Abstract
Menopause is a significant milestone in a woman’s life, characterized by decreasing estradiol (E2) and increasing follicle-stimulating hormone (FSH) levels. Growing evidence suggests that air pollution may affect reproductive health and disrupt hormone profiles, yet the associations in women undergoing menopausal transition (MT) remains underexplored. We examined the associations between annual air pollutant exposures and repeated measures of E2 and FSH in 1,365 women with known final menstrual period (FMP) date from the Study of Women’s Health Across the Nation. Air pollution was calculated as the annual averages of 24-hour average PM2.5 levels, daily one-hour maximum NO2 levels, and daily 8-hour maximum O3 levels. Linear mixed models with piece-wise linear splines were used to model non-linear trajectories of E2 and FSH in three distinct time periods: up to 2 years before the FMP (early MT), within 2 years before and 2 years after FMP (transmenopause), and more than 2 years post-FMP (postmenopause). In the transmenopausal period, an interquartile (5 µg/m3) increase in PM2.5 was associated with a significant decrease in E2 levels (−15.7%, 95% CI: −23.7, −6.8), and a 10 ppb increase in NO2 was associated with a significant decrease in E2 levels (−9.2%, 95% CI: −16.2, −1.7). A higher PM2.5 was also associated with an accelerated rate of decline in E2. Regarding FSH, a 10 ppb increase in NO2 was associated with decline in FSH levels (−11.7%, 95% CI: −21.8, −0.1) in the early MT and accelerated rates of decline in the postmenopause (−1.1% per year, 95% CI: −2.1, −0.1). Additionally, inverse associations between O3 and FSH were observed in the transmenopause and postmenopause. Our study suggests that increases in PM2.5, NO2, and O3 exposures are linked to significant declines in E2 and FSH levels across menopausal stages, suggesting the detrimental impact of air pollutants on women’s reproductive hormones.
Keywords: Estradiol, Follicle-stimulating hormone, Air pollution, PM2.5, Women
1. Introduction
Menopause is a critical life event for women as it marks the end of reproductive capacity and a dramatic change in ovarian hormone secretion. It is characterized by a sharp decline in estradiol (E2) and a gradual rise in follicle-stimulating hormone levels (FSH), resulting in permanent changes in ovarian function (Randolph et al., 2011). These hormone changes have significant implications for women’s health in mid- and later-life, particularly cardiometabolic disorders (El Khoudary et al., 2019) and bone loss (Karlamangla et al., 2021). Therefore, gaining a comprehensive understanding of factors associated with changes in hormone levels during the menopausal transition (MT), including sociodemographic, lifestyle, and environmental factors (Randolph et al., 2011; Wang et al., 2022a; Ding et al., 2023) is crucial to promoting healthy aging and preventing adverse chronic health outcomes in women.
Air pollution, including fine particulate matter (PM2.5), nitrogen dioxide (NO2), and ozone (O3), is a widespread environmental exposure that has been associated with a range of adverse health outcomes including cardiovascular (Franklin et al., 2015), respiratory (Laumbach and Kipen, 2012), and cognitive disorders (Clifford et al., 2016). Biologic evidence is accumulating that suggests that exposure to PM2.5, NO2, and O3 may also affect reproductive health. For example, exposure to NO2 is linked to disrupted estrus cycles, reduced fertility, and elevated post-implantation loss rates (Veras et al., 2009). Exposure to PM2.5 and O3 may directly cause ovarian damage by induced oxidative stress and inflammation (Ogliari et al., 2013; Shi et al., 2016; Gai et al., 2017; Qiu et al., 2018), leading to disruption of hormone profiles. A growing number of epidemiologic studies have also reported associations between air pollutant exposure and adverse reproductive outcomes including premature birth and pre-eclampsia (Pedersen et al., 2017; Arroyo et al., 2019). However, most studies of air pollution’s impact on reproductive hormones have been cross-sectional and limited to either men (Radwan et al., 2016; Zheng et al., 2022) or women undergoing in vitro fertilization (Wang et al., 2019; Shi et al., 2021b; Wu et al., 2021, 2022). To our knowledge, the associations between air pollution and sex hormone levels have not been explored in midlife women during the MT, particularly in a longitudinal study with repeated measures of hormone levels across the MT.
To inform the hypothesis that air pollution exposure may have an adverse effect on sex hormone levels during this critical midlife stage, we examined the associations of the annual levels of PM2.5, NO2, and O3 with repeated measures of E2 and FSH before and after natural menopause using data from the Study of Women’s Health Across the Nation (SWAN). We hypothesized that higher levels of exposure to these pollutants would be associated with lower levels of E2 and higher levels of FSH, as well as accelerated rates of changes in these hormones.
2. Material and Methods
2.1. Study population
SWAN is an ongoing longitudinal study that aims to investigate the physical and psychological changes experienced by midlife women during the MT and their potential long-term health effects. The study began in 1996, with a total of 3,302 women between the ages of 42 to 52 years who met the eligibility criteria, which included having an intact uterus and at least one ovary, at least one menstrual cycle in the previous three months, and no use of menopausal hormone therapy. Recruitment took place across seven sites in the United States, with each site enrolling White women and a specific racial/ethnic group, including: Black women in Boston, MA, Pittsburgh, PA, southeast Michigan, and Chicago, IL; Hispanic women in Newark, NJ; Chinese women in Oakland, CA; and Japanese women in Los Angeles, CA.
