1 Department of Occupational Medicine, University Hospital of Aarhus, Aarhus Sygehus, Norrebrogade 44, building 2C,2 Danish Epidemiology Sciences Centre, University of Aarhus, Vennelyst Boulevard 6, build. 260, DK-8000 Aarhus C, Denmark, 3 Department of Growth and Reproduction, National University Hospital, Rigshospitalet, section GR 5064, Blegdamsvej 9, DK-2100 Copenhagen,4 Department of Environmental Medicine, University of Southern Denmark, Winslowparken 17, DK-5000 Odense C, Denmark and 5 Scanian Andrology Centre, Malmö University Hospital, SE 205 02 Malmö, Sweden
6 To whom correspondence should be addressed. E-mail: anemt{at}akh.aaa.dk
![]() |
Abstract |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|
Key words: fertility/male reproduction/inhibins/sperm density/time to pregnancy
![]() |
Introduction |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|
Semen quality is often used as the biological marker in environmental epidemiology of male reproductive health, as it provides the fluid of interest and other information on testicular function. However, semen quality studies are often encumbered with low response rates and high risk of selections bias (Larsen et al., 1998), partly because of low participation rates. Furthermore, sperm count and motility exhibit very large variability between and within persons, attributable to sample collection and factors such as abstinence time, season and sampling condition (Makler et al., 1979
; Schwartz et al., 1979
; Levine et al., 1990
; Cooper et al., 1992
; Bonde et al., 1996
; Cagnacci et al., 1999
). Therefore, it is important to identify more reliable markers of fecundity for epidemiological research, and inhibin B may be a promising candidate. Inhibin B is a gonadal peptide hormone produced in the Sertoli cells. It is the principal circulating form of inhibin in men and the concentration in plasma is a quantitative measure of spermatogenesis (Anawalt et al., 1996
; Illingworth et al., 1996
; Jensen et al., 1997
; Kolb et al., 2000). An inverse relation between inhibin B and FSH has been found, through a negative feedback loop to the pituitary gland (Anawalt et al., 1996
; Illingworth et al., 1996
).
Other studies have examined the use of inhibin B as a marker of successful retrieval of sperm in patients with non-obstructive azoospermia (Pierik et al., 1998; Ballesca et al., 2000
; Vernaeve et al., 2002
), and a recent review on inhibin B and spermatogenesis concludes that inhibin B holds promise for further research (Pierik et al., 2003
).
Inhibin B may therefore be a feasible and reliable marker of male fecundity (reproductive ability), not only because blood samples are easier to obtain than ejaculates, but also because it may be a better indicator of the expected semen quality at the time of conception. Although semen quality is the endpoint of choice in studies of effects on male fecundity, the sample we have for examination is by definition not taken at the right time. Given the large intra-individual variation, this sample may be a poor marker of semen quality at the time of conception. Measuring the validity of inhibin B in this respect requires studies on the correlation between inhibin B and semen quality as well as studies on actual conception. It is the latter we provide in this study.
In a Danish population-based prospective study, we had the opportunity to investigate the fecundability of 343 Danish first pregnancy planners according to semen characteristics and reproductive hormones, including inhibin B.
The aim of the present study was to assess the predictive value of male inhibin B as a marker of fecundability, investigate a possible threshold value of inhibin B, and to compare the predictive value of inhibin B with sperm density and FSH.
![]() |
Methods |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|
When the couples had notified the research team that they had stopped using contraceptives to attain a pregnancy, both partners filled in a questionnaire on demographic, medical, reproductive, occupational and lifestyle factors. They were enrolled at two centres: the Department of Occupational Medicine in Aarhus (West Center) and the Department of Growth and Reproduction in Copenhagen (East Center), where they were invited to a consultation by a physician or a nurse who provided instructions and obtained a blood sample. At baseline the males provided a semen sample (n = 418, 97.2% of those invited) and a blood sample for inhibin B measurement. The inhibin B measurement was not part of the initial study. Therefore, only 343 inhibin B measurements were performed (79.8% of the invited). The couples were followed until a pregnancy was diagnosed by their general practitioner or for a maximum of six menstrual cycles. Pregnancy was diagnosed by consecutive determination of HCG in urine samples from day 1 to day 10 in each menstrual cycle. When one of the values during a 10-day period was above 1.0 IU followed by a decline, the woman was regarded as having an early embryonic loss. If the woman had no menstrual bleeding, a self-reported ongoing pregnancy was registered after a visit to the general practitioner.
