We conducted a case-control analysis, a family-based population analysis, and a meta-analysis to assess the role of family history of cancer and kidney cancer in association with the risk of renal cell carcinoma (RCC). A total of 325 cases and 329 controls were identified from an ongoing case-control study of RCC. Study variables were assessed through 45-minute structured face-to-face interviews. In the case-control analysis, a family history of any cancer (in first-degree relatives) was associated with a nonsignificant 1.2-fold increase in RCC risk [95% confidence interval (95% CI), 0.8-1.6]. The risk increased to 1.7 and became significant when the relative was a sibling (95% CI, 1.1-2.5). A family history of kidney cancer (kidney cancer in first-degree relatives) was associated with a 4.3-fold significantly increased risk of RCC (95% CI, 1.6-11.9). The cases reported a total of 2,536 first-degree relatives of which 21 (0.8%) had kidney cancer, and the controls reported a total of 2,333 first-degree relatives of which 5 (0.2%) had kidney cancer (P = 0.003). In the family-based population analysis, a family history of kidney cancer was associated with a 2.8-fold increased risk of RCC (95% CI, 1.0-7.8). The meta-analysis further confirmed this significant association with a 2.2-fold increased risk of RCC (95% CI, 1.6-2.9). To our knowledge, this is the first study to use three analytic strategies to investigate the association between a family history of kidney cancer and risk of RCC, and the first systematic evaluation of the relative risk for developing RCC associated with family history.(Cancer Epidemiol Biomarkers Prev 2008;18(3):801–7)

Malignant tumors of the kidney account for ∼4% of all new primary cancer cases diagnosed in the United States, with an estimated 54,390 cases occurring in 2008 (1). Renal cell carcinoma (RCC) accounts for 85% of all renal cancers (2). Numerous epidemiologic studies have identified cigarette smoking, obesity, and hypertension as the main risk factors for RCC, potentially accounting for 50% of all U.S. cases (3, 4).

Family history of kidney cancer was first associated with the risk of developing RCC through a series of early case reports (5, 6). Research since has reported associations ranging from none to a 5-fold increase in risk (7-15). Results from a relatively large Icelandic study indicated that almost 60% of RCC cases had a first-degree or second-degree relative with kidney cancer (9), suggesting a strong genetic component to risk. Among previous population-based case-control studies, a Canadian study including 518 cases and 1,381 controls did not identify an association between family history and RCC (11) whereas others reported significant increases in risk (7, 12, 15). Among the hospital-based case-control studies, one observed a significantly increased risk of RCC (particularly when the affected relative was a sibling; ref. 13) whereas two further studies reported a nonsignificant increased risk for subjects with at least one first-degree relative with kidney cancer (10, 14).

To date, however, a systematic evaluation of the relative risk (RR) for developing RCC associated with family history has not been reported. In this study, we aim to address the inconsistent results reported in previous studies, thereby improving the estimate of the familial component of RCC risk. In the first instance, a case-control approach was applied to compare self-reported family history of kidney cancer in first-degree relatives of RCC cases and their matched controls. Next, a family-based population analysis was applied to investigate the familial risk of RCC after controlling for confounding effects among the relatives. Lastly, to address the heterogeneity of previous studies, we conducted a meta-analysis (including our own data) to provide an overall estimate of effect.

Study Population

Incident RCC cases were recruited from The University of Texas M. D. Anderson Cancer Center in Houston, Texas. M. D. Anderson Cancer Center staff interviewers identified RCC cases through a daily review of computerized appointment schedules for the Departments of Urology and Genitourinary Medical Oncology. All cases were individuals with newly diagnosed histologically confirmed RCC. There were no age, gender, ethnicity, or cancer-stage restrictions on recruitment. However, to be eligible, the cases must have been residents of Texas. Healthy control subjects without a history of cancer, except nonmelanoma skin cancer, were identified and recruited with the use of the random digit dialing (RDD) method. In RDD, randomly selected phone numbers from households were used to contact potential control volunteers in the same residency of cases according to the telephone directory listings. To be eligible, controls must have lived for at least 1 y in the same county or socio-economically matched surrounding counties that the case resided in. The controls were frequency matched to the cases by age (±5 y), sex, ethnicity, and county of residence. We examined the comparability of case and control characteristics within and outside of the immediate geographic areas of the study center, and no significant differences were observed (data not shown), suggesting that case and control characteristics within and outside of the immediate geographic areas of M. D. Anderson are comparable.

