Skip to main content
NCBI home page
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2019 May 1.
Published in final edited form as: Environ Res. 2018 Feb 22;163:228–236. doi: 10.1016/j.envres.2018.02.004

Ambient air pollution and semen quality

Carrie J Nobles a, Enrique F Schisterman a, Sandie Ha a,b, Keewan Kim a, Sunni L Mumford a, Germaine M Buck Louis c, Zhen Chen d, Danping Liu d, Seth Sherman e, Pauline Mendola a
PMCID: PMC5878741  NIHMSID: NIHMS944325  PMID: 29459305

Abstract

Background

Ambient air pollution is associated with systemic increases in oxidative stress, to which sperm are particularly sensitive. Although decrements in semen quality represent a key mechanism for impaired fecundability, prior research has not established a clear association between air pollution and semen quality. To address this, we evaluated the association between ambient air pollution and semen quality among men with moderate air pollution exposure.

Methods

Of 501 couples in the LIFE study, 467 male partners provided one or more semen samples. Average residential exposure to criteria air pollutants and fine particle constituents in the 72 days before ejaculation was estimated using modified Community Multiscale Air Quality models. Generalized estimating equation models estimated the association between air pollutants and semen quality parameters (volume, count, percent hypo-osmotic swollen, motility, sperm head, morphology and sperm chromatin parameters). Models adjusted for age, body mass index, smoking and season.

Results

Most associations between air pollutants and semen parameters were small. However, associations were observed for an interquartile increase in fine particulates ≤2.5 microns and decreased sperm head size, including −0.22 (95% CI −0.34, −0.11) µm2 for area, −0.06 (95% CI −0.09, −0.03) µm for length and −0.09 (95% CI −0.19, −0.06) µm for perimeter. Fine particulates were also associated with 1.03 (95% CI 0.40, 1.66) greater percent sperm head with acrosome.

Conclusions

Air pollution exposure was not associated with semen quality, except for sperm head parameters. Moderate levels of ambient air pollution may not be a major contributor to semen quality.

Keywords: air pollution, semen quality, reproductive health

Introduction

Air pollution has been associated with a broad array of health effects, including cardiovascular disease morbidity and mortality (Langrish et al., 2012; Shah et al., 2015), even at the relatively moderate levels observed in the United States (Kaufman et al., 2016). Ambient air pollution exposure is associated with systemic increases in inflammation and oxidative stress (Chin, 2015), which may underlie prior reproductive health findings regarding the relation between air pollution and pregnancy loss and reduced fecundability (Checa Vizcaino et al., 2016). However, prior research has not determined whether reproductive effects of air pollution are due to male and/or female factors, as partners are typically exposed to similar levels of ambient air pollution. Semen quality is a sensitive marker for exposure to environmental pollution, as evidenced by associations with exposure to heavy metals, pesticides and phthalates (Gabrielsen and Tanrikut, 2016), and air pollution may affect male fecundability through adverse effects on semen quality.

While reactive oxygen species play a crucial role in spermatogenesis and fertilization, sperm are particularly sensitive to adverse impacts of oxidative stress (Du Plessis et al., 2015). Excesses in oxidative stress are associated with damage to sperm chromatin, peroxidation of sperm membranes, impaired motility and increases in apoptosis (Aitken et al., 2015; Du Plessis et al., 2015). Although systemic inflammation may directly lead to poorer semen quality through obstruction of sperm transport, impaired accessory gland functions and dysregulation of spermatogenesis, a key mechanism linking inflammation to poorer semen quality is the production of free radicals by leukocytes, contributing to oxidative stress (Azenabor et al., 2015). In a review of randomized trials of antioxidant supplementation among men with abnormal semen parameters, Ahmadi et al. noted a strong possibility of efficacy of supplementation in improving semen quality parameters (Ahmadi et al., 2016).

Prior epidemiologic research has reported inconsistent associations between air pollution and semen quality. A recent systematic review noted suggestive evidence of an association between air pollution and poorer sperm morphology, a weak association with DNA fragmentation and inconclusive associations with sperm motility and sperm count (Lafuente et al., 2016). Another meta-analysis found no statistically significant associations between air pollutants and semen quality parameters, although air pollutants appeared to be associated with a trend of decreased semen concentration and motility (Deng et al., 2016). Although several individual studies have reported impaired semen quality associated with particulate matter (Hammoud et al., 2010; Radwan et al., 2016; Wu et al., 2017; Zhou et al., 2014) and occupational exposure to vehicle exhaust (Calogero et al., 2011; Guven et al., 2008), the variation between studies in geographic region, study population and methods of assessment of both air pollution and semen quality parameters limits generalizability of findings.

More research is needed to determine whether air pollution adversely affects semen quality, and whether moderate levels of air pollution in countries such as the United States that fall below the World Health Organization air quality guidelines (World Health Organization, 2006) may be associated with decrements in semen quality. To date, studies in areas with moderate air pollution exposure have mostly utilized data from ambient air monitoring to assess exposure, which may lead to misclassification, and have only assessed the association of air pollution with a few semen quality endpoints (Hammoud et al., 2010; Hansen et al., 2010; Sokol et al., 2006). To address the need for additional research, we evaluated the association of residential exposure to ambient air pollutants, estimated using the Community Multiscale Air Quality (CMAQ) model, with 35 semen quality measures among men enrolled in a longitudinal time-to-pregnancy study exposed to moderate levels of air pollution. We hypothesized that greater exposure to ambient air pollutants would be associated with adverse effects on semen quality.

Materials and methods

The Longitudinal Investigation of Fertility and the Environment (LIFE) Study was conducted among 501 couples attempting pregnancy between 2005–2009 in Michigan (n=104) and Texas (n=397). Details of the study are fully described elsewhere (Buck Louis et al., 2011). Eligibility criteria include being married or in a committed relationship, being age 18 years or older, being able to communicate in English or Spanish and not having physician-diagnosed infertility. Couples were followed until pregnancy or up to one year of actively trying to become pregnant. This study was approved by the institutional review boards for all collaborating institutions, and couples provided written informed consent.

Criteria air pollutants and constituents of particulate matter

Ambient air pollution levels were estimated for each participants’ residence. Participants’ residential addresses were geocoded using ArcGIS software (Redlands, CA). Mean daily exposure to criteria air pollutants (sulfur dioxide [SO2], nitrogen oxides [NOX], nitrogen dioxide [NO2], carbon monoxide [CO], ozone [O3], particulate matter <10 microns [PM10] and fine particulate matter <2.5 microns [PM2.5]) and constituents of PM2.5 (elemental carbon [AEC], organic compounds [AOC], sulfate [ASO4], ammonium [ANH4] and nitrate [ANO3]) was estimated using modified Community Multiscale Air Quality (CMAQ) models and linked to residential address (Foley et al., 2010). The CMAQ models use the United States Environmental Protection Agency (US EPA) National Emissions Inventory and meteorological data generated by the Weather Research and Forecasting model to predict raw hourly estimates of pollution levels after adjusting for atmospheric photochemical properties of pollutants. To reduce measurement error, raw CMAQ estimates were fused with observed data from air monitors within the US EPA Air Quality System using inverse distance weighting. Performance of the modified CMAQ model has been previously reported (Chen et al., 2014).

