Menstrual and reproductive history of mothers of galactosemic children*

Transcription

1 FERTILITY AND STERILITY Vol. 65, No.3, March 1996 Copyright IQ 1996 American Society for Reproductive Medicine Printed on acid-free paper in U. S. A. Menstrual and reproductive history of mothers of galactosemic children* Amy E. Sayle, M.P.H. t Glinda S. Cooper, Ph.D.t David A. Savitz, Ph.D.t University of North Carolina at Chapel Hill, Chapel Hill, and National Institute of Environmental Health Sciences, Research Triangle Park, North Carolina Objective: To test the hypothesis that female carriers of the galactosemia gene (mothers of galactosemic children) experience impaired fecundability and earlier age at natural menopause. Design: Retrospective cohort study comparing experiences of mothers of galactosemic children with mothers of children with phenylketonuria (PKU). Data collection consisted of a structured telephone interview with sections on menstrual, reproductive, and medical histories. Participants: One hundred three galactosemia carriers and 116 PKU carriers (mean age 35.0 and 38.4 years, respectively), recruited from 38 metabolic disease clinics throughout the United States. Main Outcome Measures: Time-to-pregnancy (a measure offecundability), menstrual cycle patterns, and natural menopause. Results: No difference was seen between carriers of the galactosemia and the PKU gene in history of 2:12 months of unprotected intercourse without a pregnancy, physician-diagnosed infertility, or amenorrhea lasting >6 weeks. The fecundability ratio, adjusting for age and smoking before most recent pregnancy, was 1.17 (95% confidence interval [CI] 0.85 to 1.61). By age 45, 11% of galactosemia carriers who had not had a hysterectomy reported a natural menopause compared with 4% ofpku carriers (risk difference 7.2%, 95% CI -12.7% to 27.1%). Conclusion: Fecundability and other measures of fertility did not differ between mothers of galactosemic children and mothers of children with PKU. Suggestions of an earlier age at natural menopause can not be addressed definitively in this cohort because of the small number of participants who were currently age 45 years or older but may warrant additional assessment as the cohort ages. Fertil Steril 1996;65:534-8 Key Words: Heterozygous galactosemia carriers, menopausal status, infertility, ovarian toxicity Galactosemia is an inherited metabolic disease that results in a complete deficiency of galactose-1- phosphate uridyl transferase (transferase). In adult Received August 3, 1995; accepted September 14, * Supported in part by Reproductive Hazards in the Workplace, Home, Community, and Environment Research grant (15-FY and 15-FY ), March of Dimes Birth Defects Foundation, White Plains, New York. Ms. Sayle is a Howard Hughes Medical Institute Predoctoral Fellow. t Department of Epidemiology, University of North Carolina at Chapel Hill. :j: Epidemiology Branch, National Institute of Environmental Health Sciences. Reprint requests: David A. Savitz, Ph.D., Department ofepidemiology, McGavran-Greenberg CB No. 7400, University of North Carolina at Chapel Hill, Chapel Hill, North Carolina (FAX: ). females, effects include primary and secondary amenorrhea, infertility, and premature ovarian failure (before age 30 years) (1). In 1989, Cramer et al. (2) reported evidence of a lower age at natural menopause and higher prevalence of infertility among women with low transferase activity (e.g., heterozygous carriers of the galactosemia allele or another low-activity allele, the Duarte variant). Kaufman et al. (3) examined this question using data from a survey of 108 mothers of galactosemic children (i.e., known galactosemia carriers) and did not see increases in premature menopause or pregnancy loss compared with expected population rates. However, this study did not include a comparison group and did not address specifically other measures of reproductive function, such as time-to-pregnancy (a measure offecundability) or clearly defined 534 Sayle et al. Reproductive Hx of galactosemia carriers Fertility and Sterility

2 amenorrhea or natural menopause. We used structured telephone interviews to investigate the reproductive and medical history of mothers of children with galactosemia and a comparison group, mothers of children with phenylketonuria (PKU). MATERIALS AND METHODS We contacted 51 metabolic disease clinics throughout the United States that currently were treating galactosemic patients, and 38 agreed to participate. The clinics sent a standardized letter describing the study to the mothers of their patients with classic galactosemia. Mothers of adopted children and women who could not speak English were not included. The women who were interested in participating contacted our study office and arranged a time for the telephone interview. The recruitment procedure was limited to this one letter, and we were not given the names of any patients who did not volunteer to participate. Eleven PKU clinics were selected from this list in order to cover a broad geographic area, and PKU mothers