Skip to main content
NCBI home page
HHS Author Manuscripts logoLink to HHS Author Manuscripts
. Author manuscript; available in PMC: 2018 Jul 1.
Published in final edited form as: Environ Res. 2017 Apr 12;156:420–425. doi: 10.1016/j.envres.2017.04.004

In utero exposure to atrazine analytes and early menarche in the Avon Longitudinal Study of Parents and Children Cohort

Gonza Namulanda a,b,*, Ethel Taylor a, Mildred Maisonet c, Dana Boyd Barr d, W Dana Flanders a,d, David Olson a, Judith R Qualters a, John Vena e, Kate Northstone f, Luke Naeher a,b
PMCID: PMC5679269  NIHMSID: NIHMS916028  PMID: 28410519

Abstract

Background

Evidence from experimental studies suggests that atrazine and its analytes alter the timing of puberty in laboratory animals. Such associations have not been investigated in humans.

Objective

To determine the association between in utero exposure to atrazine analytes and earlier menarche attainment in a nested case-control study of the population-based Avon Longitudinal Study of Parents and Children.

Methods

Cases were girls who reported menarche before 11.5 years while controls were girls who reported menarche at or after 11.5 years. Seven atrazine analyte concentrations were measured in maternal gestational urine samples (sample gestation week median (IQR): 12 (8–17)) during the period 1991–1992, for 174 cases and 195 controls using high performance liquid chromatography-tandem mass spectrometry. We evaluated the study association using multivariate logistic regression, adjusting for potential confounders. We used multiple imputation to impute missing confounder data for 29% of the study participants.

Results

Diaminochlorotriazine (DACT) was the most frequently detected analyte (58% > limit of detection [LOD]) followed by desethyl atrazine (6%), desethyl atrazine mercapturate (3%), atrazine mercapturate (1%), hydroxyl atrazine (1%), atrazine (1%) and desisopropyl atrazine (0.5%). Because of low detection of other analytes, only DACT was included in the exposure–outcome analyses. The adjusted odds of early menarche for girls with DACT exposures≥median was 1.13 (95% Confidence Interval [95% CI]:0.82, 1.55) and exposure < median was 1.01 (95% CI: 0.73, 1.42) compared to girls with exposure < LOD (reference). In the subset that excluded girls with missing data, the adjusted odds of early menarche for girls with DACT exposures≥median was 1.86 (95% CI: 1.03, 3.38) and exposure < median was 1.26 (95% CI: 0.65, 2.24) compared to the reference.

Conclusions

This study is the first to examine the association between timing of menarche and atrazine analytes. We found a weak, non-significant association between in-utero exposure to atrazine metabolite DACT and early menarche, though the association was significant in the subset of girls with complete confounder information. Further exploration of the role of these exposures in female reproduction in other cohorts is needed.

Keywords: ALSPAC, Endocrine disrupting compounds, Puberty, Menarche, Atrazine

1. Introduction

Atrazine is used in more than 70 countries worldwide; most frequently in the United States (U.S.), Brazil, Argentina, Mexico and China to control broadleaf and grassy weeds in agricultural crops, mainly corn (LeBaron et al., 2008; EPA, 2012). Human exposure to atrazine can occur through several routes. Agricultural workers and those living near farms can be exposed to atrazine dispersed in air through spraying (ATSDR, 2003). Exposure can also occur through contact with contaminated soils or ingestion of contaminated agricultural products (ATSDR, 2003). Water run-off from crops and lawn applications can get into ground and surface waters and contaminate drinking water wells (Munger et al., 1997). In water, atrazine breaks down to the following primary metabolites: desethyl atrazine (DEA), desisopropyl atrazine (DIA) and diaminochlorotriazine (DACT) (EPA, 2012).

In areas where atrazine is used, the general population can be exposed to atrazine analytes (i.e., atrazine parent compound or any of its metabolites) through contaminated drinking water (Munger et al., 1997; Ochoa-Acuna et al., 2009; Villanueva et al., 2005). Developing fetuses and children can be exposed in utero or through breast milk (Balduini et al., 2003; Whyatt et al., 2003). Atrazine is broken down rapidly in the body and eliminated primarily in urine, within 24–48 h (Catenacci et al., 1993). Although it does not bioaccumulate appreciably in humans, atrazine and its related compounds can persist in ground water, contaminating surface and drinking water sources (ATSDR, 2003). This persistent contamination of drinking water can result in continuous exposure in humans, and therefore can be of concern (UNEP and WHO, 2012).

In Great Britain, atrazine has mainly been used on maize and sweet corn crops (Fera Science Ltd, 2016). During the years 1990–1992, about 42,000 kg of atrazine was applied annually in Great Britain, with about 17,000 kg applied to about 10,000 ha annually in the South Western region of England (Fera Science Ltd, 2016). The European Union (E.U.) banned the use of atrazine in 2003 because of its persistent contamination of drinking and ground water above the E.U. recommended limit of 0.1 parts per billion (ppb) (Sass and Colangelo, 2006). However, several countries still use this herbicide including the US where it is a restricted use pesticide (ATSDR, 2003).

Animal studies have associated exposure to atrazine with hormone-related tumors and with alterations in reproductive functions suggesting potential endocrine disrupting effects from these exposures. In rats, postnatal exposure to atrazine induced an earlier onset (Cooper et al., 2007; Wetzel et al., 1994) and increased incidence of mammary gland tumors by disrupting the ovarian function (Cooper et al., 2007). Atrazine postnatal exposure also caused a disruption in estrus cycle in rats by impairing the ovulatory surge of luteinizing hormone (Wetzel et al., 1994). Lactational exposure to atrazine or DACT (Laws et al., 2000, 2003) and prenatal exposure to atrazine (Davis et al., 2011) in female rats resulted in delayed puberty (vaginal opening) in the offspring. Further, DACT was demonstrated to be as potent as atrazine in delaying puberty (Laws et al., 2003). However, the doses associated with delayed puberty in these studies were high (50,000–200,000 ppb daily).

