Semen analyses in 1,283 men from the United States over a 25-year period: no decline in quality

Transcription

1 FERTILITY AND STERILITY 1996 American Society for Reproductive Medicine Printed on acid free paper in U. S. A. Semen analyses in 1,283 men from the United States over a 25-year period: no decline in quality Harry Fisch, M.D.*t:J: Erik T. Goluboff, M.D.* John H. Olson, M.S.T. Joseph Feldshuh, M.D.II Stephen J. Broder, B.S.~ David H. Barad, M.D.t Columbia-Presbyterian Medical Center; Albert Einstein College of Medicine, and Idant Laboratories, New York, New York; Cryogenic Laboratories, Roseville, Minnesota; and California Cryobank, Inc., Los Angeles, California Objective: To determine whether semen quality has changed in the United States over the last 25 years. Design: Retrospective review. Setting: Three U.S. sperm banks, Cryogenic Laboratories, Inc. (Roseville, Minnesota), Idant Laboratories (New York, New York), and California Cryobank, Inc. (Los Angeles, California). Patients: Twelve hundred eighty-three consecutive men who banked sperm before vasectomy from 1970 to Intervention: None. Main Outcome Measures: Age at sample collection, sperm concentration, volume, motility, and days of abstinence before sample collection were determined for each man. Linear and multiple regression analyses were used to assess changes in these characteristics over time. Results: Controlling for the effects of age and duration of abstinence, there was a slight but significant increase in mean sperm concentration but no change in either motility or semen volume over the 25-year period. Both sperm motility and semen volume decreased with increasing age at sample collection. Both sperm concentration and semen volume increased as a function of duration of abstinence. There were significant differences in mean (::!:::SEM) sperm concentrations (l06 spermlml) and motilities between the different sperm banks with California lowest (72.7 ::!::: 3.1,51.4% ::!::: 1.1%, respectively), Minnesota higher (100.8 ::!::: 2.9,56.0% ::!::: 0.5%, respectively), and New York highest (131.5 ::!::: 3.5,58.2% ::!::: 0.5%, respectively). Conclusions: Our data show no decline in sperm counts over a 25-year period in 1,283 men who banked sperm before vasectomy at three distinct geographical sites in the United States. Fertil SteriI1996;65: Key Words: Sperm counts, geographic variations and changes over time in semen characteristics Received August 24, 1995; revised and accepted November 3, * Department of Urology, Columbia-Presbyterian Medical Center. t Department of Obstetrics and Gynecology, Albert Einstein College of Medicine. :j: Reprint requests: Harry Fisch, M.D., Department of Urology, Columbia-Presbyterian Medical Center, 944 Park Avenue, New York, New York (FAX: ). Cryogenic Laboratories, Inc. An international decline in semen quality over the last several decades has been suggested (1-13). Environmental toxins and increases in rates of genitourinary congenital abnormalities have been implicated as possible causes for this supposed decline. In 1995, Auger et a1. (13), from Paris, France, reported a 30% decrease in sperm concentration over a 20-year period among screened, highly selected, fertile sperm donors from a single sperm bank. In 1992, Carlsen et a1. (8) reported a meta-analysis of 61 articles published between 1938 and 1991 that included sperm concentrations from presumably fer- II Idant Laboratories. ~ California Cryobank Inc. Fisch et al. No decline in semen quality in the U.S. 1009

2 tile men. Comparing studies from different geographic locations, these authors concluded that mean sperm concentrations had decreased worldwide by almost 50% from 1940 to 1990 (8). The only U.s. study to address the issue of changes in semen quality over time was that of Leto and Frensilli in 1981 (12). This study analyzed semen quality of men attempting to donate sperm from 1973 to During this time period, there was a 52% decline in the number of men who met the semen criteria for artificial insemination donors established by the Reproductive Council of the American Association of Tissue Banks (14). These authors concluded that semen quality in the donor population had decreased over this 8-year period. To determine whether semen quality had indeed changed over the last 25 years in the United States and to minimize selection, geographic, and methodological biases, we studied all men who banked sperm before vasectomy in the years 1970 to 1994 at three of the oldest sperm banks in the United States. MATERIALS AND METHODS A retrospective review of all men banking sperm before vasectomy from 1970 to 1994 at three U.S. sperm banks was conducted. The three sperm banks were Cryogenic Laboratories, Inc. (Roseville, Minnesota), Idant Laboratories (New York, New York), and California Cryobank, Inc. (Los Angeles, California). One thousand