The overall response rate for RDD screening was 51% and, among those who agreed to participate, the response rate was 88%. The response rate for the eligible cases was 87%. This RCC case-control study started subject recruitment in 2002 and is currently ongoing.

Data Collection

After written informed consent was obtained, the study participants completed a 45-min structured in-person interview conducted by trained M. D. Anderson staff interviewers. Data was collected on demographic characteristics, tobacco-use history, alcohol consumption, and family history of cancer. Family history data included cancer history for all first-degree relatives (biological parents, siblings, and offspring). Specifically, the information collected included whether the relative ever had cancer (yes or no), the type of cancer (site), age at diagnosis, current age or age of death, vital status (dead or alive), smoking status (yes or no), years smoked, number of cigarettes smoked, and whether the relative ever had high blood pressure (yes or no). An individual who had never smoked at least 100 cigarettes in his or her lifetime was defined as a never smoker. An individual who had smoked at least 100 cigarettes in his or her lifetime but had quit at least 12 mo before diagnosis (for cases) or before the interview (for controls) was coded as a former smoker. Current smokers were those who were currently smoking or had quit <12 mo before diagnosis (for cases) or before the interview (for controls). Both former and current smokers were also defined as ever smokers. Body mass index (BMI; kg/m2) was calculated through self-reported usual height and weight. High blood pressure was assessed by whether a subject answered yes or no to ever having been told by a doctor that they had hypertension or high blood pressure. The study protocol was approved by the M. D. Anderson Institutional Review Board.

Selection of Studies Included in the Meta-Analysis. A comprehensive literature review was carried out to identify studies on risks for RCC associated with a family history of kidney cancer. We conducted a computerized search of PubMed for literature published in any language between 1950 and April 2008, and expanded by a review of previously cited references. To limit publication bias, the search criteria were not limited to “kidney cancer” and “family history” but included “RCC” and all suspected risk factors. In this meta-analysis, we included studies that fulfilled the following criteria: (a) presented original data from either case-control or cohort studies, (b) had family history of cancer or kidney cancer among first-degree relatives as the exposure of interest, (c) had RCC as the outcome of interest, and (d) provided RR estimates with confidence intervals (CI) or enough data to calculate them. We extracted the following information from each publication: the author's last name, publication year, study design, type of control (among case-controls studies), study location, sample size, family history assessment (type of first-degree relative), variables controlled for in the analysis, and RR estimates with CIs for RCC associated with a family history of cancer and kidney cancer.

We identified 11 studies: 7 case-control studies and 4 cohort studies (7-17). One cohort study (17) was excluded because the outcome of interest was defined as familial papillary RCC, and one case-control study (12) was excluded because the participants were included in an international study (15). Including the current case-control and family-based population studies, a total of 11 met our inclusion criteria. The 7 case-control studies were published between 1993 and 2007. Including data from the current case-control study, the number of cases and controls in the meta-analysis were 5,470 and 9,126, respectively. The 4 cohort studies were published between 1994 and 2002. The total number of subjects included in the cohort studies was 29,771.

Statistical Analysis

Descriptive analyses were conducted to characterize the study population of RCC cases, controls, and their first-degree relatives. Pearson's χ2 test was used to test for differences between cases and controls in the distribution of gender, ethnicity, smoking status, and history of hypertension. Student's t test was used to test for differences between subjects for continuous variables, including age, smoking pack-years, and BMI.