Several studies have reported a secular trend towards earlier onset of puberty among girls in Europe and the United States (Aksglaede et al., 2008; Euling et al., 2008; McDowell et al., 2007; Parent et al., 2003; Semiz et al., 2008) with some reporting earlier age at onset of menarche (McDowell et al., 2007; Semiz et al., 2008). Early puberty is a risk factor for childhood risky behaviors (e.g., smoking, alcohol consumption and drug abuse) and adult-onset diseases (e.g., breast and ovarian cancer) (Gail et al., 1989; Moorman et al., 2009; vanJaarsveld et al., 2007; Walvoord, 2010). Epidemiologic studies of other endocrine disrupting chemicals (EDCs) suggest that in utero exposure to the estrogenic effect of EDCs is associated with early onset of puberty (Blanck et al., 2000; Vasiliu et al., 2004). Studies have also associated prepubertal exposure to EDCs with later onset of puberty (Windham et al., 2015; Wolff et al., 2015). Evidence from experimental studies suggests that atrazine analytes alter the timing of puberty in laboratory animals, however, such associations have not been investigated in humans. This study uses a nested case-control study design to measure atrazine analytes in maternal gestational urine as a proxy for in utero exposure of cases and controls to atrazine analytes, and examines the association with early menarche, as a marker of early puberty in girls.

2. Methods

2.1. Study population

Pregnant women living in the Bristol area, in the south west of England, United Kingdom, with an expected date of delivery from 1st April, 1991 to 31st December, 1992, were recruited to participate in the Avon Longitudinal Study of Parents and Children (ALSPAC) (Fraser et al., 2013). A total of 14,775 live births were included in the ALSPAC cohort (Boyd et al., 2013). Details about recruitment and the study participants have already been published (Fraser et al., 2013; Boyd et al., 2013). Beginning at the age of 8, a puberty questionnaire titled “Growing and Changing” was mailed to the participants each year until age 17 (except age 12) to collect information about the onset and progression of puberty in enrolled girls and boys (Rubin et al., 2009). The number of singleton girls available for follow-up was 5756 (Christensen et al., 2011), and only 3938 returned at least one valid and complete questionnaire (Christensen et al., 2010). The onset of menarche information in the questionnaire included the age and date of first menstrual period. The parents, the girls, or both completed the questionnaires (Rubin et al., 2009). The median age at onset of menarche for the girls who returned at least one puberty questionnaire (n=3938) was determined using parametric survival analysis, and reported as 12.87 years (95% confidence interval [95% CI]): 10.82, 12.91) (Christensen et al., 2010). The questionnaire is available on the study website at http://www.bristol.ac.uk/media-library/sites/alspac/migrated/documents/ques-cb16a-mum-and-daughter-at-8.pdf. The website also contains detailed descriptions of available data accessible through a fully searchable online data dictionary (http://www.bris.ac.uk/alspac/researchers/data-access/data-dictionary/).

The selection of cases and controls for the nested case-control study has been previously described (Christensen et al., 2011). In brief, only the 3682 singleton girls who returned at least 2 valid and complete “Growing and Changing” questionnaires from ages 8–13 years were considered for the study. From this number 218 were identified as having menarche before 11.5 years of age, and had one maternal gestational serum sample available for analysis. A subset of 174 cases also had one maternal gestational urine sample available for analysis. A random sample of 394 girls that had menarche at or after 11.5 years of age were selected as controls; 230 of these had one maternal gestational serum sample and 195 also had one maternal gestational urine sample available for analysis. We included only the 174 cases and 195 controls with analyzable maternal gestational urine samples in this study. The urine samples were collected from mothers as part of routine antenatal care at random times during pregnancy (between 1st April 1991 and 31st December 1992), (Fraser et al., 2013; Boyd et al., 2013) therefore, stage of gestation and timing of urine collection may vary among mothers’ urine samples. Each urine sample was divided into small aliquots to maximize efficiency of the sample use, and banked at the University of Bristol (Boyd et al., 2013). The samples were stored at +4 or −20 °C 0–6 days after initial collection then stored between −10 and −20 °C until analysis. Only one aliquot per mother was used for this study. The ALSPAC Law and Ethics Committee, the Local Research Ethics Committees, and the Centers for Disease Control and Prevention (CDC) Institutional Review Board assessed and approved human subject protection (Christensen et al., 2011).

2.2. Exposure assessment

Banked urine samples from the University of Bristol were transferred to CDC’s Division of Laboratory Sciences within the National Center for Environmental Health in 2008, and analyzed during the same year. The following atrazine analytes were measured using on-line solid phase extraction-isotope dilution-high performance liquid chromatography-tandem mass spectrometry (Panuwet et al., 2008, 2010): Atrazine, DACT, DEA, DIA, atrazine mercapturate, desethyl atrazine mercapturate, and hydroxyl atrazine. The limits of detection (LODs) were calculated as three times the standard deviation of the noise at zero concentration (Panuwet et al., 2008, 2010). An estimate of the noise was based on the variation in precision at concentrations close to the LODs (Panuwet et al., 2008, 2010). Creatinine was measured for each urine sample on a Roche Hitachi Modular P Chemistry Analyzer (Hitachi, Pleasanton, CA, USA) using the Creatinine Plus Assay, as described in Roche’s Creatinine Plus Product Application # 04903773003.

We calculated the median, and the 25th and 75th percentile only for analytes detected in more than 50% of sample measurements. Survival analyses methods are recommended when estimating distributions using data with values < LOD instead of methods that replace values < LOD with LOD/2 or LOD/√2 (Gillespie et al., 2010). A reason is that survival analyses methods do not assume a distribution of values < LOD, and instead use all available information when estimating values < LOD (Gillespie et al., 2010). Therefore, to calculate the median and percentiles, we used the reverse Kaplan-Meier estimator, a non-parametric survival analysis estimator, using the LIFETEST procedure in SAS software version 9.3 (SAS Institute Inc., Cary, NC, USA) to account for the left censored values (< LOD) after transforming the data for left censoring.

