Association of peripubertal blood lead levels with reproductive hormones and semen parameters in a longitudinal cohort of Russian men

Abstract

STUDY QUESTION

Are peripubertal blood lead levels (BLLs) associated with semen parameters and serum reproductive hormones among young Russian men?

SUMMARY ANSWER

We observed a suggestion of lower ejaculate volume with higher peripubertal BLL but no associations of BLLs with reproductive hormones measured throughout adolescence or with other sperm parameters measured at adulthood.

WHAT IS KNOWN ALREADY

Lead is a known reproductive toxicant and endocrine disruptor. Previous literature has shown associations between high lead exposure and poorer semen quality both in occupationally and environmentally exposed men. However, to our knowledge, no longitudinal studies have explored the association of childhood lead exposure with semen parameters and reproductive hormones in young men.

STUDY DESIGN, SIZE, DURATION

The Russian Children’s Study is a prospective cohort study that enrolled 516 boys at age 8–9 years in 2003–2005 and followed them annually for 10 years. BLLs were measured at entry and lifestyle and health questionnaires were completed. Reproductive hormones were measured in blood samples collected every 2 years.

PARTICIPANTS/MATERIALS, SETTING, METHODS

Among the 516 boys enrolled, 481 had BLLs measured at entry. Of these, 453 had at least one measurement of serum testosterone, follicle stimulating hormone (FSH) or luteinizing hormone (LH) (median = 5 samples per boy) and 223 had semen samples collected ∼10 years after enrolment. Semen assessment included ejaculated volume, sperm concentration, progressive motility and total sperm count, and parameters were categorized using published andrology standards for low semen quality based on sperm count and motility. Linear mixed models were used to examine the associations of log-transformed BLLs (and BLL categories) with reproductive hormones and semen parameters, adjusting for potential confounders.

MAIN RESULTS AND THE ROLE OF CHANCE

Among the 223 young men with peripubertal BLLs and at least one semen sample (total samples = 438), the median (interquartile range) BLL was 3 (2, 5) µg/dl and 27% had BLL ≥5 µg/dl. Overall, 49% of the semen samples fell below reference levels for sperm count and/or motility. Men with peripubertal BLL ≥5 µg/dl had significantly lower ejaculated volume than those with BLL <5 µg/dl (mean = 2.42 vs 2.89 ml, P = 0.02), but this difference was attenuated in adjusted models (mean = 2.60 vs 2.83 ml, P = 0.25). No associations were observed between BLL measured at age 8–9 years and reproductive hormone levels or sperm parameters, including sperm concentration, total count, progressive motility and total progressive motile sperm count, or with the probability of having low semen quality based on sperm count/motility.

LIMITATIONS, REASONS FOR CAUTION

Only a subset of the original cohort participated in the semen quality portion of the study, although inverse probability weighting was used to account for possible selection bias. BLLs were only measured at a single time in peripuberty, and other exposure time periods, including later or longer-term childhood exposure, may be more predictive of semen quality. The young men were also exposed to other chemical contaminants before and during pubertal development.

WIDER IMPLICATIONS OF THE FINDINGS

While semen volume often receives less attention than other sperm parameters, it is an important component of male fertility. Additional prospective studies covering different exposure windows and including other seminal plasma biomarkers are warranted to explore our finding of potentially lower ejaculated volume with higher BLLs and to confirm the lack of associations for other semen parameters among youth exposed to environmental BLLs.

STUDY FUNDING/COMPETING INTEREST(s)

Funding was provided through grants R01ES0014370 and P30ES000002 from the National Institute of Environmental Health Sciences, grant R82943701 from the U.S. Environmental Protection Agency, and grant 18-15-00202 from the Russian Science Foundation (O.S and Y.D.). All authors report no competing interests.

TRIAL REGISTRATION NUMBER

N/A.

Introduction

Toxic effects of lead (Pb) exposure on male reproductive health have been noted for decades, with both animal and epidemiological studies in occupational settings demonstrating associations of higher lead levels with reduced growth, delayed pubertal maturation and decreased semen quality (Schwartz et al., 1986; Shukla et al., 1991; Kim et al., 1995; Ballew et al., 1999; Williams et al., 2010). Animal studies have indicated that lead exposure decreases concentrations of growth hormone, insulin-like growth factor 1 (IGF-1), testosterone and other hormones responsible for growth and pubertal development, suggesting an endocrine mechanism for its effects on male reproductive health (Ronis et al., 1996, 1998a, 1998b; Sokol et al., 2002).