Daily air pollution levels were calculated for all days in which the participant was enrolled in the study, and for 30 days prior to enrollment for couples attempting pregnancy for 0–1 months prior to enrollment (409, 81.8%) and for 60 days prior to enrollment for couples attempting pregnancy for two months prior to enrollment (91, 18.2%). Daily air pollution levels were then averaged across the 72 day window prior to ejaculation to evaluate chronic air pollution exposure during spermatogenesis, as well as for 0–14, 15–29, 30–44, 45–59 and 60–72 day windows prior to ejaculation to evaluate timing of exposure during spermatogenesis (Heller and Clermont, 1963). One couple’s address could not be geocoded, leaving 500 couples available for analyses.

Semen quality parameters

A total of 473 participants provided at least one semen sample, and 378 provided a second semen sample approximately one month later. An established at-home collection protocol was used (Buck Louis et al., 2014). Participants were asked to ejaculate into a glass collection jar through masturbation without use of lubricants after abstaining from intercourse for two days. Participants recorded the date and time of collection, their last day of ejaculation and any spillage, and were asked to place a sperm migration straw (Vitrotubes #3520 VintroCom) plugged at one end and filled with hyaluronic acid into the ejaculate to capture sperm motility at the time of sample collection. Semen samples were shipped overnight in insulated containers with cold packs and a temperature data logger (I-Button, Maxim Integrated) and analyzed within 24 hours by the National Institute of Occupational Safety and Health (NIOSH) Andrology Laboratory (Cincinnati, OH).

Thirty-five semen quality parameters were assessed. Six general semen quality parameters included semen volume (mL); 24-hour motility (%), measured using the HTM-IVOS (Hamilton Thorne) computer-assisted semen analysis system; distance traveled in the migration straw (mm) as a measure of motile sperm at collection; sperm count (millions/mL), measured using the IVOS system and the IDENT stain; total sperm count (millions), calculated as sperm count multiplied by semen volume; and hypo-osmotic swollen (%), using the hypo-osmotic swelling assay. Seven additional motility measures, including average path velocity (µm/sec), straight-line velocity (µm/sec), curvilinear velocity (µm/sec), amplitude of lateral head (µm), beat cross frequency (Hz), straightness (%) and linearity (%), were assessed using the HTM-IVOS. Six sperm head measures, including sperm head length (µm), area (µm2), width (µm), perimeter (µm), elongation factor width/length (%) and sperm head with acrosome (%), were measured using the IVOS METRIX system.

Fourteen morphology measures, including strict criteria (%), World Health Organization (WHO) normal criteria (%), amorphous (%), round (%), pyriform (%), bicephalic (%), taper (%), megalo head (%), micro head (%), neck and midpiece abnormal (%), coiled tail (%), other tail abnormalities (%), cytoplasmic droplet (%) and immature sperm (# immature) were measured by Fertility Solutions (Rothmann et al., 2013; World Health Organization, 1992). Two sperm chromatin stability assay measures, DNA fragmentation (%) and high DNA stainability (%), were assessed using the sperm chromatin stability assay following methods of Evenson et al. (Evenson et al., 2002). Quality assurance and quality control procedures included implementation of the Westgard Rules for ensuring absence of laboratory drift and batch-related differences (Westgard et al., 1981).

All semen quality parameters were assessed in the first sample, and analyses were conducted in a second sample with the purpose of confirming azoospermia, with restriction of measures assessed to semen volume, sperm concentration, total sperm count, hypo-osmotic swollen, motility and sperm head parameters due to budgetary reasons. Five participants who were azoospermic (sperm concentration=0 million/mL) were excluded from analysis.

Covariates

Covariates were assessed at baseline. Participants self-reported age (years), parity (ever vs. never fathered a live birth), household income (<$40,000, $40,000–$69,999, $70,000–$99,999, ≥$100,000) and education (high school graduate or less, some college, college graduate). Anthropometric measures (height and weight) were used to calculate body mass index (BMI, kg/m2). Smoking status was assessed using blood cotinine (>40.35 vs. ≤40.35 ng/mL) (Jeemon et al., 2010). Season of semen sample collection was assessed as cold (October 1st–March 31st) versus hot (April 1st–September 30th), and study site as Michigan versus Texas.

Statistical Analysis

Descriptive statistics were summarized as frequencies and percentages for categorical variables and either means and standard deviations (SDs) or medians and interquartile (IQR) ranges for continuous variables. Student’s t-test was used to evaluate differences in the distribution of variables across groups for continuous variables, and Pearson’s chi-square for categorical variables. Correlations across air pollutants were calculated for the 72-day average exposure preceding the first semen sample collection using Pearson product-moment correlation coefficients.

An available-case analysis approach was used in the main analyses. We excluded missing data on semen quality parameters (98 samples for distance traveled in straw, 32 for morphology, 9 for chromatin stability, 4 for sperm head, 2 for hypo-osmotic swollen and 1 for motility) and missing data on covariates (6 for smoking and 5 for BMI). In the first sample, 334 participants (71.5%) did not have daily air pollution levels for the full 72-day window, with a median of 58 days of air pollution (IQR 47–72); in the second sample, 32 participants (8.6%) did not have the full 72-day air pollution window. The distribution of missing air pollution levels is presented in Supplemental Figure 1. For participants with missing daily air pollution values, average air pollution exposure was estimated by averaging the observed daily pollutant data for a given window of exposure in the main analyses.

Generalized estimating equations were used to model the association between an IQR increase in the 72-day average exposure to each air pollutant and mean change in semen quality parameter, and allowed for inclusion of all semen samples obtained for each participant. Robust standard errors accounted for dependency between the first and second samples and the skewed distribution of several semen quality parameters. Multipollutant models adjusted for potential confounding between co-pollutants. For criteria air pollutants, one multiple pollutant model was fitted including all six pollutants: SO2, O3, NOX, CO, PM10 and PM2.5. For the five constituents of particulate matter, individual models were fitted for each constituent, adjusting for PM2.5. Individual-level covariates included in the models were chosen based on their association with semen quality parameters and inclusion in prior research (Radwan et al., 2016; Zhou et al., 2014), and included age, BMI, active smoking and season. Although several past studies have adjusted for abstinence time (Deng et al., 2016), we chose to not include abstinence time in our multivariable models as it does not meet criteria for confounding. Although abstinence time is associated with several semen quality parameters (Agarwal et al., 2016), it is not a direct or indirect cause of ambient air pollution level and therefore cannot confound the relationship of air pollution and semen quality (Michels et al., 2017).

