-
PDF
- Split View
-
Views
-
Cite
Cite
Anne Marie L. Thomsen, Anders H. Riis, Jørn Olsen, Bo A.G. Jönsson, Christian H. Lindh, Niels H. Hjollund, Tina Kold Jensen, Jens Peter Bonde, Gunnar Toft, Female exposure to phthalates and time to pregnancy: a first pregnancy planner study, Human Reproduction, Volume 32, Issue 1, 1 January 2017, Pages 232–238, https://doi.org/10.1093/humrep/dew291
Close - Share Icon Share
Is female exposure to phthalate metabolites associated with reduced fecundity, as estimated by prolonged time to pregnancy (TTP)?
Female exposure to monoethyl phthalate (MEP) but not monobutyl phthalate (MBP), monobenzyl phthalate (MBzP) and monoethylhexyl phthalate (MEHP) was associated with a longer TTP.
Male exposure to phthalates is potentially associated with adverse effects on human fecundity in epidemiological studies, but little is known about the potential effects on female reproduction.
A cohort study with prospective data based on 229 women from a Danish cohort of 430 first pregnancy planning couples enrolled in 1992–1994. In 2009, urinary analyses of phthalate metabolites were performed on stored urine samples from this cohort.
We analyzed MEP, MBP, MBzP and MEHP in female morning spot urine samples collected daily during the first 10 days of menstrual cycles after discontinuation of contraception. The exposure assessment was based on the mean of two measurements from each woman collected in a period of 6 menstrual cycles. We used Cox regression with discrete time to estimate fecundability ratios (FRs) and 95% CI in relation to the average urine metabolite concentration exposure level, controlled for age and BMI, and the time-varying variables smoking and alcohol.
Urinary concentration of MEP was associated with a decreased fecundity (adjusted FR 0.79; 95% CI: 0.63; 0.99) corresponding to a 21% decreased probability of conception for each natural log (ln) unit increase in MEP. No significant association with TTP was found for MBP, MBzP and MEHP.
Subfertile women were overrepresented in the study population due to exclusion of 77 high fertile women who became pregnant in the first cycle when urine collection began.
Our results suggest that female exposure to MEP may have an adverse effect on female fecundity, but these findings need to be replicated in a larger and newer cohort study with sufficient exposure contrast if the use of diethyl phthalate (DEP) and thereby MEP in the future potentially should be regulated in cosmetics and industrial consumer products.
The original data collected were founded by Aarhus University Research Foundation, the Danish Medical Research Council and the Danish Medical Health Insurance Foundation. There are no conflicts of interest to be declared.
N/A
Introduction
There is increasing evidence to indicate that environmental exposure to phthalates is associated with adverse effects on human fecundity (Swan, 2008; Meeker et al., 2009; Buck Louis, 2011). Phthalates are a group of non-persistent chemicals which are used in a variety of industrial consumer products and produced in high volume annually worldwide. The high molecular weight phthalate di-(2-ethylhexyl) phthalate (DEHP) is mainly used to add flexibility to polyvinyl chloride and other plastics products (Wormuth et al., 2006; Cao, 2010). Low molecular weight phthalates such as dibutyl phthalate (DBP) and diethyl phthalate (DEP) are used as solvents in consumer and personal care products to hold color and scent (Duty et al. 2005). Phthalates easily migrate from the products to the external environment because they are not covalently bound to the plastic materials (Wittassek et al., 2011). Humans are primarily exposed to phthalates via ingestion of contaminated food from packaging materials or dermally via personal care products (cosmetics, lotion and perfumes) (Wormuth et al., 2006; Bradley et al., 2013). As a result, human exposure is widespread and starts in fetal life. Phthalate metabolites have been measured in >95% of random urine samples despite the short half-lives for most phthalate metabolites (Koch et al., 2006; Koch and Calafat, 2009).
During the last decades, it has been suspected that some phthalates and their phthalate metabolites may have endocrine-disrupting effects. A study evaluating the different aspects of the biological impact of phthalates indicated that several mechanisms of action may be involved in the reproductive toxicity of the DEP metabolite; monoethyl phthalate (MEP) including endocrine disruption (increased estradiol and decreased testosterone level), aryl hydrocarbon receptor induction, increased oxidative stress and decreased anti-oxidative defense (Mankidy et al., 2013). Most of the epidemiological literature has focused on male reproductive health. In utero and perinatal exposures to phthalates have been associated with signs of antiandrogenic effects manifested as decreased anogenital distance, hypospadias and altered endogenous reproductive hormone levels, but the results are not all consistent (Main et al., 2006; Swan, 2008; Huang et al., 2009; Ormond et al., 2009; Nassar et al., 2010). In adults, urinary metabolites concentrations of DEP, DBP and DEHP have been related to decreased semen quality (Hauser et al., 2006, 2007).