For the general population, the most common routes of lead exposure are diet, lead-containing plumbing and paints, historical use of leaded gasoline, tobacco smoking and other industrial activities (Benoff et al., 2000; Tong et al., 2000; Hsu et al. 2009; Liu et al., 2010; Apostolou et al., 2012). Lead exposure can also occur in occupational settings. Airborne exposures to particulates containing the lead ion Pb²⁺ have been documented in metalworkers, smelters and alkaline battery production workers, among others (Benoff et al., 2000). Despite initiatives to decrease the lead exposure by eliminating leaded gasoline, implementing community water monitoring programs, remediating environmental lead contamination, and changing national guidelines for defining levels of concern, lead levels have remained persistently high in some subgroups and contribute to issues of environmental injustice (Tong et al., 2000; Brulle and Pellow, 2006; Jones et al., 2009; Pasetto et al., 2019). Declines in semen quality have been observed over the past century, and environmental exposures including metals may have contributed to these declines (Swan et al., 2000; Hauser et al., 2003; Woodruff et al., 2010; Levine et al., 2017; Meeker et al., 2008; Mínguez-Alarcón et al 2018).

Epidemiological literature has demonstrated associations between high lead exposure and poorer semen quality in occupationally exposed men (Lancranjan et al., 1975; Tuohimaa and Wichmann, 1985; Assennato et al., 1987; Lerda, 1992; Alexander et al., 1996, 1998; Bonde and Kolstad, 1997; Robins et al., 1997; Telisman et al., 2007; Hosni et al., 2013). For example, among smelter workers, total sperm count and concentration were observed to decrease with increasing blood lead levels (BLLs) (Alexander et al., 1996, 1998). Occupational lead exposure has also been associated with other sperm alterations, including asthenospermia, hypospermia and teratozoospermia (Lancranjan et al., 1975). However, these occupational studies have evaluated men with very high levels of lead exposure (often >50 µg/dl). A small number of studies have explored the association of lower lead levels measured in blood and/or semen with semen parameters in men without occupational exposure, and while less consistent, some have observed adverse reproductive effects including reduced sperm count and abnormal sperm morphology (Telisman et al., 2000, 2007; Jurasović et al., 2004; Meeker et al., 2008; Wirth and Mijal, 2010; Wu et al., 2012; Fumarewa and Ugawuja, 2017; Ren et al., 2020). However, these studies have all been cross-sectional and have often recruited men from infertility clinics. No large prospective studies, which we are aware of, have been conducted in the general population, and this has been identified as a critical research gap (Wirth and Mijal, 2010).

Using data from a longitudinal cohort of Russian boys residing in an industrial region with elevated background lead exposures (Snakin and Prisyazhnaya, 2000; Rubin et al., 2002), we have previously reported associations of higher peripubertal BLLs with decreased growth, reduced IGF-1 and delays in pubertal onset and maturation (Sokol et al., 2002; Fleisch et al., 2013; Burns et al., 2017; Sergeyev et al., 2017; Williams et al., 2019). Although remediation efforts in the study area have resulted in decreased levels of many environmental contaminants (Sergeyev et al., 2017), a high percentage of children in certain industrial Russian cities had elevated BLLs in the late 1990s (23% with BLL > 10 µg/dl) (Rubin et al., 2002). In our full cohort in which samples were collected in 2003–2005, the median BLL was 3 µg/dl, with 28% ≥5 µg/dl. These levels are higher than those in the USA (1.3% with BLL > 5 µg/dl in 2003–2006) and Germany (geometric mean = 1.57 µg/dl in 2003–2006), but lower than those in China (43% >5 µg/dl in 2000–2004) (Schulz et al., 2007; Li et al., 2014; Egan et al., 2021). Sources also differed, with less evidence of lead exposure via lead-based paint or leaded water pipes in Russia compared to the USA; the primary sources in Russia continue to be lead-contaminated soil and dust from industrial and automotive emissions (Rubin et al., 2002). In addition to concerns regarding the effects of lead exposure, we have reported that higher peripubertal serum dioxin concentrations were prospectively associated with poorer semen parameters in this cohort (Mínguez-Alarcón et al., 2017). Motivated by our own previous findings suggesting a strong rationale for potential effects of lead on both hormone levels and semen quality, we evaluated whether higher levels of peripubertal BLLs measured in Russian boys at age 8–9 years were associated with reproductive hormones measured every 2 years throughout the study follow-up and with semen parameters measured ∼10 years later, by leveraging the longitudinal Russian Children’s Study (RCS) cohort study.