To ensure our findings were robust, we conducted several secondary analyses to explore how key potential biases may have affected our results. First, we adjusted for site in a separate model, as site may both control for area-level confounding and over-adjust for air pollution differences between sites. To account for potential poor model fit due to the skewed distribution of semen quality parameters, a secondary analysis was conducted using the Box-Cox transformations (Box and Cox, 1964). After offsetting the original values by 0.001 to allow for transformation of zero values, a range of transformations were applied (λ=−3 to 3, by 0.25) to each semen quality parameter in the adjusted multipollutant model. The transformation associated with the lowest mean square error for each semen quality parameter was selected.

To evaluate the possible bias due to missing daily air pollution levels, a secondary analysis was conducted imputing missing daily air pollution levels and missing data on covariates. Markov chain Monte Carlo models (Schafer, 1997) were used to generate twenty imputed datasets, implemented with proc MI in SAS (Cary, NC). As missing data on air pollution was based on length of the enrollment menstrual cycle and number of cycles attempting pregnancy prior to enrollment, we were able to predict missingness based on covariates in our study and therefore were able to assume data were Missing at Random (White and Carlin, 2010). Covariates for the imputation model included length of the enrollment menstrual cycle and number of cycles attempting pregnancy prior to enrollment, as well as study id, days prior to ejaculation, sample number (1 or 2), abstinence time, season, having fathered a liveborn infant, study site, age, body mass index, race/ethnicity, household income, education and smoking status.

A secondary analysis was also conducted evaluating smaller 15-day windows of exposure during spermatogenesis (0–14, 15–29, 30–44, 45–59 and 60–72 days prior to ejaculation), to identify whether semen quality parameters may be more susceptible to air pollution exposure during specific windows of spermatogenesis. Smaller windows of exposure were evaluated using both the main adjusted multipollutant model and the model incorporating imputed data. Although we lacked data on the 90–72 day window prior to ejaculation for the majority of participants in the first sample (n=336, 71.0%), we evaluated the association of exposure to air pollution in the 90-day window prior to ejaculation among the semen quality parameters measured in the second sample (general, motility and sperm head parameters) to include key processes of epididymal storage and the development of sperm motility (Johnson, 1997) for comparison to results from the 72-day window. Finally, to account for multiple comparisons, we adjusted all models for the false discovery rate using the Benjamini-Hochberg procedure (Benjamini and Hochberg, 1995), within categories of semen quality parameters (general, motility, sperm head, morphology and sperm chromatin measures). All analyses were conducted in SAS 9.4 (Cary, NC).

Results

A total of 473 male participants provided at least one semen sample. After excluding one participant whose address could not be geocoded and five azoospermic participants, 467 participants remained in the analysis. Of those, 372 (79.7%) provided two samples. Mean age was 31.8 (SD 4.8), and the majority were non-Hispanic white (81.1%) (Table 1). Mean BMI was approximately 30 kg/m2, and one-quarter were smokers. Almost half had previously fathered a pregnancy. The majority were college graduates (63.9%) and more than half had a household income over $70,000/year (67.2%). Mean air pollutant levels for the 72-days preceding the first semen sample were low to moderate (Table 2). Although most air pollutants were positively correlated, O3 was negatively correlated with all criteria air pollutants and AEC (Supplemental Table 1). Additionally, PM10 was negatively correlated with SO2, AEC and ANO3; CO with ANO3; and ASO4 with SO2, NOX and ANO3. The first semen sample was provided at a median of 6 (IQR 3, 12) days after enrollment, and the second sample at a median of 60 (IQR 46, 86) days after enrollment, and reported abstinence time ranged from 1 to 91 days, with a median of 3 (IQR 2, 4) days. Semen quality parameters were overall not suggestive of impairments in fertility (Table 3).

Table 1.

Participant characteristics; LIFE Study (n=467)a

n (%)
Age (years) (mean, SD) 31.8 (4.8)
Parity and gravidityb
  Nulligravid 202 (43.4)
  Gravid, nulliparous 39 (8.4)
  Parous 224 (48.2)
Body mass index (kg/m2) (mean, SD)b 29.9 (5.6)
Race/ethnicityb
  Non-Hispanic white 337 (81.1)
  Non-Hispanic black 20 (4.3)
  Hispanic 38 (8.2)
  Other 30 (6.5)
Incomeb
  <$40,000 38 (8.3)
  $40–<$70,000 112 (24.5)
  $70–<$100,000 144 (31.4)
  >$100,000 164 (35.8)
Highest level of educationb
  High school or lower 36 (7.8)
  Some college 131 (28.3)
  College graduate 296 (63.9)
Study site
  Michigan 98 (21.0)
  Texas 369 (79.0)
Number of semen samples
  1 95 (20.3)
  2 372 (79.7)
Abstinence time (days) (median, IQR) 3 (2,4)
Open in a new tab
a

Excludes one participant who could not be geocoded and 5 participants with azoospermia in all samples

b

Missing data: Parity and gravidity (n=2), BMI (n=5), race/ethnicity (n=2), income (n=9) and education (n=4)

Table 2.

Distribution of average ambient air pollution levels in the 72 days prior to sample 1; the LIFE Study (n=467)

unit IQR Min 25% Median 75% Max
Criteria pollutants
  Sulfur dioxide (SO2) ppb 1.04 0.30 0.97 1.39 2.01 5.23
  Ozone (O3) ppb 8.3 9.8 23.7 28.0 32.0 41.3
  Nitrogen oxides (NOX) ppb 8.3 0.96 7.7 11.2 16.0 49.9
  Nitrogen dioxide (NO2) ppb 4.6 0.87 5.3 7.3 9.9 24.7
  Carbon monoxide (CO) ppb 126.7 83.1 184.7 257.0 311.4 484.8
  Particulate matter (PM10) µg/m3 11.7 6.2 16.7 24.4 28.4 39.7
  Fine particular matter (PM2.5) µg/m3 3.1 5.4 10.3 12.1 13.4 18.9
Particulate constituents
  Elemental carbon (AEC) µg/m3 0.31 0.0 04 0.12 0.27 0.43 1.17
  Ammonium (ANH4) µg/m3 0.38 0.34 1.01 1.21 1.39 2.46
  Nitrate (ANO3) µg/m3 0.74 0.003 0.56 0.8 1.3 5.17
  Organic compounds (AOC) µg/m3 1.02 0.35 1.48 1.89 2.50 5.74
  Sulfate (ASO4) µg/m3 1.15 1.42 2.51 3.10 3.66 6.05
Open in a new tab

Table 3.

