Table 3.

The results of the studies on air pollution and male fertility

Source	Exposure assessment	Outcome measured	Confounders	Statistical analysis	Main faindings	Limitations	Strengths
Santi, et al. 2016 [25]	PM10 (μg/m3)	Volume, concentration (× 106/mL), total number, typical forms (%), atypical forms (%), progressive motility (%), non-progressive motility (%), total motility (%), leucocytes	Temperature, humidity	Bivariate and multivariate logistic regression models	PM2.5 was directly related to total sperm number (p = 0.001); PM10 was directly related to semen volume (p < 0.001) and typical forms (p < 0.001), inversely related to atypical forms (p < 0.001)	No information about smoking; Daily PM exposure registered through the monitoring network of the quality of the air for the province of Madena; the relationship between environmental PM and semen quality was evaluated irrespective of the life-style and risk factors of each men	The cohort is highly representative of the entire population; Choosing the coldest season of the year in this area (the possible negative influence of high ambient temp. on sperm quality was excluded);
	PM2.5 (μg/m3)
Wu, et al. 2016 [27]	PM10 (μg/m3)	Concentration(×106/mL), sperm count, total motility (%), progressive motility (%),	Age, BMI, ethnic, education, smoking, drinking, abstinence, season, average ambient temperature	Multivariate linear mixed models	PM2.5 was inversely associated with sperm concentration (β = −0.20; (95%CI: − 0.34, − 0.07) and sperm count (β = − 0.22; 95%CI: − 0.35, − 0.08); PM10 was inversely associated with sperm concentration and count (p < 0.05)	The used of ambient PM exposures as a proxy for individual exposures; No investigation of the association between exposure and sperm morphology; the results cannot be directly generalized to populations in which PM exposure levels are too low; Participants from infertility clinic	Big sample size; A wide concentration range of PM exposure; the use of IDW interpolation to more precisely estimate individual PM exposure; the non-linear exposure-response relationships
	PM2.5 (μg/m3)
Radwan, et al. 2016 [23]	PM10	Concentration (× 106/mL), motility (%), sperm with abnormal morphology (%), DFI(%), HDS(%), CASA parameters: VSL, VCL, LIN; FSH, E2, T	Age, smoking, temperature (men from 90 days), past disease, time of sexual abstinence, season	Multivariate linear regression	Exposure to air pollutants (PM10, PM2,5, SO2, NOx, CO) increased the abnormalities in sperm morphology (p = 0.0002, p = 0.0001, p = 0.0001, p = 0.01, p = 0.0001, respectively); Negative association were found between air pollutants and testosterone levels (p < 0.05); There were a positive associations between PM10 and PM2,5 and HDS (p = 0.002, p = 0.0001, respectively)	participants from an infertility clinic; single semen sample; exposure levels assessed by considering the ZIP code of each participants;	Assessment of many different semen parameters, reproductive hormones levels and sperm chromatin structure; detailed questionnaire information allowed for control confounding factors in the analysis; cotinine measured in saliva to verified smoking status;
	PM2.5
	SO2
	CO
	NOx
Jurewicz, et al. 2015 [34]	PM10 (μg/m3)	Sperm aneuploidy, sperm concentration (×106/mL), total motility (%), abnormal morphology (%)	Age, smoking, drinking, temperature, season, past disease, abstinence interval, distance from monitoring station, concentration, motility, morphology, PM10, SO2	multivariate analysis	There was a positive association between exposure to PM2.5 and disomy Y (p = 0.001), sex chromosome disomy (p = 0.05), disomy of chromosome 21 (p = 0.03); xposure to PM10 was associated with disomy 21 (p = 0.02)	Exposure levels assessed by considering the ZIP code of each participants; participants from an infertility clinic; single semen sample;	Many confounders included to the analysis; cotinine measured in saliva to verified smoking status; detailed questionnaire information allowed for control confounding factors in the analysis,
	PM2.5 (μg/m3)
	O3 (μg/m3)
	SO2 (μg/m3)
	CO (μg/m3)