Blood and semen collection and analysis
Inhibin B was measured by an enzyme immunometric assay, which is specific for the inhibin B dimer (-
B) (Groome et al., 1996
). The assay uses an immobilized monoclonal capture antibody raised against a sequence from the human
B-subunit and a secondary enzyme-conjugated monoclonal antibody raised against a sequence from the human
-subunit. The sensitivity of the inhibin B assay was 20 pg/ml, and the intra- and interassay coefficients of variation were <12 and <17%, respectively. Serum concentrations of FSH were measured using a time-resolved immunofluorometric assay from Wallac, Turku, Finland. The sensitivity of the FSH assay was 0.06 IU/l, and both the intra- and interassay coefficients of variation were <10%. A detailed description is given elsewhere (Jensen et al., 1997
).
Semen samples were obtained by masturbation and were collected at home. The participants were encouraged to obtain the sample after 3 days of abstinence, but it was emphasized that samples obtained outside this period would be accepted. The exact period of abstinence and information on recent fever were noted in a questionnaire filled in on the day of semen collection. The ejaculates were collected directly into a 50-ml polyethylene container, and 93.8% of the samples were examined within the first 2 h.
Of the available semen quality parameters, only sperm density was compared with inhibin B for predicting fecundability, since sperm density is the previously best known predictor of male fecundability (Bonde et al., 1998a).
The sperm density analysis was performed in accordance with the WHO 1992 guidelines (World Health Organization, 1992) as described elsewhere (Bonde et al., 1998a
).
Statistical analysis
We examined the probability of a clinically recognized pregnancy in a menstrual cycle during follow-up according to the male serum concentration of inhibin B and sperm density at baseline. The analyses include 343 couples, contributing a total of 1327 menstrual cycles, where the couples had reported sexual intercourse between cycle days 11 and 17. The unit of observation was a menstrual cycle and each cycle was numbered, starting with the first unprotected cycle (numbered 1).
We investigated the characteristics of the study population. Data are presented in three categories: less than 100 pg/ml; 100200 pg/ml; and above 200 pg/ml. The concentration 200 pg/ml is approximately the median value.
The association between male serum inhibin B and fecundability was estimated using a logistic regression model while controlling for potential confounding factors (Hosmer and Lemeshow, 2000). Inhibin B was entered as a continuous variable in the logistic regression analysis. This model is equivalent to the discrete version of the Cox proportional hazards model (Weinberg, 1994
) and provides fecundability odds ratios, which are the estimated odds of conception in a menstrual cycle for the group of interest. We analysed the probability of conception in a menstrual cycle, given that conception had not occurred in any of the previous cycles. Technically, we use logistic regression with adjustment for menstrual cycles (dummy variables for cycles 2, 3, 4, 5 and 6, with cycle 1 as the implicit reference). Thus our analytical technique estimates the probability of conception in one menstrual cycle at the couple level.
Potential confounding factors of fecundability were selected by their biological relevance with respect to fecundity. Only factors that were disproportionally distributed by 10% or more in the three categories were included in the model, and none of the variables that were not included changed the estimate by more than 2%. We adjusted for centre (according to design), female age, use of oral contraceptives at enrolment and time when the blood sample was taken. The serum concentration of inhibin B was transformed by the logarithmic function to approximate normality as the distribution of inhibin B was skewed towards the right. Menstrual cycle numbers were entered as dummy variables to adjust for the dependency between menstrual cycles.
