TERATOLOGY 44:629-634 (1991)
Parental Fertility and Infant Hypospadias: An International Case-Control Study BENGT KALLfiN, EDUARDO E. CASTILLA, METTE KRINGELBACH, PAUL A.L. LANCASTER, MARfA LUISA MARTfNEZ-FRf AS, PIERPAOLO MASTROIACOVO, OSVALDO MUTCHINICK. AND ELISABETH ROBERT Department of Embryology, University o f l u n d , Sweden (B.K.); ECLAMCIGeneticalFiocruz,Rio de Janeiro, Brazil, and IMBICE, La Plata, Argentina (E.E.C.); National Board of Health, Copenhagen, Denmark (M.K.); National Perinatal Statistics Unit, University of Sydney, Australia (P.A.L.L.); Hospital Uniuersitario Sun Carlos, ECEMC, Facultad de Medicina, Universidad Complutense, Madrid, Spain (M.L.M-F.); Clinica Pediatrica, Universita Cattolica, Rome, Italy (P.M.); Departmento de Gendtica, Instituto Nacional de Nutricion Salvador Zubiran, Mexico D.F., Mexico ( O M . ) ;Institut Europeen des Genomutations, Lyon, France (E.R.)
ABSTRACT
The authors have performed a n international case-control study on the significance of exogenous hormones for the origin of hypospadias (Kallen et al., 1991a,b). Using data from this study on 846 infants with isolated hypospadias and equally many controls (next male infant born in the same hospital as the case), variables that might indicate a n increased rate of fertility problems in couples who had a boy with hypospadias were studied. Cases had slightly fewer previous pregnancies, there was practically no difference in the rate of previous induced abortions, and no demonstrable difference in menstrual history. Infertility periods of at least 6 months were more likely among cases than controls but the difference did not reach statistical significance in a twotailed test. There was no demonstrable difference in the time to conceive between cases and controls. The difficulties of directly studying subfertility problems are stressed.
Hypospadias develops as a result of faulty fusion of the urethral folds. The fusion process is controlled by androgens from the fetal testicle; it is thought that hypospadias results from a reduced androgen production in the fetal testicle, perhaps because of inadequate gonadotropin stimulation (Roberts and Lloyd, '73) andlor a reduced capacity of the urethral folds to respond to androgens (Svensson and Snochowski, '79). Commonly associated other abnormalities with hypospadias are undescended testicles and inguinal hernias (Svensson, '79; Shima et al., '79). Varying degrees of genetic contribution to the origin of hypospadias have been suggested, although the actual genetic mechanism is not agreed upon; autosomal dominant (Page, '79), autosomal recessive (Frydman et al., '85), and polygenic (Czeizel and Tusdany, '84) have been suggested. Undescended testicles are more common among the fathers of hypospadiac boys than 0 1991 WILEY-LISS, INC.
in the general population; other testicular abnormalities have also been demonstrated, including reduced spermatogenesis (Sweet et al., '74). Also, maternal endocrine disturbances have been suggested to occur with hypospadias. Polednak and Janerich ('83) and Czeizel and Toth ('90) found that women whose infants had hypospadias had a slightly later menarche than did control women, but Calzolari et al. ('86) found the opposite. Polednak and Janerich also described fewer asymptomatic menstrual cycles among case mothers than among controls, indicating a higher rate of anovulatory cycles. Czeizel and Toth ('90) compared the interval between discontinuation of contracep-
Received January 30, 1991; accepted August 27, 1991. Address reprint requests to Dr. Ben@ KaIIen, Department of Embryology, Biskopsgatan 7, 5-223 62 Lund, Sweden.
