Case-control and case-only designs with genotype and family history data: estimating relative risk, residual familial aggregation, and cumulative risk

In case-control studies of inherited diseases, participating subjects (probands) are often interviewed to collect detailed data about disease history and age-at-onset information in their family members. Genotype data are typically collected from the probands, but not from their relatives. In this article, we introduce an approach that combines case-control analysis of data on the probands with kin-cohort analysis of disease history data on relatives. Assuming a marginally specified multivariate survival model for joint risk of disease among family members, we describe methods for estimating relative risk, cumulative risk, and residual familial aggregation. We also describe a variation of the methodology that can be used for kin-cohort analysis of the family history data from a sample of genotyped cases only. We perform simulation studies to assess performance of the proposed methodologies with correct and mis-specified models for familial aggregation. We illustrate the proposed methodologies by estimating the risk of breast cancer from BRCA1/2 mutations using data from the Washington Ashkenazi Study.     

Approaches to familial aggregation: hypothesis testing and estimation when families are selected through parent probands under a variant of single ascertainment

Epidemiologic approaches to testing and estimating familial aggregation of a disease consist of comparing rates of disease in relatives of individuals with the disease (known as case probands) with rates of disease in relatives of individuals without the disease (known as control probands). Gold et al. (J Am Stat Ass 1967;62: 409-420) derived an explicit mathematical model and sampling methods, under which this approach is equivalent to testing the null hypotheses that the disease risk in families is homogenous. A basic assumption of this model is that every family member has the same risk of disease and that disease status is independent among family members, although the disease risk may vary between families. When the disease is suspected of having a genetic component, rather than being purely environmental, this model has been shown to be appropriate for detecting disease aggregation in siblings, when relatives are siblings of probands. This model however is unrealistic for use in nuclear families when the affected status of offspring is not independent of the affected status of parents, and these families are selected through an affected or an unaffected parent, so that a parent is the proband and relatives are offspring of probands. We extend the Gold et al. model to allow for the disease risk in offspring to vary with the affected status of the parent. We assume that families are selected through affected and unaffected parents, under a variation of single ascertainment. Under this study design, we show that the usual test of association between affected status of probands and relatives, performed by comparing sample proportions of affected relatives of affected and unaffected probands, respectively, is no longer equivalent to a test of homogeneity of disease risk in offspring. Instead, it is equivalent to testing that the disease risk in offspring is independent of the number of affected parents. This test reduces to a test of homogeneity if and only if one assumes that the variation in disease risk in offspring, between families, is solely due to the variation in the number of affected parents. As a result, we show that under this study design, the standard chi2 test must be modified in order to obtain a valid test of familial aggregation. In addition the sample proportions of affected relatives of case and control probands, respectively, are shown to provide unbiased estimates of the expected risk of disease in an offspring given an affected/unaffected parent. We apply these results to methods of sample selection and discuss the practical implications of these findings.     

Familial Aggregation of Morbid Obesity

Recent studies have shown major gene effects for obesity in randomly ascertained families. To investigate the familial aggregation of a specific subset of obesity, which is particularly prone to medical complications, families with morbid obesity were studied. This condition occurs in 1%-2%of the population and is defined as 45.5 kg (100 pounds) or more over ideal weight. First-degree relatives of 221 morbidly obese probands (1560 adults) were identified, and height and weight (current and greatest) were obtained from each family member. Morbid obesity occurred in the family members of the probands 8 times more often than in the general population. Of the morbidly obese probands, 48% had one or more first-degree relatives who were also morbidly obese compared to a 6% population estimate. By the ages of 20–24, 12% of the morbidly obese probands were already 45.5 kg or more overweight, and 45% were 22.7 kg (50 pounds) or more overweight. There was little difference in the prevalence of familial morbid obesity by the gender of the probands: 47% of the male probands and 48% of the female probands had another morbidly obese relative, while 67% and 53% of the early onset (before age 25) male and female probands, respectively, had one or more first-degree relatives who were also morbidly obese. In addition to the extreme degree of familial aggregation, the prevalence of morbid obesity in parent-offspring sets was calculated within the morbidly obese families. Morbidly obese families who have one or two morbidly obese parents have a 2.6 times increased risk (p<0.002) of having one or more morbidly obese adult offspring, compared to families who have neither parent morbidly obese. Evidence for trimodality of the body mass index distribution was found for each gender (p = 0.0006 for male relatives and p = 0.075 for female relatives). The strong familial aggregation of morbid obesity indicates the need for further understanding of the genetic determinants of this extreme clinical disorder and how environmental factors affect the genetic expression of the trait. (OBESITY RESEARCH 1993;1:261–270)     

