ARTICLE OUTLINE
	Abstract
	Introduction
	Materials and Methods
	Results
	Discussion
	Acknowledgement
	References
	Tables

Objective:  Family history of cancer is recognized as one of the most important risk factors in predicting personal cancer risk.  Nevertheless,
there are few published population-based estimates of family history prevalence by age categories.

Methods:  We used responses of female controls (n=4754) from the population-based Cancer and Steroid Hormone study (1980-1982) to
estimate the frequency of family history of various cancers among female relatives.  We determined the age- and race-specific prevalence of family
history of breast, ovarian, endometrial, and other cancers in first-degree female relatives of women aged 20-54 years.  To evaluate changes in reporting of family history over time, we also analyzed responses of control women (n=1544) from the Women's Interview Study on Health (WISH) (1990-1992) to estimate prevalence of family history of breast cancer.

Results:  The prevalence of a first-degree family history of breast, ovarian, endometrial, and cervical cancers was 6.4% (95% CI 5.7-7.1%), 1.1% (0.8-1.4%), 3.5% (3.0-4.0%), and 2.1% (1.7-2.5%), respectively.  Among first-degree female relatives, the prevalence of family history of colon, lung, and thyroid cancers was 2.4% (2.1-2.9%), 1.5 % (1.2-1.8%), and 0.5% (0.3-0.7%), respectively.  Prevalence of family history of breast and colon cancers increased significantly with respondent's age.  Similar results for family history of breast cancer were obtained from an analysis of responses from the WISH.

Conclusions:  In addition to providing a point of reference for research and health policy, these results may be useful to providers who care for female patients because of the usefulness of information about family history of cancer for assessing lifetime risk of cancer. needed.

Key Words: cancer risk, family history, prevalence

A positive family history of cancer, often used as a proxy for an unmeasured, common genetic background within a family, is recognized as one of the most important risk factors in predicting personal cancer risk.  The relationship of first-degree family history to cancer risk has been determined for numerous cancers [1-28] and has been used to predict the risk of cancer at other sites [7, 27, 29-35].  A family history score derived from the genders, ages, and numbers of relatives has also been used to predict cancer mortality [36].  Overall, cancer in a first-degree relative increases one's personal risk of cancer by 2- to 4-fold, depending upon type of cancer, compared to persons who report no family history of cancer.  The number of affected relatives [9, 14, 19, 20, 37, 38] and having relatives with an early age of cancer onset [5, 7, 8, 18-20, 32, 33, 39] further increase this risk.  The magnitude of the risk estimate is less when only second- or third-degree relatives are affected [5, 12, 15].

Family history is of value in identifying individuals at increased risk of cancer [40] and evaluating lifetime cancer risk.  Estimates of the frequency of a family history of cancer in a population can be useful to physicians and health planners in assessing the expected caseload of patients at higher risk for cancer and the demand for increased screening and follow-up [41] such as referral for genetic counseling.  More recently, a number of reports suggest that associations with modifiable reproductive and dietary risk factors as well as underlying genetic pathways for some cancers may vary by family history status [39, 42, 43].  Thus, evaluation of family history status is important to patient management, medical care program planning and may have implications for response to treatment.

Despite the utility of information about family history of cancer to primary care providers, there is a dearth of published estimates of the population-based prevalence of family history of cancer by age categories.  Most studies to date have not been population-based, focused on one cancer type, or did not stratify results by age or race.  One study from the United Kingdom determined the prevalence of family history of cancer within the caseload of a single physician [44].  Two prior reports from the Cancer and Steroid Hormone (CASH) study have reported on the prevalence of family history of breast cancer among women [5, 38].  Schwartz et al [38] examined the prevalence of breast cancer among female relatives of women from Detroit.  Sattin et al [5] recounted the number of study participants who reported a positive family history of breast cancer by case or control status of the respondent.  In addition, Schildkraut and Thompson (1988) [15] reported the number of CASH controls who indicate having a relative with ovarian cancer.  To our knowledge, no population-based studies have examined the prevalence of family history of various cancers.  To help fill the gap in our knowledge, we used available population-based data to determine the age- and race-specific prevalence of family history of breast, ovarian, and endometrial cancer in first- and second-degree female relatives of women aged 20-54 years.  Age- and race-specific prevalence of first-degree family history only was examined for cervical, colorectal, lung, and thyroid cancer.  A further goal was to evaluate any changes in reporting of family history over time.