630
B. KALLQN ET AL.
tive activity and conception in 186 women with planned pregnancies and whose infants had hypospadias with that of 193 control women with planned pregnancies. The rate of unknown intervals was rather high (20% among cases and 12% among controls), but the authors found that a lower percentage of index women than control women conceived within 3 months and a higher percentage after 13 months. There was no difference in the rate of ovulation stimulation between the two groups. If parental gonad anomalies are more common for hypospadias than expected, it should result in a n increased rate of subfecundity or subfertility. In a n international epidemiological study of hypospadias (Kallen et al., ’86), it was found that the prevalence a t birth of hypospadias apparently varied markedly between different populations and that there was a n inverse relationship with fertility. That is, there was a high rate of hypospadias in populations in which fertility was low (judged from mean parity), and vice versa. In populations with a low reproductive rate, the percentage of couples with fertility problems among those having a n infant, should be larger than in populations with a high reproductive rate. The effect of maternal age and parity was also studied. A markedly increased risk was seen in women aged 40 + years having their first child-3-4 times the average risk. In other age classes of parity 1, or in other parity classes aged 40+ years, no increased risk was seen. The possible association between parental subfertility and hypospadias has been suggested as a n explanation (Kallen and Winberg, ’82; Czeizel, ’85; Czeizel and Toth, ’90) for the increase in hypospadias rate, seen in many countries. Equally, this increase might be due to improving ascertainment of a malformation that has varying degrees of severity and therefore recognition. There are many difficulties in the epidemiological identification of fertility problems and their association with a specific malformation. One possibility is to use indirect measurements such as the number of previous pregnancies or deliveries or the number of previous induced abortions (which should be reduced with subfertility); another is to try to record the number of years of unwanted infertility; and a third is to use the concept “time to pregnancy,”
which measures the number of months from when a couple tries to conceive until conception actually occurs. The latter concept has become very popular during recent years in discussions on possible hazards at working places (Rachootin and Olsen, ’83; Baird et al., ’86). In the present study, we have used data collected in a n international case-control study on factors associated with hypospadias. The study was aimed primarily at the question of hormone treatments (Kallen et al., ’91a) and use of oral contraceptives (Kallen et al., ’91b), but in order to analyze these features, data were collected also on factors associated with fertility. MATERIALS AND METHODS
The study was conducted in eight malformation monitoring programs that participate in the International Clearinghouse for Birth Defects Monitoring Systems. An agreed questionnaire was used when interviewing the 846 mothers of singleton infants with isolated hypospadias and 846 controls (mothers of next singleton male infants born a t the same hospital a s the case). Most interviews were conducted shortly after birth. Details of the study have been given (Kallen et al., ’91a). The following questions were included (which may shed light on fertility problems): 1. Age at menarche 2. Menstrual history during the year before the pregnancy among women who did not use oral contraceptives 3. Interval between menstrual cycles in days 4. Previous pregnancy outcomes 5. Fertility problems (defined as inability to achieve pregnancy for at least 6 months); duration of the period of infertility 6. Examination and treatment for fertility problems, male + female factors 7. Planned pregnancies: number of months required to achieve pregnancy. Information on maternal age, parental occupations, and family history of hypospadias was recorded. Heterogeneity in frequency tables was tested with chi-square tests without Yates’ continuity correction. Odds ratios (OR) were also calculated by comparing discordant pairs (case exposed, control nonexposed and vice versa) with the expected binomial distribution (P = 0.5) after normal approxima-
HYPOSPADIAS AND PARENTAL FERTILITY
63 1
TABLE 1. Social characteristics of cases and controls based on maternal and paternal occupations Occupation category No known occupation (including housewives) Workers (including farming) Clerks Professionals (including administration and executives) Chi-square for heterogeneity, 3 df
