Skip to main content
NCBI home page
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2024 Aug 15.
Published in final edited form as: Environ Res. 2023 May 4;231(Pt 1):115990. doi: 10.1016/j.envres.2023.115990

In Utero Exposure to Diethylstilbestrol and Blood DNA Methylation in Adult Women: Results from a Meta-analysis of Two Cohort Studies

Clara Bodelon 1, Gretchen L Gierach 1, Elizabeth E Hatch 2, Emily Riseberg 3, Amy Hutchinson 4, Meredith Yeager 4, Dale P Sandler 5, Jack A Taylor 5,6, Robert N Hoover 7, Zongli Xu 5, Linda Titus 8, Julie R Palmer 2,9, Rebecca Troisi 7
PMCID: PMC10442904  NIHMSID: NIHMS1900925  PMID: 37149030

Abstract

Background:

Prenatal exposure to diethylstilbestrol (DES) is associated with several adverse health outcomes. Animal studies have shown associations between prenatal DES exposure and DNA methylation.

Objective:

The aim of this study was to explore blood DNA methylation in women exposed and unexposed to DES in utero.

Methods:

Sixty women (40 exposed and 20 unexposed) in the National Cancer Institute’s Combined DES Cohort Study and 199 women (99 exposed and 100 unexposed women) in the Sister Study Cohort were included in this analysis. Within each study, robust linear regression models were used to assess associations between DES exposure and blood DNA methylation. Study-specific associations were combined using fixed-effect meta-analysis with inverse variance weights. Our analysis focused on CpG sites located within nine candidate genes identified in animal models. We further explored whether in utero DES exposure was associated with age acceleration.

Results:

Blood DNA methylation levels at 10 CpG sites in six of the nine candidate genes were statistically significantly associated with prenatal DES exposure (P<0.05) in this meta-analysis. Genes included EGF, EMB, EGFR, WNT11, FOS, and TGFB1, which are related to cell proliferation and differentiation. The most statistically significant CpG site was cg19830739 in gene EGF, and it was associated with lower methylation levels in women prenatally exposed to DES compared with those not exposed (P<0.0001; false discovery rate<0.05). The association between prenatal DES exposure in utero and age acceleration was not statistically significant (P=0.07 for meta-analyzed results).

Conclusions:

There are few opportunities to investigate the effects of prenatal DES exposure. These findings suggest that in utero DES exposure may be associated with differential blood DNA methylation levels, which could mediate the increased risk of several adverse health outcomes observed in exposed women. Our findings need further evaluation using larger data sets.

Keywords: Diethylstilbestrol, endocrine disruptor, methylation, epigenetics, in utero exposure

Background

Women who were exposed to the synthetic estrogen diethylstilbestrol (DES) in utero have a significantly increased risk of developing cancers of the vagina, breast, and precancerous lesions of the cervix (Herbst et al., 1971; Hoover et al., 2011), along with adverse reproductive and other health outcomes. DES is considered an environmental endocrine disruptor during fetal development (Gore et al., 2015) and has been classified as a carcinogen by the International Agency for Research on Cancer (Newbold et al., 2007).

The reasons for the increased risk of cancer and adverse reproductive outcomes associated with in utero DES exposure are not well-understood. Decades of studies in laboratory animals have identified multiple possible biological mechanisms including that DES exposure induces DNA methylation changes (Bromer et al., 2009; Coyle et al., 2020; Ho et al., 2017; Li et al., 2003; Li et al., 1997; Li et al., 2018; Singh et al., 2018). Mouse models have shown differential gene expression by DES exposure; genes include those associated with immune response, cell proliferation, differentiation, and apoptosis (Ma, 2009; Newbold, 1995; Sato et al., 2004). At least nine genes (EMB, WNT11, TGFB1, ERBB2, EGFR, LTF, EGF, FOS, and JUN) have previously been shown to have differential gene expression in mice following pre- or perinatal exposure to DES (Ho et al., 2017; Kamiya et al., 1996; Li et al., 1997; Miyagawa et al., 2004b; Nakamura et al., 2021; Nakamura et al., 2012; Nelson et al., 1994; Newbold et al., 2007; Yamashita et al., 2001). Animal models have further shown that DES exposure leads to altered DNA methylation dynamically during development and that 60% of genes with altered methylation have changes in expression (LeBaron et al., 2010; Li et al., 1997; Li et al., 2018).

A previous study conducted within the Sister Study Cohort investigated whether blood DNA methylation patterns differ among adult women exposed and unexposed to DES in utero. Although four potential CpG sites were identified across the genome, the CpG sites were not statistically significant after accounting for multiple testing (Harlid et al., 2015). This study further investigated methylation differences in 75 CpG sites located in the 5’ region of the nine candidate genes (Harlid et al., 2015), as the 5’ region has been previously shown to control gene expression (Suzuki and Bird, 2008); however, results were not statistically significant after accounting for multiple testing for this subset for CpG sites. In utero DES exposure was self-reported in that study, and therefore might have contributed to misclassification, leading to reduced statistical power.

Understanding the biological effects of prenatal DES exposure is important because of the large number of women exposed to DES in utero and the excess risk associated with that exposure. In addition, daughters of women exposed in utero to DES may be at increased risk of poor health outcomes (Titus et al., 2019; Titus-Ernstoff et al., 2008). Moreover, studies of women exposed to DES in utero presents a rare opportunity to investigate the effects of a well-documented endocrine disruptor. Here we explore the association between in utero DES exposure and blood DNA methylation in adult women by meta-analyzing results from the Sister Study Cohort (Harlid et al., 2015) along with newly generated results from the National Cancer Institute (NCI) Combined DES Cohort Follow-up Study. We focused on CpG sites located within nine candidate genes identified in the animal models. We further explored the relationships of in utero DES exposure and methylation at other sites in the genome and with measures of age acceleration.

Methods

Study Population

NCI’s Combined DES Cohort Follow-up Study.

