Environmental Research 138 (2015) 264–270

Contents lists available at ScienceDirect

Environmental Research journal homepage: www.elsevier.com/locate/envres

Prenatal phthalate exposure and reproductive function in young men Jonatan Axelsson a,n, Lars Rylander b, Anna Rignell-Hydbom b, Christian H. Lindh b, Bo A.G. Jönsson b, Aleksander Giwercman a a b

Molecular Reproductive Medicine, Skåne University Hospital Malmö, Lund University, 205 02 Malmö, Sweden Division of Occupational and Environmental Medicine, Lund University, 221 85 Lund, Sweden

art ic l e i nf o

a b s t r a c t

Article history: Received 14 October 2014 Received in revised form 20 February 2015 Accepted 21 February 2015

Background: Prenatal exposure to phthalates is suggested to negatively impact male reproductive function, but human data are lacking. Objectives: To study associations between prenatal exposure to diethylhexyl phthalate (DEHP) and diisononyl phthalate (DiNP), and reproductive parameters of adolescent men. Methods: Using linear regression models adjusted for potential confounders, we studied associations between levels of DEHP- and DiNP metabolites in maternal sera from mean 12 weeks of pregnancy, and testicular size, semen quality and reproductive hormones in 112 adolescent sons, recruited from the general population. Results: Men in the highest exposure tertile of one DiNP metabolite [mono-(carboxy-iso-octyl) phthalate], compared with men in the lowest tertile had: 4.3 mL (95% CI: 0.89, 7.6 mL; po 0.001) lower total testicular volume; 30% (95% CI: 3.6, 63%; p ¼0.02) higher levels of follicle-stimulating hormone; and 0.87 mL (95% CI: 0.28, 1.5 mL; p¼ 0.004) lower semen volume. Men in the highest exposure tertile of one DEHP metabolite [mono-(2-ethyl-5-hydroxylhexyl) phthalate] had 0.70 mL (95% CI: 0.090, 1.3 mL; p¼ 0.03) lower semen volume than men in the lowest exposure tertile. The levels of two DiNP metabolites [mono-(hydroxy-iso-nonyl) phthalate and mono-(oxo-iso-nonyl) phthalate] were linearly associated with luteinizing hormone (po 0.01). Conclusion: Prenatal levels of some metabolites of DEHP and DiNP seemed negatively associated with reproductive function of adolescent men. & 2015 Elsevier Inc. All rights reserved.

Keywords: Diisononyl phthalate Diethylhexyl phthalate Late effects Prenatal exposure Semen quality hormones

1. Introduction Phthalates are constituents of consumer products which makes humans continuously exposed (Wittassek et al., 2011). Exposure to phthalates during fetal life has been suggested to cause reproductive disorders in men through endocrine disruption (Sharpe and Skakkebaek, 2008). In male rats, prenatal exposure to diethylhexyl phthalate (DEHP) and diisononyl phthalate (DiNP) induce genital malformations (Gray

Abbreviations: DEHP, diethylhexyl phthalate; DFI, DNA fragmentation index; DiNP, diisononyl phthalate; FSH, follicle-stimulating hormone; HDS, High DNA stainability; LC/MS/MS, liquid chromatography–tandem mass spectrometry; LH, luteinizing hormone; LOD, limit of detection; MECPP, mono-(2-ethyl-5-carboxypentyl) phthalate; MCiOP, mono-(carboxy-iso-octyl) phthalate; MEHHP, mono-(2ethyl-5-hydroxylhexyl) phthalate; MHiNP, mono-(hydroxy-iso-nonyl) phthalate; MEOHP, mono-(2-ethyl-5-oxohexyl) phthalate; MOiNP, mono-(oxo-iso-nonyl) phthalate; MPW, masculinization programming window; SHBG, sex hormonebinding globulin; T, testosterone n Correspondence to: Clinical Research Centre, Molecular Reproductive medicine, Skåne University Hospital Malmö, 205 02 Malmö, Sweden. Fax: þ 46 40 338266. E-mail address: [email protected] (J. Axelsson). http://dx.doi.org/10.1016/j.envres.2015.02.024 0013-9351/& 2015 Elsevier Inc. All rights reserved.

et al., 2000). This is suggested to be due to a reduced fetal testosterone (T) production (Hu et al., 2009), which is reported for both DEHP and DiNP (Borch et al., 2004; Hannas et al., 2011). Human studies have reported that prenatal phthalate exposure was associated with a shorter anogenital distance in boys, indicating a reduced masculinization (Bornehag et al., 2015; Bustamante-Montes et al., 2013; Suzuki et al., 2012; Swan et al., 2005) and perhaps a reduced future fertility (Dean and Sharpe, 2013). Similar exposure has additionally been associated with cryptorchidism (Swan, 2008; Wagner-Mahler et al., 2011), altered reproductive hormone levels (Araki et al., 2014; Main et al., 2006), and with an increased risk of hypospadias (Ormond et al., 2009). However, other studies were inconsistent with those above (Chevrier et al., 2012; Huang et al., 2009; Jensen et al., 2015; Lin et al., 2011a; Main et al., 2006) and only two studies (Bornehag et al., 2015; Huang et al., 2009) have focused on exposure in early pregnancy, considered as the most sensitive period for the developing male reproductive organs (Welsh et al., 2008). In addition, the so called secondary metabolites, which are suggested as the best exposure markers (Wittassek et al., 2011) were only measured in some of these studies.