To study familial aggregation, two analytic approaches were utilized. First, a case-control analysis was done with the use of unconditional logistic regression to estimate the odds ratio (OR) associated with family history while adjusting for age, gender, ethnicity, smoking status, BMI, and history of hypertension. However, an inherent limitation in studying family history through a case-control analysis is that the family clustering may be explained by shared exposures among relatives and probands, including smoking habits (18). Subsequently, a family-based population analysis was applied to determine whether there was an excess of kidney cancer among first-degree relatives of case probands compared with control probands after controlling for confounding effects, including shared exposures, among the relatives. Probands themselves were excluded from this analysis. Instead, each first-degree relative was treated as a study subject, and his/her cancer status was treated as the outcome variable in the model. The family history variable was defined as positive if the study subject was related to a case proband, and the variable was defined as negative if the study subject was a relative of a control proband. ORs and 95% CIs associated with a family history of kidney cancer were calculated by fitting a generalized estimating equations model (19, 20) with binomial link function (logit) and exchangeable correlation structure, and then adjusted by age, gender, ethnicity, smoking, BMI and hypertension status, and proband age, smoking, and hypertension status. The generalized estimating equations were used to account for dependence within family members through the specification of the covariance structure (or within-group correlation structure) for measurements from members in the same family. All statistical tests were two-sided with a type I error rate of 5%. All statistical analyses were done with the Stata 8.2 statistical software package (Stata Corporation).

For the meta-analysis, we included studies reporting different measures of RR: case-control studies (ORs), cohort studies (rate ratios), and cohort studies with external population comparisons (standardized incidence ratios). Such an approach is warranted because these three measures of effect yield similar estimates of RR due to the absolute risk of RCC being low. Summary RRs were estimated with the statistical program Stata version 8.2 by inverse variance weighting with the use of fixed effects and random effects models that included a term for heterogeneity. All reported summary estimates in this study were based on the random effects model. The χ2 test of heterogeneity (Q test), which is based on a weighted sum of the squares of the log ORs, was estimated in the individual studies. Publication bias was assessed with the use of the funnel plot method of Begg and Mazumdar (21) and the Egger regression asymmetry test (22).

Family Aggregation Analyses

A total of 325 RCC cases and 329 controls were available for this analysis. Approximately 64% of the cases and controls were male (Table 1). There was no significant difference between the cases and the controls in terms of age (P = 0.780), ethnicity (P = 0.138), or smoking status (P = 0.160). Among ever smokers, cigarette consumption defined through median pack-years revealed no significant differences between cases and controls (P = 0.283). However, significant differences were observed between cases and controls for BMI, with cases having significantly higher BMI than controls (29.6 versus 28.4; P = 0.013). In cases, 56.3% had a history of hypertension/high blood pressure compared with only 40.6% in controls (P < 0.001).

In the case-control analysis, there was a nonsignificant increased risk of RCC associated with a positive family history of any cancer (adjusted OR, 1.15; 95% CI, 0.81-1.62). When the analysis was stratified by relative type, we observed a 1.67-fold increase in RCC if the family member with cancer was a sibling (95% CI, 1.13-2.47; Table 2). We did not observe an elevated risk for parents or offspring. When the analysis was limited to a family history of kidney cancer, a 4.32-fold increased risk of RCC was observed among those with at least one family member with kidney cancer (95% CI, 1.57-11.88; Table 2).

A second analytic approach was applied to control for risk factors of kidney cancer among the first-degree relatives. In this approach, each first-degree relative was treated as a study subject, and cancer status of first-degree relatives was treated as the outcome variable. The 325 case probands reported a total of 2,536 first-degree relatives, and the 329 control probands reported 2,333 first-degree relatives (Table 3). The case probands reported 362 (14.6%) first-degree relatives with any type of cancer, and the control probands reported 327 (14.2%; P = 0.712). First-degree relatives of cases with kidney cancer (0.8%) differed significantly from first-degree relatives of the controls (0.2%; P = 0.003). A family history of kidney cancer was associated with a borderline significant increased risk of kidney cancer (OR, 2.76; 95% CI, 0.98-7.78) after adjusting for age, sex, smoking status of relatives, proband age, proband sex, proband smoking status, proband history of hypertension, and proband BMI (results not shown).