Distribution of semen quality parameters in sample 1; the LIFE Study

n mean ± SD
General
Semen volume (mL) 467 3.42 ± 1.69
24-hour motility (%) 467 12.3 ± 12.5
Distance travelled in straw (mm) 369 10.5 ± 6.2
Sperm count (millions/mL) 467 73.9 ± 56.4
Total count (millions) 467 231.5 ± 187.8
Hypo-osmotic swollen (%) 466 67.1 ± 10.1
Motility
Average path velocity (µm/sec) 467 36.3 ± 13.4
Straight-line velocity (µm/sec) 467 27.1 ± 11.0
Curvilinear velocity (µm/sec) 467 62.2 ± 23.1
Amplitude of lateral displacement (µm) 467 3.13 ± 1.48
Beat cross frequency (Hz) 467 19.6 ± 7.7
Straightness (%) 467 67.4 ± 21.6
Linear (%) 467 40.8 ± 14.2
Sperm head parameters
Sperm head length (µm) 464 4.88 ± 0.28
Sperm head area (µm2) 464 12.19 ± 0.93
Sperm head width (µm) 464 3.17 ± 0.19
Sperm head perimeter (µm) 464 13.24 ± 0.53
Elongation factor length/width (%) 464 65.9 ± 5.3
Sperm head with acrosome (%) 464 25.7 ± 5.3
Morphology
Strict criteria (%) 435 20.15 ± 9.98
WHO normal criteria (%) 435 30.5 ± 12.4
Amorphous (%) 435 30.6 ± 11.1
Round (%) 435 1.16 ± 1.59
Pyriform (%) 435 6.28 ± 6.1
Bicephelic (%) 435 1.18 ± 1.67
Taper (%) 435 2.79 ± 2.67
Megalo head (%) 435 2.41 ± 1.84
Micro head (%) 435 1.48 ± 1.20
Neck and mid-piece abnormalities (%) 435 26.50 ± 9.81
Coiled tail (%) 435 24.2 ± 11.2
Other tail abnormalities (%) 435 5.21 ± 4.16
Cytoplasmic droplet (%) 435 10.21 ± 5.26
Immature (#) 435 6.14 ± 16.74
Sperm chromatin
DNA fragmentation index (%) 458 15.3 ± 10.4
High DNA stainability (%) 458 7.36 ± 5.21
Open in a new tab

In the main analysis evaluating the association of an IQR increase in mean air pollution exposure in multipollutant models in the 72 days preceding ejaculation and change in semen quality parameters, we observed few significant associations. There were only 38 associations among 385 total comparisons where the 95% confidence interval did not include the null value, and only 7 after adjusting for multiple comparisons (Table 4; Supplemental Table 2). Among general semen quality parameters, we observed associations between PM2.5 and lower percent 24-hour motility (β 2.13, 95% CI −3.63, −0.62, p=0.006) and percent hypo-osmotic swollen (β −1.98, 95% CI −3.27, −0.70, p=0.002). However, we also observed an association between SO2 and greater percent hypo-osmotic swollen (β 1.30, 95% CI 0.21, 2.40, p=0.019). We observed associations between CO and greater distance travelled in straw (an indicator of motility immediately following sample collection) (β 1.68, 95% CI 0.35, 0.30, p=0.013), and AEC, ANH4 and ANO3 and shorter distance travelled in straw (β −1.11, 95% CI −2.18, −0.05 mm, p=0.040; β −1.33, 95% CI −2.08, −0.58 mm, p=0.001; β −0.92, 95% CI −1.39, −0.45 mm, p<0.001; respectively). Only the associations between ANH4 and ANO3 with shorter distance travelled in straw survived adjustment for multiple comparisons. Only one association was observed for the motility parameters, SO2 and greater percent linear (β 1.65, 95% CI 0.09, 3.22, p=0.038).

Table 4.

Key associations for air pollution and semen quality parametersa, adjusted multipollutant modelb; the LIFE Study

Category semen
quality		 Semen quality parameter Air
pollutantc 		β (95% CI)d P
General 24-hour motility (%) PM2.5 −2.13 (−3.63, −0.62) 0.006
Distance traveled in straw (mm) CO 1.68 (0.35, 3.00) 0.013
AEC −1.11 (−2.18, −0.05) 0.040
ANH4 −1.33 (−2.08, −0.58) 0.001
ANO3 −0.92 (−1.39, −0.45) <0.001
Hypo-osmotic swollen (%) SO2 1.30 (0.21, 2.40) 0.019
PM2.5 −1.98 (−3.27, −0.7) 0.002
Motility Linear (%) SO2 1.65 (0.09, 3.22) 0.038
Sperm head parameters Sperm head length (µm) O3 0.031 (0.001, 0.061) 0.040
PM2.5 −0.056 (−0.086, −0.025) <0.001
Sperm head area (µm2) PM2.5 −0.221 (−0.335, −0.108) <0.001
ANO3 −0.100 (−0.162, −0.038) 0.002
Sperm head width (µm) PM2.5 −0.025 (−0.046, −0.003) 0.026
ANO3 −0.012 (−0.025, −0.0003) 0.045
AOC 0.020 (0.002, 0.037) 0.026
Sperm head perimeter (µm) PM2.5 −0.123 (−0.185, −0.060) <0.001
ANO3 −0.046 (−0.082, −0.009) 0.015
Sperm head with acrosome (%) PM2.5 1.029 (0.396, 1.661) 0.001
ANO3 0.331 (0.002, 0.660) 0.049
Morphology Amorphous (%) O3 2.11 (0.16, 4.07) 0.034
Bicephelic (%) CO 0.201 (0.032, 0.370) 0.020
Megalo head (%) SO2 −0.223 (−0.442, −0.004) 0.046
ANH4 −0.229 (−0.419, −0.040) 0.018
Micro head (%) NOX −0.241 (−0.411, −0.071) 0.006
Neck and mid-piece abnormalities (%) ANH4 −1.22 (−2.39, −0.06) 0.040
Coiled tail (%) SO2 −1.77 (−3.27, −0.27) 0.021
ASO4 −2.31 (−4.20, −0.42) 0.017
Cytoplasmic droplet (%) PM2.5 −0.93 (−1.77, −0.09) 0.030
ANH4 −0.66 (−1.29, −0.02) 0.042
Immature (#) O3 −0.93 (−1.51, −0.35) 0.002
NOX −0.66 (−1.22, −0.09) 0.022
ANH4 −0.49 (−0.89, −0.09) 0.017
AOC −0.56 (−1.01, −0.11) 0.014
ASO4 −0.68 (−1.26, −0.10) 0.021
Sperm chromatin High DNA stainability (%) PM10 0.97 (0.19, 1.76) 0.015
PM2.5 −0.71 (−1.35, −0.07) 0.030
ANO3 −0.38 (−0.69, −0.07) 0.018
ASO4 0.79 (0.05, 1.52) 0.035
Open in a new tab
a

