Endocrine Disrupting Chemicals in Seminal Plasma and Couple Fecundity

Abstract

Growing evidence supports the importance of men’s exposure to non-persistent endocrine disruptors (EDCs) and couple fecundability, as measured by time-to-pregnancy (TTP). This evolving literature contrasts with the largely equivocal findings reported for women’s exposures and fecundity. While most evidence relies upon urinary concentrations, quantification of EDCs in seminal plasma may be more informative about potential toxicity arising within the testes. We analyzed 5 chemical classes of non-persistent EDCs in seminal plasma for 339 male partners of couples who were recruited prior to conception and who were followed daily until pregnant or after one year of trying. Benzophenones, bisphenols, parabens, and phthalate metabolites and phthalate diesters were measured using high performance liquid chromatography-tandem mass spectrometry (HPLC-MS/MS) except for phthalate diesters, which were analyzed using gas chromatography-mass spectrometry. Cox regression with discrete-time was used to estimate fecundability odds ratios (FORs) and 95% confidence intervals (CIs) for each chemical to estimate the probability of pregnancy. While most EDCs were detected in seminal plasma, concentrations were lower than urinary concentrations previously analyzed for the cohort. None of the EDCs were significantly associated with fecundability even after covariate adjustment, though benzophenones consistently yielded FORs <1.0 (ranging from 0.72 to 0.91) in couple-adjusted models suggestive of diminished fecundity (longer TTP). The findings underscore that a range of EDCs can be quantified in seminal plasma, but the lower concentrations may require a large cohort for assessing couple fecundability, as well as the need to consider other fecundity outcomes such as semen quality.

Introduction

Endocrine disrupting chemicals (EDCs) are exogenous chemicals capable of interfering with any aspect of hormone action (Zoeller et al. 2012). Both persistent and non-persistent EDCs or those with long and short half-lives, respectively, have been associated with diminished fecundity, which is defined as the biologic capability of men and women for reproduction (Buck Louis et al. 2011a). To date, much of the existing research focusing on EDCs and fecundity has relied upon measured concentrations in women trying for pregnancy. Currently, there is evidence suggesting diminished fecundability, as measured by a longer time-to-pregnancy, with increasing concentrations of various classes of persistent EDCs, including dioxins, organochlorine pesticides, perfluorochemicals, polybrominated diphenyl ethers, and polychlorinated biphenyls (Axmon et al. 2005; Eskenazi et al. 2010; Fei et al. 2009; Gesink Law et al. 2005, Harley et al. 2010). Less attention has focused on non-persistent EDCs and fecundability, and findings reported to date are largely equivocal. For example, specific parabens (Smarr et al. 2016) and phthalates (Thomsen et al. 2017) have been reported to be associated with reduced fecundability or a longer TTP in prospective cohort studies with preconception enrollment of women or couples, whereas other such studies reported no associations (Buck Louis et al. 2014a; Jukic et al. 2016; Velez et al. 2015). Also of note are reported associations between non-persistent EDCs and other fecundity endpoints such as alterations in hormonal profiles or menstrual cycles and poorer in vitro fertilization (IVF) outcomes, as recently reviewed (Mínguez-Alarcón and Gaskins 2017).

Since human fecundity is a couple dependent outcome, it is important to assess EDC exposures in both partners of the couple. Recently, we summarized the findings from the Longitudinal Investigation of Fertility and the Environment (LIFE) Study and noted that male partners’ concentrations of both persistent and non-persistent EDCs were significantly associated with diminished couple fecundability or a longer TTP even in the absence of findings for female partners (Buck Louis et al. 2016). These collective findings underscore the importance of studying males when focusing on the relation between EDCs and couple fecundity. Findings from prospective IVF cohort studies also affirm the importance of studying the male partner for non-persistent EDCs such as phthalates, since negative associations have been observed between EDCs and implantation and live birth rates (Dodge et al. 2015).

Seminal fluid is a unique matrix for studying the reproductive effects of EDCs, since it is assumed to provide a more direct measure of within testes exposure (Vitku et al. 2015; whereas, urinary concentrations may be relatively more informative about total body burdens. Various classes of non-persistent EDCs have been detected in seminal plasma, such as benzophenones, bisphenol A, parabens, and phthalates (Bloom et al. 2015; Frederiksen et al. 2010, 2011; Leon et al. 2010; Vitku et al. 2016), and distributions have been compared across biologic media. For example, concentrations of 13 phthalate metabolites and 5 parabens were measured in the urine, serum and seminal plasma of 60 young Danish men. Urinary concentrations were higher than those in other matrices with relatively low correlations between urine and seminal plasma (Frederiksen et al. 2010, 2011). Another descriptive study found higher mean concentrations of 5 phthalates in the semen of 79 infertile men in comparison to 94 matched fertile men (Wang et al. 2015), and other authors have reported negative associations between seminal plasma concentrations of EDCs and semen quality (Chang et al. 2017; Vitku et al. 2016). These findings highlight the importance of assessing non-persistent EDCs in seminal plasma relative to human fecundity. Prompted by no previous research exploring this relation as known to us, we assessed a range of non-persistent EDCs measured in seminal plasma and couple fecundability, as measured by TTP. We compare the results to earlier findings for this cohort based upon urinary concentrations to assess the consistency of findings by biologic media.

