Association between exposure to polycyclic aromatic hydrocarbons and reproductive health outcomes: a systematic review and meta-analysis

Abstract

Background

Polycyclic aromatic hydrocarbons (PAHs) are widespread environmental pollutants with known harmful effects on human health. However, their specific impact on reproductive outcomes, both cancer-related and non-cancer-related, has not been comprehensively assessed. This study systematically reviews and synthesizes existing evidence on PAH exposure and reproductive health in men and women.

Methods

A comprehensive literature search was conducted in Web of Science, Cochrane Library, PubMed/MEDLINE, ProQuest, and Scopus through May 31, 2024, following PRISMA guidelines. Study quality was assessed using the Newcastle-Ottawa Scale, and the strength of evidence was evaluated using the GRADE framework. The certainty of evidence was rated as moderate for male reproductive organ cancers and low for female reproductive organ cancers, based on GRADE assessment, primarily due to imprecision in the latter.

Results

Of the 4,546 articles screened, 30 met the inclusion criteria for the systematic review. Among these, 9 studies were included in a meta-analysis. In men, PAH exposure was consistently associated with reduced semen quality, including lower sperm count, motility, viability, and morphology, as well as increased DNA damage and hormone disruption. The meta-analysis of seven studies found a 13% increased risk of male reproductive cancers, primarily prostate cancer, associated with PAH exposure (standardized incidence ratio [SIR]: 1.13; 95% CI: 1.04–1.23; P < 0.001). For women, PAH metabolites were linked to infertility and possibly endometriosis, but these associations were weakened after controlling for confounders. No significant association was found between PAH exposure and female reproductive cancers (SIR: 1.01; 95% CI: 0.91–1.12; P = 0.91).

Conclusion

PAH exposure may be associated with adverse male reproductive outcomes, including impaired semen quality and a potential increase in reproductive cancer risk. However, the evidence is limited by methodological heterogeneity, observational designs, and imprecision. In contrast, evidence in females is sparse and of very low certainty, underscoring the need for more rigorous, targeted research on female reproductive outcome.

Supplementary Information

The online version contains supplementary material available at 10.1186/s41043-025-01104-w.

Introduction

Polycyclic aromatic hydrocarbons (PAHs) are a class of over 100 organic compounds composed of multiple fused aromatic rings [1, 2]. They are generated through incomplete combustion of organic materials such as coal, petroleum, wood, and tobacco, and are widespread in both indoor and outdoor environments [3–5]. Human exposure to PAHs occurs via inhalation of polluted air, ingestion of contaminated food or water, and dermal contact with PAH-containing substances [6–9]. Due to their persistence, bioaccumulation potential, and toxicity, PAHs are classified as persistent organic pollutants (POPs) [10–14].

The carcinogenic potential of PAHs is well established, with several compounds classified as probable or possible human carcinogens by the International Agency for Research on Cancer (IARC) [15]. Their mechanisms of action include DNA adduct formation, oxidative stress, and disruption of genetic integrity [16–19]. In addition, PAHs function as endocrine-disrupting chemicals (EDCs), altering hormonal signaling and reproductive function in both males and females [20–22].

In males, exposure to PAHs has been associated with reduced semen quality, increased sperm DNA fragmentation, and hormonal imbalances such as decreased testosterone levels [23–26]. In females, PAHs have been linked to ovulatory dysfunction, menstrual irregularities, infertility, and endometriosis, although the evidence remains inconsistent [27, 28]. Several studies have also explored associations between PAH exposure and reproductive organ cancers, with varying conclusions [28–32].

Knowledge gaps and rationale

Despite growing recognition of PAH-related reproductive toxicity, key gaps in the literature persist. First, there is significant heterogeneity in study designs, exposure assessment methods, outcome definitions, and populations studied, which limits comparability and generalizability. Second, the differential effects of PAH exposure on male versus female reproductive health have not been systematically examined. Third, few reviews have simultaneously addressed both carcinogenic and non-carcinogenic reproductive outcomes. Finally, recent epidemiological studies from the past two years have not been incorporated into existing synthese [18, 27, 33]

This systematic review and meta-analysis aims to address these limitations by critically evaluating the association between PAH exposure and human reproductive health outcomes including both carcinogenic and non-carcinogenic effects, in males and females. By integrating findings from diverse populations and study designs, this review seeks to clarify existing evidence, highlight critical research needs, and support evidence-based public health strategies

Methods

Design and registration

The study protocol was meticulously registered in the PROSPERO, International Prospective Register of Systematic Reviews, under the unique code CRD42024548548. Adhering to methodological rigor, this systematic review followed the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) 2020 statement [33].

Search strategy

A systematic literature search was conducted in PubMed/MEDLINE, Scopus, Web of Science (WoS), Cochrane Library, and ProQuest up to May 31, 2024. The search strategy incorporated both Medical Subject Headings (MeSH) and relevant keywords. The complete search string used for PubMed/MEDLINE is presented in Supplementary File 1.

To ensure a comprehensive and unbiased retrieval of relevant studies, we also conducted an extensive manual and grey literature search. This included: reviewing reference lists of all eligible full-text articles.Searching the national grey literature repository (https://allcatsrgrey.org.uk). Screening preprint platforms such as medRxiv (https://www.medrxiv.org) and PsyArXiv (https://psyarxiv.com), Conducting an incognito Google Scholar search, examining the first 200 results. Reviewing the EThOS thesis repository (https://ethos.bl.uk). The grey literature search methodology, including screening rules, inclusion criteria, and rationale for limits applied is presented in Supplementary File 1, Table S1. These manual and supplementary efforts identified 55 additional studies, which were screened and included in the total count of retrieved articles, as documented in Fig. 1. All searches were performed independently by two authors (AVA and SA), with discrepancies resolved through consensus.

Fig. 1.

The literature searches results and the screening process based on PRISMA 2020 flowchart

Eligibility criteria

The eligibility criteria for studies included in the analysis were determined using the PECOs strategy, which stands for Participants, Exposure, Comparison, Outcome, and Study Design [34]. Study selection followed the PECOs framework: Participants: Adult males and females (≥ 18 years). Exposure: PAHs exposure (biomarkers in urine, blood, or semen), including environmental, occupational, or dietary sources. Comparison: Groups with vs. without PAH exposure, or fertile vs. infertile participants. Primary Outcomes: Carcinogenic (prostate/testicular cancer, breast/cervical/uterine cancer) and non-carcinogenic (semen quality, fertility, endometriosis) reproductive health outcomes. Secondary Outcomes: Specific types and concentrations of PAHs. Study Design: Observational studies (cohort, case-control, cross-sectional). Exclusions: animal studies, clinical trials, letters, conference abstracts, or non-English articles lacking full text or methodology details.

We excluded studies without access to the full text, animal studies, clinical trials, letters editor, conference papers, and posters if the main article was inaccessible or lacked methodology information and also those published in a non-English language.

Data collection

In order to improve the efficiency of managing search results and screening for relevant studies, the studies were transferred to EndNote software (version: 20.2.1). This process involved removing duplicate entries and ensuring proper referencing. Initially, two independent researchers (AVA & SA) screened titles and abstracts to identify studies that addressed the research questions. Following this, the same researchers carefully evaluated full-text articles to select studies that met the inclusion criteria. Articles meeting the criteria were included, while those that did not were excluded. Inter-rater agreement was assessed using Cohen’s kappa (κ). For title and abstract screening (n = 1,827), κ = 0.84 (95% CI: 0.81–0.87), indicating almost perfect agreement. For full-text screening (n = 143), κ = 0.79 (95% CI: 0.71–0.87), indicating substantial agreement. Discrepancies were resolved through discussion or consultation with a third reviewer (FA). The final search results were illustrated in a flow diagram according to the guidelines set out in the PRISMA statement.

Quality appraisal

The quality of studies included in this systematic review and meta-analysis was assessed by two independent researchers (AVA & SA) using the Newcastle–Ottawa Scale (NOS) a tool designed to evaluate the methodological quality of non-randomized studies, such as observational studies [35]. The NOS consists of three sections: selection of study groups, comparability of groups, and ascertainment of outcomes or exposure (Supplementary 1). The NOS assigns up to a maximum of nine points for the lowest risk of bias across the three domains for case-control and cohort studies studies. This breakdown includes four points for Selection, two points for Comparability, and three points for Outcomes, totaling a possible nine points overall [36]. For cross-sectional studies, we used the validated adaptation of the NOS [37], which modifies the selection and outcome domains to better reflect cross-sectional design features and uses an 8-point scale; 4 for Selection, 2 for Comparability, and 2 for Outcome. The total NOS score was converted to the Agency for Healthcare Research and Quality (AHRQ) quality rating as follows: Good: ≥6 points; Fair: 4–5 points; Poor: ≤3 points [38].

Moreover, the credibility of evidence from various types of included studies was assessed using the Grading of Recommendations, Assessment, Development, and Evaluations (GRADE) framework [39]. This framework evaluated the evidence by considering factors such as risk of bias, imprecision, inconsistency, indirectness, and other relevant considerations. Given that all included studies were observational, the initial certainty of evidence was rated as low, in accordance with the GRADE framework. This could be downgraded further for limitations in risk of bias, inconsistency, indirectness, or imprecision, or upgraded for factors such as large effect size, dose-response relationship, or plausible confounding that would diminish a demonstrated association.

Data extraction

Two researchers independently reviewed the full texts of all eligible studies and extracted data using a pre-established checklist, which was then recorded in an electronic file (Excel sheets). The extracted information included the first author’s name, year of publication, study design, country, types of PAHs, participants, gender, sample size, biological sample (urine, semen, or blood), outcomes that contain non-carcinogenic outcomes (semen quality and infertility for men and infertility for women) and carcinogenic outcomes (prostate and testicular cancers in men and breast, cervical, and uterine cancers in women), effects of the chemical agents on reproductive outcomes and key findings.

Synthesis of results and statistical analysis

While the primary outcomes included both carcinogenic and non-carcinogenic reproductive effects as defined a priori in the PECOs framework, only those with statistically compatible effect measures were eligible for meta-analysis. Specifically, studies reporting standardized incidence ratios (SIRs) for reproductive cancers were pooled quantitatively. In contrast, non-carcinogenic outcomes (semen quality, infertility, endometriosis) exhibited substantial methodological heterogeneity in outcome definitions, biological matrices, and effect metrics (e.g., β-coefficients, odds ratios, correlation coefficients, mean differences), precluding meta-analysis. we performed a narrative synthesis using the Summary of Findings in Meta-analyses (SWiM) framework [40]. This approach avoids the inappropriate visual pooling of incomparable effect sizes and instead presents findings in structured tables that summarize the direction, magnitude, and statistical significance of associations, along with study design and adjustment for confounders. These were therefore synthesized narratively, with key findings visually summarized using forest plots and structured tables in accordance with best practices for qualitative synthesis. We used GraphPad Prism 9© (GraphPad Software Inc., La Jolla, CA) [41] and Excel program to draw plots and made a Table [42].

For carcinogenic reproductive outcomes, the standardized incidence ratio (SIR) was extracted as the primary effect measure from cohort studies. Given that these studies compared cancer incidence in exposed populations to expected rates in the general population, SIR was the most methodologically appropriate and comparable metric. SIRs and their 95% CIs were pooled using a random-effects model with inverse-variance weighting. Case-control and cross-sectional studies that reported odds ratios (ORs) were not pooled with SIRs and were instead summarized narratively.

Given the expected heterogeneity and limited number of studies, we employed a random-effects model for all meta-analyses, using inverse-variance weighting to pool effect sizes. Between-study variance (τ²) was estimated using restricted maximum likelihood (REML), which provides less biased estimates in small meta-analyses. Analyses were conducted in STATA version 17 using the metan command: metan log_sir se_log_sir, random reml lcols(study) xlabel(“SIR”)” where log_sir and se_log_sir are the natural logarithm of the SIR and its standard error, respectively. Heterogeneity between studies was investigated using Cochran’s Q test and quantified by applying the I-squared (I²) statistic [43]. Although subgroup meta-analyses based explicitly on NOS scores were limited by sample size, study quality was indirectly accounted for through leave-one-out sensitivity analyses and subgroup comparisons based on publication year, which often aligns with methodological rigor.

