A Case–Control Study on the Effects of Plasticizers Exposure on Male Fertility

Male infertility is a serious concern for public health, and the possible role of exposure to plasticizers such as phthalates and bisphenol A in contributing to the condition is widely debated. We have herein enrolled 155 infertility cases attending an infertility center and 211 controls (fathers of a spontaneously conceived newborn) to investigate this issue. The urinary levels of seven phthalates and BPA were analyzed through HPLC/MS/MS. All data were statistically elaborated considering information about clinical situation, life habits, occupational activity, and, for cases, semen parameters (volume, sperm concentration, total count of spermatozoa, and sperm motility). Results showed significantly higher urinary concentrations for all the phthalates in cases compared to controls, except for monoethylphthalate and BPA. In total, 90.07% of cases had sperm motility lower than the WHO reference value (2010), while 53.69%, 46.31%, and 16.56% had sperm total number, concentration, and volume, respectively, out of the reference range. Regarding the possible source of exposure, the use of scents seems to be a significant source of DEP (diethylphthalate). When considering occupational settings, industrial workers, dental technicians, artisans, and farmers using chemicals showed higher risk (OR = 2.766, 95% CI 1.236–6.185), particularly in relation to DnBP (di-n-butyl phthalate) and DEHP (di-ethyl-hexyl phthalate) exposure. No clear quantitative correlation between specific plasticizers and sperm parameters could be demonstrated but these findings call for future studies about the risks associated with exposure to their mixture.


Introduction
Infertility is usually defined as the lack of ability to conceive a child after 12 months of unprotected intercourse [1]. It has been estimated that infertility affects about 37-70 million couples worldwide [2], and in 50 % of cases it can be ascribed to a male factor.
A meta-analysis of epidemiological data has shown that the quality of human semen has declined significantly over the last 40 years, with a severe reduction in sperm counts [3]. Similarly, there has been a decrease in sperm motility (15% in 17 years of follow-up) and an increase in morphological alterations in sperm (16% in 17 years of follow-up) [4].
During spermatogenesis, gametes are created daily in the seminiferous tubules, and spermatozoa, through the epididymis, subsequently reach the epididymal tail where they are collected until ejaculation [5]. All phases of the process are hormonally regulated, in particular by testosterone, 17-β-estradiol, follicle-stimulating hormone (FSH), and luteinizing hormone (LH). The mechanisms leading to male infertility are not completely clear, though changes in hormonal balance, particularly during sperm cell development, might affect the quality of spermatozoa. Risk factors that predispose an individual to male infertility can be various, ranging from congenital malformations to genetic problems, iatrogenic issues, and lifestyle factors [6]. Nevertheless, some environmental risk factors have been reported [7][8][9].

Study Design and Recruitment of the Sample Population
This is a retrospective case-control study. The Institutional Review Board of the IRCCS San Raffaele, Scientific Institute in Milan, Italy, approved the research protocol (identification code 73/INT/2017).
Male partners from couples with infertility problems (defined as more than 12 months of unprotected intercourse without a pregnancy) (n = 274) were recruited at the Reproductive Center of San Raffaele Scientific Institute in Milan (Italy) between January 2018 and February 2019.
Inclusion criteria for enrollment as "cases" were infertility status and the presence of at least one abnormal semen parameter based on the recommendations of the World Health Organization (WHO, 2010). Participants were asked to collect a seminal fluid sample after at least 2 days of abstinence from intercourse.
Exclusion criteria included endocrine diseases with known etiology (e.g., adrenal disorders), testis injury, vasectomy, epididymitis, orchiditis, and previously diagnosed neoplasms that required specific chemotherapy and/or radiotherapy in the genital area. Subjects unable to provide a semen or urine sample and those with creatininuria out of the normal range set by the WHO for biological monitoring [48] were excluded.
Other variables were considered as possible confounding factors, including age, BMI, smoking, alcohol, and specific clinical situations potentially affecting the quality of the semen (e.g., varicocele or unresolved cryptorchidism).
After application of all the above-mentioned criteria, n = 155 subjects were included as cases.
Controls were recruited at the Obstetrics and Gynecology Unit of the IRCCS San Raffaele Hospital in Milan. The criteria observed for recruitment were the following: volunteers aged >18, male partners from couples with spontaneous conception (without hormonal therapy or assisted reproduction treatment), and a time to pregnancy lower than 12 months. From an initial number of 226 men who volunteered to participate in the study, after application of the inclusion criteria the sample was reduced to n = 211 subjects.
A questionnaire was administered by trained personnel to collect data from each participant on health status, life habits (diet, drugs, smoking, use of plastic containers for storage of fatty food, use of scents, or other possible source of phthalates), and working activity. Each participant also provided a urine spot sample so the concentration of phthalate metabolites and urinary BPA could be analyzed. Therefore, the study protocol was the same for cases and controls, with the only exception being semen collection and analysis, which was carried out only for cases. Finally, participants provided signed informed consent to participate in the study.