Materials and Methods

1.1 Design and study population

A prospective cohort design with preconception recruitment of couples (n=501) was used to recruit 501 eligible couples who were discontinuing contraception to try for pregnancy from 16 counties in Michigan and Texas between 2005-2009. Given the absence of established sampling frameworks for identifying couples planning pregnancies, we used fishing/hunting license registries and marketing databases for these interests to develop samples. Households were called and residents were screened for eligibility. By design, the eligibility criteria for the male partner were minimal: aged 18+ years of age, able to communicate in English or Spanish and no history of clinically diagnosed infertility. Couples were followed daily until pregnancy or 12 months of trying without pregnancy. The study cohort for this analysis was restricted to male partners of couples with an observed TTP while participating in the Longitudinal Investigation of Fertility and the Environment (LIFE) Study, and who had residual semen samples of sufficient volume for quantifying non-persistent EDCs in seminal plasma (n=339; 68%). A complete description of the LIFE Study’s design and methods is provided elsewhere (Buck Louis et al. 2011b).

1.1.1 Data and biospecimen collection

Male partners were interviewed upon enrollment into the cohort to capture lifestyle and medical and reproductive history, and were subsequently trained in the completion of daily journals focusing on lifestyle while the couple was trying for pregnancy. Trained research assistants weighed men and measured their height using standardized methods and calibrated scales and measuring tapes for the calculation of body mass index (BMI; weight in kg/height in m²). After the baseline interview, all men provided blood and urine samples for the quantification of persistent and non-persistent EDCs, respectively. In addition, men were instructed in the collection of two at home semen samples with the intent of assessing semen quality. The first sample was obtained the day following the interview and the second sample approximately 1 month later. The second sample was used for an abbreviated semen analysis in part to corroborate azoospermia found in the first and more in-depth semen analysis, and for the quantification of EDCs. Men were instructed to collect the sample without the use of lubricants following 2 days of abstinence, and to return the sample to the andrology laboratory using overnight delivery, as previously described (Buck Louis et al. 2014b). Residual samples were stored as a pellet and then thawed and separated into sperm and seminal plasma for the quantification of non-persistent EDCs in seminal plasma by laboratory personnel experienced in the processing of semen. Specifically, seminal plasma was separated from sperm by centrifuging samples at 3000 rpm for 10 minutes. Seminal plasma was then pipetted for analysis. The interval between urine and semen collection was on average 2 months. Full human subjects’ approval was obtained from all participating institutions, and men gave informed consents prior to any data or biospecimen collection.

1.1.2 Toxicological analysis

The following EDCs were quantified in approximately 1.5 mL seminal plasma: 3 bisphenols [bisphenol A (BPA), bisphenol F (BPF) and bisphenol S (BPS)]; 5 benzophenones [2,4-dihydroxybenzophenone (BP-1), 2,2′,4,4′-tetrahydroxybenzophenone (BP-2), 2-hydroxy-4-methoxybenzophenone (BP-3), 2,2′-dihydroxy-4-methoxybenzophenone (BP-8), and 4-hydroxybenzophenone (4-OH-BP)]; 9 environmental phenols [triclosan (TCS), methyl-paraben (MeP), ethyl-paraben (EtP), propyl-paraben (PrP), butyl-paraben (BuP), heptyl-paraben (HeP), benzyl-paraben (BzP), and metabolites 4-hydroxybenzoic acid (4-HB) and 3,4-dihydroxybenzoic acid (3,4-DHB)]; 15 phthalate metabolites [mono-(2-ethyl-5-carboxypentyl) phthalate (MECPP), mono-[(2-carboxymethyl) hexyl] phthalate (MCMHP), mono-(2-ethyl-5-oxohexyl) phthalate (MEOHP), mono-(2-ethyl-5-hydroxyhexyl) phthalate (MEHHP),mono-(3-carboxypropyl) phthalate (MCPP), monomethyl phthalate (MMP), mono-ethyl phthalate (MEP), mono-n-butyl phthalate (MnBP), mono-2-isobutyl phthalate (MIBP), mono-hexyl phthalate (MHxP), mono-cyclohexyl phthalate (MCHP), mono-octyl phthalate (MOP), mono-isononyl phthalate (MINP), mono-benzyl phthalate (MBzP),and mono-(8-methyl-1-nonyl) phthalate (MIDP)]; and 9 phthalate diesters [dimethyl phthalate (DMP), diethyl phthalate (DEP), diisobutyl phthalate (DIBP), di-n-butyl phthalate (DBP), benzyl butyl phthalate (BzBP), di-n-hexyl phthalate (DNHP), dicyclohexyl phthalate (DCHP), di(2-ethylhexyl) phthalate (DEHP), and di-n-octyl phthalate (DOP)].