We performed measurement of inhibin B at baseline. To account for potential change of inhibin B over time, we also performed a restricted analysis based on baseline inhibin B and cycle 1 only, again adjusting for centre, female age, time of sampling, and use of oral contraceptives.
To evaluate if inhibin B could independently predict fecundability, we calculated the adjusted fecundability odds ratio of inhibin B increment by 1 log pg/ml, in a model that also included sperm density.
Since FSH is a known, and often used, endocrine marker in reproductive studies, we also investigated the fecundability odds ratio according to the baseline serum concentration of FSH.
Stratification of inhibin B into 50 pg/ml intervals indicated a threshold value of approximately 150 pg/ml under which the fecundability odds ratio was reduced.
A possible threshold of inhibin B was examined by shifting the threshold value through 20 integer values in the interval from 0 to 325 pg/ml in 20 iterative logistic regressions, where the most likely threshold was given by the model with the lowest 2 log L value (Olsen, 1992). The threshold value was used to dichotomize the data for further analysis.
Finally, we evaluated the outcome of pregnant/not pregnant after 6 months in couples according to inhibin B level by receiver operating characteristic (ROC) analysis. ROC graphs are plots of the sensitivity (y-axis) and 1 specificity (x-axis) pairs that are possible for all levels of a particular parameter. The sensitivity in this study is defined as the proportion of the men with low (100 pg/ml) inhibin B who did not achieve a pregnancy and the specificity is the proportion of men with inhibin B values above 100 pg/ml who achieved a pregnancy. Calculation of the area under the curve (AUC) provides a quantitative measure of accuracy, namely the ability of a certain parameter to discriminate between two conditions. The closer a ROC plot lies to the upper left corner, where the true positive rate is 1 (perfect sensitivity) and the false positive rate is 0 (perfect specificity), the closer the test is to perfect discrimination. The theoretical plot for a test with no discrimination (equal distribution of inhibin B in the two groups, pregnancy/no pregnancy) is a diagonal line at an angle of 45° from the lower left corner (0% true-positive rate and 0% false-positive rate) to the upper right corner (100% true-positive rate and 100% false-positive rate) with an area under the curve (AUC) of 0.5 (Zweig and Campbell, 1993
).
The correlation between inhibin B and sperm density and the correlation between inhibin B and FSH were determined using the Pearson correlation coefficient following adequate transformation to obtain approximate normality.
All statistical analyses were performed using SAS software, version 8.02 (SAS Institute, Cary, NC, USA).
![]() |
Results |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|
|
There was a moderate correlation between logarithmically transformed inhibin B and sperm density, of 0.34 (P < 0.0001, continuous variables), and a rather strong correlation between inhibin B and FSH after logarithmic transformation, of 0.62 (P < 0.0001, continuous variables).
Table I displays the characteristics of the study population according to the chosen cut-off values of inhibin B. All characteristics had a balanced distribution across groups, except female age, use of oral contraceptives and the time when the blood sample was taken. According to the design, we also adjusted for centre.
Table II shows the fecundability odds ratio by inhibin B values categorized into 50 pg/ml intervals. The probability of pregnancy was mainly low for inhibin B values below 150 pg/ml. With restriction to inhibin B below 150 pg/ml, the inhibin B increment by 1 log pg/ml was 2.28 (0.707.42). This may indicate that low inhibin B is a marker of reduced fecundability, and we therefore used a slopethreshold logistic regression model to identify a threshold inhibin B concentration, if any. The lowest 2 log L value was 94 pg/ml, pointing to a value of approximately 100 pg/ml as the most likely threshold (Figure 1). In the present population a total of 29 men had an inhibin B value equal to or less than 100 pg/ml, and only 10 of the 195 identified pregnancies were found in this stratum (34.5 versus 58.9%). Using the threshold value, we found an adjusted fecundability odds ratio of 0.51 (95% confidence interval 0.261.00) for the population with inhibin B less or equal to 100 pg/ml compared with the population with inhibin B above 100 pg/ml.
|
|
The area under the inhibin B ROC curve in the present population (n = 343) was 0.56 and the area under the sperm density curve was 0.60. Both are close to 0.5, which is the area for no discrimination.
