Investigating the association between alcohol intake and male reproductive function: A current meta-analysis

Background Alcohol use and alcohol-related health problems are on the rise in developing countries. This meta-analysis was conducted to determine the effects of alcohol consumption on human male reproductive function through semen parameters, antioxidants in semen, sperm DNA fragmentation, and sex hormones. Methods Studies regarding the effects of alcohol consumption on male reproductive function were searched on databases. Based on the random-effects model, STATA software was used to analyze and synthesize the selected studies. Alcoholics, moderate alcoholics, heavy alcoholics, and no alcoholics values were compared using the standard mean difference. Publications were assessed for publication bias by the Egger test. Result Forty studies were selected from databases examining the effect of alcohol consumption on male reproductive health in 23,258 people on five continents of the world. The meta-analysis revealed that alcohol intake reduced semen volume during each ejaculation (SMD = −0.51; 95% CI −0.77, −0.25). However, there were no significant associations with other semen indicators such as density, mobility, and normal and abnormal sperm count from this analysis. In addition, drinking alcohol lowered antioxidant enzymes in semen (SMD = −7.93; 95% CI −12.59, −3.28) but had no effect on sperm DNA fragmentation. Finally, the results showed a decrease in general testosterone levels (SMD = −1.60; 95% CI −2.05, −1.15), Follicle Stimulating Hormone (SMD = −0.47; 95% CI −0.88, −0.05), Luteinizing Hormone (SMD = −1.35; 95% CI −1.86, −0.83), but no effect in other sex hormones named as estradiol, Inhibin B and Sex Hormone-Binding Globulin. Furthermore, when analyzing subgroups at different drinking levels, the results showed that the moderate alcoholic group (less than 7 units/week) had no change in the semen index. Meanwhile, the group of heavy alcoholics (more than 7 units/week) harmed the semen index and sex hormones, especially by increasing estradiol. Conclusion There is evidence that alcohol consumption affected semen volume and antioxidant, reproductive hormones thus negatively affecting male reproductive function. This study might be necessary to make recommendations regarding alcohol consumption for men.


Introduction
The use of alcoholic beverages in daily life is considered a cultural feature of many countries in the world. They were used in several religions because they bring excitement and addiction to consumers [1]. According to a recent WHO study, one of the alarming current issues is that while alcohol use is declining in developed countries, it is increasing in developing countries [2][3][4]. From 1993 to 2000, alcohol consumption per capita among Vietnamese adults increased by roughly 2.5 times [5]. Excessive or long-term alcohol consumption has negative consequences not only for the person who consumes it but also for the family and society [6][7][8][9]. Reports suggest that unhealthy alcohol use affects nearly every organ system [10][11][12][13][14].
Alcohol consumption affects the entire hypothalamic-pituitary-gonadal region of the male reproductive system and disrupted the production of Gonadotropin-Releasing Hormone (GnRH), Follicle Stimulating Hormone (FSH), and Luteinizing Hormone (LH), the natural oestradiol levels due to changes in free testosterone, impaired function of Leydig and Sertoli cells, resulting in reduced sperm quality to normal morphology and sperm maturation [15][16][17]. Consumption of alcohol with a high concentration and frequency has led to negative effects on the concentration of sperm; normal morphology of spermatozoa; total sperm count, reduced testosterone and SHBG, and increased serum testosterone levels [18,19]. Scientific evidence shows that it causes impotence, infertility, and spermatogenic arrest (Pajarinen JT and Karhunen 1994, [20][21][22][23][24][25]. Many studies showed negative impacts based on evidence that drinking alcohol leads to an increase in enzymatic antioxidants, sperm DNA damage [26], and high levels of DNA fragmentation [27]. However, several reports give the opposite findings. Barratt et al. have shown that moderate alcohol consumption did not affect semen quality despite higher testosterone levels [28]. The reports of Teijon et al. and Hansen et al. show no statistically significant difference in semen parameters between the alcoholics and no-alcoholics [29] and no association between alcohol consumption and sperm quality [15]. Moreover, a meta-analytical conducted by Ying Li et al. found drinking alcohol did not affect sperm parameters [20].
For a more general and accurate assessment of existing information, we conducted a meta-analysis of published literature that assessed the effects of alcohol intake and male reproductive function. This report shows a general opinion on drinking alcohol and parameters related to male reproductive quality and provides recommendations on alcohol consumption with concentrations and levels defined.
The selection criteria include articles in English, having full text, and providing data of interest. Exclusion criteria reviewed articles that provided no data or only in the form of graphs, cell, and animal studies. Studies were selected when they: (a) evaluated the association between alcohol consumption and male reproductive health parameters; (b) defined drinking levels; (c) effect of alcohol consumption on sex hormones; (d) represented original data; (e) used a cohort study design, and (f) were written in English. Papers were excluded when they: (a) described a review, case report, or conference abstract; (b) did not contain original data; (c) did not provide data.