Quantification of all EDCs was completed using high performance liquid chromatography-tandem mass spectrometry (HPLC-MS/MS) with multiple reaction monitoring mode except for phthalate diesters, which were analyzed using gas chromatography-mass spectrometry (GC-MS). Specifically, parent phthalates were analyzed using a method described in Wang et al. (2015). Seminal plasma was placed in clean glass tubes and spiked with an isotope-labeled internal standard mixture. Samples were acidified with phosphoric acid, and hexane extraction was performed twice. The samples were analyzed using a GC-MS system (G1530A/5973; Agilent, Santa Clara, CA, US) with a HP-5MS column (30m × 0.25mm × 0.25μm; Agilent) under electron impact selected ion monitoring (SIM) mode. Phthalate metabolites were analyzed using the method described elsewhere (Asimakopoulos et al. 2016). Seminal plasma was placed into polypropylene tubes (PP) and spiked with an isotope-labeled internal standard mixture. After deconjugation by β-glucuronidase, target analytes were extracted by solid phase extraction (SPE) (ABS ELUT-Nexus, Varian, Walnut Creek, CA, US). The samples were then analyzed by HPLC (Agilent 1260 Series; Agilent, Santa Clara, US) coupled MS/MS (ABSCIEX 4500 QTRAP; Sciex, Foster City, CA, US) and an Ultra AQ C18 column (100 mm × 2.1 mm, 3 μm; Restek, Bellefonte, PA, US) under ESI-negative, multiple reaction monitoring (MRM) mode. Environmental phenols were analyzed using the method described by Kunisue et al. (2010) and Smarr et al. (2017). Seminal plasma was placed into PP tubes and spiked with an istope-labelled internal standard mixture. After deconjugation by β-glucuronidase, methyl tert-butyl ether/ethyl acetate (50/50, v/v) extraction was performed twice. Samples were analyzed by UPLC (I-class; Waters, Milford, MA, US) coupled with MS/MS (ABSCIEX 5500 Triple Quad; Sciex, Foster City, CA, US) and the Acquity UPLC BEH C18 column (1.7 μm, 50 × 2.1 mm, Waters) under ESI-negative, MRM mode. We used labelled internal standard for all target analytes and, therefore, the matrix effect was accounted for from the response of individual compounds.

All analyses were completed at the Laboratory of Organic Analytical Chemistry at the Wadsworth Center’s Division of Environmental Health Sciences.

1.1.3 Statistical analysis

Data analysis was completed in two phases commencing with descriptive inspection of the data for completeness followed by comparison of male partners’ characteristics by couples’ pregnancy status. TTP denoted the number of menstrual cycles couples required to become pregnant, with a menstrual cycle defined as the interval between the onset of bleeding lasting ≥2 days from one cycle to the next, as recorded by female partners in daily journals. Pregnancy was ascertained by a positive home pregnancy test with demonstrated accuracy (Johnson et al. 2015) on the day of expected menstruation, as reported in daily journals.

Cox discrete-time survival analytic techniques were utilized to estimate fecundability odds ratios (FORs) and corresponding 95% confidence intervals (CIs), with separate models run for each chemical consistent with our goal to fully explore non-persistent EDCs and couple fecundability (Cox 1972). FORs estimate the odds of pregnancy across the trying attempt relative to exposure conditional on not having achieved pregnancy in the previous cycle. These models accommodate the discrete time nature of TTP and allow for a cycle-varying intercept. FORs <1.0 denote diminished fecundability or a lower probability of pregnancy as reflected in a longer TTP, whereas FORs >1.0 denote enhanced fecundability or a higher probability of pregnancy as reflected in a shorter TTP. The models accounted for left truncation or the time interval between discontinuation of contraception and enrolled into the cohort, which largely reflects the logistics in arranging an enrollment home visit. Also, eligibility criteria required that couples had recently (≤2 months) discontinued contraception with the goal of becoming pregnant. Chemical concentrations (ng/mL seminal plasma) were log (1+x) transformed for analysis and rescaled by their standard deviations (SD) to aid interpretation, and without substituting values below laboratory limits of detection or quantification to minimize bias when assessing human health outcomes (Richardson and Ciampi 2003; Schisterman et al. 2006). In estimating adjusted FORs, we included a priori defined covariates in most models, i.e., male age, body mass index and serum cotinine (Auger and Jouannet 2005; Bolumar et al. 1996; Dunson et al. 2002; Merken et al. 1986; Sundaram et al. 2017), and added other variates that were included when assessing EDCs in urinary models (i.e., household income and season of enrollment for benzophenones given temporal patterns in sunscreen usage). BMI (kg/m²) was calculated from measured height (converted to centimeters) and weight (converted to kilograms) and categorized as: normal or underweight (< 25.0), overweight (25.0-29.9), obese class I (30.0-34.9), and obese class II (≥ 35) for analysis. Of note, we kept models consistent with our earlier ones where EDCs where measured in urine to facilitate comparison of findings across biologic media (Buck Louis et al. 2014a, 2015c; Smarr et al. 2017). All analyses were run using SAS version 9.4 (SAS Institute, Inc., Cary, North Carolina).

