Association Among Sleep Behaviors, Circadian System And Semen Parameters In Males

Background The male reproductive health was suggested to be in an unfavorable situation in the past decades. Studies indicated that there was regional or global decline of semen quality, and the prevalence of infertility stay high, half of which could be attributed to the males. Identification of the hazardous factors may promote the prevention and treatment of semen damage. The circadian system has been suggested to play a substantial role in the regulation of male reproductive function, and it is well known that sleep behavior could influence the running of the circadian system. However, it is still not clear whether improper sleep behavior could induce male reproductive damage via the circadian system. The few existing researches have some major limitations: the quantitative measurement of sleep behavior; appropriate outcome; investigation of the circadian system’s role in the sleep-semen relation. The aim of the present study was to investigate the association between sleep behaviors and semen parameters, as well as the effect of circadian system on the association.Methods In an adult male reproductive cohort, the Munich Chronotype Questionnaire(MCTQ) was used to quantitatively measure the sleep behaviors(time points and durations) and the chronotype(an indicator of individual circadian-system characteristics). The associations between sleep behaviors and semen parameters as well as reproductive hormones were investigated, with the consideration of the interaction of sleep behavior and chronotype. The diurnal variation of the semen parameters was also investigated, especially in the subjects of different chronotypes. The investigations were then replicated in two other populations and animal study was implemented, too. 1. MARHCS(Male Reproductive Health in Chongqing College Students) dataset In 2013, a total of 796 male adult college students were recruited in College Town of Chongqing, China. The socio-psycho-behavioral characteristics of the subjects were measured by a composite questionnaire,within which the Munich Chronotype Questionnaire(MCTQ) was used to estimate the sleep time point, sleep duration and chronotype. Other information including age, body mass index(BMI), abstinence period, smoking, drinking, consumption of tea, coffee and cola was also collected. Semen sample was collected and analyzed following the WHO-recommended methods. The time point of ejaculation was recorded. The following parameters were measured: semen volume, sperm concentration, total sperm count, progressive motility, total motility and DFI(DNA Fragment Index). Peripheral blood sample was collected for serum reproductive hormones measurement(follicle-stimulating hormone, luteinizing hormone, estradiol, progesterone, testosterone, and prolactin). The subjects were followed up twice in 2013(n=656, 82.4%) and 2014(n=568, 71.4%). This dataset was used to investigate the association between sleep behaviors(sleep time point and sleep duration) and semen parameters, the interaction of sleep behaviors and chronotype, the association between semen parameters and time point of ejaculation(the analysis of circadian rhythm in semen parameters). The association between sleep behavior and reproductive hormones was also analyzed. 2. Six Towns Study dataset In 2007, a total of 1346 male adult residents were recruited from six geographically and demographically representative towns(3 districts and 3 counties) in the Chongqing-section of the Three Gorges Reservoir Region. Questionnaire was used to collected the socio-psycho-behavioral information including sleep duration and work type(day work, evening work, shit work, etc.). Other characteristics such as age, abstinence period, smoking, drinking, consumption of tea and coffee were also measured. Semen sample was collected for analysis of the semen parameters(semen volume, sperm concentration, total sperm count, progressive motility and total motility) according to the WHO-recommended methods. Height and weight were measured to estimation of BMI. This dataset was used to investigate the difference of semen parameters among subjects of different work types(indicating the relation of sleep time point and semen quality) and the association between sleep duration and semen parameters. 3. Hubei Sperm Bank dataset This dataset includes the records of 16845 semen samples from 8122 sperm donors who visited the human sperm bank of Hubei province from 2010 to 2015. Age and abstinence period were recorded. Semen samples were collected and analyzed(semen volume, sperm concentration, total sperm count, progressive motility and total motility) following the WHO-recommended method and the standard of sperm bank. This dataset was used to investigate the association between ejaculation time point and semen parameters. 