Urine Analysis
Analysis of the analytes of interest, phthalate metabolites and BPA in urine, was carried out using previously published methods [49,50] involving HPLC-MS/MS determination after pre-treatment of samples. Each urine sample was analyzed in duplicate; quantification was obtained using the linear regression equation identified from the calibration curve for the single metabolite, expressed in µg/L urine. The following metabolites were determined: mono-n-butyl phthalate (MnBP) with a limit of detection (LOD) of 0.8 µg/L, monoethyl phthalate (MEP) with an LOD of 3 µg/L, mono-benzyl phthalate (MBzP) with an LOD of 1.2 µg/L, mono-n-octyl phthalate (MnOP) with an LOD of 0.1 µg/L, mono(2ethylhexyl)phthalate (MEHP) and mono(2-ethyl-5-hydroxyhexyl) phthalate (MEHHP), both metabolites of DEHP, with respective LODs of 0.15 and 0.05 µg/L, and urinary BPA with an LOD of 0.02 µg/L. In cases with values lower than the limit of detection (LOD), phthalate metabolite concentrations and urinary BPA concentrations were replaced with LOD/ √ 2. Urinary samples were also analyzed for their creatinine concentrations, which were determined using Jaffè's method with alkaline picrate tests and spectrophotometric detec-tion at a 490 nm wavelength [51]. Cases with values of urinary creatinine higher (more than 3.0 g/L) or lower (less than 0.3 g/L) than thresholds were excluded, as recommended by the American Conference of Governmental Industrial Hygienists (ACGIH) [52] and the WHO [48].
The final concentration of analytes was expressed in µg/g creatinine.

Semen Collection and Analysis
Each case subject provided a semen sample, which was collected at IRCCS San Raffaele Hospital in Milan (Italy) after a minimum of 2 days and a maximum of 7 days of ejaculatory abstinence.
After liquefaction in 30 min at 37 • C, the sample was characterized for volume, concentration, and motility of spermatozoa using a light microscope equipped with phase-contrast optics (Eclipse E200, Nikon) according to WHO criteria [53]. The total number of spermatozoa was then calculated considering the product of concentration and semen volume.
Sample volume was measured by weighing the sample in the container in which it had been collected, assuming the density of semen to be 1.0 g/mL.
The concentration of sperm was obtained using a Neubauer Improved (Bioanalytic GnbH, Umkirch, Germany) hemocytometer grid, with Gilson Microman M25, M50, or M250 positive displacement pipettes (Gilson UK, Luton, UK) used for dispensing the semen volume where appropriate for a dilution, and the analytical procedure was carried out according to WHO criteria [53].
The motility of spermatozoa was assessed at 37 • C and evaluated by counting at least 200 spermatozoa in a total of at least five fields in each replicate in order to achieve an acceptably low sampling error. Moreover, the average percentage of progressive, nonprogressive, and immobile sperm was calculated. The SEMQUA (seminal QUAlity) studies checklist was used to improve the accuracy and clearness of the study [54].
An internal and external quality control program, in accordance with the guidelines of the European Society of Human Reproduction and Embryology (ESHRE) [54], was set up in the laboratory to evaluate possible interlaboratory differences.
The laboratory staff were trained in accordance with the ESHRE Special Interest Group in Andrology Basic Semen Analysis Course.

Statistical Analysis
Statistical analysis was conducted with SPSS ® software version 25 (IBM, Armonk, NY, USA). All concentration levels were elaborated using creatinine-corrected data.
Based on data distribution, non-parametric tests were used. Differences between the two groups were assessed using the Mann-Whitney test and the χ2 test for qualitative variables.
The Mann-Whitney test was also used to highlight case-control differences in metabolite values considered as a whole or stratified for possible exposure sources. Furthermore, a multivariate regression model was applied to evaluate the possible correlation between urinary metabolite concentrations and semen quality.
A further stratification of the values was carried out for working activities in order to point out possible differences between cases and controls using the Kruskal-Wallis test.
Finally, the Odds Ratios (ORs) were calculated based on raw data or adjusted for confounding factors (age, BMI, smoking, alcohol, and male morbidities) to express the risk of infertility associated with professional activities.

Characteristics of the Study Population
The main characteristics of the enrolled population are shown in Table 1. The two groups of men were overall comparable; the only two parameters that represented a statistically significant difference were age (Mann-Whitney test p = 0.000, cases were on average slightly older than controls) and working activities (Mann-Whitney test p = 0.011).
A potential difference in terms of chemical exposure at workplaces emerged between the two investigated populations.