A semiparametric method for analyzing matched case-control family studies with a continuous outcome and proband sampling

We consider matched case-control familial studies which match a group of patients, called "case probands," with a group of disease-free subjects, called "control probands," using a set of family-level matching variables. Family members of each proband are then recruited into the study. Of interest here is the familial aggregation of the response variable and the effects of subject-specific covariates on the response. We propose an estimating equation approach to jointly estimate the main effects and intrafamilial correlations for matched family studies with a continuous outcome. Only knowledge of the first two joint moments of the response variable is required. The induced estimators for the main effects and intrafamilial correlations are consistent and asymptotically normally distributed. We apply the proposed method to sleep apnea data. A simulation study demonstrates the usefulness of our approach.     

Type II diabetes of early onset: a distinct clinical and genetic syndrome?

The inheritance of non-insulin-dependent (type II) diabetes was studied by a continuous infusion of glucose test in all available first degree relatives of 48 diabetic probands of various ages and with differing severity of disease. In an initial study of 38 type II diabetic subjects and their first degree relatives six islet cell antibody negative patients with early onset disease (aged 25-40 at diagnosis) were found to have a particularly high familial prevalence of diabetes or glucose intolerance. Nine of 10 parents available for study either had type II diabetes or were glucose intolerant. A high prevalence of diabetes or glucose intolerance was also found in their siblings (11/16;69%). In a second study of the families of a further 10 young diabetic probands (presenting age 25-40) whose islet cell antibody state was unknown a similar high prevalence of diabetes or glucose intolerance was found among parents of the five islet cell antibody negative probands (8/9; 89%) but not among parents of the five islet cell antibody positive probands (3/8;38%). Islet cell antibody negative diabetics with early onset type II disease may have inherited a diabetogenic gene or genes from both parents. They commonly need insulin to maintain adequate glycaemic control and may develop severe diabetic complications. Early onset type II diabetes may represent a syndrome in which characteristic pedigrees, clinical severity, and absence of islet autoimmunity make it distinct from either type I diabetes, maturity onset diabetes of the young, or late onset type II diabetes.     

Familial Correlations of Onset Age of Hepatocellular Carcinoma: A Population-Based Case-Control Family Study

Background

There was lack of evidence for familial aggregation in onset age of hepatocellular carcinoma (HCC) in Chinese population. We conducted a population-based case-control family study to examine familial correlation of age of HCC onset in Taixing, China.

Methods

A total of 202 cases and 202 matched controls as well as their relatives were included in the study. Lifetime cumulative risks of HCC were estimated using the Kaplan-Meier approach. Cross ratios (CRs) were obtained from stratified Cox proportional hazard models, to assess the familial correlation of onset age.

Results

The mean age of HCC onset was decreased as increasing number of HCC cases in a family. The onset age was the earliest for first-degree relatives, intermediate for second-degree relatives, and latest for non-blood relatives (spouse) (log-rank test, P<0.01). The onset age was significantly correlated between probands and their relatives. In stratified Cox proportional hazard models, the CRs for the probands versus their fathers, mothers, siblings and uncles/aunts were 6.25 (95% confidence interval (CI): 1.84–21.25), 9.81 (95% CI: 1.24–77.56), 6.22 (95% CI: 1.37–28.36) and 3.24 (95% CI: 1.26–8.33), respectively. After adjustment for hepatitis B virus infection, the CRs remained significant.

Conclusion

This current study suggested a significant correlation of onset age for HCC among blood relatives. Familial HCC cases yielded earlier age of onset and their relatives have higher HCC risk in early age, highlighting intensive surveillance should be start at an earlier age for individuals with family history of HCC.     

Genetic analysis of kifafa, a complex familial seizure disorder.

