Value Time‑Weighted‑Average (TLV‑TWA) from 20 ppm to 10 ppm; the cut was formally adopted in the 2020 TLV list (STEL unchanged at 20 ppm). Several European countries have since moved toward comparable or stricter limits: Sweden and Spain now apply a 10 ppm national OEL, Norway has adopted 5 ppm, while others, such as Germany and the Netherlands, still retain 20 ppm. These tightening or impending benchmarks underscore the need for contemporary exposure–response data that reflect current workplace conditions.
Large-scale‑ environmental studies hint at short-term‑ neuro‑irritative consequences: among 21,962 Gulf Coast residents, those in the highest quartile of ambient styrene showed a 12–17% excess prevalence of neurologic symptoms compared with the lowest quartile. However, occupational investigations that simultaneously capture breathing zone‑ air, urinary metabolites, and quantified dermal uptake and relate them to real-time‑ health outcomes remain scarce, particularly in Asia. This gap is striking, given South Korea’s sizable and growing FRP-manufacturing workforce.
Therefore, we (i) characterized the total internal styrene dose by measuring concurrent air, urinary, and dermal metrics in Korean FRP workers; (ii) modelled dose-dependent‑ relationships with self-reported‑ and clinically verified neuro‑irritative symptoms; and (iii) derived a biomarker-based‑ action level of 0.38 mg/g creatinine to guide workplace interventions. We hypothesized that adding dermal fluorescence intensity to urinary mandelic acid levels would outperform airborne metrics alone in symptom prediction. The following sections describe our study design, present exposure and health findings, and discuss their implications for occupational hygiene and regulatory policies.
2. Methods
2.1. Study design and setting
This cross-sectional study was conducted at a prominent FRP boat manufacturing facility in Incheon, South Korea, from March to September 2023. The facility employs approximately 150 workers and specializes in producing recreational boats, commercial vessels, and marine components using various FRP-manufacturing processes. The study protocol was approved by the Institutional Review Board of Seoul National University of Science and Technology (IRB No. 2023-0036-01, approved March 6, 2023) and was conducted in accordance with the Declaration of Helsinki [15]. All participants provided written informed consent prior to enrollment.
2.2. Study population
The eligibility criteria for participation included direct involvement in resin handling for a minimum of 2 h per shift or 8 h per week, employment in the current position for at least 6 months, and willingness to participate in all the study procedures. The exclusion criteria were pregnancy, recent use of medications known to interfere with styrene metabolism, and acute illness at the time of the assessment. Of the 92 eligible workers, 85 (92.4%) consented to participate and completed all the study procedures.
2.3. Exposure assessment
2.3.1. Airborne styrene.
Monitoring Personal breathing-zone air samples were collected using calibrated low-flow sampling pumps (SKC Inc., Eighty Four, PA, USA) equipped with activated charcoal sorbent tubes (SKC 226−01). Sampling was conducted over the entire work shift (8 h) at a flow rate of 0.2 L/min. Gas chromatographic analysis was performed using an Agilent 7890A GC system equipped with a flame ionization detector (FID) in accordance with the National Institute for Occupational Safety and Health Method 1501 [16]. The method detection limit was 0.03 mg/m³, and the desorption efficiency was 95.2 ± 2.8%, exceeding the acceptance criterion of ≥90%. Additional details of the GC-FID analytical method are provided in the Supplementary Methods (Section S1.1 in S1 File).
2.3.2. Biological monitoring.
Post-shift urine samples were collected on two consecutive days and analyzed for mandelic acid (MA) and phenylglyoxylic acid (PGA) using high-performance liquid chromatography with UV detection (HPLC-UV) following a validated acid hydrolysis procedure employing 6 M HCl at 100°C for 16 h [17,18]. The acid hydrolysis method was validated against enzymatic hydrolysis and demonstrated excellent agreement (MA: 98 ± 5%; PGA: 96 ± 6%). The complete HPLC-UV analysis protocol and quality control measures are available in Supplementary Section S1.2 in S1 File. Urinary creatinine was measured using the Jaffe reaction to adjust for urine dilution [19].
2.3.3. Dermal exposure assessment.
To quantify dermal styrene uptake, we employed a fluorescent-tracer technique adapted from validated protocols. A non-toxic optical brightener (Uvitex OB, 0.05 wt%) was premixed with the styrene resin each sampling day. At shift-end, workers’ hands and forearms were photographed under 365 nm UV light (Canon EOS R6, 420 nm long-pass filter). Fluorescent tracer–positive skin area (cm²) was quantified using ImageJ (NIH, USA) and converted to mass-based dermal styrene load (µg cm ⁻ ²) via laboratory calibration (R² = 0.98). Duplicate blanks confirmed a method detection limit of 5 µg cm ⁻ ².