Air pollution data were available for 2,452 participants from six out of seven study sites between 2000 and 2008, as the Boston site lacked residential history necessary for exposure data. Furthermore, we restricted our analyses to 1,436 participants with a known final menstrual period (FMP) date to model hormone changes in relation to the number of years before and after the FMP. Women who had a documented date of the FMP experienced a natural menopausal transition without hysterectomy and bilateral oophorectomy, and without hormone therapy. This approach allowed us to demonstrate the effects of air pollution exposures on hormone changes during the MT more precisely and support its effect on ovarian aging if associations were observed (Greendale et al., 2019). We further excluded 71 participants who lacked key covariate information, resulting in a final sample of 1,365 women with 7,600 observations (Figure S1). In a comparison detailed in Table S1, participants without a known FMP date generally shared similar baseline characteristics with those included. The study protocol was approved by the institutional review board at each site, and all participants provided written informed consent during each study visit.
2.2. Air pollution exposure
Each study participant was assigned the annual averages of 24-hour average PM2.5 level, daily one-hour maximum NO2 level, and daily 8-hour maximum O3 level. These estimates were obtained at a high spatial resolution of 1 km2 based on the participants’ residential addresses between 2000 and 2008 derived from the ensemble of three machine learning models, including neural network, random forest, and gradient boosting over 100 predictors. The ensemble demonstrated high predictive performance, with an average 10-fold cross-validated coefficient of determination (R2) of 0.89 for PM2.5, 0.84 for NO2, and 0.86 for O3. More detailed information on the assessment of air pollution exposure can be found in other sources (Di et al., 2019, 2020; Requia et al., 2020; Wang et al., 2022b). Spearman correlations between PM2.5, NO2, and O3 were calculated.
2.3. Sex hormones
Serum samples were obtained from all participants at each study visit to analyze E2 and FSH levels. Samples were collected from fasting participants before 10:00 A.M. between days 2–5 of a spontaneous menstrual cycle occurring within 60 days of recruitment at the baseline visit and approximately once per year thereafter. If such a sample was not attainable, a random fasting sample was obtained within 90 days of the anniversary of the baseline visit. E2 was assayed in duplicate using a modified, offline ACS-180 (E2-6) immunoassay with inter- and intra-assay coefficients of variation of 11% and 6%, respectively. FSH was assayed in singlet using a two-site chemiluminometric immunoassay on the automated Ciba Corning Diagnostics ACS-180 analyzer (Bayer Diagnostics Corp., Norwood, MA) with inter- and intra-assay CVs of 12% and 6%, respectively.
2.4. Covariates
Information on demographic characteristics such as age, self-defined race/ethnicity (White, Black, Chinese, Japanese, or Hispanic) and education (high school or less, some college, or college degree or higher) was collected using a self-administered questionnaire. Smoking status (never smoker, former smoker, or current smoker), exposure to environmental tobacco smoke (0, 1–4 person-hours/week, or ≥5 person-hours/week), physical activity, and cycle day were assessed through standardized interviews. Physical activity from the past year was assessed using an adapted Kaiser Physical Activity Survey (Sternfeld et al., 2000). This survey comprises 38 questions, primarily with Likert-scale answers, addressing activity across 3 domains including sports/exercise, household chores/caregiving, and daily tasks. Averages of the ordinal answers in each area were taken to create domain-specific scores, which ranged from 1 to 5. As a result, the overall activity score varied between 3 and 15, where a score of 15 indicates the most intense activity level. The physical activity score was not measured at two follow-up visits, so we imputed the value by computing the mean score for the visits preceding and following the unmeasured visit. The total caloric intake (kCal/day) was quantified using data from a modified version of the Block Food Frequency Questionnaire (Block et al., 1986).
2.5. Statistical analysis
We employed a three-step method to model the trajectories of hormone levels in relation to air pollution. First, we used generalized additive mixed models with penalized splines to model the trajectories of E2 and FSH in relation to the years before and after FMP. Both E2 and FSH displayed piece-wise linear trajectories with three segments (Figure S2). Next, we used linear mixed models with piece-wise linear splines to identify the optimal knot placements for the hormone trajectories based on Bayesian Information Criterion values. We varied the knot placements in six-month intervals from 15 years before the FMP to 10.5 years after the FMP and selected the knot positions that resulted in the lowest BIC values, which were located at FMP minus 2 years and FMP plus 2 years for both E2 and FSH, in line with our smoothing plot (Figure S2) and prior research findings in SWAN (Randolph et al., 2011). In the final step, we used piece-wise linear mixed-effect models to evaluate the associations of air pollution levels with hormone levels and their rates of change across time periods spanning the three segments- up to 2 years before the FMP (early MT), within 2 years before and 2 years after FMP (transmenopause), and more than 2 years post-FMP (postmenopause). Interaction terms between air pollution levels and time before and after FMP were included to estimate the change rates. Both primary terms and interaction terms related to time were included. To account for the right-skewed distributions of hormone levels, we log-transformed the values and then back-transformed the effect estimates to interpret the results as percent changes in outcomes associated with an interquartile range (IQR) increase in levels of each air pollutant. The covariates included in the final models were selected based on a priori knowledge from the literature and included the non–time-varying variables of age at FMP, race/ethnicity, study site, education, and time-varying variables included smoking status, environmental tobacco smoke exposure, body mass index (BMI), cycle day, and physical activity, and total calorie intake. Additionally, we included random intercepts to account for intra-participant correlations of repeated hormone measures.