two hundred eighty-three men were studied: 662 men banked in Minnesota from 1970 to 1994,400 in New York from 1972 to 1994, and 221 in California from 1978 to No man was excluded from analysis. Age, sperm concentration, semen volume, motility, and duration of abstinence were recorded for each man. Age was recorded at sample collection. Where more than one specimen was banked, the mean value for each characteristic was used. In Minnesota and in New York, the men were asked, on a prevasectomy form, whether they had ever fathered a child. Men who answered the question were categorized as "proven" if they had fathered a child in the past and "unproven" if they had not. Those who did not answer the question were labeled "unknown" fertility status. In California, fertility status before vasectomy was not asked, and this information was not available. In all sperm banks, men were instructed to abstain from ejaculation for at least 3 but no more than 10 days before specimen collection by masturbation into a widemouthed collection vessel. Specimens were brought to the laboratory within 1 hour of collection Fisch et al. No decline in semen quality in the U.S. Semen Analysis Techniques Minnesota: Cryogenic Laboratories, Inc. In brief, after liquefaction, volume was determined by drawing up the entire sample into a 6.0- ml syringe with an 18-gauge needle. A drop of semen was placed on a microscope slide and was covered with a cover slip. Percent progressive motility was determined at high power (x40) magnification by observing 100 spermatozoa. Sperm concentration was determined by diluting semen 1:50. A O.Ol-mL aliquot ofthis mixture was then placed in a Neubauer hemacytometer (Hausser Scientific, Inc., Horsham, PA) and was allowed to stand for 2 minutes. Under high power, the number of sperm cells in 8 of the 16 large squares on the corners of the grid were counted. This count was then repeated once more on 8 different square'& The average of these two values is the sperm concentration in x10 6 /ml. California: California Cryobank, Inc. In brief, the specimen was allowed to liquefy at room temperature while on an orbital mixer for a minimum of 15 and a maximum of 60 minutes. Semen volume was determined by using a graduated bulb pipette. Five microliters of semen was placed in the center of a Makler chamber (Sefi-Medical Instruments Ltd., Haifa, Israel) and covered with a cover glass. Using x20 magnification, motile and nonmotile sperm were counted in 10 squares of the counting chamber grid. This process was then repeated with 10 other squares. Sperm concentration was the sum of the number of sperm counted in 20 squares divided by 2 in X 106/mL. Percent motility was determined by dividing the number of progressively motile sperm by the total number of sperm in 20 squares. New York: Idant Laboratories From 1972 to 1977, semen analyses were performed using a Neubauer hemacytometer, similar to the technique used in Minnesota as described above. From 1978 to 1994, semen analyses were performed using a Makler chamber (similar to the technique used in California as described above. Statistical Methods The data were analyzed using SPSS statistical computer software (SPSS, Inc., Chicago, IL). Means and standard errors were calculated for age, sperm concentration, semen volume, motility, and duration of abstinence before sample collection. Analysis of variance (ANOVA) was used to assess differences Fertility and Sterility

3 Table 1 Demographic and Semen Characteristics* Sperm Semen Sperm Duration of Group nt Age concentrationt volumet motility abstinence years lo6/ ml ml % days Total ± ± ± ± ± 0.08 By sperm bank Minnesota ± ± ± ± ± 0.1 New York ± ± ± ± ± 0.2 California ± ± ± ± ± 0.3 By fertility status II Proven 684 (64%) 35.4 ± Unproven 254 (24%) 30.2 ± Unknown 124 (12%) 33.2 ± ± ± ± ± 0.1 ± ± ± ± 0.2 ± ± ± ± 0.3 * Values are means ± SE. A symbol ( or II) for the mean of a single sperm bank or fertility status indicates that there is a significant difference between that mean and the other two means in the group, which do not differ significantly from one another. A symbol for all three means in a group indicates that each differs significantly from the other two. t Values in parentheses are percentages. +P < P < II California men are excluded from the fertility status data. among sperm banks and among groups of men with differing fertility status. When a significant main effect was obtained, the Scheffe test was used to determine which means differed significantly. To obtain normal distributions for semen characteristics, data for sperm concentration, semen volume, and duration of abstinence were logarithmically (base 10) transformed (9, 15). Linear and multiple regression analyses were used to assess the changes in semen characteristics. Age at the time of donation and number of days abstinent before donation