The population from which the current sample was derived has been described previously (Troisi et al., 2018). Briefly, this is a feasibility within the NCI’s Combined DES Cohort Follow-up Study at Boston University (Boston, Massachusetts), which includes daughters from the National Cooperative Diethylstilbestrol Adenosis Project (DESAD) (Labarthe et al., 1978) and the Women’s Health Study Daughters Cohort (Greenberg et al., 1984). Women in the NCI’s Combined DES Cohort Follow-up Study were defined as exposed to DES if maternal medical records indicated prescription of DES during pregnancy. Questionnaires including questions regarding reproductive health were mailed to participants approximately every 5 years starting in 1994. Information on dose and gestational age at the time of DES exposure was obtained from medical records. Women eligible for inclusion in the feasibility study were those who responded to the most recent study questionnaire in 2011, lived within 30 miles of the Boston University Medical School General Clinical Research Unit (GCRU), were postmenopausal, never had a cancer diagnosis and had never taken hormone supplements (Supplemental Figure 1). Of the 300 women who met the inclusion criteria, 180 were invited to participate. Women with reported vaginal epithelial changes (VEC) were all invited to participate, as women with VEC are at higher risk for many of the DES-associated adverse outcomes (Hoover et al., 2011). As this was a feasibility study, recruitment stopped once 60 women, 40 exposed to DES in utero and 20 unexposed, agreed to participate. At the time of blood collection, women completed a short questionnaire that included demographic and behavioral information, including alcohol use in the day prior to blood collection and smoking status (ever/never). Height and weight were measured at the time of blood collection, and body mass index (BMI, kg/m2) was calculated. All participants in the analysis were self-identified as non-Hispanic white women, ages 51-67.

Informed consent was obtained from all participants prior to enrollment. This study was approved by the Institutional Review Boards (IRBs) of the National Cancer Institute and Boston University.

Sister Study Cohort.

The Sister Study is a prospective cohort of cancer-free women with at least one sister diagnosed with breast cancer (http://sisterstudy.niehs.nih.gov). Women completed questionnaires that queried information related to family history, lifestyle, reproductive factors, and prenatal exposures, and provided a blood sample at baseline. A case-control analysis of prenatal DES exposures and methylation has been described in detail (Harlid et al., 2015) and included 99 non-Hispanic white women, ages 40-59 years who reported by questionnaires that they had prenatal exposure to DES and 100 unexposed women.

Informed consent was obtained from all participants prior to enrollment, and the IRBs of the National Institutes of Environmental Health Sciences (NIEHS) and the Copernicus Group approved the study.

DNA Extraction and Methylation Array

NCI’s Combined DES Cohort Follow-up Study.

Whole blood was collected in EDTA tubes, stored at −80° at Boston University Medical School’s GCRU and later shipped to the Cancer Genomics Research Laboratory in the Division of Cancer Epidemiology and Genetics at NCI. Genomic DNA was isolated from buffy coat samples using the QIAsymphony (Qiagen, Germantown, MD) automated extraction instrument. 400 ng of sample DNA, according to Quant-iT PicoGreen dsDNA quantitation (ThermoFisher Scientific, Waltham, MA), was treated with sodium bisulfite using the EZ-96 DNA Methylation MagPrep Kit (Zymo Research, Irvine, CA) according to the manufacturer-provided protocol. Bisulfite-treated samples were denatured and neutralized, then whole genome amplified, isothermally, to increase the amount of DNA template. Methylation was measured using the Infinium MethylationEPIC BeadChip (Illumina Inc., San Diego, CA), which interrogates over 850,000 CpG sites in the genome. Samples were run in a single batch. A laboratory internal control, NA07057 (Coriell Cell Repositories, Camden, NJ), was utilized to confirm the efficiency of bisulfite conversion. In addition, three samples were run in duplicate, and correlations of methylation values for these duplicates were greater than 0.99.

Sister Study Cohort.

The information regarding the DNA extraction and methylation has been previously described (Harlid et al., 2015). Briefly, one microgram (μg) of DNA extracted from frozen whole blood samples was treated with sodium bisulfite using the EZ-96 DNA Methylation MagPrep Kit (Zymo Research, Irvine, CA) according to the manufacturer-provided protocol. Methylation was measured using the Illumina HumanMethylation450 BeadChip (Illumina Inc., San Diego, CA), which interrogates over 450,000 CpG sites in the genome with >90% of them also present in the Infinium MethylationEPIC BeadChip (Moran et al., 2016).

Processing of Methylation Data

Processing of the methylation data was similar in both studies. The Illumina iScan was used to obtain the raw intensity data (iDAT files). Quality control and normalization of the raw data were carried out using the ENmix R package (Xu et al., 2016; Xu et al., 2021a). After quality control analysis, CpG sites were excluded before background correction if the quality of the methylation in over 5% of the samples was low. The REgression on Logarithm of Internal Control probes (RELIC) method (Xu et al., 2017) was used for dye-bias correction, the ENmix model with out-of-band (oob) probes were used for background correction, and normalization was performed using the quantile normalization method. CpG sites on the Y chromosome were also excluded. Missing methylation values were imputed using the k-nearest neighbor method with default settings. We use the nmode function of ENmix to identify and exclude multimodal CpG sites, which are likely caused by SNP effects or unknown artifacts in the design of the probe (Xu et al., 2016). Cell type proportions were estimated using the Houseman et al. method (Houseman et al., 2012). Surrogate variables were estimated using non-negative internal control probes and these variables were used to adjust for unwanted variations present in high-throughput experiments.

Information on the CpG sites was obtained from the Illumina annotations according to the CRCh37/hg19 assembly. CpG sites in the candidate genes were included in the analysis if the gene names were listed in the UCSC or GENECODE target name in the manifest file. These were CpG sites within the gene and the promoter region. CpG sites in promoter regions were those located within 200 bp or 1500 bp upstream of the transcription start site (TSS), 5′UTRs, or exon 1 (Moran et al., 2016).

Statistical Analysis

Separate analyses were conducted within each of the two cohort studies. Robust linear regression models were used to compute associations between in utero DES exposure (independent variable) and methylation levels at individual CpG sites (outcome). The regression models included terms for age at blood draw, BMI, smoking status, cell type proportions (CD8 T cells, CD4 T cells, B cells, natural killer cells and monocytes) and control surrogate variable analysis (SVA) variables. Fixed effects meta-analysis with inverse-variance weights was conducted to combine the associations from the Sister Study and the NCI’s Combined DES Cohort. Heterogeneity in the associations across studies was assessed using the I2 statistic. Meta-analytic associations for the a priori candidate genes were considered statistically significant if P<0.05.