In addition to the primary models, we evaluated an alternative specification including dermal exposure; diagnostics and results are provided in Tables S4–S5 in S1 File.
2.4. Health outcome
Acute neuro-irritative symptoms were assessed using a validated questionnaire adapted from previous occupational health studies [20,21]. The questionnaire evaluated the presence and severity of eye, throat, headache, dizziness, and nasal irritation experienced during the current work shift. Symptoms were considered present if they occurred during working hours and were not attributable to other factors. A subset of participants (n = 21) underwent objective clinical measurements, including tear film break-up time (TFBUT) assessment(see Supplementary Methods S1.4 in S1 File) [22].
2.5. Statistical analysis
Statistical analyses were performed using R (version 4.3.0). Continuous variables were assessed for normality using the Shapiro-Wilk test and are reported as median (interquartile range) for data not following a normal distribution or as mean ± standard deviation for data that were normally distributed. Group comparisons were performed using the Kruskal-Wallis test for continuous variables and the chi-square test for categorical variables. The correlations among the exposure variables were evaluated using Spearman’s rank correlation coefficients. Spearman correlations between dermal styrene load and urinary MA/PGA were also computed, and dermal load was entered into multivariable regression models alongside airborne exposure. Logistic regression models were employed to investigate the exposure-response relationships for acute symptoms. Exposure quartiles were established based on the distribution of the TWA of styrene concentrations. Multivariable models were adjusted for potential confounders, including age, smoking status, and the ventilation status. The model fit was evaluated using the Hosmer-Lemeshow goodness-of-fit test. Statistical significance was set at p < 0.05.
Smoking status was ascertained by questionnaire and coded as current/former/never (primary models: current vs. non-current). In a non-smoker–only sensitivity analysis, effect directions and magnitudes were materially unchanged (Table S3 in S1 File).
Airborne and dermal metrics were correlated, and dermal uptake can lie on the pathway to urinary metabolites; to avoid multicollinearity and potential over-adjustment, we prioritized parsimonious models. Nevertheless, we report correlation/VIF diagnostics (Table S4 in S1 File) and an alternative model including dermal exposure (Table S5 in S1 File).
Sensitivity analyses additionally assessed link functions (logit/probit/clog-log), restricted cubic splines (3–5 knots), cluster-robust SEs/mixed-effects, and outlier influence (top/bottom 1% winsorizing). For transport validation, we used temporal hold-out and leave-one-group-out by task/site, summarizing AUC, Brier score, and calibration intercept/slope (Tables S2–S5; Fig S1 in S1 File).
3. Results
3.1. Study population characteristics
The study ultimately included 85 male employees, with a median age of 41 years (IQR: 36–46 years; range: 28–55 years). These individuals had worked at the facility for a median duration of 15 years (IQR: 8.5–22 years) and had been in their current roles for a median of 4 years (IQR: 2–6.5 years). Among the participants, 34.1% (29 out of 85) were current smokers, and an additional 21.2% had previously smoked. A significant proportion of workers (63.5%) had completed at least a high school education. (Table 1) provides a detailed overview of demographic and occupational characteristics.
3.2. Exposure assessment results
3.2.1. Airborne styrene concentrations.
There was a notable variation in the personal TWA styrene levels across the different task categories (Kruskal-Wallis p < 0.001). The highest median concentration was observed during spray-up operations at 18.3 ppm (IQR: 14.2–22.1 ppm), followed by hand lay-up tasks at 11.2 ppm (IQR: 8.9–14.6 ppm) and closed-mold operations at 7.1 ppm (IQR: 5.2–9.8 ppm).
A majority of the workers (61.2%) had time‑weighted average styrene concentrations above the 10 ppm occupational exposure limit now adopted by the ACGIH (8‑h TLV‑TWA, formalized in 2020 after the 2018 proposal) and already enforced in several European countries such as Sweden and Spain [23,24]. Fig 1 illustrates the distribution of airborne styrene concentrations by task category.
The box-and-whisker diagrams illustrate the range of time-weighted average (TWA) styrene levels for the three tasks. The spray-up category exhibited the highest median exposure level at 18.3 ppm, followed by hand lay-up at 11.2 ppm and closed-mold operations at 7.1 ppm. Horizontal lines serve as markers for the important occupational exposure limits.