Cases 436 142 112 156
Maternal Controls 436 164 114 132 3.60
Cases 70 529 101 140
Paternal Controls 66 533 101 146 0.24
tion. Stratification for program was made with the Mantel-Haenszel procedure without Yates’ correction. The 95% confidence intervals (95%CI) of odds ratios were calculated according to Miettinen’s approximation. RESULTS
Social situation As some of the variables that were studied may co-vary with education and social situation a check was first made that no major differences exist between cases and controls from the point of view of social conditions evaluated from maternal and paternal occupations. Table 1 shows the distributions recorded-there were no significant differences seen between cases and controls. Menstrual history Figure 1 shows the distribution of the age at menarche in cases and controls-no major differences are seen. Information was available on 798 cases-31 had a menarche before the age of 11 and 31 after the age of 15. Among 800 controls with information, 34 had menarche before the age at 11 and 35 after the age of 15. Thus, cases had slightly fewer instances of very early or late menarche than did controls. Women who had not used oral contraceptives the year before the pregnancy were asked about irregular menstruation during that year. In the total material, there is no difference in the rate of irregular menstruations between cases and controls (Table 2). This is true also for the individual programs except for Denmark, where 10 cases among 50 (20%)reported irregular menses but only 3 among 57 (5.3%) controls (restricted to women who did not use oral contraceptive). In Italy, the opposite is seen: 46 among 148 (31.1%) cases and 60 among 148 (40.5%) controls. Both variations are probably random. There is no difference in cycle length between cases and controls. Information was
7
9
II
13
15
17
19
RGE RT MENRRCHE
Fig. 1. Distribution of age at menarche in the total and controls (...). material for cases (-)
available for 822 pairs-the mean difference in cycle length between case and control was -0.1 month with a SEM of 0.1. Nine cases and 8 controls had a cycle length 32 days.
Previous pregnancies There were slightly fewer previous pregnancies among cases than controls, but the difference was not significant. Among the 846 cases, 477 (56.4%)women had a total of 967 previous pregnancies (mean 2.0 per woman); among the 846 controls, 513 (60.6%)women had a total of 1,002 previous pregnancies (mean 2.0 per woman). The difference in the number of women with previous pregnancies is not significant (chi-square = 3.2, NS), neither is the number of previous pregnancies (chi-square = 0.6, NS). Induced abortions among previous pregnancies could be a n indicator of fecundability, as it is reasonable to suppose that subfertile women will have fewer induced abortions than fertile women will. Induced abortions were not reported from Mexico or
632
B. KALLfiN ET AL. TABLE 2 . Menstrual irregularities the year before conceDtion in cases and controls All pairs No heredity Heredity Cases Controls Cases Controls Cases Controls
Year before urennancv No information on oral contraceptive use Oral contraceptives No oral contraceptives No information on type Regular menses Irregular menses
24 221 601 25 455 121
South America and these two programs were therefore removed from this part of the analysis. Among the remaining programs, there were 90 induced abortions among 728 previous pregnancies among the cases (12.4%) and 103 induced abortions among 779 previous control pregnancies (13.2%). The difference is not statistically significant (after stratification for program as induced abortion rate varies strongly between programs: OR = 0.97, 95% CI 0.71-1.33).
20 207 619 27 462 130
22 193 541 23 409 109
17 181 558 24 421 113
2 28 40 1 46 12
3 26 61 3 41
17
thers (17 cases, 17 controls) and in 78 mothers (40 cases and 38 controls). Some sort of medical treatment for infertility had been given to 48 women: 27 cases and 21 controls. Time to conceive Figure 2 demonstrates another variable indicating fertility problems: time to pregnancy in planned pregnancies. Cases with no reported family occurrence of hypospadias (“no heredity”) behave exactly as controls, but among cases with a positive family history (“heredity”), the cumulative graph lies higher than that of controls up to 6 months. This is, however, based on small numbers. Among 49 hereditary cases with known time to pregnancy, 43 conceived within 6 months (88%), while among 442 such controls, 342 conceived within 6 months (77%). The difference may well be random (chi-square = 2.8, NS).