Publication bias was assessed using Begg’s test, Egger’s test, Funnel plots and trim-and-fill method [44–46]. To assess the sources of heterogeneity, the meta-regression analysis was performed using the year of publication as the independent variable and the effect size (ES) as the dependent variable in the model. In this study, the statistical significance level was set at a P-value < 0.05. STATA version 17 (Stata Corp; College Station; TX, USA) were used to analyze the statistical data [47].

Results

Search results

Our systematic search across five major databases, Scopus, PubMed/MEDLINE, Web of Science (ISI), Cochrane Library, and ProQuest, initially identified 4,491 articles. An additional 55 articles were located through manual searches, including reference list screening and grey literature repositories, bringing the total to 4,546 records. After the removal of 2,719 duplicates, 1,827 unique records remained for title and abstract screening. Of these, 1,684 were excluded due to irrelevance, leaving 143 full-text articles assessed for eligibility. Following full-text review, 113 articles were excluded: 56 for not reporting reproductive outcomes and 57 for not addressing PAH exposure. Ultimately, 30 studies met the inclusion criteria and were included in the systematic review [23, 24, 26–32, 48–67]. Of these, 9 studies [29–31, 63–68] were eligible for meta-analysis, based on outcome consistency and sufficient statistical data (Fig. 1).

Studies characteristics

Of the 30 studies included in this review [23, 24, 26–32, 48–67], only 9 studies were eligible for meta-analysis [29–31, 63–68], all of which addressed carcinogenic reproductive outcomes. The remaining studies, primarily focused on non-carcinogenic outcomes such as semen quality, male and female infertility, or hormone levels were excluded from quantitative synthesis due to considerable heterogeneity in effect metrics (e.g., adjusted β-coefficients, odds ratios, correlation coefficients), inconsistent outcome definitions (e.g., use of different WHO thresholds for semen parameters), and variation in exposure assessment methods (urine, blood, semen). Additionally, several studies lacked variance estimates necessary for meta-analytic modeling. These studies were instead synthesized narratively, with key findings visually summarized in forest plots and structured tables. A full listing of excluded studies and the rationale for exclusion is provided in Supplementary File 1, Table S2.

In contrast, all studies included in the meta-analysis reported effect sizes as standardized incidence ratios (SIRs) for carcinogenic outcomes, using statistically compatible formats that allowed for valid quantitative pooling. The primary outcomes across all 30 studies included: (a) carcinogenic reproductive outcomes: prostate, testicular, breast, cervical, uterine, and ovarian cancers and (b) non-carcinogenic reproductive outcomes: semen quality, infertility, and endometriosis. Secondary outcomes involved the identification and quantification of specific PAH metabolites across biological matrices such as urine, semen, and blood.

The studies employed a range of observational designs: (a) cross-sectional (n = 18) [23, 24, 27, 28, 48–61], (b) case-control (n = 6) [26, 29, 31, 32, 62, 63], and (c) retrospective cohort (n = 6) designs [30, 64–67]. The studies were conducted across diverse geographical regions, including China (n = 9) [23, 26–28, 49–52, 54], United States (n = 5) [29, 31, 57, 60, 63], Australia (n = 4) [30, 56, 64, 66], Poland (n = 3) [24, 48, 59], Canada (n = 2) [65, 68], Taiwan (n = 1) [53], Italy (n = 1) [55], India (n = 1) [61], Norway (n = 1) [58], Sweden (n = 1) [67], Egypt (n = 1) [32], and France (n = 1) [62].

Studies addressed both male and female reproductive health, with most male-focused research assessing idiopathic infertility, semen quality parameters (including sperm concentration, motility, morphology, and viability), sperm DNA fragmentation, and hormone levels [23, 24, 26, 30–32, 48–62, 64–67]. Female-focused studies investigated outcomes such as infertility, endometriosis, and reproductive organ cancers [27–30, 63, 65].

PAH exposure was assessed via urinary metabolites, blood samples, semen PAH-DNA adducts, and occupational/residential exposure histories. Urinary biomarkers, particularly 1-hydroxypyrene (1-OHP), 1- and 2-naphthol, and hydroxylated phenanthrenes and fluorenes, were the most frequently used exposure metrics. Table 1 presents a structured overview of the included studies, detailing study characteristics including authorship, publication year, design, country, sample demographics, PAH exposure assessment method, and primary reproductive outcomes assessed.

Table 1.

Characteristics of included studies that examining the relationship between PAH exposure and human reproductive health outcomes

