Detecting specific genotype by environment interactions using marginal maximum likelihood estimation in the classical twin design

Environment-by-PGS Interaction in the Classical Twin Design: An Application to Childhood Anxiety and Negative Affect

One type of genotype-environment interaction occurs when genetic effects on a phenotype are moderated by an environment; or when environmental effects on a phenotype are moderated by genes. Here we outline these types of genotype-environment interaction models, and propose a test of genotype-environment interaction based on the classical twin design, which includes observed genetic variables (polygenic scores: PGSs) that account for part of the genetic variance of the phenotype. We introduce environment-by-PGS interaction and the results of a simulation study to address statistical power and parameter recovery. Next, we apply the model to empirical data on anxiety and negative affect in children. The power to detect environment-by-PGS interaction depends on the heritability of the phenotype, and the strength of the PGS. The simulation results indicate that under realistic conditions of sample size, heritability and strength of the interaction, the environment-by-PGS model is a viable approach to detect genotype-environment interaction. In 7-year-old children, we defined two PGS based on the largest genetic association studies for 2 traits that are genetically correlated to childhood anxiety and negative affect, namely major depression (MDD) and intelligence (IQ). We find that common environmental influences on negative affect are amplified for children with a lower IQ-PGS.

Bias in Gene-by-Environment Interaction Effects with Sum Scores; An Application to Well-being Phenotypes

In the current study, we investigated the influence of using skewed sum scores on estimated gene-by-environment interaction effects (GxE) for life satisfaction and happiness with perceived social support. To this end, we analyzed item-level data from a large adult twin sample (Ns between 3610 and 11,305) of the Netherlands Twin Register. Item response theory (IRT) models were incorporated in unmeasured (univariate) GxE models, and measured GxE models (with social support as moderator). We found that skewness introduced spurious GxE effects, with the largest effect for the most skewed variable (social support). Finally, in the IRT model for life satisfaction, but not for happiness, heritability estimates decreased with higher social support, while this was not observed when analyzing sum scores. Together, our results indicate that IRT can be used to address psychometric issues related to the use of sum scores, especially in the context of GxE, for complex traits like well-being.

Maximizing the value of twin studies in health and behaviour

In the classical twin design, researchers compare trait resemblance in cohorts of identical and non-identical twins to understand how genetic and environmental factors correlate with resemblance in behaviour and other phenotypes. The twin design is also a valuable tool for studying causality, intergenerational transmission, and gene-environment correlation and interaction. Here we review recent developments in twin studies, recent results from twin studies of new phenotypes and recent insights into twinning. We ask whether the results of existing twin studies are representative of the general population and of global diversity, and we conclude that stronger efforts to increase representativeness are needed. We provide an updated overview of twin concordance and discordance for major diseases and mental disorders, which conveys a crucial message: genetic influences are not as deterministic as many believe. This has important implications for public understanding of genetic risk prediction tools, as the accuracy of genetic predictions can never exceed identical twin concordance rates.

Genetic and Environmental Variation in Continuous Phenotypes in the ABCD Study®

Twin studies yield valuable insights into the sources of variation, covariation and causation in human traits. The ABCD Study® (abcdstudy.org) was designed to take advantage of four universities known for their twin research, neuroimaging, population-based sampling, and expertise in genetic epidemiology so that representative twin studies could be performed. In this paper we use the twin data to: (i) provide initial estimates of heritability for the wide range of phenotypes assessed in the ABCD Study using a consistent direct variance estimation approach, assuring that both data and methodology are sound; and (ii) provide an online resource for researchers that can serve as a reference point for future behavior genetic studies of this publicly available dataset. Data were analyzed from 772 pairs of twins aged 9-10 years at study inception, with zygosity determined using genotypic data, recruited and assessed at four twin hub sites. The online tool provides twin correlations and both standardized and unstandardized estimates of additive genetic, and environmental variation for 14,500 continuously distributed phenotypic features, including: structural and functional neuroimaging, neurocognition, personality, psychopathology, substance use propensity, physical, and environmental trait variables. The estimates were obtained using an unconstrained variance approach, so they can be incorporated directly into meta-analyses without upwardly biasing aggregate estimates. The results indicated broad consistency with prior literature where available and provided novel estimates for phenotypes without prior twin studies or those assessed at different ages. Effects of site, self-identified race/ethnicity, age and sex were statistically controlled. Results from genetic modeling of all 53,172 continuous variables, including 38,672 functional MRI variables, will be accessible via the user-friendly open-access web interface we have established, and will be updated as new data are released from the ABCD Study. This paper provides an overview of the initial results from the twin study embedded within the ABCD Study, an introduction to the primary research domains in the ABCD study and twin methodology, and an evaluation of the initial findings with a focus on data quality and suitability for future behavior genetic studies using the ABCD dataset. The broad introductory material is provided in recognition of the multidisciplinary appeal of the ABCD Study. While this paper focuses on univariate analyses, we emphasize the opportunities for multivariate, developmental and causal analyses, as well as those evaluating heterogeneity by key moderators such as sex, demographic factors and genetic background.