Materials and methods

Study population

The RCS is a prospective cohort study of 516 boys in Chapaevsk, Russia, enrolled in 2003–2005 at ages 8–9 years, as described in detail elsewhere (Williams et al., 2010; Burns et al., 2017; Sergeyev et al., 2017). The study was approved by the Human Studies Institutional Review Boards of the Chapaevsk Medical Association (Chapaevsk, Russia), Harvard T.H. Chan School of Public Health, Brigham and Women’s Hospital (Boston, MA, USA) and Nemours Health Care System (Wilmington, DE, USA). The parent or guardian signed an informed consent and the boy signed an assent form. At study entry, the boy’s mother or guardian completed a nurse-administered health and lifestyle questionnaire that included birth and medical history, demographic characteristics and socio-economic status indicators. Each boy had blood drawn for the measurement of BLLs. A validated Russian Institute of Nutrition semi-quantitative food frequency questionnaire was used to ascertain the child’s dietary intake (Martinchik et al., 1998). Participants underwent a complete physical exam at entry, and this initial assessment was followed by yearly physical exams and questionnaires as previously described (Williams et al., 2010; Burns et al., 2017; Sergeyev et al., 2017). Once participants were 18 years of age or older, they signed a consent form for their own participation, including a separate consent for providing semen samples. After restricting to 481 of the 516 boys (93%) enrolled in the RCS with baseline lead measures, we evaluated reproductive hormones within a subset of 453 boys (92%) with at least one measurement of testosterone, FSH or LH during follow-up, and separately evaluated semen quality in a subset of 223 men (46%) who had semen samples collected at age 18 years or older and were not excluded due to chronic health conditions, which could impact semen quality (see Supplementary Fig. S1).

Blood lead assessment

A 3.0-ml venous blood sample was collected at study entry in a trace metal-free vacutainer tube (Becton-Dickinson, Franklin Lakes, NJ, USA) after cleaning the site with alcohol. Whole blood samples were diluted with a matrix modifier solution and analysed by Zeeman background-corrected flameless graphite furnace atomic absorption (ESA Laboratories, Chelmsford, MA, USA). The limit of detection (LOD) was 1.0 µg/dl and measured BLLs below the LOD were set to 0.5 µg/dl (Cole et al., 2009); this represented 2.9% of boys in the overall study population, and 3.6% of the 223 included in the current analysis.

Assessment of semen parameters

Each participating young man was asked to contribute a semen sample upon reaching 18 years of age or older, with a recommended sexual abstinence of 2–7 days prior to semen collection. After consent as described above, they each provided up to two semen samples (∼1 week apart) by masturbation inside a dedicated room next to the study Andrology Lab. The samples were stored inside an incubator at 37°C. Semen parameters were evaluated within 1 h after sample collection at maximum, although most samples (99%) were analysed within a half hour of the collection. A technician was present on the premises and recorded self-reported information regarding any viral/bacterial illness or fever in the months prior to the semen collection. Abstinence time was calculated based on the date/time of the last recalled ejaculation and the date/time when the participant provided their sample. All samples were assessed by a single andrology technician and analysed according to criteria of the Nordic Association for Andrology (NAFA) and European Society of Human Reproduction and Embryology-Special Interest Group in Andrology (ESHRE-SIGA) (Kvist and Björndahl, 2002; Björndahl et al., 2010; WHO, 2010). The technician was blinded to the BLLs of the participants providing semen samples.

Semen volume was measured using a 1–10-ml disposable pipette. Sperm motility was evaluated by microscopic examination of the semen sample (at least 200 sperm) in duplicate at 400 times magnification. Motility was reported following the 1999 WHO manual for the examination and processing of human semen, with the classification of samples as rapidly progressive motile (Class A), slowly progressive motile (Class B), locally motile (Class C) or immotile (Class D). The progressive percent motile sperm was calculated by summing the individual percentages of the WHO Classes A and B (Björndahl et al., 2010). Sperm concentration was quantified using an Improved Neubauer Chamber Hemacytometer at 200 times magnification under a phase contrast microscope. Total sperm count (volume × sperm concentration) and total progressive motile sperm count (total sperm count × % progressive motile sperm) were calculated. Duplicates for sperm concentration and motility were assessed to ensure that difference between duplicates did not exceed the acceptance limit; the within-observer mean coefficient of variation (CV) in duplicates was 6.4% for sperm concentration and 4.9% for progressive motility.