Ref	Year	Study type	Country	Gender (n)	Study population	Biological and agent types	Fertility parameters	Findings
Wu et al. [27]	2024	Cross-sectional	China	Female N: 729	-Infertility women (n = 103) -Fertile women (n = 626)	Urinary PAHs	Infertility	-The associations between PAH metabolite with female infertility, adjusted regression: -Third tertile of 2-OHFLU shows a significant association with female infertility, with (OR: 2.84, 1.24–6.53,P: 0.015) when compared with the first tertile. −1-OHNAP: • Tertile 2 vs. 1: (OR: 0.54, 0.26–1.13, P = 0.100) • Tertile 3 vs. 1: (OR: 1.25, 0.61–2.54, P = 0.536)(P: 0.617) −2-OHNAP: • Tertile 2 vs. 1: (OR: 0.69,0.28–1.68, P = 0.396) • Tertile 3 vs. 1: (OR:1.46,0.43–4.89, P = 0.532)(P: 0.490) −3-OHFLU: • Tertile 2 vs. 1: (OR: 1.20,: 0.62–2.31, P = 0.574) • Tertile 3 vs. 1: (OR: 1.99, 0.87–4.56, P = 0.099)(P: 0.104 −2-OHFLU: • Tertile 2 vs. 1: (OR: 1.57, 0.70–3.55, P = 0.263) • Tertile 3 vs. 1: (OR: 2.84, 1.24–6.53, P = 0.015*)(P: 0.014*) −1-OHPHE: • Tertile 2 vs. 1: (OR: 1.23, 0.59–2.58, P = 0.569) • Tertile 3 vs. 1: (OR: 2.29, 0.92–5.74, P = 0.074)(P: 0.057) −1-OHPYR: • Tertile 2 vs. 1: (OR: 0.61,0.30–1.28, P = 0.185) • Tertile 3 vs. 1: (OR: 1.21, 0.51–2.87, P = 0.658)(P: 0.576) −2-OHPHE & 3-OHPHE: • Tertile 2 vs. 1: (OR: 1.01,0.52–1.97, P = 0.971) • Tertile 3 vs. 1: (OR: 1.47,0.63–3.38, P = 0.358)(P: 0.345)
Zhang et al. [28]	2024	Cross-sectional	China	Female N:1291	Women aged 20–54 years which 90 (6.97%) had endometriosis and 1201 (93.03%) did not have endometriosis.	Eight urinary PAH metabolites	Endometriosis	Concentration of urinary PAH metabolites based on median (IQR): −1-OHNAP: • Total: 1741.00 (762.60, 5981.00) • Non-endometriosis : 1683.40 (758.00, 5483.00) • Endometriosis: 3325.75 (1095.00, 12167.00)(P: 0.004*) −2- OHNAP: • Total: 3434.50 (1525.50, 8717.00) • Non-endometriosis : 3317.10 (1514.00, 8548.70) • Endometriosis: 4787.85 (1911.00, 13118.90)(P: 0.058) −2- OHFLU : • Total: 257.00 (129.00, 594.90) • Non-endometriosis: 247.00 (126.30, 578.20) • Endometriosis: 345.65 (193.10, 1109.50)(P: 0.007*) −3- OHFLU : • Total: 87.00 (42.10, 247.00) • Non-endometriosis: 85.00 (42.10, 232.00) • Endometriosis: 134.50 (53.00, 609.00)(P: 0.015*) −1- OHPHE • Total: 152.00 (82.00, 291.70) • Non-endometriosis : 150.50 (82.00, 284.90) • Endometriosis : 188.15 (93.10, 395.00)(P: 0.044*) −2-OHPHE: • Total: 60.00 (30.20, 126.70) • Non-endometriosis: 59.60 (30.00, 124.70) • Endometriosis : 74.10 (39.00, 165.50)(P: 0.032)
Nayak et al. [61]	2023	Cross-sectional	India	Male N:103	43 fertile semen donors and 60 idiopathic male infertility (IMI)	Seminal PAH exposure	-sperm concentration -Sperm motility -sperm viability -sperm morphology	-Semen parameters in fertile donor vs. IMI; (p-value < 0.0001): Motility(%): 56.30 ± 8.40 vs. 43.42 ± 11.37 Concentration(106/mL):123.33 ± 22.30 vs. 86.77 ± 40.88 viability (%):53.45 ± 10.23 vs. 39.75 ± 9.45 morphology (%):12.5 ± 1.5 vs. 20.5 ± 5.3 -Concentration of PAH ng/mL in the semen of fertile donor and IMI; (p-value < 0.0001): Anthracene: 9.00 ± 19.59 vs. 415.60 ± 327.97 Benzo (A) Anthracene:20.79 ± 66.24 vs. 509.97 ± 486.47 Benzo (A) Pyrene: 0.35 ± 1.17 vs. 43.37 ± 38.57 Benzo (GHI) Perylene: 7.26 ± 26.82 vs. 196.10 ± 291.58 Chrysene: 0.26 ± 1.18 vs. 15.77 ± 22.30 Dibenzo (AH) Anthracene:0.16 ± 1.07 vs. 16.15 ± 20.81 Fluorene: 49.67 ± 169.37 vs. 1857.60 ± 2,381.74 Fluoranthene: 1.09 ± 3.78 vs. 27.58 ± 29.77 Indol (123CD) Pyrene: 0.74 ± 2.13 vs. 7.32 ± 8.00 Napthalene: 199.65 ± 299.65 vs. 1752.88 ± 1728.33 2Bromonapthalene:68.21 ± 185.99 vs. 317.62 ± 385.50 Pyrene: 4.40 ± 11.68 vs. 54.33 ± 54.68 -Impact of the different PAHs on fertility are Anthracene < benzo (a) pyrene < benzo [b] fluoranthene < Fluoranthene < benzo (a) anthracene < indol (123CD) pyrene < pyrene < naphthalene < dibenzo (AH) anthracene < fluorene < 2bromonaphthalene < chrysene < benzo (GH1) perylene as revealed by ROC Curve analysis (AUCROC). Benzo [a] pyrene is invariably present in all infertile patients while naphthalene is present in both groups
Engelsman et al. [56]	2023	Cross-sectional	Australia	Male and female N:29	Occupationally exposure of firefighter(23 men and 6 women) to PAHs;ƸOH-NAP and ƸOH-FLU	Urinary, 1-OH-PYR 1-OHNAP 2-OH-NAP 2-OH-FLU 3-OH-FLU 1-OHPHEN 2-OHPHEN 4-OH-PHEN 9-OHPHEN	Semen parameters and female fertility	- PAH exposure linked to reduced semen quality -increased 1-OH-PYR levels associated with sperm abnormalities. - Firefighters presented with higher sperm concentration (73 million/mL vs. 43 million/mL)and total motility (56% vs. 42%) than the Chinese cohort but were lower for progressive motility (46% vs. 42%), volume (2.0 mL vs. 3.0 mL) and normal forms (9.0% vs. 21%). - Urinary PAH concentrations in female firefighters were associated with changes in ovarian function.
Saad et al. [32]	2019	Case-control	Egypt	Male N:117	GI: healthy fertile males (control = 15); GII: primary idiopathic infertile men patients (case = 51) which includes; GII-a: environmentally exposed(n = 23) GII-b: occupationally exposed(n = 28)	Urine, Semen 1-OHP 1-naphthol 2- naphthol	-Oxidative stress by measuring lipid peroxidation and antioxidant activity of glutathione and glutathione-s-transferase -hormonal activity of FSH, testosterone, prolactin. -semen quality -sperm count -sperm motility	-PAH concentration in GII; (nmol/mmol) based on mean (SD); 1-naphthol: 6 (5.11), 2-naphthol: 19.3 (31.71), 1-OHP: 1.7 (1.68), Total metabolites: 13.6 (23.06). -Glutathione (mg/dl) in the blood mean (SD); G I: 26.8 (3.67), G II-a: 24 (3.02), G II-b: 23.2 (3.05) P (I vs. II-a): 0.048*, P (I vs. II-b): 0.005* -Blood Erythrocytes Glutathione-s-transferase Enzymatic A (U/l); mean (SD) G I: 1.7 (0.28), G II-a: 1.9 (0.38), G II-b: 2.3 (0.43) P (I vs. II-a): 0.620, P (I vs. II-b): 0.001* -Testosterone (ng/dl) Levels; mean (SD) G I: 547.2 (170.06), G II-a: 320.6 (94.54), G II-b: 365.5 (137.87), P (I vs. II-a): 0.001*, P (I vs. II-b): 0.008* Serum FSH (mIU/ml) Levels; mean (SD) G I: 2.7 (0.73), G II-a: 4 (1.92), G II-b: 5.1 (2.71), P (I vs. II-a): 0.189, P (I vs. II-b): 0.006* Prolactin (ng/ml) Levels; mean (SD) G I: 5.5 (1.93), G II-a: 11.1 (3.56), G II-b: 9.1 (4.4), P (I vs. II-a): 0.003*, P (I vs. II-b): 0.015* Sperm count (n × 106), mean (SD) G I: 87.5 (15.45), G II-a: 26.8 (25.78), G II-b: 23.7 (20.7), P (I vs. II-a): 0.003*, P (I vs. II-b): 0.003* Motility (%):G I: 69.5%, G II-a: 29.5%, G II-b: 39.1% P (I vs. II-a): 0.001*, P (I vs. II-b): 0.001* *The data provide strong evidence of a statistical threshold for semen samples containing 30% sperm DNA fragmentation resulting in a reduced level of pregnancy success.
Yang et al. [51]	2017	Cross-sectional	China	Male N: 793	infertile men which includes; 405 men with sperm DNA damage and 366 men with spermatozoa apoptosis	Ten urinary OH-PAH metabolites, including: 1-OHP 1-OHNa 2-OHNa 2-OHFlu 9-OHFlu 1-OHPh 2-OHPh 3-OHPh 4-OHPh 9-OHPh	-Sperm DNA damage parameters: -tail -tail length -tail-distributed moment -comet length	-The median values for tail %, tail length, TDM, and comet length were 33.71%, 68.00 μm, 13.51 μm, and 139.10 μm. The median values for percentages of Annexin V−/PI-, Annexin V+/PI-, and PI + spermatozoa were 74.40%, 3.10%, and 8.80%. -All 10 OH-PAH metabolites were detectable in > 90% of the urine samples. GM concentrations of urinary ΣOHNa were the highest, followed by ΣOHFlu and ΣOHPh. The reproducibility of urinary OH-PAHs was poor (ICCs < 0.40), except for 1-OHNa (ICC = 0.61). -NS associations observed for urinary OH-PAHs with tail % and TDM. However, urinary 9-OHFlu was associated with increased tail length and comet length, with estimated mean increases of 8.65% (2.53–15.03%) and 7.14% (2.33–12.19%), for the highest vs. lowest tertile (FDR-corrected P for trends = 0.05 and 0.01). -suggestive associations between urinary 9-OHPh and ΣOHFlu and increased comet length were observed (both FDR-corrected P for trends = 0.09).
Yang et al. [23]	2017	Cross-sectional	China	Male N: 933	Infertile male	Twelve urinary metabolites of PAHs, including: 1-OHP 1-OHNa 2-OHNa 1-OHPh 2-OHPh 3-OHPh 4-OHPh 9-OHPh 2-OHFlu 9-OHFlu 3-OHBaP 6-OHChr	-semen quality -sperm concentration -sperm count -total motility -progressive motility -semen volume -morphology; -normal and abnormal head -sperm motion; -VSL, VCL, LIN	-Distribution of semen parameters The median sperm count, concentration, progressive motility, total motility, and semen volume were 119.65 million, 43.39 million/mL, 42.22%, 49.12%, and 3.00 mL. -Distribution of PAHs metabolites GM mean concentrations of ΣOHNa were the highest (8.06 µg/L), followed, in decreasing order, by ΣOHPh (6.75 µg/L), ΣOHFlu (5.60 µg/L), and 1-OHP (1.05 µg/L). There were significant correlations between the urinary OH-PAH metabolites (P < 0.05), except for 1-OHP, 3-OHPh, and 9-OHPh. -Associations of the Semen Parameters with Urinary PAH Metabolites −1-OHNa and ΣOHNa were related to a decreased sperm concentration with estimated mean decreases of 22.66% (−34.49%, −8.70%) and 16.47% (−29.39%,−1.19%) for the fourth vs. first quartile (p < 0.05). −1-OHNa and ΣOHPh were associated with a decreased sperm count with estimated mean decreases of 19.99% (−33.83%, −3.34%) and 21.02% (−33.83%, −5.82%)for the fourth vs. first quartile (p < 0.05). -Inverse associations between urinary 4-OHPh and 9-OHPh and semen volume were found with estimated mean decreases of 0.31 mL (−0.61, −0.01) and 0.43 mL (−0.74, −0.12), respectively, for the fourth vs. first quartile (p < 0.05). -no associations between the urinary PAH metabolites and the total sperm motility and progressive sperm motility. −1-OHNa was associated with a decreased percentage of normal morphology (−2.35%; −4.24%, −0.46% for the fourth vs. first quartile; p = 0.046 −9-OHPh was correlated with a decreased sperm VSL (−1.34 μm/sec; −2.37, −0.31 for the fourth vs. first quartile; p = 0.019) and VCL (−2.30 μm/sec; −4.06, −0.54 for the fourth vs. first quartile; p = 0.041
Jurewicz et al. [48]	2016	Cross-sectional	Poland	Male N:194	men who attended infertility clinics	Urinary; PAH (1-OHP)	Semen concentration	-The unadjusted geometric mean of 1-OHP in urine: 0.31 mg/l. -No associations between the semen concentration and 1-OHP (β: 0.09, 95%CI:−0.11 to 0.28, P = 0.385)
Mordukhovich et al. [29]	2016	Case-control	USA	Female N: 3064	case: Residential histories of 1508 participants with breast cancer control :1556 participant’s with no breast	Residential exposure to vehicular traffic (benzo[a]pyrene (B[a]P), as a proxy for traffic-related PAHs)	Brest cancer	-Mean (SD) estimated residential vehicular traffic benzo[a]pyrene exposures were consistently higher among case participants than control participants -Years (1995); Cases: 1.03 (0.62) Controls: 0.97 (0.55) -Years (1960–1990); Cases: 227.42 (125.31) Controls: 196.71 (122.06) -Associations between varying time ranges(percentile) of exposure to benzo[a]pyrene from residential vehicular traffic and breast cancer incidence -Years (1995); 50 to < 75th vs. <50th : (OR: 0.78,0.63, 0.98) 75 to < 95th vs. <50th : (OR: 1.02,0.80, 1.30) ≥ 95th vs. <50th : (OR: 1.06, 0.70, 1.60) -Years (1960–1990); 50 to < 75th vs. <50th : (OR: 0.97,0.66, 1.42) 75 to < 95th vs. <50th : (OR: 0.92, 0.61, 1.39) ≥ 95th vs. <50th : (OR: 1.47, 0.70, 3.08)
Radwan et al. [59]	2015	Cross-sectional	Poland	Male N:181	Infertile male	Environmental exposure to PAHs; urinary 1-1-OHP	Semen quality: -sperm concentration -Total motility -Normal sperm morphology	-PAH concentration: • 1-OHP µg L−1: 0.35 ± 0.35 • 1-OHP µg g−1 creatinine: 0.29 ± 0.27 -Semen quality • concentration (106 ml−1): 50.71 ± 49.33 • Total motility: 56.00 ± 51.30 • Normal sperm morphology: 46.00 ± 78.60 -The association between 1-OHP and semen parameters did not report.
Jeng et al. [53]	2013	Cross-sectional	Taiwan	Male N:51	Nonsmoking coke oven male workers includes: -High exposure group(n = 16) -Low exposure group (n = 20) -control group (n = 15)	Urinary; 16 species of PAHs, and 1-OHP	-Semen concentration -sperm motility -sperm viability -sperm morphology -sperm DNA damage	-The means of all targeted PAH species concentrations in the high exposure group were significantly higher than those in the low exposure group except acenaphthene (P = 0.098), anthracene (P = 0.613), chrysene (P = 0.252), benzo(g, h,i) perylene (P = 0.067), benzo(k) fluoranthene, and dibenzo(a, h) anthracene (P = 0.370). -sperm concentrations, vitality, and DNA fragmentation between the exposed and control groups: NS -The high exposure group experienced significantly lower percentages of normal morphology as compared with the control (P = 0.0001). -Bulky DNA adducts were detected in the exposed group that were significant higher than the control (P = 0.04). -Exposure to PAHs from coke-oven emissions could contribute to increased levels of bulky DNA adducts in sperm.
Ji et al. [52]	2013	Cross-sectional	China	Male N:433	Infertile male	Semen PAH-DNA adducts	-sperm volume -sperm concentration -sperm count -sperm motility -sperm motion parameters VSL, VCL], and linearity [LIN], -DNA Fragmentation	-Total concentration of PAH-DNA adducts:60.5323 ± 22.14607 * Compared with men who had the lowest sperm PAH-DNA adducts category (tertile 1), men with the highest (tertile 4) sperm PAH-DNA adducts level had a suggestive decline in sperm concentration, count, motility and VCL. Trend P-values of sperm concentration, count, motility, and VCL were < 0.001, < 0.001, 0.004 and < 0.001. Aside from suggestively negative associations with these semen parameters, categories of sperm PAH-DNA adducts were associated with a suggestive increasing trend in sperm DNA fragmentation (P for trend < 0.001). Adjusted regression associated with sperm DNA-PAH adducts categories and Seminal volume (ml): tertile 2 vs. 1: β: −0.31 (− 0.74, 0.13) tertile 3 vs. 1: β: −0.18 (− 0.62, 0.25) tertile 4 vs. 1: β: −0.11 (− 0.55, 0.33) Concentration (106/ml): tertile 2 vs. 1: β: −0.33 (− 0.65, −0.01)* tertile 3 vs. 1: β: −0.37 (− 0.70, −0.04)* tertile 4 vs. 1: β: −0.69 (− 1.02, −0.36)* Sperm count (10 6 /ml): tertile 2 vs. 1: β: −0.24 (− 0.60, 0.10) tertile 3 vs. 1: β: −0.36 (− 0.72, 0.01) tertile 4 vs. 1: β: −0.90 (− 1.26, −0.53)* Motility (% motile): tertile 2 vs. 1: β: −4.63 (− 11.16, 1.90) tertile 3 vs. 1: β: −4.04 (− 10.83, 2.76) tertile 4 vs. 1: β: −10.87 (− 17.74, −4.00)* VSL (µm/s): tertile 2 vs. 1: β: −1.30 (− 4.08, 1.47) tertile 3 vs. 1: β: −1.93 (− 4.87, 1.01) tertile 4 vs. 1: β: −5.58 (− 11.26, 0.10) VCL (µm/s): tertile 2 vs. 1: β: −1.62 (− 5.15, 1.91) tertile 3 vs. 1: β: −1.95 (− 5.60, 1.70) tertile 4 vs. 1: β: −8.48 (− 11.97, −5.00)* LIN (%): tertile 2 vs. 1: β: −1.82 (− 5.11, 1.46) tertile 3 vs. 1: β: 0.86 (− 2.45, 4.16) tertile 4 vs. 1: β: −2.89 (− 6.22, 0.44) DNA Fragmentation (%): tertile 2 vs. 1: β: 0.25 (− 0.02, 0.52) tertile 3 vs. 1: β: 0.21 (− 0.07, 0.49) tertile 4 vs. 1: β: 0.48 (0.20, 0.76)*
Song et al. [54]	2013	Cross-sectional	China	Male N:53	Infertile male which includes; -non-Smoker:61.4% -smoker: 38.6% -non-drinker:61.2% -drinker:38.8%	Blood PAHs (16 metabolites)	-semen volume -semen concentration -semen motility	-Concentration: -The top 3 highest PAHs detected were benzo[a]anthracene (4653 ± 3129 ng/g), fluoranthene (2344 ± 2385 ng/g), and benzo[k]fluoranthene (1583 ± 972 ng/g). -The 3 types of R40⁎ PAH concentrations detected in the volunteers’ blood were naphthalene (311 ± 374 ng/g), phenanthrene (234 ± 113 ng/g), and benzo[ghi]perylene (569 ± 365 ng/g). -the 6 types of R45⁎ PAHs detected were benzo[k]fluoranthene (1583 ± 972 ng/g), chrysene (66 ± 48 ng/g), fluoranthene (2344 ± 2385 ng/g), benzo[b]fluoranthene (122 ± 83 ng/g), benzo[a]anthracene (4653 ± 3129 ng/g), and benzo[a]pyrene (335 ± 310 ng/g). Con centration of the sum The Σ16 PAHs was 11,279 ± 6692 (ng/g). -Logistic regression analyses for semen motility with total PAHs in blood; (OR: 0.961,0.852–1.805).
Jurewicz et al. [24]	2013	Cross-sectional	Poland	Male N:277	healthy and under 45 years of age, infertile male	Environmental exposure to PAHs; urinary 1-OHP	-DFI -morphology -motility -VAP -VSL -VCL -BCF	Distribution of semen parameters and the level of 1-OHP in urine (mean ± SD) 1-OHP (µg/l): 0.33 ± 0.3, 1-OHP/creat (µg/g creat): 0.27 ± 0.24, DFI (%): 15.84 ± 10.99, Sperm head abnormalities (%): 29.61 ± 18.26, Sperm neck abnormalities (%): 14.51 ± 8.58, Sperm tail abnormalities (%): 6.25 ± 6.25, Volume (ml): 3.46 ± 1.46, Semen concentration (mln/ml): 49.65 ± 54.02, Motility (%): 56.07 ± 20.61, Static sperm (%):24.78 ± 18.99, Atypical sperm (%):48.44 ± 20.56, VAP (um/s): 52.70 ± 11.32, VSL (um/s): 43.61 ± 10.58, VCL (um/s): 78.34 ± 16.90, ALH (um): 3.55 ± 0.76, BCF (Hz): 26.37 ± 3.81 Association between the level of 1-OHP in urine and the semen parameters DFI: (β: −0.04,−0.16−0.08, P = 0.521) Sperm head abnormalities: (β: 0.61, −2.34−3.55, P = 0.687) Sperm neck abnormalities: (β: 2.12, 0.84–3.40, P = 0.001*) Sperm tail abnormalities: (β: −0.01, −0.19−0.16, P = 0.895) Association between the level of 1-OHP in urine and the semen parameters: Volume: (β: −0.06, −0.01−0.11, P = 0.014*) concentration: (β: 0.09, −0.11−0.28, P = 0.385) Motility: (β: −8.33, −5.07−11.6, P = 0.0001*) Static: (β: 0.16, 95%CI: 0.03–0.29, P = 0.018*) Atypical: (β: 0.38, −3.12−3.88, P = 0.832) VAP: (β: 0.88, −0.94−2.71, P = 0.344) VSL: (β: 0.71, −1.01−2.42, P = 0.419) VCL: (β: 0.98, −1.64−3.61, P = 0.463) BCF: (β: −0.05, −0.67−0.56, P = 0.868)
Xia et al. [50]	2009	Cross-sectional	China	Male N:542	Men with wives not diagnosed as infertile	Urinary, four PAH metabolites, 1-N 2-N 1-OHP 2-OHF	-Semen quality -semen volume -sperm concentration -sperm number per ejaculum -sperm motility	-median Creatinine (CR)-adjusted concentrations of 1-N, 2-N, 1-OHP, 2-OHF were 2.35, 4.05, 1.14, 2.89 µg/g of CR. -Men in the highest quintiles (quintile 5 vs.1) of 1-OHP exposure were more likely to have below-average sperm concentration (OR: 2.17, 1.04–4.53, P = 0.028) and sperm count (OR: 2.13, 1.03–4.41, P = 0.021) compared to those in the lowest quintiles. -NS differences observed for other metabolites affecting semen quality. −1-OHP levels in control subjects (1.16 µg/g of CR) were lower than in those with below-average sperm concentration (1.30 µg/g of CR), below-reference sperm number per ejaculum (1.40 µg/g of CR) and below-reference sperm motility (1.21 µg/g of CR).
Xia et al. [26]	2009	Case-control	China	Male N:1299	Case: infertile male (n = 513) Case I: idiopathic infertile men with normal semen quality (n = 291) Case II: idiopathic infertile men with abnormal semen quality (n = 222) Control: fertile men (n = 273)	Urinary concentrations of four PAH metabolites, including : 1-N 2-N 1-OHP 2-OHF, which were adjusted by urinary creatinine (CR)	Men fertility	-ΣPAH levels in participants urine; mean (95%CI): Control group: 10.94 (10.09–11.87) Case I group: 12.01 (11.09–13.00) Case II group: 12.53 (11.44–13.73) -The mean CR-adjusted concentrations of 1-N, 2-N, 1-OHP, 2-OHF and Sum PAH metabolites of the control group were lower than those of the case groups. −2-N, 1-OHP, 2-OHF and ΣPAH metabolites means of the Case II group were higher than Case I group. AORs for the relationships between idiopathic male infertility and CR-adjusted PAH metabolite tertiles: 1-N: tertile 2 vs. 1: (OR: 1.05, 0.73–1.51, P = 0.582) tertile 3 vs. 1: (OR: 1.11, 0.77–1.59, P = 0.582) 2-N: tertile 2 vs. 1: (OR: 1.20, 0.84–1.71, P = 0.082) tertile 3 vs. 1: (OR: 1.38, 0.96–1.98, P = 0.082)
De Fleurian et al. [62]	2009	Case-control	France	Male N:402	Case: patients with impaired sperm quality (n = 314) Control: patients with normal sperm quality (n = 88)	Self-reported physical or chemical occupational exposure to PAHs	-Semen quality -sperm count -sperm motility -sperm morphology	-The mean exposure index to PAH was significantly higher in men with altered semen than in men with normal semen (26.9 ± 23.7 vs. 16.8 ± 13.9, P = 0.016) -In adjusted regression for age and known risk factor variables, men with altered semen were significantly more exposed to PAH than men with normal semen: (OR: 1.9,1.1–3.5; P = 0.026) -higher risk of oligospermia (OR: 1.6,1.03–2.6; P = 0.038) and total sperm count impairment (P = 0.010) in subjects exposed to PAHs. -no significant association between teratospermia (OR: 1.4, 0.9–2.1, P = 0.155) and sperm morphology (P = 0.069)
Gun et al. [64]	2006	Retrospective Cohort	Australia	Male N:16,547	employment in the petroleum industry (n = 16,547)vs.general Australian population (not reported the number of cases)	Occupational exposure to petroleum industry	Incidence rate of prostate and testicular cancers standardized incidence ratios (SIRs)	-Prostatic cancer was the commonest cancer in exposed males (251 cases), and the incidence was significantly elevated (SIR:1.18: 1.04–1.34). -correlation was discovered between prostate cancer and type of work, decade of hire, employment duration, time since hire, or hydrocarbon exposure ranking -There were 34 cases with testicular cancer in exposed males, the incidence was not significantly increased (SIR:1.33: 0.80–2.08).
Rybicki et al. [31]	2006	Case-control	USA	Male N:881	Men with prostate cancer with respiratory exposure to petroleum (n = 637) vs.men without cancer with occupational respiratory exposure to petroleum (n = 244)	Occupational exposure to PAH	Prostate cancer risk associated with occupational PAH exposure,	-No difference was discerned between respiratory exposure PAH and prostate cancer (OR:1.17: 0.76–1.81; P = 0.47).Results based on PAH types: (A) Petroleum; (OR: 1.12: 0.73–1.73, P = 0.61) (B) Coal; (OR: 1.29: 0.73–2.30, P = 0.39) (C): Wood: (OR: 0.86: 0.36–2.07, P = 0.74) (D) Others: (OR: 0.79: 0.51–1.23, P = 0.30) -No difference was discerned between cutaneous PAH exposure and prostate cancer (OR: 0.76:0.50–1.17; P = 0.22).Results based on PAH types; (A) Petroleum; (OR: 0.74: 0.48–1.13, P = 0.16) (B) Coal; (OR: 1.48: 0.68–3.20,P = 0.32) (C): Wood: (OR: 0.97: 0.24–3.87, P = 0.97) (D) Others: (OR: 0.77: 0.37–1.60, P = 0.48)
Bonner et al. [63]	2005	Case-control	USA	Female N: 3271	Cases(n = 1166):women with histologically confirmed, primary, incident breast cancer. Controls (n = 2105): frequency matched by age, race, and county of residence to cases	Air monitors; occupational and residential exposure to PAHs in early life	-Reproductive organ cancer -Brest cancer	-Exposure to high concentration of mixture of PAHs at the time of birth was associated with an increase in the OR for pre-menopausal women; however, there was no exposure-response relationship and the P for trend was not significant. -Premenopausal: 84–114 vs. <84 (µg/m3): (OR: 1.96, 0.64–3.01) 115–140 vs. <84 (µg/m3): (OR: 2.23, 0.77–6.44) >140 vs. <84 (µg/m3): (OR: 1.78, 0.62–5.10)(P: 0.38) -Exposure to high concentration of mixture of PAHs at the time of birth was associated with an increase in the OR for post-menopausal women; however, there was no exposure-response relationship but the P for trend was significant (P = 0.01). -Postmenopausal: 84–114 vs. <84 (µg/m3): (OR: 2.32, 0.89–6.10) 115–140 vs. <84 (µg/m3): (OR: 1.94, 0.77–4.86) >140 vs. <84 (µg/m3): (OR: 2.42, 0.97–6.09)(P: 0.01*)
Gun et al. [30]	2004	Retrospective Cohort	Australia	Male N: 15,956 Female N:867	employment in the petroleum industry vs.general Australian population (not reported the number of cases)	Occupational exposure to petroleum industry	Incidence rate and mortality rate of: -prostate cancer -testicular cancer -cervical cancer -breast cancer by SIRs	- incidence of prostate cancer was significantly elevated in the exposed group; (SIR: 1.19; 1.00-1.40). - incidence rate of testicular cancer was not significantly different in the exposed group compared with the general population(SIR:1.24: 0.68–2.08) -incidence of cervical cancer (SIR: 1.61; 0.33 to 4.71) and breast cancer (SIR: 1.02; 0.53 to 1.79)were not significantly elevated in the exposed group
Meeker et al. [60]	2004	Cross-sectional	USA	Male N:272	Infertile male	Urine, Environment exposure to 1-N	-semen quality -sperm concentration -sperm motility -sperm morphology -motion parameters -VSL, VCL, LIN	-Concentration of 1-N in urine : SG-adjusted (mean): 3.13 µg/L SG-adjusted (median): 3.19 µg/L -For increasing 1-N tertiles, AORs were significantly elevated for below-reference sperm concentration (OR (95%CI) for low, medium, and high tertiles = 1.0, 4.2 (1.4–13), 4.2 (1.4–12.6), P = 0.01) and sperm motility (1.0, 2.5 (1.3–4.7), 2.4 (1.2–4.5); P = 0.01). -sperm morphology was not significantly related (1.0, 1.4 (0.6-3), 1.6 (0.8–3.5); P = 0.20). -sperm motility and VSL were significantly decreased with exposure to 1-N; Motility; β: −3.87 (−7.28, −0.45)* VSL; β: −1.64 (−2.99, −0.27)* -NS association between others semen quality parameter and 1-N; Concentration; β: 0.84 (0.71, 1.01) Morphology; β: −0.15 (−0.79, 0.49) VCL; β: −1.98 (−4.33, 0.35) LIN; β: −0.79 (−1.79, 0.22)
Lewis et al. [65]	2003	Retrospective Cohort	Canada	Male N: 17,230 Female N:8062	Canadian petroleum industry men and female workers vs. Incidence rates of general population	Occupationally exposed; employed in the Canadian petroleum industry	-Prostate cancer -testicular cancer -Cervical cancer -Uterine cancer -Ovarian cancer -fallopian tube cancer -broad ligament cancer -breast cancer	No difference was found in case of prostate (SIR = 0.67; 0.41–1.03) and testicular cancers (SIR 0.82; 0.45–1.37) between petroleum company workers and the general population -decreased rates of cervical (SIR: 0.42; 0.17 to 0.86) and uterine (SIR: 0.31; 0.06 to 0.89) cancers in occupationally exposed female petroleum employees compared with national rates. -no difference in ovarian, fallopian tube, or broad ligament cancer incidence (grouped) between exposed female employees and the general population (SIR 1.40; 0.78–2.30). -no difference between petroleum employees and general population in breast cancer(SIR 1.02; 0.80–1.28).
Gaspari et al. [55]	2003	Cross-sectional	Italy	Male N:182	Infertile male; -current smokers: 39% -current drinkers: 72%	Semen PAH-DNA adducts	-Sperm count -Physiologic forms -Abnormal head and neck morphology -Abnormal tail morphology	-PAH-DNA adducts (mean ± S.D.) exposure: Fertile vs. infertile: (1.51 ± 0.62 vs. 1.82 ± 0.61, P < 0.05) -Sperm morphology (Correlation coefficient): • Sperm count: − 0.09 (P = 0.2) • Physiologic forms: − 0.18 (P = 0.016) • Abnormal head morphology: 0.30 (P = 0.0001) • Abnormal neck morphology: − 0.21 (P = 0.009) • Abnormal tail morphology: − 0.10 (P = 0.1)
Wang et al. [49]	2001	Cross-sectional	China	Male N:198	Four groups of workers at a petrochemical complex; (A) exposed workers and smokers (n = 45); (E/S) (B) exposed workers and nonsmokers (n = 23); (E/NS) (C) unexposed workers and smokers (n = 81); (NE/S) (D) unexposed workers and nonsmokers (n = 49) (NE/NS)	Occupational exposure; workers in the petrochemical plant as exposed to petrochemical agents, whereas the office staff employees in the fabric factory served as NE subjects.	-Sperm concentration -Sperm count -Sperm motility -Sperm viability	−31% decrease in the sperm concentration of the E/S group (41.49 million/mL) compared with the NE/NS group(60.07 million/mL, P < 0.01). -no difference was seen in relation to the E/NS group (52.52 million/mL) or the NE/S group (55.32 million/mL). -sperm concentration was negatively correlated to years of smoking and years of exposure combined(r = − 0.28; P < 0.05) -approximate decrease of 24–29% in the sperm counts of ES compared with the NE groups: • NE/NS = 152 million/ejaculate; • NE/S = 141.86 million/ejaculate; • E/NS = 127.02 million/ejaculate; • ES = 108.48 million/ejaculate P < 0.05) −16% lower sperm motility for NS(P < 0.05) and 18% lower for smokers (P < 0.01) -no difference in sperm viability between petrochemical-exposed workers (NS = 63.41%, smokers = 60.78%) and NE workers (NS = 61.44%, smokers = 60.28%)
Bull et al. [58]	1999	Cross-sectional	Norway	Male N:81	Offshore operators (non-exposed, n = 51) vs. offshore mechanics (exposed, n = 30)	Occupational exposure to offshore oil processes among men	Men fertility	-no effect on fecundity ratios.Fecundity Ratio for offshore mechanics = 0.8 (0.49–1.32) -no effect on fecundity ratios.Fecundity Ratio for offshore drilling personnel = 0.89(0.61–1.29)
Järvholm et al. [67]	1997	Retrospective Cohort	Sweden	Male N: 4128	Swedish petroleum industry workers Vs. incidence rates of general population	Occupationally exposed men employed in the Swedish petroleum industry	Prostate cancer	-no increased risk of cancer in the exposed group compared with the general population (SIR: 1.1; 95% CI: 0.78–1.5)
Christie et al. [66]	1991	Retrospective Cohort	Australia	Male N: 15,000	employment in the petroleum industry (n = 15,000)vs.general Australian population(not reported the number of cases)	Occupational exposure to petroleum industry	Incidence rate of prostate and testicular cancers, by means of standardized mortality ratios (SMRs) and SIRs	-no difference in prostate cancer (SIR: 1.0: 0.4–1.9) and testicular cancer (SIR: 1.0: 0.2–2.8) incidence between a cohort of over 15,000 Australian petroleum industry employees and the national rates
Schechter et al. [68]	1989	Retrospective Cohort	Canada	Male N: 1126	Rural residents of Southern Alberta, Canada living near natural gas refineries vs. Expected cancer rates of three rural areas in Southern Alberta selected that were socio-demographically similar to cohort area	Residential exposure; rural population living downwind from natural gas refineries	Prostate cancer	No difference was found between the exposed and reference populations for prostate cancer (SIR: 1.76; 95% CI: 0.84–4.38).
Rosenberg et al. [57]	1985	Cross-sectional	USA	Male (A):116 (B):113	Men who worked in the wastewater treatment plant of a petroleum refinery during the six months before this study (n = 42 vs.39) vs. Other petroleum refinery workers (n = 74)	Occupational exposure by working at wastewater treatment plant of a petroleum refinery	-Sperm concentration - Sperm morphology	-The sperm concentration of exposed individuals was approximately 17% lower than that of unexposed individuals (unexposed = 80.8 million/mL; exposed = 66.9 million/mL). However, this reduction was not significant (P = 0.16). -The mean proportion of abnormally shaped sperm was similar in both unexposed and exposed groups(49.1% and 44.5%, P = 0.94) and was not altered by adjustment for abstinence or other factors. -no correlation with hours worked in the past 6 months and sperm morphology (P = 0.08)

