Residual Familial Correlations Research Articles

FamEvent: An R Package for Generating and Modeling Time-to-Event Data in Family Designs.

FamEvent is a comprehensive R package for simulating and modelling age-at-disease onset in families carrying a rare gene mutation. The package can simulate complex family data for variable time-to-event outcomes under three common family study designs (population, high-risk clinic and multi-stage) with various levels of missing genetic information among family members. Residual familial correlation can be induced through the inclusion of a frailty term or a second gene. Disease-gene carrier probabilities are evaluated assuming Mendelian transmission or empirically from the data. When genetic information on the disease gene is missing, an Expectation-Maximization algorithm is employed to calculate the carrier probabilities. Penetrance model functions with ascertainment correction adapted to the sampling design provide age-specific cumulative disease risks by sex, mutation status, and other covariates for simulated data as well as real data analysis. Robust standard errors and 95% confidence intervals are available for these estimates. Plots of pedigrees and penetrance functions based on the fitted model provide graphical displays to evaluate and summarize the models.

Power, Validity, Bias and Robustness of Family‐based Association Analysis Methods in the Presence of Linkage for Late Onset Diseases

This simulation-based report compares the performance of five methods of association analysis in the presence of linkage using extended sibships: the Family-Based Association Test (FBAT), Empirical Variance FBAT (EV-FBAT), Conditional Logistic Regression (CLR), Robust CLR (R-CLR) and Sibship Disequilibrium Test (SDT). The two tests accounting for residual familial correlation (EV-FBAT and R-CLR) and the model-free SDT showed correct test size in all simulated designs, while FBAT and CLR were only valid for small effect sizes. SDT had the lowest power, while CLR had the highest power, generally similar to FBAT and the robust variance analogues. The power of all model-dependent tests dropped when the model was misspecified, although often not substantially. Estimates of genetic effect with CLR and R-CLR were unbiased when the disease locus was analysed but biased when a nearby marker was analysed. This study demonstrates that the genetic effect does not need to be extreme to invalidate tests that ignore familial correlation and confirms that analogous methods using robust variance estimation provide a valid alternative at little cost to power. Overall R-CLR is the best-performing method among these alternatives for the analysis of extended sibship data.

A genome-wide search replicates evidence of a quantitative trait locus for circulating angiotensin I-converting enzyme (ACE) unlinked to the ACE gene

BackgroundAngiotensin I-converting enzyme (ACE) plays an important role in cardiovascular homeostasis. There is evidence from different ethnic groups that circulating ACE levels are influenced by a quantitative trait locus (QTL) at the ACE gene on chromosome 17. The finding of significant residual familial correlations in different ethnic groups, after accounting for this QTL, and the finding of support for linkage to a locus on chromosome 4 in Mexican-American families strongly suggest that there may well be QTLs for ACE unlinked to the ACE gene.MethodsA genome-wide panel of microsatellite markers, and a panel of biallelic polymorphisms in the ACE gene were typed in Nigerian families. Single locus models with fixed parameters were used to test for linkage to circulating ACE with and without adjustment for the effects of the ACE gene polymorphisms.ResultsStrong evidence was found for D17S2193 (Zmax = 3.5); other nearby markers on chromosome 17 also showed modest support. After adjustment for the effects of the ACE gene locus, evidence of "suggestive linkage" to circulating ACE was found for D4S1629 (Zmax = 2.2); this marker is very close to a locus previously shown to be linked to circulating ACE levels in Mexican-American families.ConclusionIn this report we have provided further support for the notion that there are QTLs for ACE unlinked to the ACE gene; our findings for chromosome 4, which appear to replicate the findings of a previous independent study, should be considered strong grounds for a more detailed examination of this region in the search for genes/variants which influence ACE levels.The poor yields, thus far, in defining the genetic determinants of hypertension risk suggest a need to look beyond simple relationships between genotypes and the ultimate phenotype. In addition to incorporating information on important environmental exposures, a better understanding of the factors which influence the building blocks of the blood pressure homeostatic network is also required. Detailed studies of the genetic determinants of ACE, an important component of the renin-angiotensin system, have the potential to contribute to this strategic objective.