2.3. Exposure and early menarche analyses

We classified exposure to analytes detected in more than 50% of sample measurements into the following 3 categories: < LOD, LOD to < median, and ≥median. We determined the median using the distribution of controls ≥LOD. We also dichotomized exposure as “Yes” (≥LOD) or “No” (< LOD).

We selected the following covariates a priori from the literature, and evaluated them for their potential to confound the outcome–exposure association: Mother’s age at delivery (< 20, 20–24, 25–29, 30–39, or ≥40 years) (Adair, 2001), mother’s pre-natal body mass index (BMI) (< 18.5 [underweight], 18.5–24.9 [normal], 25–29.9 [over-weight], or ≥30 [obese]) (Rubin et al., 2009), mother’s age at menarche (8–11, 12–14, or ≥15 years) (Rubin et al., 2009), mother’s education (Certificate of secondary education (CSE)/None, Vocational, O-level, A-level, or Degree) (Adair, 2001), child’s birth weight (< 2500 g or ≥2500 g) (Adair, 2001; Ruder et al., 2010), child’s birth order (first born, second born, or third born or later) (Rubin et al., 2009), child’s BMI at age 7 determined using actual measurements of the child’s height and weight (Rubin et al., 2009; Biro et al., 2010), child’s duration of breast feeding (never, < 6 months, or ≥6 months) (Balduini et al., 2003), trimester when gestational urine sample was collected (< 12 weeks, 13–≤28 weeks, or ≥29 weeks) and month of urine sample collection (December–February, March–May, June–August, or September–November), or (March–August or September–February) (Whyatt et al., 2003; Winchester et al., 2008).

We used logistic regression methods in SAS 9.3 to evaluate the covariate–exposure and the covariate–outcome association. We considered covariates with a p-value (p) < 0.25 (Mickey and Greenland, 1989) for both associations to be confounders.

Twenty nine percent of the study participants (52 cases and 53 controls) were missing data for one or more of the confounders identified above. To minimize bias due to systematic differences between cases and controls with complete data and those with missing data, we applied multiple imputation methods using the Fully Conditional Specification procedure in SAS 9.3 to impute values for missing covariate data (Liu and De, 2015). Five datasets were generated with imputed values for the following categorical covariates: mother’s age at delivery, mother’s prenatal BMI, mother’s age at menarche, mother’s education, child’s birth order, child’s duration of breast feeding, and month urine sample was collected, and the following continuous covariates: child’s BMI at age 7 and mother’s urine creatinine. We used these datasets with imputed covariate values to estimate the outcome–exposure association. We used logistic regression to analyze the association between early menarche and in utero exposure to atrazine analytes, and then used the MIANALYZE procedure to combine the results from imputed datasets. We adjusted for urine creatinine by including it as a covariate in the multivariate model.

3. Results

Child’s birth order, child’s BMI at age 7, mother’s prenatal BMI and mother’s age at menarche significantly (p < 0.05) differed between cases and controls (Table 1). Cases were more likely than controls to be first born children (61.5% vs. 52.1%), have a mother who was overweight (BMI 25–29.9) or obese (BMI > 30) (17.7% and 10.4%, respectively vs. 8.3% and 3.9% of controls), or have a mother with onset of menarche between the ages 8–11 years (32.9% vs. 13.4% of controls) (Table 1). Cases had significantly higher BMIs at age 7 compared to controls (p < 0.0001). Cases were also more likely to have a mother with either no education or a CSE (15.5% vs. 9.4% of controls) or to have been breastfed for less than 6 months (47.9% vs. 37.7% of controls), and less likely to have been breastfed for 6 months or more (32.2% vs. 40.9% of controls) although these differences were not statistically significant. The median age at menarche for cases was 11 years compared to 12.8 years for controls (Table 1).

Table 1.

Study population characteristics.

Characteristic Cases (N=174)
Controls (N=195)
p-valuea
N % N %
Mother’s ethnicity 0.88
 White 167 98.2 189 98.4
 Other Ethnicity 3 1.8 3 1.6
 Missing 4 3
Mother’s educationb 0.45
 CSE/None 26 15.5 18 9.4
 Vocational 12 7.1 14 7.3
 O-Level 56 33.3 61 31.9
 A-Level 48 28.6 64 33.5
 Degree 26 15.5 34 17.8
 Missing 6 14
Mother’s age at delivery (yrs) 0.36
 < 20 1 0.6 6 3.1
 20–24 32 18.5 28 14.4
 25–29 72 41.6 74 38
 30–39 65 37.6 84 43.1
 > 40 3 1.7 3 1.5
 Missing 1
Mother’s prenatal 0.004
BMI
 < 18.5 5 3.3 11 6.2
 18.5–24.9 109 69.0 148 81.8
 25–29.9 28 17.7 15 8.3
 > 30 16 10.4 7 3.9
 Missing 16 14
Mother’s age at menarche (yrs) 0.0003
 8–11 50 32.9 22 13.4
 12–14 92 60.5 126 76.8
 ≥15 10 6.6 16 9.8
 Missing 22 31
Child’s birth order 0.03
 First born 102 61.5 98 52.1
 Second born 36 21.7 65 34.6
 Third born or later 28 16.9 25 13.3
 Missing 8 7
Child’s birth weight 0.86
 < 2500 g 4 2.3 5 2.6
 ≥2500 g 168 97.7 187 97.4
 Missing 2 3
Child’s duration of breast feeding 0.16
 Never 28 16.9 39 21.3
 < 6 months 79 47.9 69 37.7
 ≥6 months 58 32.2 75 40.9
 Missing 9 12
Median Quartile Median Quartile < 0.0001
Child’s BMI at age 7 17 15.7–18.9 15.9 15.2–16.8
Child’s age at menarche (yrs) 11 10.7–11.3 12.8 12.3–13.3
Urine Sample gestation week 12 8–17 13 9–18 0.23
Open in a new tab
a

Comparison of cases and controls using logistic regression.

b

CSE=certificate of secondary education, O-level=ordinary level, A-level=advanced level.