Study participants, study design, and data collection

We analyzed data from two population-based case-control studies that focused primarily on oral contraceptive use and reproductive cancers in younger women, The Cancer and Steroid Hormone (CASH) study and the Women's Interview Study on Health (WISH).  Cases who participated in the CASH study, a large, multi-center study [45] were women aged 20 to 54 years who were newly diagnosed with breast, ovarian, or endometrial cancer between December 1, 1980 and December 31, 1982, and resided in any of eight reporting areas of the Surveillance, Epidemiology, and End Results (SEER) program of the National Cancer Institute: the metropolitan areas of Atlanta, Detroit, San Francisco, and Seattle; the states of Connecticut, Iowa, and New Mexico; and the four urban counties of Utah.  A 88% response rate to the telephone screener was obtained for the 61337 telephone numbers assessed as residential. Controls, selected by random digit dialing, were matched to breast cancer cases by 5-year age group and SEER area.  Only data from controls were used in the present analysis.  In-person interviews that focused on reproductive, contraceptive, and medical histories, including family history of cancer were completed within 6 months of selection by 4,754 (83%) of 5,698 eligible controls.  For these analyses, the CASH dataset consisted of 3,965 (83.4%) white, 487 (10.2%) black, and 302 (6.4%) women of various other racial backgrounds.

The collection of family history data was previously reported [5].  Briefly, respondents were asked for the vital status and current age (or age at death) of each natural mother, sister, daughter, or grandmother and for the total number of natural aunts.  Each participant was shown a list of site-specific cancers and asked if each relative had been diagnosed with any cancer on the list.  Age at cancer onset was collected for each female relative identified as having breast, ovarian, or endometrial cancer.  For cervical, colorectal, lung, and thyroid cancer, only information about mothers, sisters, daughters, or grandmothers was collected, without information about age of onset.  Any cancers, other than the aforementioned, reported in relatives were categorized as “other”.

Data from WISH, the second study, were obtained from participants recruited in Atlanta, Seattle/Puget Sound, and central New Jersey [46].  The study was confined to women aged 20 to 44 years in Seattle and New Jersey and to women aged 20 to 54 years in Atlanta.  Cases were women newly diagnosed with breast cancer between May 1, 1990 and December 31, 1992.  A 90.5% response rate to the telephone screener was obtained for the 16254 telephone numbers assessed as residential. Controls selected for study inclusion were identified by random digit dialing and were frequency-matched to cases by 5-year age group and geographic area of residence.  In-person interviews that collected information on reproductive, medical, and contraceptive histories (including family history of breast cancer among first-degree female relatives) were completed by 2,009 (78.1%) of 2,571 eligible controls.  For this analysis, we restricted participants to control women from Atlanta and Seattle, because New Jersey had not been included in the CASH study.  There was a 70.7% response rate among controls.  The final WISH dataset consisted of 1,186 (76.8%) white, 278 (18.0%) black, and 80 (5.2%) women of various other racial backgrounds.  In the comparison of CASH and WISH, the CASH study was also restricted to women living in Atlanta and Seattle.

Statistical analysis

We defined family history of cancer as “first-degree” if the respondent reported a mother, sister, or daughter with cancer; “second-degree” if cancer was reported in a grandmother or aunt; or “no family history” if no first- or second-degree female relatives had cancer.  If the status of first-degree female relatives was unknown,  we classified family history as or “unknown”, regardless of the status of second-degree relatives.  For respondents with more than one female relative with cancer, we classified family history as first-degree if at least one first-degree relative had cancer and second-degree if only second-degree relatives had cancer.

The overall frequency of reported family history was determined for each cancer.  We then stratified on 10-year age groups and race to generate age- and race-specific prevalence estimates.  CASH study data stratified by race were age-adjusted to the 1990 SEER standard mortality population using the direct adjustment method to better estimate current age distributions.  Confidence intervals for proportions were derived using the equation p + 1.96√standard error .  All analyses were conducted using SAS (SAS version 6.12, Cary, NC).  Tests for linear trend in proportions across age were determined using the c2 statistic in EpiInfo (EpiInfo version 6.04b, Atlanta, GA) and differences in proportions by race strata were evaluated using the χ2 statistic of the Mantel-Haenszel extension test.