were included in the multiple regression models as modifiers. RESULTS Table 1 depicts the demographic and semen analysis characteristics for the entire group of 1,283 men as well as for men separated by location of sperm bank and fertility status. Age was significantly lower in Minnesota than in New York or California. Sperm concentration and motility differed significantly between different locations with New York> Minnesota > California. Semen volume and duration of abstinence were significantly lower in California than in the other locations. Age differed significantly among the fertility status groups with Proven> Unknown> Unproven. Semen characteristics were similar among the different fertility status groups, except for sperm motility, which was significantly higher only in the Unknown group. Figure 1 shows the linear regression analysis of sperm concentration for the total population by year of specimen collection. There was a statistically significant increase in sperm concentration over the 25- year period from a mean of 77 X 10 6 /ml in 1970 to 89 X 10 6 /ml in 1994 (P = 0.04). The concomitant percent increase in mean sperm concentration was 0.65% per year. Data analyzed through linear regression by individual sperm bank also revealed a similar increase in sperm concentration over time in New York (r = 0.15, P = 0.002) and Minnesota (r = 0.11, P = 0.006). No statistically significant change was evident in California (r = 0.003, P = 0.06) where the sample size was small. Linear regression analysis by fertility status revealed an increase in sperm concentration over time in the Proven fertility status group (r = 0.012, P = 0.002) 1000., , 8' eli! i 0 Ċ. 100 I i ~ I: I ii, ~ e : ~ f Z W U z 8 10 ~ ~ CJ) o 8 I I : ; 0 CI B 8 ~ i I,: I I o. I i I! i i!! I!!j I, i! 1 I: 0 ' 8 0 g, i,:, 0;1 g ij 1~~~~~~~~~~~~~T-~_r~~~~ YEAR OF SPECIMEN COLLECTION Figure 1 Linear regression analysis of sperm concentration by year of specimen collection for the total population (r = 0.06, P = 0.04, n = 1283). Fisch et al. No decline in semen quality in the U.S. 1011!

4 Table 2 Multiple Regression Analyses of Year of Specimen Collection, Age, and Abstinence Period on Sperm Concentration, Motility, and Volume from 1,283 Men* Sperm Semen Sperm Variable concentration volume motility Year of sample collection 0.07t Age * -0.17t Days abstinent O.lM 0.22* 0.01 *All values represent regression coefficient. Statistical significance: t p = 0.03, * P < 0.001, not significant. and the Unproven fertility status group (r = 0.17, P = 0.007), while no statistically significant change was evident in the smaller Unknown fertility status group (r = 0.10, P = 0.26). Linear regression analysis for the total population showed no significant change in sperm motility over the 25-year period, but semen volume decreased slightly (r = -0.07, P = 0.001). Age at specimen collection increased significantly over the study period (r = 0.32, P < 0.001) from a mean of 30 to 38 years. Duration of abstinence did not change significantly over the study period or with increasing age. Age and duration of abstinence have been shown to influence semen characteristics (16, 17). Therefore, we used these variables as modifiers in the multiple regression models (Table 2). With increasing age, motility and volume both significantly decreased (P < 0.001), whereas with increasing duration of abstinence, sperm concentration and volume both significantly increased (P < 0.001). With age and duration of abstinence as modifiers, sperm concentration still showed a significant increase (P = 0.03), whereas motility and volume did not change over the study period. To rule out the possibility that the significant increase in mean sperm concentration over time was an artifact of intersite differences in mean semen characteristics and the changing proportion of subjects reported from each site from year to year, we conducted an additional series of multiple regression analyses. In each analysis, a set of dummy variables was entered to represent the three sites together with the predictors of age, abstinence, and year (18). The results of these analyses are consistent with those shown in Table 2: controlling for intersite differences, there was still a significant increase in sperm concentration (P = 0.004) over time, but no change in either motility (P = 0.67) or volume (P = 0.38). DISCUSSION Our study of 1,283 men who banked sperm before vasectomy during the years 1970 to 1994 at three 1012 Fisch et al. No decline in semen quality in the U.S. geographically distinct U.S. sperm banks shows a statistically significant increase in sperm concentration and no change in sperm motility or semen volume. i Many studies in the literature have suggested a decline in semen quality over time (1-13). Our data and other reports, however, do not support this conclusion (19-23). In