Atrazine was detected in only 0.8% of the study participants. DACT was the only analyte detected above the LOD in over 50% of the participants (58.3%) (Table 2). Other less frequently detected analytes were DEA (6.2%), desethyl atrazine mercapturate (3.3%), atrazine mercapturate (1.4%), hydroxyl atrazine (1%) and DIA (0.5%).

Table 2.

Gestational urine concentrations (ng/mL).

Analyte Total

N (%) > LODa 		Casesc
Controlsc
p-value
N (%) > LOD Median (quartiles) N (%) > LOD Median (quartiles)
Diaminochlorotriazine (DACT) 369 (58.3) 174 (60.3) 0.27 (0.16–0.44) 195 (56.4) 0.24 (0.16–0.41) 0.31b
Desethyl atrazine (DEA) 369 (6.2) 174 (4.6) 195 (7.7)
Desethyl atrazine mercapturate 369 (3.3) 174 (2.3) 195 (4.1)
Atrazine mercapturate 369 (1.4) 174 (0.6) 195 (2.1)
Hydroxyl atrazine 369 (1.1) 174 (0.6) 195. (1.5)
Atrazine 369 (0.8) 174 (0.6) 195 (1)
Desisopropyl atrazine (DIA) 369 (0.5) 174 (0.6) 195 (0.5)
Open in a new tab
a

The LODs were: atrazine, 0.05 ng/mL; DACT, 0.19 ng/mL; DEA, 0.10 ng/mL; DIA, 0.12 ng/mL; atrazine mercapturate, 0.08 ng/mL; desethyl atrazine mercapturate, 0.09 ng/mL; and hydroxyl atrazine, 0.14 ng/mL.

b

Comparison between cases and controls using Wilcoxon Rank Sum test.

c

Urine creatinine median (IQR) μg/g: Cases 81 (48–134) μg/g; Controls 80 (52–118) μg/g.

Because of low detection of the other atrazine analytes, only DACT was included in the exposure–outcome analyses. Mothers prenatal BMI, child’s BMI at age 7, and child’s duration of breastfeeding were identified as potential confounders of the in utero exposure to DACT and early menarche association.

For the association between early menarche and dichotomized DACT exposure, the unadjusted odds ratio was 1.18 (95% CI: 0.78, 1.78), and the adjusted odds ratio was 1.12 (95% CI: 0.90, 1.40) (Table 3). For the association between early menarche and DACT exposure categories, the unadjusted odds ratio was 1.01 (95% CI: 0.61, 1.67) for the < median category and 1.34 (95% CI: 0.83, 2.18) for the ≥median category compared to the reference (< LOD) (Table 3). The adjusted odds ratio was 1.03 (95% CI: 0.74, 1.45) for the < median category and 1.13 (0.82, 1.55) for the ≥median category compared to the reference (Table 3). Even though the odds ratios were > 1, none of the associations were statistically significant (the 95% CI included the null).

Table 3.

Association between DACT gestational urine concentrations (ng/mL) and early menarche (using data with imputed values).

DACT N cases/controls Unadjusted OR (95%CI) Adjusted OR (95% CI)
Dichotomous exposure
 < LOD 69/85 1 Reference 1 Reference
 ≥LOD 105/110 1.18 (0.78–1.78)a 1.12 (0.90–1.40)
Categorical exposure
 < LODd 69/85 1 Reference 1 Reference
 LOD– < Mediand 45/55 1.01 (0.61–1.67)b 1.03 (0.74–1.45)a,cb,c
 ≥Mediand 60/55 1.34 (0.83–2.18)b 1.13 (0.82–1.55)b,c
Open in a new tab
a

Odds of early menarche for DACT exposure ≥LOD compared to exposure < LOD.

b

Odds of early menarche by DACT exposure category compared to reference (< LOD).

c

Adjusted for Mother’s pre-natal BMI, childhood BMI at age 7, duration of breastfeeding, and urine sample creatinine.

d

DACT LOD=0.19 ng/mL and Median of controls ≥LOD=0.4 ng/mL.

When we excluded the cases and controls without confounder data from the analyses, the adjusted odds ratio for dichotomous DACT exposure was 1.58 (95% CI: 0.94–2.66) (Table 4). The association remained non-significant. For the categorical exposure, the adjusted odds of early menarche for girls with DACT exposures < median was 1.26 (95% CI: 0.65, 2.42) and ≥median was 1.86 (95% CI: 1.03, 3.38) compared to the reference (Table 4). The association was significant for in utero exposure to DACT≥median compared to the reference.

Table 4.

Association between DACT gestational urine concentrations (ng/mL) and early menarche (using complete case data).

DACT N cases/controls Unadjusted OR (95%CI) Adjusted OR (95% CI)
Dichotomous exposure
 < LOD 45/68 1 Reference 1 Reference
 ≥LOD 77/74 1.57 (0.96–2.58)a,c 1.58 (0.94–2.66)a,c
Categorical exposure
 < LODd 45/68 1 Reference 1 Reference
 LOD– < Mediand 28/38 1.11 (0.60–2.06)b 1.26 (0.65–2.42)b,c
 ≥Mediand 49/36 2.06 (1.16–3.64)b 1.86 (1.03–3.38)b,c
Open in a new tab
a

Odds of early menarche for DACT exposure ≥LOD compared to exposure < LOD.

b

Odds of early menarche by DACT exposure category compared to reference (< LOD).

c

Adjusted for Mother’s pre-natal BMI, childhood BMI at age 7, duration of breastfeeding, and urine sample creatinine.

d

DACT LOD=0.19 ng/mL and Median of controls ≥LOD=0.4 ng/mL.

4. Discussion

To our knowledge, this study is the first to analyze the association between exposure to atrazine analytes and timing of menarche in humans. Our results suggest that earlier menarche may be associated with in utero exposure to DACT, and though the association is weak with wide confidence intervals that include the null, it was statistically significant in higher exposures (≥median) for the subset of girls with complete data.