In females, an association between DEHP and endometriosis has been suggested (Cobellis et al., 2003). Moreover, a potential association between DEHP among pregnant women and gestational duration has been reported but with inconsistent results (Latini et al., 2003; Wolff et al., 2008; Adibi et al., 2009). The DEHP metabolite monoethylhexyl phthalate (MEHP) was in one study associated with an increased odds of pregnancy loss (Toft et al., 2012), and in the EARTH Study urinary concentrations of DEHP metabolites were associated with pregnancy loss among women undergoing assisted reproduction technologies (ART) (Messerlian et al., 2016a). Another recent study found on the contrary that DEHP metabolites were associated with reduced risk of pregnancy loss (Jukic et al., 2016). The EARTH Study reported that urinary concentrations of DEHP metabolites were inversely associated with clinical pregnancy and live birth following ART (Hauser et al., 2016). Another EARTH Study found a negative association between higher urinary concentrations of DEHP metabolites and antral follicle count among women seeking infertility treatment (Messerlian et al., 2016b). Regarding female fecundity, a few studies have been performed, and they do not conclude an adverse effect of phthalate exposure. Shorter time to pregnancy (TTP) was associated with DEHP metabolites in a pregnancy-based TTP study from the INUENDO cohort (Specht et al., 2015). In two cohort studies, the LIFE Study (Buck Louis et al., 2014), and the North Carolina Early Pregnancy Study (Jukic et al., 2016), and in one pregnancy-based TTP study, the MIREC Study (Velez et al., 2015), no association between phthalate exposure and TTP was found. When interpreting these results, change in exposure over time and between populations should be taken into consideration.
Since there are few studies with prospective data on the association between phthalate metabolites in female urine and TTP, the aim of the current study was to investigate this potential relationship further. We used data from the Danish First Pregnancy Planner Study to perform a detailed evaluation of the association between high and low molecular weight urinary phthalate metabolite concentrations and time-varying covariates with TTP at a time period with a relatively high exposure.
Materials and Methods
Study subjects
The study subjects were recruited as part of the Danish First Pregnancy Planner Study. Briefly, from 1992 to 1994 a total of 430 couples were enrolled in the study after nationwide mailing of personal letters to >50 000 trade union members (nurses, office workers, day-care workers and metal workers) that were 20–35 years of age, childless and living with a partner of the opposite sex. Only couples without previous pregnancy, no prior knowledge of own fertility, and who for the first time intended to discontinue contraception to become pregnant were eligible. Follow-up started when birth control stopped and lasted for a maximum of six menstrual cycles, or until a clinical pregnancy was achieved within the follow-up time. The couples were enrolled at the Department of Occupational Medicine in Aarhus (West Center) and the Department of Growth and Reproduction in Copenhagen (East Center). At enrollment, the couples filled out a baseline questionnaire on occupation and health. On the 21st day of each menstrual cycle, the couples completed a questionnaire, which allowed for the collection of time-specific data on covariates among others smoking and alcohol consumption where 85% of the women had complete data on these variables. Missing values in the monthly recording were replaced with the answer in the nearest cycle with non-missing response. If the persons had missing information for all cycles the missing response were replaced by 0 for smoking and alcohol consumption. Furthermore, the women recorded sexual intercourse and vaginal bleeding in a structured diary and collected first-morning urine samples from Day 1 to Day 10 in each cycle, until a pregnancy was achieved or whatever came first. The urine specimens were stored in the women's home freezers until they were picked up every third month. The women contacted the research team when a pregnancy was diagnosed by their general practitioner, and data on TTP was thereby recorded (Bonde et al., 1998).
Study design and urine sample collection
In 2009, stored urine samples from 242 women were available from the total of 430 women who were enrolled in the Danish First Pregnancy Planner Study. Out of the 430 women, 77 of the women had no urine samples collected because conception occurred during the first cycle, and additionally urine samples from 111 women were lost during storage for unknown reasons. Of the 242 women who had urine samples available, 13 women had variables on TTP missing for unknown reasons. Thus, 229 women were included in this cohort study.
The study subjects had up to two measurements of phthalate metabolites available. The women, who became pregnant, had the first measurement of phthalate metabolites from Day 10 in the cycle before a pregnancy was conceived for the majority of these women. The second measurement was taken Day 10 in the cycle, where a pregnancy was conceived. Similarly, the women, who did not achieve a pregnancy, had the first measurement of phthalate metabolites from Day 10 in the fifth cycle of the follow-up period. The second measurement was taken from a urine sample Day 10 in the sixth cycle for the majority of these women (Toft et al., 2012). Since phthalate measurements have high variability and the potential causal time window is unknown, analysis based on the mean of the first and second measurement was performed as a proxy for the level of phthalate exposure during the period of trying to become pregnant. In a secondary analysis, we analyzed the association to the first measurement alone to evaluate if potential bias due to pregnancy changed behavior of the women during the follow-up period affected the results.