Kifafa is the Swahili name for an epileptic seizure disorder, first reported in the early 1960s, that is prevalent in the Wapogoro tribe of the Mahenge region of Tanzania in eastern Africa. A 1990 epidemiological survey of seizure disorders in this region reported a prevalence in the range of 19/1,000-36/1,000, with a mean age at onset of 11.6 years; 80% of those affected had onset prior to 20 years of age. A team of investigators returned to Tanzania in 1992 and collected data on > 1,600 relatives of 26 probands in 20 kifafa families. We have undertaken a genetic analysis of these data in order to detect the presence of familial clustering and whether such aggregation could be attributed to genetic factors. Of the 127 affected individuals in these pedigrees, 23 are first-degree relatives (parent, full sibling, or offspring) of the 26 probands; 20 are second-degree relatives (half-sibling, grandparent, uncle, or aunt). When corrected for age, the risk to first-degree relatives is .15; the risk to second-degree relatives is .063. These risks are significantly higher than would be expected if there were no familial clustering. Segregation analysis, using PAP (rev.4.0), was undertaken to clarify the mode of inheritance. Among the Mendelian single-locus models, an additive model was favored over either a dominant, recessive, or codominant model. The single-locus model could be rejected when compared with the mixed Mendelian model (inclusion of a polygenic background), although the major-gene component tends to be recessive.(ABSTRACT TRUNCATED AT 250 WORDS)     

Sampling Considerations for Disease Surveillance in Wildlife Populations

Abstract Disease surveillance in wildlife populations involves detecting the presence of a disease, characterizing its prevalence and spread, and subsequent monitoring. A probability sample of animals selected from the population and corresponding estimators of disease prevalence and detection provide estimates with quantifiable statistical properties, but this approach is rarely used. Although wildlife scientists often assume probability sampling and random disease distributions to calculate sample sizes, convenience samples (i.e., samples of readily available animals) are typically used, and disease distributions are rarely random. We demonstrate how landscape-based simulation can be used to explore properties of estimators from convenience samples in relation to probability samples. We used simulation methods to model what is known about the habitat preferences of the wildlife population, the disease distribution, and the potential biases of the convenience-sample approach. Using chronic wasting disease in free-ranging deer (Odocoileus virginianus) as a simple illustration, we show that using probability sample designs with appropriate estimators provides unbiased surveillance parameter estimates but that the selection bias and coverage errors associated with convenience samples can lead to biased and misleading results. We also suggest practical alternatives to convenience samples that mix probability and convenience sampling. For example, a sample of land areas can be selected using a probability design that oversamples areas with larger animal populations, followed by harvesting of individual animals within sampled areas using a convenience sampling method.     

Genetic epidemiology of vitiligo: multilocus recessivity cross-validated.

Vitiligo is a dermatological disorder characterized by hypopigmentary patches that tend to become progressive over time. There are reports of extensive familial aggregation. A genetic model for this disorder was earlier proposed by us. This model postulates that recessive alleles at multiple unlinked autosomal loci interact epistatically in the pathogenesis of vitiligo. The present family study was primarily undertaken to cross-validate the proposed genetic model. Data on 194 families from the United States were collected. Each family was ascertained through an affected proband. Analyses of these data reveal that approximately 20% of probands have at least one first-degree relative afflicted with vitiligo. All types of first-degree relatives of probands show a significant risk of developing vitiligo. Results of segregation and robustness analyses reveal that the genetic model postulated by us previously is the most parsimonious model for the present family data set.     

Etiologic heterogeneity in the familial aggregation of congenital cardiovascular malformations.

Recent data indicate that there is increased risk of congenital cardiovascular malformations (CCVM) within families of probands diagnosed with congenital cardiovascular malformations that are due to altered embryonic blood flow (flow lesions). In the present study, regressive models recently developed by Bonney were used to compare specific models of inheritance and to test for etiologic heterogeneity among three subgroups of 375 flow-lesion families identified by the Baltimore-Washington Infant Study. When all families were analyzed as a single group, the best-fitting model was a simple recessive model with Mendelian transmission; race did not have a significant effect on estimated risk. Separate analyses of families of probands with left heart defects, right heart defects, and ventricular septal defects (VSD) confirmed this simple Mendelian recessive model as the model of choice. However, when race was included as a covariate in these genetic models, there was evidence for significant heterogeneity among the three subgroups. There was an increased risk to relatives of white probands with right heart defects and to relatives of black probands with VSD, while there was no effect of race among relatives of probands with left heart defects. These results strongly suggest that there is etiologic heterogeneity in the control of CCVM among flow-lesion families and that the patterns of familial aggregation differ among the races.     

