The classic EDCs, phthalate esters and organochlorines, in relation to abnormal sperm quality: a systematic review with meta-analysis

The association between endocrine disrupting chemicals (EDCs) and human sperm quality is controversial due to the inconsistent literature findings, therefore, a systematic review with meta-analysis was performed. Through the literature search and selection based on inclusion criteria, a total of 9 studies (7 cross-sectional, 1 case-control, and 1 pilot study) were analyzed for classic EDCs (5 studies for phthalate esters and 4 studies for organochlorines). Funnel plots revealed a symmetrical distribution with no evidence of publication bias (Begg’s test: intercept = 0.40; p = 0.692). The summary odds ratios (OR) of human sperm quality associated with the classic EDCs was 1.67 (95% CI: 1.31–2.02). After stratification by specific chemical class, consistent increases in the risk of abnormal sperm quality were found in phthalate ester group (OR = 1.52; 95% CI: 1.09–1.95) and organochlorine group (OR = 1.98; 95% CI: 1.34–2.62). Additionally, identification of official data, and a comprehensive review of the mechanisms were performed, and better elucidated the increased risk of these classic EDCs on abnormal sperm quality. The present systematic review and meta-analysis helps to identify the impact of classic EDCs on human sperm quality. However, it still highlights the need for additional epidemiological studies in a larger variety of geographic locations.


Results
Study characteristics. Through the literature search and selection based on inclusion criteria, 9 articles were identified by reviewing potentially relevant articles (Fig. 1), including seven cross-sectional, one case-control, and one pilot study for the present meta-analysis. The characteristics of the selected studies are shown in Table 1. Seven studies were conducted in the US 16,21,22,[29][30][31][32] and there were two in China 33,34 . The literature was divided into two categories after stratification by specific chemical class; one class for phthalate esters and the other class for organochlorines. Funnel plots revealed a symmetrical distribution with no evidence of publication bias (Begg's test, intercept = 0.40; p = 0.692) (Fig. 2). There was one case-control study with a total of 25 male infertility cases, seven cross-sectional studies involving 2016 male partners of subfertile couples, and one pilot study involving 45 male partners of subfertile couples (Table 1).

Meta-analysis.
Nine studies contributing a total of 26 odds ratio (OR) estimators met the inclusion criteria and were taken into consideration. The summary OR of abnormal sperm quality associated with exposure to classic EDCs was 1.67 (95% CI: 1.31-2.02) by both fixed-and random-effects models (Fig. 3). The heterogeneity Q statistic was 13.50 (p = 0.973, p > 0.05) and the I 2 was 0.00%, indicating no statistical evidence for heterogeneity. Furthermore, in the subgroup analysis based on specific chemical class, the overall association between phthalate ester and abnormal sperm quality was statistically significant for the five studies (OR: 1.52, 95% CI: 1.09-1.95, p = 0.989), and an obvious increase in the risk of abnormal sperm quality was found in the organochlorine group (OR = 1.98, 95% CI: 1.34-2.62, p = 0.656). In addition, a forest plot of the 9 studies displayed the weights applied in each study for the overall meta-analysis (Fig. 3). The class of phthalate esters and organochlorine pesticides contributed 68.70% and 31.30% of the total weight, respectively. No single study contributed more than 30% of the total weight, thus, the overall estimate risk contributed the largest number of cases, which was not directly affected by a single study with a methodological difference.
Review of official data. Present results suggested that exposure to organochlorine and phthalate esters appeared to be associated with an increased risk of abnormal sperm quality. To validate the findings we performed an additional systematic review using official research data, reports, and relevant literature reviews on the relationship between classical EDCs and sperm quality.
(1) PCBs and sperm quality. As shown in Table 2, PCBs and their congeners (PCB 153, PCB 138) were associated with abnormal sperm motility and morphology in humans, and PCB 132, PCB 118, PCB 77, PCB 126 induced a reduction in sperm number and daily sperm production in rats. Such relationships have been consistently reported across studies performed in different countries. For example, an epidemiology study from the United States reported that high-doses of PCBs from accidental food contamination presented a dose-response relationship with sperm motility (ORs per tertile of PCB 138, 1.00, 1.68, 2.35) and morphology (1.00, 1.36, 2.53) 35 . Environmental exposure to lower doses of PCBs also supported an association with reduced sperm quality, specifically sperm motility 36 . Richthoff et al. 37 found a weak, but statistically significant negative correlation between PCB 153 levels and both the ratio of testosterone:SHBG (sex hormone-binding globulin) (r = − 0.25), and sperm motility (r = − 0.13). Among 195 Swedish fishermen, the subjects in the highest quintile of PCB 153 exposure (> 328 ng/g lipid) tended to have decreased sperm motility compared with those in the lowest quintile (< 113 ng/g lipid), but this association was weak after age-adjusted (p = 0.08) 20 .