the most recent report, Auger et al. (13) studied 1,351 fertile men from a single sperm bank in Paris, France. This study showed a decline in sperm concentration from 89 X 10 6 /ml to 60 X 10 6 /ml over a 20-year period and concluded that there was "a decline in semen quality among fertile men in Paris..." This study, however, contained a significant selection bias in that only fertile men who had been accepted as sperm donors were included for study and therefore excluded important data from fertile men who had attempted to become sperm donors but were rejected. Their conclusions are therefore limited to men accepted as sperm donors and not the general fertile population. Leto and Frensilli (12), in 1981, retrospectively reported on 275 men who attempted to donate semen in the United States for future use by couples with male factor infertility. Of men who attempted to donate in 1973, 77% met the 1980 semen criteria ofthe Reproductive Council of the American Association of Tissue Banks (14) for artificial insemination donors whereas only 37% of those who donated in 1980 met these criteria. These authors concluded that semen quality in the population of men attempting to become sperm donors had decreased over this 8-year period. Although we agree with the conclusions that semen quality in the donor or potential donor population may have decreased over time, we disagree with drawing conclusions from these data about the general or fertile male populations. Sperm donors represent a highly selected and screened group of men. Changes in donor sperm quality may simply be related to changes in the methods of recruitment or screening of donors, particularly as demand for donor sperm increases and the number of men attempting to become sperm donors increases (12). In our study no exclusion criteria existed; all men who banked sperm before vasectomy at each of three sperm banks during the years 1970 to 1994 were included. In addition, our population consisted of men with both proven and unproven fertility; our population may be more representative of the general population than studies concerning themselves solely with fertile men. Interestingly, we found no significant differences in semen characteristics between those with proven or unproven fertility. As well, there were similar increases in sperm concentration over time in the proven and unproven fertility status groups. Fertility and Sterility

5 Carlsen et al. (8), in 1992, reported a meta-analysis of 61 different studies published from 1938 to 1991 concerning semen analyses in fertile men from 22 different countries. With 14,947 men included and using linear regression alone, these authors found that mean sperm concentration and mean semen volume had significantly decreased from 113 X 1Q6/mL and 3.4 ml, respectively, in 1940, to 66 X 1Q6/mL and 2.75 ml, respectively, in An analysis of Carlsen's work, however, shows that significant biases could have skewed their results. These biases included failing to account for age and duration of abstinence, methodological inconsistencies, and geographic variations in semen quality. Age and duration of abstinence have been shown to have effects on semen characteristics (16, 17) and were not accounted for in Carlsen's meta-analysis (8). In our study, as age increased, sperm motility and semen volume both decreased, whereas as duration of abstinence increased, sperm concentration and semen volume also increased. Using linear regression analysis, we found a statistically significant decrease in semen volume over the study period just as Carlsen had reported. However, when age and duration of abstinence were used as modifiers in the multiple regression models, no such change was evident. The lack of standardized methods for performing semen analysis is an inherent problem of any study involving the counting of sperm, including ours. In Carlsen's meta-analysis, there were 61 different laboratories, with varying sets of protocols and "different types of counting chambers" (8). In our study, only the two most common counting chambers were used. The Makler chamber was used exclusively in California, and in New York from 1978 to The hemacytometer was used exclusively in Minnesota, and in New York from 1972 to Differences in semen characteristics in both Carlsen's study and ours may, in part, be due to methodological inconsistencies. We found marked differences in semen characteristics between the New York, Minnesota, and California sperm banks. MacLeod and Wang (22), as well as Smith and Steinberger (24), have reported marked geographic variations in sperm concentrations from New York, Houston, Iowa, and Philadelphia, with New York values being highest. Our study shows similar variations with sperm concentration and motility highest in New York and lowest in California. In examining the papers selected for inclusion