	NOx (μg/m3)
Radwan, et al. 2015 [35]	1 PAH metabolite in urine (1-OHP)	Sperm aneuploidy, sperm concentration (× 106/mL), total motility (%), normal sperm morphology (%),	Abstinence, age, smoking, season, past disease,	Multivariate analysis	Positive associations between level of 1-OHP in urine and total sex-chromosome disomy (p = 0.03); An increase in the frequency of disomy 18 was related to the level of 1-OHP in urine (p = 0.03)	participants from an infertility clinic; single semen sample; one biomarker of PAHs exposure- 1-hydroksypyrene (1-OHP)	cotinine measured in saliva to verified smoking status; detailed questionnaire information allowed for control confounding factors in the analysis,
Zhou, et al. 2014 [21]	PM10 (μg/m3)	Volume, concentration (×106/mL), progressive motility (%), total motility, morphology (normal forms %), CASA sperm motility parameters (VCL, VSL, VAP, BCF, ALH, LIN, STR)	Age, education, smoking, drinking, BMI, abstinence, season,	Multivariate regression models	PM10, SO2, NO2 were negatively associated with a normal sperm morphology percentage (p < 0.001); there were inverse associations between sperm VCL and VSL value and PM10, SO2, NO2 (p < 0.001); PM10 was positively associated with sperm concentration (p = 0.031)	Used of monitoring data for the study sites to measure ambient air pollution; the measurement and prediction of PM10, SO2, NO2 exposure were performed outside; the used only a single semen sample	Big sample size; Exposure to air pollutants in both urban and rural areas; The use of mean concentration of each pollutant during the 90 days before sampling; Many confounders including to the analyses
	SO2 (μg/m3)
	NO2 (μg/m3)
Jurewicz, et al. 2013 [24]	1 PAH metabolite in urine (1-OHP)	Volume, concentration (×106/mL), motility (%), atypical sperm (%),static sperm (%), DFI (%), CASA parameters: VAP, VSL, VCL, BCF, ALH,	Age, smoking, past disease, season, sexual abstinence,	Multivariate regression models	A positive associations were observed between the level of 1-OHP in urine and sperm neck abnormalities (p = 0.001) and a negative between the semen volume (p = 0.014), %motility (p = 0.0001) and %static sperm (p = 0.018);	Participants from an infertility clinic; single urine and semen sample; one biomarker of PAHs exposure- 1-hydroksypyrene (1-OHP)	cotinine measured in saliva to verified smoking status; detailed questionnaire information allowed for control confounding factors in the analysis,
Song, et al. 2013 [22]	16 PAHs in blood	Concentration (× 106/mL), volume, motility (grade A, grade B, grade C)	–	The Pearson correlation analysis	Significant correlations between PAHs in blood and semen motility were observed (p < 0.01)	Small sample size; participants from infertility clinic; No investigation of the association between PAH exposure and semen morphology;	16 PAH metabolites as a biomarkers of exposure to PAH;
Han, et al. 2011 [31]	4 PAH metabolites in urine	Apoptotic marker (Annexin V−/PI− spermatozoa %, Annexin V+/PI− spermatozoa %, PI+ spermatozoa %, comet parameters (tail%, tail length, TDM)	Age, BMI, abstinence, smoking, drinking, grilled and smoked foods ingestions	Multivariate regression models	2-OHNa levels were associated with increased comet parameters including tail% (β = 13.26% per log unit 2-OHNa; 95%CI: 7.97–18.55), tail length (β = 12.25; 95%CI: 0.01–24.52) and tail distribution (β = 7.55; 95%CI: 1.28–18.83); 1-OHP was associated with increased tail% (β = 5.32; 95%CI: 0.47–10.17); urinary PAH metabolites were associated with decreased vital Annexin V negative sperm count;	Single urine sample; the use of biomarkers didn’t allow for determination of primary exposure sources;	Urinary PAH metabolites as biomarkers of PAH exposure; Assessing 4 metabolites individually; Participants recruited only during the winter when air pollution is higher); many confounders including in the analyses;
	1-OHP
	9-OHPh
	2-OHFIu
	2-OHNa
