Click on the links below to access all the ArticlePlus for this article.
Please note that ArticlePlus files may launch a viewer application outside of your web browser.
Fertility, infertility, and reproductive characteristics are associated with variations in the incidence of epithelial ovarian carcinoma. 1–4 Nulliparous women are at increased risk, relative to those who have given birth at least once, and many studies have shown that increasing parity is associated with a progressive reduction in risk. 5 Use of oral contraception is strongly protective; weaker protective effects are seen with hysterectomy and tubal sterilization; and somewhat inconsistent relations have been seen with age at first birth, recent pregnancy, incomplete pregnancies, and age at menarche.
Involuntary infertility has been consistently associated with a small increase in the risk of epithelial ovarian carcinoma, specifically among nulliparae. Relative to other nulliparae, women who report a history of involuntary infertility have shown approximately a 1.5-fold increase in risk. 6 Some have suggested that this effect is associated with unexplained or refractory infecundity, or with treatment to induce ovulation, but this subject is controversial. 2 Associations seem to be stronger with borderline tumors and may be explained, at least partly, by diagnostic bias. Studies that have taken into account ethnic and socio-economic differences in women who have undergone treatment to induce ovulation have not found the risk of ovarian cancer to be greatly increased, relative to other patients with infertility. 2,6
It is not clear why parity, infertility, or fertility alter the risk of ovarian carcinoma. There is a growing recognition that their effects are not completely congruent with any of the hypotheses that have been proposed traditionally to explain the etiology of this carcinoma, which are “incessant ovulation,” “excess gonadotrophins,” or “retrograde carcinogens.”5 Nor are they adequately explained by the more recently proposed hypotheses that relate this carcinoma to inflammation, 4 androgens, or progesterone. 3 Relations of ovarian carcinoma to parity, age at first birth, and recency of pregnancy may be more complex than previously recognized, 7,8 and effects of parity may differ between sporadic and familial ovarian carcinoma. 9,10
If the risk of ovarian carcinoma could be modified by childbearing, this knowledge would be important in advising women at risk (eg, those with a family history of ovarian or breast cancers). Therefore, a better understanding is needed of the reasons for effects of parity and/or infertility on this carcinoma. One hypothesis posits that these variables merely reflect a sum of heritable factors (not amenable to modification) associated with ability to conceive and maintain a pregnancy. If this theory were true, then one would expect to observe decreased fecundability (likelihood of conceiving) and fertility (actual births) in the relatives of women with the disease. Our study was designed to test this hypothesis by measuring the fertility of the parents of women with epithelial ovarian carcinoma, compared with the parents of controls.
This case-control study was done in Manhattan, New York City, in the years 1994–1998. Eligible patients had incident, fully invasive epithelial ovarian carcinomas diagnosed within the past year. They were recruited at Memorial Sloan-Kettering Cancer Center (N = 122) or at the adjacent New York Hospital-Cornell University Medical College (N = 46). Borderline tumors were excluded because they are rarely seen at tertiary care centers and because diagnostic bias tends to associate them specifically with infertility. We picked community controls from two sources, intending them to resemble the cases in socio-economic characteristics. We recruited one group from the cases’ own neighborhoods using random digit dialing (N = 81), frequency matching them for age. The other group was recruited from areas that were socio-economically similar to the cases’ neighborhoods, and in the same counties, using a commercial mailing list (N = 78). We have described the methods, response rates and characteristics of these controls in detail elsewhere and as they were similar, we combined them. 11 We also obtained controls at the hospitals, but have excluded these from the present analysis because the methods used to recruit them invalidate their use for estimating effects of numbers of siblings. We excluded any controls reporting a history of bilateral oophorectomy.