Quantification of hormone levels

Fasting whole blood samples were collected biennially at scheduled study visits. Collection, sample processing and analysis took place at the Chapaevsk Medical Association laboratory, located at the Chapaevsk Hospital in Russia. Immediately after collection, the blood was centrifuged and the serum fraction was aliquoted into 0.8–2 ml of samples and then frozen at a temperature of −80°C. For the measurement of hormones, serum was thawed at room temperature and analysed in batches of 25–35 samples within 130 min of thawing. Serum levels of testosterone, FSH, LH and sex hormone binding globulin (SHBG) were analysed using the Architect i1000SR and chemoluminescent microparticle immunoassay (Abbott Laboratories, Lake Bluff, IL, USA). All analyses were conducted by the same laboratory technician blinded to lead levels. Detection limits were 0.05 nmol/l for total testosterone, 0.05 mIU/ml for FSH, 0.01 mIU/ml for LH and 0.1 nmol/l for SHBG. For each assay, calibration curves were generated using manufacturer-supplied calibrators containing low, medium and high hormone levels. The mean intra-assay CVs were 2.7% for total testosterone, 1.6% for FSH, 3.0% for LH and 5.0% for SHBG. As an additional QA/QC, we analysed three levels of manufacturer-supplied control samples on a daily basis, measuring a different level each day (MCC Liquichek Immunoassay Plus Quality Control, Bio-Rad Laboratories, USA). The mean intra-assay CVs for all three levels were 5.4% for total testosterone, 4.3% for FSH, 5.2% for LH and 6.2% for SHBG.

Statistical analysis

We calculated medians and interquartile ranges (IQR) for continuous demographic, reproductive and dietary characteristics as well as counts and percentages for categorical variables, and we descriptively compared these characteristics by higher vs lower BLL (<5 vs ≥5 µg/dl). We also compared background characteristics between the participants included as compared to those enrolled but not included in our analysis, to assess representativeness of the sample. Semen parameters were reported as medians (IQR), and as the number (%) meeting criteria for poor semen quality recommended by Björndahl et al. based on lower sperm counts (≤80 million) and lower total motility (≤60%) (Kvist and Björndahl, 2002; Björndahl et al., 2010). The distribution of BLLs was summarized as median (IQR) and as the percent with low, moderate or high (≤2, 3–4, ≥5 µg/dl) BLLs.

Our primary analyses of reproductive hormones and semen parameters compared boys with ‘higher’ (≥5 µg/dl) vs ‘lower’ (<5 µg/dl) peripubertal BLLs; secondary analyses were also conducted using the three categories of low, moderate and high BLLs noted above and by treating log-transformed BLL as a continuous predictor. Total sperm count, sperm concentration and total progressive motile sperm count were log-transformed to more closely approximate a normal distribution, while semen volume was approximately normally distributed and not transformed. Linear regression models were used to examine the relation between BLLs and semen parameters with a random subject effect to account for repeated semen samples within the same man. In addition to evaluating each semen parameter individually, we also fit a log-binomial model using generalized estimating equations to estimate the relative risk of low semen quality for sperm counts and total motility as defined by NAFA-ESHRE, accounting for the correlation between multiple semen samples per participant. Inverse probability weights (IPW) were included in models to account for potential selection bias in the subset of eligible young men with available semen samples. Variables included in the IPW models included boy’s BLLs, birthweight, total caloric intake and percent intake of carbohydrates and fats, breastfeeding duration and parental education. Predicted marginal means for the semen parameters were estimated, adjusted for confounders at the mean level for continuous variables, weighted according to their frequencies for categorical variables and back-transformed to allow presentation of results on the original scale. Tests for linear trends were conducted using the three categories of BLLs as ordinal levels.

Potential confounders included in the models were selected primarily based on a priori evidence from the literature and supported empirically by associations in our cohort (Weng et al., 2009; Evans et al., 2012). Based on these criteria, we included boys’ birthweight (kg), breastfeeding duration (weeks), total caloric intake (kcal), monthly household income (<5000 rubles/month) and height z-score at enrolment. We also included abstinence time (<2, 2–5 and ≥5 days) in adjusted models regardless of statistical significance, as a known predictor of semen parameters. To assess the robustness of the findings, we conducted a sensitivity analysis excluding boys who were diagnosed with cryptorchidism (n = 4), varicocele (n = 7) or orchiditis (n = 1).