J. Axelsson et al. / Environmental Research 138 (2015) 264–270

In male rats, the development and size of the reproductive organs are programmed in an early masculinization programming window (MPW) (Macleod et al., 2010; Welsh et al., 2008) during which exposure to the compound dibutyl phthalate negatively affects the future reproductive organ size (Macleod et al., 2010). From a previous study of reproductive function in men from the general population (Axelsson et al., 2013), we had access to maternal serum samples from early pregnancy through a Swedish screening program for rubella. Most of these samples were collected within the suggested corresponding MPW in humans, between the 8th and 14th gestational week (Welsh et al., 2008). Our aim was to study associations between prenatal exposure to DEHP and DiNP, and male reproductive parameters, by measuring secondary metabolites in maternal samples.

2. Subjects and methods This work was carried out in accordance with The Code of Ethics of the World Medical Association (WMA, 2013) and the Uniform Requirements for Manuscripts Submitted to Biomedical Journals. International Committee of Medical Journal Editors (1997). 2.1. Study population Between 2008 and 2010, we invited 1681 men presenting for the military health board. Among them, 241 (14%) accepted participation. In order to expand the participant number, we included another 73 men through announcement in schools, giving initially 314 participants (Axelsson et al., 2013). Inclusion criteria were: living within 60 km from the city of Malmö in southern Sweden; 17–20 years of age; participant and mother born and raised in Sweden. The men signed an informed consent, filled in questionnaires regarding height and weight, previous genital diseases, current smoking and paternal smoking during pregnancy, and finally delivered semen and blood samples. They were paid 500 SEK (55 EURO) for participation. The study was approved by the regional ethical review board. The extent of maternal smoking during pregnancy was assessed through cotinine levels in serum (see Section 2.3). Fortynine of the 314 initially included men were excluded since they or their mother declined further participation, leaving 265 men for search of maternal samples. 2.2. Maternal sampling In Sweden, screening for rubella antibodies in serum is routinely done in early pregnancy. Unless the woman declines, part of the sample is stored in a biobank. We retrieved maternal samples of 112 men born 1989–1992, on whom this study was based. There was no difference in reproductive parameters between these men and the remaining 153 for whom no maternal samples were found (data not shown). The samples were obtained from the 6th to the 35th week of pregnancy (mean 12 weeks). Seventy-seven men (69%) had mothers sampled between 8 and 14 completed gestational weeks. 2.3. Analyses of exposure markers Phthalate metabolites and the nicotine metabolite cotinine were analyzed in maternal serum by liquid chromatography–tandem mass spectrometry (LC/MS/MS). The analyzed phthalate metabolites were the secondary metabolites of DEHP: mono-(2ethyl-5-hydroxylhexyl) phthalate (MEHHP), mono-(2-ethyl-5-oxohexyl) phthalate (MEOHP) and mono-(2-ethyl-5-carboxypentyl)

265

phthalate (MECPP) and of DiNP: mono-(hydroxy-iso-nonyl) phthalate (MHiNP), mono-(oxo-iso-nonyl) phthalat (MOiNP) and mono-(carboxy-iso-octyl) phthalate (MCiOP). For the analysis, aliquots of 100 ml serum were added with isotopically labeled internal standards for all evaluated compounds. The samples were digested with glucuronidase, and proteins were precipitated using acetonitrile. The samples were prepared in 96-well plates and analyzed using a triple quadrupole linear ion trap mass spectrometer (QTRAP 5500; AB Sciex, Foster City, CA, USA) coupled to a liquid chromatography system (UFLCXR, Shimadzu Corporation, Kyoto, Japan; LC/MS/MS). The analysis was performed in negative ion mode. All data acquisition was performed using Analyst 1.6.1 software, and data was processed using Multiquant 2.1 (AB Sciex). Limits of detection (LOD) for metabolite analyses was defined as the concentration corresponding to three times the standard deviation of the ratio of the peak at the same retention time as the analyzed compounds and the corresponding internal standards determined in the chemical blank samples. Coefficients of variation (CV) of a quality control sample were analyzed in all sample batches. A more detailed description of the method and the validation can be found in the Supplementary data. 2.4. Genital examination All men were genitally examined by a physician for total testicular volume using Prader's orchidometer. 2.5. Semen analysis We asked the men to keep 48–72 h of abstinence (which 42% fulfilled) but registered the actual length. Semen was analyzed for sperm concentration, total sperm count, progressive sperm motility and proportion of morphologically normal sperm according to WHO guidelines (World Health Organization, 1999). We additionally analyzed sperm DNA fragmentation index (DFI) and High DNA stainability (HDS) using the Sperm Chromatin Structure Assay (Evenson et al., 2002). 2.6. Analyses of reproductive hormones Serum samples obtained from the men before noon, were analyzed at the laboratory of clinical chemistry, Skåne University Hospital, Sweden. Levels of T, follicle-stimulating hormone (FSH), luteinizing hormone (LH) and sex hormone-binding globulin (SHBG) were analyzed with ElectroChemiLuminiscenceImmunoassay (Roche Cobas), and estradiol by use of an immunofluorometric method (Delfia, Perkin-Elmer). For T, CV was 3.8% at 3.2 nmol/L and 1.6% at 25 nmol/L, and LOD 0.087 nmol/L; FSH had CV 5.5% at 5.0 IU/L and LOD 0.10 IU/L; LH had CV 3.2% at 7.0 IU/L and LOD 0.10 IU/L; SHBG had CV 1.2% at 16 nmol/L, 1.4% at 34 nmol/L, and LOD 0.35 nmol/L; and estradiol had CV 20% at 30 pmol/L, 10% at 280 pmol/L, and LOD 8 pmol/L. Concentration of free T was calculated according to Vermeulen et al. (1999). 2.7. Statistics We used SPSS v 20-22 for statistical analyses, and defined po 0.05 as statistically significant. Correlations between phthalate metabolites were studied using Spearman's rank correlation test. To better fulfill model assumptions regarding normal distribution of residuals, we transformed DFI, HDS, T, free T, LH, FSH and SHBG by the natural logarithm, and sperm concentration and total sperm count by the cubic root. Likewise, we transformed phthalate metabolite concentrations and fetal age by the natural logarithm to increase statistical prediction (Tabachnick and Fidell, 2013). We