To determine the fertility of the subjects’ parents we counted the numbers of siblings. The raw data were derived from a closed, structured, machine-readable family history questionnaire modified from one developed by Memorial Sloan-Kettering Cancer Center’s Department of Genetics. This instrument provided rubrics for up to 5 siblings of each gender, as well as parents, offspring, and limited numbers of more distant relatives. Information on ethnic ancestry was obtained via questions about race, religion, and country of birth of the subject, her parents, and grandparents. We assigned a Jewish ancestry to any woman reporting one or more grandparents whose religion was Jewish (of those with one or more, most reported four). Our findings relevant to ethnic variables are reported in detail elsewhere. 12
We analyzed the data using SAS (SAS Institute Inc, Cary, NC). As we consider the study to be unmatched, we used unconditional logistic regression models, employing the LOGISTIC procedure of SAS. 13 Results are given as odds ratios (ORs, approximating to relative risks) and 95% confidence limits. We adjusted results for variables related to the method of recruitment, which were categories of age (<45, 45–54, and 65+ versus the dummy category of 55–64), area of residence (Manhattan, Connecticut, others), education (up to high school versus some college), and religion of father (Catholic versus others). We also included in the models variables for parity (numbers of births), oral contraceptives (years of use), age at menarche (in years), and Jewish ancestry.
After excluding five cases who did not complete the family history questionnaire, there were 163 cases and 159 controls. Table 1 shows the distributions of selected risk factors and gives their estimated odds ratios. Effects of parity, oral contraceptive use, family history of ovarian/breast cancer, and Jewish ethnicity are similar to those in previous studies. 1 Age at menarche was strongly associated with variations in risk (Table 1) in this study but associations with height and weight were compatible with chance. Effects of Jewish and other ancestries are reported in detail elsewhere. 12 Only 12.5% of cases and 11.9% of controls reported nonwhite ancestors, with 5.3% of cases and 6.2% being African-American. Some 6.8% of cases and 2.5% of controls were Latinas.
The numbers of siblings varied from 0 to 9, with 2.5% of subjects reporting 5 brothers and 1.2% reporting 5 sisters. Parents’ fertility was related to religion, as expected. Catholic women reported somewhat more siblings than Protestants and Jewish women reported considerably fewer; nearly 80% of Jewish controls reported two or fewer siblings, compared with only 49% of non-Jewish controls; and 80% of Jewish cases had two or fewer siblings, compared with 68% of other cases.
Parental fertility was lower in cases than controls (Table 2). Estimated odds ratios of epithelial ovarian carcinoma showed a progressive gradient in risk associated with increasing numbers of siblings. Compared with the reference group of women with 0 or 1 sibling, those with four or more siblings were estimated to have approximately half the risk.
Table 3 treats number of siblings as an ordinal variable and investigates the gender of the siblings. On average, each additional sibling was associated with a 20% reduction in risk, ie, an odds ratio of 0.80. Effects of brothers seemed slightly stronger than effects of sisters, even when each was adjusted to the numbers of siblings of the other gender; however, the difference between the genders was readily compatible with chance variability.
Table 4 examines the effects of siblings within subgroups of other variables. A clearly protective effect associated with increasing numbers of siblings was detected within each subgroup of age and age at menarche, as well as in Jewish and non-Jewish and Catholic and non-Catholic subgroups.
In women with no first-degree family history of breast or ovarian cancer the estimated OR was 0.75 (0.61–0.94); and in those with a family history it was 0.73 (0.37–1.44). Considering only the mothers’ histories, the estimate for women with no family history was 0.81 (0.66–0.98) based on 145 cases and 152 controls. The data were too sparse to allow an estimate of effects on siblings when the mother had breast or ovarian cancer.
Further adjustment for other ethnic groups did not appreciably change these results.
Our results suggest a difference in the family structure of patients with ovarian carcinoma, compared with controls. Our estimated OR of 0.80 per sibling is similar to the effects of parity of women themselves, as previously estimated in population-based studies. Odds ratios/relative risks of 0.78, 0.81, 0.84, and 0.87 per offspring were observed in studies in Toronto, 14 Sweden, 8 the Nurses Health Study 15 andWhittemore et al’ s combinedanalysisof several case-control studies. 5 Our study, though not population-based, shows a similar effect of parity.