When looking at the subgroup with the lowest (100 pg/ml) inhibin B measurements, corresponding to the group with expected subfecundity (n = 29) according to the calculated threshold value, the AUC was 0.79 (Figure 2). In the stratum with the lowest (
100 pg/ml) inhibin B (n = 29), the area under the sperm density curve was 0.91 (Figure 3) and the area under the FSH curve was 0.80 (Figure 4).
|
|
|
In a fecundability odds ratio analysis including cycle 1 only, we found a point estimate of 1.40 (0.662.96). The point estimate of inhibin B increment by 1 log pg/ml in the adjusted model, including adjustment for sperm density, was 1.33 (0.941.88), and the fecundability odds ratio of FSH increment by 1 log IU/l in the adjusted model was 0.75 (0.570.99).
![]() |
Discussion |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|
The inhibin B assay used in this study specifically measures the inhibin B dimer. Inhibin B is the main circulating inhibin form in adult men, and an inverse relationship exists between circulating inhibin B and FSH, both in men with normal spermatogenesis and in those with abnormal spermatogenesis (Anawalt et al., 1996; Illingworth et al., 1996
). Our study corroborates this association.
FSH has previously been suggested to be a reliable marker of male fecundity (Bergmann et al., 1994), which is supported by the significant inverse relation between serum FSH and couple fecundability in our study. But taking into account that more recent studies show that FSH levels are affected by GnRH, oestradiol and testosterone, and that FSH is a product of the pituitary gland and is not produced in the testes (Klingmuller and Haidl, 1997
; Pierik et al., 1998
; Yaman et al., 1999
), inhibin B should probably be seen as a direct marker of male fecundity. Although from a statistical point of view FSH is as strong a marker of fecundity as inhibin B, the latter is to be preferred for biological reasons.
In contrast, sperm density is affected by factors such as seasonal variation and abstinence time. Variation in methodology makes it difficult to compare between samples collected at different laboratories (Makler et al., 1979; Schwartz et al., 1979
; Levine et al., 1990
; Bonde et al., 1996
; Cooper et al., 1992
; Cagnacci et al., 1999
). Since inhibin B seems to be unaffected by these factors, it could be a potential competitor to sperm density in the determination of male fecundity. On the other hand, there could be situations, e.g. obstruction of the seminal ducts, where semen quality is affected and inhibin B is unaffected, why this in particular could be of interest when considering urogenital disorders. Another limitation of the inhibin B measurement is the fact that this hormone is a marker of the early, premeiotic part of spermatogenesis. The postmeiotic events seem to be mainly androgen-regulated (De Gendt et al., 2004
).
The percentage of females over the age of 30 was larger in the stratum with inhibin B between 0 and 100 pg/ml and the number of recognized pregnancies was smaller in that stratum, which would bias the point estimate towards too high values. In addition, there was a slightly larger percentage of men who had the blood sample drawn before 12 a.m. in the strata with the highest inhibin B, which is consistent with the fact that inhibin B in the blood peaks in the morning hours (Carlsen et al., 1999).The resulting misclassification is expected to attenuate the point estimate. Finally, the percentage of women using oral contraceptives at enrolment was highest in the stratum with the highest inhibin B values, causing bias, probably towards null values. However, we controlled for all these potential confounders in the adjusted and final model and the adjusted point estimate was generally slightly lower, but highly significant. All other factors we controlled for were similar in the two strata and inhibin B was not related to smoking or other lifestyle factors, as seen in other studies (Jensen et al., 1998a
,b
,c
; Andersson et al., 2003
, 2004
). For this reason we did not include them in the adjusted model. Several known or potential risk factors for reduced female fecundity were equally distributed in the two groups.
The study was based on a sample from the general population with no prior knowledge of their fecundity. Such a sampling strategy oversamples the less fecund, and the average monthly fecundability was 15%, which is below the expected values of 2030% (Bonde et al., 1998b; Wilcox et al., 1988
). Except for bringing more power to the study, it is not expected to change the effect size.