PAH: Polycyclic Aromatic Hydrocarbons; 1-OHNa: 1-Hydroxynaphthalene; 2-OHNa: 2-Hydroxynaphthalene; 1-OHP: 1-Hydroxypyrene;2-OHF: 2-Hydroxyfluorene; 1-OHph: 1- Hydroxyphenanthren; 1-OHPy: 1-hydroxypyrene; FSH: follicle stimulating hormone; BPDE: Benzo(A)Pyrene Diolepoxide, CI: Confidence Interval, DFI: DNA Fragmentation Index, DNA: Deoxyribonucleic Acid, G: Group, LIN: Linearity, L: liter, LOD: Limit of Detection,, N: Number, Na: naphthalene, NS: Not Significant, OR: Odds Ratio, VCL: Curvilinear Velocity, VSL: Straight-Line Velocity; GM: geometric mean; VAP: velocity average path; BCF: beat cross frequency; 1-N :1-hydroxynapthalene; 2-N :2-hydroxynapthalene; AOR: Adjusted ORs; SG: specific gravity; SIRs: standardized incidence ratios; 2-OHFLU :2-hydroxyfluorene; 1-OHNAP:1-hydroxynaphthalene; 1-OHPHE:1-hydroxyphenanthrene; 1-OHPYR:1-hydroxypyrene; ƸOH-NAP: Sum hydroxy-naphthalene; ƸOH-FLU: sum hydroxy-fluorene