(2) Organochlorine pesticides and sperm quality. Organochlorine pesticides may affect human sperm abnormalities including sperm count, morphology, motility, and seminal volume (Table 3). Two recent studies found associations between pesticide representatives and reduced sperm quality in the general population 38,39 . An increased risk for poor sperm quality was found in relation to urinary concentrations of several pesticides such as alachlor, 2-isopropoxy-4-methyl-pyrimidinol, atrazine, 1-naphthol, and 3,5,6-trichloro-2-pyridinol 38 . A cross-sectional study in Limpopo, South Africa showed that serum DDT derivatives (six compounds) induced morphology scores 84% below either the World Health Organization (WHO) or the Tygerberg criteria 40 . Another cross-sectional study in Limpopo that found that higher plasma DDE was associated with reduced sperm motility, reduced ejaculate volume, and oligozoospermia. Both DDT and DDE were significantly associated with asthenozoospermia 18 . A study in Mexico demonstrated that higher plasma DDE levels were associated with the decreased percentage of motile sperm and the increased percentage of sperm with morphological tail defects 19 . Another study from infertility clinics in Michigan, USA indicated significant associations between p,p′ -DDE serum levels and reduced sperm concentration, motility, and morphology (depending on DNA polymorphisms), and the risk for low sperm concentration was significantly increased by high DDE and DDT serum concentrations 31 . Additionally, the increased p,p′ -DDE in serum was associated with a moderate, but significant increase in chromatin defects in the sperm of 209 young men from DDT-sprayed dwellings in South Africa 41 . Ayotte et al.   Table 4. The results show that phthalates might be associated with decreased sperm concentration, motility, morphology, and sperm DNA damage in humans or rats. In Andhra Pradesh, India, an epidemiological study showed that phthalates might be instrumental in the deterioration of semen quality in infertile men, especially increased abnormal sperm morphology 2 . In the general population of young Swedish males, Jonsson et al. 26 suggested that subjects within the highest quartile for MEP had fewer motile sperm and more immotile sperms. Hauser 43 from the National Institute of Environmental Health Sciences (NIEHS) reported MEP and MEHP were associated with increased sperm DNA damage. Another study in Massachusetts General Hospital, demonstrated that urinary MEP, at environmental levels, was associated with the increased DNA damage in sperm 44 .

Discussion
This systematic review with meta-analysis extracted the estimates of association of two classic EDCs, phthalate esters and organochlorines, from nine independent studies from two countries (majority in North America), and the results showed consistent evidence of positive associations of these classic EDCs with abnormal sperm quality (Fig. 3). The heterogeneity Q statistic was 13.50 (p = 0.973, p > 0.05) and I 2 was 0.00%, indicating no statistical evidence for heterogeneity among the selected studies. Begg's funnel plot and Egger's test were usually performed to assess the publication bias of literature. Particularly, the Egger's test was used to provide statistical evidence of funnel plot symmetry. Our results showed no evidence of obvious asymmetry in the shape of the funnel plot, and no publication bias (Egger's test, p = 0.691), indicating that the present results in meta-analysis were reliable. Among the studies in this meta-analysis (Table 1), Hauser et al. 30 found that PCB-153 in relation to sperm motility, and there were dose-response and inverse relationships among PCB-138 and sperm motility and morphology in their another study 16 . Sperm motility may be the target effect of PCBs. High DDE-DDT exposure adversely affected some sperm parameters (sperm count, motility and morphology) 31 . However, there was no significant association between organochlorines and semen quality 32 . Phthalate esters seem to show the weak adverse effects on sperm quality. Hauser et al. reported near significant (highest quartile of MBzP vs low sperm concentration, p = 0.13) or significant associations between urinary levels of monobutyl phthalate and low sperm counts (p = 0.02) and/or motility (p = 0.04) 21 , and similar trends were found for additional phthalate metabolites (monobenzyl phthalate) in other studies 22,29,33 . But a study from general population in Chongqing, China, showed that exposure to the environmental level of phthalate had weak adverse effects on the sperm concentration 34 .