Stated fertility problems The women were asked to report whether they had tried to conceive without success for more than 6 months (Table 3). This information was lacking in 10% of the cases and 11%of the controls: 135 of the 179 reports lacking information came from the Italian program and 24 from the Spanish program. Among those answering the question, there were somewhat more cases DISCUSSION (16.2%) than controls (14.1%) stating a period of infertility. The difference is not staAs discussed in the Introduction, there tistically significant: odds ratio is 1.36 with are in the literature some indications that a 95% CI of 0.89-1.57. hypospadias may be related to parental subThe period of infertility was known in 120 fertility. Most evidence refers to paternal cases and 104 controls. More cases than con- subfertility (Sweet et al., ’74), but abnortrols reported a n infertility period of 6-12 malities in the menstrual history of the months, while there was no difference for mothers have also been described (Polednak infertility lasting >1 year. In 59 pairs, the and Janerich, ’83). case but not the control were infertile for Our study is the largest case-control between 6 and 12 months and in 43 pairs study of hypospadias published so far. We the opposite was true. This difference is not tried to identify menstrual irregularities in significant in a two-tailed test (OR = 1.37, the mothers and found no difference be95% CI = 0.93-2.03). tween cases and controls. The age at menIf the material is divided into two groups, arche was similarly distributed; in the preone with a known family history of hypo- vious literature, Polednak and Janerich spadias and the other without such a his- (’83) and Czeizel and T6th (’90) found that tory, the difference in infertility rate is sta- cases had a slightly later menarche than tistically nearly significant in the former controls, but Calzolari e t al. (’86) found the group (OR 1.37, 95% CI 0.97-1.94), but no opposite. Probably no difference exists. We association is seen in the latter group (8 found no increased rate of irregular menamong 83 cases, 14 among 85 controls, OR struations in cases compared with controls. = 0.56, 95% CI 0.19-1.66). Medical examiWe did not study the variable asymptomatic nation for infertility was reported in 34 fa- cycles used by Polednak and Janerich as we
HYPOSPADIAS AND PARENTAL FERTILITY
633
TABLE 3. Stated fertility problems (trted to conceive 6 months or more) in cases and controls’ All pairs No heredity Heredity Fertility status Cases Controls Cases Controls Cases Controls 7 5 87 92 80 87 No information 75 71 636 648 561 577 No fertility problem 115 92 8 14 Fertility problem 123 106 0 1 3 2 3 1 Time not stated 5 6 73 55 68 49
Parental Fertility and Infant Hypospadias: An International Case-Control Study BENGT KALLfiN, EDUARDO E. CASTILLA, METTE KRINGELBACH, PAUL A.L. LANCASTER, MARfA LUISA MARTfNEZ-FRf AS, PIERPAOLO MASTROIACOVO, OSVALDO MUTCHINICK. AND ELISABETH ROBERT Department of Embryology, University o f l u n d , Sweden (B.K.); ECLAMCIGeneticalFiocruz,Rio de Janeiro, Brazil, and IMBICE, La Plata, Argentina (E.E.C.); National Board of Health, Copenhagen, Denmark (M.K.); National Perinatal Statistics Unit, University of Sydney, Australia (P.A.L.L.); Hospital Uniuersitario Sun Carlos, ECEMC, Facultad de Medicina, Universidad Complutense, Madrid, Spain (M.L.M-F.); Clinica Pediatrica, Universita Cattolica, Rome, Italy (P.M.); Departmento de Gendtica, Instituto Nacional de Nutricion Salvador Zubiran, Mexico D.F., Mexico ( O M . ) ;Institut Europeen des Genomutations, Lyon, France (E.R.)
ABSTRACT
The authors have performed a n international case-control study on the significance of exogenous hormones for the origin of hypospadias (Kallen et al., 1991a,b). Using data from this study on 846 infants with isolated hypospadias and equally many controls (next male infant born in the same hospital as the case), variables that might indicate a n increased rate of fertility problems in couples who had a boy with hypospadias were studied. Cases had slightly fewer previous pregnancies, there was practically no difference in the rate of previous induced abortions, and no demonstrable difference in menstrual history. Infertility periods of at least 6 months were more likely among cases than controls but the difference did not reach statistical significance in a twotailed test. There was no demonstrable difference in the time to conceive between cases and controls. The difficulties of directly studying subfertility problems are stressed.