Analytic method
In 2009, urine samples were analyzed for the phthalate metabolites; MEP of DEP, monobutyl phthalate (MBP) mainly of DBP and benzyl butyl phthalate (BBzP), monobenzyl phthalate (MBzP) of BBzP, MEHP of DEHP, and two secondary metabolites of DEHP; 5-hydroxy-MEHP and 5-oxo-MEHP. Standards for other phthalate metabolites were not commercially available at that time. The four primary phthalate metabolites: MEP, MBP, MBzP and MEHP were selected for the primary analysis in the current study. The two secondary DEHP metabolites; 5-hydroxy-MEHP and 5-oxo-MEHP were analyzed in a secondary analysis and not included in the main results presented. The urine samples were processed using an automated solid-phase extraction technique and analyzed by liquid chromatography tandem–mass spectrometry (LC–MS-MS). After adding internal standards, the samples were treated with glucuronidase to remove glucuronic acid and acidified, and the metabolites were extracted using Oasis HLB (Waters, Milford, MA, USA) 3 ml (60 mg) on an Aspec XL4 automated solid-phase extraction equipment (Gilson, Middleton, WI, USA). The samples were then evaporated and dissolved in water containing acetic acid. 2H4-MEP, 2H4-MBP, 2H44-MEHP and 13C44-MBzP were used as internal standards. An internal quality control urine sample was analyzed in each series. No temporal variation was measured in the concentrations of the phthalate metabolites in the control samples. The coefficient of variation of concentrations measured in the quality control samples varied between 7 and 11% and is shown in Toft et al., 2012, Supplementary data, Table S1 together with detection limits and MS parameters (http://ehp.niehs.nih.gov/wp-content/uploads/120/3/ehp.1103552.s001.pdf). The values for the phthalate metabolite MBP were within the tolerance limits obtained in urine samples from a round-robin interlaboratory control program from the University of Erlangen-Nürnberg, Germany. Moreover, the laboratory at Lund University that performed the analyses was acknowledged as a reference laboratory in the Democophes study (http://www.eu-hbm.info/democophes) for all the analyzed phthalate metabolites (Schindler et al., 2014). We used a hand refractometer to determine the urine density. The corrected concentration was calculated based on the following formula: Ccorr = Cobs × 0.016/(ρ − 1), where (shall be written instead) Ccorr = the corrected concentration, Cobs = the observed concentration and ρ = the specific density of the urine samples (Boeniger et al., 1993). In seven samples from the first measurement, MEHP levels were below the limit of detection (LOD; 2 ng/ml), and in all the remaining urine samples, all measured phthalate metabolites were detected. No women had samples below LOD on both first and second measurement. Finally, actual readout from LC–MS-MS was used for values below the LOD (Toft et al., 2012). All recordings were done blinded to the outcomes.
Data analysis and statistics
Descriptive statistics according to low and high levels of urinary concentration of MEP were calculated for demographic variables and median for urine phthalate metabolite concentrations (ng/ml) as the mean of the first and second measurement. Spearman correlation coefficient was used to determine correlations between individual phthalate metabolites in the first and second measurements and also between MEP and the other phthalate metabolites MBP, MBzP and MEHP based on the mean of the first measurement and second measurement.
Statistical analyses of the associations between each of the measured phthalate metabolites and TTP were performed based on the mean of the first and second phthalate measurement, and in secondary analyses with the first measurement alone as originally planned. The associations between TTP and phthalate metabolites were modeled on a continuous natural logarithm transformed scale. We estimated fecundability ratios (FRs) and 95% CI for TTP using a discrete-time Cox regression model which allows for cycle-varying intercept. The FRs estimate the probability of becoming pregnant each cycle, given exposure to the specific phthalate, conditional not being pregnant in the previous cycle. Thus, FR <1 denotes a reduction in fecundity or longer TTP and FR >1 denotes a shorter TTP. The women were censored in analyses upon withdrawal from the study or upon six cycles of pregnancy attempts. The FR analyses were adjusted for a priori established confounders, which were the variables; age (years), BMI (kg/m2) and the two time-specific variables smoking (cigarettes/day) and alcohol (units/week) (Bolumar et al., 1996; Augood et al., 1998; Dunson et al., 2002; Ramlau-Hansen et al., 2007; Wise et al., 2010). We evaluated the interactions between the exposure variables and the potential confounders, and no statistically significant effect modification was found on the multiplicative scale. The testable modeling assumptions were also evaluated, and the proportional hazard assumption and the linearity of covariates were verified. Statistical analyses were performed using STATA statistical version 12.0 (StataCorp, College Station, TX, USA).
Ethical approval
This study and the use of the additional urinary analysis of phthalates were approved by the ethics committee of the Central Region, Denmark. All participants gave written informed consent at enrollment in the study.