Although an inverse relation between high parity and risk of ovarian carcinoma has been observed consistently in many previous studies, there may be paradoxical relations in familial disease. In the Utah pedigree database, Kerber and Slattery found that the familial cases of ovarian cancer were not protected by increasing parity, although the usual relation with parity was seen in women with no family history. 10 In Canada, Narod et al9 studied cases in women carrying mutations in BRCA1, finding an increased penetrance, ie, ovarian cancer incidence, in association with higher parity. In Washington DC, however, a prevalence study of 5318 Jewish people found no rela-tion of BRCA1/2 mutations to family size. 16
The association of lower parity with ovarian carcinoma has supported the hypothesis of a relation between this disease and ovulation. 17 Another hypothesis posits that pregnancy clearstransformed cells (eg, by increasing apoptosis); yet another draws on an analogy with breast cancer, for which pregnancy induces terminal differentiation and reduces the number of breast stem cells susceptible to DNA damage. 18 Another hypothesis, the “fetal antigen hypothesis,” was originally proposed to explain the association of lower parity with breast cancer. 19 This hypothesis suggested that pregnancies might immunize a woman against the fetuses’ foreign antigens (eg, blood group substances and other cell surface molecules), increasing the likelihood of her being able to mount an immune response to inappropriate (non-self) epitopes on mutated cells. The hypothesis has been developed to propose that pregnancies conceived with different partners would be more protective against ovarian carcinoma than those conceived with the same partner; but a recent test of this theory found no support for it. 20 On the other hand, multiparous women have been shown to possess antibodies to onco-fetal antigens expressed by ovarian 21 and endometrial 22 carcinomas, antibodies not shared by nulliparous women, or men.
Yet another hypothesis posits that during their offspring’s childhood, parous women will be exposed to infectious diseases more frequently than nulliparae, leading to a larger immune repertoire and more frequent immune stimulation. This mechanism might also account for our findings on siblings. Similarly, our finding might be consistent with some cases of ovarian carcinoma being related to the later acquisition of some infectious disease, such as mumps 23 at an age when the ovary might be more susceptible to damage; women born into larger families would be exposed to infectious diseases at an earlier age than those with few or no siblings.
There are a number of drawbacks to our study that may limit the validity of our results. One drawback is that we cannot distinguish between parents whose low fertility was by choice, as opposed to those with reduced fecundability (probability of conception) or increased fetal loss. A majority of pregnancies and a large proportion of births are unintended, however, 24 so that parents whose fertility was low (ie, with few actual births) would have tended to be less fecundable than those with larger families. Of course, no parents were infertile.
Another drawback is the study’s small size and its hospital-based design. The cases presenting at our hospitals - tertiary-care cancer centers - are likely to differ, in many respects, from those occurring in the community. Our strategies for selecting community controls were designed to recruit women who were socially and economically similar to the cases and should, theoretically, have reduced the effects of such hospital biases in the cases. Numerous factors influence the ways people choose a hospital, however, and we are unable to exclude the possibility of a systematic bias between cases and controls vis-a-vis numbers of siblings. Another drawback, in common with other studies in New York City, is that response rates for both cases and controls were relatively low, as we have previously reported. 11 Although a low response rate per se will not have invalidated our results, one cannot be sure, theoretically, that some unrecognized difference in the responses of cases and controls is related to family size. Against this, however, our comparisons of cases with community controls gave ORs estimating the effects of other variables, including family history, parity, oral contraceptive use, and Jewish ancestry, that agree closely with those derived from population-based studies. 5,25 Therefore, our community controls, on the face of it, seem well chosen.