The fact that we used couple fecundability as the reference probably resulted in underestimation of inhibin Bs true predictive value for male fecundity, because couple fecundability is also influenced by female fecundity. It was more optimal to include only fecund females in the study. Furthermore, we followed the couples for a maximum of 6 months. This was done for feasibility reasons. We believe that a longer follow-up period would have lowered the participation rate. Also, the couples could be expected to change behavior after 6 months in order to obtain pregnancy. Therefore, we think that it was better to study more couples for 6 months rather than fewer couples for e.g. 1 year.
We only obtained a baseline blood sample. According to a study on inhibin B and seasonal variation, the day-to-day levels of inhibin B are relatively constant (Andersson et al., 2003). Nevertheless if inhibin B does change over time, the true inhibin B value at time of conception could be different from the value measured at baseline. Therefore, we analysed the data only obtained in the first cycle to examine this possible bias. The point estimate of the crude model was close to the adjusted model estimate, indicating that lack of cycle-specific inhibin B measurements is not a substantial problem.
The analyses including sperm density indicated that inhibin B provides information on male fecundity beyond that given by sperm density. This confirms the recent findings of Andersson and colleagues (Andersson et al., 2004) that inhibin B measurement should be considered a valuable additional tool in the elucidation of male fecundity.
In the inhibin B and sperm density ROC curves based on the whole study population, the ability to discriminate between pregnancy and no pregnancy was not very good, since the AUC for inhibin B was 0.562 and the AUC for sperm density was 0.600.
By using the threshold value of inhibin B of 100 pg/ml, corresponding to the group with higher probability of subfecundity, the AUC was 0.79. The area under the sperm density curve and the FSH curve, based on the same population, was 0.91 and 0.8 respectively. The larger AUC for sperm density illustrates that sperm density seems to be a more accurate predictor of fecundity than inhibin B, but both AUCs show the possibility of discriminating between pregnancy and no pregnancy.
It has been demonstrated previously that inhibin B can be used as an endocrine marker of spermatogenesis (Anawalt et al., 1996; Illingworth et al., 1996
; Jensen et al., 1997
; Kolb et al., 2000
) and therefore it is of great interest to investigate whether male inhibin B can be used as a marker of fecundability.
Other studies have examined the use of inhibin B as a biomarker; for instance, to predicting the likelihood of finding of sperm by surgical sperm retrieval. One study (Pierik et al., 1998) found that inhibin B levels were clearly correlated with the total sperm count and testicular volume. This finding is supported by the findings of another smaller (n = 17) study (Ballesca et al., 2000
), which found that serum inhibin B is a predictor of successful testicular sperm extraction in men with non-obstructive azoospermia. By forming a ROC curve, they found that the area under the inhibin B curve was 0.98. These findings are, however, not consistent with the findings in a study among 185 non-obstructive azoospermic patients (Vernaeve et al., 2002
). They also formed a ROC curve, and found the area under inhibin B curve to be 0.51.
The differences in the results of these studies might be due to differences in patient selection, differences in the number of patients included, and differences in endpoint.
In conclusion, our findings do not support the replacement of sperm density with male inhibin B when obtaining sperm data is an option. However, because inhibin B seems not to be confounded by lifestyle factors and other factors related to reproduction, it may be an alternative marker of male fecundity for epidemiological studies and may offer some practical advantages compared with sperm density.
![]() |
Acknowledgements |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|
![]() |
References |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|
Andersson AM, Carlsen E, Petersen JH and Skakkebaek NE (2003) Variation in levels of serum inhibin B, testosterone, estradiol, luteinizing hormone, follicle-stimulating hormone and sex hormone-binding globulin in monthly samples from healthy men during a 17-month period: possible effects of seasons. J Clin Endocrinol Metab 88,932937.