Homocysteine levels – before and after methionine loading – in 51 Dutch families

Elevated levels of homocysteine are a risk factor for vascular disease, thrombosis, neural tube defects and dementia. The 677C>T polymorphism in the methylenetetrahydrofolate reductase (MTHFR) gene appears to be the most important single determinant of plasma homocysteine concentration. In the current study, we estimated heritability and fit a series of models of inheritance for both fasting and postmethionine-load homocysteine levels in the HOFAM-study (HOmocysteine in FAMilies study), which included 306 participants from 51 pedigrees, ascertained through a hyperhomocysteinemic proband. The crude heritability was 21.6% for fasting and 67.5% for postloading homocysteine. After adjustment for MTHFR 677C>T genotype, heritability dropped to 5.2 and 63.9%, respectively. Segregation analysis revealed that a nongenetic model with equal transmission was the best fitting and most parsimonious model for fasting homocysteine levels, while a two-distribution, Mendelian model with residual familial correlation was best for postmethionine-load homocysteine levels. This study shows that postload homocysteine levels have a stronger genetic determination than do fasting homocysteine levels. The heritability of postload homocysteine levels were not strongly affected by adjustment for MTHFR 677C>T genotype, in contrast to fasting homocysteine levels. Further studies are needed to identify the genes responsible for the inheritance of postload homocysteine levels.

Segregation analysis of phenotypic components of learning disabilities. II. Phonological decoding.

Dyslexia is a common, complex disorder, which is thought to have a genetic component. The study of the genetics of dyslexia is complicated by a lack of consensus on diagnostic criteria, and the probability of genetic heterogeneity-it is possible that deficits in different language processes are caused by different underlying genes. In order to address these difficulties, we study continuous phenotypes that are part of the psychometric test batteries often used to diagnose dyslexia. Prior to embarking on a linkage study, it is helpful to employ segregation analysis, both to identify phenotypes that may be amenable to mapping by linkage analysis, and to determine the best models to use for model based analyses. We study 409 people in 102 nuclear families, and employ (1) oligogenic segregation analysis to estimate the number of quantitative trait loci (QTLs) contributing to each phenotype, and (2) complex segregation analysis in order to identify the most parsimonious inheritance model. In this paper, we consider two measures of phonological decoding ability-word attack and phonemic decoding efficiency. We find evidence for one or two genes of at least modest effect contributing to phonemic decoding efficiency, and the best fitting model is a dominant major gene model with residual familial correlations. For word attack, we find evidence for one or two genes of at least modest effect, and the variation in the trait is best explained by a polygenic model.

Detection of putative functional angiotensinogen (AGT) gene variants controlling plasma AGT levels by combined segregation-linkage analysis.

Previous studies have suggested that angiotensinogen (AGT) gene variants are associated with increased plasma AGT levels, and may also contribute towards the inherited component of predisposition to essential hypertension in humans. To explore the potential functionality of several AGT polymorphisms and estimate their effects, together with other sources of familial correlations, on plasma AGT, we undertook a large study involving 545 healthy French volunteers in 130 nuclear families that include 285 offspring. Plasma AGT levels were measured in all participants, and bi-allelic AGT variants were analysed as candidate functional variants at three sites in the 5'-flanking region (C-532T, A-20C, G-6A), two sites in exon 2 (M235T, T174M) and two newly identified variant sites in the untranslated sequence of exon 5 and the 3'-flanking region (C+2054A, C+2127T) of the gene. Analysis with the class D regressive model showed significant effects influencing plasma AGT levels of all AGT polymorphisms tested, with the exception of T174M. The most significant result was found at C-532T (P=0.000001), which accounts for 4.3% of total plasma AGT variability in parents and 5.5% in offspring, with substantial residual familial correlations. Maximum likelihood estimates of haplotype frequencies and tests of linkage disequilibrium between each AGT polymorphism and a putative QTL are in agreement with a complete confounding of C-532T with the QTL, when taking into account sex and generation specific effects of the QTL. However, further combined segregation-linkage analyses showed significant evidence for additional effects of G-6A, M235T and C+2054A polymorphisms after accounting for C-532T, which supports a complex model with at least two functional variants within the AGT gene controlling AGT levels.