Results

Study participants were largely Non-Hispanic White (83%) men who were nonsmokers (84%) with a mean age of approximately 31 years and BMI of 29.5 (overweight). Couples becoming pregnant were more likely to have male partners who did not smoke and to reside in households with higher incomes than couples not becoming pregnant (Table 1).

Table 1.

Description of male participants by couple’s pregnancy status, (n=339).

Characteristics	Pregnant (n=246) n (%)	Not pregnant (n=93) n (%)	P-value
Age in years, mean (SD)	31.4 (0.3)	32.2 (0.6)	0.354
Serum cotinine concentration (ng/mL):
No active smoking (<9.99)	212 (87)	68 (75)	0.028
Passive exposure (10–99.9)	5 (2)	4 (4)
Active smoking (≥100)	27 (11)	19 (21)
Serum cotinine, mean (SD)	39.0 (8)	65.1 (15.4)	0.002
Body mass index (weight kg/height m2):
Underweight (<18.5)	2 (1)	0 (0)	0.806
Healthy weight (18.5–24.9)	50 (20)	17 (18)
Overweight (25.0–29.9)	101 (41)	39 (42)
Obese (≥30)	93 (38)	37 (40)
BMI, mean (SD)	29.3 (0.3)	29.8 (0.6)	0.728
Household income:
<$70,000	57 (24)	42 (46)	0.000
≥$70,000	183 (76)	49 (54)
Race/ethnicity:
Non-Hispanic white	204 (83)	76 (82)	0.737
Non-White	41 (17)	17 (18)
Research site:
Michigan	32 (13)	27 (29)	0.001
Texas	212 (87)	66 (71)
Season of enrollment:
Winter	53 (21)	28 (30)	0.060
Spring	78 (32)	20 (22)
Summer	63 (26)	18 (19)
Fall	52 (21)	27 (29)

NOTE: Significance determined using the Chi-square test statistic for categorical variables and the non-parametric median test for continuous variables.

SD, standard deviation

The distributions of EDCs by biospecimens are presented in Table 2 and reflect lower concentrations for all chemicals when measured in seminal plasma in comparison to urine, with the exception of BP-3, BzP and MCHP (though actual differences were small). Contrarily, MeP, TCS, MEP, MnBP, MiBP, DIBP, DBP, and DEHP were universally detected in all seminal plasma. Figure 1 illustrates the low correlations between EDCs with ≥50% of concentrations above laboratory limits of quantification when measured in seminal plasma and urine. The highest observed correlations were for creatinine adjusted MnBP (r=0.44; p<0.0001) and TCS (0.33; p<0.0001).

Table 2.

Distribution of endocrine disrupting chemicals (EDCs) in seminal plasma and urine. Concentrations are presented in ng/ml seminal plasma.