266

J. Axelsson et al. / Environmental Research 138 (2015) 264–270

controlled improvements in distributions through values of skewness and kurtosis, tests of normality and histograms, and for residuals also through plots assessing normality, linearity and scedasticity. Associations between exposures and reproductive parameters were studied using regression models adjusted for the men's BMI, current smoking (yes/no), paternal smoking during pregnancy (yes/no), fetal age in days and maternal age and cotinine concentration at sampling as potential confounders based on the literature. Additionally, semen parameters were adjusted for abstinence time [categorized: r48 h (n¼ 34), 49–72 h (n¼ 47), 73– 96 h (n ¼16), 97–120 h (n¼ 10) and Z121 h (n ¼4), (missing in one man)], and hormone levels for time of day at sampling as a continuous variable. Subsequently, the analyses were re-performed without adjustments. Metabolite levels below LOD, were given the value of the half of LOD. We modeled exposure levels both as continuous variables and categorized in tertiles. For categorized exposure, we studied differences in reproductive parameters between the first (lowest) and the third (highest) exposure tertile, and for adjusted statistically significant findings also trends, modeling tertiles as if they were linear variables numbered 1, 2 and 3. To minimize the risk of adult exposures to DiNP or DEHP as an explanation of the associations between prenatal levels of any metabolite and a reproductive outcome, statistically significant findings were also adjusted for urinary levels of the particular metabolite measured in urine collected at the time of semen sampling. The urinary phthalate levels were analyzed according to a previously published method (Toft et al., 2012), and were adjusted for the creatinine level.

3.1. Phthalate metabolite levels and testicular and seminal parameters 3.1.1. Adjusted For categorized exposure, men in the highest tertile of prenatal MCiOP exposure had 4.3 mL (95% CI: 0.89, 7.6 mL) lower testicular volume than men in the lowest tertile (41 mL vs. 45 mL) (p ¼0.01, ptrend ¼0.02). For MCiOP, additionally, semen volume was 0.87 mL (95% CI: 0.28, 1.5 mL) lower (ptrend ¼0.004) in the highest exposure tertile, and for MEHHP 0.70 mL (95% CI: 0.090, 1.3 mL) lower (ptrend ¼0.03), than in lowest respective tertile (2.9 mL and 2.8 mL vs. 3.8 mL and 3.5 mL, p ¼0.004 and 0.03). These results are summarized in Table 4. As continuous variables, both MEHHP and MCiOP had negative linear associations with semen volume (p ¼ 0.005 and 0.048, regression coefficients [B]¼  0.62 [95% confidence interval (CI):  1.0,  0.19] and  0.30 [95% CI:  0.59,  0.003]). These results are summarized in Supplementary material, Table 2. 3.1.2. Unadjusted For unadjusted analyses, similar results as above were seen for MEHHP (categorized, and as continuous [Fig. 1]) in relation to semen volume, and for MCiOP (categorized) in relation to semen volume, and testicular volume (Fig. 2). In addition, total sperm count was lower in the highest than in the lowest exposure tertile of MEHHP. For continuous exposure markers, the statistical significance of the adjusted association between MCiOP and semen volume was not seen without adjustment, whereas the opposite was found for MEOHP and MECPP. These results are summarized in Supplementary material, Tables 3 and 4.

3. Results 3.2. Phthalate metabolite levels and reproductive hormones Background data are given in Table 1, and exposure data in Table 2. Previous reproductive tract problems are shown in Supplementary material, Table 1. Twenty-seven (24%) men were smokers and thirty-three (30%) men reported paternal smoking (missing in eight men). Correlations between metabolites (Table 3) were statistically significant within the DINP and DEHP families (rho: 0.38–0.86) but not between.

3.2.1. Adjusted In the adjusted analyses, men in the highest tertile of MCiOP had 30% (95% CI: 3.6, 63) higher FSH than men in the lowest tertile (3.5 IU/L vs. 2.7 IU/L, p ¼0.02, ptrend ¼0.03). These results are summarized in Table 5. When testing exposure markers as continuous variables, both MCiOP and MOiNP were linearly associated with FSH (p ¼0.02 and

Table 1 Background information.

Age (years) BMI (kg/m2) Testicular size, left þ right (mL) Sperm concentration (  106/mL) Total sperm count (  106) Semen volume (mL) Progressive sperm (%) Morphologically normal (%) DNA fragmentation index (%) High DNA stainability (%) Testosterone (nmol/L) Free testosterone (nmol/L) Lutenizing hormone (IU/L) Sex-hormone binding globulin (nmol/L) Follicle-stimulating hormone (IU/L) Estradiol (pmol/l) Time of day at men's sampling (hh:mm) Fetal age at sampling (days) Maternal age at sampling (years) Maternal cotinine (ng/mL)

n

Mean 7SD

Minimum

Median

Maximum

112 112 112 111 111 112 112 112 108 108 112 112 112 112 112 112 112 112 112 112

18.3 7 0.41 237 3.2 437 7.0 697 59 1907 210 2.7 7 1.3 52 7 17 8.8 7 5.6 117 5.7 117 7.0 187 5.7 0.36 7 0.10 4.7 7 1.8 337 11 3.4 7 1.7 88 7 23 09:34 7 00:35 87 7 34 28 7 5.2 42 7 73