4. Animal study The sleep behavior of the male C57BL/6 mice was controlled by artificial schedule of light exposure. Two groups of mice(8 weeks old; n=7 for each group) were kept in12 h light: 12 h dark cycle for 2 weeks. Then the light/dark schedule of the intervention group was shifted(lights on at the previous light-off time point, vice versa) every 6 day while the schedule of the control group was kept unchanged. Eighteen days later, animals were sacrificed and comparison of testis weight, epididymis weight and semen parameters(sperm concentration, progressive motility and total motility) between the two groups was done. This experiment aimed to investigate the effect of change in sleep time point on semen parameters.Results1. Association between sleep time point and semen parameters(1) In the MARHCS dataset, univariate analysis showed that sleep midpoint was negatively correlated with semen volume and total sperm count(P=0.032 and 0.009, respectively). However, these associations were not significant after adjustment for potential confounders including age, abstinence period, BMI, smoking, drinking, consumption of tea, cola and coffee. Then we calculate the difference of sleep midpoint between work days and free days, and investigate the association between the midpoint difference and the semen parameters. Although the midpoint difference was negatively correlated with total sperm count in univariate analysis(P=0.029), the result of multivariate analysis was not statistically significant. No association was found between sleep midpoint(or midpoint difference) and any reproductive hormone.(2) In the Six Towns Study dataset, the semen parameters of evening workers were not significantly different from those of the day works. As to the shift works, their semen volume and total sperm count were lower than those of the day workers(P=0.025 and 0.034, respectively), but that difference become non-significant after adjustment for potential confounders.(3) In the animal study, spermconcentration(1.6×106/m L vs. 2.6×106/m L, P=0.035), progressive motility(5.6% vs. 9.4%, P=0.047) and total motility(17.4% vs. 28.3%, P=0.029) were found to be lower in the intervention group than in the control group, while the weight of testis and epididymis was not significantly different between the two groups. 2. Association between sleep duration and semen parameters(1)Significant inverse U-shaped association was found between semen parameters and sleep duration. Either excessive or restricted sleep duration was associated with reduced semen volume and total sperm count. The association remained significant after adjustment for potential confounders. Compared with those who slept 7.0-7.5h/d, subjects who slept >9.0h/d was associated with a 21.5%(95% confidential interval: 9.2%, 32.2%) reduction of semen volume and a 39.4%(23.3%, 52.1%) reduction of total sperm count; those who slept ≤6.5h/d was associated with 4.6%(-10.5%, 22.3%) and 25.7%(-1.2%, 60.1%) reduction of the two parameters. When comparing the sleep duration and semen parameters between 2013 and 2014, those who slept ≤7.0h/d(median: 6.7) in 2013 was found to have a longer sleep(median: 7.3) in 2014, while those slept >7.5h/d(median: 8.2) in 2013 had a shorter sleep(median: 8.0) in 2014. On the other hand, semen volume and total sperm count of the subjects increased from 2013 to 2014(P=0.002 and 0.025 respectively) while no significant change was found in other semen parameters. When integrating the data of the two years by mixed model, each hour of departure from 7.0-7.5h/d sleep was associated with a 5.7%(2.1%,9.2%; P=0.002) reduction of semen volume and a 9.3%(2.2%, 16.1%; P=0.011) reduction of total sperm count. No association was found between sleep duration and reproductive hormones.(2) In the Six Towns Study dataset, similar association between sleep duration and the two semen parameters was found.(3) The subjects of MARHCS dataset were determined as there chronotypes(early type, intermediate type and late type) and the association between sleep duration and semen parameters were reanalyzed in each chronotype respectively. Inverse U-shaped association was found in each of the three groups but the turning-point of the association was different among them: in the early type subjects, those who slept 6.5-7.0h/d had the highest semen parameters; as to the intermediate type subjects and the late type subjects, highest semen parameters were found in those who slept 7.0-7.5 h/d and 8.0-8.5 h/d respectively.