Hypospadias develops as a result of faulty fusion of the urethral folds. The fusion process is controlled by androgens from the fetal testicle; it is thought that hypospadias results from a reduced androgen production in the fetal testicle, perhaps because of inadequate gonadotropin stimulation (Roberts and Lloyd, '73) andlor a reduced capacity of the urethral folds to respond to androgens (Svensson and Snochowski, '79). Commonly associated other abnormalities with hypospadias are undescended testicles and inguinal hernias (Svensson, '79; Shima et al., '79). Varying degrees of genetic contribution to the origin of hypospadias have been suggested, although the actual genetic mechanism is not agreed upon; autosomal dominant (Page, '79), autosomal recessive (Frydman et al., '85), and polygenic (Czeizel and Tusdany, '84) have been suggested. Undescended testicles are more common among the fathers of hypospadiac boys than 0 1991 WILEY-LISS, INC.
in the general population; other testicular abnormalities have also been demonstrated, including reduced spermatogenesis (Sweet et al., '74). Also, maternal endocrine disturbances have been suggested to occur with hypospadias. Polednak and Janerich ('83) and Czeizel and Toth ('90) found that women whose infants had hypospadias had a slightly later menarche than did control women, but Calzolari et al. ('86) found the opposite. Polednak and Janerich also described fewer asymptomatic menstrual cycles among case mothers than among controls, indicating a higher rate of anovulatory cycles. Czeizel and Toth ('90) compared the interval between discontinuation of contracep-
Received January 30, 1991; accepted August 27, 1991. Address reprint requests to Dr. Ben@ KaIIen, Department of Embryology, Biskopsgatan 7, 5-223 62 Lund, Sweden.
630
B. KALLQN ET AL.
tive activity and conception in 186 women with planned pregnancies and whose infants had hypospadias with that of 193 control women with planned pregnancies. The rate of unknown intervals was rather high (20% among cases and 12% among controls), but the authors found that a lower percentage of index women than control women conceived within 3 months and a higher percentage after 13 months. There was no difference in the rate of ovulation stimulation between the two groups. If parental gonad anomalies are more common for hypospadias than expected, it should result in a n increased rate of subfecundity or subfertility. In a n international epidemiological study of hypospadias (Kallen et al., ’86), it was found that the prevalence a t birth of hypospadias apparently varied markedly between different populations and that there was a n inverse relationship with fertility. That is, there was a high rate of hypospadias in populations in which fertility was low (judged from mean parity), and vice versa. In populations with a low reproductive rate, the percentage of couples with fertility problems among those having a n infant, should be larger than in populations with a high reproductive rate. The effect of maternal age and parity was also studied. A markedly increased risk was seen in women aged 40 + years having their first child-3-4 times the average risk. In other age classes of parity 1, or in other parity classes aged 40+ years, no increased risk was seen. The possible association between parental subfertility and hypospadias has been suggested as a n explanation (Kallen and Winberg, ’82; Czeizel, ’85; Czeizel and Toth, ’90) for the increase in hypospadias rate, seen in many countries. Equally, this increase might be due to improving ascertainment of a malformation that has varying degrees of severity and therefore recognition. There are many difficulties in the epidemiological identification of fertility problems and their association with a specific malformation. One possibility is to use indirect measurements such as the number of previous pregnancies or deliveries or the number of previous induced abortions (which should be reduced with subfertility); another is to try to record the number of years of unwanted infertility; and a third is to use the concept “time to pregnancy,”
which measures the number of months from when a couple tries to conceive until conception actually occurs. The latter concept has become very popular during recent years in discussions on possible hazards at working places (Rachootin and Olsen, ’83; Baird et al., ’86). In the present study, we have used data collected in a n international case-control study on factors associated with hypospadias. The study was aimed primarily at the question of hormone treatments (Kallen et al., ’91a) and use of oral contraceptives (Kallen et al., ’91b), but in order to analyze these features, data were collected also on factors associated with fertility. MATERIALS AND METHODS