Genetics of the apnea hypopnea index in Caucasians and African Americans: I. Segregation analysis.

Differences in age of presentation and anatomic risk factors for obstructive sleep apnea (OSA) in Caucasians and African Americans suggest possible racial differences in the genetic underpinnings of the disorder. In this study, we assess transmission patterns in a Caucasian sample consisting of 177 families (N = 1,195) and in an African American sample consisting of 125 families (N = 720) for two variables: 1) apnea hypopnea index (AHI) log transformed and adjusted for age, and 2) AHI log transformed and adjusted for age and body mass index (BMI). We allowed for residual familial correlations and sex-specific means in all models. Analysis of the Caucasian sample showed transmission patterns consistent with that of a major gene that were stronger in the age-adjusted variable than in the age- and BMI-adjusted variable. However, in the African American families, adjusting for BMI in addition to age gave stronger evidence for segregation of a codominant gene with an allele frequency of 0.14, accounting for 35% of the total variance. These results provide support for an underlying genetic basis for OSA that in African Americans is independent of the contribution of BMI.

Association and aggregation analysis using kin-cohort designs with applications to genotype and family history data from the Washington Ashkenazi Study.

When a rare inherited mutation in a disease gene, such as BRCA1, is found through extensive study of high-risk families, it is critical to estimate not only age-specific penetrance of the disease associated with the mutation, but also the residual effect of family history once the mutation is taken into account. The kin-cohort design, a cross-sectional survey of a suitable population that collects DNA and family history data, provides an efficient alternative to cohort or case-control designs for estimating age-specific penetrance in a population not selected because of high familial risk. In this report, we develop a method for analyzing kin-cohort data that simultaneously estimate the age-specific cumulative risk of the disease among the carriers and non-carriers of the mutations and the gene-adjusted residual familial aggregation or correlation of the disease. We employ a semiparametric modeling approach, where the marginal cumulative risks corresponding to the carriers and non-carriers are treated non-parametrically and the residual familial aggregation is described parametrically by a class of bivariate failure time models known as copula models. A simple and robust two-stage method is developed for estimation. We apply the method to data from the Washington Ashkenazi Study [Struewing et al., 1997, N Engl J Med 336:1401-1408] to study the residual effect of family history on the risk of breast cancer among non-carriers and carriers of specific BRCA1/BRCA2 germline mutations. We find that positive history of a single first-degree relative significantly increases risk of the non-carriers (RR = 2.0, 95% CI = 1.6-2.6) but has little or no effect on the carriers.

A marginal likelihood approach for estimating penetrance from kin-cohort designs.

The kin-cohort design is a promising alternative to traditional cohort or case-control designs for estimating penetrance of an identified rare autosomal mutation. In this design, a suitably selected sample of participants provides genotype and detailed family history information on the disease of interest. To estimate penetrance of the mutation, we consider a marginal likelihood approach that is computationally simple to implement, more flexible than the original analytic approach proposed by Wacholder et al. (1998, American Journal of Epidemiology 148, 623-629), and more robust than the likelihood approach considered by Gail et al. (1999, Genetic Epidemiology 16, 15-39) to presence of residual familial correlation. We study the trade-off between robustness and efficiency using simulation experiments. The method is illustrated by analysis of the data from the Washington Ashkenazi Study.