Associations presented in table reached statistical significance, and all associations are presented in Supplemental Table 2. Associations in bold remained significant after adjustment for the false discovery rate (per category of semen quality).

b

Models adjusted for age (continuous), BMI (<25, 25–<30 and ≥30 kg/m2), smoking (blood cotinine ≥40.35 vs. <40.35 ng/mL) and season (cold [Oct. 1st–March 31st] vs. hot [April 1st–September 30th]).

c

SO2, sulfur dioxide; O3, ozone; NOX, nitrogen oxides; CO, carbon monoxide; PM10, particulate matter <10 µg/m3; PM2.5, fine particulate matter <2.5 µg/m3; AEC, elemental carbon; ANH4, ammonium; ANO3, nitrate; AOC, organic compounds; ASO4, sulfate; DNA, deoxyribonucleic acid

d

Linear regression with robust standard errors for sperm morphology and chromatin stability measures and generalized estimating equations with robust standard errors for volume, count, percent hypo-osmotic swollen, 24-hour motility and sperm head measures.

The strongest associations were found between air pollutants and sperm head parameters. PM2.5 was associated with smaller sperm head length (β −0.056, 95% CI −0.086, −0.025 µm, p<0.001), whereas both PM2.5 and ANO3 were associated with smaller sperm head area (β −0.221, 95% CI −0.335, −0.108 µm2, p<0.001; β −0.100, 95% CI −0.162, −0.038 µm2, p=0.002, respectively), sperm head width (β −0.025, 95% CI −0.046, −0.003 µm, p=0.026; β −0.012, 95% CI −0.025, −0.0003 µm, p=0.045, respectively), sperm head perimeter (β −0.123, 95% CI −0.185, −0.060 µm, p=<0.001; β −0.046, 95% CI −0.082, −0.009 µm, p=0.015, respectively) and greater percent sperm head with acrosome (β 1.029, 95% CI 0.396, 1.661, p=0.001; β 0.331, 95% CI 0.002, 0.660, p=0.049, respectively). Only the associations for PM2.5 with sperm head length, PM2.5 and ANO3 with smaller sperm head area, PM2.5 with smaller sperm head perimeter and PM2.5 with greater sperm head percent with acrosome persisted after adjustment for multiple comparisons.

A general trend was observed between air pollutants and fewer impairments in morphology, although none of the associations survived the false discovery rate. No associations were observed for DNA fragmentation index. Conflicting associations were found for high DNA stainability, with PM10 and ASO4 associated with a greater percent sperm with high DNA stainability (β 0.97, 95% CI 0.19, 1.76, p=0.015; β 0.79, 95% CI 0.05, 1.52, p=0.035, respectively), and PM2.5 and ANO3 associated with a lower percent sperm with high DNA stainability (β −0.71, 95% CI −1.35, −0.07, p=0.030; β −0.38, 95% CI −0.69, −0.07, p=0.018, respectively). None survived adjustment for the false discovery rate. Observations were similar for unadjusted models (Supplemental Table 3) and single-pollutant models adjusted for covariates (Supplemental Table 4).

Similar to our main analyses, the findings from secondary analyses suggested that air pollutant levels had little impact on semen quality. In a secondary analysis adjusting for study site (Supplemental Table 5), the effect estimates remained similar, but precision was reduced. Distance travelled in straw is an exception with significant differences by site (3.7 mm greater distance in Texas vs. Michigan). After adjusting for site, carbon monoxide (β −0.18, 95% CI −2.06, 1.71 mm, p=0.85), ammonium (β −0.36, 95% CI −1.28, 0.56 mm, p=0.45) and nitrates (β −0.08, 95% CI −0.82, 0.67 mm, p=0.84) were no longer associated with distance travelled in straw. Applying Box-Cox transformations to normalize semen quality parameters did not alter findings (data not shown). The models incorporating multiply imputed daily air pollution values also produced similar findings to the main model (Supplemental Table 6). The main differences in findings for the model including imputed data versus the main model were an attenuated association of PM2.5 with 24-hour motility (β −1.48, 95% CI −3.20, 0.25, p=0.09) and SO2 with percent hypo-osmotic swollen (β 0.99, 95% CI −0.31, 2.29, p=0.14), and stronger associations for percent sperm head with acrosome, including between PM10 and ASO4 and lower percent sperm head with acrosome (β −1.25, 95% CI −2.17, −0.32, p=0.008; β −0.95, 95% CI −1.72, −0.19, p=0.016, respectively). When evaluating the association of exposure during a 90-day window in the second sample, we observed additional associations between PM2.5 and motility, including beat cross frequency (β −1.19, 95% CI −1.98, −0.41 Hz, p=0.003), although these associations did not survive adjustment for multiple comparisons (Supplemental Table 7). Finally, in an evaluation of smaller windows of exposure to air pollution during spermatogenesis and semen quality for both the main model (Figures 1 and 2) and the model with multiple imputed data (Supplemental Figures 2 and 3), we observed no clear patterns of association for early versus late exposure to air pollutants.

Figure 1.

Figure 1

Figure 1

Open in a new tab

Associations for air pollution and semen quality parameters that reached statistical significance by 15-day windows of exposure, adjusted multipollutant model. Arrows indicate the direction of association between the air pollutant (SO2, sulfur dioxide; O3, ozone; NOX, nitrogen oxides; CO, carbon monoxide; PM10, particulate matter <10 µg/m3; PM2.5, fine particulate matter <2.5 µg/m3; AEC, elemental carbon; ANH4, ammonium; ANO3, nitrate; AOC, organic compounds; ASO4, sulfate) and general, motility and sperm head parameters.

Figure 2.

Figure 2

Figure 2

Open in a new tab

Associations for air pollution and semen quality parameters that reached statistical significance by 15-day windows of exposure, adjusted multipollutant model. Arrows indicate the direction of association between the air pollutant (SO2, sulfur dioxide; O3, ozone; NOX, nitrogen oxides; CO, carbon monoxide; PM10, particulate matter <10 µg/m3; PM2.5, fine particulate matter <2.5 µg/m3; AEC, elemental carbon; ANH4, ammonium; ANO3, nitrate; AOC, organic compounds; ASO4, sulfate) and morphology and sperm chromatic parameters.