17.57 18 27 0.08 0.28 0.1 2.0 0.0 4.0 4.0 5.9 0.16 2.0 10 1.1 37 08:20 39 17 o LOD

18.3 22 45 53 140 2.6 58 9.0 10 10 17 0.35 4.2 31 3.3 87 09:34 81 27 0.74

20.5 37 43 300 1400 9.1 86 21 38 47 41 0.73 12 67 11 140 11:10 240 41 370

J. Axelsson et al. / Environmental Research 138 (2015) 264–270

267

Table 2 Levels of metabolites (ng/mL) in maternal sera, detection limits and imprecision. Mother compound

Metabolite

n

Mean 7SD

DEHP

MEHHP MEOHP MECPP

112 112 112

0.117 0.076 0.26 7 0.087 0.78 7 0.60

DINP

MHiNP MOiNP MCiOP

112 112 112

0.029 7 0.054 0.023 7 0.020 0.247 0.40

Minimum

Median

Maximum

LOD

0.026 0.10 0.12

0.92 0.26 0.64

0.61 0.54 5.2

0.02 0.02 0.005

o LOD o LOD 0.030

0.016 0.020 0.12

0.44 0.19 0.24

0.01 0.01 0.01

Proportiono LOD (%)

Coefficient of variation (%)

0 0 0

9 7 3

22 6.3 0

4 8 2

Abbreviations: LOD, limit of detection; SD, standard deviation.

Table 3 Correlations (Spearman's rho) between phthalate metabolites. MEHHP MEHHP MEOHP MECPP MHiNP MOiNP MCiOP n

1

MEOHP 0.42 1

n

MECPP nn

0.78 0.38nn 1

MHiNP

MOiNP

MCiOP

0.17 0.13 0.059 1

0.18 0.12 0.064 0.86nn 1

0.17 0.067 0.10 0.79nn 0.62nn 1

po 0.05. p o0.01.

nn

0.01, B ¼0.13 [95% CI: 0.022, 0.24] and 0.21 [95% CI: 0.049, 0.37]), and MHiNP and MOiNP with LH (p ¼0.01 for both, B ¼0.093 [95% CI: 0.023, 0.16] and 0.15 [95% CI: 0.033, 0.26]). These results are summarized in Supplementary material, Table 5. 3.2.2. Unadjusted With respect to LH, statistically significant associations were seen as indicated above, whereas for FSH, this was only true for MOiNP as a continuous variable (see Supplementary material, Tables 6 and 7). In addition, most DEHP metabolites as continuous variables had negative linear associations with T and free T (see Supplementary material, Table 7). Finally, free T was lower in the highest than in the lowest tertile of MECPP. 3.3. Additional adjustment for adult levels The statistically significant mean differences and regression coefficients for the relevant metabolites remained similar (largest change 6.4%) and statistically significant after additional adjustment for adult urinary levels (in Supplementary material, Table 8) of the metabolite (data not shown).

4. Discussion The most notable result of this study was a finding of negative associations between levels in pregnant women's serum of one metabolite of DiNP and one of DEHP, and the semen volume of their sons. In addition, some metabolites of DiNP had positive associations with levels of FSH or LH, and one metabolite a negative association with testicular volume. Thus, our study found that fetal exposure to these compounds seemed negatively associated with adolescent males' reproductive function. The associations seemed not to be confounded by their exposure to the same compounds in adulthood. One of the strengths of this study is the access to maternal samples from early pregnancy, a critical period for the male reproductive system (Welsh et al., 2008). Most of the samples were collected within a window suggested for human fetal

programming of the reproductive function and organ size (Macleod et al., 2010; Welsh et al., 2008). Despite a low participation rate, a previous study indicated that men recruited at the medical examination prior to military service mirror the general population in terms of reproductive function (Andersen et al., 2000). Thus, it seems unlikely that our results were caused by selection bias. Multiple testing (6 metabolites and 14 outcomes), might have increased the risk of chance findings (Ioannidis, 2005) but considering our exploratory purpose, we avoided to adjust for the many comparisons (Bender and Lange, 2001; Rothman, 1990). We also omitted adjustment for previous cryptorchidism, since this may be one of the pathogenic mechanisms that links prenatal phthalate exposure to a deteriorated male reproductive function (Gray et al., 2000). The limited size of our study may have reduced the power of the analyses and underestimated any true effects (Hodgson et al., 2010). Simultaneously, levels of DEHP- and DiNP metabolites vary considerably within individuals (Baird et al., 2010; Frederiksen et al., 2013; Fromme et al., 2007; Peck et al., 2010), including in pregnant women (Adibi et al., 2008; Cantonwine et al., 2014; Irvin et al., 2010), which might have decreased the power even further (Frederiksen et al., 2013). Still, these secondary metabolites are considered as the best available exposure markers (Calafat et al., 2013; Wittassek et al., 2011), and have been reported fairly reproducible by others (Kim et al., 2014; Townsend et al., 2013), also during pregnancy (Suzuki et al., 2009). Phthalate exposure is usually determined by measuring metabolite levels in urine. In this study, exposure was measured in serum samples, which are available in large biobanks, whereas urinary samples generally are not. Thus, the present study was not possible to perform by analyses of urine. There are however some disadvantages with the use of serum instead of urine. E.g., if the serum sample post-collection becomes contaminated with the ubiquitous di-esters of phthalate, the lipase activity of serum will split the di-esters into primary metabolites (Wittassek et al., 2011). The serum concentrations of these metabolites are therefore not reliable measures of exposure, contrary to concentrations in urine which has no lipase activity. Therefore, we did not investigate any primary metabolite. In addition, the metabolite levels in serum are much lower than in urine which sets special demands on the used method. However, compared with other reports in the field, the performance of our LC/MS/MS-based analysis allowed measurement of phthalate metabolite concentrations 2–1000 times lower than previous studies on levels in serum (Frederiksen et al., 2010; Hines et al., 2009; Hogberg et al., 2008; Specht et al., 2014) or amniotic fluid (Jensen et al., 2015). These studies mostly detected MECPP (Frederiksen et al., 2010; Hines et al., 2009) and MCiOP (Frederiksen et al., 2010), at similar levels as in our study. Levels of secondary metabolites in serum, further, seem correlated with levels in urine (Frederiksen et al., 2010) like maternal levels in urine with levels in cord blood of the newborn (Lin et al., 2011b). In addition, since our maternal serum levels were similar to those