The distribution profile of the analytes detected in the present study population suggests environmental exposure to atrazine, i.e., through drinking water or food, as opposed to occupational exposures (Barr et al., 2007). Assessments of streams and ground water in the US showed that as atrazine persists in ground water, it breaks down to its degradates (e.g., DEA, DACT and DIA) such that the degradates are detected as frequently as (Barbash et al., 1999), and at higher concentrations than the parent compound (Gilliom et al., 2006). Barr et al. (2007), also reported that DACT is the analyte predominantly detected (77%) in populations exposed environmentally to atrazine, followed distantly by DEA (15%), DIA (6%) and then atrazine mercapturate (2%). In our study, we detected DACT at 58%, followed by DEA (6%). We detected atrazine mercapturate and DIA in 1.4% and 0.5% of our study participants respectively. The Barr et al. (2007) study had a small number of participants (N=24), and even though DACT was the predominant atrazine analyte detected in the different exposure categories (environmental, occupational and acute), the inter-person variations were about 30%. Further evaluation in larger studies is needed (Barr et al., 2007).

In our study, a single sample of urine collected during pregnancy from each mother was used. Since atrazine has a short half-life of 8 h with elimination from the body occurring in 24–48 h, (Catenacci et al., 1993) atrazine analytes measured in the urine sample would most likely only describe exposure in the past 24–48 h, and may not be an accurate representation of exposure throughout the pregnancy. However, the use of atrazine, e.g., agricultural applications, can lead to contaminated surface and ground water, which can result in persistent contamination of drinking water supplies (ATSDR, 2003; Munger, 1997). If exposure to atrazine in this population occurred regularly through drinking water as suggested by the environmental exposure profile, the measurement of atrazine analytes in this study may be an adequate approximation of long-term exposure. Although the use of only one urine sample may introduce error to the classification of exposure, this classification error is likely to be non-differential for cases and controls.

In previous human exposure studies in the U.S., the analyte mainly measured has been atrazine mercapturate (Arcury, 2007; Bakke, 2009). As atrazine is still in use in several countries including U.S., other metabolites, including DACT, need to be measured to better assess exposure (Barr, 2007; Chevrier, 2011). However, it is necessary to note that DACT, DEA and DIA can also result from the metabolism of other chlorotriazine pesticides such as simazine and propazine, or from their breakdown in the environment (ATSDR, 2003). Therefore, DACT exposure in our study could include contributions from other chlorotriazines besides atrazine. Barr et al. (2007) state that “measuring atrazine compound or atrazine mercapturate in urine is the unequivocal indication that a person was exposed to atrazine and not an environment degradate”. Future exposure studies could also assess how concentrations of atrazine and analytes measured in environmental media such as surface or drinking water can be used to predict human exposure, e.g., using physiologically based pharmacokinetic models.

Our study results conflict with animal studies suggesting that prenatal or lactational exposure to atrazine and DACT may delay the onset of puberty by altering the hypothalamic-pituitary activity. In most of these studies the doses used to induce the observed outcomes are typically very high (50,000–200,000 ppb), and do not compare to environmental levels that human populations are usually exposed to. Epidemiologic studies suggest that in utero exposure to the estrogenic effect of some EDCs e.g., dichlorodiphenyldichloroethene (DDE), a metabolite of the pesticide dichlorodiphenyltrichloroethane (DDT), is associated with early puberty (Blanck, 2000; Vasiliu, 2004). Animal studies suggest that atrazine does not bind to the estrogen receptor, and that the hypothalamic function is the main target of atrazine exposure (Cooper et al., 2007). A suggested endocrine disrupting mechanism of action associated with atrazine exposure is aromatase induction, which results in an increase in estrogen (Hayes et al., 2002; Sanderson et al., 2001). This increase may contribute to some estrogen-mediated toxicities observed in animal studies (e.g., induction and earlier onset of the incidence of mammary gland tumors, and lengthening of estrus cycle in rats) (Sanderson et al., 2001). More experimental studies conducted using doses of atrazine analytes at background levels, i.e., doses that populations are more likely to be exposed to, are needed to determine health outcomes, including onset of puberty, and mechanisms of action.

The exclusion of cases and controls with missing data on one or more confounders may bias the results presented in Table 4. First, the distribution of DACT exposure differed between participants with complete data and those with missing data. Cases with complete data had a median DACT concentration (25th–75th percentile) of 0.32 ng/mL (0.16–0.45 ng/mL) while cases with missing data had a median concentration of 0.19 ng/mL (0.15–0.39 ng/mL). Controls with complete data had a median DACT concentration of 0.18 ng/mL (0.16–0.50 ng/mL) while controls with missing data had a median concentration of 0.32 ng/mL (0.17–0.43 ng/mL). Second, some levels of confounders differed between participants with complete data compared to those with missing data (see Supplement). For example, cases with complete data were less likely than cases with missing data to have a mother who was overweight before giving birth (3% vs. 22%) and more likely to have a mother who was obese before giving birth (22% vs. 7%). Compared to cases with complete data, cases with missing data were less likely to have been breast fed for over 6 months (26% vs. 39%) and more likely never to have been breast fed (23% vs. 15%). Controls with missing data were less likely than controls with complete data to have been breast fed for over 6 months (22% vs. 47%) and more likely to have been breast fed for less than 6 months (51% vs. 34%).

A strength of this study is the ability to link to several demographic and health-related characteristics of the child and mother. A limitation of this study is use of only one urine sample during the pregnancy period. A second limitation is that age at menarche was self-reported by children or their parents, which could be subject to recall bias. Another limitation is that 29% of the participants were missing data on one or more of the confounders. Lastly, this study analyzed the association between in utero exposure to DACT and early menarche, not onset of puberty, which is marked by Tanner stage 2 breast development.

5. Conclusion

Our results show a non-significant association between early menarche and in utero exposure to the atrazine analyte DACT in the ALSPAC cohort. In the subset of girls with complete data for confounders, the association was statistically significant for girls with higher exposures. Further exploration of the role of exposure to atrazine analytes in female reproduction in other large independent cohorts is needed. To comprehensively assess atrazine exposure in human populations, several analytes including DACT, may need to be measured along with atrazine parent compound and atrazine mercapturate.