Since CASH and WISH were conducted at different points in time, we evaluated changes in the reported overall, age-specific, and race-specific prevalence of family history of cancer among first-degree relatives.  In this comparison, we restricted data from the two studies to the Atlanta and Seattle SEER areas.  For the race-specific comparisons, age-adjustment was made to the 1990 SEER standard population using the direct method.  In addition, we compared across studies the strata-specific odds ratios (ORs) derived from separate logistic regression models fitted for each study to adjust for variation in distribution of age and race.  Finally, a model that combined data from both studies was used to determine whether the reporting of family history across studies differed substantially.

In analysis of age-specific prevalence of reported family history of various cancers by age of respondents, 268 women aged 20 to 29 years were excluded as there were too few who reported a positive family history to generate stable estimates.  Except for thyroid cancer, the prevalence of first-degree family history increased with age for each cancer, although some increases were slight and only two were statistically significant:  breast cancer (p<0.001) and colorectal cancer (p<0.01)(Table 2).  In contrast, prevalence of reported second-degree family history of breast, ovarian, and endometrial cancer decreased across the age groups, although none of the trends was significant.  The percentage of persons with an unknown family history increased significantly across the age groups for each cancer examined (all p<0.05).

In analysis of family history of cancer by race, only the estimates for breast and endometrial cancer resulted in stable estimates for nonwhites.  Significant differences were observed in prevalence of family history of breast cancer across racial groups:  compared with white women, both black women and women of other races were less likely to report second-degree family history and more likely to report no family history (all p<0.001)
(Table 3).  In addition, compared with white women, women of other races were less likely to report first-degree family history (p<0.05).  We observed no significant differences in prevalence of family history of endometrial cancer across racial categories.

In a comparison of the prevalence of reported family history of breast cancer from the CASH study and WISH, restricted to controls from the Atlanta and Seattle SEER sites, we observed no substantial difference in the overall prevalence (Table 4).  The only significant change in age- or race-specific strata across studies was that the prevalence was significantly increased among 20 to 29 year-olds interviewed in 1990-1992 in WISH compared with women of the same age interviewed in the 1980-82 CASH study (p<0.05).  No other statistically significant differences were noted across age strata between the two studies.  In comparing prevalence of first-degree family history of breast cancer across studies by race, no statistically significant differences were observed after age-adjustment (data not shown).

We ran logistic regression models to simultaneously adjust for age and race within each study. Because none of the control women 20-29 years of age in the CASH study reported a first-degree family history of breast cancer, we excluded them from the CASH model.  We used the odds ratio as an estimate of the relative risk, since the outcome of interest was an uncommon event in the study population.  These data show that in the 1980-1982 CASH study, women aged 40-54 years were more than twice as likely to report a first-degree family history of breast cancer than women aged 30-39 years, and that black women were significantly less likely than white women to report this (Table 4).  Analysis of the 1990 WISH data revealed no significant differences by age or race.  Even so, women aged 20-29 years were more likely than women of all other ages to report a positive family history of breast, and women of other racial backgrounds were the group least likely to do so.  As with the comparison of prevalences across studies previously reported, comparison of strata-specific odds ratios indicated no significant differences in prevalence of reporting of family history over time.  Finally, results of the model using combined data from both studies indicated no significant difference in reporting of first-degree family history of breast cancer overall.

For most cancers, risk increases with age [55].  In addition, as individuals age so do their relatives, making it more likely that the relatives of older respondents will have developed cancer.  Correspondingly, for each cancer examined except thyroid, our results showed that prevalence of first-degree family history of cancer increased with increasing respondent age.  Several studies have found that self-reported family history of cancer other than breast often results in an underestimation of cancer in relatives, especially second-degree [51-53].  We found that where second-degree family history could be evaluated, prevalence generally decreased with increasing age.  Also, the prevalence of unknown first-degree family history of cancer increased with age;  the magnitude of increase was similar across age strata for each cancer examined.  These findings may reflect problems with recall among older women in the study or may be due to a cohort effect in which the older women were less knowledgeable about cancer in their grandmothers and aunts or less aware of cancer in their families.

