Familial Resemblance and Heritability

https://doi.org/10.1016/S0065-2660(07)00402-6Get rights and content

Abstract

Familial resemblance arises when family members are more similar than unrelated pairs of individuals, and may be estimated in terms of correlations or covariances among family members. Multifactorial heritability (or generalized heritability) quantifies the strength of the familial resemblance and represents the percentage of variance that is due to all additive familial effects including additive genetic and those of the familial environment. However, the traditional concept of heritability, which may be more appropriately called the genetic heritability, represents only the percentage of phenotypic variance due to additive genetic effects. Resolving the various sources of familial resemblance entails other issues. For example, there may be major gene effects that are largely or entirely nonadditive, temporal or developmental trends, and gene–gene (epistasis) and gene–environment interactions. The design of a family study determines which of these sources are resolvable. For example, in nuclear families consisting of parents and offspring, the genetic and familial environmental effects are not resolvable because these relatives share both genes and environments. However, extended pedigree and twin and adoption designs allow separation of the heritable effects and, possibly, more complex etiologies, including interactions. Various factors affect the estimation and interpretability of heritabilities, for example, assumptions regarding linearity and additivity, assortative mating, and the underlying distribution of the data. Nonnormality of the data can lead to errors in hypothesis testing, although it yields reasonably unbiased estimates. Fortunately, these and other complications can be directly modeled in many of the sophisticated software packages available today in genetic epidemiology.

Introduction

One of the earliest statistical concepts, correlations, was conceived and advanced by Karl Pearson and Francis Galton almost simultaneously with the idea of the quantifying genetic resemblance in close relatives (see time line in Table 2.1). These two ideas fed off and complemented one another even before the term heritability was coined, as Fisher recounted in his seminal 1925 work:

One of the earliest and most striking successes of the method of correlation was in the biometrical study of inheritance. At a time when nothing was known of the mechanism of inheritance, or of the structure of the germinal material, it was possible by this method to demonstrate the existence of inheritance, and to “measure its intensity;” and this in an organism in which experimental breeding could not be practiced, namely, Man. (Fisher, 1925, p. 175)

Although the fundamental concepts of familial resemblance and heritability reviewed in this chapter were created for an earlier era of genetics when it was not possible to directly measure genes in humans, these ideas remain at the cornerstone of genetic epidemiology even today. Indeed, some of the most recent and powerful developments in linkage equilibrium and disequilibrium analyses rely heavily on the estimation and partitioning of heritable components due to different measured sources in the form of variance components models, such as a linked genetic marker and a measured candidate gene. Even approaches that do not rely upon the estimation of heritability per se often find it useful to “translate” effect sizes onto the heritability scale because it is such a convenient concept. Here we review the methods for estimating the magnitudes of unmeasured and measured effects and for testing hypotheses. We also review some of the relevant study designs and the corresponding statistical methods, factors affecting the estimation, and extensions developed to model more complex etiologic factors. This provides a foundation for the extensions that incorporate linkage and association developed in later chapters.

Section snippets

Family resemblance

Familial resemblance arises when relatives who share genes and/or environmental factors exhibit greater phenotypic similarity than do unrelated individuals. The extent of the familial resemblance can be measured by familial correlations (e.g., sibling, parent–offspring, and spouse). In general, biological relatives such as siblings share both genes and familial environments in common. Thus, familial resemblance can be a function of shared genes, shared environments, or both. In contrast, under

Study Designs and Multifactorial Models

The specificity of a given model in terms of what parameters can be estimated depends on the study design or the types of relatives included (e.g., twin, nuclear families, pedigrees, adoptions). Under some designs, the familial components (G and C) are not resolvable and are estimated as a single heritable component that represents all additive effects that are transmitted from parents to offspring. The assumptions underlying these study designs are also critical and are reviewed here.

Discussion

In 1986, Morton (1986) wrote: “Genetic epidemiology is now assimilating the rapid advances in molecular biology that promise a complete linkage map and molecular definition of disease loci…. When they are complete our morning will have passed into the afternoon of molecular epidemiology.”