(1) The sperm motility(progressive motility and total motility) was found to fluctuate both in the MARHCS dataset and in the Hubei Sperm Bank dataset. In the MARHCS dataset it began to rise at 14:00-(P=0.007 for progressive motility, P=0.011 for total motility) and reached its peak at 18:00-, then it declined until after 20:00(P=0.017 and 0.009 respectively). According to the standard of Chinese sperm bank(progressive motility over 60% was eligible for donation), ejaculating each hour earlier or later than 18:00- was associated with a 14%(4%, 22%) lower probability to meet that standard. In the Hubei Sperm Bank dataset, the sperm motility began to rise at 12:00-(P=0.001 for progressive motility, P=0.001 for total motility) and reached its peak at 14:00-, then it declined until after 17:00(P=0.001 and P<0.001 respectively). Ejaculating each hour earlier or later than 14:00- was associated with a 9%(3%, 14%) lower probability to meet the standard of progressive motility for donation.(2) The fluctuation of progressive motility was reanalyzed in MARHCS subjects of different chronotype separately. The fluctuation was found in each group(P=0.003, 0.016 and 0.035 respectively) with distinct time point of beginning, peak and ending. In the early type subjects, it began at 14:00- and reached its peak at 18:00-, then declined until after 20:00-; in the intermediate type subjects, it began at 9:00- and reached its peak at 11:00-, then declined until 15:00-; as to the late type subjects, it began at 13:00-and reached its peak at 16:00-, then declined until 19:00-.(3) No significant 24-h circadian rhythm of any semen parameter was found and replicated in both dataset when investigating by the Cosinor software although semen volume(P=0.035) and total sperm count(P=0.057) in the MARHCS dataset seemed to fit that distribution. 3. Circadian rhythm of the semen parameters in human beingsDiscussions The results of the present study showed that(1) Sleep duration but not sleep time point was associated with semen parameters. There was an inverse U-shaped association between sleep duration and semen volume and total sperm count, either restricted or excessive sleep was associated with declined semen parameters. There was also interaction of sleep duration and chronotype on semen parameters: the “most suitable” sleep duration was distinct among the subjects of different chronotypes.(2) The sperm motility(progressive motility and total motility) of the subjects was found to fluctuate during certain time points, showing an inverse-U shaped process. Although no substantial 24-hour circadian rhythm was found in any semen parameter, the fluctuation of sperm motility seemed associated with chronotype:the time point of beginning, peak and ending was distinct among different chronotypers. Circadian rhythm of male reproductive function such as the release of spermatozoa has been found in model animals. However, it is still not clear whether there was also circadian rhythm in the semen parameters of human beings. The only two studies just investigated the difference of semen parameters between morning and afternoon or evening and showed controversial results. In the present study we found the sperm motility fluctuating during certain time points depending on the chronotype of the subject, indicating that semen of human being was under the regulation of circadian system. No substantial 24-hour circadian rhythm was found in any semen parameter, indicating that the long period of spermatogenesis and sperm maturation(about 90 days) in human lead to a more complicate pattern of semen. The chronotype-dependent fluctuation may also help to understand the discordance between the results of former studies. The present study indicates a potential strategy to promote the probability of impregnation and collection of eligible sperm samples in the sperm banks. As to the research of sleep behavior and male reproductive phenotypes, there have been several studies about the association between sleep duration and reproductive hormones or association between shift work and infertility, however, to our knowledge there is no report that investigated the relation of sleep behaviors and semen parameters in human beings. Our study showed an inverse-U shaped association between sleep duration and two semen parameters and an interaction of sleep duration and chronotype on the semen parameters, suggesting that sleep behavior may affect the semen quality via the circadian system, which was in concordance with our hypothesis. This finding may deserve serious concern considering the prevalence of untraditional sleep behaviors nowadays around the world. The results suggest that circadian rhythm is involved in the semen regulation of human beings and could be the pathway by which sleep behavior influences the semen quality. This study may promote the understanding of the relation among sleep behavior, circadian system and the male reproductive phenotypes. Further researches would be needed to confirm the findings in the present study.