Semen volume (mL)

0.95 0.83 0.56 0.55 0.13 0.49 9.9 10 10 9.2 9.4 9.4 9.8 9.8 9.3 10 12 10 0.96 0.78 0.51 0.75 0.77 0.40 11 11 11 12 11 11 11 11 11 11 11 12

p 3rd 1st 3rd 1st

DNA frag-mentation index (%)c

p

J. Axelsson et al. / Environmental Research 138 (2015) 264–270

High DNA stainability (%)c

268

10 8 R2 = 0.10

6 4

p < 0.001

2 0

0.70 0.26 0.43 0.83 0.26 0.89

-3

-2

-1

0

Ln maternal MEHHP

c

b

Testicular volume (mL)

Adjusted for BMI, abstinence time (except testicular size), own smoking, paternal smoking, fetal age, and maternal age and cotinine level. Back-transformed from cubic-root transformation. Back-transformed from ln transformation.

Fig. 1. Semen volume and maternal DEHP metabolite concentration. Unadjusted semen volume according to maternal MEHHP levels in early pregnancy (ln transformed).

a

9.7 9.9 9.2 8.9 9.9 8.4 150 180 140 190 170 150 45 43 43 44 43 45 MEHHP MOEHP MCEPP MHiNP MOiNP MCiOP

42 44 43 43 42 41

0.12 0.77 0.96 0.86 0.59 0.01

65 51 56 48 54 51

59 66 53 66 60 55

0.63 0.26 0.76 0.19 0.64 0.73

220 160 180 150 160 190

0.09 0.61 0.26 0.41 0.69 0.37

3.5 3.2 3.4 3.2 3.1 3.8

2.8 2.9 2.9 3.1 3.0 2.9

0.03 0.40 0.15 0.57 0.86 0.004

51 52 53 53 47 51

52 54 54 55 54 52

0.90 0.58 0.88 0.66 0.09 0.78

9.1 8.2 8.0 8.6 8.3 8.6

p 3rd 1st p 3rd 1st p 3rd 1st p 3rd 1st p 3rd 1st p 3rd 1st Exposure tertile

Semen volume (mL) Total sperm count (x106)b Sperm concentration (  106/mL)b Total testicular volume (mL)

Table 4 Comparisons of adjusteda mean values of testicular volume and semen parameters in first and third exposure tertile.

Progressive sperm (%)

Proportion normal sperm (%)

-4

70 60 50 40 30 20

------ p = 0.02 ------

10 0

Low

Middle

High

Maternal MCiOP tertile Fig. 2. Testicular volume according to maternal DiNP metabolite level. Unadjusted total testicular volume and categorized MCiOP in maternal serum from pregnancy [3.7 mL (95% CI: 0.50, 6.9 mL; p ¼ 0.02) lower in 3rd than in 1st exposure tertile; ptrend 0.02]. Boxes depict inter-quartile ranges (25th–75th percentile), and bold horizontal lines median values. T-bars extend 1.5 times the height of the box or if less, to minimum and maximum.

cord blood levels, and to levels in amniotic fluid (Jensen et al., 2012), this indicates that fetuses are not protected from maternal exposure, and that maternal serum levels could be used as markers of fetal exposure. Although it has been suggested that phthalate exposure does not imply a decreased T production in the human fetus (Jensen et al., 2015; Johnson et al., 2012; Spade et al., 2014), DEHP has been reported to decrease T production in the human adult testis (Desdoits-Lethimonier et al., 2012; Hallmark et al., 2007). In addition, other mechanisms than an altered T production could be implicated in the effects of phthalates on the male reproductive development (Anand-Ivell and Ivell, 2014; Kim et al., 2010). Thus, rat studies have indicated that prenatal DEHP- and DiNP exposure reduce fetal T production, increase LH (Borch et al., 2004; Hannas et al., 2011) and disturb the formation of testicles and seminal vesicles (Andrade et al., 2006; Gray et al., 2000; Macleod et al., 2010; Stroheker et al., 2005). Seminal vesicles produce the predominant part of the seminal fluid (Mortimer, 1994), why a disturbed growth of these vesicles might lead to a lower semen

J. Axelsson et al. / Environmental Research 138 (2015) 264–270

269

Table 5 Comparisons of adjusteda mean hormone levels in first and third exposure tertile. T (nmol/L)b

Free T (nmol/L)b

LH (IU/L)b

FSH (IU/L)b

SHBG (nmol/L)b

Estradiol (pmol/l)