Meta-Analysis

The studies included in the meta-analysis are summarized in Table 4. An overall combined estimate was generated for a family history of kidney cancer for both genders and all first-degree relatives combined (Fig. 1). One study was excluded because it reported RRs stratified by type of relative and did not report enough information to compute a combined RR (9). As shown in Fig. 1, the overall combined RR for the development of RCC associated with a positive family history of kidney cancer was 2.21 (95% CI, 1.55-2.87). When the current case-control and family-based studies were included separately, the RRs increased to 2.43 (95% CI, 1.73-3.12) and 2.27 (1.62-2.89), respectively. There was evidence of between-study heterogeneity (Q = 74.14; P < 0.001). The heterogeneity increased to 97.64 (P < 0.001) when the current case-control analysis was included and to 77.80 (P < 0.001) when the current family-based analysis was included.

Stratification by study design showed that a family history of kidney cancer was associated with a 2.32-fold increase in RCC (95% CI, 1.27-3.36) among case-control studies and a 1.93-fold increase in RCC (95% CI, 1.07-2.79) among cohort studies. When the current analyses were included, a 2.59-fold increase in RCC (95% CI, 1.55-3.64) was observed among case-controls studies and a 2.17-fold increase (95% CI, 1.35-2.99) among cohort studies. Further stratification showed a 2.87-fold increase (95% CI, 0.39-5.36) among hospital-based studies and a 2.37-fold increase in RCC (95% CI, 1.28-3.45) among population-based studies, including the current study. Evidence of heterogeneity fluctuated when stratified by type of study (cohort: Q = 5.03 and P = 0.025; case-control: Q = 84.74 and P < 0.001; hospital-based case-control: Q = 47.97 and P < 0.001; population-based case-control: Q = 28.72 and P < 0.001).

When the analysis was stratified by type of relative, the summary estimate among siblings with kidney cancer was 4.02 (95% CI, 2.48-5.56) including the current case-control study and 3.91 (95% CI, 2.30-5.51) including the current family-based study. The summary estimates for studies conducted in North America were slightly lower than those conducted in Europe (RR = 2.22; 95% CI, 1.53-2.90 and RR = 2.57; 95% CI, 1.31-3.84, respectively). Likewise, studies published in or before 2000 observed slightly lower estimates than those published after 2000 (RR = 1.75; 95% CI, 1.18-2.32 and RR = 2.56; 95% CI, 1.49-3.63, respectively). Although decreased, there still existed statistically significant heterogeneity within most subgroups.

The funnel plot showed slight asymmetry, reflecting the relative absence of studies with small sample sizes and negative effects. However, both the Begg's and Egger's tests showed no indication of significant publication bias (P = 0.174 and P = 0.188, respectively).

To our knowledge, this is the first study to use three analytic strategies (case-control analysis, family-based analysis, and meta-analysis) to investigate the association between a family history of kidney cancer and risk of RCC, and the first systematic evaluation of the relative risk for developing RCC associated with family history. Through a case-control analysis and a family-based population analysis, we observed significant associations between a family history of kidney cancer and RCC. The results of the meta-analysis further confirmed that a family history of kidney cancer is associated with a significant increase in RCC risk. Of utmost interest is the finding that the observed risks were higher when the affected relative was a sibling rather than a parent or child, suggesting the presence of low-penetrance genes (23).