were enrolled in a similar manner. Letters were sent to 259 galactosemia mothers and 103 (40%) completed interviews; 238 PKU mothers were contacted, and 116 (49%) were interviewed. Interviews were conducted in 1993 and 1994, and each lasted approximately 20 to 30 minutes. Information was obtained using a structured questionnaire and included demographic characteristics, menstrual and reproductive history, medical history, oral contraceptive and estrogen replacement therapy use, smoking history, and family medical history. Detailed information about use of contraceptives before pregnancy and the number of cycles it took to get pregnant was obtained for the participant's most recent pregnancy. The study protocol was approved by the Institutional Review Board at the University of North Carolina and at each of the clinics as required. Univariate differences in means between the two groups (galactosemia and PKU carriers) were analyzed using t-tests, and relative risks (RR) with 95% confidence intervals (Cl) were calculated for categorical outcome measures. Logistic regression was used to control for age, with the calculated odds ratio (OR) from this procedure taken as an approximation of the relative risk. We used a discrete form of a Cox proportional hazards model, as described by Weinberg et al. (4), to calculate the probability (or risk) of conception per menstrual cycle. The fecundability ratio produced by this analysis is analogous to a risk ratio in that it represents a comparison of the conception probabilities between groups; values < 1.0 indicate reduced fecundability. A Kaplan-Meier life Vol. 65, No.3, March 1996 table analysis was used to analyze age at natural menopause, with age at hysterectomy or current age used as censoring events (5). RESULTS Participants ranged in age from 19 to 64 years, but galactosemia carriers were younger than PKU carriers (mean age 35.0 versus 38.4 years, P = 0.002). Approximately 85% were currently married, 27% had completed college, 97% were white, and 42% were current or former smokers. There were no notable differences in these characteristics between the groups. Galactosemia and PKU carriers did not differ with respect to mean age at menarche (12.6 and 12.5 years, respectively, P = 0.47) or the proportion who experienced menarche at age 15 years or later (11.7% and 9.5%, RR 1.2, 95% CI 0.56 to 2.51) (Table 1). There was little difference in the number of years after menarche until cycles became regular, and similar proportions in each group reported a history of >6 weeks of amenorrhea. Galactosemia carriers were almost twice as likely as PKU carriers to report that their cycle length varied by a week or more (age adjusted estimate, OR = 1.93, 95% CI 0.75 to 4.98), although this difference was not statistically significant. There were few differences with respect to reproductive history between galactosemia and PKU carriers. Mean gravidity was 2.9 in both groups (P = 0.94), and mean number oflive births was 2.2 and 2.3 in the galactosemia and PKU carriers, respectively. One galactosemia carrier reported a stillbirth. There was no difference in the risk of spontaneous abortion between groups (Table 1); adjusting for age, the relative risk of a spontaneous abortion was 1.17 (95% CI 0.72 to 1.88) among galactosemia carriers. Galactosemia carriers showed little evidence for an increased risk of infertility, defined as :2::12 months of unprotected intercourse without a pregnancy (RR 1.1, 95% CI 0.78 to 1.63). Similar proportions in each group had seen a physician because of concern about infertility (RR 1.0, 95% CI 0.55 to 1.83) and, among these women, galactosemia carriers were somewhat less likely to have been diagnosed with an ovulatory problem (35% of 17 galactosemia carriers who had seen a physician compared with 53% of 19 PKU carriers, RR 0.67, 95% CI 0.31 to 1.45). Thirty-two percent of galactosemia carriers and 37% of PKU carriers reported that their most recent pregnancy had occurred when they were using some form of birth control (RR 0.86, 95% CI 0.60 to 1.25). Among couples who were not using any contraceptive method, the median time to pregnancy for the Sayle et al. Reproductive Hx of galactosemia carriers 535

3 Table 1 Menstrual Cycle Characteristics and History of Spontaneous Abortion in Carriers of Galactosemia and of PKU* Galactosemia carriers PKU carriers 95% confidence (n = 103) (n = 116) Relative risk interval Menarche age (y),,;;11 19 (18.4) 12 to (59.2) (10.7) ;,,15 12 (11.7) Years after menarche before menstrual cycles became regulart <1 to 2 80 (78) 3 to 4 4 (4) ;,,5 9 (9) Never 9 (9) Amenorrhea lasting >6 weeks:j: Yes 26 (25) No 77 (75) Usual menstrual cycle length (d),,;;25 16 (21) 26 to (59) ;,,36 1 (1) Varies by > 1 week 14 (18) Spontaneous abortions Percent of pregnancies 39 (13) Percent of women with one or more spontaneous abortion 28 (27) * Values in parentheses are percentages. t One galactosemia carrier and two PKU carriers excluded because of missing data. 26 (22.4) to (58.6) 1.0 Reference 11 (9.5) to (9.5) to (73) 1.0 Reference 5 (4) to (15) to (8) to (23) to (77) 1.0 Reference 17 (22) to (67) 1.0 Reference 0(0) 8 (10) to (12) to (26) to 1.63 :j: Excluding pregnancy and lactation. Current postmenopausal women and users of oral contraceptives, estrogen replacement therapy, or other hormones excluded. Relative risk not calculated for strata with zero observations. most recent pregnancy (three cycles) was identical for galactosemia and PKU carriers (Fig. 1). The fecundability ratio comparing galactosemia with PKU carriers and adjusting for age (at the time the pregnancy attempt began) and smoking (during the year before the pregnancy) was 1.17 (95% CI 0.85 to 1.61). Further, there was no evidence of reduced fecundability among galactosemia carriers when galactose consumption in the year before the preg- C <II c: 01 ~ Il.. C Ql ~ Ql Il Menstrual Cycle Figure 1 Cumulative percent pregnant across menstrual cycle, among noncontracepting women. 0, Galactosemia carriers (n = 70);., carriers of PKU (n = 72). 536 Sayle et al. Reproductive Hx of galactosemia carriers nancy began was included as an interaction variable in the analysis. Data were censored at six cycles in these analyses. One galactosemia carrier (currently age 33 years) reported that she had been diagnosed with premature ovarian failure at age 26; no PKU carriers reported this condition. By age 45, 11% of galactosemia carriers and 4% of the PKU carriers who had not had a hysterectomy reported a natural menopause. However, there were only 10 galactosemia carriers at this age, so these estimates are imprecise: the risk difference was 7.2% (95% CI -12.7% to 27.1%). Surgical menopause was more common among PKU carriers (15%) than among galactosemia carriers (2%). Because galactosemia often results in cataract formation, we also asked about the prevalence of this condition among the study group. One galactosemia carrier reported a history of physician-diagnosed cataracts (at age 29 years) compared with 3 (3%) PKU carriers (ages 26, 47, and 59 years). Approximately equal proportions of galactosemia and PKU carriers reported ever having ovarian cysts (23% versus 22%). No women reported ever having been diagnosed with polycystic ovarian disease. Mter excluding data from women who did not provide information about family medical history, 5 of 98 (5%) galactosemia carriers reported that their mothers had had ovarian cancer compared with 3 of Fertility and Sterility

4 112 (3%) PKU carriers (Fisher's exact two-tailed P = 0.48). Among women who had at least one sister, 2 of 68 galactosemia carriers and 2 of 78 PKU carriers reported having sisters who had had ovarian cancer. DISCUSSION We did not see an increased risk of spontaneous abortion, infertility (defined as 2: 12 months of unprotected intercourse without conception) or reduced fecundability (measured by time-to-pregnancy among noncontracepting couples) among these mothers of galactosemic children compared with mothers of children with PKU. These results with respect to reproductive history generally are consistent with the study reported by Kaufman et al. (3) of members of Parents of Galactosemic Children. Although there was a trend toward earlier age at menopause among the galactosemia carriers, the number of women age 2: 45 years was too small to ascertain definitively the patterns of menopause in these two groups at this time. Cigarette smoking decreases age at menopause by 1 to 2 years (6), but none of the galactosemia carriers and only one PKU carrier reported smoking cigarettes within the 5 years preceding menopause. Thus, it is unlikely that our observation represents a spurious association with smoking. The one case of premature ovarian failure (age 26 years) in a galactosemia carrier is interesting, but the overall risk of premature ovarian failure (1%) in this group is similar to that seen in a population-based study from Rochester, Minnesota (7). Assessment of the menopausal status of these women after 2:5 years of follow-up may be warranted. As with the previous study by Kaufman et al. (3), our study sample consisted of women who were known to have had at least one live birth. This may not represent the actual fertility experience among galactosemia (or PKU) carriers because women who could not get pregnant would not have been included in our sample. An alternative design would be to include all carriers, identified through some kind of population-based screening program. Because the estimated prevalence of the galactosemia gene in the general population is < 1 %, this type of study design would require screening at least 10,000 blood samples in order to find 100 carriers. We chose a comparison group, mothers of children with PKU, that we believed would be similar to the galactosemia carriers with respect to recall of reproductive history and access to the clinics through which we recruited the galactosemia carriers. However, ifpku carriers are at higher risk of menstrual and reproductive problems than the general population, then comparison of the galactosemia group Vol. 65, No.3, March 1996 with the PKU group would produce