EDC	LODs	% <LOD Seminal Plasma	n	Median (semen)	IQR (semen)	% <LO D Urin e		n	Median (urine)	IQR (urine)
	Phenols
BPA	0.048	16	339	0.16	0.08, 0.37	2		439	0.50	0.23, 1.13
BPF	0.035	50	339	0.038	<LOD, 0.11	–		–	–	–
BPS	0.018	25	339	0.11	0.02, 0.28	–		–	–	–
	Benzophenones
BP-1	0.008	4	339	0.35	0.10, 1.66	1		439	2.83	0.92, 15.68
BP-2	0.045	82	339	<LOD	<LOD, 0.02	28		439	0.14	0.07, 0.30
BP-3	0.042	10	339	1.42	0.41, 7.98	2		439	1.22	0.41, 6.75
BP-8	0.035	99	339	<LOD	<LOD	27		439	0.07	0.02, 0.35
4-OH-BP	0.006	59	339	<LOD	<LOD, 0.02	4		439	0.05	0.02, 0.15
	Phthalate Metabolites
MeP	0.010	<1	339	0.84	0.40, 1.87	<1		439	6.55	2.13, 26.44
EtP	0.008	18	339	0.07	0.02, 0.16	11		439	0.37	0.17, 1.26
PrP	0.010	4	339	0.21	0.08, 0.71	4		439	1.45	0.49, 5.55
BuP	0.012	45	339	0.02	0.00, 0.07	67		439	0.03	0.01, 0.17
BzP	0.011	9	339	0.07	0.03, 0.17	88		439	0.02	0.00, 0.04
HeP	0.010	100	339	<LOD	<LOD	100		439	0.00	0.00, 0.00
TCS	0.043	0	339	10.61	3.86, 42.75	13		439	17.75	4.42, 77.12
TCC	0.008	77	339	<LOD	<LOD, 0.01	88		439	0.01	0.00, 0.03
MEP	0.106	<1	333	1.26	0.43, 1.30	<1		439	87.20	32.10, 277.19
MCPP	0.071	87	333	0.33	<LOD, 0.03	3		439	5.58	2.46, 12.13
MEHHP	0.107	26	333	0.43	0.10, 0.44	1		439	14.32	5.54, 37.80
MBzP	0.120	29	333	0.29	0.08, 0.41	4		439	3.70	1.52, 8.52
MOP	0.149	100	333	<LOD	<LOD, 0.00	96		439	<LOD	<LOD, 0.03
MMP	0.122	7	333	0.63	0.24, 0.58	61		439	0.57	0.03, 2.01
MCHP	0.111	100	333	0.01	0.00, 0.02	96		439	0.00	<LOD, 0.01
MIDP	0.090	90	333	<LOD	<LOD, 0.26	–		–	–	–
MHxP	0.127	94	333	0.01	0.00, 0.00	–		–	–	–
MnBP	0.099	0	333	2.97	1.72, 3.65	1		439	7.40	3.35, 14.87
MIBP	0.111	<1	333	1.15	0.68, 1.47	2		439	4.37	1.83, 9.08
MECPP	0.087	33	333	0.47	0.01, 0.57	<1		439	20.26	8.53, 46.25
MCMHP	0.054	63	333	0.30	<LOD, 0.42	<1		439	18.42	6.61, 47.15
MEOHP	0.084	52	333	0.21	0.03, 0.21	2		439	6.94	3.04, 17.77
MINP	0.165	92	333	0.01	<LOD, 0.00	95		439	0.01	<LOD, 0.08
	Phthalate Diesters
DMP	1.032	96	292	0.15	0.00, 0.00		–	–	–	–
DEP	0.190	4	296	1.80	0.73, 2.66		–	–	–	–
DIBP	0.155	<1	298	1.13	0.67, 1.30		–	–	–	–
DBP	0.083	0	287	2.06	1.17, 2.62		–	–	–	–
DNHP	0.122	97	299	0.01	0.00, 0.00		–	–	–	–
BzBP	0.324	42	299	0.39	0.00, 0.58		–	–	–	–
DCHP	0.096	99	299	0.00	0.00, 0.00		–	–	–	–
DEHP	0.119	0	299	4.20	1.23, 3.45		–	–	–	–
DOP	0.335	88	299	0.32	0.00, 0.00		–	–	–	–

NOTE: All concentrations are rounded to 2 decimal places and percentages are rounded to nearest whole number.

IQR, interquartile range

<LOD, below limits of detection.

(–) denotes not measured in urine

Figure 1.

Correlations between endocrine disrupting chemicals measured in seminal plasma and urine.

We observed no association between any of the phenols and phthalates and couple fecundability, as measured by prospectively observed TTP when replicating our previous models based on urinary EDC concentrations (Table 3). FORs were both below and above 1.0 for individual chemicals and remained so even after covariate adjustment. All FORs were consistently below 1.0 for benzophenones even with covariate adjustment, ranging from 0.72 to 0.91 in couple-adjusted models suggestive of a 9% to 28% reduction in fecundability (Table 4). BPS was consistently associated with FORs <1.0 reflecting an 8% reduction in the couple-adjusted model as were most phthalate metabolites and diesters (ranging from 1% to 13% reduction in couple-adjusted models), though none achieved significance. Parabens and TCS were associated with FORs both below and above 1.0 and patterns remained relatively consistent across adjusted models (Table 5). None of the findings achieved significance, as evident from CIs inclusive of 1. Of note, we did not observe any EDCs to be positively associated with fecundability, as reflected in a shorter TTP.

Table 3.