Our findings are in agreement with previous research examining the association between a family history of kidney cancer and RCC risk. Consistent with our case-control estimate, which suggests a significant positive association, a population-based case-control study in Denmark with a similar sample size observed a statistically significant OR of 4.1 (95% CI, 1.1-14.9) in men and 4.8 in women (95% CI, 1.0-23), and a combined risk of 3.6 (95% CI, 1.6-8.2; ref. 12). However, an inherent limitation in studying family history through a case-control analysis is that the family clustering may be explained by shared exposures among relatives and probands, including smoking habits. We therefore applied a family-based population analysis that makes it possible to control for such confounding effects. In our family-based population analysis, we observed a borderline significantly increased risk associated with a family history of kidney cancer (OR, 2.76; 95% CI, 0.98-7.78). Consistent with our result, in a Utah study a familial relative risk of 2.5 (95% CI, 1.1-4.5) among first-degree relatives was observed (8).

Although the vast majority of RCC occurs sporadically, several hereditary conditions, including von Hippel–Lindau syndrome (24), hereditary papillary renal cancer related to germline mutations, activation of the MET and the FH gene (25), and Birt-Hogg-Dube syndrome (26), have been linked to RCC development. However, these syndromes are rare, and probably most of the familial risk is not due to these highly penetrant genes (27). Other lower-penetrance genes may exist with higher frequency in the population and may account for more cases of RCC (28).

The genetic interpretation of the familial risks is that dominant effect is reflected in offspring risk whereas recessive effect is signaled by elevated sibling risk (23). Previous studies examining the familial risk of RCC observed higher risks among siblings than among parents and offspring (9, 10, 12, 13, 16). The present study also observed a higher risk of RCC among siblings with any type of cancer than among parents and offspring. The same difference in risk was not observed when our analyses were focused on kidney cancer. A limitation of our familial aggregation analyses is small sample size and, thus, inability to study stratified analyses comprehensively. However, our meta-analysis did render a combined RR of 3.52 (95% CI, 2.23-4.82) for siblings, which was higher than the overall combined estimate of 2.37 (95% CI, 1.70-3.04). This observation among siblings supports the importance of recessive effects in familial RCC, in contrast with von Hippel–Lindau and other identified dominant familial RCC syndromes, and supports the existence of lower-penetrance susceptibility genes in RCC etiology.

A limitation of the current study is that information on family history of first-degree relatives was self-reported, which could have resulted in recall bias. However, research evaluating the accuracy and completeness of reporting family history has proven to be accurate for first-degree relatives (29, 30). By limiting family history to include only first-degree relatives, this study hoped to increase the validity of our exposure measurement. There exists the possibility of information bias due to proband cases reporting relatives with kidney cancer more than proband controls. However, several studies have reported that case or control status was not important for accurate reporting of first-degree relative family history (31-33). Specifically, Soegaard et al. observed no significant differences between cases and controls in the sensitivity (0.81 and 0.83, respectively), specificity (0.95 and 0.95, respectively), or κ (0.72 and 0.75, respectively) of reporting of familial cancer.

In literature, the relatively low response rates of the RDD screening have been extensively discussed (34, 35). A study comparing ovarian cancer controls selected through RDD and through a commercial database observed similar overall response rates to our study (36). Although the relatively low response rates of RDD screening in potential controls could have resulted in selection bias, researches have shown that controls selected by RDD usually represent characteristics of target population. For example, a study evaluating RDD suggests that control groups selected by RDD are representative of the general population (37). Brogon et al. (2001) reported that controls selected by RDD had characteristics similar to U.S. Census demographic characteristics of the target population (34). Nevertheless, there still exists the possibility of selection bias among the current case-control analysis, and the results should be interpreted with caution. Meta-analyses of observational studies are prone to biases and confounding factors that are inherent in the original studies. Nonetheless, as the first systematic assessment of its kind, the current meta-analysis suggests a significant association between a family history of kidney cancer and RCC, and an increased risk among siblings, indicating a need for the further examination of recessive effects. Lastly, the results of the current meta-analysis should be interpreted with caution. Although substantially decreased, significant heterogeneity remained after the current meta-analysis was stratified by type of study, region, year of publication, and relative type, indicating that other factors, such as differences in RCC histology, may be contributing to the heterogeneity.