In the sensitivity analysis, we investigated the potential impact of exposure to O3 during warm seasons. Specifically, we calculated the average daily maximum 8-hour levels of O3 from May to October and examined their associations with E2 and FSH (Shi et al., 2021a). Additionally, we evaluated the associations among never and former smokers given smoking is considered an important source of air pollution exposure. All analyses were conducted using R, version 4.2.2 (www.R-project.org). We employed the “lcmm” package for generalized additive mixed models and the “nlme” package for linear mixed-effect models.
3. Results
3.1. Participant characteristics
Participant characteristics at the exposure assessment baseline (1999–2000) included White (42.6%), Black (25.4%), Chinese (12.5%), Japanese (14.5%), and Hispanic (5.0%) women (Table 1). Participants were from six different locations across the United States: Southeast Michigan (19.4%), Chicago (15.5%), Oakland (20.5%), Los Angeles (22.2%), Newark (7.2%), and Pittsburgh (15.2%). Overall, median exposure levels at exposure assessment baseline were 15.6 µg/m3 for PM2.5, 38.5 ppb for NO2, and 32.0 ppb O3 at baseline. PM2.5, NO2, and O3 were positively correlated, with the strongest correlation observed between PM2.5 and NO2 (r = 0.72) (Figure S3).
Table 1.
Characteristics at exposure assessment baseline, 1999–2000.
| Characteristics | Median (IQR) or N (%) |
|---|---|
| Air pollution | |
| PM2.5, µg/m3 | 15.6 (14.2, 16.7) |
| NO2, ppb | 38.5 (34.3, 44.7) |
| O3, ppb | 32.0 (29.2, 34.2) |
| Sex hormone | |
| E2, pg/mL | 33.9 (19.4, 75.2) |
| FSH, mIU/mL | 27.5 (14.4, 64.5) |
| Covariate | |
| Age at FMP, year | 52.3 (50.4, 54.0) |
| BMI, kg/m2 | 26.9 (22.9, 32.6) |
| Site | |
| Southeast MI | 264 (19.4%) |
| Chicago, IL | 212 (15.5%) |
| Oakland, CA | 280 (20.5%) |
| Los Angeles, CA | 303 (22.2%) |
| Newark, NJ | 98 (7.2%) |
| Pittsburgh, PA | 208 (15.2%) |
| Race/ethnicity | |
| White | 582 (42.6%) |
| Black | 346 (25.4%) |
| Chinese | 170 (12.5%) |
| Japanese | 199 (14.5%) |
| Hispanic | 68 (5.0%) |
| Education | |
| Less than high school | 326 (23.9%) |
| High school | 442 (32.4%) |
| Some college | 290 (21.2%) |
| College degree or higher | 307 (22.5%) |
| Smoking status | |
| Never | 828 (60.7%) |
| Former | 343 (25.1%) |
| Current | 194 (14.2%) |
| Environmental tobacco smoke exposure | |
| None | 768 (56.3%) |
| 1–4 person-hours/week | 314 (23.0%) |
| ≥5 person-hours/week | 283 (20.7%) |
| Cycle day | 1 (1, 2) |
| Physical activity score | 7.6 (6.4, 8.8) |
| Total calorie intake, kcal | 1715 (1340, 2237) |
| Time to FMP, year | -2.4 (−4.9, −0.1) |
Abbreviations: E2: estradiol; FSH: follicle-stimulating hormone; FMP: final menstrual period; BMI: body mass index.
The participants’ median age at FMP during follow-up was 52.3 years, and most women were premenopausal at exposure baseline (median time to FMP of −2.4 years). Median levels of E2 and FSH at exposure baseline were 33.9 pg/mL and 27.5 mIU/mL, respectively. A majority (60.7%) of participants were never smokers and 56.3% reported no exposure to environmental tobacco smoke. These participants had a median BMI of 26.9 kg/m2 and a median physical activity score of 7.6, pointing to a moderate intensity of physical activity at exposure baseline. The median total calorie intake was 1715 kcal/day.
3.2. Air pollution and trajectory of E2
We examined the associations of PM2.5, NO2, and O3 levels with changes in E2 levels and categorized the examined associations into three distinct time periods: early MT (more than 2 years before the FMP), transmenopause (within 2 years before and after FMP), and postmenopause (more than 2 years post-FMP). Our analyses demonstrated that exposure to these pollutants was statistically significantly associated with longitudinal changes in E2, though the extent and direction of these changes varied (Table 2).
