Associations between occupation exposure to Formaldehyde and semen quality, a primary study

Formaldehyde (FA), a ubiquitous environmental pollutant, has long been suspected of having male reproductive toxicity. However, FA male reproductive toxicity was inconclusive due to dearth of human studies. Therefore, we sought to investigate whether occupational exposure to FA affects semen quality. Semen quality including five conventional parameters and seven kinematics parameters were compared between 114 male workers occupationally exposed to FA and 76 referents. FA exposure index (FEI) was measured and calculated. Our results showed that sperm progressive motility, total sperm motility, VCL, VSL and VAP were statistically significant decreased in FA exposure workers compared with the referents. Moreover, FEI was significantly negative associated with sperm progressive motility (β = −0.19, P = 0.01) and total sperm motility (β = −0.23, P = 0.004). In addition, a significant elevated risk of abnormal sperm progressive motility were observed in both low- (OR = 2.58; 95% CI: 1.11–5.97) and high-FA-exposed group (OR = 3.41; 95% CI: 1.45–7.92) respectively. Furthermore, a significant increased risk was also estimated for abnormal total sperm motility in both low- (OR = 3.21; 95% CI: 1.24–8.28) and high-FA-exposed group (OR = 4.84; 95% CI: 1.83–12.81) respectively. In conclusion, our study revealed the adverse effects of FA occupation exposure on semen quality, especially on sperm motion parameters.


Materials and Methods
Subject recruitment. The experimental protocols were approved by the Institutional Medical Ethics Committee of Xi'an Jiaotong University. The experiment methods were carried out in accordance with the approved guidelines. A recruitment campaign to enroll participants from seven wood processing industries (plywood production) was organized by Pathology Department of Xi'an Jiaotong Unversity. The purpose of this study was informed to all subjects, and written informed consent forms were obtained from all subjects.
Inclusion criteria for FA exposure group were as follows: men aged from 23 to 40 years old at the time of inclusion; Chinese Han ethnicity; and men who had worked in the FA exposure environment for at least 24 months. On the other hand, men who had lived in a newly built or recently decorated house; men who had genital malformations or other chronic diseases were excluded form the study. At the same time, age-matched male Han population volunteers who had lived in same place of residence were chosen as reference group with respect to educational level and socioeconomic status. For the reference group, we only selected those who exposure to FA or other reproductive toxicants was avoided. The reference group mainly consisted of salesmen and clerks.
Questionnaire. According to the methods used in our previous study 26 , a semi-structured interview questionnaire was completed by all participants, including baseline information about sociodemographics (age, nationality, education level, and income), lifestyle (e.g. smoking status and alcohol consumption), abstinence duration, previous or current diseases, and occupational exposure (time and duration of occupational exposure). All participants claimed that their lifestyles and environments had not changed for at least 6 months prior to semen collection.
Physical examination. A physical examination, including height and weight, was performed. Body mass index (BMI) was calculated for every subjects. In addition, two infertility specialists performed the genital examination of all subjects; the possible presence of genital malformations and diseases was investigated.
Scientific RepoRts | 5:15874 | DOi: 10.1038/srep15874 FA exposure Assessment. According to the method of our previous study, with minor modification 26 , FA exposure index (FEI) was measured and calculated for every worker. Firstly, the workers were asked to give information about their workplace, work tasks, work duration and time. Secondly, the FA concentration in the workplace of each worker was measured by stationary sampling on the day when the investigation was done. In detail, the formaldehyde detector (4160-2; Interscan, Chatsworth, CA) was stable for about 15 minutes in the environment before the zero adjustment. Next zero was set at the sampling mode, and then removed off the filter. The stable reading on the screen is the on-site FA concentration. The total sampling time in single measurement lasts about 25 minutes, and we monitored the FA concentration at three different time points (9:00 AM, 12:00 AM and 15:00 PM, respectively) during one workday. The mean value of the three measurements was used as FA exposure concentration of the worker. Due to the limitations of the measurement time points and frequency, no data were available for potential short-term exposure peaks. Finally, FEI was calculated as follows: FA exposure index (FEI) = concentration of FA (mean value of three measurements, in mg/m 3 ) × exposed work time during a workday × exposure duration (year). According to the FEI level, two exposure classes were further defined so that the number of men in the low and high FA-exposed groups had the ratio of 1:1.