We evaluated the association between BLLs and hormone levels using mixed effect linear regression models for the log-transformed hormone levels. The correlations between each boy’s repeated hormone measures were modelled using a heterogeneous Toeplitz or first-order autoregressive covariance structure, as guided by goodness-of-fit criteria for each separate hormone. Models were fit both for log-transformed BLL and for higher vs lower BLL. All longitudinal hormone models were fit as a function of age throughout follow-up (allowing higher order age terms for flexibility) and adjusted for boys’ baseline BMI z-score, total caloric intake and percent calories from fat at enrolment, birthweight, breastfeeding history, highest parental education at baseline, presence of biological father in the home and maternal alcohol consumption during boys’ gestation. Models for total testosterone were additionally adjusted for SHBG. Graphical summaries were created to illustrate trends in hormone levels between ages 9 and 18 years. All analyses were conducted using SAS (version 9.4; SAS Institute Inc., Cary, NC, USA), and two-sided P-values ≤0.05 were considered statistically significant.

Results

We evaluated semen quality in a subset of 223 of the 516 boys (43%) initially enrolled and followed in the RCS, as summarized in Supplementary Fig. S1. The demographic and dietary characteristics of the 223 young men included are summarized in Table I, overall and by BLL. All participants were Caucasian males, with a median age at time of semen collection of 18.5 years (IQR: 18.1–19.1) and a median BMI of 20.9 kg/m² (IQR: 19.0–23.1). The median peripubertal BLL measured at entry (age 8–9 years) was 3 µg/dl (IQR: 2–5); 26% had BLL≤ 2 µg/dl, 47% had BLL 3–4 µg/dl and 27% had BLL of at least 5 µg/dl. Six boys (3%) had BLL ≥ 10 µg/dl (max = 17 µg/dl). Boys with higher BLL at entry tended to be shorter, had longer duration of breastfeeding, and more often had lower socioeconomic status as reflected by parental education and household income than those with lower BLL. The background characteristics and BLLs of those included in the analysis were similar to those not included, although those included were, on average, breastfed for a longer duration and weighed more at birth (Supplementary Table SI). In the overall study population of all 481 boys with blood lead measured, the median BLL was 3 µg/dl, with 28% ≥ 5 µg/dl (range: 0.5–31 µg/dl) (Williams et al., 2010; Burns et al., 2017).

Table I.

Demographics, dietary and parental characteristics among 223 young men in the Russian Children’s Study, overall and by BLL.

	Overall N = 223	Higher BLL (≥5 µg/dl) N = 60	Lower BLL (<5 µg/dl) N = 163
Demographic and reproductive characteristics at semen collection
Age, years	18.5 (18.1, 19.1)	18.6 (18.2, 19.2)	18.5 (18.1, 19.1)
Body Mass Index (BMI), kg/m2	20.9 (19.0, 23.1)	20.9 (19.0, 23.2)	20.9 (19.0, 23.1)
Smoking*, n (%)	109 (49)	31 (52)	78 (49)
Beer intake, n (%)	139 (63)	41 (69)	98 (61)
Liquor intake, n (%)	70 (32)	22 (37)	48 (30)
History/current cryptorchidism, n (%)	4 (2)	1 (2)	3 (2)
History/current varicocele, n (%)	7 (3)	1 (2)	6 (4)
History/current orchiditis, n (%)	1 (1)	0 (0)	1 (1)
Perinatal characteristics
Breastfeeding duration, weeks	17.2 (8.7, 39.0)	21.7 (8.7, 52.0)	13.0 (8.7, 34.7)
Gestational age, weeks	40 (38, 40)	39 (38, 40)	40 (39, 40)
Birth weight, kg	3.40 (3.14, 3.70)	3.30 (2.90, 3.60)	3.50 (3.20, 3.70)
Nutritional and Sociodemographic Characteristics at entry
Age, years	8.2 (8.0, 9.0)	8.1 (8.0, 9.0)	8.9 (8.1, 9.0)
Body Mass Index, kg/m2	15.4 (14.4, 16.6)	15.3 (14.5, 16.5)	15.4 (14.4, 16.7)
Height Z-score	0.19 (-0.59, 0.75)	−0.17 (-0.96, 0.31)	0.38 (-0.37, 0.81)
Low income (<5000 rubles/mon), n (%)	74 (33)	26 (43)	48 (29)
Parental education, n (%)
High school or less	11 (5)	6 (10)	5 (3)
Technical degree/some college	131 (61)	34 (59)	97 (61)
College graduate or higher	76 (35)	18 (31)	58 (36)
Household smoking during pregnancy, n (%)	31 (14)	6 (10)	21 (15)
Total calorie intake (kcal)	2788 (2139, 3585)	3177 (2329, 3783)	2660 (2093, 3495)
Carbohydrates (% calories)	53.9 (49.6, 58.1)	53.7 (50.1, 57.2)	54.1 (49.3, 58.4)
Fat (% calories)	34.6 (30.9, 37.6)	34.9 (31.5, 37.4)	34.5 (30.5, 37.7)
Protein (% calories)	11.4 (10.5, 12.5)	11.6 (10.2, 12.6)	11.4 (10.6, 12.5)

Values presented are medians and interquartile ranges (25^th, 75^th percentiles) unless otherwise noted.