Overview of CAPICE-Childhood and Adolescence Psychopathology: unravelling the complex etiology by a large Interdisciplinary Collaboration in Europe-an EU Marie Skłodowska-Curie International Training Network

The Roadmap for Mental Health and Wellbeing Research in Europe (ROAMER) identified child and adolescent mental illness as a priority area for research. CAPICE (Childhood and Adolescence Psychopathology: unravelling the complex etiology by a large Interdisciplinary Collaboration in Europe) is a European Union (EU) funded training network aimed at investigating the causes of individual differences in common childhood and adolescent psychopathology, especially depression, anxiety, and attention deficit hyperactivity disorder. CAPICE brings together eight birth and childhood cohorts as well as other cohorts from the EArly Genetics and Life course Epidemiology (EAGLE) consortium, including twin cohorts, with unique longitudinal data on environmental exposures and mental health problems, and genetic data on participants. Here we describe the objectives, summarize the methodological approaches and initial results, and present the dissemination strategy of the CAPICE network. Besides identifying genetic and epigenetic variants associated with these phenotypes, analyses have been performed to shed light on the role of genetic factors and the interplay with the environment in influencing the persistence of symptoms across the lifespan. Data harmonization and building an advanced data catalogue are also part of the work plan. Findings will be disseminated to non-academic parties, in close collaboration with the Global Alliance of Mental Illness Advocacy Networks-Europe (GAMIAN-Europe).

Incorporating Polygenic Risk Scores in the ACE Twin Model to Estimate A-C Covariance

The assumption in the twin model that genotypic and environmental variables are uncorrelated is primarily made to ensure parameter identification, not because researchers necessarily think that these variables are uncorrelated. Although the biasing effects of such correlations are well understood, a method to estimate these parameters in the twin model would be useful. Here we explore the possibility of relaxing this assumption by adding polygenic scores to the (univariate) twin model. We demonstrate that this extension renders the additive genetic (A)—common environmental (C) covariance (σ_AC) identified. We study the statistical power to reject σ_AC = 0 in the ACE model and present the results of simulations.

Dependence of Gene-by-Environment Interactions (GxE) on Scaling: Comparing the Use of Sum Scores, Transformed Sum Scores and IRT Scores for the Phenotype in Tests of GxE

Estimates of gene-environment interactions (GxE) in behavior genetic models depend on how a phenotype is scaled. Inappropriately scaled phenotypes result in biased estimates of GxE and can sometimes even suggest GxE in the direction opposite to its true direction. Previously proposed solutions are mathematically complex, computationally demanding and may prove impractical for the substantive researcher. We, therefore, evaluated two simple-to-use alternatives: (1) straightforward non-linear transformation of sum scores and (2) factor scores from an appropriate item response theory (IRT) model. Within Purcell's (2002) GxM framework, both alternatives provided less biased parameter estimates, and improved false and true positive rates than using a raw sum score. These approaches are, therefore, recommended over using raw sum scores in tests of GxE. Circumstances under which IRT factor scores versus transformed sum scores should be preferred are discussed.

Genes, Culture and Conservatism-A Psychometric-Genetic Approach

The Wilson−Patterson conservatism scale was psychometrically evaluated using homogeneity analysis and item response theory models. Results showed that this scale actually measures two different aspects in people: on the one hand people vary in their agreement with either conservative or liberal catch-phrases and on the other hand people vary in their use of the “?” response category of the scale. A 9-item subscale was constructed, consisting of items that seemed to measure liberalism, and this subscale was subsequently used in a biometric analysis including genotype–environment interaction, correcting for non-homogeneous measurement error. Biometric results showed significant genetic and shared environmental influences, and significant genotype–environment interaction effects, suggesting that individuals with a genetic predisposition for conservatism show more non-shared variance but less shared variance than individuals with a genetic predisposition for liberalism.

Analysis of Behavioral and Emotional Problems in Children Highlights the Role of Genotype × Environment Interaction

This study tested for Genotype × Environment (G × E) interaction on behavioral and emotional problems in children using new methods that do not require identification of candidate genes or environments, can distinguish between interaction with shared and unique environment, and are insensitive to scale effects. Parental ratings of problem behavior from 14,755 twin pairs (5.3 years, SD = 0.22) indicated G × E interaction on emotional liability, social isolation, aggression, attention problems, dependency, anxiety, and physical coordination. Environmental influences increased in children who were genetically more predisposed to problem behavior, with ~20% of the variance due to G × E interaction (8% for anxiety to 37% for attention problems). Ignoring G × E interaction does not greatly bias heritability estimates, but it does offer a comprehensive model of the etiology for childhood problems.