Risk of bias within included studies

The NOS was used to assess the risk of bias in the included studies (Supplementary File 1, Tables S3, S4 and S5). Based on AHRQ standards, study quality was classified as good (≥ 6 points), fair (4–5 points), or poor (≤ 3 points). Of the 30 included studies, 19 were rated as good quality and 11 as fair quality. No studies were rated as poor. Among the case-control studies (n = 6), four were rated as good quality [26, 29, 32, 62], and two as fair quality [31, 63]. For cohort studies (n = 6), two were rated as good quality [30, 64], and four as fair quality [65–68]. Of the 18 cross-sectional studies, 13 were rated as good quality [23, 24, 27, 28, 48–56], and five as fair quality [57–61].

Although several fair-quality studies were included in the meta-analyses, we applied a series of methodological safeguards to ensure that their inclusion did not compromise the validity of the findings: Sensitivity Analyses: We performed leave-one-out sensitivity analyses for each meta-analysis to evaluate the robustness of the pooled effect estimates. These analyses demonstrated that the exclusion of any single study, including those rated as “fair”, did not materially alter the overall results. Subgroup Analyses: While subgroup analyses based specifically on NOS scores were limited by sample size, we conducted subgroup comparisons by publication year, which served as a reasonable proxy for study quality in many cases. These analyses yielded consistent findings, further supporting the stability of our results.

In addition, all meta-analyzed studies underwent further appraisal using the GRADE framework. The certainty of evidence was rated as moderate for male reproductive organ cancers (starting from low due to observational design, upgraded for strong association and consistency) and low for female reproductive organ cancers (starting from low, further downgraded for serious imprecision), based on GRADE assessment (Supplementary File 1, Tables S6 and S7). These ratings reflect the observational design of the studies, the presence of residual confounding, and moderate heterogeneity across studies. Together, these quality assessment measures and sensitivity checks support the inclusion of fair-quality studies in the meta-analyses and provide confidence that our conclusions are both methodologically sound and appropriately cautious.

Non-carcinogenic reproductive outcomes

In the analysis of non-carcinogenic reproductive outcomes, a total of twenty-one studies were reviewed [23, 24, 26–28, 32, 48–62]. For non-meta-analyzed outcomes, visual forest plots were used only when effect measures were of the same type (e.g., all ORs or all β-coefficients), and no plots combined heterogeneous metrics. Where multiple metrics were reported, results are presented in separate panels to preserve interpretability. Among these, seventeen studies [23, 24, 32, 48–57, 59–62]. focused on the impact of PAHs metabolite exposure on semen characteristics in males, while two studies [26, 58], investigated infertility outcomes. Saad et al., [32] provided insights into both semen characteristics and infertility associated with PAHs metabolites. For females, two studies examining the association between PAHs metabolite exposure and non-carcinogenic female reproductive health outcomes [27, 28]. The extracted data from each of these 23 articles, segregated by outcome, is available in a separate sheet in Supplementary 2.

Semen characteristics

Eighteen studies investigated the relationship between PAHs metabolites with semen characteristics. These characteristics included semen concentration (12 studies) [23, 24, 48, 50, 52, 53, 56, 59, 61], sperm count (7 studies) [23, 32, 49, 50, 52, 55, 62], sperm motility (12 studies) [23, 24, 32, 49, 50, 52–54, 56, 59–61], sperm viability (3 studies) [49, 53, 61], sperm morphology (9 studies) [23, 24, 53, 55, 57, 59–62], semen volume (5 studies) [23, 24, 50, 52, 54], sperm DNA damage (comet assay parameters: tail, tail distributed moment, and tail length) (4 studies) [24, 51–53], and sperm motion parameters (straight line velocity, curvilinear velocity, linearity, beat cross frequency, and average path velocity) (4 studies) [23, 24, 52, 60]. Fifteen of these studies were cross-sectional [23–25, 48–57, 59–61], and two were case-control [32, 62]. PAHs levels were measured in semen (7 studies) [26, 49, 52, 55, 57, 61, 62], urine (5 studies) [27, 28, 48, 53, 56], blood (one study) [54] or both of urine and semen (7 studies) [23, 24, 26, 32, 51, 59, 60]. Twelve studies assessed environmental exposure [23, 24, 48, 50–52, 54, 55, 59–61], while six focused on occupational exposure [32, 49, 53, 56, 57, 62]. Semen parameters were assessed based on World Health Organization (WHO) reference values for semen volume (< 2 mL), sperm concentration (< 20 × 10⁶/mL), sperm count per ejaculum (< 40 × 10⁶), sperm motility (< 50% motile sperm), and sperm morphology (< 4% normal morphology) [69].

PAHs exposures and semen concentration

The association between PAHs exposure and semen concentration was reported in 12 studies [24, 48–53, 56, 57, 59–61]. Out of these, six studies found a significant inverse association [23, 49, 50, 52, 60, 61]. Wang et al. [49], reported a statistically significant difference in semen concentration between smokers occupationally exposed to PAHs (n = 45) and non-exposed/non-smokers (n = 49) (41.49 ± 1.87 vs. 60.07 ± 1.8, P = 0.001), as well as between exposed/smokers (n = 45) and non-exposed/smokers (n = 81) (41.5 ± 1.8 vs. 55.3 ± 1.7, P = 0.04). However, no significant difference was found between exposed/smokers (n = 45) and exposed/non-smokers (n = 23) (41.5 ± 1.8 vs. 52.5 ± 2.1, P = 0.08). Nayak et al. [61], reported significantly higher PAH concentrations and lower semen concentrations in idiopathic infertile patients (n = 60) compared to fertile donors (n = 43) (86.7 ± 40.8 vs. 123.3 ± 22.3, P = 0.001).