3.2.2. Urinary biomarker concentrations.
Significant differences in urinary MA concentrations were observed across the different task categories (Kruskal-Wallis, p < 0.001). The median concentrations were 0.412 mg/g creatinine for spray-up workers, 0.342 mg/g creatinine for hand lay-up workers, and 0.278 mg/g creatinine for closed-mold workers. Similarly, the PGA concentrations showed a comparable trend, with median values of 0.089, 0.076, and 0.062 mg/g creatinine. In comparison to the ACGIH BEI of 0.400 mg/g creatinine for combined MA and PGA, 47.1% of all workers surpassed this limit, including 72.4% of spray-up workers, 50.0% of hand lay-up workers, and 17.9% of closed-mold workers [25]. Fig 2 illustrates the distribution of urinary biomarker concentrations according to the task category. (Table 2) presents the detailed exposure assessment results categorized by task. Additional exposure assessment by work area is shown in Supplementary (Table S1 in S1 File).
Urinary mandelic acid (mg/g creatinine) by task—Closed‑mold (n = 28), Hand lay‑up (n = 28), Spray‑up (n = 29). Boxes show IQR and medians (0.269, 0.332, 0.410 mg/g creatinine); whiskers = 1.5 × IQR, circles = outliers. Dashed line marks the ACGIH BEI (0.400 mg/g creatinine).
3.2.3. Dermal exposure.
Workers engaged in spray-up tasks exhibited the highest dermal exposure metrics, with a mean fluorescent tracer-positive skin area of 15.2 ± 4.8 cm² and a corresponding dermal styrene load of 220 ± 65 µg cm ⁻ ². The median dermal styrene load was 212 µg cm ⁻ ² (IQR: 170–270).
In contrast, hand lay-up and closed-mold workers had significantly lower exposure, with mean dermal loads of 130 ± 44 and 40 ± 17 µg cm⁻², respectively.
To standardize inter-task comparisons, a dermal exposure index (DEI) was calculated, setting spray-up as the reference (DEI = 1.00). The relative DEIs were 0.59 ± 0.07 for hand lay-up and 0.18 ± 0.05 for closed-mold operations.
These findings highlight a substantial gradient in dermal exposure intensity aligned with task characteristics and proximity to resin application(Table 3).
3.3. Exposure-biomarker relationships
Strong positive correlations were observed between the levels of styrene in the air and the metabolite levels in urine. The correlation coefficient between TWA styrene and urinary MA was r = 0.78 (p < 0.001), and for PGA, it was r = 0.77 (p < 0.001). The correlation between MA and PGA was even more pronounced, with a coefficient of r = 0.78 (95% CI: 0.67–0.86; p < 0.001). Dermal styrene load showed a moderate correlation with urinary MA (r = 0.42, p < 0.001). Linear regression analysis using log-transformed data showed that a tenfold increase in airborne styrene concentration corresponded to a 6.8-fold increase in urinary MA concentration (95% CI: 4.2–11.0; p < 0.001) [26,27]. Fig 3 illustrates the connection between airborne styrene and urinary mandelic acid. (Table 4) presents the full correlation matrix for all exposure variables and health outcomes.
Scatter plot illustrating the correlation between airborne styrene concentrations (TWA, ppm) and urinary mandelic acid levels (mg/g creatinine) among fiberglass-reinforced plastic workers (n = 85). Each dot represents an individual measurement, color-coded by task category: closed-mold (blue), hand lay-up (green), and spray-up (red). A positive correlation was observed (r = 0.61, p ≤ 0.001), and the fitted regression line indicates a linear trend between airborne exposure and internal dose.
3.4. Health outcome assessment
3.4.1. Acute symptom prevalence.
The overall occurrence of acute neuro-irritative symptoms was 41.4% among the workers (35 of 85). The prevalence of these symptoms showed a significant increase across exposure quartiles: 19.0% in the first quartile (less than 8.2 ppm), 27.3% in the second quartile (8.2–12.1 ppm), 47.6% in the third quartile (12.2–19.5 ppm), and 71.4% in the fourth quartile (> 19.5 ppm) (χ² trend test, p < 0.001). The most frequently reported symptoms were eye irritation (28.2%), headache (24.7%), and throat irritation (21.2%). Among the workers who underwent objective clinical assessments (n = 21), 57.1% of those with symptoms had an abnormal tear film‑ break‑up time (TFBUT < 10 s; see Supplementary Methods S1.4 in S1 File) compared to 12.5% of those without symptoms (p = 0.02). [28,29] Fig 4 illustrates the dose-response relationship between styrene exposure and the prevalence of acute symptoms in workers.
This bar chart illustrates the occurrence of acute neuro-irritative symptoms across the different styrene exposure quartiles. The chart revealed a distinct dose-response pattern, with symptom prevalence rising from 19.0% in the lowest quartile (Q1: ≤ 8.2 ppm) to 71.4% in the highest quartile (Q4: ≥ 19.5 ppm). Error bars indicate the 95% confidence intervals.