Discussion

In an evaluation of the association between air pollution and semen quality parameters among healthy men in areas with low to moderate exposure to air pollution, we observed mostly null results. Associations that remained after adjustment for multiple comparisons included fine particulates with smaller sperm head size, greater percent sperm head with acrosome and shorter distance travelled in migration straw, which may warrant investigation in future research. Overall, however, among a study population of healthy men exposed to moderate levels of ambient air pollution, there is no clear evidence to suggest an association between ambient air pollution exposure and adverse changes in semen quality. This suggests that air pollution may not be a major contributor to differences in semen quality among men with no indication for infertility and who are exposed to moderate levels of air pollution.

The levels of air pollution observed in our study are comparable to those of previous studies conducted in the United States where few associations for air pollution and differences in semen quality have been observed. Among 228 male partners of pregnant women in North Carolina, Tennessee and Texas, no associations were found for average exposure to O3 or PM2.5 in the 90 days prior to ejaculation with sperm concentration, total sperm count, morphology, DNA integrity or DNA chromatin maturity (Hansen et al., 2010). Among 1,699 men evaluated for infertility in Salt Lake County, Utah, PM2.5 was found to be negatively correlated with motility and the 2nd and 3rd months prior to sample date (r=−0.330, p=0.009) after adjustment for ambient temperature and season, but not to be associated with sperm concentration or morphology (Hammoud et al., 2010). Among 48 non-smoking men who donated semen samples in Los Angeles, O3 was found to be associated with lower sperm count 0–9 days prior to donation (2.80% decrease, p=0.04) and 10–14 days prior to donation (2.36%, p=0.04). No associations were observed for NO2, CO or PM10 (Sokol et al., 2006).

Our air pollution exposures are lower than those observed in much of the research conducted outside the United States. In two European studies (Radwan et al., 2016; Santi et al., 2016), both PM2.5 (mean 34.52 µg/m3 [SD 22.13] Poland and mean 27.5 µg/m3 [SD 17.34] Italy) and PM10 (mean 41.85 µg/m3 [SD 34.17] Poland and mean 40.69 µg/m3 [SD 22.04] Italy) were approximately 2–3 times higher than our observed values of 12.2 and 24.4 µg/m3 for PM2.5 and PM10, respectively. Both studies observed more consistent associations for particulates and semen quality. In a study in Lodz, Poland among 327 men attending an infertility clinic, PM2.5 was associated with a greater percent sperm with abnormal morphology (β 33.4, 95% CI 24.0, 42.8) and greater percent sperm with high DNA stainability (β 0.36, 95% CI 0.17, 0.55) (Radwan et al., 2016). In a study among 406 men presenting for semen analysis in a clinic in Modena Provence, Italy, PM10 was positively correlated with semen volume (ρ=0.211, P=0.001) (Santi et al., 2016).

In two studies conducted in China, levels of PM10 were several times higher than our observed median value of 24.4 µg/m3 (IQR 16.7, 28.4), with observed median values of PM10 ranging from 82.5 (5%–95% 49.1–132.9) to 104.0 (5%–95% 28.0–245.9) µg/m3 in rural and urban areas in Chongqing (Zhou et al., 2014) and 104.0 µg/m3 (5%–95% 35.0, 222.0) in Wuhan (Wu et al., 2017). In contrast to our findings, the study in Chongqing, conducted among 1,346 men with no history of infertility, found that PM10, SO2 and NO2 were associated with lower normal morphology (β −0.212, p<0.001; β −0.378, p<0.001; and β −0.381, p<0.001, respectively) (Zhou et al., 2014). Among 1,759 male partners of women undergoing assisted reproduction in Wuhan, average PM2.5 in the 90 days prior to the sample was associated with lower sperm concentration (β −0.20, 95% CI −0.34, −0.07) and sperm count (β −0.22, 95% CI −0.35, −0.08) (Wu et al., 2017). Given this past research, it is possible that a threshold effect exists for the association of air pollution and semen quality parameters and/or the variability at lower levels of air pollution leads to differences difficult to detect in semen quality parameters. In either scenario, generalizability of air pollution findings across populations with varying exposure levels may not be advisable.

Due to the hypothesized effects of air pollution on semen quality through oxidative stress mechanisms (Du Plessis et al., 2015), it is notable that we observed no clear associations between air pollutants and DNA fragmentation index, percent hypo-osmotic swollen, and motility, as these parameters are particularly sensitive to oxidative stress. We did, however, observe relatively strong associations between constituents of fine particulate matter, notably ANH4 and ANO3, and motility immediately following collection as measured by the placement of a migration straw in the semen sample (“distance travelled in straw”). However, these associations were severely attenuated after adjustment for site, and may reflect site-level differences in air pollution or other unmeasured factors.

The association between air pollution and sperm head measures has not been evaluated in prior research, although findings from the LIFE study suggest that sperm head measures may be sensitive to environmental pollutants (Bloom et al., 2015; Louis et al., 2015). Despite an association with smaller sperm head measures, we observed no associations between fine particulates and semen morphology measures. Additionally, exposure to fine particulates was associated with greater percent of sperm head with acrosome, a feature associated with improved fertilization. This apparent protective association may be due to the role of reactive oxygen species in the acrosome reaction, with moderate levels of air pollution exposure leading to only small changes in oxidative stress and a potential subsequent augmentation of the acrosome reaction, rather than the decrements in mature spermatozoa observed with large increases in oxidative stress (Du Plessis et al., 2015). Although these associations persisted after adjustment for multiple comparisons, they need to be replicated due to the potential for chance findings.

Due to few common causes for individual-level behaviors and ambient air pollution levels, it is unlikely that unmeasured participant-level factors biased our findings, but we cannot rule out confounding by either broader area-level risk factors or co-pollutants, which may have biased our results towards the null. It is reassuring that findings were similar for the single- and multipollutant models, suggesting that adjustment for correlated co-pollutants had little effect on estimates. It is also possible that misclassification of exposure to air pollution may have biased our results towards the null. However, use of the modified CMAQ model allowed for a relatively precise estimation of ambient air pollution exposure around each participant’s residence as compared to sparse ambient monitoring employed in most prior research. Our group has additionally observed biologically relevant associations between air pollution and both pregnancy loss (Ha et al., 2017) and fecundability (Nobles et al., 2017) in the LIFE study, suggesting a reasonable degree of precision of our air pollution assessment.

We had missing air pollution values for a portion of the days used to calculate mean exposure for the windows of interest prior to ejaculation. Although we averaged the non-missing values for each participant over those windows, it is possible that this led to imprecision in our estimation of the full 72-day window of exposure. Missing data was particularly a concern in assessing smaller windows of exposure, as participants who were missing all daily measures in a given window were excluded from analysis. To address this concern for selection bias in the analysis of finer windows of exposure, we used multiple imputation to estimate missing air pollution data, and observed no major differences in the observed patterns of associations between the available-case models and the models incorporating imputed data. We additionally lacked data for the window of 90-72 days prior to ejaculation, during which key processes of epididymal storage and development of sperm motility occur (Johnson, 1997). However, in a secondary analysis using data from the second sample only, the results were generally consistent with our main study findings. As noted above, our findings may not be generalizable to individuals who live in areas with higher levels of air pollution.