Supplementary Material

Supplemental Material

Acknowledgments

We are extremely grateful to all the families who took part in this study, the midwives for their help in recruiting them, and the whole ALSPAC team, which includes interviewers, computer and laboratory technicians, clerical workers, research scientists, volunteers, managers, receptionists and nurses. The UK Medical Research Council and the Wellcome Trust (Grant ref: 102215/2/13/2) and the University of Bristol provide core support for ALSPAC. This work was specifically funded by Centers for Disease Control and Prevention. KN is supported by the National Institute for Health Research (NIHR) Collaboration for Leadership in Applied Health Research and Care West (CLAHRC West) at University Hospitals Bristol NHS Foundation Trust. The findings and conclusions in this report are those of the authors and do not necessarily represent the views of the Centers for Disease Control and Prevention, the NHS, the NIHR or the Department of Health.

Appendix A. Supporting information

Supplementary data associated with this article can be found in the online version at doi:10.1016/j.envres.2017.04.004.

References

  1. Adair L. Size at birth predicts age at Menarche. Pediatrics. 2001;107:4. doi: 10.1542/peds.107.4.e59. [DOI] [PubMed] [Google Scholar]
  2. Aksglaede L, Olsen LW, Sorensen TIA, Juul A. Forty years trends in timing of pubertal growth spurt in 157,000 Danish school children. PLoS One. 2008;3:7. doi: 10.1371/journal.pone.0002728. [DOI] [PMC free article] [PubMed] [Google Scholar]
  3. Arcury T, Grzywacz JG, Barr DB, Tapia J, Chen H, Quandt SA. Pesticide urinary metabolite levels of children in Eastern North Carolina farmworker households. Environ Health Perspect. 2007;15(7):1254–1260. doi: 10.1289/ehp.9975. [DOI] [PMC free article] [PubMed] [Google Scholar]
  4. ATSDR (Agency for Toxic Substances and Disease Registry) Toxicological Profile for Atrazine. U.S. Centers for Disease Control and Prevention; Atlanta, GA: 2003. [PubMed] [Google Scholar]
  5. Bakke B, De Roos AJ, Barr D, Stewart PA, Blair A, Freeman LB, et al. Exposure to atrazine and selected non-persistent pesticides among corn farmers during a growing season. J Expo Sci Environ Epidemiol. 2009;19:544–554. doi: 10.1038/jes.2008.53. [DOI] [PMC free article] [PubMed] [Google Scholar]
  6. Balduini L, Matoga M, Cavalli E, Seilles E, Riethmuller D, Thomassin M, et al. Triazine herbicide determination by gas chromatography-mass spectrometry in breast milk. J Chromatogr B. 2003;794:389–395. doi: 10.1016/s1570-0232(03)00455-0. [DOI] [PubMed] [Google Scholar]
  7. Barbash JE, Thelin GP, Kolpin DW, Gilliom RJ. Distribution of major herbicides in ground water of the United States. USGS EPA Water-Resour Investig Rep. 1999:98–4245. [Google Scholar]
  8. Barr D, Panuwet P, Nguyen JV, Udunka S, Needham LL. Assessing exposure to atrazine and its metabolites using biomonitoring. Environ Health Perspect. 2007;115(10):1474–1478. doi: 10.1289/ehp.10141. [DOI] [PMC free article] [PubMed] [Google Scholar]
  9. Biro FM, Galvez MP, Greenspan LC, Succop PA, Vangeepuram N, Pinney SM, et al. Pubertal assessment method and baseline characteristics in a mixed longitudinal study of girls. Pediatrics. 2010;126(3):e583–e590. doi: 10.1542/peds.2009-3079. [DOI] [PMC free article] [PubMed] [Google Scholar]
  10. Blanck H, Marcus M, Tolbert PE, Rubin C, Henderson AK, Hertzberg VS, et al. Age at menarche and tanner stage in girls exposed in utero and postnatally to polybrominated biphenyl. Epidemiology. 2000;11:641–647. doi: 10.1097/00001648-200011000-00005. [DOI] [PubMed] [Google Scholar]
  11. Boyd A, Golding J, Macleod J, Lawlor DA, Fraser A, Henderson J, et al. Cohort profile: the ‘Children of the 90s’ – the Avon longitudinal study of parents and children. Int J Epidemiol. 2013;42:1. doi: 10.1093/ije/dys064. [DOI] [PMC free article] [PubMed] [Google Scholar]
  12. Catenacci G, Barbieri F, Bersani M, Fereoli A, Cottoca D, Maroni M. Biological monitoring of human exposure to atrazine. Toxicol Lett. 1993;69(2):217–222. doi: 10.1016/0378-4274(93)90107-9. [DOI] [PubMed] [Google Scholar]
  13. Chevrier C, Limon G, Monfort C, Rouget F, Garlanteezec R, Petit C, et al. Urinary biomarkers of prenatal atrazine exposure and adverse birth outcomes on the PELAGIE birth cohort. Environ Health Perspect. 2011;119(7):1034–1041. doi: 10.1289/ehp.1002775. [DOI] [PMC free article] [PubMed] [Google Scholar]
  14. Christensen KY, Maisonet M, Rubin C, Holmes A, Calafat AM, Kato K, et al. Exposure to polyfluoroalkyl chemicals during pregnancy is not associated with offspring age at menarche in a contemporary British cohort. Environ Int. 2011;37:129–135. doi: 10.1016/j.envint.2010.08.007. [DOI] [PMC free article] [PubMed] [Google Scholar]
  15. Christensen KY, Maisonet M, Rubin R, Holmes A, Flanders WD, Heron J, et al. Pubertal pathways in girls enrolled in a contemporary British cohort. Int J Pediatr. 2010 doi: 10.1155/2010/329261. [DOI] [PMC free article] [PubMed] [Google Scholar]