Effects of the ApoE Polymorphism on Plasma Lipoproteins in Mexican Americans

PURPOSE: To evaluate the impact of apolipoprotein E (apoE) genotypes on lipoprotein measurements relative to that of other known cardiovascular risk factors in participants of a large population-based family study. METHODS: We measured concentrations of apoE, the major constituents of HDL (cholesterol, apoAI), LDL-C (cholesterol and apoB), and fraction of apoE in lipoprotein size classes in 859 participants of the San Antonio Family Heart Study, and then tested the association between the three common apoE genotypes ( ϵ2ϵ3, ϵ3ϵ3, and ϵ3ϵ4) and lipoprotein traits using the measured genotype approach to account for residual familial correlations. RESULTS: Allele frequencies in this population for ϵ2, ϵ3, and ϵ4 were 3.5%, 89.6%, and 6.9%, respectively. As expected, adjusted apoE concentrations were highest in those with ϵ2ϵ3, intermediate in those with ϵ3ϵ3, and lowest in those with ϵ3ϵ4. The concentrations of total cholesterol, LDL-C and apoB were lowest in those with ϵ2ϵ3, intermediate in those with ϵ3ϵ3, and highest in those with ϵ3ϵ4. There was no significant effect of apoE genotypes on triglycerides, HDL-C, or apoAI levels. Compared to subjects with ϵ3ϵ4, subjects with ϵ2ϵ3 had relatively less apoE in LDL and HDL 1, and relatively more in HDL 2 and HDL 3 size fractions. The effect of apoE genotypes was significantly greater on apoB in women than in men sex × apoE interaction p- value = 0.02 . ApoE genotypes accounted for 4.5%, 12.3%, and 4.7% of the total genetic variation in apoB, apoE, and LDL-C, respectively. CONCLUSION: ApoE genotypes account for a modest, albeit significant, proportion of phenotypic variation in concentrations of LDL-C, apoB, and apoE, and distributions of apoE among lipoproteins in this population; these genotypes have a greater effect on apoB levels in women than in men.

Testing Linkage Disequilibrium in Sibships

We describe the use of multivariate regression for testing allelic association in the presence of linkage, using marker genotype data from sibships. The test is valid, provided that the correct mean structure is modeled but does not require the correlation structure within families to be specified. The test can be implemented using standard statistical software such as the SAS programming language. In a simulation study, we evaluated this new test in comparison with one from a standard, matched-case-control analysis. First, we noted that the genetic effect needed to be quite extreme before residual familial correlation due to linkage led to false inference using the standard, matched-pair analysis. Second, we showed that under examples of extreme residual familial correlation, the new test had the correct test size. Third, we found that the test was more powerful than the sibship disequilibrium test of Horvath and Laird. Finally, we concluded that although the standard analysis may lead to correct inference for practical purposes, the new test is valid, even under extreme residual familial correlation and with no cost in power at the causal locus.

A genetic analysis of smoking behavior in family members of older adult males.

To conduct a genetic study of smoking behavior in 493 three-generation families. Complex segregation analysis and maximum likelihood statistics were used to describe the familial clustering of ever-smoking under several transmission models. The Western Collaborative Group Study, an ageing and health study currently in its 39th year of follow-up. Probands were male participants who were of mean age 71.6 years at the time of the family history interview in 1986-88. Data were collected via an interview that focused on the family smoking history of participants. Smoking histories of all first-degree relatives were obtained from probands. Evidence for genetic transmission was indicated by rejection of both the environmental and sporadic models in favor of a Mendelian genetic model with residual familial effects from spouses and both parents. The best-fitting model was that of a dominant major gene with low estimated frequency and residual familial correlations. This is the first study to date to model the familial transmission of ever-smoking in three-generation families.

Evidence of major gene control of cortical bone loss in humans.