^*Smoking status was based on the response to the question: ‘Have you smoked a cigarette, even a few puffs, within the past year?’.

Of the 223 young men, 215 (96%) provided two semen samples and 8 (4%) provided one sample (total N = 438 samples). The median (IQR) values for semen parameters are summarized in Table II. The median (IQR) abstinence time was 2.71 days (1.88, 3.96). Overall, 49% of semen samples were below NAFA-ESHRE reference cut-offs for sperm count and/or motility.

Table III.

Adjusted mean semen parameters (and 95% CIs) and prevalence of low semen quality by blood lead levels (BLLs) among 223 young men contributing 438 semen samples in the Russian Children’s Study.

BLL (µg/dl)	Semen Volume (ml)	Sperm Concentration (mill/ml)	Total Sperm Count (mill)	Progressive Sperm Motility (%)	Total Progressive Motile Sperm Count (mill)	Low semen quality (probability)
Higher vs Lower Blood Lead Levels
Lower (<5 µg/dl)	2.83 (2.61, 3.06)	47.0 (41.3, 53.4)	111 (95.6, 129)	53.2 (51.7, 54.7)	57.7 (48.9, 68.1)	0.51 (0.44, 0.58)
Higher (≥5 µg/dl)	2.60 (2.27, 2.93)	49.0 (37.8, 63.4)	107 (80.0, 143)	53.1 (50.9, 55.2)	55.7 (40.6, 76.4)	0.49 (0.39, 0.59)
P-value	0.25	0.78	0.83	0.90	0.85	0.76
Blood Lead Categories
Low [≤2 µg/dl]	2.92 (2.50, 3.34)	41.3 (33.2, 51.3)	99 (76.4, 128)	51.2 (48.6, 53.9)	49.4 (36.8, 66.2)	0.43 (0.31, 0.55)
Medium [3-4 µg/dl]	2.79 (2.52, 3.06)	50.3 (42.8, 59.0)	118 (98.5, 141)	54.3 (52.5, 56.1)	62.6 (51.3, 76.5)	0.55 (0.46, 0.63)
High [≥5 µg/dl]	2.60 (2.27, 2.93)	49.1 (38.0, 63.6)	107 (80.3, 143)	53.2 (50.9, 55.3)	56.0 (40.8, 76.8)	0.49 (0.39, 0.59)
P-trend	0.24	0.33	0.68	0.29	0.57	0.43
Continuous Blood Lead (Log-transformed)
β (95% CI)	−0.40 (−0.82, 0.03)	0.09 (−0.13, 0.31)	−0.02 (−0.27, 0.23)	1.77 (−0.55, 4.08)	0.01 (−0.27, 0.29)	0.20 (−0.22, 0.65)
P−value	0.06	0.43	0.85	0.14	0.94	0.35

Estimates presented as marginal means (95% CI), adjusted for boys’ birthweight, total caloric intake, height z-score at entry, breastfeeding duration, monthly household income and abstinence time. Beta estimates for continuous blood lead estimates reflect changes in log-transformed measures for sperm concentration, total sperm count and motile sperm count.

Table II.

Semen parameters and reproductive characteristics among 223 young men (18–19 years) contributing 438 semen samples in the Russian Children’s Study.

	Mean (SD)	Median (IQR)
Mean testicular volume by Prader orchidometer (ml)	27.4 (4.2)	30.0 (25.0, 30.0)
Ejaculated volume (ml)	2.75 (1.64)	2.40 (1.60, 3.50)
Sperm concentration (mil/ml)	67.9 (58.2)	53.5 (28.3, 88.5)
Total sperm count (mil/ejaculate)	174 (165)	128 (63.2, 230)
Progressive sperm motility (%)	53.1 (10.2)	55.0 (48.0, 60.0)
Total progressive motile count (%/ejaculate)	96.5 (92.4)	71.6 (31.5, 131)
Abstinence time (days)	4.92 (10.9)	2.71 (1.88, 3.96)