Statistical analysis
A meta-analysis of the studies was done using the statistical analysis software STATA 15 based on the random effects model (Random Effects Model). Heterogeneity between studies was assessed through the heterogeneity index I-squared. The values groups were compared using the standard mean difference (SMD). Publications are evaluated for publication bias using the Egger test.

Subgroup analysis
We employed subgroup analysis by drinking levels definition (according to alcohol concentration and frequency consumption). For that there are three groups of people based on their drinking habits: no-alcoholics if they drink less than 1 unit of alcohol a week; T. Nguyen-Thanh et al. moderate alcoholics if they use at least 1 unit/week to 7 units/week; and heavy alcoholics when they take more than 7 units of alcohol/ week. One unit of alcohol is equivalent to 7.9 g of ethanol or 10 mL of ethanol contained in an oral solution [33]. Thus, 1 unit of alcohol is equivalent to 3/4 bottles/cans of 330 mL beer (5% alcohol); 1 cup of draft beer 330 mL; 1 glass of wine 100 mL (13.5% alcohol); or 1 cup of spirits 30 mL (40% alcohol).

Characterization of eligible studies
Studies were searched on scientific databases MEDLINE, EMBASE, and scientific websites. The process of searching and filtering the studies for the meta-analysis is shown in Fig. 1. The results of 1296 studies were found, of which 818 were omitted because the title and abstract did not match the selection criteria to conduct the review. The remaining 478 full-text articles continued to be evaluated. Of these, 438 articles were excluded for reasons such as duplicate content, some reviews, animal studies, in vitro studies, studies with graphical data only, and articles that had not provided complete data. In the end, we obtained 40 studies that were eligible for inclusion in the systematic review, of which 20 reported on alcohol consumption and semen volume (mL), 25 articles included data on sperm density (million/mL), 24 articles had information on sperm motility of (percentage of motile sperm), 14 articles show results on sperm morphology. The remaining studies involved alcohol use and other indicators such as DNA fragmentation (9 studies), testosterone (18 studies), estradiol (9 studies), FSH (13 studies), LH (13 studies), inhibit B (5 studies), SHBG (6 studies) [15,18,26, ( Fig. 1 and Table 1).

Association between alcohol consumption and semen parameters
The effects of alcohol consumption on semen volume were analyzed across 18 studies involving 17,144 male adults are shown in Fig. 3a. The studies were analyzed based on the random-effects model. The difference between studies was large, with the I-squared heterogeneity index being 96.9%. Each study was weighted (%weight) from 3.33 to 6.22. The results of the meta-analysis showed that the standard mean difference (SMD) between the no-alcoholics and alcoholic groups was − 0.51 (95% CI: − 0.77, − 0.25). This result suggests that drinking alcohol reduces the volume of semen in each ejaculation. Fig. 3b presents the results of the assessment of publication bias in publications on the effects of alcohol consumption on semen volume using the Egger test. The results showed no publication bias with p for bias being 0.068. Fig. 4 presents an analysis of the effect of alcohol consumption on sperm density (23 studies, with 19,436 men), motility (22 studies, with 18,549 men), normal morphology (12 studies, with 14,296 men), and abnormal morphology (9 studies, with 2071 men). The results of the analysis showed that the SMD between the no-alcoholics and alcoholics groups was 0.15 (95% CI: − 0.14, 0.43), sperm motility 0.11 (95% CI: − 0.17, 0.40), SMD of normal sperm morphology is − 0.43 (95% CI: − 1.35, 0.49). SMD of abnormal sperm morphology was 0.10 (95% CI: − 0.58, 0.77). This result shows that drinking alcohol did not significantly affect the parameters of sperm density, motility, normal morphology, and abnormal morphology.