Hammoud, et al. 2010 [20]	PM2.5 (μg/m3)	Motility, concentration (×106/mL), morphology (normal forms %)	Temperature, season	Multivariate regression models	PM2.5 was negatively associated with sperm motility 2 months and 3 months following the recording of the PM2.5 values (p = 0.010 and p = 0.044, respectively)	Participants from andrology laboratory; occupational exposure didn’t include in the analysis	Repeated semen samples; big sample size; The analyses of daily levels of PM2.5 over the 5 years;
Hansen, et al. 2010 [19]	O3 (ppb)	Concentration (× 106/mL), count, morphology (normal forms %), abnormal morphology (%), abnormal head (%), abnormal midsection (%), abnormal tail (%), cytoplasmic droplets (%), CMA (%), DFI (%)	Age, abstinence, education levels, smoking, season, temperature	Multivariate regression models	Exposures to O3 or PM2.5 at least below the current National Ambient Air Quality Standards were not associated with decrements in sperm outcomes.	Small sample size; single semen sample;	Many confounders including in the analysis;
	PM2.5 (μg/m3)
Xia, et al. 2009a [28]	4 PAH metabolites in urine	Volume, concentration (×106/mL), sperm number per ejaculum, sperm motility	Age, abstinence time,	Logistic regression analysis	Men with higher urinary concentrations of 1-OHP, 2-OHP and Sum PAH metabolites were more likely to have idiopathic male infertility (p for trend = 0.034)	Single semen sample; No investigation of the association between PAH exposure and semen morphology;	Big sample size; Urinary PAH metabolites as biomarkers of PAH exposure; men with idiopathic infertility were divided into ‘normal’ and ‘abnormal’ semen quality group
	1-N
	2-N
	1-OHP
	2-OHF
Xia, et al. 2009b [29]	4 PAH metabolites in urine	Volume, concentration (× 106/mL), sperm number per ejaculum, sperm motility	No information-based on abstract	No information-based on abstract	Men with higher 1-OHP (assessed as quintiles) were likely to have below-reference sperm concentration and sperm number per ejaculum.	No information-based on abstract	No information-based on abstract
	1-N
	2-N
	1-OHP
	2-OHF
Rubes, et al. 2007 [32]	SO2 (μg/m3)	%DFI GSTM1 genotype	Smoking	Mixed models	Association between GSTM1 null genotype and increased %DFI (beta = 0.309; 95% CI: 0.129, 0.489). Furthermore, GSTM1 null men also showed higher %DFI in response to exposure to intermittent high air pollution (beta = 0.487; 95% CI: 0.243, 0.731)	Small sample size; Single semen sample; only one confounder included to the analysis;	Novel evidence for a gene-environment interaction between GSTM1 and air pollution;
	NOX (μg/m3)
	PM10 (μg/m3)
	PAH (ng/m3)
Sokol, et al. 2006 [18]	O3 (ppb)	Concentration (×106/mL), motility (×106)	Temperature, seasonality, age of donation, date of birth	Linear mixed-effects model	Negative association between ozone and sperm concentration (p < 0.01);	Small sample size; No investigation of the association between air pollution exposure and semen morphology;	Repeated semen samples;
	NO2 (ppb)
	CO (ppm)
	PM10 (μg/m3)
Rubes, et al. 2005 [26]	SO2 (μg/m3)	Count, concentration (× 106/mL), volume, motility (%), normal sperm head morphology (%), normal morphology (%), straight line velocity, curvilinear velocity, linearity, DFI%	Smoking, drinking, caffeine, abstinence, fever, briefs,	Mixed models for repeated measures	Significant association between high air pollution and %DFI (β = 0.19; 95%CI: 0.02–0.36); Other semen measures were not associated with air pollution;	Small sample size;	Urine sample were assayed for cotinine to confirm self-reported smoking status; Blood sample was collected for analysis of lead, mercury and cadmium as an indication of possible exposure to metals; Air pollution categorized as low and high; semen samples classify as a “winter” or “summer” sample
	NOx (μg/m3)
	PM10 (μg/m3)
	PAH (ng/ m3)