3.4.2. Exposure-response relationships.
Logistic regression analysis identified strong exposure-response relationships with acute symptoms in workers. In the unadjusted model, workers in the top exposure quartile had 9.15 times the odds of experiencing symptoms compared with those in the lowest quartile (95% CI: 2.89–28.9; p < 0.001). After adjusting for age, smoking status, and ventilation status, the odds ratio remained significant at 5.60 (95% CI: 2.60–12.00; p < 0.001). The use of local exhaust ventilation showed a notable protective effect, reducing the risk of acute symptoms by 68% (OR=0.32; 95% CI: 0.12–0.82; p = 0.018). The detailed results of the logistic regression analysis are presented in (Table 5). These results are visually summarized in the forest plot in Fig 5.
Forest plot showing adjusted odds ratios and 95% confidence intervals for acute neuro-irritative symptoms. The plot displays the results of the multivariable logistic regression model adjusted for age, smoking, and ventilation status. Exposure quartiles Q2-Q4 show increasing odds ratios compared to Q1 (reference), with Q4 demonstrating a statistically significant association (OR=5.60, 95% CI: 2.60-12.00). Ventilation ON shows a protective effect (OR=0.32, 95% CI: 0.12-0.82).
3.5. Biomarker-based prediction model
3.5.1. Model development and performance.
A prediction model based on biomarkers was created using multivariable logistic regression, with urinary MA concentration as the main predictor while accounting for age and ventilation status. The model showed good apparent discrimination (in-sample AUC 0.93, 95% CI 0.88–0.99); however, internal cross-validation yielded more modest performance (AUC 0.46–0.76; mean 0.606 ± 0.118), with acceptable calibration (slope 0.871; intercept −0.084) and a Brier score of 0.241. The Hosmer–Lemeshow test indicated adequate fit (p = 0.67). Cross-validated ROC and calibration plots are provided in Supplementary Fig S1 in S1 File.
3.5.2. Optimal cut-point analysis.
ROC analysis identified a urinary MA threshold of 0.38 mg/g creatinine that best separated cases from non-cases, yielding sensitivity 85.7% (95% CI 69.7–95.2%) and specificity 88.0% (95% CI 76.2–95.6%) in-sample; the positive and negative predictive values were 81.1% (95% CI 64.8–92.0%) and 91.5% (95% CI 79.6–97.6%), respectively. This candidate threshold (MA-only) is numerically lower than the ACGIH BEI of 0.400 mg/g creatinine (MA + PGA, end-of-shift), which is an exposure index rather than a health-effect threshold; it should be interpreted as an operational monitoring trigger pending external validation, not a replacement for the BEI. Fig 6 illustrates the ROC curve, and Table 6 summarizes the performance metrics.
Apparent (in sample) AUC = 0.93. Cross validated performance is summarized in Section 3.6. ROC curve of the symptom‑prediction model based on urinary mandelic acid (mg g ⁻ ¹ creatinine). The model shows excellent discrimination (AUC = 0.93). The optimal threshold, 0.38 mg g ⁻ ¹, yields 85.7% sensitivity and 88.0% specificity; the diagonal line represents chance performance (AUC = 0.50).
3.6. Sensitivity and validation
Findings were robust across sensitivity analyses. Spline tests indicated no material non-linearity (global spline p = 0.266). Validation yielded AUCs of 0.46–0.76 (mean 0.606 ± 0.118) with calibration slope 0.871 and intercept −0.084 based on out-of-fold predictions; the average Brier score was 0.241 [30,31]. Results remained consistent across alternative model specifications using the available variables (Tables S2–S5; Fig S1 in S1 File).
4. Discussion
This comprehensive investigation provides robust evidence of exposure-dependent acute neuro-irritative symptoms among fiberglass-reinforced plastic (FRP) workers in South Korea, representing one of the most extensive integrated assessments of styrene exposure in Asian industrial settings to date. This study combined environmental monitoring, biological markers, and health outcome evaluations to advance our understanding of the complex relationship between styrene exposure and occupational health outcomes in workers.
The hierarchy of styrene exposure risk was clear: spray‑up operations produced the highest concentrations (median 18.3 ppm), well above contemporary occupational limits. This level exceeds the ACGIH 10 ppm 8‑h TLV‑TWA proposed in 2018 and adopted in the 2020 TLV list as well as the 10 ppm national OELs now enforced in several European countries such as Sweden and Spain. By contrast, Germany and the Netherlands still apply a 20 ppm limit, underscoring how urgently high‑exposure tasks require improved controls [32,33].