Our race-specific estimates of the prevalence of family history of cancer were limited by the small number of nonwhite women who reported a family history of cancer other than breast and endometrial.  White and black women reported a statistically similar prevalence of first-degree family history of breast cancer.  Among women of other backgrounds, we noted a high prevalence of unknown family history of breast cancer relative to the prevalence of first-degree family history; a similar pattern was seen for endometrial cancer.

The prevalence of family history of lung cancer among first-degree female relatives of CASH study controls was 1.5%.  This number seems low because lung cancer is the second most common cancer among U.S. women today [54].  Our analysis includes many women whose mothers may have been born more than 75 years ago.  If we assume an average maternal age of 22 years, then for the 1980-1982 CASH study respondents aged 40-54, their mothers would belong to the approximate birth cohort of 1904-1918.  SEER data show that the lung cancer incidence rates among women 65 years or older was three times higher in 1992 (birth cohort up to 1927) than in 1973 (birth cohort up to 1908) [55].  Thus, we expect there to have been even lower rates in the 1904 birth cohort.  As a result, we expected that few CASH study respondents would report a first-degree family history of lung cancer.

Conversely, the prevalence of family history of endometrial cancer appears high.  We cannot rule out that cervical cancer or hysterectomy was inappropriately reported as endometrial cancer, inflating the prevalence estimate.

A limitation of using the CASH study data for this analysis is its restriction to younger women.  The study was designed primarily to evaluate effects of oral contraceptive use on reproductive cancers and thus, focused on women younger than 55 years of age who had the opportunity for oral contraceptive exposure during their reproductive years.  Thus, our analysis does not include information about the prevalence of family history of cancer among women at older ages.  A further limitation is that cancer among relatives was not verified.  Although some studies have found self-reported family history of breast cancer to be relatively accurate [52, 56-59], reliance on self-reported data can result in misclassification of family history for other cancers [52, 57, 60] and underestimation of cancer in second-degree relatives [51-53].  It is not feasible to validate the reported family history of a population-based, cross-sectional study.  We were further limited by the lack of information about family history among male relatives.  It is especially pertinent to extend studies of family history to male relatives because a family history of prostate cancer may increase the risk of breast cancer [29, 33, 34] and a family history of colorectal cancer increases the risk of endometrial and ovarian cancers [27, 35].  In addition, male breast cancer is a hallmark of germline BRCA2 mutations in a family [61].  Therefore, evaluation of family history of cancer among male relatives will likely be important to the determination of a more accurate risk estimate for female patients and should be examined in future prevalence estimates.

Our analysis is also limited by the use of 20-year-old data.  Data collection occurred when awareness of breast and ovarian cancers was not as great as it is now, and before efforts to screen for breast cancer were begun.  In view of this limitation, we evaluated the prevalence of reported first-degree family history of breast cancer in a subgroup of the more recent WISH and found no significant differences in overall, age-, or race-specific prevalence of family history of breast cancer over the intervening 10-year span among controls sampled from Atlanta and Seattle, with the exception that the prevalence of family history of breast cancer had increased substantially over time among the youngest women.

Taken overall, the estimates of prevalence presented here are likely to be conservative compared with actual current prevalence because of the age of the data and their restriction to younger women and to female relatives. In addition, in the past decade, great strides have been made in screening for and detecting breast cancer.  This recent gain may have impacted current public awareness and resulted in an increased reporting of family history for some cancers.  More recent estimates of family history of cancer, based on national samples that include larger numbers of nonwhite men and women are needed.  Plans are under way to address some of these issues in the 2000 Cancer Control Supplement of the National Health Interview Study.

This project was supported under a cooperative agreement from the Centers for Disease Control and Prevention through the Association of Teachers of Preventive Medicine.  We are grateful to Dr. Ralph Coates and Dr. Kathleen E. Malone for kindly providing access to the WISH data from Atlanta and Seattle and for discussions on how the data were collected and coded.  We also acknowledge Bob Uhler for technical assistance.