  16. Cooper RL, Laws SC, Das PC, Narotsky MG, Goldman JM, Tyrey EL, et al. Atrazine and reproductive function: mode and mechanism of action studies. Birth Defects Res (Part B) 2007;80:98–112. doi: 10.1002/bdrb.20110. [DOI] [PubMed] [Google Scholar]
  17. Davis LK, Murr AS, Best DB, Fraites MJP, Zorrilla LM, Narotsky MG, et al. The effects of prenatal exposure to atrazine on pubertal and postnatal reproductive indices in the female rat. Reprod Toxicol. 2011;32:43–51. doi: 10.1016/j.reprotox.2011.04.004. [DOI] [PubMed] [Google Scholar]
  18. EPA (U.S. Environmental Protection Agency) Atrazine Background Fact Sheet. 2012 〈 http://www.epa.gov/pesticides/factsheets/atrazine_background.htm〉. (accessed October 2013)
  19. Euling SY, Herman-Giddens ME, Lee PA, Selevan SG, Juul A, Sorenson TIA, et al. Examination of US puberty-timing data from 1940 to 1994 for secular trends: panel findings. Pediatrics. 2008;121 doi: 10.1542/peds.2007-1813D. [DOI] [PubMed] [Google Scholar]
  20. Fera Science Ltd. Pesticide Usage Surveys. 2016 Available: 〈 https://secure.fera.defra.gov.uk/pusstats/index.cfm〉. (accessed April 2016)
  21. Fraser A, Macdonald-Wallis C, Tilling K, Boyd A, Golding J, Smith GD, et al. Cohort profile: the Avon longitudinal study of parents and children: ALSPAC mothers cohort. Int J Epidemiol. 2013;42:1. doi: 10.1093/ije/dys066. [DOI] [PMC free article] [PubMed] [Google Scholar]
  22. Gail MH, Brinton LA, Byar DP, Corle DK, Green SB, Schairer C, Mulvihill JJ. Projecting individualized probabilities of developing breast cancer for white females who are being examined annually. J Natl Cancer Inst. 1989;81:24. doi: 10.1093/jnci/81.24.1879. [DOI] [PubMed] [Google Scholar]
  23. Gillespie BW, Chen QX, Reichert H, Franzblau A, Hedgeman E, Lepkowski J, et al. Estimating population distributions when some data are below a limit of detection by using a reverse Kaplan-Meier estimator. Epidemiology. 2010;21(4):S64–S70. doi: 10.1097/EDE.0b013e3181ce9f08. [DOI] [PubMed] [Google Scholar]
  24. Gilliom RJ, Barbash JE, Crawford CG, Hamilton PA, Martin JD, Nakagaki N, et al. Pesticides in the nation’s streams and groundwater, 1992–2001. US Geol Surv Circ. 2006;1291:172. [Google Scholar]
  25. Hayes TB, Collins A, Lee M, Mendoza M, Nortega N, Stuart AA, et al. Hermaphroditic, demasculinized frogs after exposure to the herbicide atrazine at low ecologically relevant doses. Proc Natl Acad Sci USA. 2002;99(8):5476–5480. doi: 10.1073/pnas.082121499. [DOI] [PMC free article] [PubMed] [Google Scholar]
  26. Laws SC, Ferrell JM, Stoker TE, Cooper RL. Pubertal development in female wistar rats following exposure to propazine and atrazine biotransformation byproducts, diamino-s-chlorotriazine and hydroxyatrazine. Toxicol Sci. 2003;76:190–200. doi: 10.1093/toxsci/kfg223. [DOI] [PubMed] [Google Scholar]
  27. Laws SC, Ferrell JM, Stoker TE, Schmid J, Cooper RL. The effects of atrazine on female wistar rats: an evaluation of the protocol for assessing pubertal development and thyroid function. Toxicol Sci. 2000;58:366–376. doi: 10.1093/toxsci/58.2.366. [DOI] [PubMed] [Google Scholar]
  28. LeBaron H, McFarland J, Burnside O, editors. The Triazine Herbicides; 50 Years Revolutionizing Agriculture. Elsevier; Oxford, UK: 2008. [Google Scholar]
  29. Liu Y, De A. Multiple imputation by fully conditional specification for dealing with missing data in a large epidemiologic study. Int J Stat Med Res. 2015;4:287–295. doi: 10.6000/1929-6029.2015.04.03.7. [DOI] [PMC free article] [PubMed] [Google Scholar]
  30. McDowell MA, Brody DJ, Hughes JP. Has age at menarche changed? Results from the national health and nutrition examination survey (NHANES) 1999–2004. J Adolesc Health. 2007;40:227–231. doi: 10.1016/j.jadohealth.2006.10.002. [DOI] [PubMed] [Google Scholar]
  31. Mickey R, Greenland S. The impact of confounder selection criteria on effect estimation. Am J Epidemiol. 1989;129(1):125–137. doi: 10.1093/oxfordjournals.aje.a115101. [DOI] [PubMed] [Google Scholar]
  32. Moorman PG, Palmieri RT, Akushevich L, Berchuck A, Schildkraut JM. Ovarian cancer risk factors in African-American and white women. Am J Epidemiol. 2009;170:5. doi: 10.1093/aje/kwp176. [DOI] [PMC free article] [PubMed] [Google Scholar]
  33. Munger R, Isacson P, Hu S, Burns T, Hanson J, Lynch CF, et al. Intrauterine growth retardation in Iowa communities with herbicide-contaminated drinking water supplies. Environ Health Perspect. 1997;105(3):308–314. doi: 10.1289/ehp.97105308. [DOI] [PMC free article] [PubMed] [Google Scholar]
  34. Ochoa-Acuna H, Frankenberger J, Hahn L, Carbajo C. Drinking-water herbicide exposure in Indiana and prevalence of small for gestational age and preterm delivery. Environ Health Perspect. 2009;117:10. doi: 10.1289/ehp.0900784. [DOI] [PMC free article] [PubMed] [Google Scholar]