Three studies specifically investigated the relationship between semen concentration changes and PAHs levels [50, 52, 60]. Meeker et al. [60], found an increased risk of decreased semen concentration (< 20 × 10^6/mL) in those with medium (OR: 4.2, 95% CI: 1.4–13, P = 0.01) and high (OR: 4.2, 95% CI: 1.4–12.6, P = 0.01) exposure to 1-naphthol (1-N) compared to low exposure. Xia et al. [50], reported a higher risk of reduced semen concentration based on 1-hydroxypyrene (1-OHP) levels in tertile-4 compared to tertile-1 (OR: 2.17, 95% CI: 1.04–4.53, P = 0.028). Ji et al. [52], showed a significant inverse correlation between PAH-DNA adduct levels and semen concentration, with tertile-2 vs. tertile-1 (β: −0.33), tertile-3 vs. tertile-1 (β: −0.37), and tertile-4 vs. tertile-1 (β: −0.69). Additionally, Yang et al. [23], found a significant inverse relationship between 1-hydroxynaphthalene (1-OHN) and Σ-OHN (sum of 1-OHNa and 2-OHNa) with sperm concentration.

PAHs exposures and sperm count

Seven studies investigated the relationship between PAHs metabolite exposure on sperm count per ejaculum [23, 32, 49, 50, 52, 55, 62]. Overall, six studies concluded that PAHs metabolite exposure was associated with reduced sperm count [23, 32, 49, 50, 52, 62]. Wang et al. [49], demonstrated a significant reduction in mean sperm count among smokers occupationally exposed to PAHs compared to non-exposed/non-smokers and exposed/smokers compared to non-exposed/smokers (108.48 ± 2.05 vs. 152.08 ± 1.87, P = 0.001; 108.48 ± 2.05 vs. 141.86 ± 1.94, P = 0.04), respectively. However, no statistically significant difference was observed between exposed/smokers and exposed/non-smokers (108.48 ± 2.05 vs. 127.02 ± 2.16, P = 0.08) in sperm count. Saad et al. [32], showed a highly significant decrease in sperm count for two idiopathic infertile patients environmentally or occupationally exposed to 1-Hydroxypyrene (1-OHP), 1-2-naphthol compared to healthy fertile males (26.8 ± 25.7 vs. 87.5 ± 15.4, P = 0.003; 23.7 ± 20.1 vs. 87.5 ± 15.4, P = 0.003). De Fleurian et al. [62], found a higher risk of reduced sperm count in patients with impaired sperm quality compared to those with normal sperm quality (OR: 1.6, 95%CI: 1.03–2.6, P = 0.038). Additionally, Yang et al. [23], found a significant inverse relationship between 1-OHN and Σ-OHN concentrations with sperm count in 933 infertile men.

Two studies examined the association between changes in sperm count and PAHs concentrations [50, 52]. Xia et al. [50], reported a significantly higher risk of reduced semen count (< 40 per ejaculum) related exposure to urinary 1-OHP for individuals in tertile-5 compared to those in tertile-1 (OR: 2.13, 95%CI: 1.03–4.41, P = 0.021). Ji et al. [52], reported a statistically significant inverse correlation between the amount of PAH-DNA adduct and sperm count in individuals in tertile-4 versus tertile-1 (β: −0.90, P < 0.05). However, there was no statistically significant correlation between tertile-2 versus tertile-1 (β: −0.24, P >0.05) or tertile-3 versus tertile-1 (β: −0.36, P >0.05).

PAHs exposures and sperm motility

Among the studies we reviewed, 12 assessed the relationship between PAHs metabolites and sperm motility [23, 24, 32, 49, 50, 52–54, 56, 59–61]. Wang et al. [49], demonstrated a significant reduction in mean sperm motility among smokers occupationally exposed to PAHs compared to non-exposed/non-smokers and exposed/smokers (2.01 ± 0.6 vs. 2.41 ± 0.7, P = 0.01; 2.01 ± 0.6 vs. 2.45 ± 0.6, P = 0.01), respectively. However, no statistically significant difference was observed between exposed/smokers and exposed/non-smokers (2.01 ± 0.6 vs. 2.02 ± 0.63, P = 0.08) in sperm motility. Saad et al. [32], showed a highly significant decrease in sperm motility for idiopathic infertile patients environmentally or occupationally exposed to 1-OHP and 1-2-naphthol compared to healthy fertile males. Nayak et al. [61], reported a significantly higher concentration of PAHs and lower sperm motility in idiopathic infertile patients vs. fertile donors (43.4 ± 11.3 vs. 56.3 ± 8.4, P < 0.001). Engelsman et al. [56], found that firefighters occupationally exposed to PAHs had higher levels of total and progressive sperm motility compared to the general population. Meeker et al. [60], found that individuals with medium (OR: 2.5, 95% CI: 1.3–4.7, P = 0.01) and high (OR: 2.4, 95% CI: 1.2–4.5, P = 0.01) levels of exposure to 1-naphthol (1-N) had a higher risk of reduced sperm motility (< 50% motile) compared to those with low exposure levels. Moreover, exposure to the 1-OHP metabolite showed a statistically significant inverse correlation with sperm motility (β: −8.33, P < 0.001) [24]. Ji et al. [52], demonstrated a significant correlation between the amount of PAH-DNA adduct in subjects in tertile-4 compared to tertile-1 with sperm motility (β: −10.9, P < 0.05). Jeng et al. [53], found significant correlations between Benzo(b)fluoranthene (β: −0.21, P = 0.045), Benzo(g, h,i)pyrene (β: −0.21, P = 0.045), Benzo(k)fluoranthene (β: −0.23, P = 0.044), Dibenzo(a, h)anthracene (β: −0.20, P = 0.049), and Naphthalene (β: −0.22, P = 0.021) with sperm motility in non-smoking coke oven workers.

PAHs exposures and sperm viability

The association between exposure to PAHs metabolites and sperm viability was reported in three studies [49, 53, 61], three of which concluded that exposure to PAHs metabolites could be associated with reduced sperm viability [25, 53, 61]. Nayak et al. [61]., demonstrated a significantly higher concentration of PAHs and lower sperm viability in idiopathic infertile patients (n = 60) compared to fertile donors (n = 43); (39.7 ± 9.4 vs. 53.4 ± 10.2, P < 0.001). Jeng et al. [53], found significant correlations between Benzo(b) fluoranthene (β: −0.23, P = 0.043), Benzo(g, h,i)pyrene (β: −0.21, P = 0.045) and Benzo(k)fluoranthene (β: −0.21, P = 0.041) with sperm viability in non-smoking coke oven workers.

PAHs exposures and sperm morphology

Nine studies evaluated the association between PAHs metabolites and sperm morphology [23, 24, 53, 55, 57, 59–62]. Gaspari et al. [55], found a significant correlation between PAH-DNA adducts and abnormal morphology of the sperm head (β: 0.3, P < 0.001), neck (β: −0.21, P = 0.009), and physiologic forms (β: −0.18, P = 0.016) in infertile men. Jurewicz et al. [24], showed a significant correlation between exposure to 1-OHP/creatinine (µg/g creat) and sperm neck abnormalities (β: 2.12, P = 0.001). Yang et al. [23], reported a significant inverse correlation between 1-OHNa and normal morphology (β: −2.35, P = 0.046). Nayak et al. [61], showed that the percentage of abnormal morphology was significantly higher in idiopathic male infertility (IMI) than fertile donor due to higher exposure to PAH (P < 0.001). Jeng et al. [53], found a significant correlation between Anthracene (β: 0.503), Benzo(a)anthracene (β: −0.501), Benzo(a)pyrene (β: −0.516), Benzo(b)fluoranthene (β: −0.516), Benzo(g, h,i)pyrene (β: −0.557), Benzo(k)fluoranthene (β: −0.528), Chrysene (β: 0.536), Dibenzo(a, h)anthracene (β: −0.532), Fluoranthene (β: −0.454), Fluorene (β: 0.468), Indeno(1,2,3-cd)pyrene (β: −0.524), Naphthalene (β: −0.276), and 1-OHP (β: −0.281), with normal sperm morphology.

PAHs exposures and semen volume

The association between PAHs metabolites and semen volume was evaluated in five [23, 24, 50, 52, 54], summarized in Fig. 2 A , B. Xia et al. [50], focused on adjusted ORs for below-reference semen volume (< 2 mL) across different tertiles of creatinine-adjusted urinary PAHs metabolites in 542 men (Fig. 2A: ORs adjusted between PAH metabolites and semen volume). The examined metabolites included 1-N, 2-hydroxynaphthalene (2-N), 1-OHP, and 2-hydroxyfluorene (2-OHF). None of these metabolites showed statistically significant associations with reduced semen volume, with ORs ranging from 0.64 to 2.46. The other four studies presented adjusted regression coefficients (β) for changes in semen volume in relation to specific PAHs metabolites (Fig. 2B: adjusted regression coefficients between PAH metabolites and semen volume) [23, 24, 52, 54]. Significant negative associations were observed for 4-hydroxyphenanthrene (4-OHP, β = −0.31), 9-hydroxyphenanthrene (9-OHP, β = −0.43), and 1-OHP (β = −0.19) [23, 24]. Additionally, the sum of 16 PAHs (Σ16 PAHs) indicated a statistically significant negative association (β = −0.17) with semen volume [54]. However, Ji et al. [52], analyzed PAH-DNA adducts across three tertiles, showing β values from − 0.31 to −0.11, none of which were statistically significant.

Fig. 2.

Narrative summary of associations between PAH metabolites and semen volume from five studies. (A) Forest plot of adjusted odds ratios (ORs) for below-reference semen volume (<2 mL) by quintiles of PAH exposure. (B) Forest plot of adjusted regression coefficients (β) for change in semen volume per unit increase in PAH metabolites. Effect measures are not pooled across panels due to differing scales and interpretations. Studies included are those reporting either ORs or β-coefficients, but not both in the same model. Abbreviation; 1-N (1-hydroxynaphthalene), 2-N (2-hydroxynaphthalene), 1-OHP (1-hydroxypyrene), 2-OHF (2-hydroxyfluorene), 4-OHP (4-hydroxyphenanthrene), 9-OHP (4-hydroxyphenanthrene).

PAHs exposures and sperm DNA damage

The association between exposure to PAHs metabolites and sperm DNA damage was reported in four studies [23, 24, 52, 53], two of which found a statistically significant associations [23, 52]. A study by Ji et al. [52], found there was a significant negative correlation between semen PAH-DNA adducts and sperm DNA fragmentation (P < 0.001). Higher PAH-DNA adduct levels were associated with increased DNA fragmentation. Tertile-4 had the highest DNA fragmentation compared to tertile-1 (β: 0.48, 95% CI: 0.20, 0.76). Additionally, Yang et al. [23], found an association between OH-PAH metabolites and sperm DNA damage parameters.

PAHs exposures and sperm motion parameters

The association between exposure to PAHs metabolites and sperm motion parameter was reported in four studies [23, 24, 52, 60]. A study by Meeker et al. [60], reported that environmental exposure to 1-N was significantly associated with reduced straight-line velocity (VSL) of sperm motion (β: −1.64, 95% CI: −2.99, −0.27). However, there was no statistically significant association with curvilinear velocity (VCL) or linearity (LIN) (P > 0.05). Jurewicz et al. [24], found no significant associations between urinary 1-OHP, and various sperm motion parameters, including VSL, velocity average path (VAP), VCL, and beat cross frequency (BCF) (P > 0.05). A study by Ji et al. [52], showed that PAH-DNA adducts did not significantly influence on sperm motion parameters, including VSL, VCL, and LIN, across tertiles (P > 0.05), except for VCL in tertile-4, which showed a statistically significant reduction in VCL compared to tertile-1 (β: −8.48, 95% CI: −11.97, −5.00). Yang et al. [23], reported that urinary 9-OHPh was correlated with decreased VSL (β: −1.34, 95% CI: −2.37, −0.31, P = 0.019) and VCL (β: −2.30, 95% CI: −4.06, −0.54, P = 0.041) in infertile males.