We have also considered whether our results might be explained by confounding due to variables that we cannot control. Family size might be merely a marker for differences in socio-economic status (SES) during childhood and might reflect differences in resources (eg, food or educational opportunities) or lifestyle that have influenced exposure to carcinogens. Although in developing countries high fertility is a marker for lower SES, in the United States, during the past 1–2 generations, it has been the families with more money who have chosen to have more children. Furthermore, our data do not suggest a relation between attained height (as a marker of nutrition in childhood and adolescence) and number of siblings, nor between height and ovarian carcinoma.
Jewish ancestry might have been another important confounder, as it is associated with a higher prevalence of mutations in BRCA1 and BRCA2, which increase susceptibility to the disease. 16,26 Since the beginning of this century, demographers have observed Jewish people in Europe and the United States to have voluntarily smaller families than Gentiles. 27 We, too, see that the Jewish women in our study had grown up in significantly smaller families. Our study has controlled for Jewish ancestry, however. Furthermore, we detected effects of numbers of siblings on ovarian cancer risk within both Jewish and non-Jewish groups. Jewish ancestry could be a source of confounding and/or misclassification only if subjects concealed it, or were unaware of it. While we think this possibility is unlikely, we have no way of assessing it.
With these caveats, our data are consistent with the hypothesis that the relation of fertility/infertility to ovarian carcinoma may be mediated partly through heritable factors, rather than be directly causal. Janerich and Thompson studied patterns of reproduction in the mothers of breast cancer patients in the Utah registry. 28 They found more fetal losses and smaller completed family sized in case mothers, compared with controls. Similarly, Gruber and Thompson 29 reported that young women with endometrial carcinoma in the CASH study had fewer sisters than controls. So far as we are aware, however, no previous study has investigated numbers of siblings in ovarian carcinoma.
Nieto et al30 questioned patients being treated for infertility about histories of ovarian cancer in their female relatives. They found an excess of this cancer reported by patients with refractory infertility, compared with patients who were successfully treated. Nieto et al’ s data, though preliminary, are also consistent with the hypothesis that there is a heritable component to the association of ovarian carcinoma with infertility. In future studies this hypothesis would be further supported if the disease were found to be increased in the relatives of infertile males.
In the subjects’ mothers, relevant genes might include those contributing to conditions such as endometriosis or polycystic ovary, both of which carry some familial component and are also associated with ovarian carcinoma. 31,32 The fertility of the subjects’ parents, however, should not be assumed to be attributable only to the mothers, and future researchers should consider investigating patterns in both parents’ families. Similarly, in studies based on couples undergoing treatment for infertility, it would be useful to know the incidence of ovarian cancer in the male partner’s relatives. Many genes known to be important in carcinogenesis have been shown to alter fertility, fetal development or extra-embryonic fetal membranes in transgenic mice, 33 and most can be inherited equally from mothers and fathers. These include many genes involved in control of cell cycle, apoptosis, angiogenesis, DNA repair, and immunity.
Another reason for focusing on fathers as well as mothers is the speculative possibility that our findings might be explained, in part, by de novo germline mutations. The strong association in our data of an effect of numbers of siblings in women with no family history is consistent this mechanism, because new mutations that have occurred in the subjects themselves (ie, at or after their conception) would not normally be associated with any unusual family history of cancer. New mutations in the mothers’ or fathers’ germ cells might be inherited by siblings, but would not be associated with any unusual cancer occurrence in the parents’ generation or earlier ancestors. Consistent with this possibility, others have shown that the risk of ovarian carcinoma between sisters is stronger than the risk associated with a maternal history. 10
The main mechanism for the introduction of new germline mutations is via inheritance from elderly fathers, who acquire mutations in clones of spermatogonia (as in somatic cells) with increasing age; relatively few de novo mutations are transmitted by mothers. 34,35 Studies of new mutations contributing to various birth defects, neurologic and other diseases have shown a strong association with paternal age. 36,37 Recent studies in cancers of brain 38 and prostate 39 have shown paternal age effects; the latter is associated with epithelial ovarian carcinoma in some populations 40,41 but not in others. 42–44 Some women in our study might have had fewer siblings because their fathers were older, either because the fathers married late in life, or because one or other parent, or both parents, were sub-fecund. Our data do not allow us to estimate the age of the parents at the time of the subjects’ birth or to determine the position of each subject within her sibship.