Semen collection and analysis. All subjects were not taken any medical or surgical treatments in the three months prior to the semen examination. The semen analysis was performed within 2 weeks after the FA exposure measurements. Subjects were requested to provide a semen sampling after 2-7 days of abstinence. Semen samples were collected by masturbation in a sterile wide-mouthed calibrated container and semen were immediately analyzed within 60 min after collection. All the semen analyses were conducted by two well-trained technicians using the same apparatus in a blinded manner, in order to reduce the variance of assessment of sperm characteristics.
Semen analyses were performed using a computer-automated semen analysis system (CASA) (WLJY-9000, China). According to the manufacturer's guidelines, freshly collected semen samples were allowed to liquefy for 20 min at 37 °C. Conventional semen parameters included semen volume, sperm concentration, total sperm count, sperm progressive motility and total sperm motility was assessed. Moreover, the seven kinematic parameters were detected according to the WHO laboratory manual (2010) 27  Statistical analysis. At first, a descriptive analysis was performed following data collection. Means and standard deviations were calculated for continuous variables with checking for normality. Frequency was evaluated for the different categorical variables. Moreover, because conventional semen parameters follow markedly skewed (non-normal) distribution, the 25th, median and 75th percentiles were computed.
Secondly, sociodemographic characteristics and semen parameters were compared between the FA exposure groups and the referent group by using t-test or One-way ANOVA for normally distributed continuous variables, Kruskal-Wallis test for non-normally distributed continuous variables, and Chi-Square-Fisher's exact test for categorical variables respectively.
Thirdly, we used multiple linear regression models to estimate the associations between FA exposure and semen parameters. Semen parameters were log-normal (In) transformed to improve the normality as dependent variables in the linear models. In addition, logistic regression analysis was used to further calculate the crude and adjusted odds ratios (ORs) of abnormal semen parameters with 95% confidence intervals (CIs). A full model that included all possible confounding factors to be examined in regression was used. Selection of confounding factors for the final model was based on their importance in the literature and biological plausibility [28][29][30] , the variables entered into the regression model included age, body mass index, education, income, smoking, drinking and abstinence duration.
All statistical analyses were carried out using SPSS statistical software version 16.0 (SPSS, Chicago, USA), and P < 0.05 was considered statistically significant.

Results
Subjects characteristics. Totally 124 (62.3%) male worker occupationally exposed to FA were recruited and eligible for the study whereas 75 workers refused to provide semen. However, in the phase of semen collection, 8 workers were excluded for failing to collect their semen samples and 2 workers were excluded for spillage of the sample semen. Finally, a total of 114 male worker occupationally exposed to FA were eligible and completed all the steps of the study. On the other hand, 81 referents (40.5%) were recruited and screened for the study. Of these, 3 referents were excluded for failing to collect the semen samples and 2 were excluded for spillage of the sample semen. Finally, a total of 76 eligible referents participated in this study.
By interviewing, the participants in the FA exposure and reference group were not engaged in heavy physical labor, and didn't work and live in high temperature environment. In addition, the participants didn't experience extra mental stresses in both groups. Table 1 shows the sociodemographic characteristics and abstinence duration of FA exposure workers and referents. Compared with the referents, the FA-exposed subjects were similar in age, body mass index (BMI), education level, income, cigarette intake, alcohol consumption and duration of abstinence. The mean age in both groups was around 30, the average body mass index (BMI) was about 24, and the majority of the subjects had completed primary education. The average monthly income in both groups was around 1800 RMB. More than half of the subjects used tobacco and alcohol in both groups. In addition, the mean duration of abstinence was about 4 days in both FA exposure group and reference group.