In summary, we found few associations between exposure to ambient air pollution and semen quality measures. Although we did observe associations between PM2.5 and sperm head measures that were robust to adjustment for multiple comparisons, our findings generally suggest that, among our healthy study population with relative low exposure, there is not enough evidence to suggest an association between ambient air pollution exposure and adverse changes in semen quality. Further investigation of semen quality among populations exposed to higher levels of ambient air pollution may be warranted.

Supplementary Material

supplement

Highlights.

  • Sperm development is sensitive to oxidative stress, a key effect of air pollution

  • Prior evidence is inconclusive for an effect of air pollution on semen quality

  • Ambient air pollution was not associated with most semen quality parameters

  • Fine particulate matter was associated with smaller sperm head measures

  • Findings are generally reassuring for healthy men exposed to moderate air pollution

Acknowledgments

Funding: This work was supported by the Intramural Research Program of the Eunice Kennedy Shriver National Institute of Child Health and Human Development (LIFE study contract nos. #N01-HD-3-3355, NO1-HD-#-3356, N01-HD-3-3358 and the Air Quality and Reproductive Health Study Contract No. HHSN275200800002I, Task Order No. HHSN27500008).

Role of the funding source: The funding source had no involvement in the study design; in the collection, analysis or interpretation of data, in the writing of the report; or in the decision to submit the article for publication.

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 citable 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.

Declarations of interest: none.