biased (attenuated) estimates of the effects of being a carrier for the galactosemia gene. However, comparison with other sources of data suggests that the characteristics of PKU carriers were similar to expected values with respect to age at menarche, age at natural menopause, time to pregnancy, and spontaneous abortions (8-10). Thus, it seems that the bias introduced by our choice of comparison group is likely to be small. A response bias could have affected our results, for example, if the PKU carriers who have had reproductive or menstrual problems were more likely to participate than those who have not had problems. Thus, the similarity of characteristics of the PKU carriers in our sample to those expected in the general population group also is reassuring, because this indicates that a response bias is not likely to have occurred. We did not confirm reported medical history by reviewing medical records so that inaccurate information may have been obtained. The percentages of galactosemia and PKU carriers who reported that their mothers had had ovarian cancer were higher than found in the Contraceptive and Steroid Hormone (CASH) study (11). Among 2,465 populationbased controls, age 20 to 54 years, 15 (0.8%) reported that their mother had had ovarian cancer. The CASH data were also based on self-report, without medical record review. We believe that the high prevalence (in both groups) reported in our study is most likely due to false-positive reports. The galactosemia carriers may have had a heightened concern about ovarian cancer because of a previous study that found an association between reduced transferase activity and ovarian cancer risk (12), so differential over-reporting should also be considered. We recruited patients through a large number of metabolic disease clinics throughout the United States. The median number of galactosemic patients seen by clinics in our sample was four. These patients may be more representative of the galactosemic population than those seen in the large metabolic disease clinics, which may, by the nature of their program, see a more selected population. Our response rates (40% and 49%, respectively, for galactosemia and PKU carriers) were low but were what we expected given that potential participants were contacted only once about the study. It is likely that the response rate would have increased with additional recruitment efforts, such as follow-up letters and phone calls to nonresponders. However, we felt that this additional burden to the clinics would have resulted in fewer clinics agreeing to participate and thus would have created additional problems with respect to generalizability of the results. Sayle et al. Reproductive Hx of galactosemia carriers 537

5 In summary, our results indicate no evidence of impairment of fertility or increased risk of spontaneous abortion among carriers of the galactosemia gene who have had at least one live birth. The suggestion of an earlier age at natural menopause seen in these data warrants further study. Although a median difference in age at menopause of 1 to 3 years is important on a population level, individual women who carry the galactosemia gene can be reassured that their menopausal experience most likely will fall within what is considered to be the usual age range for menopause. Acknowledgment. We thank Dianne Frazier, Ph.D., M.P.H., R.D., for her work as scientific advisor for this project. REFERENCES 1. Waggoner DD, Buist NRM, Donnell GN. Long-term prognosis in galactosaemia: results of a survey of 350 cases. J Inherit Metab Dis 1990;13: Cramer DW, Harlow BL, Barbieri RL, Ng WG. Galactose-1- phosphate uridyl transferase activity associated with age at menopause and reproductive history. Fertil Steril 1989; 51: Kaufman FR, Devgan S, Donnell GN. Results of a survey of carrier women for the galactosemia gene. Fertil Steril 1993; 60: Weinberg CR, Wilcox AJ, Baird DD. Reduced fecundability in women with prenatal exposure to cigarette smoking. Am J Epidemiol 1989; 129: Colton T. Statistics in medicine. Boston: Little, Brown and Company. 1974: Midgette AS, Baron JA. Cigarette smoking and the risk of natural menopause. Epidemiology 1990;1: Coulam CB, Adamson SC, Annegers JF. Incidence of pre mature ovarian failure. Obstet Gynecol 1986;67: Treloar AE. Menarche, menopause, and intervening fecundability. Hum Bioi 1974;46: Baird DD, Wilcox AJ. Cigarette smoking associated with delayed conception. JAMA 1985;253: Mosher WD, Praff WF. Fecundity, infertility, and reproductive health in the United States, National Center for Health Statistics, Vital and health statistics, series 23, Data from the National Survey of Family Growth, No. 14. DHHS publication no. (PHS) Schildkraut JM, Thompson WD. Familial ovarian cancer: a population-based case-control study. Am J Epidemiol 1988; 128: Cramer DW, Harlow BL, Willett WC, Welch WR, Bell DA, Scully RE, et al. Galactose consumption and metabolism in relation to the risk of ovarian cancer. Lancet 1989; 2: Sayle et at Reproductive Hx of galactosemia carriers Fertility and Sterility