The strong correlations between airborne styrene concentrations and urinary metabolite levels (r=0.78 for mandelic acid and r=0.77 for phenylglyoxylic acid) provide compelling evidence for biological monitoring in this occupational population [34,35]. Although inhalation predominated, dermal uptake independently predicted urinary MA levels, underscoring the need for combined respiratory and skin protection [36]. Linear regression analysis revealed that a ten-fold increase in airborne styrene corresponded to a 6.8-fold increase in urinary mandelic acid (95% CI: 4.2-11.0; p<0.001), providing a quantitative framework for predicting the internal dose from environmental measurements. The exceptionally strong correlation between mandelic acid and phenylglyoxylic acid (r=0.78) confirms that both metabolites reflect the same underlying exposure processes, supporting the use of either biomarker as a reliable exposure indicator.
The clear dose-response relationship for acute neuro-irritative symptoms, with prevalence increasing from 19.0% in the lowest exposure quartile to 71.4% in the highest, provides compelling epidemiological evidence of acute health effects at levels commonly encountered in occupational settings. Multivariable logistic regression analysis demonstrated that workers in the highest exposure quartile had 5.60 times the odds of experiencing acute symptoms compared to those in the lowest quartile (95% CI: 2.60-12.00; p<0.001). This finding is particularly significant because many exposure levels associated with symptoms were below the current occupational exposure limits, suggesting that the existing regulatory guidelines may not provide adequate protection against acute health effects. The most frequently reported symptoms—eye irritation (28.2%), headache (24.7%), and throat irritation (21.2%)— were consistent with the known neurotoxic and irritant properties of styrene. Objective clinical assessment of tear film break-up time provided additional validation, with 57.1% of symptomatic workers showing abnormal results compared with only 12.5% of asymptomatic workers (p=0.02).
External validity is limited by sampling from a single FRP facility and an all-male workforce; replication across multi-site, mixed-sex cohorts with differing process controls is warranted. Additionally, detailed respirator/PPE use was unavailable; any unmeasured protection would likely bias exposure–outcome associations toward the null, suggesting our estimates may be conservative.
The significant protective effect of local exhaust ventilation, which reduced the risk of acute symptoms by 68% (OR=0.32; 95% CI: 0.12-0.82; p=0.018), underscores the critical importance of implementing such engineering control measures. The differential effectiveness across task categories, with 78.6% of closed-mold workers having access to functioning ventilation compared to only 41.4% of spray-up workers, highlights a critical gap in protective measures for the highest-risk operations in the industry. This disparity likely contributes to the observed differences in exposure and suggests that targeted interventions focusing on spray-up operations may yield substantial health benefits.
In the dermal-included alternative model (Table S5 in S1 File), the association for Ventilation ON was attenuated and non-significant (adjusted OR 0.62, 95% CI: 0.23–1.64; p = 0.333), whereas it was protective in the primary model (OR 0.32, 95% CI: 0.12–0.82; p = 0.018)
The development of a biomarker-based prediction model with exceptional discriminatory performance (AUC=0.93, 95% CI: 0.88-0.99) represents a significant advancement in occupational health surveillance methodology. Although dermal exposure was a significant contributor to the total internal dose, it was not retained as an independent predictor in the final multivariable symptom model after accounting for urinary MA levels. The identification of an optimal cut-off point for urinary mandelic acid at 0.38 mg/g creatinine, corresponding to 85.7% sensitivity and 88.0% specificity, provides a health-based action level that is more protective than current regulatory guidelines. This threshold is particularly valuable because it is derived from actual health outcomes rather than arbitrary safety factors applied to toxicological data. The finding that 47.1% of workers exceeded the ACGIH Biological Exposure Index (BEI) of 0.400 mg/g creatinine for combined mandelic acid and phenylglyoxylic acid raises important questions regarding the adequacy of the current international guidelines. Our health-based action level of 0.38 mg/g creatinine for mandelic acid alone, which is lower than the combined BEI, suggests that the existing guidelines may not provide sufficient protection against acute health effects [37–40].
These findings have significant implications for the global FRP industry, which has experienced substantial growth driven by demand in the automotive, marine, construction, and renewable energy sectors. The widespread use of open-mold processes, particularly in developing countries, suggests that millions of workers worldwide may be at risk of developing styrene-related health problems. The high prevalence of acute symptoms (41.4% overall, rising to 71.4% in the highest exposure group) indicates that styrene exposure may significantly contribute to reduced work performance and quality of life in FRP workers. The economic impact extends beyond direct medical costs to include productivity losses, absenteeism, and worker turnover, suggesting that implementing effective control measures can yield substantial economic benefits through improved worker health and productivity of the organization.