In addition, through the analysis of official research data, reports, and relevant literature, we found that occupational and environmental exposure to the aforementioned EDCs are closely related to male reproductive   (Table 2, Table 3, Table 4). For example, both animal studies and epidemiological evidence supported an inverse association between PCBs and sperm quality 35,37,[45][46][47][48][49][50] . Serum levels of DDE/DDT were associated with reduced sperm concentration, motility, and morphology 18,19,25,38,42,51 . Still, there were significant associations between phthalate esters and poor sperm quality 2,26,43,44,52 . Notably, interactions between MBP and MBzP with the PCB-153 congener, in relation to sperm motility were found 30 . Despite these findings, some studies have shown no evidence that supports these negative effects in sperm parameters or fertility in p,p′ -DDE 53,54 , MMP, and MBzP 29 . Therefore, due to the limited amount of literature in the present meta-analysis, and inconsistent findings among studies related to EDC exposure and human sperm quality, it further highlights the need for additional epidemiological studies in a large variety of geographic locations.
To address these concerns, we summarized the mechanism of action for organochlorines and phthalates on sperm quality from several in vivo and in vitro studies (Fig. 4). Animal, clinical and epidemiological studies have demonstrated that exposure to EDCs disrupts male reproductive health, traditionally through male steroidogenesis to disrupt spermatogenesis. Phthalates and its metabolite, DEHP, were observed to exert an effect on Leydig or Sertoli cell structure and functions through the activation of peroxisome proliferator activated receptors (PPARs). DEHP also affected the binding of LH to G-protein coupled LH receptors, thereby influencing steroid hormone biosynthesis in fetal rat testes 55 , and DEHP was found to inhibit testosterone production resulting in the dysfunction of StAR, 3β -HSD, CYP17, and 17β -HSD. In addition, phthalates have been shown to disrupt the patterns of gene expression that regulate cholesterol and lipid homeostasis or insulin signaling, resulting in lower testosterone synthesis 56 . Still, phthalates could decrease testosterone through the induction of cytochrome P450 aromatase (AROM), which converts testosterone to estrogen in cultured rat cerebellar granule cells 56 . Similarily, PCB 153 and Aroclor 1254 promoted the down-regulation of StAR, 3β -HSD, CYP 17, and 17β -HSD in 14-year-old boys from a birth cohort in the Faroe Islands 57 . Prenatal PCB exposure was associated with lower serum concentrations of LH and testosterone in rats 58 . Select phthalate monoesters may interfere with the ability of Sertoli cells    to respond to their normal endogenous ligand, FSH 59 , thereby affecting the downstream secretion of androgen binding protein (ABP) which could induce reproductive dysfunction. The spermatogenic process is usually regarded as both a source and a target of reactive oxygen species (ROS). EDCs could interfere with testicular functions by breaking the dynamic balance of this antioxidant system, and subsequently affect sperm quality. Phthalates, mainly DEHP, DBP, DEP, MEHHP, MEOHP, and MEHP, are associated with specific events in spermatogenesis including the induction of ROS, lipid peroxidation, and apoptosis of spermatocytes in mice 60 . In neonatal Sertoli cell/gonocyte coculture system, MEHP was found to induce germ cell apoptosis through the Fas/FasL pathway 61 . PCB 132 impaired sperm function and altered testicular apoptosis-related gene expression through the generation of ROS, activation of caspase-3 and -9, and down-regulation of Fas, Bax, bcl-2, and p53 genes in rat offspring 14 . In other studies, PCB 153 and p,p′ -DDE were found to induce ROS and cellular apoptosis in a Sertoli cell/gonocyte co-culture system 62,63 . Moreover, exogenous estrogen (e.g. EDCs) and endogenous estrogen can induce germ cell apoptosis and blood-testis barrier (BTB) dysfunction via PI3K/FAK or PI3K/AKT and MAPK/ERK signaling pathways in mice 64 . For example, phthalates have the ability to directly damage the BTB integrity by allowing the BTB to "open" and disrupt spermatogenesis in rats 65 . Notably, local biosynthesis of estrogen and androgen occurs in the testis when aromatase is expressed in Leydig cells and some populations of germ cells 66 . EDCs can interfere with the binding of the hormone and the receptor, and disrupt the normal process of spermatogenesis. Some organochlorines, such as p,p′ -DDE, and other PCBs, are regarded as ER antagonists, antiestrogens, or AR antagonists 67,68 . Phthalates, DDT, and DDT metabolites (o,p′ -DDT, and p,p′ -DDE) can inhibit endogenous ligands from binding estrogen and androgen receptors 69 . PCBs can disrupt estrogen receptor function by mimicking the natural ligand and acting as an agonist 70 . Additionally, human semen MMP, MEP, and urinary phthalate metabolites (MBzP, MBP, MEHP, and MEP) were associated with increased sperm DNA damage and sperm aneuploidy 71,72 . A majority of organophosphate pesticides, including PCB, DDT, and p,p′ -DDE, affect the male reproductive system by inducing sperm DNA damage 73 . DEHP can induce changes in DNA methylation within CpG islands, resulting in testicular toxicity 74 .