in Carlsen's meta-analysis (8), it was evident that the earlier papers were from the United States, particularly from New York. Before 1970, 94% of men studied were from the United States. Of these men, 87% were from New York and represented the highest sperm concentrations of all the studies. Interestingly, our mean sperm concentration for the total population studied from 1970 to 1994 (105.5 X 10 6 / ml) is similar to the average sperm concentration of the United States studies from 1938 to 1970 cited in Carlsen's study (110.1 X 1Q6/mL). Mter 1970, only 50% of the men in Carlsen's study were from the United States with only 25% of these from New York. Given the reported geographic variability in semen quality across the United States (22, 24) and the possibility of variability across the world, Carlsen's meta-analysis is significantly skewed in terms of geographic distribution of studies. Carlsen's finding of declining semen quality over time is not surprising based on the preponderance of N ew York studies early on, and the preponderance of non-new York non-u.s. studies later on. Prior studies suggesting decreases in the semen quality over the last several decades have been fraught with selection, methodological, and geographic biases (1-13). By attempting to minimize these biases, we found no such decline in semen quality. In fact, we found a statistically significant increase in sperm counts in 1,283 men who banked sperm before vasectomy in the United States from 1970 to Acknowledgments. The authors are indebted to Howard Andrews, Ph.D., Ms. Kristin A. Olson, Russell C. Bierbaum, M.S.T., Eric K. Seamen, M.D., Patricia E. Barg, M.D., Carl A. Olsson, M.D., and Karen S. Fisch, B.S. REFERENCES 1. Nelson CMK, Bunge RG. Semen analysis: evidence for changing parameters of male fertility potential. Fertil Steril1974; 25: Naghma-E-Rehan, Sobrero AJ, Fertig JW. The semen offertile men: statistical analysis of 1300 men. Fertil Steril 1975; 26: SheriffDS. Setting standards of male fertility. I. Semen analyses in 1500 patients-a report. Andrologia 1983; 15: Lapido OA. Seminal analysis in fertile and infertile Nigerian men. J Natl Med Assn 1980;72: Wang C, Chan SYW, Leung A, Ng RP, Ng M, Tang LCH. Cross-sectional study of semen parameters in a large group of normal Chinese men. lnt J Androl 1985;8: Aribarg A, Kenkeerati W, Vorapaiboonsak V, Leepipatpaiboon S, Farley TMM. Testicular volume, semen profile and serum hormone levels in fertile Thai males. lnt J Androl 1986; 9: Kirei BR. Semen characteristics in 120 fertile Tanzanian men. East Mr Med J 1987;64: Carlsen E, Giwercman A, Keiding N, Skakkebaek NE. Evidence for decreasing quality of semen during the past 50 years. Br Med J 1992;305: Bostofte E, Serup J, Rebbe H. Has the fertility of Danish men declined through the years in terms of semen quality? A comparison of semen qualities between 1952 and nt J Fertil 1983;28:91-5. Fisch et al. No decline in semen quality in the U.S. 1013

6 10. Bendvold E. Semen quality in Norwegian men over a 20 year period. Int J Fertil1989;34: Menkveld R, Van Zyl JA, Kotze TJW, Joubert G. Possible changes in male fertility over a 15 year period. Arch Androl 1986; 17: Leto S, Frensilli FJ. Changing parameters of donor semen. Fertil Steril 1981; 36: Auger J, Kuntsmann JM, Czyglik F, Jouannet P. Decline in semen quality among fertile men in Paris during the last 20 years. New Engl J Med 1995;332: Sherman JK. Guidelines for the banking of human semen. Newsletter of the American Association of Tissue Banks, Rockville, MD. 1979;3: Bromwich P, Cohen J, Stewaret I, Walker A. Decline in sperm counts: an artifact of changed reference range of "normal"? Br Med J 1994;309: Schwartz D, Mayaux M-J, Spira A, Moscato M-L, Jouannet P, Czyglik E. Semen characteristics as a function of age in 833 fertile men. Fertil Steril 1983;39: Jouannet P, Czyglik F, David G, et al. Study of a group of 484 men. I. Distribution of semen characteristics. Int J Androl 1981;4: Cohen J, Cohen P. Applied multiple regression/correlation analysis for the behavioral sciences. 2nd ed. Hillsdale, NJ: Lawrence Erlbaum Associates, 1986: Suominen J, Vierula M. Semen quality of Finnish men. Br Med J 1993;306: Tummon IS and Mortimer D. Decreasing quality of semen. Br Med J 1992;305: Brake A and Krause W. Decreasil).g quality of semen. Br Med J 1992;305: MacLeod J, Wang Y. Male fertility potential in terms of semen quality: a review of the past, a study of the present. Fertil SteriI1979;31: Olsen GW, Bodner KM, Ramlow JM, Ross CE, Lipshultz LI. Have sperm counts been reduced 50 percent in 50 years? A statistical model revisited. Fertil SteriI1995;63: Smith KD, Steinberger E. What is oligospermia? In: Troen P, Nankin HR, editors. The testis in normal and infertile men. New York: Raven Press; 1977: Fisch et al. No decline in semen quality in the U.S. Fertility and Sterility