Analytical Relationship between Family History and Genetic and Environmental risks,with Application to Female Breast Cancer

It is well established in genetic epidemiology that family history is an important indicator of familial aggregation of disease in a family. A strong genetic risk factor or an environmental risk factor with high familial correlation can result in a strong family history. In this paper, family history refers to the number of first‐degree relatives affected with the disease. Cui and Hopper (Journal of Epidemiology and Biostatistics 2001; 6 : 331–342) proposed an analytical relationship between family history and relevant genetic parameters. In this paper we expand the relationship to both genetic and environmental risk factors. We established a closed‐form formula for family history as a function of genetic and environmental parameters which include genetic and environmental relative risks, genotype frequency, prevalence and familial correlation of the environmental risk factor. The relationship is illustrated by an example of female breast cancer in Australia. For genetic and environmental relative risks less than 10, most of the female breast cancer cases occur between the age of 40 and 60 years. A higher genetic or environmental relative risk will move the peak of the distribution to a younger age. A more common disease allele or more prevalent environmental risk factor will move the peak to an older age. For a proband with breast cancer, it is most likely (with probability ≥80%) that none of her first‐degree relatives is affected with the disease. To enable the probability of having a positive family history to reach 50%, the environmental relative risks must be extremely as high as 100, the familial correlation as high as 0.8 and the prevalence as low as 0.1. For genetic risk alone, even the relative risk is as high as 100, the probability of having a positive family history can only reach about 30%. This suggests that the environmental risk factor seems to play a more important role in determining a strong family history than the genetic risk factor. (© 2004 WILEY‐VCH Verlag GmbH & Co. KGaA, Weinheim)     

Interval Estimation of Simple Difference under Independent Negative Binomial Sampling

When the sample size is not large or when the underlying disease is rare, to assure collection of an appropriate number of cases and to control the relative error of estimation, one may employ inverse sampling, in which one continues sampling subjects until one obtains exactly the desired number of cases. This paper focuses discussion on interval estimation of the simple difference between two proportions under independent inverse sampling. This paper develops three asymptotic interval estimators on the basis of the maximum likelihood estimator (MLE), the uniformly minimum variance unbiased estimator (UMVUE), and the asymptotic likelihood ratio test (ALRT). To compare the performance of these three estimators, this paper calculates the coverage probability and the expected length of the resulting confidence intervals on the basis of the exact distribution. This paper finds that when the underlying proportions of cases in both two comparison populations are small or moderate (≤0.20), all three asymptotic interval estimators developed here perform reasonably well even for the pre-determined number of cases as small as 5. When the pre-determined number of cases is moderate or large (≥50), all three estimators are essentially equivalent in all the situations considered here. Because application of the two interval estimators derived from the MLE and the UMVUE does not involve any numerical iterative procedure needed in the ALRT, for simplicity we may use these two estimators without losing efficiency.     

Estimating the magnitude of genetic factors by calculating the genetic relative risk of stroke in first-ever lacunar stroke patients

Background

Positive family history of stroke is an independent risk factor for lacunar stroke. However, the magnitude of familial aggregation of a certain disease is better evaluated by the genetic relative risk. This is calculated by dividing the prevalence of specific disease in family members of patients by the prevalence of this disease in the general population. In a cohort of lacunar stroke patients, who were subtyped clinically and radiologically, we determined the genetic relative risk of stroke.

Methods

By questionnaire and additional interview, we obtained a complete first-degree family history of stroke. The prevalence of stroke in first-degree relatives of these lacunar stroke patients was compared to the self-reported prevalence of stroke in a Dutch community based cohort of elderly volunteers. Secondly, the influence of proband characteristics and family composition on parental and sibling history of stroke were evaluated.

Principal Findings

We collected data of 1066 first-degree relatives of 195 lacunar stroke patients. Strokes occurred in 13.5% of first-degree relatives. The genetic relative risk was 2.94 (95%CI 2.45–3.53) for overall first-degree relatives, 4.52 (95%CI 3.61–5.65) for patients'' parents and 2.10 (95%CI 1.63–2.69) for patients'' siblings. Age of proband and proband status for hypertension influenced the chance of having a parent with a history of stroke whereas the likelihood of having a concordant sibling increased with sibship size.