Results
Table I shows the demographic characteristics of the study subjects according to women with low and high level of urinary concentrations of MEP. The distributions of these covariates were generally unrelated to the exposure. Women with a high level of urinary concentrations of MEP had a higher proportion of current smokers than the women who were low exposed. Similar findings for the distributions of the covariates were found for the other phthalate metabolites MBP, MBzP and MEHP (data not shown). During the follow-up period, 2 (1%) women became pregnant after one cycle, 54 (24%) after three cycles, 101 (44%) after five cycles and 115 (50%) after six cycles, respectively.
Demographic characteristics at baseline of the 229 study subjects according to below and above median level of urinary concentrations of MEP.
| Demographic characteristics . | MEP (low) (n = 114) . | MEP (high) (n = 115) . |
|---|---|---|
| Mean (min–max) | ||
| Age, years | 26.0 (19; 35) | 25.2 (19; 32) |
| BMI, kg/m2 | 22.8 (15.2; 35.1) | 22.6 (17.0; 37.6) |
| Alcohol (drinks/week) | 4.0 (0; 16) | 3.9 (0; 25) |
| Menstrual cycle (days) | 31.6 (21; 62) | 31.4 (24; 68) |
| Coffee intake (cups/day) | 2.0 (0; 10) | 2.1 (0; 10) |
| N (%) | ||
| Smoking | ||
| Never | 58 (51) | 54 (47) |
| Now | 23 (20) | 43 (37) |
| Former | 33 (29) | 18 (16) |
| Education | ||
| 9–10 grade | 3 (3) | 6 (5) |
| High school | 2 (2) | 7 (6) |
| Vocational | 57 (50) | 60 (52) |
| Further education | 52 (45) | 42 (37) |
| Demographic characteristics . | MEP (low) (n = 114) . | MEP (high) (n = 115) . |
|---|---|---|
| Mean (min–max) | ||
| Age, years | 26.0 (19; 35) | 25.2 (19; 32) |
| BMI, kg/m2 | 22.8 (15.2; 35.1) | 22.6 (17.0; 37.6) |
| Alcohol (drinks/week) | 4.0 (0; 16) | 3.9 (0; 25) |
| Menstrual cycle (days) | 31.6 (21; 62) | 31.4 (24; 68) |
| Coffee intake (cups/day) | 2.0 (0; 10) | 2.1 (0; 10) |
| N (%) | ||
| Smoking | ||
| Never | 58 (51) | 54 (47) |
| Now | 23 (20) | 43 (37) |
| Former | 33 (29) | 18 (16) |
| Education | ||
| 9–10 grade | 3 (3) | 6 (5) |
| High school | 2 (2) | 7 (6) |
| Vocational | 57 (50) | 60 (52) |
| Further education | 52 (45) | 42 (37) |
MEP, monoethyl phthalate.
Demographic characteristics at baseline of the 229 study subjects according to below and above median level of urinary concentrations of MEP.
| Demographic characteristics . | MEP (low) (n = 114) . | MEP (high) (n = 115) . |
|---|---|---|
| Mean (min–max) | ||
| Age, years | 26.0 (19; 35) | 25.2 (19; 32) |
| BMI, kg/m2 | 22.8 (15.2; 35.1) | 22.6 (17.0; 37.6) |
| Alcohol (drinks/week) | 4.0 (0; 16) | 3.9 (0; 25) |
| Menstrual cycle (days) | 31.6 (21; 62) | 31.4 (24; 68) |
| Coffee intake (cups/day) | 2.0 (0; 10) | 2.1 (0; 10) |
| N (%) | ||
| Smoking | ||
| Never | 58 (51) | 54 (47) |
| Now | 23 (20) | 43 (37) |
| Former | 33 (29) | 18 (16) |
| Education | ||
| 9–10 grade | 3 (3) | 6 (5) |
| High school | 2 (2) | 7 (6) |
| Vocational | 57 (50) | 60 (52) |
| Further education | 52 (45) | 42 (37) |
| Demographic characteristics . | MEP (low) (n = 114) . | MEP (high) (n = 115) . |
|---|---|---|
| Mean (min–max) | ||
| Age, years | 26.0 (19; 35) | 25.2 (19; 32) |
| BMI, kg/m2 | 22.8 (15.2; 35.1) | 22.6 (17.0; 37.6) |
| Alcohol (drinks/week) | 4.0 (0; 16) | 3.9 (0; 25) |
| Menstrual cycle (days) | 31.6 (21; 62) | 31.4 (24; 68) |
| Coffee intake (cups/day) | 2.0 (0; 10) | 2.1 (0; 10) |
| N (%) | ||
| Smoking | ||
| Never | 58 (51) | 54 (47) |
| Now | 23 (20) | 43 (37) |
| Former | 33 (29) | 18 (16) |
| Education | ||
| 9–10 grade | 3 (3) | 6 (5) |
| High school | 2 (2) | 7 (6) |
| Vocational | 57 (50) | 60 (52) |
| Further education | 52 (45) | 42 (37) |
MEP, monoethyl phthalate.