Chemical Exposure and Elaboration
A visual comparison between the mean values of phthalate metabolites in cases and controls is presented in Figure 1. The analytical results of urinary analysis are reported in detail in Table 2, together with indications of significant differences between the two groups. Considering that data were not normally distributed, a non-parametric test was used (Mann-Whitney).
A statistically significant difference between cases and controls emerged for all the metabolites measured. Infertile men showed significantly higher urinary values for metabolite concentrations, with the only exceptions being MEP and BPA. Regarding MEP, even if the mean concentration value was higher for cases, the wide distribution of data and the high standard deviation did not reach statistical significance. Conversely, BPA urinary levels were similarly low in both groups. controls is presented in Figure 1. The analytical results of urinary analysis are reported in detail in Table 2, together with indications of significant differences between the two groups. Considering that data were not normally distributed, a non-parametric test was used (Mann-Whitney).
A statistically significant difference between cases and controls emerged for all the metabolites measured. Infertile men showed significantly higher urinary values for metabolite concentrations, with the only exceptions being MEP and BPA. Regarding MEP, even if the mean concentration value was higher for cases, the wide distribution of data and the high standard deviation did not reach statistical significance. Conversely, BPA urinary levels were similarly low in both groups.  Table 3 shows the seminal fluid parameters (volume, total number of spermatozoa, sperm concentration, and motility) in the population of infertility cases [53]. Sperm motility was the most significantly affected parameter, being below the WHO reference values in more than 90% of the infertile population. Total sperm count was abnormal in 53.69% of cases and sperm concentration in 46.3%. Conversely, semen volume appeared to be the less critical parameter, as only 16.56% of the studied population had values below the normal range. Table 4 shows the results of multivariate analysis considering the analyzed metabolites and seminal parameters. No specific correlations were found for any of the studied metabolites in relation to any of the semen parameters after adjustment for confounding factors (age, smoke, BMI, alcohol use, genital pathologies) and considering a significance level of 95%.
In consideration of previously published results [50] describing possible exposure risks, the same data are shown in Table 5 after a stratification based on the habits of using scents and of using plastic containers for fat food storage, suspected to be possible relevant sources. A significantly more frequent use of perfumes was associated with an increased exposure to DEP, and therefore with a higher urinary MEP level. This evidence did not emerge for any of the other plasticizers. No significant results were conversely found in relation to the habit of using plastic containers for the storage of fat foods.
In order to clarify whether certain types of working activities could be associated with exposure to a specific plasticizer, a stratification of data in relation to this aspect was carried out (Table 6). This distribution of the data did not actually show any specific exposure for the occupational activities reported by the population under study (Kruskal-Wallis test).
Finally, to better investigate the possible role of occupational activity on reproductive health, ORs for infertility were calculated considering differences between cases and controls ( Table 7). This risk index represents the "cumulative" effects of occupational risk factors. From calculation of these indices, we found that activities that involve remaining for long periods of time in sitting positions were definitely at lower risk in terms of reproductive health issues, whereas professional activities involving possible exposure to chemical agents to some degree presented a higher risk as expected.     Table 6. Urinary levels of phthalate metabolites and BPA (average ± standard deviation in µg/g creatinine), stratification for working activity, and comparison through Kruskal-Wallis tests.