FA exposure level. FA concentrations in 76 referents ranged from 0.0 to 0.02 mg/m 3 . These values can be neglected because they were far lower than FA occupation exposure limit. Moreover, the concentrations of FA measured in air ranged from 0.22 to 2.91 mg/m 3 in all 114 workers. The FA exposure index (FEI) calculated in each FA-exposed workers ranged from 4.54 to 195.08. We further divided the workers into low and high-FA exposed groups in accordance with the median level (56.55) of FEI so that the number of workers in two FA-exposed groups had the ratio of 1:1. Semen parameters. The outcome of semen conventional parameters was summarized in Table 2. No statistically significant differences were found in semen volume, sperm concentration, total sperm count between FA exposure groups and reference group. However, the results clearly suggest that there has been a statistically significant decrease in sperm progressive motility (38.6% and 36.0% vs. 49.9%) and sperm total motility (49.8% and 43.8% vs. 59.5%) in low-and high-FA exposure groups when compared with reference group. Moreover, these reductions are in a dose-dependent trend.
In addition, the data of seven semen kinematic parameters were reported in Table 3. There were no significant differences in terms of LIN, STR, MAD and ALH between FA exposure group and reference group. However, there was statistically significant decrease in VCL, VSL and VAP in FA exposure groups when compared with reference group, where a dose-dependent trend was also observed.
Associations between FA exposure and semen quality. To clarify the effects of FA exposure on semen quality, the associations between FA exposure index (FEI) and semen parameters (conventional and kinematic parameters) was calculated by multiple linear regression and the results were summarized in Table 4 and 5 respectively. After adjusting for confounding factors, FEI showed significantly negative associations with sperm progressive motility (β = − 0.19, P = 0.01) and total sperm motility (β = − 0.23, P = 0.004) , respectively. Table 6 provides information on crude and adjusted ORs for the risks between abnormal (below-normal values) semen parameters and FA occupational exposure. After correction for confounding factors, a significant elevated risk of abnormal sperm progressive motility were found in low-FA-exposed group (OR = 2.58; 95% CI: 1.11-5.97) and high-FA-exposed group (OR = 3.41; 95% CI: 1.45-7.92) respectively compared with the reference group. Moreover, a significant increased risk was also observed for abnormal total sperm motility in low-FA-exposed group (OR = 3.21; 95% CI: 1.24-8.28) and high-FA-exposed Characteristics Reference group (n = 76) Low FA-exposed group (n = 57) High FA-exposed group (n = 57)    Table 4, Supplementary Table 5 and  Supplementary Table 6 respectively. In addition, slightly increased risks of abnormal semen volume, sperm concentration and total sperm count was found in FA exposure groups, but this association was not reached statistical significance even after adjusting for confounding factors (Table 6).

Discussion
Our study is the first to reveal the adverse effects of FA occupation exposure on semen quality parameters in a Chinese population. Our study found that semen parameters included sperm progressive motility, total sperm motility, VCL, VSL and VAP were statistically significant decreased in FA exposure groups at a dose-dependent trend compared with reference group. Moreover, personal FEI (FA exposure index) showed significantly negative associated with sperm progressive motility (β = − 0.19, P = 0.01) and sperm total motility (β = − 0.23, P = 0.004) after adjusting for confounding factors. In addition, compared with the reference group, a significant elevated risk of abnormal sperm progressive motility were observed in both low-(OR = 2.58; 95% CI: 1.11-5.97) and high-FA-exposed group (OR = 3.41; 95% CI: 1.45-7.92) , respectively, after correction for confounding factors. Furthermore, a significant increased  Table 5. Associations of Formaldehyde exposure index (FEI) with semen kinematic parameters before and after adjusting for confounding factors. Regression coefficients were adjusted for age, cigarette intake, alcohol consumption, body mass index, income, education and abstinence duration.