The study was conducted in eight malformation monitoring programs that participate in the International Clearinghouse for Birth Defects Monitoring Systems. An agreed questionnaire was used when interviewing the 846 mothers of singleton infants with isolated hypospadias and 846 controls (mothers of next singleton male infants born a t the same hospital a s the case). Most interviews were conducted shortly after birth. Details of the study have been given (Kallen et al., ’91a). The following questions were included (which may shed light on fertility problems): 1. Age at menarche 2. Menstrual history during the year before the pregnancy among women who did not use oral contraceptives 3. Interval between menstrual cycles in days 4. Previous pregnancy outcomes 5. Fertility problems (defined as inability to achieve pregnancy for at least 6 months); duration of the period of infertility 6. Examination and treatment for fertility problems, male + female factors 7. Planned pregnancies: number of months required to achieve pregnancy. Information on maternal age, parental occupations, and family history of hypospadias was recorded. Heterogeneity in frequency tables was tested with chi-square tests without Yates’ continuity correction. Odds ratios (OR) were also calculated by comparing discordant pairs (case exposed, control nonexposed and vice versa) with the expected binomial distribution (P = 0.5) after normal approxima-
HYPOSPADIAS AND PARENTAL FERTILITY
63 1
TABLE 1. Social characteristics of cases and controls based on maternal and paternal occupations Occupation category No known occupation (including housewives) Workers (including farming) Clerks Professionals (including administration and executives) Chi-square for heterogeneity, 3 df
Cases 436 142 112 156
Maternal Controls 436 164 114 132 3.60
Cases 70 529 101 140
Paternal Controls 66 533 101 146 0.24
tion. Stratification for program was made with the Mantel-Haenszel procedure without Yates’ correction. The 95% confidence intervals (95%CI) of odds ratios were calculated according to Miettinen’s approximation. RESULTS
Social situation As some of the variables that were studied may co-vary with education and social situation a check was first made that no major differences exist between cases and controls from the point of view of social conditions evaluated from maternal and paternal occupations. Table 1 shows the distributions recorded-there were no significant differences seen between cases and controls. Menstrual history Figure 1 shows the distribution of the age at menarche in cases and controls-no major differences are seen. Information was available on 798 cases-31 had a menarche before the age of 11 and 31 after the age of 15. Among 800 controls with information, 34 had menarche before the age at 11 and 35 after the age of 15. Thus, cases had slightly fewer instances of very early or late menarche than did controls. Women who had not used oral contraceptives the year before the pregnancy were asked about irregular menstruation during that year. In the total material, there is no difference in the rate of irregular menstruations between cases and controls (Table 2). This is true also for the individual programs except for Denmark, where 10 cases among 50 (20%)reported irregular menses but only 3 among 57 (5.3%) controls (restricted to women who did not use oral contraceptive). In Italy, the opposite is seen: 46 among 148 (31.1%) cases and 60 among 148 (40.5%) controls. Both variations are probably random. There is no difference in cycle length between cases and controls. Information was
7
9
II
13
15
17
19
RGE RT MENRRCHE
Fig. 1. Distribution of age at menarche in the total and controls (...). material for cases (-)
available for 822 pairs-the mean difference in cycle length between case and control was -0.1 month with a SEM of 0.1. Nine cases and 8 controls had a cycle length 32 days.
Previous pregnancies There were slightly fewer previous pregnancies among cases than controls, but the difference was not significant. Among the 846 cases, 477 (56.4%)women had a total of 967 previous pregnancies (mean 2.0 per woman); among the 846 controls, 513 (60.6%)women had a total of 1,002 previous pregnancies (mean 2.0 per woman). The difference in the number of women with previous pregnancies is not significant (chi-square = 3.2, NS), neither is the number of previous pregnancies (chi-square = 0.6, NS). Induced abortions among previous pregnancies could be a n indicator of fecundability, as it is reasonable to suppose that subfertile women will have fewer induced abortions than fertile women will. Induced abortions were not reported from Mexico or
632
B. KALLfiN ET AL. TABLE 2 . Menstrual irregularities the year before conceDtion in cases and controls All pairs No heredity Heredity Cases Controls Cases Controls Cases Controls
Year before urennancv No information on oral contraceptive use Oral contraceptives No oral contraceptives No information on type Regular menses Irregular menses
24 221 601 25 455 121
South America and these two programs were therefore removed from this part of the analysis. Among the remaining programs, there were 90 induced abortions among 728 previous pregnancies among the cases (12.4%) and 103 induced abortions among 779 previous control pregnancies (13.2%). The difference is not statistically significant (after stratification for program as induced abortion rate varies strongly between programs: OR = 0.97, 95% CI 0.71-1.33).