Our analyses focused on the associations for CpG sites located in the nine candidate genes (EMB, WNT11, TGFB1, ERBB2, EGFR, LTF, EGF, FOS, and JUN). Because multiple CpG sites are located in the same genes, we assessed whether differentially methylated regions (DMR) could be detected in these genes by combining the meta-analytic P-values in these genes. We used two algorithms ipDMR (Xu et al., 2021b) and Comb-p (Pedersen et al., 2012) with a false discovery rate (FDR) of 0.05, both available in the ENmix R package. These two algorithms have shown to have good performance (Mallik et al., 2019; Xu et al., 2021b).

We also conducted exploratory analyses of the association between in utero DES exposure and methylation levels at other sites present in the methylation arrays outside of the candidate genes. Meta-analysis was conducted for the associations with the CpG sites common to both the Illumina HumanMethylation450 and the Infinium MethylationEPIC BeadChip arrays. For the CpG sites present only in the Infinium MethylationEPIC BeadChip array, results are restricted to NCI’s Combined DES Cohort. The Bonferroni correction level for the 812,969 CpG sites available for analysis in the MethylationEPIC was used to account for multiple comparisons in this exploratory analysis and CpG sites with P<6.15×10−8 were considered statistically significant.

Finally, in each of the two cohorts, we investigated associations between in utero DES exposure and measures of age acceleration derived from the Horvath, Hannum, PhenoAge (also known as Levine) and GrimAge epigenetic clocks (Oblak et al., 2021). Both Horvath and Hannum are predictors of chronological age (Hannum et al., 2013; Horvath, 2013), while PhenoAge and GrimAge are predictors of lifespan and health span (Levine et al., 2018; Lu et al., 2019). Measures of epigenetic age and age acceleration were computed using the online calculator https://dnamage.genetics.ucla.edu/ by submitting normalized methylation values. Age acceleration measures are defined as the residuals of the regression of the age epigenetic measures on chronological age. Robust linear regression was used to compute the associations between in utero DES (independent variable) and age acceleration measures (outcome) adjusted for BMI and smoking for each study independently and meta-analyzed as described above.

Results

Table 1 shows the characteristics of the participants from the NCI’s Combined DES Cohort and the Sister Study Cohort included in this analysis. The mean age at blood donation among women who were prenatally exposed and unexposed to DES were 60.3 years and 61.2 in the NCI’s Combined DES Cohort and 50.6 and 50.7 in the Sister Study, respectively. Ages at menarche and at menopause were similar in exposed and unexposed women. The mean age at first DES exposure was 10.8 weeks in the NCI’s Combined DES Cohort (gestational age at exposure was not available for the Sister Study Cohort).

Table 1.

Characteristics of a subset of participants in the NCI’s Combined DES Cohort Study and Sister Study Cohort by prenatal exposure to diethylstilbestrol (DES).

NCI’s Combined DES Cohort Study Sister Study Cohort
Characteristics DES Exposed
(N=40)		 DES Unexposed
(N=20)		 DES Exposed
(N=99)		 DES Unexposed
(N=100)
Age (years), mean (range) 60.3 (51.0-67.0) 61.2 (55.0-67.0) 50.6 (40.0-58.0) 50.7 (40.0-59.0)
Body mass index (BMI) (kg/m2), mean (range) 30.0 (20.9-55.4) 28.3 (18.7-47.1) 25.7 (19.0-47.8) 27.7 (17.5-47.5)
Age at menarche (years), mean (range) 12.7 (9.00-15.0) 12.5 (11.0-16.0) 12.9 (9.0-16.0) 12.6 (9.0-17.0)
Age at menopause (years), mean (range) 50.9 (40.0-57.0) 51.5 (47.0-57.0) 47.1 (31-55) 47.1 (32-56)
Parous, n (%)
 No 15 (37.5) 10 (50.0) 39 (39.4) 21 (21.0)
 Yes 25 (62.5) 10 (50.0) 60 (60.6) 79 (79.0)
Alcohol*, n (%)
 No 26 (65.0) 17 (85.0) 64 (64.6) 58 (58.0)
 Yes 14 (35.0)   3 (15.0) 35 (35.4) 42 (42.0)
Ever smoked**, n (%)
 No 21 (52.5) 11 (55.0) 55 (55.6) 62 (62.0)
 Yes 19 (47.5)   9 (45.0) 44 (44.4) 38 (38.0)
Gestational age at DES exposure (weeks), mean (range) 10.8 (3.00-30.0) - Not collected Not collected
DES dose, mean (range) 8,347.3 (480-16,475) - Not collected Not collected
Vaginal Epithelial Changes (VEC), n (%)
 No   4 (10.0) 20 (100%) Not collected Not collected
 Yes 36 (90.0) - Not collected Not collected
Open in a new tab
*

For the NCI’s Combined DES Cohort study, alcohol consumption refers to consumption within 24 hours before blood collection. For the Sister Study Cohort, it refers to general alcohol consumption.

**

For the NCI’s Combined DES Cohort study, smoking data was collected in the 2006 cohort follow-up questionnaire. For the Sister Study Cohort, information was collected at study entry.

There were 204 CpG sites common across the two arrays used in the Sister Study and the NCI’s Combined DES Cohort studies within the nine candidate genes (EMB, WNT11, TGFB1, ERBB2, EGFR, LTF, EGF, FOS, and JUN) (Supplemental Table 1). In utero exposure to DES was associated with methylation levels at 10 of these CpG sites representing six of nine candidate genes included in the meta-analysis (P<0.05; Figure 1). Directions of association for these 10 CpG sites were consistent in the two cohort studies. The most significant association in the meta-analysis was found at CpG site cg19830739 in EGF (P<0.0001 and FDR<0.05); women exposed in utero to DES had on average 0.21% lower methylation levels at this site compared with unexposed women. Since there were multiple significant CpG sites in several genes (EGFR, EGF and TGFB1: Figure 1), we assessed whether there were any DMR, but none were found at the FDR level of 0.05.

Figure 1 .

Figure 1 .

Open in a new tab

Meta-analytic associations were computed using a fixed effects meta-analysis. They were adjusted for age at blood draw, BMI, smoking status, cell type proportions (CD8 T cells, CD4 T Journal Pre-proof cells, B cells, natural killer cells and monocytes) and SVA variables in the individual studies. Beta indicates the mean difference in percent methylation levels in expose versus unexposed

Next, we examined epigenome-wide associations between in utero DES exposure and methylation (Supplemental Figure 2, Supplemental Figure 3, Supplemental Table 2, and Supplemental Table 3). There were no statistically significant associations for any of the 419,919 CpG sites located in the 450k array and included for analysis in both studies at the Bonferroni correction level or FDR level of 0.05 in the meta-analysis. We observed two sites (cg06393376 and cg00726784) with statistically significant associations at the Bonferroni correction level (P<6.15×10−8) in the EPIC array, but not in the 450k array, and therefore only available for analysis in the NCI’s Combined DES Cohort Study (Supplemental Figure 3 and Supplemental Table 3). CpG site cg06393376 in gene TRIOBP was associated with 2.7% lower methylation levels in DES exposed women (P=7.49×10−10) while intergenic site cg00726784 was associated with 0.4% higher levels in exposed women (P=4.32×10−8).