Selevan, et al. 2000 [17]	PM10 (μg/m3)	Semen volume, concentration (×106/mL), total count, motility (%), total progressive, normal morphology (%), normal heads (%), VSL, VCL, LIN, sperm chromatin structure	Abstinence, wearing briefs, caffeine, high fever, work and hobbies with metal, work and hobbies with solvents, season	Multivariate regression models	Significant associations between medium air pollution and %motile sperm (β = −8.12; 95%CI: −12.95, −3.30); between medium air pollution and progressive motility (β = −0.15; 95%CI: −0.30, −0.01); Negative association between medium and high air pollution and %normal morphology (β = − 0.42; 95%CI: − 0.69, − 0.14 and β = − 0.84; 95%CI: − 1.15, − 0.53); High air pollution was associated with parameters of sperm motion – VSL, VCL, LIN (p < 0,005)	Single semen sample; Data about air pollution from the air monitoring program;	Air pollution categorized as low, medium and high; Many confounders included to the analysis
	SO2 (μg/m3)
	CO (mg/m3)
	NOx (μg/m3)
Robbins, et al. 1999 [33]	SO2	aneuploidy	Alcohol, caffeine intake, fever, laboratory variables	Poisson regression modeling	The sex chromosomal aneuploidy YY was associated with exposure to high air pollution IRR = 5.25, 95%CI: 2.5,11.0	No information-based on abstract	No information-based on abstract
Occupational exposure
Calogero, et al. 2011 [37]	NOx (μg/m3)	LH, FSH, T, sperm concentration (×106/mL), total sperm count, total motility, progressive motility (%), normal form (%), sperm chromatin integrity; DFI (%)	Age, length of occupational exposure, smoking	Multivariate regression models;	Motorway tollgate workers had a significantly higher %spermatozoa with damage chromatin (p < 0.001) compared with controls, likewise the %spermatozoa with fragmented DNA; late sign of apoptosis was also significantly higher in tollgate workers (p < 0.001); sperm concentration, total sperm count, total and progressive motility, and normal form were significantly lower in tollgate workers compared with controls (p < 0.05)	Small sample size;	blood levels of MHb, SHb, COHb, Pb as a biological biomarkers of environmental pollution; the air pollution was measured with specific analyzers 24 h/day for 30 days during summer;
	SOx (μg/m3)
Boggia, et al. 2009 [38]	NO2 (μg/m3)	FSH, LH, T, sperm count, motility, morphology,	Fuel combustion gases		Negative association between men occupationally exposed to NO2 and total motility (p < 0.05)	No individual estimations of exposure;	Cross-sectional design
Guven, et al. 2008 [36]	diesel	Concentration (×106/mL), motility, morphology	–	student’s t-test, Mann-Whitney U-test;	The differences regarding the abnormal sperm count and motility were significant between the study group and control group (p = 0.002 and p = 0.003, respectively); the ratio of sperm cells with normal morphology was significantly lower in the study group (p = 0.001)	Small sample size; No individual estimations of exposure;	Cross-sectional design
De Rosa, et al. 2003 [39]	CO (mg/m3)	FSH, LH, T, sperm count, volume, motility, morphology, sperm membrane function, forward progression, sperm kinetics: VSL, VCL, LIN, ALH,		student’s t-test; linear regression analysis	Total motility, forward progression, functional test and sperm kinetic were significantly lower in tollgate workers vs. controls (p < 0.0001)	Small sample size; No individual estimations of exposure	blood levels of MHb, SHb, COHb, Pb, Zn as a biological biomarkers of environmental pollution;
	NO (mg/m3)
	SO (μg/m3)
	Pb (μg/m3)

PM10 particulate matter < 10 μm, PM2,5 particulate matter < 2,5 μm, O₃ ozone, SO_x sulphur oxides, CO carbon monoxide, NO_x nitrogen oxides, PAH polycyclic aromatic hydrocarbons, 1-OHP 1-hydroxypyrene, Pb lead, Cd – cadmium, 1-N 1-hydroxynapthalene, 2-N 2-hydroxynapthalene, 1-OHP 1-hydroxypyrene, 2-OHF 2-hydroxyfluorene, VCL curvilinear velocity, VSL straight line velocity, VAP average path velocity, BCF beat cross frequency, ALH amplitude of lateral head displacements, %DFI percentage of sperm with fragmented DNA, %HDS high DNA stainability, LIN linearity, STR straightness, grade A sperm with progressive motility, grade B nonlinear motility, grade c non-progressive motility, TDM tail distributed moment, %CMA percentage of sperm chromatin maturity, LH luteinizing hormone, FSH follicle-stimulating hormone, T testosterone, E₂ estradiol, MHb methahaemoglobin, SHb sulphahaemoglobin, COHb carboxyhaemoglobin, IDW interpolation inverse distance weighting (Shepard’s method)