PAHs exposures and male infertility

The association between exposure to PAH metabolites and male infertility was reported in three studies [26, 32, 58]. Bull et al. [58], found no statistically significant relationship with fecundity ratios in offshore oil workers exposed to PAHs compared to non-exposed groups. Xia et al. [26], investigated various PAH metabolites, such as 1-N, 2-N, 1-OHP, and 2-OHF across different tertiles and their association with idiopathic male infertility. They found no statistically significant association between male infertility with 1-N, 2-N and the sum of 1-N and 2-N across different tertiles. However, higher urinary levels of 1-OHP (P-trend = 0.034), 2-OHF (P-trend = 0.022) and ΣPAH metabolites (P-trend = 0.022) were linked to increased risk of idiopathic male infertility. Additionally, Saad et al. [32], observed a notable decrease in testosterone levels in primary idiopathic infertile men compared to healthy fertile males, both in environmental (547.2 ± 170.1 vs. 320.6 ± 94.5, P = 0.001) and occupational (547.2 ± 170.1 vs. 365.5 ± 137.8, P = 0.008) exposure settings to 1-OHP, 1-N, 2-N.

PAH exposures and female infertility

The association between exposure to PAH metabolites and female infertility was reported in two studies [27, 28]. Wu et al. [27], investigated various PAH metabolites, such as 1-hydroxynaphthalene (1-OHNAP), 2-hydroxynaphthalene (2-OHNAP), 3-hydroxyfluorene (3-OHFLU), 2-hydroxyfluorene (2-OHFLU), 1-hydroxyphenanthrene (1-OHPHE), 1-hydroxypyrene (1-OHPYR) and 2 & 3-hydroxyphenanthrene (2-OHPHE & 3-OHPHE) across different tertiles and their association with female infertility. Results indicated that the tertile-3 of 2-OHFLU shows a significant association with female infertility when compared with tertile-1(OR: 2.84, 95% CI: 1.24–6.53, P = 0.015). Zhang et al. [28], conducted a cross-sectional study in China, associating certain PAH metabolites with endometriosis, with concentrations of urinary PAH metabolites showing significant differences between endometriosis and non-endometriosis groups. However, after adjusting for various factors, the study found no statistically significant risk of endometriosis associated with higher levels of PAH metabolites.

Carcinogenic reproductive outcomes

In the analysis of carcinogenic reproductive outcomes, a total of nine studies were reviewed, assessing five reproductive cancers: prostate cancer, testicular cancer, cervical cancer, breast cancer, ovarian cancer and uterine cancer [29–31, 63–68]. Among these, five studies investigated the association between exposure to PAH metabolites and male reproductive organ cancers [31, 64, 66–68], while two studies focused on female reproductive organ cancers [29, 63]. Additionally, two studies assessed the relationship between PAH metabolites exposure with both male and female reproductive organ cancers [30, 65]. Six studies were retrospective cohort studies [30, 64–68], and three were case-control studies [29, 31, 63]. Exposure assessment was based on environmental exposure in two studies [29, 68] and occupational exposure in eight studies [30, 31, 63–67]. The extracted data from each of these 9 articles, segregated by outcome, are accessible in separate Excel sheets included in Supplementary 3.

Male reproductive organ cancers

The association between exposure to PAH metabolites and male reproductive organ cancers was reported in seven studies [30, 31, 64–68]. Two of these studies found significant associations between occupational exposure to PAHs and prostate cancer [30, 64]. The first study reported an increased rate of prostate cancer (standard incidence ratio (SIR): 1.19; 95% CI: 1.00-1.40) for male employees compared with national rates [30]. The second study, which followed the cohort for a longer period, detected similar results (SIR: 1.18; 95% CI: 1.04–1.34) [64].

Meta-analysis results

A meta-analysis was conducted to evaluate the association between exposure to PAH metabolites and male reproductive organ cancers, encompassing data from seven studies [30, 31, 64–68]. The results indicated a statistically significant association, with an overall SIR of 1.13 (95% CI: 1.04–1.23, P < 0.001), suggesting that exposure to PAH metabolites increases the risk of developing reproductive male cancers by 13%. The low heterogeneity among the studies (I²: 9.96%, P = 0.35) implies that the results are consistent across the studies analyzed (Fig. 3: Association between exposure to PAH metabolites and male reproductive organ cancers). The symmetry of funnel plots indicated no evidence of publication bias for the total SIRs (Supplementary 4, Figure S1A). Additionally, Egger’s linear regression test (P = 0.237) and Begg’s rank test (P = 0.731) found no evidence of publication bias. The trim-and-fill method revealed that the average effect sizes for the total SIRs remained unchanged, suggesting that publication bias did not impact the results (Supplementary 4). The pooled SIR for male reproductive organ cancers was influenced by the study by Gun et al. [64], When this study was excluded, the analysis indicated that exposure to PAH metabolites increases the risk by 9%; however, this finding did not reach statistical significance. The Galbraith plot used to assess heterogeneity among the seven included studies revealed no significant heterogeneity (Supplementary 4). Subgroup analyses were conducted based on the year of publication, distinguishing studies published before 2000 and after 2000. Results indicated no significant association for studies published before 2000 (SIR: 1.14, 95% CI: 0.87–1.51, P = 0.74), with no heterogeneity among these studies (I²: 0%). However, for studies published after 2000, results indicated a statistically significant association (SIR: 1.13, 95% CI: 1.03–1.24, P < 0.001), with moderate heterogeneity among these studies (I²: 36.11%, P = 0.14) (Supplementary 4). The meta-regression analysis was conducted to identify the source of heterogeneity for the SIR of male reproductive organ cancers based on the publication year of the studies. The univariate meta-regression showed a coefficient (β) of 0.00122 with a 95% CI ranging from − 0.0248 to 0.0272, and a P-value of 0.927. This indicates that the publication year does not significantly contribute to the heterogeneity observed in the SIR of male reproductive organ cancers, as the coefficient is close to zero and the P-value is much greater than 0.05.

Fig. 3.

Association between exposure to PAH metabolites and male reproductive organ cancers

Female reproductive organ cancers

The association between exposure to PAH metabolites and female reproductive organ cancers was reported in four studies [29, 30, 63, 65]. Three of these studies found significant associations between occupational exposure to PAHs with cervical and uterine cancers [65], post-menopausal breast cancer [63], and breast cancer [29]. Lewis et al. [65], reported increased rates of cervical cancer (SIR: 0.42; 95% CI: 0.17–0.86) and uterine cancer (SIR: 0.31; 95% CI: 0.06–0.89) among 8,062 female petroleum industry workers in Canada compared to the general population. Bonner et al. [63], conducted a case-control study in the USA and found a significant trend for increased breast cancer risk in post-menopausal women exposed to high PAH levels at birth (P = 0.01), with ORs ranging from 1.94 to 2.42, suggesting a dose-response relationship. Mordukhovich et al. [29], also found significant associations in their case-control study in the USA, where higher estimated residential exposure to benzo[a]pyrene was consistently higher among breast cancer cases than controls. These findings indicate a complex relationship between PAH exposure and increased risk for certain female reproductive cancers.

Meta-analysis results

The meta-analysis of the association between exposure to PAH metabolites and female reproductive organ cancers was conducted based on four studies [29, 30, 63, 65]. The results indicated no significant association between PAH exposure and female reproductive organ cancers, with an overall SIR of 1.01 (95% CI: 0.91–1.12, P = 0.91) (Fig. 4: Association between exposure to PAH metabolites and female reproductive organ cancers). The analysis showed moderate heterogeneity among the studies (I²: 44.04%, P = 0.02), suggesting some variability in the study results. However, the overall effect estimate indicates no increased risk of female reproductive organ cancers associated with PAH exposure. Visual inspection of the funnel plot suggested a slight degree of publication bias (Supplementary 4). Egger’s linear regression test (P = 0.009) indicated evidence of publication bias, while Begg’s rank test (P = 0.956) did not. The trim-and-fill method, which adjusts for publication bias, showed that adding imputed missing studies did not change the overall results, maintaining the non-significant association between PAH exposure and female reproductive organ cancers (Supplementary 4). This suggests that publication bias did not significantly impact the results. However, it should be noted that tests for publication bias, such as Egger’s and Begg’s, have very low statistical power when the number of studies is small (k = 4). Therefore, the results of these tests should be interpreted with extreme caution, as they are unlikely to reliably detect true publication bias. The trim-and-fill method, which is less dependent on the number of studies, showed that the overall result remained non-significant after imputing potential missing studies, suggesting that the conclusion is robust to potential small-study effects. The pooled SIR for female reproductive organ cancers was influenced by the study by Mordukhovich et al. [29], Excluding this study indicated an 8% increased risk associated with PAH exposure, but this finding did not reach statistical significance (Supplementary 4). The Galbraith plot, used to assess heterogeneity among the included studies, revealed no significant heterogeneity (Supplementary 4, Figure S5B). Subgroup analyses were conducted based on the year of publication, distinguishing studies published before 2000 and after 2000. For studies published before 2000, there was no significant association (SIR: 1.16, 95% CI: 0.98–1.38), with moderate heterogeneity (I²: 48.55%, P = 0.03). For studies published after 2000, there was also no significant association between PAH exposure and female reproductive organ cancers (SIR: 0.93, 95% CI: 0.81–1.06, P < 0.001), with no heterogeneity (I²: 0%) (Supplementary 4).

Fig. 4.

Association between exposure to PAH metabolites and female reproductive organ cancers

Meta-regression analysis was performed to find the source of heterogeneity among studies based on the year of publication. The univariate meta-regression showed a coefficient (β) of −0.015 with a 95% CI ranging from − 0.033 to 0.002, and a P-value of 0.081. This suggests a potential, though not statistically significant, negative relationship between the publication year and the SIR of female reproductive organ cancers. The negative coefficient implies that more recent studies might report slightly lower SIRs compared to older studies, but this trend is not strong enough to reach statistical significance.

Discussion

Summary of key findings

This systematic review and meta-analysis evaluated the relationship between exposure to PAHs and human reproductive outcomes, both carcinogenic and non-carcinogenic. A total of 30 studies were included in the systematic review [23, 24, 26–32, 48–67], of which nine were eligible for meta-analysis [29–31, 63–68]. Our findings suggest that PAH exposure may be associated with declines in semen quality, including reduced sperm concentration, motility, viability, and morphology, as well as increased sperm DNA damage and altered hormone levels in males. The meta-analysis found a statistically significant association between PAH exposure and male reproductive organ cancers, particularly prostate cancer. However, given the predominance of cross-sectional designs, variability in exposure and outcome definitions, and the inclusion of studies with moderate risk of bias, these findings should be interpreted with caution. In contrast, evidence for female reproductive health outcomes, including infertility, endometriosis, and reproductive cancers, remains limited and inconclusive.

Male reproductive outcomes

Seventeen studies investigated semen characteristics, including semen concentration [23, 24, 48, 50, 52, 53, 56, 59, 61], sperm count [23, 32, 49, 50, 52, 55, 62], sperm motility [23, 24, 32, 49, 50, 52–54, 56, 59–61], sperm viability [49, 53, 61], sperm morphology [23, 24, 53, 55, 57, 59–62], semen volume [23, 24, 50, 52, 54], sperm DNA damage [24, 51–53], and sperm motion parameters [23, 24, 52, 60]. These measures were generally consistent with WHO guidelines and demonstrated significant inverse associations with PAH metabolites. However, the majority were cross-sectional, limiting causal inference.

The association with male infertility was evaluated in three studies [26, 32, 58], showing hormonal disruption (e.g., reduced testosterone) and increased idiopathic infertility risk. Although these studies provide valuable clinical relevance, small sample sizes and multifactorial etiology limit generalizability.