Discussion
The results obtained herein demonstrate significantly higher urinary concentrations for all the phthalates analyzed in cases compared to controls, except for MEP and BPA. In particular, the infertile population demonstrated, on average, values for MnBP, MBzP, and DEHP that were almost double those of controls, as well as a 50% higher value for MnOP.
The analytes were detectable in most of the samples, except for MBzP, which was measured in about half of the samples. It is noteworthy that DEHP metabolites could be detected in all samples, despite DEHP being one of the phthalates with increased marketing restrictions due to its well-known toxicity.
Comparing our concentrations of urinary metabolites with data in the literature, sometimes significant differences emerge that are probably related to geographical variability. Specifically, the mean value for MEP detected in the present study (182.47 µg/g creatinine) is comparable to findings collected in other Western countries, such as the US (200.96 µg/L) [55] and Sweden (190 µg/L) [56], while it is considerably higher than that reported in some Asian populations (15.2 µg/L and 18.8µg/L) [57]. The urinary concentrations of this metabolite, known to be strongly linked to the use of personal care products, always present important variability. According to the present findings, the frequent use of scents represents the only route of exposure to MEP clearly identified in daily life (p = 0.000). DEP is used in numerous types of formulations during perfume production. A previous study [58] analyzed 47 branded perfumes and highlighted the presence of DEP in all samples, with concentration levels ranging from 1621.6 ppm to 23,649.2 ppm. In general, the high concentration of MEP detected in urine samples is due to the absence of specific laws that place restrictions on the use of the compound in industrial processes and marketing, although the health effects associated with human exposure are not yet clearly understood.
For the other metabolites, the order of magnitude seems to be similar to those pre- For the metabolites of DEHP, published values range from 17 to 40 µg/L, comparable with our results (16.56 µg/g creatinine). For MnOP, it is difficult to find historical data because it has been rarely assessed. Pan et al. [57] failed to detect the metabolite, finding concentrations lower than the LOD.
To the best of our knowledge, we were the first to have clearly demonstrated higher levels of several metabolites in cases compared to controls, which is in disagreement with some previous studies. In a previous report [60], the assessment of 16 different free and conjugated metabolites led to the detection of significant differences in only MEP (43.4 vs. 28.9 µg/g creatinine) and MEHP (3.3 vs. 2.5 µg/g creatinine), with values lower in order of magnitude than those found in our study for MEP. Den Hond et al. [61] did not observe statistical differences between cases and controls when analyzing MEHP (2.9 vs. Discrepancies when compared to our study could be explained by the lower number of subjects previously considered (80 cases and 40 controls), as well as the different criteria used for the selection of controls (who were recruited from a fertility center among men with a higher total motility count).
Notably, most of the investigations did not normalize urinary concentrations for creatinine levels. This choice implies, in general, the imprecision of the data obtained, linked to a possible bias due to the dilution of the analytical matrix. For this reason, comparison to our data can be correctly performed only in terms of orders of magnitude and not exact levels.
In the population of infertile men under study, the sperm motility parameter was found to be lower than the reference value for 90.07% of subjects, thus confirming it to be the most critical parameter. This was followed by the total number of spermatozoa (in 53.69% of cases it was lower than the reference value), sperm concentration (in 46.31% of cases it was out of range), and volume (in only 16.56% of cases was it abnormal).
Previous research has documented a correlation between exposure to specific phthalates and alterations in semen quality. In particular, MnBP has been positively associated with a significant decrease in semen concentration [62,63] and sperm count (p < 0.05). Along this line, an increase in MEHP levels has been associated with a significant increase in "abnormal heads" of spermatozoa (p < 0.01) [62], as well as with an increase in oxidative stress at the sperm level [31].
Overall, there is good evidence to suggest that exposure to phthalates may be involved in sperm alterations and may therefore affect reproductive health, although it has to be admitted that a clear quantitative relationship between exposure and the reduced quality of seminal parameters is not easily assessable [65].
Indeed, when we adjusted the variables for confounding factors, the correlation between urinary analyte levels and semen parameters lost its statistical significance, thus limiting the possibility of clarifying whether one specific substance rather than another was more involved.
The combination of these data suggests that the risk should be addressed in terms of multiple exposure, since exposure to several phthalates at the same time could lead to a cumulative or even synergistic effect and result in clearer detectable changes. In this context, we are encouraged by the fact that all cases had higher values for the measured analytes compared to controls, which is not a secondary aspect.
By studying workplaces and possible sources of exposure, it emerged that groups engaged in activities involving exposure to chemical agents were, as expected, at higher risk. Farmers, artisans, dental technicians, and industrial workers showed an OR = 2.766 (95% CI 1.236-6.185) for infertility, and exposure to DnBP and DEHP in particular seems to be highly involved. Other working activities did not have an impact on male reproductive health.
A limit of this study is related to the determination of analytes on a single spot of urine, which allowed for exposure during the previous 24-48 h to be assessed. Another limitation is represented by the evaluation of exposure related to living and working environments relying exclusively on self-reported data. We could not directly verify the absence or presence of specific sources of exposure.
On the other hand, a very strict selection of cases and controls represents a strength of the investigation, which allowed important differences to be detected and helped confirm data from the literature. The frequent use of perfumes has been identified as a life habit that possibly represents a route of significant exposure to DEP. Furthermore, the focus on working environments was another point in favor of the validity of the present study.

Conclusions
Male infertility is a serious concern for public health because of its widespread prevalence and serious consequences. Numerous factors could lead to reproductive problems, both in environmental and occupational settings. Experimental data both in vitro and in vivo highlighted the involvement of phthalates in negatively affecting spermatogenesis and increasing oxidative stress and DNA damage, resulting in poor quality semen [66,67].
The results of the present study confirmed higher exposure to several phthalates in a population of infertile men compared to a fertile population. Even if a correlation between specific metabolites and semen parameters did not emerge, we cannot exclude the impact of multiple exposure on the observed health effects, particularly on sperm motility (which is the most critical parameter).
Regarding the route of environmental exposure, the use of scents was shown as a source of DEP exposure, confirming previous observations [50].
Considering possible occupational exposure, for men working in industrial plants, artisans, or other professions implying the use of chemicals, such as dentists, technicians, and farmers, evidence of the type of workplace contributing to infertility status emerged (OR = 2.766, 95% CI 1.236-6.185).
These findings call for novel investigations regarding the context of occupational settings in order to elucidate the specific environmental exposure involved, with the final aim of protecting workers from reproductive risks.