Associations between alcohol consumption and antioxidant enzymes, sperm DNA fragmentation
The results of an analysis of 1400 men from seven studies on the effects of alcohol consumption on antioxidant enzymes are shown in Fig. 5a. The studies were analyzed based on the random-effects model. The studies have a large variation in research results, the heterogeneity index I-squared heterogeneity was 96.9%. The results of the meta-analysis show that the difference in SMD between the no-alcoholics and alcoholics groups was − 7.93 (95% CI: − 12.59, − 3.28). This result suggests that drinking alcohol reduces antioxidant enzymes in semen. Fig. 5c presents the results of the assessment of publication bias in publications on the impact of alcohol consumption on antioxidant enzymes using the Egger test. The results show no publication bias with p for bias was 0.303.
The effect of alcohol intake on sperm DNA fragmentation in 1883 men from 9 meta-analyzed studies (Fig. 5b). The studies were analyzed based on the random-effects model. The studies have a large variation in research results, the heterogeneity index I-squared heterogeneity was 96.9%, The results of the meta-analysis show that the standard means difference in SMD between the drinking and no-drinking groups was 0.59 (95% CI: − 0. 17, 1.34). This result shows that drinking alcohol had not to change the sperm DNA fragmentation index. Fig. 5d presents the results of the assessment of publication bias in publications on the impact of alcohol consumption on sperm DNA fragmentation using the Egger test. The results show publication bias with p for bias being 0.005.   4. Non-observed associations between alcohol consumption and sperm concentration, motility, normal and abnormal morphology. a. Forest plot for the effect of alcohol consumption and sperm density, assessment of publication bias using the Egger publication bias plot; b. Forest plot for the effect of alcohol consumption and sperm motility and assessment of publication bias using the Egger publication bias plot; c. Forest plot for the effect of alcohol consumption and normal morphology and assessment of publication bias using the Egger publication bias plot; d. Forest plot for the effect of alcohol consumption and abnormal morphology and assessment of publication bias using Egger publication bias plot.

Associations between alcohol consumption and sex hormones
The results of the study on the effects of alcohol consumption on the hormone testosterone from 17 studies involving 13,373 men were meta-analyzed and presented in Fig. 6a. The studies were analyzed based on the random-effects model. The studies have a large variation in research results, the heterogeneity index I-squared heterogeneity was 96.9%, The results of the meta-analysis show that the standard means difference in SMD between the drinking and non-drinking groups was − 1.60 (95% CI: − 2.05, − 1.15). These results suggest that alcohol consumption lowers the testosterone hormone. Fig. 6c presents the results of the assessment of publication bias in publications on the effects of alcohol consumption on testosterone using the Egger test. The results show publication bias with p for bias was 0.004.
The results of the estradiol hormone survey on 4143 men from 8 studies show that the alcoholic group did not change the estradiol hormone compared with the no-alcoholic group with SMD values of − 0.03 (95% CI: − 0.47, 0.41) in Fig. 6b. Evaluation of publication bias in publications on the impact of alcohol consumption on the hormone estradiol using the Egger test. The results show no publication bias p for bias was 0.871 in Fig. 6d.
Results of the FSH hormone survey in 9375 men from 12 studies and LH hormone in 11,458 men from 12 studies show that the standard means difference in SMD were − 0.47; (95% CI: − 0.88, − 0.05) and − 1.35; (95% CI: − 1.86, − 0.83), respectively. This suggests that drinking alcohol lowers the hormones FSH and LH ( Fig. 7a and b). Evaluation of publication bias in publications on the effects of alcohol consumption on hormones FSH and LH by the Egger test. The results showed no publication bias for both FSH or LH with p for bias being 0.472 and 0.097, respectively ( Fig. 7c and d).
Inhibin B hormone survey from 5 studies including 10,782 men and SHBG2 hormone from 6 studies including 11,215 men are shown in Fig. 8. The standard means difference SMD for inhibin B was 0.04 (95% CI: − 0.08, 0.16); SMD for SHBG2 was − 0.05; (95% CI: − 0.31, 0.2). Thus, using alcohol has not changed the hormone Inhibin B and SHBG2 ( Fig. 8a and b). The assessment of publication bias in the publications on the impact of alcohol consumption on inhibin B and SHBG2 using the Egger test. The results show no publication bias for both inhibin B and SHBG2 with p for bias being 0.275 and 0.474, respectively ( Fig. 8c and d).