Overall, our data suggest a significant positive association between a family history of kidney cancer and risk of RCC, and confirms this association through a significant combined estimate from our meta-analysis. With higher risks among siblings, our data also predict the existence of genes with low-penetrance germline mutations and warrant further examination of the genetic factors associated with RCC risk.

No potential conflicts of interest were disclosed.

Grant support: National Cancer Institute grants CA098897 and CA057730.

The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

1
Jemal A, Siegel R, Ward E, et al. Cancer statistics, 2008.
CA Cancer J Clin
2008
;
58
:
71
–96.
2
Chow WH, Devesa SS, Warren JL, Fraumeni JF, Jr. Rising incidence of renal cell cancer in the United States.
JAMA
1999
;
281
:
1628
–31.
3
Moore LE, Wilson RT, Campleman SL. Lifestyle factors, exposures, genetic susceptibility, and renal cell cancer risk: a review.
Cancer Invest
2005
;
23
:
240
–55.
4
Benichou J, Chow WH, McLaughlin JK, Mandel JS, Fraumeni JF, Jr. Population attributable risk of renal cell cancer in Minnesota.
Am J Epidemiol
1998
;
148
:
424
–30.
5
Cohen AJ, Li FP, Berg S, et al. Hereditary renal-cell carcinoma associated with a chromosomal translocation.
N Engl J Med
1979
;
301
:
592
–5.
6
Franksson C, Bergstrand A, Ljungdahl I, Magnusson G, Nordenstam H. Renal carcinoma (hypernephroma) occurring in 5 siblings.
J Urol
1972
;
108
:
58
–61.
7
Gago-Dominguez M, Yuan JM, Castelao JE, Ross RK, Yu MC. Family history and risk of renal cell carcinoma.
Cancer Epidemiol Biomarkers Prev
2001
;
10
:
1001
–4.
8
Goldgar DE, Easton DF, Cannon-Albright LA, Skolnick MH. Systematic population-based assessment of cancer risk in first-degree relatives of cancer probands.
J Natl Cancer Inst
1994
;
86
:
1600
–8.
9
Gudbjartsson T, Jonasdottir TJ, Thoroddsen A, et al. A population-based familial aggregation analysis indicates genetic contribution in a majority of renal cell carcinomas.
Int J Cancer
2002
;
100
:
476
–9.
10
Hung RJ, Moore L, Boffetta P, et al. Family history and the risk of kidney cancer: a multicenter case-control study in Central Europe.
Cancer Epidemiol Biomarkers Prev
2007
;
16
:
1287
–90.
11
Kreiger N, Marrett LD, Dodds L, Hilditch S, Darlington GA. Risk factors for renal cell carcinoma: results of a population-based case-control study.
Cancer Causes Control
1993
;
4
:
101
–10.
12
Mellemgaard A, Engholm G, McLaughlin JK, Olsen JH. Risk factors for renal cell carcinoma in Denmark. I. Role of socioeconomic status, tobacco use, beverages, and family history.
Cancer Causes Control
1994
;
5
:
105
–13.
13
Negri E, Foschi R, Talamini R, et al. Family history of cancer and the risk of renal cell cancer.
Cancer Epidemiol Biomarkers Prev
2006
;
15
:
2441
–4.
14
Randi G, Pelucchi C, Negri E, et al. Family history of urogenital cancers in patients with bladder, renal cell and prostate cancers.
Int J Cancer
2007
;
121
:
2748
–52.
15
Schlehofer B, Pommer W, Mellemgaard A, et al. International renal-cell-cancer study. VI. The role of medical and family history.
Int J Cancer
1996
;
66
:
723
–6.
16
Czene K, Hemminki K. Kidney cancer in the Swedish Family Cancer Database: familial risks and second primary malignancies.