An association of ovarian carcinoma with parental fertility, if confirmed by others, would have a number of implications. Our results draw attention to the possibility that certain families at risk may be unusually small, making it unlikely that there would be an affected female relative. Such families would escape notice by geneticists and would not normally be recruited into genetic epidemiology studies, so that the relevant susceptibility genes would be less easily detected.
Many previous studies might have under estimated the risk associated with a family history of the disease. It is well recognized by geneticists that the probability of a family history being manifest (if the family carries a disease-causing mutation) is conditional on the size of the family. 45 A corollary of this is the problem that attempts to control for family history will inevitably lead to biased estimates of effects of siblings. This problem can be seen in the obviously biased results we obtained within subgroups defined by family history. For this reason we did not adjust for family history. The problem may ameliorated in future studies by restricting the family history to that of the mother. One cannot resolve this problem, completely, however, since a mother’s cancer diagnosis at an early age would have been a cause of curtailment of fertility.
Another implication of our results, if confirmed, is that effect estimates previously made for Jewish versus non-Jewish women 26 might have been over-estimates, as until now they have not taken into account the considerably smaller size of Jewish families in the United States A related problem is that many previous studies, which have not controlled for Jewish ethnicity, may have overestimated the protective effects of higher parity.
We are grateful to the women who participated in the study and also thank C. Aghajanian, C.L. Brown, D.S. Chi, R.R. Barakat, K. Economos, H. Gretz, W.J. Hoskins, W.B. Jones, V. Leary, J.L. Lewis, K. Lloyd, L. McGuinn, M. Melo, K. Offit, E. Poynor, P.J. Sabbatini, D.R. Spriggs, and D. Wuest.
1. Westhoff C. Ovarian cancer. Annu Rev Public Health 1996; 17: 85–96.
2. Riman T, Persson I, Nilsson S. Hormonal aspects of epithelial ovarian cancer: review of epidemiological evidence. Clin Endocrinol (Oxf) 1998; 49: 695–707.
3. Risch HA. Hormonal etiology of epithelial ovarian cancer, with a hypothesis concerning the role of androgens and progesterone. J Natl Cancer Inst 1998; 90: 1774–1786.
4. Ness RB, Cottreau C. Possible role of ovarian epithelial inflammation in ovarian cancer. J Natl Cancer Inst 1999; 91: 1459–1467.
5. Whittemore AS, Harris R, Itnyre J. Characteristics relating to ovarian cancer risk: collaborative analysis of 12 US case-control studies. IV. The pathogenesis of epithelial ovarian cancer. Collaborative Ovarian Cancer Group. Am J Epidemiol 1992; 136: 1212–1220.
6. Mosgaard BJ, Lidegaard O, Kjaer SK, Schou G, Andersen AN. Infertility, fertility drugs, and invasive ovarian cancer: a case-control study. Fertil Steril 1997; 67: 1005–1012.
7. Albrektsen G, Heuch I, Kvåle G. Reproductive factors and incidence of epithelial ovarian cancer: a Norwegian prospective study. Cancer Causes Control 1996; 7: 421–427.
8. Adami HO, Hsieh CC, Lambe M, Trichopoulos D, Leon D, Persson I, Ekbom A, Janson PO. Parity, age at first childbirth, and risk of ovarian cancer. Lancet 1994; 344: 1250–1254.
9. Narod SA, Goldgar D, Cannon-Albright L, Weber B, Moslehi R, Ives E, Lenoir G, Lynch H. Risk modifiers in carriers of BRCA1 mutations. Int J Cancer 1995; 64: 394–398.