1. Ghadirian P, Lacroix A, Perret C, Maisonneuve P, Boyle P.  Sociodemographic characteristics, smoking, medical and family history, and breast cancer.  Cancer Detect Prev 1998;22:485-494.
2. Ghadirian P, Maisonneuve P, Perret C, Lacroix A, Boyle P.  Epidemiology of sociodemographic characteristics, lifestyle, medical history, and colon cancer: a case-control study among French Canadians in Montreal.  Cancer Detect Prev 1998;22:396-404.
3. Fioretti F, Tavani A, Bosetti C, La Vecchia C, Negri E, Barbone F, Talamini R, Franceschi S.  Risk factors for breast cancer in nulliparous women.  Br J Cancer 1999;79:1923-1928.
4. Magnusson C, Colditz G, Rosner B, Bergstrom R, Persson I.  Association of family history and other risk factors with breast cancer risk (Sweden).  Cancer Causes Control 1998;9:259-267.
5. Sattin RW, Rubin GL, Webster LA, Huezo CM, Wingo PA, Ory HW, Laude PM.  Family history and the risk of breast cancer.  JAMA 1995;253:1908-1913.
6. Lynch HT, Watson P, Conway T, Fitzsimmons ML, Lynch J.  Breast cancer family history as a risk factor for early onset breast cancer.  Breast Canc Res Treat 1988;11:263-267.
7. Negri E, Braga C, La Vecchia C, Franceschi S, Parazzini F.  Family history of cancer and risk of breast cancer.  Int J Canc 1997;72:735-738.
8. Bhatia S, Pratt CB, Sharp GB, Robison LL.  Family history of cancer in children and young adults with colorectal cancer.  Med Ped Oncol 1999;33:470-475.
9. Newcomb PA, Taylor JO, Trentham-Dietz A.  Interactions of familial and hormonal risk factors for large bowel cancer in women.  Int J Epidemiol 1999;28:603-608.
10. Kerber RA, Slattery ML, Potter JD, Caan BJ, Edwards SL.  Risk of colon cancer associated with a family history of cancer or colorectal polyps:  the diet, activity, and reproduction in colon cancer study.  Int J Cancer 1998;78:157-160.
11. Le Marchand L, Zhao LP, Quiaoit F, Wilkens LR, Kolonel LN.  Family history and risk of colorectal cancer in the multiethnic population of Hawaii.  Am J Epidemiol 1996;144:1122-1128.
12. Mink PJ, Folsom AR, Sellers TA, Kushi LH.  Physical activity, waist-to-hip ratio, and other risk factors for ovarian cancer: a follow-up study of older women.  Epidemiol 1996;7:38-45.
13. Kerber RA, Slattery ML.  The impact of family history on ovarian cancer risk.  The Utah Population Database.  Arch Intern Med 1995;155:905-912.
14. Stratton JF, Pharoah P, Smith SK, Easton D, Ponder BA.  A systematic review and meta-analysis of family history and risk of ovarian cancer.  Br J Obstet Gynaecol 1998;105:493-499.
15. Schildkraut JM, Thompson WD.  Familial ovarian cancer:  a population-based case-control study.  Am J Epidemiol 1988;128:456-466.
16. Glover FE Jr, Coffey DS, Douglas L, Russell H, Cadigan M, Tulloch T, Wedderburn K, Wan RL, Baker TD, Walsh PC.  Familial study of prostate cancer in Jamaica.  Urology 1998;52:441-443.
17. Cerhan JR, Parker AS, Putnam SD, Chiu 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.
18. Bratt O, Kristoffersson U, Lundgren R, Olsson H.  Familial and hereditary prostate cancer in southern Sweden.  A population-based case-control study.  Eur J Cancer 1999;35:272-277.
19. Rodriguez C, Calle EE, Miracle-McMahill HL, Tatham LM, Wingo PA, Thun MJ, Heath CW Jr.  Family history and risk of fatal prostate cancer.  Epidemiol 1997;8:653-657.
20. Lesko SM, Rosenberg L, Shapiro S.  Family history and prostate cancer risk.  Am J Epidemiol 1996;144:1041-1047.
21. Schuurman AG, Zeegers MP, Goldbohm, RA, van den Brandt PA.  A case-cohort study on prostate cancer risk in relation to family history of prostate cancer.  Epidemiol 1999;10:192-195.
22. Whittemore AS, Wu AH, Kolonel LN, John EM, Gallagher RP, Howe GR, West DW, Teh CZ, Stamey T.  Family history and prostate cancer risk in black, white, and Asian men in the United States and Canada.  Am J Epidemiol 1995;141:732-740.