Finally, we investigated whether exposure in utero to DES was associated with age acceleration in these adult women. Average levels of age acceleration in women exposed in utero to DES were greater compared to those who were not exposed in the NCI’s Combined DES Cohort study, but not in the Sister Study (Table 2). Horvath, PhenoAge, and GrimAge measures of age acceleration were consistent in both studies in the direction of the association, but with varying effect sizes. Although in utero DES exposure was statistically significantly associated with GrimAge (β=1.78, 95% CI: 0.05, 3.51; P=0.043) in the NCI’s Combined DES Cohort Study, none of the epigenetic measures of age acceleration were statistically significant in the meta-analysis (P=0.53, 0.73, 0.52, and 0.07 for Horvath, Hannum, PhenoAge and GrimAge, respectively).

Table 2.

Association between in utero exposure to diethylstilbestrol (DES) and epigenetic measures of age acceleration in adult women.

NCI’s Combined DES Cohort Feasibility Study Sister Study Cohort Fixed effect meta-analysis
Epigenetic measures of age acceleration Exposed
(N=40)
Mean (SD)		 Unexposed
(N=20)
Mean (SD)		 β (95% CI)* Exposed
(N=99)
Mean (SD)		 Unexposed
(N=100)
Mean (SD)		 β (95% CI)* β (95% CI) I2
Horvath 0.30 (4.24) −0.60 (4.34) 0.56
(−1.99, 3.10)		 −0.05 (4.25) 0.05 (4.47) 0.31
(−0.95,1.58)		 0.36
(−0.77, 1.50)		 0%
Hannum 0.37 (3.88) −0.74 (4.77) 1.16
(−1.26, 3.59)		 −0.39 (4.34) 0.30 (4.26) −0.05
(−1.22, 1.12)		 0.19
(−0.87, 1.24)		 0%
PhenoAge 0.71 (5.67) −1.41 (6.23) 1.85
(−1.65, 5.35)		 −0.25 (5.74) 0.25 (6.03) 0.18
(−1.41, 1.77)		 0.47
(−0.97, 1.91)		 0%
GrimAge 0.71 (3.68) −1.41 (2.66) 1.78
(0.05, 3.51) 		−0.03 (4.03) 0.03 (3.81) 0.40
(−0.35, 0.30)		 0.63
(−0.06, 1.31)		 0.53%
Open in a new tab
*

Associations were computed using robust linear regression models with epigenetic measures of age acceleration as the outcome. Beta indicates the mean difference in methylation age in expose versus unexposed. Models were adjusted for BMI and smoking.

Discussion

DES is an endocrine disruptor and carcinogen that adversely affects reproductive health and increases risk of cancer in women exposed in utero. This study presents novel findings on the association between adult women exposed to DES in utero and differential blood DNA methylation. We meta-analyzed data from two cohort studies and found several statistically significant associations in genes associated with in utero DES exposure in animal models. Patterns of association were consistent across both cohorts, suggesting the findings are robust. In an exploratory analysis, we also found some novel, associations of DES exposure with CpG sites located outside of the candidate genes.

Within six of the nine candidate genes, ten CpG sites were statistically significantly associated with in utero DES exposure. The most significant finding was at cg19830739, located in an enhancer region of an intron in gene EGF (epidermal growth factor), with lower levels of methylation associated with DES exposure. Lower methylation in this enhancer region may be related to higher gene expression. Vaginal tissues from neonatally DES exposed mice have higher EGF expression than unexposed mice (Miyagawa et al., 2004a). Another site, cg18905856, located within 1500 base pairs of the TSS in the promoter region of EGF was associated with higher methylation levels in the DES exposed compared with the unexposed. While these higher levels of methylation could be related to lower expression of EGF, this CpG site was the most distal to the TSS. We looked at multiple CpG sites in the promoter regions of EGF and the association with DES at the site most proximal to the TSS (cg24818418) was associated with lower methylation levels, although it did not reach statistical significance.

We observed three CpG sites significantly associated with DES in the EGFR (epidermal growth factor receptor) gene, which is highly expressed in normal placenta (Fagerberg et al., 2014) and is involved in cell proliferation. Two of the sites, cg09378783 and cg01461514, were in enhancer regions in an intron portion of the gene, while the other, cg02003682, was located in an exon. DES exposure was related to lower methylation levels at all three of these sites. Signal transduction via EGFR has been related to the estrogen-induced vaginal changes observed with DES exposure in animals (Miyagawa et al., 2004a), and together with sustained activation of EGF may result in cancerous lesions in the lower reproductive tract (Iguchi, 1992).

There were five other significant associations including cg03179043 in the promoter region in embigin gene EMB, which is involved in cell growth and development, and broadly expressed in multiple tissues, and cg18760094 in the promoter region in gene WNT11 (embigin), which is involved in oncogenesis and multiple developmental processes (Ono et al., 2013; Vijayaragavan et al., 2009). Both had lower levels of methylation which could be related to higher expression of these genes. The site cg00773696 in the promoter region in gene FOS (Fos proto-oncogene, AP-1 transcription factor subunit), a gene associated with cell proliferation and differentiation, but also with apoptotic cell death (Smeyne et al., 1993), was associated with higher methylation in prenatally DES exposed women, possibly inhibiting the expression of this gene. Finally, two sites in gene TGFB1 (transforming growth factor beta 1), cg27540367 and cg21198010, were associated with higher and lower methylation respectively. This gene can modulate multiple growth factors and is frequently expressed in tumors (Batlle and Massague, 2019).

There were two novel findings (CpG sites cg06393376 and cg00726784) from our exploratory analyses of CpG sites outside of the candidate genes. CpG site cg06393376 is located in gene TRIOBP and data from GTEx shows that this gene is most highly expressed in normal cervical tissues. CpG site cg00726784 is intergenic, located upstream from the TSS of gene ITGB6; however, it is unclear whether this particular site influences this gene. However, they could only be investigated in the NCI’s Combined DES Cohort study, as they are not included in the Illumina HumanMethylation450, and therefore require validation.