Table 2.
Adjusted percent change in E2 associated with PM2.5, NO2 and O3 before and after FMP.
| Before the MT >2 years before FMP (Early MT) |
During the MT 2 years before and after FMP (Transmenopause) |
After the MT >2 years after FMP (Postmenopause) |
|
|---|---|---|---|
| Model for PM2.5 | Percent change in E2, % | Percent change in E2, % | Percent change in E2, % |
| PM2.5, per 5 µg/m3 increase | −2.4 (−18.2, 16.4) P=0.79 |
−15.7 (−23.7, −6.8) P=0.008 |
−7.3 (−16.0, 2.2) P=0.13 |
| PM2.5 × Time to FMP | −1.0 (−3.4, 1.6) P=0.46 |
−5.3 (−9.2, −1.2) P=0.01 |
−0.5 (−2.5, 1.5) P=0.61 |
| Model for NO2 | Percent change in E2, % | Percent change in E2, % | Percent change in E2, % |
| NO2, per 10 ppb increase | −5.5 (−19.2, 10.6) P=0.48 |
−9.2 (−16.2, −1.7) P=0.02 |
−6.9 (−14.8, 1.6) P=0.11 |
| NO2 × Time to FMP | −1.3 (−3.8, 1.2) P=0.31 |
−3.5 (−7.2, 0.4) P=0.08 |
0.4 (−1.3, 2.2) P=0.63 |
| Model for O3 | Percent change in E2, % | Percent change in E2, % | Percent change in E2, % |
| O3, per 5 ppb increase | −1.0 (−11.8, 11.0) P=0.86 |
6.6 (0.4, 13.1) P=0.04 |
1.9 (−4.9, 9.3) P=0.59 |
| O3 × Time to FMP | −0.6 (−2.6, 1.4) P=0.55 |
−1.7 (−4.6, 1.3) P=0.26 |
−0.5 (−2.0, 1.0) P=0.48 |
Abbreviations: E2: estradiol; FMP: final menstrual period; BMI: body mass index.
Models were adjusted for age at FMP, race/ethnicity, study site, education, smoking status, secondhand smoking, cycle day of blood draw, body mass index, physical activity, and total calorie intake.
For PM2.5, a non-significant association with E2 levels was found in the early MT (−2.4%, 95% CI: −18.2, 16.4); but in the transmenopause, a significant decrease in E2 levels was observed with each 5 µg/m3 increase in PM2.5 (−15.7%, 95% CI: −23.7, −6.8). This association appeared to diminish in the postmenopause (−7.3%, 95% −16.0, 2.2). In addition, each increase of PM2.5 level by 5 µg/m3 was associated with an accelerated reduction in E2 levels in the transmenopause. Specifically, we observed a −5.3% change (95% CI: −9.2, −1.2) in E2 per year, indicating a potential intensifying effect of PM2.5 on E2 decline in the transmenopause.
A similar pattern was observed for NO2, with non-significant associations with E2 levels in the early MT (−5.5%, 95% CI: −19.2, 10.6) and postmenopause (−6.9%, 95% CI: −14.8, 1.6). However, in the transmenopause, a significant decrease in E2 levels was found with each 10 ppb increase in NO2 (−9.2%, 95% CI: −16.2, −1.7). No significant association between NO2 and rates of change in E2 over time were observed.
In contrast, O3 exposure showed a different relationship with E2. No significant changes in E2 levels were observed in association with O3 in the early MT (−1.0%, 95% CI: −11.8, 11.0) or postmenopause (1.9%, 95% CI: −4.9, 9.3). In the transmenopause, each 5 ppb increase in O3 was associated with a significant increase in E2 levels (6.6%, 95% CI: 0.4, 13.1).
3.3. Air pollution and trajectory of FSH
As with E2, the magnitude and direction of these associations PM2.5, NO2, and O3 levels with changes in FSH levels were variable across the different periods (Table 3).
Table 3.
Adjusted percent change in FSH associated with PM2.5, NO2 and O3 before and after FMP.
| Before the MT >2 years before FMP (Early MT) |
During the MT 2 years before and after FMP (Transmenopause) |
After the MT >2 years after FMP (Postmenopause) |
|
|---|---|---|---|
| Model for PM2.5 | Percent change in FSH, % | Percent change in FSH, % | Percent change in FSH, % |
| PM2.5, per 5 µg/m3 increase | −10.3 (−21.8, 3.0) P=0.12 |
−6.5 (−13.3, 0.8) P=0.08 |
−3.3 (−8.9, 2.5) P=0.26 |
| PM2.5 × Time to FMP | −0.9 (−2.8, 1.1) P=0.38 |
2.8 (−0.2, 5.8) P=0.07 |
−0.4 (−1.6, 0.7) P=0.48 |
| Model for NO2 | Percent change in FSH, % | Percent change in FSH, % | Percent change in FSH, % |
| NO2, per 10 ppb increase | −11.7 (−21.8, −0.1) P=0.04 |
−3.2 (−8.8, 2.8) P=0.29 |
−4.9 (−10.0, 0.5) P=0.07 |
| NO2 × Time to FMP | −1.1 (−3.0, 0.9) P=0.27 |
0.8 (−1.9, 3.7) P=0.56 |
−1.1 (−2.1, −0.1) P=0.03 |
| Model for O3 | Percent change in FSH, % | Percent change in FSH, % | Percent change in FSH, % |
| O3, per 5 ppb increase | 4.4 (−4.5, 14.2) P=0.34 |
−5.6 (−9.8, −1.3) P=0.01 |
−5.1 (−8.9, −1.1) P=0.01 |
| O3 × Time to FMP | 1.4 (−0.2, 3.0) P=0.09 |
1.9 (−0.2, 4.0) P=0.08 |
1.1 (0.2, 1.9) P=0.02 |
Abbreviations: FSH: follicle-stimulating hormone; FMP: final menstrual period; BMI: body mass index.