References

  1. Agarwal A, Gupta S, Du Plessis S, et al. Abstinence Time and Its Impact on Basic and Advanced Semen Parameters. Urology. 2016;94:102–10. doi: 10.1016/j.urology.2016.03.059. [DOI] [PubMed] [Google Scholar]
  2. Ahmadi S, Bashiri R, Ghadiri-Anari A, et al. Antioxidant supplements and semen parameters: An evidence based review. Int J Reprod Biomed (Yazd) 2016;14:729–736. [PMC free article] [PubMed] [Google Scholar]
  3. Aitken RJ, Baker MA, Nixon B. Are sperm capacitation and apoptosis the opposite ends of a continuum driven by oxidative stress? Asian J Androl. 2015;17:633–9. doi: 10.4103/1008-682X.153850. [DOI] [PMC free article] [PubMed] [Google Scholar]
  4. Azenabor A, Ekun AO, Akinloye O. Impact of Inflammation on Male Reproductive Tract. J Reprod Infertil. 2015;16:123–9. [PMC free article] [PubMed] [Google Scholar]
  5. Benjamini Y, Hochberg Y. Controlling the false discovery rate: a practical and powerful approach to multiple testing. Journal of the royal statistical society. Series B (Methodological) 1995:289–300. [Google Scholar]
  6. Bloom MS, Whitcomb BW, Chen Z, et al. Associations between urinary phthalate concentrations and semen quality parameters in a general population. Hum Reprod. 2015;30:2645–57. doi: 10.1093/humrep/dev219. [DOI] [PMC free article] [PubMed] [Google Scholar]
  7. Box GE, Cox DR. An analysis of transformations. Journal of the Royal Statistical Society. Series B (Methodological) 1964:211–252. [Google Scholar]
  8. Buck Louis GM, Schisterman EF, Sweeney AM, et al. Designing prospective cohort studies for assessing reproductive and developmental toxicity during sensitive windows of human reproduction and development--the LIFE Study. Paediatr Perinat Epidemiol. 2011;25:413–24. doi: 10.1111/j.1365-3016.2011.01205.x. [DOI] [PMC free article] [PubMed] [Google Scholar]
  9. Buck Louis GM, Sundaram R, Schisterman EF, et al. Semen quality and time to pregnancy: the Longitudinal Investigation of Fertility and the Environment Study. Fertil Steril. 2014;101:453–62. doi: 10.1016/j.fertnstert.2013.10.022. [DOI] [PMC free article] [PubMed] [Google Scholar]
  10. Calogero AE, La Vignera S, Condorelli RA, et al. Environmental car exhaust pollution damages human sperm chromatin and DNA. J Endocrinol Invest. 2011;34:e139–43. doi: 10.1007/BF03346722. [DOI] [PubMed] [Google Scholar]
  11. Checa Vizcaino MA, Gonzalez-Comadran M, Jacquemin B. Outdoor air pollution and human infertility: a systematic review. Fertil Steril. 2016;106:897–904.e1. doi: 10.1016/j.fertnstert.2016.07.1110. [DOI] [PubMed] [Google Scholar]
  12. Chen G, Li J, Ying Q, et al. Evaluation of observation-fused regional air quality model results for population air pollution exposure estimation. Sci Total Environ. 2014:485–486. 563–74. doi: 10.1016/j.scitotenv.2014.03.107. [DOI] [PMC free article] [PubMed] [Google Scholar]
  13. Chin MT. Basic mechanisms for adverse cardiovascular events associated with air pollution. Heart. 2015;101:253–6. doi: 10.1136/heartjnl-2014-306379. [DOI] [PMC free article] [PubMed] [Google Scholar]
  14. Deng Z, Chen F, Zhang M, et al. Association between air pollution and sperm quality: A systematic review and meta-analysis. Environ Pollut. 2016;208:663–9. doi: 10.1016/j.envpol.2015.10.044. [DOI] [PubMed] [Google Scholar]
  15. Du Plessis SS, Agarwal A, Halabi J, et al. Contemporary evidence on the physiological role of reactive oxygen species in human sperm function. J Assist Reprod Genet. 2015;32:509–20. doi: 10.1007/s10815-014-0425-7. [DOI] [PMC free article] [PubMed] [Google Scholar]
  16. Evenson DP, Larson KL, Jost LK. Sperm chromatin structure assay: its clinical use for detecting sperm DNA fragmentation in male infertility and comparisons with other techniques. J Androl. 2002;23:25–43. doi: 10.1002/j.1939-4640.2002.tb02599.x. [DOI] [PubMed] [Google Scholar]
  17. Foley K, Roselle S, Appel K, et al. Incremental testing of the Community Multiscale Air Quality (CMAQ) modeling system version 4.7. Geoscientific Model Development. 2010;3:205. doi: 10.5194/gmd-10-1703-2017. [DOI] [PMC free article] [PubMed] [Google Scholar]
  18. Gabrielsen JS, Tanrikut C. Chronic exposures and male fertility: the impacts of environment, diet, and drug use on spermatogenesis. Andrology. 2016;4:648–61. doi: 10.1111/andr.12198. [DOI] [PubMed] [Google Scholar]
  19. Guven A, Kayikci A, Cam K, et al. Alterations in semen parameters of toll collectors working at motorways: does diesel exposure induce detrimental effects on semen? Andrologia. 2008;40:346–51. doi: 10.1111/j.1439-0272.2008.00867.x. [DOI] [PubMed] [Google Scholar]
  20. Ha S, Sundaram R, Buck Louis GM, et al. Ambient air pollution and the risk of pregnancy loss: a prospective cohort study. Fertil Steril. 2017 doi: 10.1016/j.fertnstert.2017.09.037. [DOI] [PMC free article] [PubMed] [Google Scholar]
  21. Hammoud A, Carrell DT, Gibson M, et al. Decreased sperm motility is associated with air pollution in Salt Lake City. Fertil Steril. 2010;93:1875–9. doi: 10.1016/j.fertnstert.2008.12.089. [DOI] [PubMed] [Google Scholar]
  22. Hansen C, Luben TJ, Sacks JD, et al. The effect of ambient air pollution on sperm quality. Environ Health Perspect. 2010;118:203–9. doi: 10.1289/ehp.0901022. [DOI] [PMC free article] [PubMed] [Google Scholar]
  23. Heller CG, Clermont Y. Spermatogenesis in man: an estimate of its duration. Science. 1963;140:184–6. doi: 10.1126/science.140.3563.184. [DOI] [PubMed] [Google Scholar]
  24. Jeemon P, Agarwal S, Ramakrishnan L, et al. Validation of self-reported smoking status by measuring serum cotinine levels: an Indian perspective. Natl Med J India. 2010;23:134–6. [PubMed] [Google Scholar]
  25. Johnson LW, T H, Jr, Wilker CE. Anatomy and physiology of the male reproductive system and potential targets of toxicants. In: Sipes CA GM, Gandolfi AJ, editors. Comprehensive Toxicology. Pergamon; New York: 1997. [Google Scholar]
  26. Kaufman JD, Adar SD, Barr RG, et al. Association between air pollution and coronary artery calcification within six metropolitan areas in the USA (the Multi-Ethnic Study of Atherosclerosis and Air Pollution): a longitudinal cohort study. Lancet. 2016;388:696–704. doi: 10.1016/S0140-6736(16)00378-0. [DOI] [PMC free article] [PubMed] [Google Scholar]
  27. Lafuente R, Garcia-Blaquez N, Jacquemin B, et al. Outdoor air pollution and sperm quality. Fertil Steril. 2016;106:880–96. doi: 10.1016/j.fertnstert.2016.08.022. [DOI] [PubMed] [Google Scholar]
  28. Langrish JP, Bosson J, Unosson J, et al. Cardiovascular effects of particulate air pollution exposure: time course and underlying mechanisms. J Intern Med. 2012;272:224–39. doi: 10.1111/j.1365-2796.2012.02566.x. [DOI] [PubMed] [Google Scholar]
  29. Louis GM, Chen Z, Schisterman EF, et al. Perfluorochemicals and human semen quality: the LIFE study. Environ Health Perspect. 2015;123:57–63. doi: 10.1289/ehp.1307621. [DOI] [PMC free article] [PubMed] [Google Scholar]
  30. Michels KA, Kim K, Yeung EH, et al. Adjusting for abstinence time in semen analyses: some considerations. Andrology. 2017;5:191–193. doi: 10.1111/andr.12285. [DOI] [PMC free article] [PubMed] [Google Scholar]
  31. Nobles CJ, Schisterman EF, Ha S, et al. Time-varying cycle average and daily variation in ambient air pollution and fecundability. Hum Reprod. 2017:1–11. doi: 10.1093/humrep/dex341. [DOI] [PMC free article] [PubMed] [Google Scholar]
  32. Radwan M, Jurewicz J, Polanska K, et al. Exposure to ambient air pollution--does it affect semen quality and the level of reproductive hormones? Ann Hum Biol. 2016;43:50–6. doi: 10.3109/03014460.2015.1013986. [DOI] [PubMed] [Google Scholar]
  33. Rothmann SA, Bort AM, Quigley J, et al. Sperm morphology classification: a rational method for schemes adopted by the world health organization. Methods Mol Biol. 2013;927:27–37. doi: 10.1007/978-1-62703-038-0_4. [DOI] [PubMed] [Google Scholar]
  34. Santi D, Vezzani S, Granata AR, et al. Sperm quality and environment: A retrospective, cohort study in a Northern province of Italy. Environ Res. 2016;150:144–53. doi: 10.1016/j.envres.2016.05.053. [DOI] [PubMed] [Google Scholar]
  35. Schafer JL. Analysis of incomplete multivariate data. CRC press; 1997. [Google Scholar]
  36. Shah AS, Lee KK, McAllister DA, et al. Short term exposure to air pollution and stroke: systematic review and meta-analysis. Bmj. 2015;350:h1295. doi: 10.1136/bmj.h1295. [DOI] [PMC free article] [PubMed] [Google Scholar]
  37. Sokol RZ, Kraft P, Fowler IM, et al. Exposure to environmental ozone alters semen quality. Environ Health Perspect. 2006;114:360–5. doi: 10.1289/ehp.8232. [DOI] [PMC free article] [PubMed] [Google Scholar]
  38. Westgard JO, Barry PL, Hunt MR, et al. A multi-rule Shewhart chart for quality control in clinical chemistry. Clin Chem. 1981;27:493–501. [PubMed] [Google Scholar]
  39. White IR, Carlin JB. Bias and efficiency of multiple imputation compared with complete-case analysis for missing covariate values. Stat Med. 2010;29:2920–31. doi: 10.1002/sim.3944. [DOI] [PubMed] [Google Scholar]
  40. World Health Organization. Laboratory manual for the examination of human semen and semen-cervical mucus interaction. 3. Cambridge University Press; Cambridge, UK: 1992. [Google Scholar]
  41. World Health Organization. Air quality guidelines: global update 2005: particulate matter, ozone, nitrogen dioxide, and sulfur dioxide. World Health Organization; 2006. [PubMed] [Google Scholar]
  42. Wu L, Jin L, Shi T, et al. Association between ambient particulate matter exposure and semen quality in Wuhan, China. Environ Int. 2017;98:219–228. doi: 10.1016/j.envint.2016.11.013. [DOI] [PubMed] [Google Scholar]
  43. Zhou N, Cui Z, Yang S, et al. Air pollution and decreased semen quality: a comparative study of Chongqing urban and rural areas. Environ Pollut. 2014;187:145–52. doi: 10.1016/j.envpol.2013.12.030. [DOI] [PubMed] [Google Scholar]

Associated Data

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

Supplementary Materials

supplement
NIHMS944325-supplement.xlsx (1.6MB, xlsx)

RESOURCES