Men with peripubertal BLL ≥5 µg/dl had significantly lower ejaculated semen volume than those with BLL <5 µg/dl (mean = 2.42 vs 2.89 ml, P = 0.02), reflecting a mean decline of 16% (Supplementary Table SII). However, after adjusting for abstinence time and other potential confounders, this difference was attenuated to an 8% decrease in volume and no longer significant (adjusted mean = 2.60 vs 2.83 ml, P = 0.25) (Table III). Similarly, while semen volume decreased over increasing categories of BLL (adjusted means of 2.92 ml for low, 2.79 ml for moderate and 2.60 for high BLLs), this declining trend did not reach significance. When considering BLLs as a continuous exposure measure, associations were slightly stronger and each log unit increase in BLL was associated with an adjusted mean decline in semen volume of 0.40 ml (95% CI: −0.82, 0.03 ml). No associations were observed of BLLs with other semen parameters, including sperm concentration, total sperm count, progressive motility and total motile sperm count, nor the probability of having low semen quality (Table III). In a sensitivity analysis excluding 12 boys diagnosed with cryptorchidism, varicocele or orchiditis, results were similar to those of the primary analyses (Supplementary Table SIII).

Among the 453 boys with serum sex hormone measurements, levels were very similar for boys with higher and lower BLLs throughout study follow-up from age 8 to 18 years (Fig. 1). In linear mixed effect models for repeated sex hormone measures, adjusting for age and other covariates, we observed no association of either log-transformed BLLs or higher vs lower BLLs with total testosterone, FSH or LH (Supplementary Table SIV).

Figure 1.

Distribution of reproductive hormones by age for boys with lower (<5 µg/dl) and higher (≥5 µg/dl) blood lead levels (BLLs).

Discussion

In this longitudinal study of 223 young Russian men contributing over 400 semen samples, we explored associations between peripubertal BLLs measured at 8–9 years of age and semen parameters measured in young adulthood ∼10 years later. We found a suggestion of lower ejaculated semen volume with higher BLLs, but this finding was not consistently observed when considering increasing BLL categories. No other associations were found between BLLs and any other semen parameters including sperm concentration. In addition, there was no association between peripubertal BLL levels and reproductive hormones: testosterone, FSH and LH. This lack of association was observed for log-transformed BLLs as well as categorical BLL measures, both in fully adjusted models and in models which adjusted only for age.

Our findings differ from those of occupationally exposed cohorts, such as smelter and battery workers, which have consistently demonstrated associations of higher lead levels with impaired semen quality (Lancranjan et al., 1975; Tuohimaa and Wichmann, 1985; Assennato et al., 1987; Lerda, 1992; Alexander et al., 1996, 1998; Robins et al., 1997; Hosni et al., 2013). Several cross-sectional studies of men with infertility have also indicated associations of BLLs with lower semen quality, particularly for progressive motility which has been suggested to be especially sensitive to the effects of lead exposure (Li et al., 2015).

The lack of a clear relationship between lead levels and semen parameters in our study may be multifactorial. First, the BLLs were much lower than in the occupational cohorts. While the BLLs themselves were consistent with studies conducted in infertility clinics, the lead measurements were obtained prior to puberty and were not contemporaneous with the semen analysis. Second, the 10-year gap between the peripubertal lead measurements and semen quality may allow repair of damage or attenuation of lead effects or may contribute to misclassification of exposure given the relatively short half-life of lead in blood (about 35–40 days). Third, the recurring differentiation and maturation of mature sperm during the normal seminiferous cycle might ameliorate the reproductive toxicity of childhood exposures. Environmental levels of lead in this community were thought to decline during the 10 years of this study and declining lead exposures closer to the time of semen sample collection might have tempered the earlier effects of the peripubertal levels.

Other studies evaluating men with lower levels of blood lead, reflective of those more commonly observed in the general population, have also sometimes failed to detect associations with semen quality, and a comprehensive review by Wirth and Mijal also noted such inconsistencies (Meeker et al., 2008; Wirth and Mijal, 2010). A recent study in 941 Chinese men reported associations of higher urinary lead levels with lower sperm concentration, sperm count and progressive motility, as well as decreased testosterone and serum follicle stimulating hormone (Ren et al., 2020). However, when evaluating associations in the two separate subgroups of 751 college students and 190 community members, the authors only observed associations of urinary Pb with sperm motility among community members; the high percentage of urinary Pb samples below the LOD (47%) and assignment of these measures to the LOD/√2 may have adversely impacted the robustness of these analyses (Cole et al., 2009). Finally, while use of BLLs is generally considered the preferred biomarker of short-term inorganic Pb exposure, several studies have observed associations of adverse semen parameters with seminal Pb concentrations but not with blood Pb (Hernandez-Ochoa et al., 2005; Mendiola et al., 2011; Sukhn et al., 2018), and seminal plasma concentrations have been suggested as having a stronger correlation with semen parameters because they represent levels in the target organ (Mínguez-Alarcón et al., 2012).