This study had several notable strengths, including a comprehensive exposure assessment approach that combined personal air monitoring, biological markers, and health outcome evaluation. The real-world occupational setting provides ecological validity, while the relatively large sample size and high participation rate (92.4%) minimize selection bias. However, the limitations include the cross-sectional design, single facility setting, focus on acute rather than chronic health effects, and all-male study population. Another significant limitation of this study was the failure to account for the use of personal protective equipment (PPE), particularly respirators, as a variable. As respirators directly reduce the inhalation of styrene vapors, their use critically impacts the internal dose. Consequently, the airborne styrene concentrations (external dose) measured in this study may not perfectly reflect the precise exposure levels of each worker. Future studies should incorporate PPE use as a covariate to accurately model exposure-response relationships.
Despite this limitation, a key strength of this study was the integration of biological monitoring. The urinary concentrations of mandelic acid (MA) and phenylglyoxylic acid (PGA) reliably reflected the total absorbed dose of styrene, regardless of whether the workers were wearing a respirator. Indeed, the symptom prediction model based on urinary MA concentration demonstrated outstanding discriminatory performance (AUC = 0.93). This provides robust evidence that biomarker data compensate for the limitations of external air monitoring and effectively explain the actual health risks. Therefore, the health-based action level proposed in this study remains highly valid and applicable for protecting workers’ health.
Urinary mandelic and phenylglyoxylic acids predominantly reflect recent exposure (hour-scale effective half-times), yet multi-compartment elimination implies potential carryover from prior shifts; accordingly, repeated biomonitoring in longitudinal designs would strengthen causal inference [41].
The demonstration that acute health effects occur at exposure levels below current regulatory limits suggests that existing standards may need to be revised. The effectiveness of engineering controls supports regulatory approaches that prioritize engineering solutions over personal protective equipment (PPE).
Future research should focus on longitudinal health outcomes, genetic susceptibility factors, and the development of innovative exposure control technologies. The integration of real-time monitoring systems with biological surveillance can provide unprecedented capabilities for protecting worker health through immediate intervention when exposure thresholds are exceeded. Studies in other geographic regions and industrial settings could assess the generalizability of these findings and identify factors that influence exposure-response relationships.
This study provides compelling evidence of exposure-dependent acute health effects among FRP workers, with spray-up operations posing the greatest risk. The health-based action level of 0.38 mg/g creatinine for urinary mandelic acid provides a more protective benchmark than existing guidelines. The biomarker-based prediction model demonstrated outstanding discriminatory capability and offers a practical tool for occupational health surveillance programs. These findings advocate the establishment of integrated environmental and biological monitoring initiatives, coupled with enhanced workplace interventions, to safeguard worker health in the expanding global FRP industry. Priority should be given to improving ventilation systems, particularly for spray-up processes, and implementing biological monitoring programs based on the health-based action level identified in this study. A breakdown of the symptom-specific trends by exposure quartile is presented in Supplementary (Table S2 in S1 File). The scientific evidence presented supports the development of more protective occupational health standards and provides a foundation for evidence-based interventions that can significantly improve worker health and safety in this industry.
5. Conclusions
This study offers compelling evidence of acute health effects that vary with exposure levels among FRP workers, with spray-up operations posing the greatest risk. The strong link between airborne concentrations and urinary biomarkers confirms that MA and PGA are reliable indicators of the total internal dose of styrene exposure. The significant protective effects of local exhaust ventilation highlight the essential role of engineering control measures.
While the ACGIH BEI for styrene is 0.400 mg/g creatinine (MA+PGA, end-of-shift)—an exposure index intended for biological monitoring rather than a health-effect threshold—we propose 0.38 mg/g (MA) as a candidate operational threshold to flag workers for closer evaluation. The apparent (in-sample) ROC AUC was 0.93, whereas internal cross-validation showed more modest discrimination (AUC 0.46–0.76; mean 0.606 ± 0.118) with acceptable calibration (slope 0.871; intercept −0.084) (see Section 3.6). This cut-point does not replace the BEI and requires external validation.
The results advocate for the establishment of comprehensive environmental and biological monitoring initiatives, along with improved workplace measures to safeguard the health of workers in the growing global FRP sector. Emphasis should be placed on enhancing ventilation systems, especially for spray-up processes, and adopting biological monitoring programs based on the suggested health-based action levels [42,43]. All supplementary materials are available in the Supporting Information file (S1 File).
References
- 1.
National Institute for Occupational Safety and Health. NIOSH Manual of Analytical Methods: Styrene, Method 1501. Cincinnati, OH: NIOSH. 2003.
- 2. Guillemin MP, Berode M. Biological monitoring of styrene: a review. Am Ind Hyg Assoc J. 1988;49(10):497–505. pmid:3055914
- 3. Banton MI, Bus JS, Collins JJ, Delzell E, Gelbke H-P, Kester JE, et al. Evaluation of potential health effects associated with occupational and environmental exposure to styrene - an update. J Toxicol Environ Health B Crit Rev. 2019;22(1–4):1–130. pmid:31284836
- 4.