Kidney Int
2002
;
61
:
1806
–13.
17
Czene K, Hemminki K. Familial papillary renal cell tumors and subsequent cancers: a nationwide epidemiological study from Sweden.
J Urol
2003
;
169
:
1271
–5.
18
Lin J, Spitz MR, Dinney CP, Etzel CJ, Grossman HB, Wu X. Bladder cancer risk as modified by family history and smoking.
Cancer
2006
;
107
:
705
–11.
19
Liang KY, Beaty TH. Statistical designs for familial aggregation.
Stat Methods Med Res
2000
;
9
:
543
–62.
20
Zeger SL, Liang KY. Longitudinal data analysis for discrete and continuous outcomes.
Biometrics
1986
;
42
:
121
–30.
21
Begg CB, Mazumdar M. Operating characteristics of a rank correlation test for publication bias.
Biometrics
1994
;
50
:
1088
–101.
22
Egger M, Davey Smith G, Schneider M, Minder C. Bias in meta-analysis detected by a simple, graphical test.
BMJ
1997
;
315
:
629
–34.
23
Risch N. The genetic epidemiology of cancer: interpreting family and twin studies and their implications for molecular genetic approaches.
Cancer Epidemiol Biomarkers Prev
2001
;
10
:
733
–41.
24
Latif F, Tory K, Gnarra J, et al. Identification of the von Hippel-Lindau disease tumor suppressor gene.
Science
1993
;
260
:
1317
–20.
25
Tomlinson IP, Alam NA, Rowan AJ, et al. Germline mutations in FH predispose to dominantly inherited uterine fibroids, skin leiomyomata and papillary renal cell cancer.
Nat Genet
2002
;
30
:
406
–10.
26
Nickerson ML, Warren MB, Toro JR, et al. Mutations in a novel gene lead to kidney tumors, lung wall defects, and benign tumors of the hair follicle in patients with the Birt-Hogg-Dube syndrome.
Cancer Cell
2002
;
2
:
157
–64.
27
Peto J, Houlston RS. Genetics and the common cancers.
Eur J Cancer
2001
;
37
:
S88
–96.
28
Hemminki K, Li X. Familial renal cell cancer appears to have a recessive component.
J Med Genet
2004
;
41
:
e58
.
29
Boyle CA, Brann EA; The Selected Cancers Cooperative Study Group. Proxy respondents and the validity of occupational and other exposure data.
Am J Epidemiol
1992
;
136
:
712
–21.
30
Theis B, Boyd N, Lockwood G, Tritchler D. Accuracy of family cancer history in breast cancer patients.
Eur J Cancer Prev
1994
;
3
:
321
–7.
31
Soegaard M, Jensen A, Frederiksen K, et al. Accuracy of self-reported family history of cancer in a large case-control study of ovarian cancer.
Cancer Causes Control
2008
;
19
:
469
–79.
32
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
–8.
33
Mitchell RJ, Brewster D, Campbell H, et al. Accuracy of reporting of family history of colorectal cancer.
Gut
2004
;
53
:
291
–5.
34
Brogan DJ, Denniston MM, Liff JM, Flagg EW, Coates RJ, Brinton LA. Comparison of telephone sampling and area sampling: response rates and within-household coverage.
Am J Epidemiol
2001
;
153
:
1119
–27.
35
DiGaetano R, Waksberg J. Commentary: trade-offs in the development of a sample design for case-control studies.
Am J Epidemiol
2002
;
155
:
771
–5.
36
Olson SH, Kelsey JL, Pearson TA, Levin B. Evaluation of random digit dialing as a method of control selection in case-control studies.
Am J Epidemiol
1992
;
135
:
210
–22.
37
Olson SH, Mignone L, Harlap S. Selection of control groups by using a commercial database and random digit dialing.
Am J Epidemiol
2000
;
152
:
585
–92.