Seminal plasma phenol and phthalate concentration and couple fecundability.

Chemical (ng/mL)	Male Unadjusted FOR (95% CI)	Male Adjusteda FOR (95% CI)	Couple Adjustedb FOR (95% CI)
Phenols (n=339)
BPA	1.08 (0.94, 1.24)	1.06 (0.92, 1.22)	1.04 (0.90, 1.20)
BPF	1.09 (0.94, 1.26)	1.08 (0.93, 1.26)	1.10 (0.94, 1.28)
BPS	0.92 (0.79, 1.06)	0.91 (0.78, 1.05)	0.92 (0.79, 1.08)
Phthalate Metabolites (n=333)
MEP	0.92 (0.80, 1.07)	0.90 (0.77, 1.06)	0.93 (0.79, 1.09)
MCPP	1.03 (0.90, 1.17)	1.01 (0.89, 1.15)	1.00 (0.63, 1.58)
MEHHP	0.91 (0.78, 1.06)	0.87 (0.75, 1.01)	0.86 (0.73, 1.01)
MBzP	0.91 (0.79, 1.05)	0.92 (0.79, 1.05)	0.90 (0.78, 1.05)
MOP	1.17 (1.02, 1.36)	1.17 (1.01, 1.37)	1.19 (1.02, 1.40)
MMP	0.97 (0.83, 1.13)	0.97 (0.81, 1.15)	0.98 (0.82, 1.17)
MCHP	1.03 (0.90, 1.18)	0.98 (0.86, 1.13)	0.98 (0.85, 1.13)
MIDP	0.98 (0.85, 1.12)	1.02 (0.89, 1.18)	1.03 (0.89, 1.19)
MHxP	0.92 (0.79, 1.07)	0.94 (0.82, 1.09)	0.93 (0.80, 1.07)
MnBP	0.93 (0.81, 1.07)	0.94 (0.82, 1.09)	0.93 (0.80, 1.07)
MIBP	0.94 (0.82, 1.08)	0.96 (0.82, 1.11)	0.97 (0.83, 1.12)
MECPP	0.95 (0.81, 1.10)	0.94 (0.81, 1.10)	0.94 (0.80, 1.10)
MCMHP	0.94 (0.81, 1.09)	0.92 (0.80, 1.06)	0.92 (0.79, 1.07)
MEOHP	0.91 (0.78, 1.07)	0.88 (0.75, 1.02)	0.88 (0.75, 1.04)
MINP	0.94 (0.81, 1.08)	0.90 (0.78, 1.03)	0.90 (0.77, 1.04)
Phthalate Diesters (n=287 to 299)
DMP	0.94 (0.80, 1.11)	1.00 (0.85, 1.19)	0.99 (0.84, 1.18)
DEP	1.06 (0.92, 1.21)	1.03 (0.91, 1.16)	1.01 (0.90, 1.14)
DIBP	1.05 (0.76, 1.46)	1.17 (0.83, 1.65)	1.11 (0.78, 1.58)
DBP	0.99 (0.85, 1.16)	1.00 (0.85, 1.16)	0.96 (0.82, 1.13)
DNHP	0.94 (0.79, 1.11)	0.93 (0.79, 1.11)	0.94 (0.79, 1.11)
BzBP	0.89 (0.77, 1.03)	0.92 (0.79, 1.07)	0.89 (0.76, 1.04)
DCHP	1.12 (0.94, 1.33)	1.10 (0.93, 1.31)	1.11 (0.93, 1.32)
DEHP	0.88 (0.75, 1.04)	0.89 (0.74, 1.06)	0.90 (0.75, 1.08)
DOP	0.86 (0.72, 1.01)	0.91 (0.76, 1.08)	0.93 (0.78, 1.11)

NOTE: Separate models were run for each chemical. Chemical concentrations were log transformed and rescaled by their standard deviation for analysis. All models accounted for left truncation or time couple was off contraception. N’s vary slightly across chemical classes due to missing covariate information.

^aMale model adjusted for male age (continuous), male BMI (continuous), male serum cotinine (continuous), and research site (Michigan/Texas).

^bCouple model adjusted for female age (continuous), difference in couples’ ages (continuous), both partners’ BMI (continuous), both partners’ serum cotinine concentrations (continuous), and research site (Michigan/Texas).

CI, 95% confidence interval; FOR, fecundability odds ratio

Table 4.

Seminal plasma benzophenone concentration and couple fecundability (n=339).