Models were adjusted for age at FMP, race/ethnicity, study site, education, smoking status, secondhand smoking, cycle day of blood draw, body mass index, physical activity, and total calorie intake.
For PM2.5, no associations with FSH were observed in the early MT (−10.3%, 95% CI: −21.8, 3.0), transmenopause (−6.5%, 95% CI: −13.3, 0.8), or postmenopause (−3.3%, 95% CI: −8.9, 2.5). In addition, no significant results were found for rates of changes in FSH over time.
NO2 exhibited a different pattern. While non-significant associations with FSH levels were observed in the transmenpause (−3.2%, 95% CI: −8.8, 2.8) and postmenopause (−4.9%, 95% CI: −10.0, 0.5), a significant decrease in FSH levels was found in the early MT with each 10 ppb increase in NO2 (−11.7%, 95% CI: −21.8, −0.1). In addition, NO2 was related to an accelerated decline in FSH in the postmenopause. Specifically, we observed a −1.1% change (95% CI: −2.1, −0.1) in FSH per year with each 10 ppb increase in NO2.
In contrast, O3 was not associated with FSH levels in the early MT (4.4%, 95% CI: −4.5, 14.2). However, in the transmenopause and postmenopause, a significant decrease in FSH levels was observed with each 5 ppb increase in O3 (−5.6%, 95% CI: −9.8, −1.3 and −5.1%, 95% CI: −8.9, −1.1, respectively). Additionally, we observed higher rates of increase in FSH (1.1% per year, 95% CI: 0.2, 1.9) with each 5 ppb increase in O3 in the postmenopause.
3.4. Sensitivity analyses
In the sensitivity analysis of warm-season O3, we found associations consistent with our primary findings (Table S2). In the sensitivity analysis in the subpopulation of never and former smokers, similar associations between air pollution and hormone levels were also observed (Table S3 and Table S4).
4. Discussion
In this prospective multi-ethnic cohort study of 1,365 midlife women, exposure to PM2.5 and NO2 pollutants was associated with decreased E2 levels in the transmenopausal period, with PM2.5 also linked to an accelerated rate of E2 decline during this period. For FSH levels, exposure to NO2 was associated with decreased levels in the early MT and an accelerated decline in the postmenpause, and O3 was associated with a significant decrease in FSH during the transmenopause and postmenopause, along with acceleration of FSH increase in the postmenopause.
The observed associations between air pollution exposure and hormonal changes in midlife women are biologically plausible. Gai et al. found that PM2.5 exposure was associated with increased levels of pro-inflammatory markers, including interleukin-6 and tumor necrosis factor-alpha, and oxidative stress marker 8-hydroxy-2’-deoxyguanosine in the ovaries of female mice, causing apoptosis-related protein expressions and ovarian damage (Gai et al., 2017). Ogliari et al. also found that exposure to PM2.5 consistent with average daily levels recommended by the World Health Organization, significantly reduced the proportion of primordial follicles during intrauterine and postnatal periods in mice (Ogliari et al., 2013). Similarly, Veras et al. observed that exposure to non-filtered air that included PM2.5 and NO2 led to changes in prolonged estrus cycles, a reduced number of cycles, and a significant decrease in the number of antral follicles in mice, which further caused reduced fertility and pregnancy indices, and an increase in post-implantation loss rates (Veras et al., 2009). While these animal studies do not specifically focus on the MT, they provided evidence of potential adverse effects of air pollution on ovarian function and thus on reproductive hormones.