Epigenetic changes, including DNA methylation, are a hallmark of aging, and are associated with health and lifespan (López-Otín et al., 2013). Several epigenetic clocks that predict age or age-related phenotypes have been derived and are widely used to characterized biological age (or age acceleration, defined as the difference between the epigenetically derived age and chronological age) and predict age-related outcomes (Oblak et al., 2021). The four measures of age acceleration used in our analysis have been associated with all-cause mortality (Chen et al., 2016; Levine et al., 2018; Lu et al., 2019; Marioni et al., 2015) and age-related phenotypes (Oblak et al., 2021). The relationship between exposure to DES in utero and any of these measures of age acceleration had not been previously investigated. We found that in utero DES exposure was not significantly associated with age acceleration in the meta-analysis. However, in utero DES exposure was associated with increased age acceleration as measured by GrimAge (denoted this way due to the “grim” implications regarding morbidity and mortality) in the NCI’s Combined DES Cohort Study. Differences in exposure assessment could contribute to the inconsistent findings between the two studies. In utero DES exposure was based on medical records in the NCI’s Combined DES Cohort study, while it was self-reported in the Sister Study, which could have led to misclassification and bias to the null. Perhaps more importantly, the women included in the NCI study were recruited to increase the probability of demonstrating biological effects of in utero DES exposure as they were from a region of the country (Boston) that used high DES doses (Palmer et al., 2006; Troisi et al., 2018). This is consistent with their high prevalence of VEC (O’Brien et al., 1979). Future studies with larger sample sizes should further assess whether in utero DES exposure is associated with age acceleration.

Our current analysis has several strengths. First, our study added new data from the NCI’s Combined DES Cohort Study of prenatally exposed women to DES with methylation. Despite the relatively small sample size of the new data, we maximized the likelihood of finding associations by including highly exposed women with biological evidence of DES exposure (as determined by medical records verifying early exposure and high doses, resulting in VEC). In addition to the new dataset, we conducted a meta-analysis leveraging the only existing opportunities to study prenatal exposure to DES and DNA methylation in humans. We report significant associations in the meta-analysis for sites that, not only reach statistical significance, but also demonstrate consistency in their direction of association. So, the two studies serve as a validation for each other. Another major strength of our approach was the focus on a priori selected genes that had been previously identified in animal studies for their biological plausibility in being affected by prenatal exposure to DES. Our results suggest differential methylation patterns by in utero DES exposure in the candidate genes that were previously found to have aberrant expression in DES exposed animal models (Li et al., 2018; Nakamura et al., 2012). We conducted a more comprehensive analysis of CpG sites located in these candidate genes and included 204 CpG sites within any region of these genes and up to 1500 base pairs from the TSS. In contrast to Harlid et al.’s analysis was restricted to 75 CpG sites located in the 5’ region of the nine candidate genes (CpG sites in the 5’ regions are the most likely to change gene expression)(Harlid et al., 2015). The use of a large methylation array for wide coverage of the genome with over 800,000 CpG sites for analysis, accounting for approximately 4% of CpG sites in the genome, allowed us to explore other possible sites in relation to prenatal exposure to DES. The Sister Study used an older array which covered approximately 2% of CpG sites in the genome, and therefore we were able to conduct a meta-analysis only on sites shared across both arrays. A limitation was the small sample sizes of the NCI study and the Sister Study which limited our ability to detect weak signals. For this reason, we largely focused on the a priori selected genes with biological plausibility and conducted a meta-analysis to increase the power, as a single study had limited power to detect associations (Harlid et al., 2015). While we explored additional associations beyond the nine candidate genes, they will require validation in larger studies. Animal studies could also investigate the biological plausibility of these novel findings. Lastly, methylation was measured in blood DNA, instead of in target tissues from the vagina, cervix or breast where adverse outcomes have been reported. Whether blood DNA methylation is a surrogate of the methylation at the target tissues is largely unknown.

In conclusion, we conducted a comprehensive meta-analysis of in utero DES exposure and blood DNA methylation in adult female offspring and observed several statistically significant associations in genes previously identified as being affected by DES in animal models. Our results suggest that prenatal environmental exposures may lead to persistent changes in DNA methylation with the potential for adverse health effects. Opportunities for validation of findings in future studies with larger sample sizes should be considered.

Supplementary Material

1

Supplemental Figure 1. Participant selection scheme of the NCI’s Combined DES Cohort Feasibility Study.

2

Supplemental Figure 2. Meta-analysis results for associations between prenatal diethylstilbestrol (DES) exposure and methylation levels at individual CpG sites.

Supplemental Figure 2 legend. A. Manhattan plot for the meta-analytic associations between DES and methylation levels at 419,919 CpG sites common to the NCI’s Combined DES Cohort Study and Sister Study Cohort in the 450K methylation array. Black horizontal line indicates the Bonferroni correction level. B. Volcano plot for the meta-analytic associations between DES and methylation levels at 419,919 CpG sites common to the NCI’s Combined DES Cohort Study and Sister Study Cohort in the 450K methylation array. Black horizontal line indicates the Bonferroni correction level.

3

Supplemental Figure 3. Associations between prenatal diethylstilbestrol (DES) exposure and methylation levels at individual CpG sites in the 850K methylation array that are not included in the 450K methylation array; NCI’s Combined DES Cohort Study.

Supplemental Figure 3 legend. A. Manhattan plot for the associations between DES and methylation levels at 393,050 CpG sites in the 850K methylation array, that are not included in the 450K methylation array, from the NCI’s Combined DES Cohort Feasibility Study. Black horizontal line indicates the Bonferroni correction level for the entire 850K array. B. Volcano plot for the meta-analytic associations between DES and methylation levels at 393,050 CpG sites in the 850K methylation array that are not included in the 450K methylation array, from the NCI’s Combined DES Cohort Feasibility Study. Black horizontal line indicates the Bonferroni correction level for the entire 850K array. C. Boxplot of the methylation levels at the most significant CpG site (cg06393376) in the analysis in women exposed in utero and unexposed to DES in the NCI’s Combined DES Cohort Feasibility Study. D. Boxplot of the methylation levels at the second most significant CpG site (cg00726784) in the analysis among women exposed in utero and unexposed to DES in the NCI’s Combined DES Cohort Feasibility Study.