Chemical (ng/mL)	Male Unadjusted FOR (95% CI)	Male Adjusteda FOR (95% CI)	Couple Adjustedb FOR (95% CI)
BP-1	0.94 (0.69, 1.29)	0.81 (0.58, 1.14)	0.78 (0.55, 1.10)
BP-2	0.84 (0.61, 1.17)	0.93 (0.66, 1.31)	0.91 (0.64, 1.29)
BP-3	0.90 (0.66, 1.24)	0.76 (0.54, 1.07)	0.72 (0.51, 1.03)
BP-8	0.97 (0.71, 1.34)	0.86 (0.62, 1.20)	0.88 (0.63, 1.23)
4-OH-BP	0.99 (0.71, 1.37)	0.87 (0.62, 1.23)	0.86 (0.60, 1.21)

NOTE: Separate models were run for each chemical. Chemical concentrations were dichotomized at the 75th percentile, with the group corresponding to lower values serving as the referent. All models accounted for left-truncation or time off contraception. N’s vary slightly across chemical classes due to missing covariate information.

^aMale model adjusted for age (years; continuous), BMI (categorical), serum cotinine concentration (categorical), season (winter, spring, summer, or fall), and research site (Michigan or Texas).

^bCouple model adjusted for female partners’ age (years; continuous), difference between the partners’ ages, couples’ body mass index (categorical), season (winter, spring, summer, or fall), male serum cotinine concentration (categorical), and research site (Michigan or Texas).

CI, 95% confidence interval; FOR, fecundability odds ratio

Table 5.

Seminal plasma paraben and antimicrobial concentration and couple fecundability (n=339).

Chemical (ng/mL)	Male Unadjusted FOR (95% CI)	Male Adjusteda FOR (95% CI)	Couple Adjustedb FOR (95% CI)
MeP	0.84 (0.73, 0.98)	0.85 (0.74, 0.98)	0.85 (0.74, 0.99)
EtP	0.96 (0.83, 1.10)	0.97 (0.83, 1.12)	0.95 (0.82, 1.11)
PrP	0.85 (0.73, 0.99)	0.85 (0.72, 1.01)	0.85 (0.72, 1.01)
BuP	1.00 (0.89, 1.12)	1.03 (0.89, 1.20)	1.03 (0.88, 1.20)
BzP	1.02 (0.89, 1.17)	1.03 (0.89, 1.18)	1.03 (0.89, 1.18)
HeP	0.90 (0.78, 1.03)	0.83 (0.71, 0.96)	0.83 (0.72, 0.97)
TCS	1.09 (0.95, 1.25)	1.09 (0.94, 1.26)	1.09 (0.94, 1.26)
TCC	1.08 (0.94, 1.24)	1.10 (0.96, 1.26)	1.11 (0.97, 1.28)

NOTE: Separate models were run for each chemical. Chemical concentrations were log transformed and rescaled by their standard deviation for analysis. All models accounted for left truncation or time couple was off contraception. N’s vary slightly across chemical classes due to missing covariate information.

^aMale model adjusted for age (years; continuous), BMI (categorical), serum cotinine concentration (categorical), race/ethnicity (dichotomized), and income (dichotomized).

^bCouple model adjusted for female age (years; continuous), difference between partners’ age (continuous), male BMI (categorical), male serum cotinine concentrations (categorical), race/ethnicity (dichotomized), and income (dichotomized).

CI, 95% confidence interval; FOR, fecundability odds ratio

Discussion

While we could detect and quantify 100% of phenols, 71% of parabens, 80% of phthalate metabolites, and 73% of phthalate diesters in seminal plasma, none was observed to be significantly associated with couple fecundability, as measured by TTP. Perhaps the strongest signal from our findings is the consistent pattern of reduced fecundability for all 5 benzophenones, as evident by FORs <1.0, though the findings were not significant.

Our findings are not readily comparable to previous studies as we found no study that assessed seminal plasma EDCs and couple fecundability. We compared our findings for seminal plasma with previous publications from the LIFE Study where EDCs were quantified in urine. Results from this earlier work focusing on male partners’ exposures and TTP identified two urinary phthalate metabolites – MMP and MBzP – to be associated with a 19% and 20% reduction in couple fecundability. This finding was only observed for the male’s and not female partner’s urinary concentration (Buck Louis et al. 2014a). We also previously reported two urinary benzophenones – BP-2 and 4-OHBP – to be significantly associated with a 31% and 26% reduction in couple fecundability. In fact, the association between BP-2 and fecundability was robust to adjustment for the female partners’ concentrations (Buck Louis 2014c). Consistent with the seminal plasma results, we found no association between male urinary bisphenol A (BPA), parabens or TCS and couple fecundability (Buck Louis et al. 2014a; Smarr et al. 2017).