23. Whiteman DC, Valery P, McWhirter W, Green AC.  Risk factors for childhood melanoma in Queensland, Australia.  Int J Caner 1997;70:26-31.
24. Cutler C, Foulkes WD, Brunet JS, Flanders TY, Shibata H, Narod SA.  Cutaneous malignant melanoma in women is uncommonly associated with a family history of melanoma in first-degree relatives:  a case-control study.  Melanoma Res 1996;6:435-440.
25. Burden AD, Newell J, Andrew N, Kavanagh G, Connor JM, MacKie RM.  Genetic and environmental influences in the development of multiple primary melanoma.  Arch Dermatol 1999;135:261-265.
26. Masback A, Westerdahl J, Ingvar C, Olsson H, Jonsson N.  Clinical and histopathological characteristics in relation to aetiological risk factors in cutaneous melanoma:  a population-based study.  Melanoma Res 1999;9:189-197.
27. Gruber SB, Thompson WD.  A population-based study of endometrial cancer and familial risk in younger women.  The Cancer and Steroid Hormone Study Group.  Canc Epid Biomark Prev 1996;5:411-417.
28. Brown LM, Linet MS, Greenberg RS, Silverman DT, Hayes RB, Swanson GM, Schwartz AG, Schoenberg JB, Pottern LM, Fraumeni JF Jr.  Multiple myeloma and family history of cancer among blacks and whites in the U.S.  Cancer 1999;85:2385-2390.
29. McCahy PJ, Harris CA, Neal DE.  Breast and prostate cancer in the relatives of men with prostate cancer.  Br J Urology 1996;78:552-556.
30. Claus EB, Risch N, Thompson WD.  The calculation of breast cancer risk for women with a first degree family history of ovarian cancer.  Breast Canc Res Treat 1993;28:115-120.
31. Thompson WD, Schildkraut JM.  Family history of gynaecological cancers: relationships to the incidence of breast cancer prior to age 55.  Int J Epidemiol 1991;20:595-602.
32. Rodriguez C, Calle EE, Tatham L, Wingo PA, Miracle-McMahill HL, Thun MJ, Heath CW Jr.  Family history of breast cancer as a predictor for fatal prostate cancer.  Epidemiol 1998;9:525-529.
33. Sellers TA, Potter JD, Rich SS, Drinkard CR, Bostick RM, Kushi LH, Zheng W, Folsom AR.  Familial clustering of breast and prostate cancers and risk of postmenopausal breast cancer.  J Natl Cancer Inst 1994;86:1860-1865.
34. Anderson DE, Badzioch MD.  Familial breast cancer risks.  Effects of prostate and other cancers.  Cancer 1993;72:114-119.
35. Schildkraut JM, Thompson WD.  Relationship of epithelial ovarian cancer to other malignancies within families.  Genetic Epidemiol 1998;5:355-367.
36. Yang Q, Khoury MJ, Rodriguez C, Calle EE, Tatham LM, Flanders WD.  Family history score as a predictor of breast cancer mortality: prospective data from the Cancer Prevention Study II, United States, 1982-1991.  Am J Epidemiol 1998;147:652-659.
37. Claus EB, Risch N, Thompson WD.  Age at onset as an indicator of familial risk of breast cancer.  Am J Epidemiol 1990;131:244-253.
38. Schwartz AG, King MC, Belle SH, Satariano WA, Swanson GM.  Risk of breast cancer to relatives of young breast cancer patients.  J Natl Cancer Inst 1985;75:665-668.
39. Egan KM, Stampfer MJ, Rosner BA, Trichopoulos D, Newcomb PA, Trentham-Dietz A, Longnecker MP, Mittendorf R, Greenberg ER, Willett WC.  Risk factors for breast cancer in women with a breast cancer family history.  Cancer Epidemiol Biomark Prev 1998;7:359-364.
40. Weber W.  Cancer control by family history.  Anticancer Res 1993;13:1197-1201.
41. Brzezinski W, Orrom WJ, Wiens E.  Prospective and retrospective analysis of colonoscopy findings in patients with a history of colorectal carcinoma in first-degree relatives.  Canadian J Surg 1990;33:314-316.