Exposure distributions for the unadjusted and density adjusted urinary concentrations for the phthalate metabolites by pregnancy status are shown in Table II. We found that density adjusted MEP concentration ranged from 21.5 to 7044 ng/ml, and the highest median concentration (258 ng/ml) was found for women, who did not become pregnant. MBP ranged from 32.2 to 1392 (ng/ml) and was also considerably higher than values of MBzP and MEHP. For MBP, MBzP and MEHP, there were only small differences between those who achieved a pregnancy and those who did not.
Achieved pregnancy by phthalate metabolite concentrations based on the mean of the first and second measurement.
| Characteristics . | Pregnancy (n = 115) . | No pregnancy (n = 114) . | Overall (n = 229) . |
|---|---|---|---|
| Phthalate metabolites (ng/ml) | |||
| Median (min–max) | |||
| MEP | 201 (21.5; 1920) | 258 (30.6; 7044) | 225 (21.5; 7044) |
| MEP unadjusted | 343 (15.5; 2907) | 446 (16.0; 12357) | 379 (15.5; 12357) |
| MBP | 178 (35.0; 924) | 176 (32.3; 1392) | 178 (32.3; 1392) |
| MBP unadjusted | 262 (10.0; 1918) | 286 (11.0; 2892) | 272 (10.0; 2892) |
| MBzP | 13.7 (2.7; 103) | 15.3 (2.5; 77.4) | 14.9 (2.5; 103) |
| MBzP unadjusted | 21.0 (1.0; 203) | 23.0 (1.0; 193) | 22.25 (1.0; 203) |
| MEHP | 11.6 (0.6; 65.7) | 10.6 (1.0; 71.1) | 11.2 (0.6; 71.1) |
| MEHP unadjusted | 13.0 (0; 348) | 16.5 (0; 316) | 14.5 (0; 348) |
| Characteristics . | Pregnancy (n = 115) . | No pregnancy (n = 114) . | Overall (n = 229) . |
|---|---|---|---|
| Phthalate metabolites (ng/ml) | |||
| Median (min–max) | |||
| MEP | 201 (21.5; 1920) | 258 (30.6; 7044) | 225 (21.5; 7044) |
| MEP unadjusted | 343 (15.5; 2907) | 446 (16.0; 12357) | 379 (15.5; 12357) |
| MBP | 178 (35.0; 924) | 176 (32.3; 1392) | 178 (32.3; 1392) |
| MBP unadjusted | 262 (10.0; 1918) | 286 (11.0; 2892) | 272 (10.0; 2892) |
| MBzP | 13.7 (2.7; 103) | 15.3 (2.5; 77.4) | 14.9 (2.5; 103) |
| MBzP unadjusted | 21.0 (1.0; 203) | 23.0 (1.0; 193) | 22.25 (1.0; 203) |
| MEHP | 11.6 (0.6; 65.7) | 10.6 (1.0; 71.1) | 11.2 (0.6; 71.1) |
| MEHP unadjusted | 13.0 (0; 348) | 16.5 (0; 316) | 14.5 (0; 348) |
MEP, monoethyl phthalate; MBP, monobutyl phthalate; MBzP, monobenzyl phthalate; MEHP, monoethylhexyl phthalate.
Achieved pregnancy by phthalate metabolite concentrations based on the mean of the first and second measurement.