Possible reasons for the lack of consistency in findings for men participating in the LIFE Study across biologic media include distributions skewed toward lower levels for seminal plasma in comparison to urine, a finding that corroborates earlier research (Frederiksen et al. 2010, 2011). Other reasons include reduced statistical power in that 32% of men participating in the LIFE Study did not have residual semen samples available for analysis, a factor that may increase the likelihood of a type 2 error, especially in the context of lower concentrations. It remains possible that the approximate 2-month interval between collection of the urine and semen samples reflect exposures at different time points, in light of the short-lived nature of the EDCs assessed in this work. If there are marked daily variations in concentrations, it is possible that the time interval may not have adequately capture exposure. Another possible explanation is that fecundability is not the ideal outcome for seminal exposures, especially given that early investigations have found negative associations between specific non-persistent EDCs measured in seminal plasma and semen quality endpoints (Chang et al. 2017; Vitku et al. 2016). The low correlations for select EDCs despite statistical significance between biologic media underscores that they are not interchangeable matrices in that measurements in one cannot be used as proxies for the other. In light of continuing concern about increasing trends of fecundity related impairments affecting men and women (Marques-Pinto and Carvalho 2013) and the call for purposeful investigation into underlying reasons (Smarr et al. 2017), continued inquiry into EDCs as measured in various biologic media may help delineate underlying mechanisms and in developing guidance for couples in protecting their fecundity.

Lastly, it is important to note that our findings do not directly assess the relation between EDCs as quantified in seminal plasma and male fecundity, as measured by semen quality. Rather, our work focused on couple fecundability. As such, we estimated TTP and cannot attribute delays to either diminished male or female fecundity, per se. Future work in this area is critical in part given the emerging mechanistic research focusing on EDCs and semen quality. For example 4 methylbenzylidene camphor (Schiffer 2014) and p,p’-DDE (Tavares 2013) have been investigated relative to sperm capacitation and motility with findings suggesting an important role for activation of the CatSper channel. As work continues in this area, it also will be important to consider male exposures and the joint modeling of semen quality and TTP to more fully inform about mechanisms underlying conception delays.

Conclusions

Despite most non-persistent EDCs being detected and quantified in seminal plasma, none was significantly associated with couple fecundability reflecting as a longer TTP. Possible reasons for the lack of consistent findings between EDCs measured in urinary and seminal plasma remain to be established, but likely reflect diminished power in the context of lower concentrations when measuring in seminal plasma. The implications of these EDCs in seminal plasma for male fecundity and health status remain to be established.

Highlights.

• Non-persistent endocrine disruptors were detected and quantified in seminal plasma

• Concentrations were lower in seminal plasma than in urine

• Benzophenones were associated with reduced fecundability, but not significantly

• Bi-directional associations for other endocrine disruptors, but not significant

Abbreviations

BP-1

2,4-dihydroxybenzophenone

BP-2

2,2′,4,4′-tetrahydroxybenzophenone

BP-3

2-hydroxy-4-methoxybenzophenone

BP-8

2,2′-dihydroxy-4-methoxybenzophenone

4-OH-BP

4-hydroxybenzophenone

BPA

bisphenol A

BPF

bisphenol F

BPS

bisphenol S

BuP

butyl-paraben

BzBP

benzyl butyl phthalate

BzP

benzyl-paraben

CI

confidence interval

DBP

di-n-butyl phthalate

DCHP

dicyclohexyl phthalate

DEHP

di(2-ethylhexyl) phthalate

DEP

diethyl phthalate

DIBP

diisobutyl phthalate

DMP

dimethyl phthalate

DNHP

di-n-hexyl phthalate

DOP

di-n-octyl phthalate

EDC

endocrine disrupting chemicals

EtP

ethyl-paraben

FOR

fecundability odds ratio

HeP

heptyl-paraben

4-HB

4-hydroxybenzoic acid

3,4-DHB

3,4-dihydroxybenzoic acid

MnBP

mono-n-butyl phthalate

MBzP

mono-benzyl phthalate

MCHP

mono-cyclohexyl phthalate

MCMHP

mono-[(2-carboxymethyl) hexyl] phthalate

MCPP

mono-(3-carboxypropyl) phthalate

MIDP

mono-(8-methyl-1-nonyl) phthalate

MECPP

mono-(2-ethyl-5-carboxypentyl) phthalate

MEHHP

mono-(2-ethyl-5-hydroxyhexyl) phthalate

MEOHP

mono-(2-ethyl-5-oxohexyl) phthalate

MEP

mono-ethyl phthalate

MeP

methyl-paraben

MHxP

mono-hexyl phthalate

MIBP

mono-2-isobutyl phthalate

MINP

mono-isononyl phthalate

MMP

mono-methyl phthalate

MOP

mono-octyl phthalate

PrP

propyl-paraben

TCC

triclocarban

TCS

triclosan

TTP

time-to-pregnancy