Cortical index (CI) is the ratio of the combined cortical thickness to the total diameter of the bone. It serves for the assessment of the geometric properties of bone and for indirect evaluation of bone mass. CI is a useful predictor of osteoporosis. The aim of the present study was to test the hypothesis of major gene control of CI variation in a large sample of pedigrees from Chuvashia, Russia. Complex segregation analysis revealed that the major gene model of CI inheritance is the best fitting and most parsimonious for the present data. Parameters of the genotype-gender specific dependence of CI variation on age were estimated simultaneously with other parameters in the segregation analysis. The results of analysis showed that not only the baseline level of CI but also the age at onset of the involutive bone changes (inflection point) and the rate of the CI decrease with age (slope coefficient) are under control of the same major gene. Non-major gene effects shared by pedigree members (residual familial correlations) were found to be statistically insignificant. Approximately 73% of inter-individual variation in CI was attributable to the effects explicitly included in the model.

Evidence for major genes influencing pulmonary function in the NHLBI family heart study.

Segregation analysis was performed on the pulmonary measures forced expiratory volume in one second (FEV1), forced vital capacity (FVC), and the ratio of FEV1/FVC in 455 randomly ascertained families from the NHLBI Family Heart Study (FHS). Gender specific standardized residuals were used as the phenotypic variable in both familial correlation and segregation analyses. These residuals represented adjustments for the effects of age, age(2), age(3), Body Mass Index (BMI, kg/m(2)), height, the ratio of waist to hip measurements (WHR), the presence of coronary heart disease, smoking history, and pack years for current smokers. Sibling correlations were not different from parent-offspring correlations for all three traits, and heritability estimates for FEV1, FVC, and the FEV1/FVC ratio were 0. 515, 0.540, and 0.449, respectively. Segregation analysis of FEV1, a trait that measures airflow, indicated that a dominant major gene best fits the data, although a residual familial correlation supports the presence of an additional polygenic or common environmental component. For FVC, a trait that measures lung volume, alternative models could not be statistically differentiated, but the transmission probabilities do not support a Mendelian major gene. The best model for FEV1/FVC ratio is a non-Mendelian codominant model, perhaps due to the mixing of the individual underlying distributions influencing airflow and lung volume. These results support the hypothesis that complex relationships exist for lung function traits and that multiple genes and environmental factors influence lung function.

A frailty approach for modelling diseases with variable age of onset in families: the NHLBI Family Heart Study.

We use frailty models to analyse the effect of latent genetic and environmental risk factors on hazard functions in nuclear families. The approach expresses latent risk factors (frailties) as functions of the effects of a single major gene and shared familial risk. The latter may result from shared polygenes and/or a common environment. Genetic frailties are modelled using a two-point distribution, and residual frailties (shared environment, polygenes) using a gamma distribution. The two-point distribution follows the laws of Mendelian transmission, under either dominant or recessive gene action. We describe a robust EM approach for the joint estimation of the magnitude of genetic, covariate, gene by covariate interaction effects while allowing residual familial correlation. We illustrate the method on coronary heart disease data from the National Heart, Lung, and Blood Institute Family Heart Study. In addition, a simulation study shows that ignoring possible residual correlation in disease status due to a shared familial environment leads to an overestimate of the relative risk associated with a latent genotype.

Measured haplotype analysis of the angiotensin-I converting enzyme gene.

Linkage and segregation analysis have shown that circulating angiotensin-I converting enzyme (ACE) levels are influenced by a major quantitative trait locus that maps within or close to the ACE gene. The D variant of a 287 bp insertion/deletion (I/D) polymorphism in intron 16 of the gene is associated with high ACE levels and may also be related to increased risk of cardiovascular disease. Multiple variants that are in linkage disequilibrium with the I/D polymorphism have been described, but it is unknown if any of these are directly implicated, alone or in combination with as yet undiscovered variants, in the determination of ACE levels. An analysis of 10 polymorphisms spanning 26 kb of the ACE gene revealed a limited number of haplotypes in Caucasian British families due to strong linkage disequilibrium operating over this small chromosomal region. A haplotype tree (cladogram) was constructed with three main branches (clades A-C) which account for 90% of the observed haplotypes. Clade C is most likely derived from clades A and B following an ancestral recombination event. This evolutionary information was then used to direct a series of nested, measured haplotype analyses that excluded upstream sequences, including the ACE promoter, from harbouring the major ACE-linked variant that explains 36% of the total trait variability. Residual familial correlations were highly significant, suggesting the influence of additional unlinked genes. Our results demonstrate that a combined cladistic/measured haplotype analysis of polymorphisms within a gene provides a powerful means to localize variants that directly influence a quantitative trait.