10. Kerber RA, Slattery ML. The impact of family history on ovarian cancer risk. The Utah Population Database. Arch Intern Med 1995; 155: 905–912.
11. Olson SH, Mignone L, Harlap S. Selection of controls using a commercial database and random digit dialing. Am J Epidemiol 2000; 152: 585–592.
12. Harlap S, Olson SH, Akhmedkhanov A, Barakat RR, Caputo TA, Sanchez D, Xue X. Epithelial ovarian carcinoma and European birthplace of grandparents. Gynecol Oncol. 2001; 81: 25–32.
13. Breslow NE, Day W. Statistical Methods in Cancer Research. The Analysis of Case-Control Studies. IARC Scientific Pub. No. 32. vol. 1. Lyon: International Agency for Research on Cancer, 1980; 192–246.
14. Risch HA, Marrett LD, Howe GR. Parity, contraception, infertility, and the risk of epithelial ovarian cancer. Am J Epidemiol 1994; 140: 585–597.
15. Hankinson SE, Colditz GA, Hunter DJ, Willett WC, Stamper MJ, Rosner B, Hennekens CH, Speizer FE. A prospective study of reproductive factors and risk of epithelial ovarian cancer. Cancer 1995; 76: 284–290.
16. Struewing JP, Hartge P, Wacholder S, Baker SM, Berlin M, McAdams M, Timmerman MM, Brody LC, Tucker MA. The risk of cancer associated with specific mutations of BRCA1 and BRCA2 among Ashkenazi Jews. N Engl J Med 1997; 336: 1401–1408.
17. Whittemore AS. Characteristics relating to ovarian cancer risk: implications for prevention and detection. Gynecol Oncol 1994; 55: S15–19.
18. Russo J, Russo IH. Cellular basis of breast cancer susceptibility. Oncol Res 1999; 11: 169–178.
19. Janerich DT. The fetal antigen hypothesis for breast cancer, revisited. Med Hypotheses 1994; 43: 105–110.
20. Mockett EJ, Rossing MA, Weiss NS. Fetal antigen hypothesis and ovarian cancer: is there an immunogenic explanation for the reduction in risk associated with parity? Epidemiology 2000; 11: 55–58.
21. Shields LB, Gercel-Taylor C, Yashar CM, Wan TC, Katsanis WA, Spinnato JA, Taylor DD. Induction of immune responses to ovarian tumor antigens by multiparity. J Soc Gynecol Investig 1997; 4: 298–304.
22. Katsanis WA, Shields LB, Spinnato JA, Gercel-Taylor C, Taylor DD. Immune recognition of endometrial tumor antigens induced by multiparity. Gynecol Oncol 1998; 70: 33–39.
23. Cramer DW, Welch WR, Cassells S, Scully RE. Mumps, menarche, menopause, and ovarian cancer. Am J Obstet Gynecol 1983; 147: 1–6.
24. Harlap S, Kost K, Forrest JD. Preventing Pregnancy, Protecting Health. A New Look at Birth Control Choices in the United States. New York: The Alan Guttmacher Institute, 1991.
25. Hulka BS. Epidemiologic analysis of breast and gynecologic cancers. Prog Clin Biol Res 1997; 396: 17–29.
26. Steinberg KK, Pernarelli JM, Marcus M, Khoury MJ, Schildkraut JM, Marchbanks PA. Increased risk for familial ovarian cancer among Jewish women: a population-based case-control study. Genet Epidemiol 1998; 15: 51–59.
27. DellaPergola S. Patterns of American Jewish fertility. Demography 1980; 17: 261–273.
28. Janerich DT, Thompson WD, Mineau GP. Maternal pattern of reproduction and risk of breast cancer in daughters: results from the Utah Population Database [see comments]. J Natl Cancer Inst 1994; 86: 1634–1639.