Andersson AM, Petersen JH, Jorgensen N, Jensen TK and Skakkebaek NE (2004) Serum inhibin B and follicle-stimulating hormone levels as tools in the evaluation of infertile men: significance of adequate reference values from proven fertile men. J Clin Endocrinol Metab 89,28732879.
Ballesca JL, Balasch J, Calafell JM, Alvarez R, Fabregues F, de Osaba MJ, Ascaso C and Vanrell JA (2000) Serum inhibin B determination is predictive of successful testicular sperm extraction in men with non-obstructive azoospermia. Hum Reprod 15,17341738.
Bergmann M, Behre HM and Nieschlag E (1994) Serum FSH and testicular morphology in male infertility. Clin Endocrinol (Oxf) 40,133136.[ISI][Medline]
Bonde JP, Giwercman A and Ernst E (1996) Identifying environmental risk to male reproductive function by occupational sperm studies: logistics and design options. Occup Environ Med 53,511519.[Abstract]
Bonde JP, Ernst E, Jensen TK, Hjollund NH, Kolstad H, Henriksen TB, Scheike T, Giwercman A, Olsen J and Skakkebaek NE (1998a) Relation between semen quality and fertility: a population-based study of 430 first-pregnancy planners. Lancet 352,11721177.[CrossRef][ISI][Medline]
Bonde JP, Hjollund NH, Jensen TK, Ernst E, Kolstad H, Henriksen TB, Giwercman A, Skakkebaek NE, Andersson AM and Olsen J (1998b) A follow-up study of environmental and biologic determinants of fertility among 430 Danish first-pregnancy planners: design and methods. Reprod Toxicol 12,1927.[CrossRef][ISI][Medline]
Cagnacci A, Maxia N and Volpe A (1999) Diurnal variation of semen quality in human males. Hum Reprod 14,106109.
Carlsen E, Olsson C, Petersen JH, Andersson AM and Skakkebaek NE (1999) Diurnal rhythm in serum levels of inhibin B in normal men: relation to testicular steroids and gonadotropins. J Clin Endocrinol Metab 84,16641669.
Cooper TG, Neuwinger J, Bahrs S and Nieschlag E (1992) Internal quality control of semen analysis. Fertil Steril 58,172178.[ISI][Medline]
De Gendt K, Swinnen JV, Saunders PT, Schoonjans L, Dewerchin M, Devos A, Tan K, Atanassova N, Claessens F, Lecureuil C et al (2004) A Sertoli cell-selective knockout of the androgen receptor causes spermatogenic arrest in meiosis. Proc Natl Acad Sci USA 101,13271332.
Groome NP, Illingworth PJ, OBrien M, Pai R, Rodger FE, Mather JP and McNeilly AS (1996) Measurement of dimeric inhibin B throughout the human menstrual cycle. J Clin Endocrinol Metab 81,14011405.[Abstract]
Hosmer DW and Lemeshow S (2000) Applied logistic regression. Wiley Interscience New York, N.Y.
Illingworth PJ, Groome NP, Byrd W, Rainey WE, McNeilly AS, Mather JP and Bremner WJ (1996) Inhibin-B: a likely candidate for the physiologically important form of inhibin in men. J Clin Endocrinol Metab 81,13211325.[Abstract]
Jensen TK, Andersson AM, Hjollund NH, Scheike T, Kolstad H, Giwercman A, Henriksen TB, Ernst E, Bonde JP, Olsen J et al (1997) Inhibin B as a serum marker of spermatogenesis: correlation to differences in sperm concentration and follicle-stimulating hormone levels. A study of 349 Danish men. J Clin Endocrinol Metab 82,40594063.