Exposure tertile

1st

3rd

p

1st

3rd

p

1st

3rd

p

1st

3rd

p

1st

3rd

p

1st

3rd

p

MEHHP MOEHP MCEPP MHiNP MOiNP MCiOP

18 18 19 18 17 17

16 16 17 17 17 17

0.21 0.15 0.12 0.62 0.77 0.85

0.37 0.36 0.39 0.37 0.35 0.35

0.33 0.34 0.35 0.35 0.33 0.35

0.11 0.38 0.09 0.44 0.46 0.74

4.7 4.9 4.5 4.3 4.0 4.2

4.0 4.2 4.5 5.0 4.6 4.7

0.08 0.07 0.98 0.07 0.06 0.19

2.8 3.0 3.0 2.9 2.7 2.7

3.0 3.0 3.2 3.2 3.3 3.5

0.47 0.99 0.52 0.31 0.05 0.02

90 92 92 85 90 84

85 87 88 92 83 87

0.29 0.30 0.47 0.23 0.18 0.53

30 33 31 29 29 28

30 29 30 31 32 32

0.94 0.05 0.79 0.49 0.35 0.18

Abbreviations: FSH, follicle-stimulating hormone; LH, luteinizing hormone; SHBG, sex hormone-binding globulin; T, testosterone. a b

Adjusted for BMI, own smoking, paternal smoking, fetal age, maternal age and cotinine level, and time of day at man's sampling. Back-transformed from ln transformed values.

volume (Sharpe and Skakkebaek, 2008), this being in accordance with our findings. In spite of the correlations we found between metabolites of the same parent compound, the associations with the reproductive parameters were most pronounced for single metabolites only. However, for the other metabolites of the same parent compound, the trend of association was in the same direction in the majority of cases, although not statistically significant. We found DiNP to be more strongly associated with the reproductive outcomes than was DEHP, although DiNP has been reported as being less toxic in rats (Gray et al., 2000). However, a recent human study reported that levels of DiNP metabolites in maternal urine from the 1st trimester were associated with a shorter anogenital distance in boys, with a similar but not statistically significant association for metabolites of DEHP (Bornehag et al., 2015). These findings may be explained by a higher diurnal variation in DEHP- than in DiNP metabolite levels (Cantonwine et al., 2014; Fromme et al., 2007), which would decrease the precision of the assessment of DEHP exposure, and thereby also decrease the chance of finding statistically significant associations with reproductive outcomes. Theoretically, due to the long half-lives (Wittassek and Angerer, 2008; Wittassek et al., 2011), the carboxylated metabolites (MECPP and MCiOP) should be the best exposure markers. However, in our study, among the metabolites of DEHP, only levels of the hydroxylated metabolite (MEHHP) were associated with reproductive outcomes in the adjusted analyses. Accordingly, MEHHP is the only secondary DEHP metabolite reported as deleterious to the fetal testis (Chauvigne et al., 2009). We found no statistically significant associations for the corresponding hydroxylated DiNP metabolite (MHiNP), which might be due to a difference between MEHHP and MHiNP as considers the position of the hydroxy group (Koch et al., 2005). Although adult human phthalate exposure may be associated with altered hormone levels or a reduced semen quality (Kay et al., 2014), including sperm motility (Axelsson et al., manuscript under preparation), our findings were largely unaffected by adjustment for the adult exposure to DEHP or DiNP. This seems to agree with animal findings of a higher sensitivity to phthalates in utero than later in life (Kay et al., 2014). Still, an early postnatal exposure could aggravate the effects of a prenatal exposure (Macleod et al., 2010), why the elevated uptakes of phthalates reported in human infants through bottle-feeding or PVC flooring (Carlstedt et al., 2013) may have played a role in our men. However, for our cohort we have no access to any data on neonatal exposure.

5. Conclusions Through maternal levels of certain metabolites, prenatal

exposure to DiNP and DEHP seemed negatively associated with the reproductive function of these adolescent men.

Disclosure statement The authors have no conflicts of interest that could inappropriately influence these results.

Funding sources This work was supported by the Swedish Cancer Society [Grant number CAN 2009/817], Swedish Governmental Funding for medical training and research, Skane University Hospital Funds, Swedish Childhood Cancer Society [Grant number NO 10/0030], Swedish Environmental Protection Agency (grant no. 2151213), Interreg (ReproHigh) and Skane County Council's Research and Development Foundation [Grant numbers REGSKANE-58821, 120081, 192641, 270931, 350291]. The sponsors had no role in study design; in collection, analysis and interpretation of data; in writing of the report; or in decision to submit the article for publication.

Acknowledgments We thank Erna Jeppson Stridsberg for help in recruiting participants, Agneta Kristensen for analyzing phthalate metabolites, and Cecilia Tingsmark, Mania Winitsky and others for analyses of semen.

Appendix A. Supplementary material Supplementary data associated with this article can be found in the online version at http://dx.doi.org/10.1016/j.envres.2015.02. 024.

References Adibi, J.J., et al., 2008. Characterization of phthalate exposure among pregnant women assessed by repeat air and urine samples. Environ. Health Perspect. 116, 467–473. Anand-Ivell, R., Ivell, R., 2014. Insulin-like factor 3 as a monitor of endocrine disruption. Reproduction 147, R87–R95. Andersen, A.G., et al., 2000. High frequency of sub-optimal semen quality in an unselected population of young men. Hum. Reprod. 15, 366–372. Andrade, A.J., et al., 2006. A dose response study following in utero and lactational exposure to di-(2-ethylhexyl) phthalate (DEHP): reproductive effects on adult male offspring rats. Toxicology 228, 85–97.