29. Gruber SB, Thompson WD. A population-based study of endometrial cancer and familial risk in younger women. Cancer and Steroid Hormone Study Group. Cancer Epidemiol Biomarkers Prev 1996; 5: 411–417.
30. Nieto JJ, Rolfe KJ, MacLean AB, Hardiman P. Ovarian cancer and infertility: a genetic link? (letter). Lancet 1999; 354: 649.
31. Brinton LA, Gridley G, Persson I, Baron J, Bergqvist A. Cancer risk after a hospital discharge diagnosis of endometriosis. Am J Obstet Gynecol 1997; 176: 572–579.
32. Franks S, Gharani N, McCarthy M. Genetic abnormalities in polycystic ovary syndrome. Ann Endocrinol (Paris) 1999; 60: 131–133.
33. Rosenberg MP. Gene knockout and transgenic technologies in risk assessment: the next generation. Mol Carcinog 1997; 20: 262–274.
34. Crow JF. The high spontaneous mutation rate: is it a health risk? Proc Natl Acad Sci U S A 1997; 94: 8380–8386.
35. Crow JF. Spontaneous mutation in man. Mutat Res 1999; 437: 5–9.
36. Malaspina D, Harlap S, Fennig S, Heiman D, Mahon D, Gorman JM, Feldman D, Susser ES. Advancing paternal age, new mutations and schizophrenia risk. Arch Gen Psychiat 2001; 58: 361–367.
37. Glaser RL, Jiang W, Boyadjiev SA, Tran AK, Zachary AA, Van Maldergem L, Johnson D, Walsh S, Oldridge M, Wall SA, Wilkie AO, Jabs EW. Paternal origin of FGFR2 mutations in sporadic cases of Crouzon syndrome and Pfeiffer syndrome. Am J Hum Genet 2000; 66: 768–777.
38. Hemminki K, Kyyronen P. Parental age and risk of sporadic and familial cancer in offspring: implications for germ cell mutagenesis. Epidemiology 1999; 10: 747–751.
39. Zhang Y, Kreger BE, Dorgan JF, Cupples LA, Meyers RH, Splansky GL, Schatzkin A, Ellison RC. Parental age at child’s birth and son’s risk of prostate cancer. The Framingham Study. Am J Epidemiol 1999; 150: 1208–1212.
40. Cerhan JR, Parker AS, Putnam SD, Chu BC, Lynch CF, Cohen MB, Torner JC, Cantor KP. Family history and prostate cancer risk in a population-based cohort of Iowa men. Cancer Epidemiol Biomarkers Prev 1999; 8: 53–60.
41. Tulinius H, Egilsson V, Olafsdottir GH, Sigvaldason H. Risk of prostate, ovarian, and endometrial cancer among relatives of women with breast cancer. BMJ 1992; 305: 855–857.
42. Damber L, Gronberg H, Damber JE. Familial prostate cancer and possible associated malignancies: nation-wide register cohort study in Sweden. Int J Cancer 1998; 78: 293–297.
43. Vazina A, Baniel J, Yaacobi Y, Striker A, Engelstein D, Leibovitch I, Zehavi M, Sidi AA, Ramon Y, Tischler T, Livne PM, Friedman E. The rate of the founder Jewish mutations in BRCA1 and BRCA2 in prostate cancer patients in Israel. Br J Cancer 2000; 83: 463–466.
44. Nastiuk KL, Mansukhani M, Terry MB, Kularatne P, Rubin MA, Gammon MD, Ittman M, Krolewski JJ. Common mutations in BRCA1 and BRCA2 do not contribute to early prostate cancer in Jewish men. Prostate 1999; 40: 172–177.
45. Thompson EA, Kravitz K, Hill J, Skolnick MH. Linkage and the power of a pedigree structure. In: Morton NE, Chung CC, eds. Genetic Epidemiology. New York:Academic Press
, 1978: 247–253.