4

Supplemental Table 1. Associations of in utero diethylstilbestrol (DES) exposure with blood DNA methylation at the 207 CpG sites common across the two arrays used in the Sister Study Cohort and the NCI’s Combined DES Cohort Study within the nine candidate genes (EMB, WNT11, TGFB1, ERBB2, EGFR, LTF, EGF, FOS, and JUN). Results from the individual studies and the fixed-effect meta-analysis.

5

Supplemental Table 2. Associations of in utero diethylstilbestrol (DES) exposure with blood DNA methylation at CpG sites in the 450K array common across the two arrays used in the Sister Study Cohort and the NCI’s Combined DES Cohort Study. Results from the individual studies and the fixed-effect meta-analysis presented for the 100 most significant CpG sites, according to the meta-analytic p-value.

6

Supplemental Table 3. Associations of in utero diethylstilbestrol (DES) exposure with blood DNA methylation at CpG sites in the EPIC array, not in the 450k array, in the NCI’s Combined DES Cohort Study. Results presented for the 100 most significant CpG sites.

Highlights.

  • Millions of women were exposed to the synthetic estrogen diethylstilbestrol (DES) and in utero exposure to DES has been linked with several cancers in the daughters of the exposed women

  • Animal studies have shown associations between prenatal DES exposure and altered DNA methylation

  • Few opportunities exist to study the effects of the exposure of this endocrine disruptor in humans

  • DNA methylation at ten sites in genes related to cell proliferation and differentiation were associated with prenatal DES exposure in a meta-analysis of two cohort studies

Footnotes

Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

Credit author statement

Clara Bodelon: Conceptualization, methodology, formal analysis, writing – original draft; Gretchen L. Gierach: Supervision, methodology, writing – original draft; Elizabeth E. Hatch: Resources, writing – review & editing; Emily Riseberg: Formal analysis, writing – original draft; Amy Hutchinson: Resources, writing – review & editing; Meredith Yeager: Resources, writing – review & editing; Dale P. Sandler: Resources, writing – review & editing; Jack A. Taylor: Methodology, formal analysis, software, writing – review & editing; Robert N. Hoover: Resources, conceptualization, methodology, writing – review & editing; Zongli Xu: Methodology, formal analysis, software, writing – review & editing; Linda Titus: Resources, writing – review & editing; Julie R. Palmer: Resources, writing – review & editing; Rebecca Troisi: Conceptualization, methodology, supervision, writing – original draft.