270

J. Axelsson et al. / Environmental Research 138 (2015) 264–270

Araki, A., et al., 2014. Association between maternal exposure to di(2-ethylhexyl) phthalate and reproductive hormone levels in fetal blood: the Hokkaido study on environment and children's health. PLoS One 9, e109039. Axelsson, J., et al., 2013. The impact of paternal and maternal smoking on semen quality of adolescent men. PLoS One 8, e66766. Baird, D.D., et al., 2010. Within-person variability in urinary phthalate metabolite concentrations: measurements from specimens after long-term frozen storage. J. Expo. Sci. Environ. Epidemiol. 20, 169–175. Bender, R., Lange, S., 2001. Adjusting for multiple testing-when and how? J. Clin. Epidemiol. 54, 343–349. Borch, J., et al., 2004. Steroidogenesis in fetal male rats is reduced by DEHP and DINP, but endocrine effects of DEHP are not modulated by DEHA in fetal, prepubertal and adult male rats. Reprod. Toxicol. 18, 53–61. Bornehag, C.G., et al., 2015. Prenatal phthalate exposures and anogenital distance in Swedish boys. Environ. Health Perspect. 123, 101–107. Bustamante-Montes, L.P., et al., 2013. Prenatal exposure to phthalates is associated with decreased anogenital distance and penile size in male newborns. J. Dev. Orig. Health Dis. 4, 300–306. Calafat, A.M., et al., 2013. Misuse of blood serum to assess exposure to bisphenol A and phthalates. Breast Cancer Res. 15, 403. Cantonwine, D.E., et al., 2014. Urinary phthalate metabolite concentrations among pregnant women in Northern Puerto Rico: distribution, temporal variability, and predictors. Environ. Int. 62C, 1–11. Carlstedt, F., et al., 2013. PVC flooring is related to human uptake of phthalates in infants. Indoor Air 23, 32–39. Chauvigne, F., et al., 2009. Time- and dose-related effects of di-(2-ethylhexyl) phthalate and its main metabolites on the function of the rat fetal testis in vitro. Environ Health Perspect 117, 515–521. Chevrier, C., et al., 2012. Maternal urinary phthalates and phenols and male genital anomalies. Epidemiology 23, 353–356. Dean, A., Sharpe, R.M., 2013. Clinical review: anogenital distance or digit length ratio as measures of fetal androgen exposure: relationship to male reproductive development and its disorders. J. Clin. Endocrinol. Metab. 98, 2230–2238. Desdoits-Lethimonier, C., et al., 2012. Human testis steroidogenesis is inhibited by phthalates. Hum. Reprod. 27, 1451–1459. Evenson, D.P., et al., 2002. Sperm chromatin structure assay: its clinical use for detecting sperm DNA fragmentation in male infertility and comparisons with other techniques. J. Androl. 23, 25–43. Frederiksen, H., et al., 2010. Correlations between phthalate metabolites in urine, serum, and seminal plasma from young Danish men determined by isotope dilution liquid chromatography tandem mass spectrometry. J. Anal. Toxicol. 34, 400–410. Frederiksen, H., et al., 2013. Temporal variability in urinary phthalate metabolite excretion based on spot, morning, and 24-h urine samples: considerations for epidemiological studies. Environ. Sci. Technol. 47, 958–967. Fromme, H., et al., 2007. Occurrence and daily variation of phthalate metabolites in the urine of an adult population. Int. J. Hyg. Environ. Health 210, 21–33. Gray Jr., L.E., et al., 2000. Perinatal exposure to the phthalates DEHP, BBP, and DINP, but not DEP, DMP, or DOTP, alters sexual differentiation of the male rat. Toxicol. Sci. 58, 350–365. Hallmark, N., et al., 2007. Effects of monobutyl and di(n-butyl) phthalate in vitro on steroidogenesis and Leydig cell aggregation in fetal testis explants from the rat: comparison with effects in vivo in the fetal rat and neonatal marmoset and in vitro in the human. Environ. Health Perspect. 115, 390–396. Hannas, B.R., et al., 2011. Dose-response assessment of fetal testosterone production and gene expression levels in rat testes following in utero exposure to diethylhexyl phthalate, diisobutyl phthalate, diisoheptyl phthalate, and diisononyl phthalate. Toxicol. Sci. 123, 206–216. Hines, E.P., et al., 2009. Concentrations of phthalate metabolites in milk, urine, saliva, and Serum of lactating North Carolina women. Environ. Health Perspect. 117, 86–92. Hodgson, S., Khaw, F.-M., 2010. Epidemiological methods for attributing illness In: Ayres, J.G., et al. (Eds.), Environmental Medicine. Hodder Education, London, pp. 25–43. Hogberg, J., et al., 2008. Phthalate diesters and their metabolites in human breast milk, blood or serum, and urine as biomarkers of exposure in vulnerable populations. Environ. Health Perspect. 116, 334–339. Hu, G.X., et al., 2009. Phthalate-induced testicular dysgenesis syndrome: leydig cell influence. Trends Endocrinol. Metab. 20, 139–145. Huang, P.C., et al., 2009. Association between prenatal exposure to phthalates and the health of newborns. Environ. Int. 35, 14–20. Ioannidis, J.P., 2005. Why most published research findings are false. PLoS Med. 2, e124. Irvin, E.A., et al., 2010. An estimate of phthalate exposure among pregnant women living in Trujillo, Peru. Chemosphere 80, 1301–1307. Jensen, M.S., et al., 2015. Amniotic fluid phthalate levels and male fetal gonad function. Epidemiology 26, 91–99. Jensen, M.S., et al., 2012. Phthalates and perfluorooctanesulfonic acid in human amniotic fluid: temporal trends and timing of amniocentesis in pregnancy. Environ. Health Perspect. 120, 897–903. Johnson, K.J., et al., 2012. Of mice and men (and rats): phthalate-induced fetal testis