Despite the lack of observed associations in our study, there is justification for concern based on experimental animal and invitro studies, which have demonstrated strong evidence of a detrimental effect of lead exposure on the male reproductive system and semen quality (Ronis et al., 1996, 1998a, 1998b). In addition to the observations of decreases in sperm motility and increased percentages with abnormal morphology in animal studies of lead exposure, increases in DNA and chromatin structure damage have also been reported (Hernandez-Ochoa et al., 2005; Li et al., 2018). These effects are thought to stem from disruption of the hypothalamus–pituitary–gonadal axis, similar to other endocrine disruptors, resulting in altered levels of reproductive hormones including follicle stimulating hormone, testosterone and growth hormone (Zawatski and Lee, 2013; Mitra et al., 2017). However, extrapolating results from animal studies to humans is always limited by physiological differences among species, exposure periods and doses, all of which lead to variation of the effect; this is particularly true of reproductive cycles, which differ among species in timing and pace. Genetic polymorphisms and epigenetic mechanisms may also contribute to variability in adverse outcomes in response to lead exposure (Mitra et al., 2017). In addition, some of the inconsistencies in reproductive health effects observed for lead exposure in human studies may be attributable to differences in assessing occupationally vs non-occupationally exposed populations.

While not statistically significant in adjusted models, we observed a suggestion of lower semen volume with higher BLLs. Even though low semen volume is often overlooked, particularly when other semen abnormalities are present, it is an important component of male infertility and has been linked to higher risk of mortality among infertile men (Hauser and Sokol, 2008; Roberts and Jarvi, 2009; Eisenberg et al., 2014). Previous findings among men from the RCS cohort showed that higher peripubertal serum organochlorine pesticide concentrations were associated with lower semen volume and progressive motility, while other organochlorine compounds such as dioxins have shown strong associations with lower sperm concentration and count, but not with ejaculated volume or progressive motility (Mínguez-Alarcón et al., 2017; Abou Ghayda et al., 2020). One possible reason why associations were not observed in this study could be that the young men were exposed before and during pubertal development to multiple other contaminants (not related with lead exposure), which are stronger drivers of semen quality than the lead levels measured in this cohort.

Our study has both strengths and certain limitations inherent to its observational design. Limitations included the fact that only a subset of the original cohort participated in the semen quality portion of the study, and that BLLs were only measured at a single time point, which may not have reflected the critical window of exposure. In addition, a single peripubertal BLL does not capture other potentially critical exposure windows or long-term childhood lead exposures (e.g. ascertainable via serial blood lead measures or bone lead levels), which may be more predictive of childhood lead toxicities that manifest at a later time (Needleman et al., 2002; Coscia et al., 2003). The main strength of this study is its prospective design, limiting the possibility of reverse causation. Other strengths include the analysis of the semen samples by a single technician who was blinded to the BLLs, enhancing consistency of semen parameter measurement and reducing the possibility of observer bias. The collection of two semen samples for most men in the RCS also reduced measurement error and provided a more robust design. We also were able to adjust for a wide range of potential confounders and used appropriate statistical methods including inverse probability weighting to account for the possibility of selection bias. These findings thus contribute to the existing literature, which has rarely included prospective studies of reproductive effects of childhood lead exposure. Further questions remain unanswered, such as whether lead exposure may be associated with an altered profile of proteins and other important components of seminal fluids needed for sperm competition and capacitation, or whether other exposure time periods, including longer-term childhood lead exposure, are more predictive of semen parameters (Puga Molina et al., 2018; Samanta et al., 2018). Thus, additional longitudinal studies investigating long-term effects of childhood lead exposure on the male reproductive system are still needed.

Conclusion

In this longitudinal Russian cohort, peripubertal BLLs measured at 8–9 years showed no significant association with reproductive hormones measured throughout adolescence or semen parameters measured ∼10 years later, after adjustment for important confounders. Additional prospective studies, including those evaluating different time windows of exposure, are warranted to confirm the lack of associations with semen quality for young men exposed to lower levels of lead during childhood. In addition, despite the largely null findings of this study, our previously demonstrated adverse effects of higher BLLs at ages 8–9 years on impaired growth and alterations in timing of pubertal onset and sexual maturity, along with a large literature on adverse effects of lead in children, warrant concern. In this context, screening of BLLs in mid-childhood and adolescence has the potential to identify youth or even communities that might benefit from interventions to reduce BLLs.

Data availability

The data underlying this article cannot be shared for ethical and privacy reasons.