Declaration of interests

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

References

  1. Batlle E, Massague J, 2019. Transforming Growth Factor-beta Signaling in Immunity and Cancer. Immunity. 50, 924–940. [DOI] [PMC free article] [PubMed] [Google Scholar]
  2. Bromer JG, et al. , 2009. Hypermethylation of homeobox A10 by in utero diethylstilbestrol exposure: an epigenetic mechanism for altered developmental programming. Endocrinology. 150, 3376–82. [DOI] [PMC free article] [PubMed] [Google Scholar]
  3. Chen BH, et al. , 2016. DNA methylation-based measures of biological age: meta-analysis predicting time to death. Aging (Albany NY). 8, 1844–1865. [DOI] [PMC free article] [PubMed] [Google Scholar]
  4. Coyle CS, et al. , 2020. Ovarian hormones induce de novo DNA methyltransferase expression in the Siberian hamster suprachiasmatic nucleus. Journal of Neuroendocrinology. 32, e12819. [DOI] [PubMed] [Google Scholar]
  5. Fagerberg L, et al. , 2014. Analysis of the human tissue-specific expression by genome-wide integration of transcriptomics and antibody-based proteomics. Mol Cell Proteomics. 13, 397–406. [DOI] [PMC free article] [PubMed] [Google Scholar]
  6. Gore AC, et al. , 2015. EDC-2: The Endocrine Society’s Second Scientific Statement on Endocrine-Disrupting Chemicals. Endocr Rev. 36, E1–E150. [DOI] [PMC free article] [PubMed] [Google Scholar]
  7. Greenberg ER, et al. , 1984. Breast Cancer in Mothers Given Diethylstilbestrol in Pregnancy. New England Journal of Medicine. 311, 1393–1398. [DOI] [PubMed] [Google Scholar]
  8. Hannum G, et al. , 2013. Genome-wide Methylation Profiles Reveal Quantitative Views of Human Aging Rates. Molecular Cell. 49, 359–367. [DOI] [PMC free article] [PubMed] [Google Scholar]
  9. Harlid S, et al. , 2015. In Utero Exposure to Diethylstilbestrol and Blood DNA Methylation in Women Ages 40–59 Years from the Sister Study. PLOS ONE. 10, e0118757. [DOI] [PMC free article] [PubMed] [Google Scholar]
  10. Herbst AL, et al. , 1971. Adenocarcinoma of the vagina. Association of maternal stilbestrol therapy with tumor appearance in young women. N Engl J Med. 284, 878–81. [DOI] [PubMed] [Google Scholar]
  11. Ho SM, et al. , 2017. Environmental factors, epigenetics, and developmental origin of reproductive disorders. Reprod Toxicol. 68, 85–104. [DOI] [PMC free article] [PubMed] [Google Scholar]
  12. Hoover RN, et al. , 2011. Adverse health outcomes in women exposed in utero to diethylstilbestrol. N Engl J Med. 365, 1304–14. [DOI] [PubMed] [Google Scholar]
  13. Horvath S, 2013. DNA methylation age of human tissues and cell types. Genome Biol. 14, R115. [DOI] [PMC free article] [PubMed] [Google Scholar]
  14. Houseman EA, et al. , 2012. DNA methylation arrays as surrogate measures of cell mixture distribution. BMC Bioinformatics. 13, 86. [DOI] [PMC free article] [PubMed] [Google Scholar]
  15. Iguchi T, Cellular Effects of Early Exposure to Sex Hormones and Antihormones. In: Jeon KW, Friedlander M, Eds.), International Review of Cytology. Academic Press, 1992, pp. 1–57. [DOI] [PubMed] [Google Scholar]
  16. Kamiya K, et al. , 1996. Expression of estrogen receptor and proto-oncogene messenger ribonucleic acids in reproductive tracts of neonatally diethylstilbestrol-exposed female mice with or without post-puberal estrogen administration. Experimental and clinical endocrinology & diabetes: official journal, German Society of Endocrinology [and] German Diabetes Association. 104, 111–122. [DOI] [PubMed] [Google Scholar]
  17. Labarthe D, et al. , 1978. Design and preliminary observations of National Cooperative Diethylstilbestrol Adenosis (DESAD) Project. Obstetrics and gynecology. 51, 453–458. [DOI] [PubMed] [Google Scholar]
  18. LeBaron MJ, et al. , 2010. Epigenetics and chemical safety assessment. Mutation Research/Reviews in Mutation Research. 705, 83–95. [DOI] [PubMed] [Google Scholar]
  19. Levine ME, et al. , 2018. An epigenetic biomarker of aging for lifespan and healthspan. Aging (Albany NY). 10, 573–591. [DOI] [PMC free article] [PubMed] [Google Scholar]
  20. Li S, et al. , 2003. Environmental exposure, DNA methylation, and gene regulation: lessons from diethylstilbesterol-induced cancers. Ann N Y Acad Sci. 983, 161–9. [DOI] [PubMed] [Google Scholar]
  21. Li S, et al. , 1997. Developmental exposure to diethylstilbestrol elicits demethylation of estrogen-responsive lactoferrin gene in mouse uterus. Cancer Res. 57, 4356–9. [PubMed] [Google Scholar]
  22. Li Y, et al. , 2018. DNA methylation and transcriptome aberrations mediated by ERalpha in mouse seminal vesicles following developmental DES exposure. Proc Natl Acad Sci U S A. 115, E4189–E4198. [DOI] [PMC free article] [PubMed] [Google Scholar]
  23. López-Otín C, et al. , 2013. The Hallmarks of Aging. Cell. 153, 1194–1217. [DOI] [PMC free article] [PubMed] [Google Scholar]
  24. Lu AT, et al. , 2019. DNA methylation GrimAge strongly predicts lifespan and healthspan. Aging (Albany NY). 11, 303–327. [DOI] [PMC free article] [PubMed] [Google Scholar]
  25. Ma L, 2009. Endocrine disruptors in female reproductive tract development and carcinogenesis. Trends Endocrinol Metab. 20, 357–63. [DOI] [PMC free article] [PubMed] [Google Scholar]
  26. Mallik S, et al. , 2019. An evaluation of supervised methods for identifying differentially methylated regions in Illumina methylation arrays. Brief Bioinform. 20, 2224–2235. [DOI] [PMC free article] [PubMed] [Google Scholar]
  27. Marioni RE, et al. , 2015. DNA methylation age of blood predicts all-cause mortality in later life. Genome Biol. 16, 25. [DOI] [PMC free article] [PubMed] [Google Scholar]
  28. Miyagawa S, et al. , 2004a. Estrogen-independent activation of erbBs signaling and estrogen receptor alpha in the mouse vagina exposed neonatally to diethylstilbestrol. Oncogene. 23, 340–9. [DOI] [PubMed] [Google Scholar]
  29. Miyagawa S, et al. , 2004b. Persistent gene expression in mouse vagina exposed neonatally to diethylstilbestrol. J Mol Endocrinol. 32, 663–77. [DOI] [PubMed] [Google Scholar]
  30. Moran S, et al. , 2016. Validation of a DNA methylation microarray for 850,000 CpG sites of the human genome enriched in enhancer sequences. Epigenomics. 8, 389–99. [DOI] [PMC free article] [PubMed] [Google Scholar]
  31. Nakamura M, et al. , 2021. Harnessing the chondroitin sulfate-binding characteristics of human lactoferrin to neutralize neurite outgrowth inhibition. Biochem Biophys Res Commun. 534, 1076–1082. [DOI] [PubMed] [Google Scholar]
  32. Nakamura T, et al. , 2012. Wnt family genes and their modulation in the ovary-independent and persistent vaginal epithelial cell proliferation and keratinization induced by neonatal diethylstilbestrol exposure in mice. Toxicology. 296, 13–19. [DOI] [PubMed] [Google Scholar]
  33. Nelson KG, et al. , 1994. Exposure to diethylstilbestrol during a critical developmental period of the mouse reproductive tract leads to persistent induction of two estrogen-regulated genes. Cell Growth Differ. 5, 595–606. [PubMed] [Google Scholar]
  34. Newbold R, 1995. Cellular and molecular effects of developmental exposure to diethylstilbestrol: implications for other environmental estrogens. Environ Health Perspect. 103 Suppl 7, 83–7. [DOI] [PMC free article] [PubMed] [Google Scholar]
  35. Newbold RR, et al. , 2007. Developmental exposure to diethylstilbestrol alters uterine gene expression that may be associated with uterine neoplasia later in life. Molecular Carcinogenesis. 46, 783–796. [DOI] [PMC free article] [PubMed] [Google Scholar]
  36. O’Brien PC, et al. , 1979. Vaginal epithelial changes in young women enrolled in the National Cooperative Diethylstilbestrol Adenosis (DESAD) project. Obstet Gynecol. 53, 300–8. [PubMed] [Google Scholar]
  37. Oblak L, et al. , 2021. A systematic review of biological, social and environmental factors associated with epigenetic clock acceleration. Ageing Res Rev. 69, 101348. [DOI] [PubMed] [Google Scholar]
  38. Ono M, et al. , 2013. Paracrine activation of WNT/beta-catenin pathway in uterine leiomyoma stem cells promotes tumor growth. Proc Natl Acad Sci U S A. 110, 17053–8. [DOI] [PMC free article] [PubMed] [Google Scholar]
  39. Palmer JR, et al. , 2006. Prenatal Diethylstilbestrol Exposure and Risk of Breast Cancer. Cancer Epidemiology, Biomarkers & Prevention. 15, 1509–1514. [DOI] [PubMed] [Google Scholar]
  40. Pedersen BS, et al. , 2012. Comb-p: software for combining, analyzing, grouping and correcting spatially correlated P-values. Bioinformatics. 28, 2986–8. [DOI] [PMC free article] [PubMed] [Google Scholar]
  41. Sato T, et al. , 2004. Involvement of growth factors in induction of persistent proliferation of vaginal epithelium of mice exposed neonatally to diethylstilbestrol. Reprod Toxicol. 19, 43–51. [DOI] [PubMed] [Google Scholar]
  42. Singh NP, et al. , 2018. Diethylstilbestrol (DES) induces autophagy in thymocytes by regulating Beclin-1 expression through epigenetic modulation. Toxicology. 410, 49–58. [DOI] [PMC free article] [PubMed] [Google Scholar]
  43. Smeyne RJ, et al. , 1993. Continuous c-fos expression precedes programmed cell death in vivo. Nature. 363, 166–9. [DOI] [PubMed] [Google Scholar]
  44. Suzuki MM, Bird A, 2008. DNA methylation landscapes: provocative insights from epigenomics. Nat Rev Genet. 9, 465–476. [DOI] [PubMed] [Google Scholar]
  45. Titus L, et al. , 2019. Reproductive and hormone-related outcomes in women whose mothers were exposed in utero to diethylstilbestrol (DES): A report from the US National Cancer Institute DES Third Generation Study. Reproductive Toxicology. 84, 32–38. [DOI] [PMC free article] [PubMed] [Google Scholar]
  46. Titus-Ernstoff L, et al. , 2008. Offspring of women exposed in utero to diethylstilbestrol (DES): a preliminary report of benign and malignant pathology in the third generation. Epidemiology. 19, 251–7. [DOI] [PubMed] [Google Scholar]
  47. Troisi R, et al. , 2018. Estrogen Metabolism in Postmenopausal Women Exposed In Utero to Diethylstilbestrol. Cancer Epidemiol Biomarkers Prev. 27, 1208–1213. [DOI] [PMC free article] [PubMed] [Google Scholar]
  48. Vijayaragavan K, et al. , 2009. Noncanonical Wnt signaling orchestrates early developmental events toward hematopoietic cell fate from human embryonic stem cells. Cell Stem Cell. 4, 248–62. [DOI] [PMC free article] [PubMed] [Google Scholar]
  49. Xu Z, et al. , 2017. RELIC: a novel dye-bias correction method for Illumina Methylation BeadChip. BMC Genomics. 18, 4. [DOI] [PMC free article] [PubMed] [Google Scholar]
  50. Xu Z, et al. , 2016. ENmix: a novel background correction method for Illumina HumanMethylation450 BeadChip. Nucleic Acids Res. 44, e20. [DOI] [PMC free article] [PubMed] [Google Scholar]
  51. Xu Z, et al. , 2021a. The ENmix DNA methylation analysis pipeline for Illumina BeadChip and comparisons with seven other preprocessing pipelines. Clin Epigenetics. 13, 216. [DOI] [PMC free article] [PubMed] [Google Scholar]
  52. Xu Z, et al. , 2021b. ipDMR: identification of differentially methylated regions with interval P-values. Bioinformatics. 37, 711–713. [DOI] [PMC free article] [PubMed] [Google Scholar]
  53. Yamashita S, et al. , 2001. Effects of neonatal diethylstilbestrol exposure on c-fos and c-jun protooncogene expression in the mouse uterus. Histol Histopathol. 16, 131–40. [DOI] [PubMed] [Google Scholar]