Previous epidemiologic studies exploring the impact of air pollution on reproductive health have focused on women of reproductive age. Shi et al., 2021 found that exposure to NO2 was associated with a 14% lower rate of biochemical pregnancy, and that PM10 exposure was associated with a 12% decrease in live birth rate; whereas the associations of PM2.5 and O3 exposure with pregnancy outcomes were not statistically significant (Shi et al., 2021b). Wang et al. 2019 examined the effects of air pollution on fresh and frozen-thawed embryo transfer outcomes in in vitro fertilization (IVF), and found O3 levels to be significantly linked with lower live birth rates in frozen-thawed embryo transfer cycles; whereas PM2.5 and NO2 showed no significant associations with IVF outcomes (Wang et al., 2019). In Wu et al. 2022, exposure to PM2.5 was associated with a higher risk of poor ovarian response (< 4 oocytes retrieved) with the adjusted OR of 1.54 (1.10, 2.14) comparing the highest to the lowest quartile of exposure (p for trend <0.001) in women who received IVF treatment (Wu et al., 2022). Furthermore, Wu et al. 2021 found that in both fresh and frozen-thawed embryo transfer cycles, exposure to the highest quartile of NO2 was associated with higher risks of ectopic pregnancy (adjusted OR = 1.92 for fresh cycles and 2.76 for frozen-thawed embryo transfer cycles), compared to the lowest quartile; and a similar inverse association with PM2.5 was also reported (Wu et al., 2021). Our findings, which specifically focus on midlife women during the MT, provide unique insights into the associations between air pollution exposure and hormonal changes in this specific population, extending beyond previous studies that primarily examined women of reproductive age.
In contrast to PM2.5 and NO2, the relationship between O3 exposure and hormonal changes in midlife women during the menopausal transition demonstrated different results in our study. Furthermore, analysis of warm-season O3 exposure did not alter the findings. The contrasting results observed for O3 compared to PM2.5 and NO2 may be attributed to several factors, including the negative correlation between O3 and PM2.5/NO2 levels, variations in exposure patterns, and the presence of confounding factors by co-pollutants not accounted for in the analysis. Further research is needed to elucidate the precise reasons for the divergent associations between O3 and the examined hormonal changes in midlife women.
Our study drawn from a multi-racial/ethnic cohort, drawn from a broad geographical and spectrum across the United States, offered a unique opportunity to evaluate the effects of varied air pollution exposure on sex hormone levels in midlife women. These findings collectively shed light on the complex relationship between air pollution and hormonal changes in women during this critical life stage. The use of reproductive age, including periods before and after the MT, provides additional strength to our findings by capturing the effects of air pollution on hormonal changes at critical periods around the FMP. However, it is important to acknowledge the limitations of our study. The findings may not be generalizable to the broader population of women in the United States or other racial/ethnic groups, as our study sample only included women with an observed FMP or natural menopause from SWAN’s community-based sample. Further research is needed to explore the effects of air pollution on hormonal changes in women from more diverse populations. Additionally, despite adjusting for several known confounders, we were unable to eliminate potential residual confounding, such as temperature and humidity, due to the lack of available data. Furthermore, the possibility of unknown residual confounding factors cannot be entirely ruled out. Factors such as co-exposure to endocrine-disrupting chemicals, which were not included in our analysis, may have influenced the observed associations, and should be considered in future studies. Finally, our study did not analyze the specific chemical composition of PM2.5 due to data unavailability. Future studies that include this detailed information will improve our understanding of the associations between various particulate components and hormone levels.
5. Conclusions
The study’s findings suggest that exposure to air pollution during the MT may contribute to hormonal changes and possibly imbalances in midlife women by significantly decreasing their levels of E2. This decrease and accelerated rate of E2 decline, primarily driven by PM2.5, could potentially intensify menopausal symptoms, such as hot flashes, night sweats, mood changes, sleep issues, and memory problems. Furthermore, NO2 exposure was also associated with lower FSH levels in the early MT and with moderate declines postmenopausal, which may have implications for subsequent bone health, and cardiovascular risks. Therefore, these findings underscore the potential public health importance of air pollution mitigation for improving women’s health during and after midlife.
Supplementary Material
Acknowledgements
The Study of Women’s Health Across the Nation (SWAN) has grant support from NIH, DHHS, NIA, the National Institute of Nursing Research (NINR) and the NIH Office of Research on Women’s Health (ORWH) (Grants U01NR004061; U01AG012505, U01AG012535, U01AG012531, U01AG012539, U01AG012546, U01AG012553, U01AG012554, U01AG012495, and U19AG063720). This study was supported by the Michigan Lifestage Environmental Exposures and Disease (M-LEEaD) NIEHS P30 Core Center (P30ES017885).The study was also supported by the SWAN Repository (U01AG017719). The content of this manuscript is solely the responsibility of the authors and does not necessarily represent the official views of the NIA, NINR, ORWH or the NIH.
Clinical Centers: University of Michigan, Ann Arbor – Carrie Karvonen-Gutierrez, PI 2021 – present, Siobán Harlow, PI 2011 – 2021, MaryFran Sowers, PI 1994–2011; Massachusetts General Hospital, Boston, MA – Sherri‐Ann Burnett‐Bowie, PI 2020 – Present; Joel Finkelstein, PI 1999 – 2020; Robert Neer, PI 1994 – 1999; Rush University, Rush University Medical Center, Chicago, IL – Imke Janssen, PI 2020 – Present; Howard Kravitz, PI 2009 – 2020; Lynda Powell, PI 1994 – 2009; University of California, Davis/Kaiser – Elaine Waetjen and Monique Hedderson, PIs 2020 – Present; Ellen Gold, PI 1994 – 2020; University of California, Los Angeles – Arun Karlamangla, PI 2020 – Present; Gail Greendale, PI 1994 – 2020; Albert Einstein College of Medicine, Bronx, NY – Carol Derby, PI 2011 – present, Rachel Wildman, PI 2010 – 2011; Nanette Santoro, PI 2004 – 2010; University of Medicine and Dentistry – New Jersey Medical School, Newark – Gerson Weiss, PI 1994 – 2004; and the University of Pittsburgh, Pittsburgh, PA – Rebecca Thurston, PI 2020 – Present; Karen Matthews, PI 1994 – 2020.