| Characteristics . | Pregnancy (n = 115) . | No pregnancy (n = 114) . | Overall (n = 229) . |
|---|---|---|---|
| Phthalate metabolites (ng/ml) | |||
| Median (min–max) | |||
| MEP | 201 (21.5; 1920) | 258 (30.6; 7044) | 225 (21.5; 7044) |
| MEP unadjusted | 343 (15.5; 2907) | 446 (16.0; 12357) | 379 (15.5; 12357) |
| MBP | 178 (35.0; 924) | 176 (32.3; 1392) | 178 (32.3; 1392) |
| MBP unadjusted | 262 (10.0; 1918) | 286 (11.0; 2892) | 272 (10.0; 2892) |
| MBzP | 13.7 (2.7; 103) | 15.3 (2.5; 77.4) | 14.9 (2.5; 103) |
| MBzP unadjusted | 21.0 (1.0; 203) | 23.0 (1.0; 193) | 22.25 (1.0; 203) |
| MEHP | 11.6 (0.6; 65.7) | 10.6 (1.0; 71.1) | 11.2 (0.6; 71.1) |
| MEHP unadjusted | 13.0 (0; 348) | 16.5 (0; 316) | 14.5 (0; 348) |
| Characteristics . | Pregnancy (n = 115) . | No pregnancy (n = 114) . | Overall (n = 229) . |
|---|---|---|---|
| Phthalate metabolites (ng/ml) | |||
| Median (min–max) | |||
| MEP | 201 (21.5; 1920) | 258 (30.6; 7044) | 225 (21.5; 7044) |
| MEP unadjusted | 343 (15.5; 2907) | 446 (16.0; 12357) | 379 (15.5; 12357) |
| MBP | 178 (35.0; 924) | 176 (32.3; 1392) | 178 (32.3; 1392) |
| MBP unadjusted | 262 (10.0; 1918) | 286 (11.0; 2892) | 272 (10.0; 2892) |
| MBzP | 13.7 (2.7; 103) | 15.3 (2.5; 77.4) | 14.9 (2.5; 103) |
| MBzP unadjusted | 21.0 (1.0; 203) | 23.0 (1.0; 193) | 22.25 (1.0; 203) |
| MEHP | 11.6 (0.6; 65.7) | 10.6 (1.0; 71.1) | 11.2 (0.6; 71.1) |
| MEHP unadjusted | 13.0 (0; 348) | 16.5 (0; 316) | 14.5 (0; 348) |
MEP, monoethyl phthalate; MBP, monobutyl phthalate; MBzP, monobenzyl phthalate; MEHP, monoethylhexyl phthalate.
As shown in Table III, MEP was associated with reduced FR (adjusted FR 0.79; 95% CI: 0.63; 0.99) indicating longer TTP at higher MEP concentration with a 21% reduction in the probability of conception during a cycle for each ln-unit increase in MEP. There was no significant association with TTP found for the other phthalate metabolites; MBP, MBzP and MEHP either in crude or adjusted models. Also, no significant association was found for the two secondary DEHP-metabolites 5-hydroxy-MEHP (adjusted FR 0.70; 95% CI: 0.48; 1.03) and 5-oxo-MEHP (adjusted FR 0.83; 95% CI: 0.58; 1.18). Crude FRs were generally consistent with adjusted FRs. In comparison, we also found in secondary analyses that MEP was associated with longer TTP (adjusted FR 0.82; 95% CI: 0.63; 0.99) and no significant association for MBP, MBZP and MEHP (data not shown) when the association was analyzed to the first measurement alone. In another secondary analysis, we excluded the women with the highest urinary concentration of MEP (upper 5%). The effect estimate for the lower 95% exposed was similar to what was found in the full cohort; however, adjusted FR 0.83; 95% CI: 0.64; 1.07 was no longer statistical significant.
FRs and 95% CI according to urine concentration of ln-transformed phthalate metabolite concentrations based on the mean of the first and second phthalate measurement among 229 Danish first pregnancy planners.
| . | Crude . | Adjusteda . |
|---|---|---|
| . | FR (95% CI) . | FR (95% CI) . |
| MEP | 0.83 (0.67; 1.04) | 0.79 (0.63; 0.99) |
| MBP | 1.04 (0.77; 1.41) | 1.03 (0.76; 1.39) |
| MBzP | 0.89 (0.67; 1.20) | 0.88 (0.64; 1.19) |
| MEHP | 0.99 (0.74; 1.32) | 0.99 (0.72; 1.35) |
| . | Crude . | Adjusteda . |
|---|---|---|
| . | FR (95% CI) . | FR (95% CI) . |
| MEP | 0.83 (0.67; 1.04) | 0.79 (0.63; 0.99) |
| MBP | 1.04 (0.77; 1.41) | 1.03 (0.76; 1.39) |
| MBzP | 0.89 (0.67; 1.20) | 0.88 (0.64; 1.19) |
| MEHP | 0.99 (0.74; 1.32) | 0.99 (0.72; 1.35) |
aModel adjusts for age, BMI, and the time-varying variables alcohol and smoking. MEP, monoethyl phthalate; MBP, monobutyl phthalate; MBzP, monobenzyl phthalate; MEHP, monoethylhexyl phthalate; FR, Fecundability ratio.
FRs and 95% CI according to urine concentration of ln-transformed phthalate metabolite concentrations based on the mean of the first and second phthalate measurement among 229 Danish first pregnancy planners.
| . | Crude . | Adjusteda . |
|---|---|---|
| . | FR (95% CI) . | FR (95% CI) . |
| MEP | 0.83 (0.67; 1.04) | 0.79 (0.63; 0.99) |
| MBP | 1.04 (0.77; 1.41) | 1.03 (0.76; 1.39) |
| MBzP | 0.89 (0.67; 1.20) | 0.88 (0.64; 1.19) |
| MEHP | 0.99 (0.74; 1.32) | 0.99 (0.72; 1.35) |
| . | Crude . | Adjusteda . |
|---|---|---|
| . | FR (95% CI) . | FR (95% CI) . |
| MEP | 0.83 (0.67; 1.04) | 0.79 (0.63; 0.99) |
| MBP | 1.04 (0.77; 1.41) | 1.03 (0.76; 1.39) |
| MBzP | 0.89 (0.67; 1.20) | 0.88 (0.64; 1.19) |
| MEHP | 0.99 (0.74; 1.32) | 0.99 (0.72; 1.35) |
aModel adjusts for age, BMI, and the time-varying variables alcohol and smoking. MEP, monoethyl phthalate; MBP, monobutyl phthalate; MBzP, monobenzyl phthalate; MEHP, monoethylhexyl phthalate; FR, Fecundability ratio.