endocrine disruption is species-dependent. Toxicol. Sci. 129, 235–248. Kay, V.R., et al., 2014. Reproductive and developmental effects of phthalate diesters in males. Crit. Rev. Toxicol. 44, 1–32. Kim, S., et al., 2014. Urinary phthalate metabolites among elementary school children of Korea: sources, risks, and their association with oxidative stress marker. Sci. Total Environ. 472, 49–55. Kim, T.S., et al., 2010. Effects of in utero exposure to DI(n-Butyl) phthalate on development of male reproductive tracts in Sprague–Dawley rats. J. Toxicol. Environ. Health A 73, 1544–1559. Koch, H.M., et al., 2005. New metabolites of di(2-ethylhexyl)phthalate (DEHP) in human urine and serum after single oral doses of deuterium-labelled DEHP. Arch. Toxicol. 79, 367–376. Lin, L.C., et al., 2011a. Associations between maternal phthalate exposure and cord sex hormones in human infants. Chemosphere 83, 1192–1199. Lin, S., et al., 2011b. Phthalate exposure in pregnant women and their children in central Taiwan. Chemosphere 82, 947–955. Macleod, D.J., et al., 2010. Androgen action in the masculinization programming window and development of male reproductive organs. Int. J. Androl. 33, 279–287. Main, K.M., et al., 2006. Human breast milk contamination with phthalates and alterations of endogenous reproductive hormones in infants three months of age. Environ. Health Perspect. 114, 270–276. Mortimer, D. (Ed.), 1994. Sperm physiology In: Practical Laboratory Andrology. Oxford University Press, New York, p. 393. Ormond, G., et al., 2009. Endocrine disruptors in the workplace, hair spray, folate supplementation, and risk of hypospadias: case-control study. Environ. Health Perspect. 117, 303–307. Peck, J.D., et al., 2010. Intra- and inter-individual variability of urinary phthalate metabolite concentrations in Hmong women of reproductive age. J. Expo. Sci. Environ. Epidemiol. 20, 90–100. Rothman, K.J., 1990. No adjustments are needed for multiple comparisons. Epidemiology 1, 43–46. Sharpe, R.M., Skakkebaek, N.E., 2008. Testicular dysgenesis syndrome: mechanistic insights and potential new downstream effects. Fertil. Steril. 89, e33–e38. Spade, D.J., et al., 2014. Differential response to abiraterone acetate and di-n-butyl phthalate in an androgen-sensitive human fetal testis xenograft bioassay. Toxicol. Sci. 138, 148–160. Specht, I.O., et al., 2014. Associations between serum phthalates and biomarkers of reproductive function in 589 adult men. Environ. Int. 66, 146–156. Stroheker, T., et al., 2005. Evaluation of anti-androgenic activity of di-(2-ethylhexyl) phthalate. Toxicology 208, 115–121. Suzuki, Y., et al., 2009. Exposure assessment of phthalate esters in Japanese pregnant women by using urinary metabolite analysis. Environ. Health Prev. Med. 14, 180–187. Suzuki, Y., et al., 2012. Foetal exposure to phthalate esters and anogenital distance in male newborns. Int. J. Androl. 35, 236–244. Swan, S.H., 2008. Environmental phthalate exposure in relation to reproductive outcomes and other health endpoints in humans. Environ. Res. 108, 177–184. Swan, S.H., et al., 2005. Decrease in anogenital distance among male infants with prenatal phthalate exposure. Environ. Health Perspect. 113, 1056–1061. Tabachnick, B.G., Fidell, L.S. (Ed.), 2013. Multiple regression In: Using Multivariate Statistics. Pearson, Boston, MA, London, pp. 117–196. Toft, G., et al., 2012. Association between pregnancy loss and urinary phthalate levels around the time of conception. Environ. Health Perspect. 120, 458–463. Townsend, M.K., et al., 2013. Within-person reproducibility of urinary bisphenol A and phthalate metabolites over a 1 to 3 year period among women in the Nurses' Health Studies: a prospective cohort study. Environ. Health 12, 80. Uniform Requirements for Manuscripts Submitted to Biomedical Journals. International Committee of Medical Journal Editors, 1997. N. Engl. J. Med. 336, 309–315. Vermeulen, A., et al., 1999. A critical evaluation of simple methods for the estimation of free testosterone in serum. J. Clin. Endocrinol. Metab. 84, 3666–3672. Welsh, M., et al., 2008. Identification in rats of a programming window for reproductive tract masculinization, disruption of which leads to hypospadias and cryptorchidism. J. Clin. Invest. 118, 1479–1490. Wagner-Mahler, K., et al., 2011. Prospective study on the prevalence and associated risk factors of cryptorchidism in 6246 newborn boys from Nice area, France. Int. J. Androl. 34, e499–e510. Wittassek, M., Angerer, J., 2008. Phthalates: metabolism and exposure. Int. J. Androl. 31, 131–138. Wittassek, M., et al., 2011. Assessing exposure to phthalates-the human biomonitoring approach. Mol. Nutr. Food Res. 55, 7–31. WMA., 2013. WMA Declaration of Helsinki-Ethical Principles for Medical Research Involving Human Subjects. Vol. 2014. World Health Organization, 1999. WHO Laboratory Manual for the Examination of Human Semen and Sperm-Cervical Mucus Interaction. Cambridge University Press, Cambridge, UK; New York, NY, Published on behalf of the World Health Organization [by].