Our morning of genetic epidemiology has revealed that complex traits generally involve additive and nonadditive interactions among several genes and environmental factors, none of which may entail large

Acknowledgments

This work was supported in part by a grant from the National Institute of General Medical Sciences (GM 28719) and a grant from the National Heart, Lung, and Blood Institute (HL54473), National Institutes of Health.

References (43)

  • L.J. Eaves et al.

    The effects of social homogamy and cultural inheritance on the covariances of twins and their parents: A LISREL model

    Behav. Genet.

    (1989)
  • D.S. Falconer

    “Introduction to Quantitative Genetics.”

    (1960)
  • R.A. Fisher

    “Statistical Methods for Research Worker.”

    (1925)
  • D.E. Goldgar

    Multipoint analysis of human quantitative genetic variation

    Am. J. Hum. Genet.

    (1990)
  • C.L. Hanis et al.

    Nonrandom sampling in human genetics: Familial correlations

    I. M. A. J. Math. Appl. Med. Biol.

    (1984)
  • A.C. Heath et al.

    Resolving the effects of phenotype and social background on mate selection

    Behav. Genet.

    (1985)
  • J.L. Hopper et al.

    Extensions to multivariate normal models for pedigree analysis

    Ann. Hum. Genet.

    (1982)
  • L.W. Konigsberg et al.

    Mixed model segregation analysis of LDL‐c concentration with genotype‐covariate interaction

    Genet. Epidemiol.

    (1991)
  • J.M. Lalouel et al.

    Complex segregation analysis with pointers

    Hum. Hered.

    (1981)
  • K. Lange et al.

    Extensions to pedigree analysis. IV. Covariance components models for multivariate traits

    Am. J. Med. Genet.

    (1983)
  • C.C. Li

    “Population Genetics.”

    (1955)
  • Cited by (39)

    • Familial resemblances in human plasma metabolites are attributable to both genetic and common environmental effects

      2019, Nutrition Research
      Citation Excerpt :

      Thus, the total phenotypic variance of a trait may be expressed as VP = VG + VC + VE. The most commonly used study designs to estimate heritability are nuclear families (parents and their biological offspring) and twins (monozygotic and dizygotic twins) [27]. The variance component method described above can be extended to a bivariate (two traits) genetic analysis.

    • Dissecting the catatonia phenotype in psychotic and mood disorders on the basis of familial-genetic factors

      2018, Schizophrenia Research
      Citation Excerpt :

      Given that this is the first study examining levels of familial aggregation of different definitions of the catatonia phenotype, our results are not readily comparable with any in the literature. However, and considering that our familiality estimates measure the maximal effect of genes (Rice, 2008), some aspects of our results merit comments in relation to previous studies reporting heritability estimates. Overall, those catatonia phenotypes maximizing the familiality estimates (i.e., mutism, rigidity, classical catatonia factor, catatonia severity at the level of moderate or higher, presence of at least 5 catatonia signs, psychosis with prominent catatonia features) do have similar h2 figures than those reported in other family studies for the diagnoses of psychotic and major mood disorders (h2 ~ 60%) (Lichtenstein et al., 2009; Wray and Gottesman, 2012) but lower than those reported in twin studies (h2 ~ 80%) (Cardno et al., 1999; McGuffin et al., 2003).

    • Twin and family studies reveal strong environmental and weaker genetic cues explaining heritability of eosinophilic esophagitis

      2014, Journal of Allergy and Clinical Immunology
      Citation Excerpt :

      A strong genetic basis for EoE is further supported by candidate and genomewide association studies that identified risk variants,9,11-13 as well as EoE-specific gene expression profiles.10 However, estimating heritability from nuclear families has limited interpretation because genes and family environment cannot be distinguished.49,50 Specifically, similar environmental exposures and risk within the common family environment mimic genetic inheritance patterns and confound heritability.

    View all citing articles on Scopus
    View full text