Moreover, correlation analysis showed moderate to strong correlation for MEP in the first and the second measurement with a correlation coefficient of 0.51. The other individual phthalate metabolites MBP, MBzP and MEHP had low to moderate correlations (r = 0.22, 0.35 and 0.10). There were low correlations between MEP and the other phthalate metabolites MBP, MBzP and MEHP based on the mean of the first and second measurement with correlation coefficients of 0.10–0.14.
Discussion
In this cohort study based on prospective data from 229 women in the Danish First Pregnancy Planner Study, we estimated the association between female exposure to phthalates and TTP. The findings indicate that exposure to the phthalate metabolite MEP at the level studied may be associated with prolonged TTP. The other phthalate metabolites MBP, MBzP and MEHP were not associated with TTP.
To our knowledge, this is the first study which has found an association between urinary concentrations of MEP and TTP. In the LIFE Study (Buck Louis et al., 2014), the MIREC Study (Velez et al., 2015) and the North Carolina Early Pregnancy Study (Jukic et al., 2016), the phthalate metabolite MEP was not associated with TTP. Similar to our results, these studies also found no association for MBzP and MEHP, and the LIFE study had negative findings for MBP and TTP. However, in the Life Study urinary concentrations of MBP and MBzP in men were associated with a reduction in couple fecundity. Although, no association was found between MEP and TTP in other epidemiological studies, findings based on animal studies could suggest that MEP may induce reproductive toxicity. In fathead minnows, the parent compound DEP exposure was associated with increased embryo toxicity (Mankidy et al., 2013), and rodent studies indicated reproductive toxicity at a high level of DEP exposure (Kay et al., 2013).
Our data from the Danish First Pregnancy Planner Study was collected >20 years ago at a time where levels of phthalate exposure often were higher than they are in contemporary populations and this could be one potential explanation for the heterogeneous findings compared to previous studies. The median level of urinary concentration of MEP in the current study was 379 ng/ml which is higher but comparable with the findings from two US studies from the same time period with 305 ng/ml in NHANES (1988–1994) (Blount et al., 2000) and 211 ng/ml (Hoppin et al., 2002), respectively. In newer studies, the median levels were 90 ng/ml in a German Study (Koch et al., 2003) and 68 ng/ml in the US NHANES (2009–2010) (Jukic et al., 2016). We still find women with high urinary MEP concentrations in present time. In the German study (Koch et al., 2003), in the 95th percentile the value was 560 ng/ml and in the US EARTH Study (Hauser et al., 2016) maximum value ranged up to above 3000 ng/ml. As a consequence of a potential accumulated effect by use of different sources of MEP, these high values could occur. It was found that women who reported the combined use of four or more products of cosmetics and personal care products showed up to four times higher average MEP concentrations than women who used fewer than four products (Berman et al., 2009). The adverse effect on TTP in our study could be due to higher concentrations of MEP, and a high exposure contrast may be needed to detect an association. However, the detrimental effects of MEP on fecundity seem not to be driven by a few highly exposed individuals shown in the secondary analysis with the exclusion of the women with the highest urinary concentrations of MEP. Due to the large number of statistical comparisons, we cannot exclude that it might be a chance finding. We did not adjust for multiple comparisons due to increased risk for a Type 2 error.
We found a moderate to strong correlation for MEP between the first and the second measurement (r = 0.51), and thereby less variation in the phthalate exposure over time compared to the weak to moderate correlation found for MBP, MBzP and MEHP (r = 0.22, 0.35 and 0.10). MEP is thus less sensitive to temporal fluctuations than the other measured compounds, and thereby the risk of misclassification of exposure is lower for MEP compared to the other measured compounds. Therefore, we cannot exclude that an association to the other phthalate metabolites are missed due to bias towards the null. However, in spite of substantial within-day and between-day variation of urinary phthalate metabolites, average levels seem rather stable across a 2-month period of repeated sampling (Preau et al., 2010). Since the specific time window for the potential adverse effect of phthalates on female fecundity is not known in detail, we chose a mean of the exposure in two available samples before conception as a proxy for the exposure in the period of planning a pregnancy. The secondary analysis based on the first available sample corroborated the overall results indicating that bias due to change in behavior during the period of planning a pregnancy may not be a major concern.