International Agency for Research on Cancer. IARC Monographs on the Evaluation of Carcinogenic Risks to Humans: Some Traditional Herbal Medicines, Some Mycotoxins. Naphthalene and Styrene. 82. 2002.
- 5.
Korean Chemical Industry Association. Annual Report on Chemical Industry Statistics. Seoul: KCIA. 2023.
- 6. Checkoway H, Pearce N, Crawford-Brown DJ, Cragle DL. Radiation doses and cause-specific mortality among workers at a nuclear materials fabrication plant. Am J Epidemiol. 1988;127(2):255–66. pmid:3337081
- 7. Nett RJ, Cox-Ganser JM, Hubbs AF, Ruder AM, Cummings KJ, Huang Y-CT, et al. Non-malignant respiratory disease among workers in industries using styrene-A review of the evidence. Am J Ind Med. 2017;60(2):163–80. pmid:28079275
- 8. Rappaport SM, Kupper LL, Lin YS. Epidemiological evidence for the threshold concept in occupational exposure to styrene. Scand J Work Environ Health. 2009;35:275–84.
- 9.
European Chemicals Agency. Restriction Report for Styrene. Helsinki: ECHA. 2019.
- 10. Occupational Safety and Health Administration. Occupational exposure to styrene; final rule. Fed Regist. 1989;54:20672–707.
- 11. Murata K, Araki S, Yokoyama K, Tanigawa T, Yamashita K, Okajima F, et al. Cardiac autonomic dysfunction in rotogravure printers exposed to toluene in relation to peripheral nerve conduction. Ind Health. 1993;31(3):79–90. pmid:8262818
- 12. Triebig G, Stark T, Ihrig A, Dietz MC. Intervention study on acquired color vision deficiencies in styrene-exposed workers. J Occup Environ Med. 2001;43(5):494–500. pmid:11382185
- 13. Manini P, De Palma G, Andreoli R, Poli D, Mozzoni P, Folesani G, et al. Environmental and biological monitoring of benzene exposure in a cohort of Italian taxi drivers. Toxicol Lett. 2006;167(2):142–51. pmid:17056211
- 14. Manno M, Viau C, in collaboration with, Cocker J, Colosio C, Lowry L, et al. Biomonitoring for occupational health risk assessment (BOHRA). Toxicol Lett. 2010;192(1):3–16. pmid:19446015
- 15.
World Medical Association. World Medical Association Declaration of Helsinki: Ethical Principles for Medical Research Involving Human Subjects. 2013.
- 16.
National Institute for Occupational Safety and Health. NIOSH Manual of Analytical Methods: Styrene, Method 1501. 4th ed. Cincinnati, OH: NIOSH. 2003.
- 17. Sabatini L, Barbieri A, Indiveri P, Righi E, Violante FS. Validation of an HPLC‑MS/MS Method for the Simultaneous Determination of Phenylmercapturic Acid, Benzylmercapturic Acid and o-Methylbenzyl Mercapturic Acid in Urine. J Chromatogr B. 2008;863(2):115–22.
- 18. Ghittori S, Fiorentino ML, Maestri L, Cordioli G, Imbriani M. Urinary excretion of unmetabolized benzene as an indicator of benzene exposure. J Toxicol Environ Health. 1993;38(3):233–43. pmid:8450555
- 19. Barr DB, Wilder LC, Caudill SP, Gonzalez AJ, Needham LL, Pirkle JL. Urinary creatinine concentrations in the U.S. population: implications for urinary biologic monitoring measurements. Environ Health Perspect. 2005;113(2):192–200. pmid:15687057
- 20. Kaukiainen A, Vehmas T, Rantala K, Nurminen M, Martikainen R, Taskinen H. Results of common laboratory tests in solvent-exposed workers. Int Arch Occup Environ Health. 2004;77(1):39–46. pmid:14600835
- 21. Edling C, Anundi H, Johanson G, Nilsson K. Increase in neuropsychiatric symptoms after occupational exposure to low levels of styrene. Br J Ind Med. 1993;50(9):843–50. pmid:8398879
- 22. Schiffman SS, Nagle HT. Effect of environmental pollutants on taste and smell. Otolaryngol Head Neck Surg. 1992;106(6):693–700. pmid:1608635
- 23.
American Conference of Governmental Industrial Hygienists. American Conference of Governmental Industrial Hygienists. TLVs and BEIs: Threshold Limit Values for Chemical Substances and Physical Agents and Biological Exposure Indices. ACGIH. https://www.acgih.org/tlv-bei-guidelines/. 2023.