Reference group(n = 76)
Low-FA-exposed group (n = 57) High-FA-exposed group (n = 57)  Table 6. ORs and 95% CI for below-normal values of semen parameters associated with FA occupational exposure before and after adjusting for confounding factors. OR: odd ratio; CI: confidence interval. Note: abnormal sperm parameters were defined by the World Health Organization (WHO, 2010) standards: semen volume < 1.5 ml, sperm concentration < 15 × 10 6 /ml, total sperm count < 39 × 10 6 / ml, total sperm motility < 40%, and progressive motility < 32%. Regression coefficients were adjusted for age, cigarette intake, alcohol consumption, body mass index, income, education and abstinence duration. *P < 0.05. risk was also estimated for abnormal total sperm motility in both low-(OR = 3.21; 95% CI: 1.24-8.28) and high-FA-exposed group (OR = 4.84; 95% CI: 1.83-12.81) after correction for confounding factors. During the last few decades a possible degradation in human semen quality has been studied intensively and many studies have suggested the declining trend of sperm quality worldwide [5][6][7][8] . With increased evidence of a decline in semen quality in recent years, public, government and scientific concerns about the effects of environmental changes on male reproductive health have grown to become a major preoccupation in many countries 31,32 . As one of the ubiquitous environmental contaminant in the world, although preventive measures aimed at reducing FA contaminant have been implemented, exposure to FA remains one of the most prominent environmental health problems 33,34 . The detrimental effects of FA on male reproduction have also been shown in many animal experiments [16][17][18][19][20][21][22][23] . However, due to species differences, the result cannot be directly extrapolated to humans. FA male reproductive toxicity was inconclusive due to the scare of detailed data in human studies. Our present results suggested the relationships between FA exposures and human sperm quality parameters. Our findings further increase the objective evidence to support the hypothesis that FA exposure has adverse effect on male reproduction.
There is a widespread human exposure to FA, the principal exposure route in environment is through inhalation. Historically, occupational exposure has been the dominant source of FA exposure in China 33 . Current Occupational Exposure Limit (OEL) for FA 35 in China is 0.5 mg/m 3 , and some countries employ time-weighted averages (TWA) and/or short-term exposure limits (STEL). For example, the USA's OEL 36 is 0.75ppm (0.92 mg/m 3 , 8-h TWA) with a STEL of 2ppm (2.46 mg/m 3 ), while the United Kingdom 37 is 2ppm (2.46 mg/m 3 ) for both TWA and STEL. Although the low OEL for FA in China compared with other countries, FA exposure levels have been high across many different Chinese industries. For example, Wong et al. 38 reported that the average FA concentration in the wood processing industry was 3.07 ± 5.83 mg/m 3 . Fan et al. 39 found the FA concentrations in several hospital pathology laboratories were as high as 5.84 mg/m 3 . The investigation of Zhang et al. 40 showed that the FA concentration in some anatomy laboratories was above 10 mg/m 3 . In the present study, we found that the concentrations of FA measured in air ranged from 0.22 to 2.91 mg/m 3 in 114 workers of wood processing industry, among of them, part FA concentrations are far higher than OEL standards in China (0.5 mg/m 3 ). We speculated that the wood processing industry has the high FA concentration, which is caused in part by unventilated workshops and a lack of employee safety precautions.
Semen quality is one of the most valuable indications of male reproductive health, and it is associated with fertility status 29,41 . In our study, semen quality parameters included both semen conventional parameters and kinematic parameters were all analyzed by Computer-Assisted Sperm Analysis (CASA). CASA not only decreases the time spended in sperm observation, but reduces intra-observer differences and improve the accuracy of final results. CASA method 42 makes the assessment of semen quality more objective, detailed and reproducible.
In our study, semen volume, sperm concentration and total sperm count were not markedly declined in FA occupational exposure groups. However, semen motion parameters included sperm progressive motility, total sperm motility, VCL, VSL and VAP were statistically significant decreased in FA exposure groups compared with the reference. Our results indicated sperm motility might be more sensitive to environmental FA exposure than other semen parameters. Our present findings is not consistent with the Finland study (only 11 autopsy worker) that reported that no effects on sperm were seen from FA or its metabolites in this population after occupational exposure 24 . However, the decline of the motility parameters due to FA exposure has been highlighted in several animal experiments which found FA caused regressive histological changes in the seminiferous tubules and epididymis resulting in the suppression of sperm maturation and motility 18,19,21 .