Current evidence has shown limitations in meta-analysis regarding the relationship between EDCs and human sperm quality. First, studies with small sample sizes may not adequately explore the potential exposure-response relationships. Secondly, differences exist between epidemiological studies, including differences in sample size, study design, study populations, life stage, data analysis approaches, strategies for exposure, and endpoints of effects. Thirdly, some of the extracted effects were unadjusted and a more precise analysis, perhaps through multivariable analysis instead of univariate, should be conducted from all the data. Finally, there are limited Figure 4. A schematic mechanism on the effects of phthalate esters and organochlorines on testosterone and sperm quality. A. Steroidogenesis. EDCs can inhibit the synthesis of testosterone through direct pathways including cholesterol, StAR, 3β -HSD, CYP 17, and 17β -HSD, or indirect pathways including the binding of LH to LH receptor and PPARγ , PKA, and StAR, or through FSH receptor, or the binding of testosterone to ABP or AR. In addition, EDCs can increase the AROM activity, which converts testosterone to estrogen, resulting in the decrease in testosterone. B. Spermatogenesis. EDCs may affect spermatogenesis through the apoptosis of spermatocytes, ROS production, or disrupting BTB integrity via the activation of PI3K/FAK or PI3K/Akt and MAPK/ERK signaling pathways. C. DNA damage and DNA methylation. CpG islands may be possible mechanisms of EDC-induced testicular toxicity and sperm quality. LHR, luteinizing hormone receptor; PPARγ , peroxisome proliferator activated receptor gamma; PKA, protein kinase A; CREB, cAMP response element; StAR, steroidogenic acute regulatory protein; TSPO, translocator protein; 3β -HSD, 3β hydroxysteroid; CYP 17, Cytochrome P450 17; 17β -HSD, 17β hydroxysteroid; AROM, cytochrome P450 aromatase; ROS, reactive oxygen species; AR, androgen receptor; ER, estrogen receptor; FSH, follicle stimulating hormone; ABP, androgen binding protein; PI3K, phosphatidylinositol 3 kinase; FAK, focal adhesion kinase; MAPK, mitogen activated protein kinase; ERK, extracellular regulated protein kinases; BTB, blood-testis barrier.
Scientific RepoRts | 6:19982 | DOI: 10.1038/srep19982 inherent epidemiological studies that evaluate the toxicity of multiple chemicals and to our knowledge, humans are exposed to a mixture of chemicals rather than a single chemical.
Nonetheless, our systematic review with meta-analysis, together with the review of possible mechanisms, provides a better explanation of the impact of classic EDCs on sperm quality. However, future research is needed to examine the following: (1) the biomarker of testis function and human fertility should be well defined in an epidemiology study, and the analysis of sperm quality parameters need to be normalized; (2) the size of adequate samples, occupational exposure to specific EDCs, longitudinal instead of cross-sectional studies, and multi-center studies need to be conducted; (3) due to potential interactions between different EDCs on sperm quality, co-exposure to mixtures of EDCs, as well as their interactions or combined effects should be investigated; (4) for a better understanding of classic EDC-induced abnormal sperm quality, mechanism studies should be focused on low-dose, long-term, and co-exposure; and (5) both human studies and animal experiments are needed on transgenerational effects (e.g. DNA methylation) of EDCs because epigenetic effects as a result of EDC exposure can subsequently change the sperm quality of future generations.