Associated Data

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

Supplementary Materials

1

Supplemental Figure 1. Participant selection scheme of the NCI’s Combined DES Cohort Feasibility Study.

NIHMS1900925-supplement-1.jpg (30.2KB, jpg)
2

Supplemental Figure 2. Meta-analysis results for associations between prenatal diethylstilbestrol (DES) exposure and methylation levels at individual CpG sites.

Supplemental Figure 2 legend. A. Manhattan plot for the meta-analytic associations between DES and methylation levels at 419,919 CpG sites common to the NCI’s Combined DES Cohort Study and Sister Study Cohort in the 450K methylation array. Black horizontal line indicates the Bonferroni correction level. B. Volcano plot for the meta-analytic associations between DES and methylation levels at 419,919 CpG sites common to the NCI’s Combined DES Cohort Study and Sister Study Cohort in the 450K methylation array. Black horizontal line indicates the Bonferroni correction level.

NIHMS1900925-supplement-2.jpg (34KB, jpg)
3

Supplemental Figure 3. Associations between prenatal diethylstilbestrol (DES) exposure and methylation levels at individual CpG sites in the 850K methylation array that are not included in the 450K methylation array; NCI’s Combined DES Cohort Study.

Supplemental Figure 3 legend. A. Manhattan plot for the associations between DES and methylation levels at 393,050 CpG sites in the 850K methylation array, that are not included in the 450K methylation array, from the NCI’s Combined DES Cohort Feasibility Study. Black horizontal line indicates the Bonferroni correction level for the entire 850K array. B. Volcano plot for the meta-analytic associations between DES and methylation levels at 393,050 CpG sites in the 850K methylation array that are not included in the 450K methylation array, from the NCI’s Combined DES Cohort Feasibility Study. Black horizontal line indicates the Bonferroni correction level for the entire 850K array. C. Boxplot of the methylation levels at the most significant CpG site (cg06393376) in the analysis in women exposed in utero and unexposed to DES in the NCI’s Combined DES Cohort Feasibility Study. D. Boxplot of the methylation levels at the second most significant CpG site (cg00726784) in the analysis among women exposed in utero and unexposed to DES in the NCI’s Combined DES Cohort Feasibility Study.

NIHMS1900925-supplement-3.jpg (35.3KB, jpg)
4

Supplemental Table 1. Associations of in utero diethylstilbestrol (DES) exposure with blood DNA methylation at the 207 CpG sites common across the two arrays used in the Sister Study Cohort and the NCI’s Combined DES Cohort Study within the nine candidate genes (EMB, WNT11, TGFB1, ERBB2, EGFR, LTF, EGF, FOS, and JUN). Results from the individual studies and the fixed-effect meta-analysis.

NIHMS1900925-supplement-4.xlsx (54.2KB, xlsx)
5

Supplemental Table 2. Associations of in utero diethylstilbestrol (DES) exposure with blood DNA methylation at CpG sites in the 450K array common across the two arrays used in the Sister Study Cohort and the NCI’s Combined DES Cohort Study. Results from the individual studies and the fixed-effect meta-analysis presented for the 100 most significant CpG sites, according to the meta-analytic p-value.

NIHMS1900925-supplement-5.csv (22.6KB, csv)
6

Supplemental Table 3. Associations of in utero diethylstilbestrol (DES) exposure with blood DNA methylation at CpG sites in the EPIC array, not in the 450k array, in the NCI’s Combined DES Cohort Study. Results presented for the 100 most significant CpG sites.

NIHMS1900925-supplement-6.csv (10.6KB, csv)

RESOURCES