Segregation analysis of pulmonary function among families in the Framingham Study.

Familial aggregation of cross-sectional pulmonary function was examined in 5,003 subjects from 1,408 families participating in the Framingham Study. Subjects, who were members of either the Original Cohort (recruited from 1948 to 1952) or the Offspring Cohort (recruited from 1971 to 1974), underwent spirometry at a mean age of 53 yr. The effects of age, height, weight, and smoking status on FEV1 were evaluated through linear-regression analysis, with separate models for men and women in each cohort. The gender- and cohort-specific standardized residual FEV1 from these models was used as the phenotypic variable in familial correlation and segregation analyses to assess inheritance patterns. In models that assumed no major gene determining FEV1, correlation of pulmonary function was greater for mothers and offspring than for fathers and offspring (p[mo] = 0.190, p[fo] = 0.112; p = 0.06), and sibling correlation exceeded parent-offspring correlation (p[sib] = 0.225; p < 0.01). By comparison with a general model, in which transmission probabilities and residual familial correlations are arbitrary, models that imposed a Mendelian gene were rejected (p < 0.001). A model with no parent-offspring transmission of a major factor, but with residual familial correlation, provided as good a fit as the general model, suggesting that environmental and/or polygenic genetic influences determine FEV1.

A multivariate logistic model (MLM) for analyzing binary family data.

We consider modeling the familial correlation between 2 related individuals using a multiple logistic regressive model. It is shown that there is a discrepancy in the marginal probability of the second individual. We investigate the conditions under which this discrepancy can be minimized and show how it can have a direct effect on handling missing values and ascertainment. We derive a functional relationship between the parameters in the model that eliminates this discrepancy, hence solving the problems that can arise in the handling of missing values and ascertainment. Because this methodology fails when there are more than 2 related individuals, we present a new model based on a multivariate logistic distribution. Residual familial correlations can be directly related to the parameters of this model. The likelihood for family data under this model is independent of the order in which the family members enter the calculation. The marginal probabilities can be easily computed.

Modeling age of onset and residual familial correlations for the linkage analysis of bipolar disorder

Modeling age of onset and residual familial correlations for the linkage analysis of bipolar disorder

Detection of a recessive major gene for high IgE levels acting independently of specific response to allergens.

The genetic control of the total IgE, the immunoglobulins E involved in allergy, remains still unclear. Although high IgE levels were found to be determined by a recessive major gene in several studies, other modes of inheritance were also reported. Moreover, at least two different genetic mechanisms controlling the IgE regulation have been suggested: one involved in the specific IgE response and the other one in the nonspecific response. To better understand the genetic mechanisms controlling IgE variation, we performed segregation analysis of IgE levels by ignoring or taking into account the specific response to allergens (SRA). Analyses were conducted using the class D regressive model, in a sample of 234 Australian nuclear families randomly selected during the winter months, when IgE levels are the lowest (basal). SRA, when included as a covariate in the model, was defined by one of the three following criteria: (1) raised specific IgE level for one or more allergens, (2) positive skin test for one or more allergens, and (3) at least one of the (1) or (2) criteria. When the presence of SRA is ignored, the familial transmission of total IgE level is compatible with the segregation of a recessive major gene and residual familial correlations. When the presence of SRA is accounted for in the analysis, whether defined by criteria (1), (2), or (3), there is still evidence for a recessive major gene controlling IgE levels but residual familial correlations are no longer significant. In addition, no interaction between this major gene and SRA is shown here. Our results suggest that this gene, which accounts for 28% of the variation of the trait, may be involved in the control of basal IgE production, independently of specific response to allergens.