Conclusions

We found an increased genetic relative risk of stroke in first-degree relatives of patients with lacunar stroke. Our data warrant further genomic research in this well-defined high risk population for stroke.     

Familial aggregation in lone atrial fibrillation

Atrial fibrillation (AF) is the most common clinical arrhythmia and a major risk factor for stroke. To investigate the role of genetic factors in a typical clinical population, we determined the extent of familial aggregation in patients with lone AF. To estimate the relative risk to family members, the prevalence of AF for each class of relative was compared to the prevalence in the comparable age and sex group from the general population. Family members had an increased relative risk of AF compared to the general population (risk ratio; 95% confidence intervals): sons (8.1; 2.0–32), daughters (9.5; 1.3–67), brothers (70; 47–102), sisters (34; 14–80), mothers (4.0; 2.5–6.5) and fathers (2.0; 1.2–3.6). Relatives of probands with lone AF are at a substantially increased risk of developing this arrhythmia suggesting a Mendelian genetic contribution to the etiology of this common trait.     

Evidence for a third genetic locus causing familial hypercholesterolemia. A non-LDLR, non-APOB kindred.

Monogenically inherited hypercholesterolemia is most commonly caused by mutations at the low density lipoprotein receptor (LDLR) locus causing familial hypercholesterolemia (FH) or at the apolipoprotein B (APOB) locus causing the disorder familial defective apoB (FDB). Probands from 47 kindreds with a strict clinical diagnosis of FH were selected from the Cardiovascular Genetics Research Lipid Clinic, Utah, for molecular genetic analysis. Using a combination of single-strand conformation polymorphism (SSCP) and direct sequencing, 12 different LDLR gene mutations were found in 16 of the probands. Three of the probands were carriers of the APOB R3500Q mutation. In five of the remaining 28 pedigrees where no mutation had been detected, samples from enough relatives were available to examine co-segregation with the LDLR region using the microsatellite marker D19S221, which is within 1 Mb centromeric of the LDLR locus, and D19S394, sited within 150 kb telomeric of the LDLR locus. In four of the families there was strong evidence for co-segregation between the LDLR locus and the phenotype of hypercholesterolemia, but in one large family with 18 living affected members and clear-cut bimodal hypercholesterolemia, there were numerous exclusions of co-segregation. Using length polymorphic markers within and outside the APOB gene, linkage of phenotype in this family to the APOB region was similarly excluded. In this large family, the degree of hypercholesterolemia, prevalence of tendon xanthomata, and occurrence of early coronary disease were indistinguishable from the other families studied. In summary, the data provide unequivocal evidence that a third locus can be etiological for monogenic familial hypercholesterolemia and should be reinvigorating to research in this field.     

Genetic Segregation Analysis of Early-Onset Recurrent Unipolar Depression

Major depression is a relatively common psychiatric disorder that can be quite debilitating. Family, twin, and adoption studies indicate that unipolar depression has both genetic and environmental components. Early age at onset and recurrent episodes in the proband each increase the familiarity of the illness. To investigate the potential genetic underpinnings of the disease, we have performed a complex segregation analysis on 832 individuals from 50 multigenerational families ascertained through a proband with early-onset recurrent unipolar major depression. The analysis was conducted by use of regressive models, to test a variety of hypotheses to explain the familial aggregation of recurrent unipolar depression. Analyses were conducted under two alternative definitions of affection status for the relatives of probands: (1) "narrow," in which relatives were assumed to be affected only if they were diagnosed with recurrent unipolar depression; and (2) "broad," in which relatives were assumed to be affected if diagnosed with any major affective illness. Under the narrow-definition assumption, the model that best explains these family data is a transmitted (although non-Mendelian) recessive major effect with significant residual parental effects on affection status. Under the broad-definition assumption, the best-fitting model is a Mendelian codominant major locus with significant residual parental and spousal effects.     

A family study of Gilles de la Tourette syndrome.

Previous studies have demonstrated that Gilles de la Tourette syndrome (TS) is a familial disorder and that chronic tics (CT) and obsessive compulsive disorder (OCD) appear to be etiologically related to the syndrome. In the present study we report the results from a study of 338 biological relatives of 86 TS probands, 21 biologically unrelated relatives of adopted TS probands, and 22 relatives of normal subjects. The 43 first-degree relatives of the adopted TS and normal probands constituted a control sample. The rates of TS, CT, and OCD in the total sample of biological relatives of TS probands were significantly greater than in the relatives of controls. In addition, the morbid risks of TS, OCD, and CT were not significantly different in families of probands with OCD when compared to relatives of probands without OCD. These findings provide further evidence that OCD is etiologically related to TS.     