  35. Panuwet P, Nguyen JV, Kuklenyik P, Udunka SO, Needham LL, Barr D. Quantification of atrazine and its metabolites in urine by on-line solid-phase extraction–high-performance liquid chromatography–tandem mass spectrometry. Anal Bioanal Chem. 2008;391:1931–1939. doi: 10.1007/s00216-008-2102-0. [DOI] [PubMed] [Google Scholar]
  36. Panuwet P, Restrepo PA, Magsumbol M, Jung KY, Montesano MA, Needham LL, et al. An improved high-performance liquid chromatography–tandem mass spectrometric method to measure atrazine and its metabolites in human urine. J Chromatogr B. 2010;878:957–962. doi: 10.1016/j.jchromb.2010.02.025. [DOI] [PubMed] [Google Scholar]
  37. Parent AS, Teilmann G, Juul A, Skakkebaek NE, Toppari J, Bourguignon JP. The timing of normal puberty and the age limits of sexual precocity: variations around the world, secular trends, and changes after migration. Endocr Rev. 2003;24(5):668–693. doi: 10.1210/er.2002-0019. [DOI] [PubMed] [Google Scholar]
  38. Rubin C, Maisonet M, Kieszak S, Monteih C, Holmes A, Flanders D, et al. Timing of maturation and predictors of menarche in girls enrolled in a contemporary British cohort. Paediatr Perinat Epidemiol. 2009;23:492–504. doi: 10.1111/j.1365-3016.2009.01055.x. [DOI] [PubMed] [Google Scholar]
  39. Ruder E, Hartman TJ, Rovine MJ, Dorgan JF. Birth characteristics and age at menarche: results from the dietary intervention study in children (DISC) Cancer Causes Control. 2010;21:1379–1386. doi: 10.1007/s10552-010-9565-y. [DOI] [PMC free article] [PubMed] [Google Scholar]
  40. Sanderson JT, Letcher RJ, Heneweer M, Giesy JP, van den Berg M. Effects of Chloro-s-Triazine herbicides and metabolites on aromatase activity in various human cell lines and on vitellogenin production in male carp hepatocytes. Environ Health Perspect. 2001;109(10):1027–1089. doi: 10.1289/ehp.011091027. [DOI] [PMC free article] [PubMed] [Google Scholar]
  41. Sass JB, Colangelo A. European Union bans atrazine, while the United States negotiates continued use. Int J Occup Environ Health. 2006;12(3):260–267. doi: 10.1179/oeh.2006.12.3.260. [DOI] [PubMed] [Google Scholar]
  42. Semiz S, Kurt DT, Zencir M, Sevinc O. Pubertal development of Turkish children. J Pediatr Endocrinol Metab. 2008;21:951–961. doi: 10.1515/jpem.2008.21.10.951. [DOI] [PubMed] [Google Scholar]
  43. UNEP (United Nations Environment Programme), WHO (World Health Organization) Interorganization programme for the sound management of chemicals. Geneva: 2012. State of the Science of Endocrine Disrupting Chemicals-2012. [Google Scholar]
  44. vanJaarsveld CHM, Fidler JA, Simon AE, Wardle J. Persistent impact of pubertal timing on trends in smoking, food choice, activity, and stress in adolescence. Psychosom Med. 2007;69:798–806. doi: 10.1097/PSY.0b013e3181576106. [DOI] [PubMed] [Google Scholar]
  45. Vasiliu O, Muttineni J, Karmau W. In utero exposure to organochlorines and age at menarche. Human Reprod. 2004;19(7):1506–1512. doi: 10.1093/humrep/deh292. [DOI] [PubMed] [Google Scholar]
  46. Villanueva C, Durand G, Coutte MB, Chevrier C, Cordier S. Atrazine in municipal drinking water and risk of low birthweight, preterm delivery, and small-for-gestational-age status. Occup Environ Med. 2005;62:400–405. doi: 10.1136/oem.2004.016469. [DOI] [PMC free article] [PubMed] [Google Scholar]
  47. Walvoord E. The timing of puberty: is it changing? Does it matter? J Adolesc Health. 2010;47:433–439. doi: 10.1016/j.jadohealth.2010.05.018. [DOI] [PubMed] [Google Scholar]
  48. Wetzel L, Luempert LG, III, Breckenridge CB, Tisdel MO, Stevens JT, Thakur AK, et al. Chronic effects of atrazine on estrus and mammary tumor formation in female Sprague-Dawley and Fischer 344 rats. J Toxicol Environ Health. 1994;43:169–182. doi: 10.1080/15287399409531913. [DOI] [PubMed] [Google Scholar]
  49. Whyatt RM, Barr DB, Camann DE, Kinney PL, Barr JR, Andrews HF, et al. Contemporary-use pesticides in personal air samples during pregnancy and blood samples at delivery among urban minority mothers and newborns. Environ Health Perspect. 2003;111(5):749–756. doi: 10.1289/ehp.5768. [DOI] [PMC free article] [PubMed] [Google Scholar]
  50. Winchester PJ, Huskins J, Ying J. Agrichemicals in surface water and birth defects in the United States. Acta Pædiatrica. 2008;98:664–669. doi: 10.1111/j.1651-2227.2008.01207.x. [DOI] [PMC free article] [PubMed] [Google Scholar]
  51. Windham GC, Pinney SM, Voss RW, Sjodin A, Biro FM, Greenspan LC, et al. Brominated flame retardants and other persistent organohalogenated compounds in relation to timing of puberty in a longitudinal study of girls. Environ Health Perspect. 2015;123:1046–1052. doi: 10.1289/ehp.1408778. [DOI] [PMC free article] [PubMed] [Google Scholar]
  52. Wolff MS, Teitelbaum SL, McGovern K, Pinney SM, Windham GC, Galvez M, et al. Environmental phenols and pubertal development in girls. Environ Int. 2015;8:174–180. doi: 10.1016/j.envint.2015.08.008. [DOI] [PMC free article] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

Supplemental Material
NIHMS916028-supplement-Supplemental_Material.docx (21.1KB, docx)

RESOURCES