Methods
Literature search. We performed a systematic electronic search on the National Library of Medicine PubMed database and Web of Science database to identify published studies from January 1990 to April 2015. The research question was defined as 'what are the associations between classic EDCs and sperm quality?' . This question was subsequently broken down to cover specific search terms such as 'endocrine disrupting chemicals' , 'male reproductive damage' , 'infant development' , 'abnormal development' , 'malformation' , 'infertility' , 'abnormal sperm' , 'sperm parameter' , 'asthenospermia' , 'aspermia' , and 'oligospermia' . Through searching and examining relevant literature to fill the missing components, each of them was cross-referenced with the following classic EDC terms: 'bisphenol A' , 'genistein' , 'cadmium' , 'lead' , 'phthalates' , 'poly chlorinated biphenyls' , 'polybrominated diphenyl ethers' , 'perfluorooctane sulfonates' , 'topical corticosteroid dependent dermatitis' , 'pesticide' , 'DDT' , and 'DDE' . However, only the term 'phthalates' , 'poly chlorinated biphenyls' , 'DDT' and 'DDE' were finally selected due to the lack of human studies or insufficient data for other EDCs. In addition, we replicated the search in the Web of Science database to identify additional pertinent references. Searches were restricted to human trials with the language restriction of English. All the references of relevant articles were scanned for additional analysis.

Selection of studies.
Studies selected for the meta-analysis met the following inclusion criteria: (1) Written and published in English; (2) Reported results from case-control, cohort, or cross-sectional epidemiology studies; (3) A relative risks (RR) and odds ratios (OR) with confidence intervals (CI) was reported, or could be calculated from provided data; (4) Referred to environmental or occupational exposure to the classic EDCs, phthalate esters and organochlorines.
Studies were excluded from the analysis if they: (1) included subjects that were already included in another more complete or more recent study; (2) did not report original results (reviews, comments, letters, editorials); (3) investigated women studies.
Data extraction. Each eligible study was classified as pilot, cross-sectional, cohort, or case-control study. We extracted the following information from the full text of each eligible publication: (1) author; (2) publication year; (3) study design (cross-sectional, case-control, or cohort); (4) exposure type; (5) source population for the controls in case-control studies; (6) number of cohort participants or number of cases and controls; (7) experimental results and health outcomes; (8) the name and category of EDCs; (9) effect estimates (RR, OR, and p-value) and CI.
Meta-analysis. The data was synthesized using both fixed-effects and random-effects models weighting each study by a measure of its precision, the inverse of the estimate variance. Heterogeneity of effects across studies was assessed by the Cochran's Q statistic 75 and was deemed significant when P < 0.05. In addition, the coefficient of inconsistency (I 2 ) as described by Higgins and Thompson 76 was also computed to assess heterogeneity. To examine the possibility that publication bias may have affected the results, a funnel plot of the natural logarithm of OR was constructed as the inverse of the variance of the studies, and regression test for the effects of small studies 77 was used for quantitative assessment of publication bias and funnel plot asymmetry. The data on ORs and 95% CI were entered into the STATA 12.0 statistical package to perform these calculations, and META command was used to calculate a summary OR, 95% CI, and heterogeneity statistics. The META-BIAS command was used to conduct the Begg's test which is used to diagnose publication bias and approximate to the fact than Egger's test if the publications are tendentiously deleted 78 .
Identification of official data. To validate the results from the meta-analysis, we performed a systematic electronic search related to phthalate esters, organochlorines, and sperm quality from the website databases of the U.S Environmental Protection Agency (EPA), WHO, U.S. Centers for Disease Control and Prevention (CDC), National Institutes of Health (NIH), and the NIEHS. These official research data and reports were then summarized in order to support our findings.

Systematic review on mechanisms from animal experiments and in vitro studies.
Recently, a large number of experimental animal studies have shown that EDCs have strong reproductive toxicity through steroid hormone synthesis, and possibly alter reproductive hormones to disrupt spermatogenesis. These studies provide new insights about other mechanisms such as oxidative stress, genetic susceptibility, and epigenetic effects. Therefore, a comprehensive review was additionally conducted through animal experiments and in vitro studies related to the mechanisms for the effects of phthalate esters and organochlorines on male reproductive damage.