We had exposure measurements from data on urinary analysis of phthalates made in 2009 with up to two measurements of phthalate metabolites available per person. The measurements were taken at different time points during the follow-up period for the women who achieved a pregnancy and for those women, who did not. However, the women were unaware of the phthalate concentrations in the follow-up period, and therefore, we only expect non-differential misclassification of the exposure. The exclusion of the 77 of the 430 women, who became pregnant in the first cycle after enrollment in the cohort and therefore were more fecund on average, limit the generalizability of the study. These women however, did not differ from the other women in the cohort according to the baseline characteristics showed in Table I (see Supplementary data, Table S1). These high fertile women, who became pregnant in the first menstrual cycle, had no urinary samples collected before a pregnancy occurred. If these women were less sensitive to phthalate exposure than the women, who participated in the current study, there is a risk of overestimating the association for the population at large. The subfertile women in our study were therefore relatively overrepresented which was linked to the low percentage (50%) of women who achieved a pregnancy during the six cycles of follow-up. Also, in 111 of the 430 women in the cohort, urine samples were lost during storage for unknown reasons, and they were most likely lost by chance, why selection bias is unlikely. TTP was collected in real time and based on clinical pregnancies determined by the general practitioner, and although some misclassification on the actual starting date of the beginning of family planning and the date of conception occur, we believe this to be independent of the exposure. The women in the current study were first pregnancy planners and had no prior knowledge about their fecundity. They were included when they stopped using contraception and therefore they had no knowledge about their potential fertility problems. Also, they had no knowledge about their phthalate exposure and did not try to avoid these in their environment because the potentially harmful health effects of phthalates were not of public interest in the early 1990s, and therefore we found that selections bias was unlikely. It is an inherent limitation that the women who have unplanned pregnancies were not eligible, and we cannot rule out that it could cause bias, if these women had higher or lower phthalate levels, and at the same time had longer or shorter TTP, than the women included in the study.
The potential confounders (age, BMI, smoking and alcohol consumption) in our analysis were self-reported and were recorded with some error. The correlations between MEP and the other phthalate metabolites MBP, MBzP and MEHP (r = 0.10–0.14) indicate that MEP is only weakly correlated to the other phthalate metabolites and therefore were not considered as potential confounding factors. In a secondary analysis, we included education as a potential confounder, and that did not change the estimates substantially (data not shown). We cannot exclude that unmeasured confounders may affect the estimate, but we are not aware of major potential confounders that should be included in the analyses.
Our results indicate that maternal MEP may be associated with TTP, but we cannot rule out that couple fecundity may be affected by male phthalate exposure as well. If maternal and paternal MEP is highly correlated, the associations observed may be due to paternal exposures. However, when the correlation between couples paired urine measurements of phthalate metabolites was investigated on the same day, only weak to moderate correlation of measurements of MEP (r = 0.27) and MEHP (r = 0.42) was found in a study with 70 couples (Meeker et al., 2012). In addition, the LIFE Study found that correlation between partners urinary concentrations of phthalate metabolites among other MEP, MBP, MBzP and MEHP was low (r = 0–0.3) (Buck Louis et al., 2014). These findings indicate that couples only to a lesser degree were similarly exposed to phthalates possibly because to a certain extent have similar exposure patterns. Since cosmetics are a major source of DEP and thereby MEP, the low correlation of maternal and paternal exposures for this compound was expected. Therefore, we find it unlikely that the observed association between MEP and TTP was caused by a male subfecundity factor.
In conclusion, our findings suggest that increased MEP exposure is associated with a decreased probability of achieving a pregnancy within the first six cycles of trying. The results of our study need to be replicated in a larger and newer cohort study with sufficient exposure contrast to fully evaluate the potential impact of phthalates on human fecundity. None of the other measured phthalate metabolites; MBP, MBzP and MEHP were associated with TTP at detectable levels.
Supplementary data
Supplementary data are available at http://humrep.oxfordjournals.org/.
Acknowledgement
None.
Authors’ roles
A.M.L.T., G.T. and J.O. conceived and designed the study, performed data analysis and data interpretation. T.K.J., N.H.H., J.P.B., B.A.G.J. and C.H.L. collected data for the project. A.H.R. performed expert statistical support. A.M.L.T. drafted the manuscript. All authors revised the manuscript critically for important intellectual content and approved the final manuscript.
Funding
The original data for the study were collected as part of the Danish First Pregnancy Planner Study supported by the Aarhus University Research Foundation, the Danish Medical Research Council and the Danish Medical Health Insurance Foundation. Phthalate and data analyses were funded by the Danish Ministry of Environment and the Lundbeck Foundation, AFA (Swedish Labor Market Insurance Company) and the Swedish Research Council.
Conflict of interest
None declared.