42. Sellers TA, Bazyk AE, Bostick RM, Kushi LH, Olson JE, Anderson KE, Lazovich D, Folsom AR.  Diet and risk of colon cancer in a large prospective study of older women:  an analysis stratified on family history.  Cancer Causes Control 1998;9:357-367.
43. Keller G, Rudelius M, Vogelsang H, Grimm V, Wilhelm MG, Mueller J, Siewert JR, Hofler J.  Microsatellite instability and loss of heterozygosity in gastric carcinoma in comparison to family history.  Am J Pathol 1998;152:1281-1289.
44. Johnson N, Lancaster T, Fuller A, Hodgson SV.  The prevalence of a family history of cancer in general practice.  Family Practice 1995;12:287-289.
45. Wingo PA, Ory HW, Layde PM, Lee NC.  The evaluation of the data collection process for a multicenter, population-based, case-control design.  Am J Epidemiol 1998;128:206-217.
46. Brinton LA, Daling JR, Liff JM, Schoenberg JB, Malone KE, Stanford JL, Coates RJ, Gammon MD, Hanson L, Hoover RN.  Oral contraceptives and breast cancer risk among younger women.  J Natl Cancer Inst 1995;87:827-835.
47. Lipkus IM, Iden D, Terrenoire J, Feaganes JR.  Relationships among breast cancer concern, risk perceptions, and interest in genetic testing for breast cancer susceptibility among African-American women with and without a family history of breast cancer.  Cancer Epidemiol Biomark Prev 1999;8:533-539.
48. Lerman C, Lustbader E, Rimer B, Daly M, Miller S, Sands C, Balshem A.  Effects of individualized breast cancer risk counseling:  a randomized trial.  J Natl Cancer Inst 1995;87:286-292.
49. Audrain J, Lerman C, Rimer B, Cella D, Steffens R, Gomez-Caminero A.  Awareness of heightened breast cancer risk among first-degree relatives of recently diagnosed breast cancer patients.  The High Risk Breast Cancer Consortium.  Cancer Epidemiol Biomark Prev 1995;4:561-565.
50. Inglehart JD, Miron A, Rimer BK, Winer EP, Berry D, and Schildkraut JM.  Overestimation of hereditary breast cancer risk.  Ann Surg 1998;228:375-384.
51. Claus EB, Risch N, Thompson WD.  Autosomal dominant inheritance of early-onset breast cancer.  Cancer 1994;73:643-651.
52. Douglas FS, O'Dair LC, Robinson M, Evans DG, Lynch SA.  The accuracy of diagnoses as reported in families with cancer:  a retrospective study.  J Med Genetics 1999;36:309-312.
53. Glanz K, Grove J, LeMarchand L, Gotay C.  Underreporting of family history of colon cancer:  correlates and implications.  Cancer Epid Biomark Prev 1999;8:635-639.
54. American Cancer Society, Cancer Facts and Figures-1999.  Atlanta, GA.
55. SEER cancer statistics review, 1973-1996.  1999.  Bethesda, MD:  National Cancer Institute, (NIH publication no. 97-2789).
56. Parent ME, Ghadirian P, Lacroix A, Perret C.  The reliability of recollections of family history:  implications for the medical provider.  J Canc Educ 1997;12:114-120.
57. Kerber RA, Slattery ML.  Comparison of self-reported and database-linked family history of cancer data in a case-control study.  Am J Epidemiol 1997;146:244-248.
58. Anton-Culver H, Kurosaki T, Taylor TH, Gildea M, Brunner D, Bringman D.  Validation of family history of breast cancer and identification of the BRCA1 and other syndromes using a population-based cancer registry.  Genetic Epidemiol 1996;13:193-203.
59.   Bergmann MM, Calle EE, Mervis CA, Miracle-McMahill HL, Thun MJ, Heath CW.  Validity of self-reported cancers in a prospective cohort study in comparisonwith data from state cancer registries.  Am J Epidemiol 1998;147:556-562.
60. Weinstock MA, Brodsky GL.  Bias in the assessment of family history of melanoma and its association with dysplastic nevi in an case-control study.  J Clin Epidemiol 1998;51:1299-1303.
61. Wooster R, Bignell G, Lancaster J, Swift S, Seal S, Mangion J, Collins N, Gregory S, Gumbs C, Micklem G.  Identification of the breast cancer susceptibility gene BRCA2.  Nature 1995;378:789-792.