Sperm motility is one of the most important indicators in assessing sperm quality 41 , it is also an important indicator for sperm fertilization. The decline in the sperm motion parameters may reflect decreased sperm functions that allow the sperm to reach the oocyte to complete fertilization 43 . Hirano et al. (2001) 44 reported that VCL, VSL, VAP have been correlated with fertilization rates in vivo and may be bioindicators of the fertilizing ability of human sperm. Our earlier study 26 found that the wives of male workers occupationally exposed to FA have increased risk of prolonged TTP (Time-to-pregnancy) and spontaneous abortion. Together with our current findings, we speculated adverse reproductive outcomes such as prolonged TTP and spontaneous abortion might be attributed to the decline of sperm motion ability. Some studies showed strong evidence that FA itself does not reach systemic circulation 45,46 . The pathogenic effect of FA exposure on sperm motility might be a result of free radical-generated oxidative stress and a distriburbance of the redox equilibrium [20][21][22][23]47 . Oxidative stress promotes a dose-dependent increase in tyrosine nitration and S-glutathionylation and alters motility and the ability of spermatozoa to undergo capacitation 48 . However, the underlying molecular mechanisms of FA exposure on sperm were still remaining unclear.
Moreover, in this study, we use FA exposure index (FEI) to reflect FA exposure level. FEI not only reflect FA exposure concentration, but also combined with exposure duration. This indicator is more precise to reflect the actual FA occupational exposure level. Our results showed that FEI was significantly negative associated with semen motion parameters included sperm progressive motility (β = − 0.19, P = 0.01) and sperm total motility (β = − 0.23, P = 0.004). That is, a 10-fold increase in FEI was found to be associated with a 1.9-fold decrease in sperm progressive motility, and a 2.3-fold decrease in total sperm motility. In addition, compared with the referent, a significant elevated risk of abnormal sperm progressive motility were observed in both low-(OR = 2.58; 95% CI: 1.11-5.97) and high-FA-exposed group (OR = 3.41; 95% CI: 1.45-7.92) respectively. Furthermore, a significant increased risk of abnormal total sperm motility in both low-(OR = 3.21; 95% CI: 1.24-8.28) and high-FA-exposed group (OR = 4.84; 95% CI: 1.83-12.81) were also detected. Our results further demonstrated the male reproductive toxicity of FA exposure might be dose-dependent.
In this study, we applied some methods to overcome the potential bias in different phases of this study. At first, in the phase of subject recruitment, strict inclusion criteria and exclusion criteria were used to screen subjects. Except the similar ethnic origin, we recruited only those subjects who had worked in the FA exposure environment for at least 24 months. Since too short exposure time was not suitable to evaluate the association between FA occupational exposure and reproductive health. In addition, modern home building and furnishings are the main sources of indoor FA pollution globally, newly built residences and recently remodeled apartment release high levels of indoor FA. Therefore, to reduce the influence of family environmental FA pollution, we excluded the subjects who live in a newly built residences and recently remodeled apartment. On the other hand, to decrease the selection bias, we chose reference group with respect to similar educational level, socioeconomic status and residential environment. Secondly, in the process of FA exposure assessment, we measure the air concentration of FA in occupational environment and calculate FA exposure index (FEI) according to exposure duration and concentration. This approach relies on personal work area monitoring data to construct an exposure assessment of FA, it is more precise to reflect the true FA occupational exposure level. Thus, it reduces the potential for exposure misclassification. Thirdly, in the process of semen analysis, semen quality parameters included both semen conventional parameters and kinematic parameters were all analyzed by Computer-Assisted Sperm Analysis (CASA) by two well-trained technicians. CASA method makes the semen assessment more objective, detailed and reproducible 42 .
Although we used many methods to reduce bias, we realize that our study was not without limitations. A potential limitation of this study was the fact that only FA concentration was measured and evaluated for each subject. In the wood processing industry, except mainly exposure to FA, there may be exposure to other organic solvents such as phenols and wood preservatives 49 . Phenol has been reported to have genotoxicity and induced sister chromatid exchange in human cells 50 . In addition, occupational exposure to organic solvents has been related to low motile sperm count in men 51 . Therefore, in the future study the concentration of other organic solvents should be monitored and calculated.

Conclusions
The present study suggested the adverse effects of FA occupation exposure on semen quality, especially on sperm motion parameters in a dose-dependent trend, which might increase the objective evidence to support the hypothesis that FA exposure has negative effects on male reproduction. Given the importance of human reproductive health and the current wide usage of FA, it is valuable to further investigate the correlations between FA exposure and semen quality in a large cohort.