Jensen TK, Henriksen TB, Hjollund NH, Scheike T, Kolstad H, Giwercman A, Ernst E, Bonde JP, Skakkebaek NE and Olsen J (1998a) Adult and prenatal exposures to tobacco smoke as risk indicators of fertility among 430 Danish couples. Am J Epidemiol 148,992997.[Abstract]
Jensen TK, Henriksen TB, Hjollund NH, Scheike T, Kolstad H, Giwercman A, Ernst E, Bonde JP, Skakkebaek NE and Olsen J (1998b) Caffeine intake and fecundability: a follow-up study among 430 Danish couples planning their first pregnancy. Reprod Toxicol 12,289295.[CrossRef][ISI][Medline]
Jensen TK, Hjollund NH, Henriksen TB, Scheike T, Kolstad H, Giwercman A, Ernst E, Bonde JP, Skakkebaek NE and Olsen J (1998c) Does moderate alcohol consumption affect fertility? Follow up study among couples planning first pregnancy. BMJ 317,505510.
Juul S, Karmaus W and Olsen J (1999) Regional differences in waiting time to pregnancy: pregnancy-based surveys from Denmark, France, Germany, Italy and Sweden. The European Infertility and Subfecundity Study Group. Hum Reprod 14,12501254.
Klingmuller D and Haidl G (1997) Inhibin B in men with normal and disturbed spermatogenesis. Hum Reprod 12,23762378.[Abstract]
Kolb BA, Stanczyk FZ and Sokol RZ (2000) Serum inhibin B levels in males with gonadal dysfunction. Fertil Steril 74,234238.[CrossRef][ISI][Medline]
Larsen SB, Abell A and Bonde JP (1998) Selection bias in occupational sperm studies. Am J Epidemiol 147,681685.[Abstract]
Levine RJ, Mathew RM, Chenault CB, Brown MH, Hurtt ME, Bentley KS, Mohr KL and Working PK (1990) Differences in the quality of semen in outdoor workers during summer and winter. N Engl J Med 323,1216.[Abstract]
Makler A, Zaidise I, Paldi E and Brandes JM (1979) Factors affecting sperm motility. I. In vitro change in motility with time after ejaculation. Fertil Steril 31,147154.[ISI][Medline]
Olsen J (1992) EUROMAC. A European concerted action: maternal alcohol consumption and its relation to the outcome of pregnancy and child development at 18 months. Recommendations. Int J Epidemiol, 21 Suppl 1, S82S83.[Medline]
Pierik FH, Vreeburg JT, Stijnen T, De Jong FH and Weber RF (1998) Serum inhibin B as a marker of spermatogenesis. J Clin Endocrinol Metab 83,31103114.
Pierik FH, Burdorf A, De Jong FH and Weber RF (2003) Inhibin B: a novel marker of spermatogenesis. Ann Med 35,1220.[CrossRef][ISI][Medline]
Schwartz D, Laplanche A, Jouannet P and David G (1979) Within-subject variability of human semen in regard to sperm count, volume, total number of spermatozoa and length of abstinence. J Reprod Fertil 57,391395.[ISI][Medline]
Vernaeve V, Tournaye H, Schiettecatte J, Verheyen G, Van Steirteghem A and Devroey P (2002) Serum inhibin B cannot predict testicular sperm retrieval in patients with non-obstructive azoospermia. Hum Reprod 17,971976.
Weinberg CR (1994) The influence of occupational activity on the menstrual cycle and fecundability. Epidemiology 5,563564.[ISI][Medline]
Wilcox AJ, Weinberg CR, OConnor JF, Baird DD, Schlatterer JP, Canfield RE, Armstrong EG and Nisula BC (1988) Incidence of early loss of pregnancy. N Engl J Med 319,189194.[Abstract]
World Health Organization (1992) WHO laboratory manual for the examination of human semen and sperm-cervical mucus interaction. Cambridge University Press, Cambridge.
Yaman O, Ozdiler E, Seckiner I and Gogus O (1999) Significance of serum FSH levels and testicular morphology in infertile males. Int Urol Nephrol 31,519523.[CrossRef][Medline]
Zweig MH and Campbell G (1993) Receiver-operating characteristic (ROC) plots: a fundamental evaluation tool in clinical medicine. Clin Chem 39,561577.
Submitted on December 8, 2004; resubmitted on April 11, 2005; accepted on April 19, 2005.
|