NIH Program Office: National Institute on Aging, Bethesda, MD – Rosaly Correa-de-Araujo 2020 - present; Chhanda Dutta 2016- present; Winifred Rossi 2012–2016; Sherry Sherman 1994 – 2012; Marcia Ory 1994 – 2001; National Institute of Nursing Research, Bethesda, MD – Program Officers.
Central Laboratory: University of Michigan, Ann Arbor – Daniel McConnell (Central Ligand Assay Satellite Services).
NIA Biorepository: Rosaly Correa-de-Araujo 2019 – Present; SWAN Repository: University of Michigan, Ann Arbor – Siobán Harlow 2013 – 2018; Dan McConnell 2011 – 2013; MaryFran Sowers 2000 – 2011.
Coordinating Center: University of Pittsburgh, Pittsburgh, PA – Maria Mori Brooks, PI 2012 - present; Kim Sutton-Tyrrell, PI 2001 – 2012; New England Research Institutes, Watertown, MA - Sonja McKinlay, PI 1995 – 2001.
Steering Committee: Susan Johnson, Current Chair
Chris Gallagher, Former Chair
We thank the study staff at each site and all the women who participated in SWAN.
Footnotes
Conflict of Interest
The authors declare they have no actual or potential competing interest.
References
- Arroyo V, Linares C, Díaz J. Premature births in Spain: Measuring the impact of air pollution using time series analyses. Science of The Total Environment 2019;660:105–114. [DOI] [PubMed] [Google Scholar]
- Block G, Hartman AM, Dresser CM, Carroll MD, Gannon J, Gardner L. A data-based approach to diet questionnaire design and testing. American Journal of Epidemiology 1986;124:453–469. [DOI] [PubMed] [Google Scholar]
- Clifford A, Lang L, Chen R, Anstey KJ, Seaton A. Exposure to air pollution and cognitive functioning across the life course – A systematic literature review. Environmental Research 2016;147:383–398. [DOI] [PubMed] [Google Scholar]
- Di Q, Amini H, Shi L, Kloog I, Silvern R, Kelly J, Sabath MB, Choirat C, Koutrakis P, Lyapustin A, et al. An ensemble-based model of PM2.5 concentration across the contiguous United States with high spatiotemporal resolution. Environment International 2019;130:104909. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Di Q, Amini H, Shi L, Kloog I, Silvern R, Kelly J, Sabath MB, Choirat C, Koutrakis P, Lyapustin A, et al. Assessing no2 concentration and model uncertainty with high spatiotemporal resolution across the contiguous united states using ensemble model averaging. Environmental Science and Technology 2020;54:1372–1384. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Ding N, Zheutlin E, Harlow SD, Randolph JF, Mukherjee B, Park SK. Associations Between Repeated Measures of Urinary Phthalate Metabolites With Hormones and Timing of Natural Menopause. Journal of the Endocrine Society 2023;7:1–11. [DOI] [PMC free article] [PubMed] [Google Scholar]
- El Khoudary SR, Greendale G, Crawford SL, Avis NE, Brooks MM, Thurston RC, Karvonen-Gutierrez C, Waetjen LE, Matthews K. The menopause transition and women’s health at midlife: A progress report from the Study of Women’s Health across the Nation (SWAN). Menopause 2019;26:1213–1227. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Franklin BA, Brook R, Arden Pope C. Air Pollution and Cardiovascular Disease. Current Problems in Cardiology 2015;40:207–238. [DOI] [PubMed] [Google Scholar]
- Gai HF, An JX, Qian XY, Wei YJ, Williams JP, Gao GL. Ovarian Damages Produced by Aerosolized Fine Particulate Matter (PM2.5) Pollution in Mice: Possible Protective Medications and Mechanisms. Chinese Medical Journal 2017;130:1400. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Greendale GA, Sternfeld B, Huang MH, Han W, Karvonen-Gutierrez C, Ruppert K, Cauley JA, Finkelstein JS, Jiang SF, Karlamangla AS. Changes in body composition and weight during the menopause transition. JCI Insight 2019;4: e124865. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Karlamangla AS, Shieh A, Greendale GA. Chapter Fifteen - Hormones and bone loss across the menopause transition. In Litwack G, editor. Vitamins and Hormones 2021;115:, p. 401–417. [DOI] [PubMed] [Google Scholar]
- Laumbach RJ, Kipen HM. Respiratory health effects of air pollution: Update on biomass smoke and traffic pollution. Journal of Allergy and Clinical Immunology 2012;129:3–11. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Ogliari KS, De Faria Coimbra Lichtenfels AJ, De Marchi MRR, Ferreira AT, Dolhnikoff M, Saldiva PHN. Intrauterine exposure to diesel exhaust diminishes adult ovarian reserve. Fertility and Sterility 2013;99:1681–1688.e2. [DOI] [PubMed] [Google Scholar]