20 207 619 27 462 130
22 193 541 23 409 109
17 181 558 24 421 113
2 28 40 1 46 12
3 26 61 3 41
17
thers (17 cases, 17 controls) and in 78 mothers (40 cases and 38 controls). Some sort of medical treatment for infertility had been given to 48 women: 27 cases and 21 controls. Time to conceive Figure 2 demonstrates another variable indicating fertility problems: time to pregnancy in planned pregnancies. Cases with no reported family occurrence of hypospadias (“no heredity”) behave exactly as controls, but among cases with a positive family history (“heredity”), the cumulative graph lies higher than that of controls up to 6 months. This is, however, based on small numbers. Among 49 hereditary cases with known time to pregnancy, 43 conceived within 6 months (88%), while among 442 such controls, 342 conceived within 6 months (77%). The difference may well be random (chi-square = 2.8, NS).
Stated fertility problems The women were asked to report whether they had tried to conceive without success for more than 6 months (Table 3). This information was lacking in 10% of the cases and 11%of the controls: 135 of the 179 reports lacking information came from the Italian program and 24 from the Spanish program. Among those answering the question, there were somewhat more cases DISCUSSION (16.2%) than controls (14.1%) stating a period of infertility. The difference is not staAs discussed in the Introduction, there tistically significant: odds ratio is 1.36 with are in the literature some indications that a 95% CI of 0.89-1.57. hypospadias may be related to parental subThe period of infertility was known in 120 fertility. Most evidence refers to paternal cases and 104 controls. More cases than con- subfertility (Sweet et al., ’74), but abnortrols reported a n infertility period of 6-12 malities in the menstrual history of the months, while there was no difference for mothers have also been described (Polednak infertility lasting >1 year. In 59 pairs, the and Janerich, ’83). case but not the control were infertile for Our study is the largest case-control between 6 and 12 months and in 43 pairs study of hypospadias published so far. We the opposite was true. This difference is not tried to identify menstrual irregularities in significant in a two-tailed test (OR = 1.37, the mothers and found no difference be95% CI = 0.93-2.03). tween cases and controls. The age at menIf the material is divided into two groups, arche was similarly distributed; in the preone with a known family history of hypo- vious literature, Polednak and Janerich spadias and the other without such a his- (’83) and Czeizel and T6th (’90) found that tory, the difference in infertility rate is sta- cases had a slightly later menarche than tistically nearly significant in the former controls, but Calzolari e t al. (’86) found the group (OR 1.37, 95% CI 0.97-1.94), but no opposite. Probably no difference exists. We association is seen in the latter group (8 found no increased rate of irregular menamong 83 cases, 14 among 85 controls, OR struations in cases compared with controls. = 0.56, 95% CI 0.19-1.66). Medical examiWe did not study the variable asymptomatic nation for infertility was reported in 34 fa- cycles used by Polednak and Janerich as we
HYPOSPADIAS AND PARENTAL FERTILITY
633
TABLE 3. Stated fertility problems (trted to conceive 6 months or more) in cases and controls’ All pairs No heredity Heredity Fertility status Cases Controls Cases Controls Cases Controls 7 5 87 92 80 87 No information 75 71 636 648 561 577 No fertility problem 115 92 8 14 Fertility problem 123 106 0 1 3 2 3 1 Time not stated 5 6 73 55 68 49