- 24. European Union. Commission Directive 2009/161/EU Establishing a Third List of Indicative Occupational Exposure Limit Values. Off J Eur Union. 2009;L338:87–9.
- 25.
American Conference of Governmental Industrial Hygienists. Documentation of the Threshold Limit Values and Biological Exposure Indices. 7th ed. Cincinnati, OH: ACGIH. 2001.
- 26. Fustinoni S, Buratti M, Campo L, Colombi A, Consonni D, Pesatori AC, et al. Urinary t,t-muconic acid, S-phenylmercapturic acid and benzene as biomarkers of low benzene exposure. Chem Biol Interact. 2005;153–154:253–6. pmid:15935823
- 27. Bonanni RC, Gatto MP, Paci E, Gordiani A, Gherardi M, Tranfo G. Biomonitoring for Exposure Assessment to Styrene in the Fibreglass Reinforced Plastic Industry: Determinants and Interferents. Ann Occup Hyg. 2015;59(8):1000–11. pmid:26180262
- 28. Korn RJ, Dockery DW, Speizer FE, Ware JH, Ferris BG Jr. Occupational exposures and chronic respiratory symptoms. A population-based study. Am Rev Respir Dis. 1987;136(2):298–304. pmid:3497594
- 29. Triebig G, Bruckner T, Seeber A. Occupational styrene exposure and hearing loss: a cohort study with repeated measurements. Int Arch Occup Environ Health. 2009;82(4):463–80. pmid:18762967
- 30. Harrell FE Jr, Lee KL, Mark DB. Multivariable prognostic models: issues in developing models, evaluating assumptions and adequacy, and measuring and reducing errors. Stat Med. 1996;15(4):361–87. pmid:8668867
- 31. Steyerberg EW, Harrell FE Jr, Borsboom GJ, Eijkemans MJ, Vergouwe Y, Habbema JD. Internal validation of predictive models: efficiency of some procedures for logistic regression analysis. J Clin Epidemiol. 2001;54(8):774–81. pmid:11470385
- 32. Schinkel J, Ritchie P, Goede H, Fransman W, van Tongeren M, Cherrie CW, et al. The Advanced REACH Tool (ART): Incorporation of an Exposure Measurement Database. Ann Occup Hyg. 2013;57(6):717–27.
- 33. Banton MI, Bus JS, Collins JJ, Delzell E, Gelbke H-P, Kester JE, et al. Evaluation of potential health effects associated with occupational and environmental exposure to styrene - an update. J Toxicol Environ Health B Crit Rev. 2019;22(1–4):1–130. pmid:31284836
- 34. Fustinoni S, Rossella F, Campo L, Mercadante R, Bertazzi PA. Urinary BTEX, MTBE and naphthalene as biomarkers to gain environmental exposure profiles of the general population. Sci Total Environ. 2010;408(14):2840–9. pmid:20417546
- 35. Manini P, De Palma G, Andreoli R, Poli D, Mozzoni P, Folesani G, et al. Environmental and biological monitoring of benzene exposure in a cohort of Italian taxi drivers. Toxicol Lett. 2006;167(2):142–51. pmid:17056211
- 36. Aitio A, Pekari K, Järvisalo J. Skin absorption as a source of error in biological monitoring. Scand J Work Environ Health. 1984;10(5):317–20. pmid:6523096
- 37. Murata K, Araki S, Yokoyama K, Okumura T, Ishimatsu S, Takasu N, et al. Asymptomatic sequelae to acute sarin poisoning in the central and autonomic nervous system 6 months after the Tokyo subway attack. J Neurol. 1997;244(10):601–6. pmid:9402534
- 38. Triebig G, Stark T, Ihrig A, Dietz MC. Intervention study on acquired color vision deficiencies in styrene-exposed workers. J Occup Environ Med. 2001;43(5):494–500. pmid:11382185
- 39.
Lauwerys RR, Hoet P. Industrial Chemical Exposure: Guidelines for Biological Monitoring. 3rd ed. Boca Raton, FL: CRC Press. 2001.
- 40. Apostoli P, Alessio L. Challenges of biomonitoring for the occupational health physician. Toxicol Lett. 2006;162:179–85.
- 41.
Lauwerys RR, Hoet P. Industrial Chemical Exposure: Guidelines for Biological Monitoring. 3rd ed. Boca Raton, FL: CRC Press; 2001.
- 42. Symanski E, Kupper LL, Rappaport SM. Comprehensive evaluation of long-term trends in occupational exposure: Part 1. Description of the database. Occup Environ Med. 1998;55(5):300–9. pmid:9764107
- 43. Rappaport SM, Kupper LL, Lin YS. Epidemiological evidence for the threshold concept in occupational exposure to styrene. Scand J Work Environ Health. 2009;35:275–84.