Subgroup analysis of the associations between alcohol consumption at different levels on male reproductive function
Subgroup analysis was conducted to assess the impact of different levels of alcohol consumption on male fertility. The results presented in Table 2 show that the alcoholic group (drinking at least 1 unit of alcohol per week) had a decrease of semen volume with the SMD was − 0.51 (95% CI: − 0.77, − 0.25); the antioxidant with SMD was − 7.93 (95% CI: − 12.59, − 3.28); testosterone with SMD was − 1.60 (95% CI: − 2.05, − 1.15); FSH with SMD was − 0.47 (95% CI: − 0.88, 0.05); LH with SMD was − 1.35 (95% CI: − 1.86, − 0.83) compared to the no-alcoholic group. Meanwhile, the moderate alcoholic group (who have 1-7 units of alcohol per week) had no statistical impact on seminal parameters (volume, density, mobility, normal morphology) and sperm abnormalities as well as sex hormones (testosterone, estradiol, FSH, LH, Inhibin B, SHBG2) in comparison to the no-alcoholic group. Comparing the heavy alcoholic group (consuming more than 7 units a week) and the no-alcoholic group, there were significant associations on male reproductive function, likely to reduce the volume of semen per ejaculation with SMD was − 0.48 (95% CI: − 0.86, − 0.11); testosterone with SMD was − 1.92 (95% CI: − 2.70, − 1.13); FSH with SMD was − 0.77 (95% CI: − 1.47, − 0.06), but increased estradiol with SMD was 0.22 (95% CI: 0.00, 0.44); and increased LH with SMD was 0.06= (95% CI: 0.01, 0.12). When analyzing two subgroups of the heavy alcoholic group and the moderate alcoholic group, there was no impact on seminal parameters, but the heavy alcoholic group was shown to maintain a significant association with the increase in estradiol (SMD = 0.54 (95% CI: 0.06, 1.02)) and LH (SMD = 0.22 (95% Fig. 6. The observed association of the decreased hormone testosterone and non-observed association of hormone Estradiol with Alcohol consumption. a. Forest plot for the effect of alcohol consumption and the hormone testosterone; b. Forest plot for the effect of alcohol consumption and the hormone β-Estradiol; c. Evaluation of publication bias about the hormone testosterone using Egger publication bias plot; d. Evaluation of publication bias using Egger's publication bias plot. CI: 0.00, 0.43)) ( Table 2).