HLA class II haplotype DRB1*04–DQB1*0301 contributes to risk of familial generalized vitiligo and early disease onset

Generalized vitiligo is a common autoimmune disorder characterized by white patches of skin and overlying hair caused by loss of pigment‐forming melanocytes from involved areas. Familial clustering of vitiligo is not uncommon, and patients and their relatives are at increased risk for a specific complex of other autoimmune diseases. Compared with sporadic vitiligo, familial vitiligo is characterized by earlier disease onset and greater risk and broader repertoire of autoimmunity, suggesting a stronger genetic component, and perhaps stronger associations with specific alleles. To determine whether the major histocompatibility complex (MHC) contributes to the familial clustering of vitiligo and vitiligo‐associated autoimmune/autoinflammatory diseases, we performed case–control and family‐based association analyses of HLA class II‐DRB1 and ‐DQB1 alleles and haplotypes in affected probands and their parents from 76 European‐American Caucasian families with familial vitiligo. Affected probands showed a significantly increased frequency of DRB1*04–DQB1*0301 and a significantly decreased frequency of DRB1*15–DQB1*0602 compared with a large sample of reference chromosomes. Family‐based association analyses confirmed these results. Probands with DRB1*04–DQB1*0301 developed vitiligo an average of 13.32 yr earlier than probands with DRB1*15–DQB1*0602. Overall, our results indicate that specific MHC‐linked genetic variation contributes to risk of familial vitiligo, although HLA does not completely explain familial clustering of vitiligo‐associated autoimmune/autoinflammatory diseases.     

Segregation analysis of cryptogenic epilepsy and an empirical test of the validity of the results.

We used POINTER to perform segregation analysis of cryptogenic epilepsy in 1,557 three-generation families (probands and their parents, siblings, and offspring) ascertained from voluntary organizations. Analysis of the full data set indicated that the data were most consistent with an autosomal dominant (AD) model with 61% penetrance of the susceptibility gene. However, subsequent analyses revealed that the patterns of familial aggregation differed markedly between siblings and offspring of the probands. Risks in siblings were consistent with an autosomal recessive (AR) model and inconsistent with an AD model, whereas risks in offspring were inconsistent with an AR model and more consistent with an AD model. As a further test of the validity of the AD model, we used sequential ascertainment to extend the family history information in the subset of families judged likely to carry the putative susceptibility gene because they contained at least three affected individuals. Prevalence of idiopathic/cryptogenic epilepsy was only 3.7% in newly identified relatives expected to have a 50% probability of carrying the susceptibility gene under an AD model. Approximately 30% (i.e., 50% x 61%) were expected to be affected under the AD model resulting from the segregation analysis. These results suggest that the familial distribution of cryptogenic epilepsy is inconsistent with any conventional genetic model. The differences between siblings and offspring in the patterns of familial risk are intriguing and should be investigated further.     

Hip joint replacement surgery for idiopathic osteoarthritis aggregates in families

In order to determine whether there is a genetic component to hip or knee joint failure due to idiopathic osteoarthritis (OA), we invited patients (probands) undergoing hip or knee arthroplasty for management of idiopathic OA to provide detailed family histories regarding the prevalence of idiopathic OA requiring joint replacement in their siblings. We also invited their spouses to provide detailed family histories about their siblings to serve as a control group. In the probands, we confirmed the diagnosis of idiopathic OA using American College of Rheumatology criteria. The cohorts included the siblings of 635 probands undergoing total hip replacement, the siblings of 486 probands undergoing total knee replacement, and the siblings of 787 spouses. We compared the prevalence of arthroplasty for idiopathic OA among the siblings of the probands with that among the siblings of the spouses, and we used logistic regression to identify independent risk factors for hip and knee arthroplasty in the siblings. Familial aggregation for hip arthroplasty, but not for knee arthroplasty, was observed after controlling for age and sex, suggesting a genetic contribution to end-stage hip OA but not to end-stage knee OA. We conclude that attempts to identify genes that predispose to idiopathic OA resulting in joint failure are more likely to be successful in patients with hip OA than in those with knee OA.