However, nearly all male reproductive studies used cross-sectional designs, limiting causal inference. In addition, the use of varied semen assessment protocols (e.g., differing WHO thresholds), heterogeneous biological matrices (urine vs. semen vs. blood), and inconsistent adjustment for confounders contributed to methodological variability. While several studies reported similar directions of effect, the methodological heterogeneity, observational design, and low-to-moderate certainty of evidence (as per GRADE) limit the strength of these associations. The observed patterns are suggestive but not conclusive of a causal relationship.

The meta-analysis found a 13% increased risk of male reproductive cancers (SIR: 1.13; 95% CI: 1.04–1.23), but with only seven studies included, the estimate is imprecise and should be considered preliminary. The certainty of this evidence is rated as moderate, downgraded from high due to the observational nature of the studies.

Female reproductive outcomes

Female reproductive outcomes were assessed in only two studies [27, 28], focusing on infertility and endometriosis. One study reported a statistically significant association before adjustment, but this association was attenuated after controlling for key confounders, suggesting potential residual confounding [27]. Given the very low certainty of evidence for female reproductive outcomes (as per GRADE), no firm conclusions can be drawn. A second study observed elevated PAH levels among women with endometriosis, though no statistically significant risk was identified after adjustment [28]. The limited number of female-focused studies, combined with small sample sizes and significant methodological heterogeneity including variation in outcome definitions, hormonal phase timing, and exposure assessment, may partly explain the inconclusive findings. Additionally, the complexity of female reproductive physiology, which involves cyclical hormonal fluctuations, oocyte maturation, and uterine receptivity, creates multiple potential points of disruption that may not be consistently captured across studies.

Overall, heterogeneity in outcome definitions (e.g., infertility measured via clinical diagnosis vs. self-report) and population characteristics (e.g., smoking status, occupational exposure) reduced comparability. This underscores the need for standardized outcome assessment in future research.

Exposure assessment

The reviewed studies used various approaches to assess PAH exposure, each with strengths and limitations. Biological monitoring using semen, urine, and blood samples provided direct internal exposure measures. Semen analysis, used in seven studies [26, 49, 52, 55, 57, 61, 62], reflects localized exposure but may be influenced by sample handling. Urine-based methods, applied in five studies [27, 28, 48, 53, 56], are non-invasive but reflect only recent exposure and are affected by individual metabolism. Blood sampling, used in one study [54], captures systemic exposure but is more invasive.

Seven studies used both urine and semen [23, 24, 26, 32, 51, 59, 60], allowing broader assessment but increasing design complexity. Twelve studies relied on environmental exposure metrics [23, 24, 48, 50–52, 54, 55, 59–61], while six focused on occupational exposure [32, 49, 53, 56, 57, 62]. Environmental assessments offer contextual exposure insight but lack individual specificity. Occupational assessments provide detailed exposure data but may not be generalizable and often rely on self-report, introducing recall bias. These methodological differences highlight the challenge of comparing exposure metrics across studies and reinforce the importance of using multiple, complementary methods where feasible.

Biological mechanisms

PAHs affect reproductive health through multiple mechanisms. One primary pathway is oxidative stress, where PAH metabolism generates reactive oxygen species (ROS), damaging sperm DNA, lipids, and proteins, ultimately impairing motility and viability [70–73].

As endocrine-disrupting chemicals (EDCs), PAHs interfere with sex hormones. In males, PAHs are linked to reduced testosterone levels and impaired spermatogenesis [74]. In females, they can disrupt ovulation and menstrual cycles, possibly contributing to infertility and endometriosis.

PAHs can also form DNA adducts, resulting in mutations when not repaired, and are associated with abnormal sperm morphology and cancer risk [75]. Furthermore, epigenetic modifications (e.g., altered DNA methylation) have been observed, which can dysregulate reproductive gene expression and affect embryo development [76].

Finally, PAHs are known to induce immunomodulatory effects, including chronic inflammation that may impair testicular function and disrupt the ovarian environment, with potential links to endometriosis [28, 77]. In females, PAHs may interfere with folliculogenesis, ovulation, and implantation by disrupting the hypothalamic–pituitary–gonadal axis and inducing chronic ovarian and uterine inflammation. Evidence from NHANES (2013–2016) shows that elevated urinary levels of metabolites such as 2‑hydroxynaphthalene and 2‑hydroxyfluorene are significantly associated with female infertility, especially when considering PAH mixtures [27]. Experimental findings in human granulosa cells demonstrate that PAHs, particularly in individuals with higher BMI, can accumulate in follicular fluid and inhibit steroidogenesis, reduce cell viability and proliferation, and increase oxidative stress, all critical for oocyte maturation and ovarian function [78]. Additionally, epidemiological and toxicological data suggest that PAH exposure alters menstrual cycle hormonal dynamics (e.g. shorter follicular phase, lower estradiol, altered LH/FSH levels), potentially impairing ovulation and fertility via HPG‑axis disruption [79]. These mechanisms support the biological plausibility of observed associations and emphasize the reproductive health risks of PAH exposure.

Strengths and limitations

The strengths of this study lie in the thorough search strategy and the rigorous assessment of study quality using the NOS and GRADE approaches. The diverse range of study designs and populations included enhances the generalizability of the findings.

However, several limitations must be acknowledged. First, the included studies demonstrated considerable heterogeneity across key methodological domains, including study design, population characteristics, exposure assessment methods, biological matrices, and outcome definitions. For example, semen parameters were measured using varying laboratory protocols and reference thresholds (e.g., different editions of WHO guidelines), and infertility was inconsistently defined (e.g., clinical diagnosis, self-report, or time-to-pregnancy). This heterogeneity limited the comparability of findings and reduced the precision of pooled estimates. As a result, quantitative synthesis was restricted to carcinogenic outcomes with compatible effect measures (SIRs), while non-carcinogenic outcomes were narratively summarized. This underscores the need for standardized and harmonized protocols in future research.

Second, the variation in biological matrices used to assess PAH exposure; urine, blood, and semen, introduces another layer of methodological complexity. Urine reflects recent exposure but may be influenced by metabolic variability and dilution. Blood provides systemic exposure levels but is more invasive and less frequently used. Semen offers insight into localized reproductive tract exposure but is subject to collection and quality variability. These differences complicate direct comparisons and limit confidence in dose-response interpretations.

Third, most studies were observational in design, which restricts causal inference due to potential residual confounding and reverse causation. While we applied rigorous quality appraisal (NOS and GRADE), adjustment for confounders varied across studies, particularly for smoking, diet, and occupational co-exposures. Furthermore, many studies did not fully account for concurrent exposure to other reproductive toxicants, such as volatile organic compounds, heavy metals, diesel exhaust, or endocrine-disrupting chemicals. This lack of adjustment for complex exposure mixtures may bias effect estimates and complicates the interpretation of PAH-specific effects.

Fourth, a significant proportion of the cohort studies were conducted in occupational settings, where the healthy-worker effect (HWE) may bias risk estimates toward the null. Since employed populations are generally healthier than the general population, this could attenuate observed associations between PAH exposure and reproductive cancers. Notably, we still observed a significant increase in male reproductive cancer risk despite this potential bias, suggesting the true effect may be even stronger.

Fifth, while biomarker-based exposure assessment (e.g., urinary OH-PAHs) offers greater precision than job-exposure matrices, such studies were largely limited to non-carcinogenic outcomes and lacked long-term cancer follow-up. Consequently, no biomarker-based studies reporting SIRs were available for inclusion in the meta-analysis, precluding a sensitivity analysis restricted to high-specificity exposure measures. This disconnect between accurate exposure data and hard health outcomes represents a critical gap in the current literature.

Sixth, the evidence base is disproportionately focused on male reproductive outcomes. The small number of studies on female infertility, endometriosis, and reproductive cancers limits the robustness of conclusions for women and reflects a broader gender imbalance in environmental reproductive research.

Seventh, most included studies originated from high-income countries, potentially limiting the generalizability of findings to populations in low- and middle-income settings, where exposure patterns, dietary habits (e.g., grilling or smoking of food), and cultural or religious dietary norms may differ significantly [80].

Finally, despite our comprehensive search strategy including multiple databases and grey literature sources, the possibility of publication bias cannot be ruled out. While funnel plots and statistical tests (e.g., Egger’s test) were used, their power is limited when fewer than 10 studies are included, as was the case for some analyses.

In our meta-analysis, combining studies with varying PAH compounds and exposure contexts increased statistical power but also introduced residual heterogeneity. Although random-effects models, subgroup, and sensitivity analyses were used to assess robustness, some variability in exposure and outcome measurement remains.

Public health implications

The findings of this review have meaningful implications for environmental health policy and clinical practice. Given the widespread presence of PAHs in urban air, industrial emissions, and consumer products, exposure is common even outside occupational settings. Reducing PAH emissions through stricter regulations and promoting awareness about dietary and environmental sources could help mitigate reproductive risks.

Recommendations for future research

Future research should aim to reduce methodological heterogeneity by applying standardized PAH exposure measurements and harmonized outcome definitions. More longitudinal cohort studies are needed to establish causal relationships and explore dose-response patterns. Importantly, future research should address the current gender imbalance by including more studies on female reproductive outcomes and exploring sex-specific biological mechanisms. Mechanistic studies are also needed to clarify how different PAHs influence hormonal pathways, DNA integrity, and fertility in both sexes.

Conclusion

This systematic review and meta-analysis synthesized evidence on the association between PAHs exposure and human reproductive health outcomes. The findings suggest that exposure to PAH metabolites may be associated with declines in semen quality such as reduced sperm concentration, motility, viability, and morphology, and an increased risk of idiopathic male infertility. Additionally, our meta-analysis found a statistically significant association between PAH exposure and male reproductive organ cancers, particularly prostate cancer. However, given the predominance of cross-sectional designs, variability in exposure and outcome definitions, and the inclusion of studies with moderate risk of bias, these findings should be interpreted with caution. In contrast, evidence for female reproductive health outcomes, including infertility, endometriosis, and reproductive cancers, remains limited and inconclusive. This reflects both the scarcity of female-focused studies and the complexity of measuring hormonal and inflammatory pathways in female reproductive biology. More robust, sex-specific data are needed to clarify these associations.

From a public health perspective, these findings support the need for strategies to reduce PAH exposure, particularly in occupational and urban environments where exposure levels may be elevated. Public awareness campaigns and regulatory measures targeting dietary and airborne sources of PAHs could help minimize cumulative exposure. Clinicians, particularly those managing cases of idiopathic infertility, should consider environmental risk factors such as PAH exposure as part of patient history-taking. Where relevant, clinicians might consider recommending screening or counseling on environmental and occupational exposures, especially for individuals with otherwise unexplained reproductive dysfunction.

Future research should prioritize longitudinal study designs, harmonized exposure assessment protocols, and standardized reproductive outcome definitions. Additionally, mechanistic studies exploring the pathways through which PAHs affect endocrine function and gametogenesis especially in women, are essential to strengthening the evidence base and guiding preventive strategies.

Author contributions

Ali Mohammad Latifi: conceptualization, methodology, writing - review & editing; Fatemeh Abdi: methodology, data curation, writing - review & editing; Mohammad Miri: methodology, literature search writing - review & editing, writing-original draft; Sara Ashtari: literature search, data extraction, data curation, risk of bias, writing - review & editing, writing-original draft; Seyedeh Noushin Ghalandarpoor-attar: writing-review & editing; Milad Mohamadzadeh: writing-review & editing; Abbas Ali Imani-Fooladi: methodology, writing - review & editing; Shahab Uddin: conceptualization, methodology, writing - review & editing; Amir Vahedian-azimi: supervisor, literature search, data extraction, data curation, risk of bias, writing - review & editing, writing-original draft.

Funding

This research did not receive any specific grant from funding agencies in the public, commercial, or not-for-profit sectors.

Data availability

Data is provided within the supplementary information files.

Declarations

Ethics approval and consent to participate

Research ethics confirmation was received from the Baqiyatallah University of Medical Sciences under the code: IR.BMSU.REC.1402.098. In addition, The study protocol was meticulously registered in the PROSPERO, International Prospective Register of Systematic Reviews, under the unique code CRD42024548548. Consent to participate was not applicable.

Consent for publication

Not applicable.

Competing interests

The authors declare no competing interests.