Discussion
In developed countries in Eastern Europe and Central Asia, the burden of disease caused by alcohol accounts for 12.1%. In several developing countries in the Asia Pacific and South America, this rate is estimated at 6.2% [5,71]. This meta-analytical study, which involves 23,258 men in 40 reports on 5 continents, once again confirms the potential danger to male reproductive quality at various concentrations of alcohol in everyday life.
There is a contradiction in conclusions of the effect of alcohol on male reproductive function. Gou et al. in a report on 66 alcoholics, drinking more than 1 year, results showed that the average sperm count, volume, motility, or percentage of sperm was significantly reduced [72]. Stutz et al. concluded that alcohol use could have had detrimental effects on seminal parameters [25]. In contrast, the findings of Govern et al. demonstrated no statistically significant differences in semen volume, sperm concentration, and percentage of motile sperm of daily drinkers [73].
Oxidative stress is implicated as a cause of sperm DNA damage and it can lead to infertility [74][75][76]. In men, a high ROS ratio leads to sperm DNA damage and sperm DNA fragmentation [77]. Sperm DNA damage has been evaluated as an important attribution of semen quality [78,79], and higher ROS concentrations may also induce autophagy of spermatozoa [80]. Therefore, to be able to diagnose and treat male infertility effectively, it is necessary to evaluate the oxidative status, antioxidant defense system and DNA damage, and sperm parameters [76]. Our study results showed that drinking alcohol reduced antioxidant enzymes in semen, but did not change sperm DNA fragmentation index. In addition, we do not have enough studies to analyze how different levels of alcohol intake affect antioxidant enzymes and sperm DNA fragmentation.
The association between total testosterone levels and alcohol drinking habits was not previously found [62,[81][82][83][84]. In contrast, reports have shown that men with alcoholism will have a decrease in serum testosterone levels accompanied by a low LH and FSH, which may result from two mechanisms: alcohol destroys Leydig cells or affects the metabolism of hypothalamic-pituitary-gonadal activity [85]. Our results have also shown the effect of alcohol on reproductive hormones, which reduces testosterone, FSH, and LH.
Performing a subgroup analysis provides a clearer view and several interesting results when looking at each group separately. The report also advises that heavy drinkers should be encouraged to cut back [71]. As previous studies only focused on studying two main groups of drinkers and non-drinkers, or heavy drinkers and drinkers, our study expanded the study groups to four subgroups, including the comparison of the alcoholic vs the no-alcoholic group, the moderate alcoholic vs the non-alcoholic groups, the heavy alcoholic vs the no-alcoholic groups, the heavy alcoholic vs the moderate alcoholic groups. This helps our study to determine exactly how alcohol consumption levels affect fertility in men. The heavy alcohol consumption groups have been shown to affect sperm quality, decreasing testosterone and increasing estradiol. Drinking alcohol at all levels does not affect Inhibin B and SHBG2 which is contrary to several previous studies [19,60]. Subgroup analysis offers interesting conclusions, moderate drinking (less than 7 alcohol units per week) has no effects on male fertility. Heavy drinking (more than 7 units/week) negatively impacts male reproductive health. Around the world, alcohol use among men should be considered an important public health issue. During the five-year period from 2000 to 2005, it was alarmed noting that the rate of alcohol use among Vietnamese people increased rapidly rate by 50% [86,87], besides that the data related to epidemiological data shows that the harmful effects of alcohol use on public health and society are increasing [88].
In conclusion, the meta-analysis of 40 studies surveying 23,258 men showed that alcohol consumption hurt ejaculate semen volume, antioxidant enzymes in semen, and sex hormone levels. However, drinking alcohol at a moderate level (less than 7 alcohol units per week) is unlikely to change the parameters of semen and sex hormones, while drinking at a heavy level (more than 7 units/per week) is likely to make negative effects on semen volume and sex hormones.

Author contribution statement
Tung Nguyen-Thanh, Ph.D.: Conceived and designed the experiments; Performed the experiments; Analyzed and interpreted the data; Contributed reagents, materials, analysis tools or data; Wrote the paper.
Ai-Phuong Hoang-Thi: Performed the experiments; Analyzed and interpreted the data; Contributed reagents, materials, analysis tools or data; Wrote the paper.
Dang Thi Anh Thu: Analyzed and interpreted the data; Contributed reagents, materials, analysis tools or data; Wrote the paper.

Table 2
Subgroup analysis of the associations between alcohol consumption at different levels on male reproductive function.

Data availability statement
Data included in article/supp. material/referenced in article.

Declaration of interest's statement
The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

Additional information
No additional information is available for this paper.