- Pedersen M, Halldorsson TI, Olsen SF, Hjortebjerg D, Ketzel M, Grandström C, Raaschou-Nielsen O, Sørensen M. Impact of Road Traffic Pollution on Pre-eclampsia and Pregnancy-induced Hypertensive Disorders. Epidemiology 2017;28:99. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Qiu L, Chen M, Wang X, Qin X, Chen S, Qian Y, Liu Z, Cao Q, Ying Z. Exposure to Concentrated Ambient PM2.5 Compromises Spermatogenesis in a Mouse Model: Role of Suppression of Hypothalamus-Pituitary-Gonads Axis. Toxicological Sciences 2018;162:318–326. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Radwan M, Jurewicz J, Polańska K, Sobala W, Radwan P, Bochenek M, Hanke W. Exposure to ambient air pollution-does it affect semen quality and the level of reproductive hormones? Annals of Human Biology 2016;43:50–56. [DOI] [PubMed] [Google Scholar]
- Randolph JF, Zheng H, Sowers MR, Crandall C, Crawford S, Gold EB, Vuga M. Change in Follicle-Stimulating Hormone and Estradiol Across the Menopausal Transition: Effect of Age at the Final Menstrual Period. The Journal of Clinical Endocrinology & Metabolism 2011;96:746–754. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Requia WJ, Di Q, Silvern R, Kelly JT, Koutrakis P, Mickley LJ, Sulprizio MP, Amini H, Shi L, Schwartz J. An Ensemble Learning Approach for Estimating High Spatiotemporal Resolution of Ground-Level Ozone in the Contiguous United States. Environmental Science and Technology 2020;54:11037–11047. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Shi L, Steenland K, Li H, Liu P, Zhang Y, Lyles RH, Requia WJ, Ilango SD, Chang HH, Wingo T, et al. A national cohort study (2000–2018) of long-term air pollution exposure and incident dementia in older adults in the United States. Nature Communications 2021 12:1 2021a;12:1–9. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Shi L, Zhang J, Lai Z, Tian Y, Fang L, Wu M, Xiong J, Qin X, Luo A, Wang S. Long-Term Moderate Oxidative Stress Decreased Ovarian Reproductive Function by Reducing Follicle Quality and Progesterone Production. PLoS One 2016;11:e0162194. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Shi W, Sun C, Chen Q, Ye M, Niu J, Meng Z, Bukulmez O, Chen M, Teng X. Association between ambient air pollution and pregnancy outcomes in patients undergoing in vitro fertilization in Shanghai, China: A retrospective cohort study. Environment International 2021b;148:106377. [DOI] [PubMed] [Google Scholar]
- Sternfeld B, Cauley J, Harlow S, Liu G, Lee M. Assessment of Physical Activity with a Single Global Question in a Large, Multiethnic Sample of Midlife Women. American Journal of Epidemiology 2000;152:678–687. [DOI] [PubMed] [Google Scholar]
- Veras MM, Damaceno-Rodrigues NR, Guimarães Silva RM, Scoriza JN, Saldiva PHN, Caldini EG, Dolhnikoff M. Chronic exposure to fine particulate matter emitted by traffic affects reproductive and fetal outcomes in mice. Environmental Research 2009;109:536–543. [DOI] [PubMed] [Google Scholar]
- Wang X, Cai J, Liu L, Jiang X, Li P, Sha A, Ren J. Association between outdoor air pollution during in vitro culture and the outcomes of frozen–thawed embryo transfer. Human Reproduction 2019;34:441–451. [DOI] [PubMed] [Google Scholar]
- Wang X, Ding N, Harlow SD, Randolph JF, Mukherjee B, Gold EB, Park SK. Exposure to heavy metals and hormone levels in midlife women: The Study of Women’s Health Across the Nation (SWAN). Environmental Pollution 2022a;317:120740. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Wang X, Karvonen-Gutierrez CA, Gold EB, Derby C, Greendale G, Wu X, Schwartz J, Park SK. Longitudinal Associations of Air Pollution With Body Size and Composition in Midlife Women: The Study of Women’s Health Across the Nation. Diabetes care 2022b;45:2577–2584. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Wu S, Hao G, Zhang Y, Chen X, Ren H, Fan Y, Zhang Y, Bi X, Du C, Bai L, et al. Poor ovarian response is associated with air pollutants: A multicentre study in China. eBioMedicine 2022;81:104084. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Wu S, Zhang Y, Wu X, Hao G, Ren H, Qiu J, Zhang Y, Bi X, Yang A, Bai L, et al. Association between exposure to ambient air pollutants and the outcomes of in vitro fertilization treatment: A multicenter retrospective study. Environment International 2021;153:106544. [DOI] [PubMed] [Google Scholar]
- Zheng P, Chen Z, Shi J, Xue Y, Bai Y, Kang Y, Xu H, Jia G, Wang T. Association between ambient air pollution and blood sex hormones levels in men. Environmental Research 2022;211:113117. [DOI] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.