Evidence for Gender-Dependent Genotype by Environment Interaction in Adult Depression

Depression in adults is heritable with about 40 % of the phenotypic variance due to additive genetic effects and the remaining phenotypic variance due to unique (unshared) environmental effects. Common environmental effects shared by family members are rarely found in adults. One possible explanation for this finding is that there is an interaction between genes and the environment which may mask effects of the common environment. To test this hypothesis, we investigated genotype by environment interaction in a large sample of female and male adult twins aged 18-70 years. The anxious depression subscale of the Adult Self Report from the Achenbach System of Empirically Based Assessment (Achenbach and Rescorla in Manual for the ASEBA adult: forms and profiles, 2003) was completed by 13,022 twins who participate in longitudinal studies of the Netherlands Twin Register. In a single group analysis, we found genotype by unique environment interaction, but no genotype by common environment interaction. However, when conditioning on gender, we observed genotype by common environment interaction in men, with larger common environmental variance in men who are genetically less at risk to develop depression. Although the effect size of the interaction is characterized by large uncertainty, the results show that there is at least some variance due to the common environment in adult depression in men.

Heteroscedastic Latent Trait Models for Dichotomous Data

Effort has been devoted to account for heteroscedasticity with respect to observed or latent moderator variables in item or test scores. For instance, in the multi-group generalized linear latent trait model, it could be tested whether the observed (polychoric) covariance matrix differs across the levels of an observed moderator variable. In the case that heteroscedasticity arises across the latent trait itself, existing models commonly distinguish between heteroscedastic residuals and a skewed trait distribution. These models have valuable applications in intelligence, personality and psychopathology research. However, existing approaches are only limited to continuous and polytomous data, while dichotomous data are common in intelligence and psychopathology research. Therefore, in present paper, a heteroscedastic latent trait model is presented for dichotomous data. The model is studied in a simulation study, and applied to data pertaining alcohol use and cognitive ability.

Assessing genotype by environment interaction in case of heterogeneous measurement error

Considerable effort has been devoted to establish genotype by environment interaction (G x E) in case of unmeasured genetic and environmental influences. Although it has been outlined by various authors that the appearance of G x E can be dependent on properties of the given measurement scale, a non-biased method to assess G x E is still lacking. We show that the incorporation of an explicit measurement model can remedy potential bias due to ceiling and floor effects. By means of a simulation study it is shown that the use of sum scores can lead to biased estimates whereas the proposed method is unbiased. The power of the suggested method is illustrated by means of a second simulation study with different sample sizes and G x E effect sizes.

Testing systematic genotype by environment interactions using item level data

Investigating genotype by environment interactions (GxE) is generally considered challenging due to the scale dependency of the interaction effect. The present paper illustrates the problems associated with testing for GxEs on summed item scores within the well-known ACE model. That is, it is shown how genuine GxEs may be masked and how spurious interactions can arise from scaling issues in the data. A solution is proposed which explicitly distinguishes between a measurement model for the ordinal item responses and a biometric model in which the GxE effects are investigated. The new approach is studied in a simulation study using both a scenario in which the measurement instrument suffers from mild scaling problems and a scenario in which the measurement instrument suffers from severe scaling problems. Results indicate that the severity of the scale problems affects the power to detect GxE, but it rarely results in false positives. We illustrate the new approach on a real dataset concerning affect.

Latent classiness and other mixtures

The aim of this article is to laud Lindon Eaves' role in the development of mixture modeling in genetic studies. The specification of models for mixture distributions was very much in its infancy when Professor Eaves implemented it in his own FORTRAN programs, and extended it to data collected from relatives such as twins. It was his collaboration with the author of this article which led to the first implementation of mixture distribution modeling in a general-purpose structural equation modeling program, Mx, resulting in a 1996 article on linkage analysis in Behavior Genetics. Today, the popularity of these methods continues to grow, encompassing methods for genetic association, latent class analysis, growth curve mixture modeling, factor mixture modeling, regime switching, marginal maximum likelihood, genotype by environment interaction, variance component twin modeling in the absence of zygosity information, and many others. This primarily historical article concludes with some consideration of some possible future developments.

Genotype by environment interactions in cognitive ability: a survey of 14 studies from four countries covering four age groups

A large part of the variation in cognitive ability is known to be due to genetic factors. Researchers have tried to identify modifiers that influence the heritability of cognitive ability, indicating a genotype by environment interaction (G×E). To date, such modifiers include measured variables like income and socioeconomic status. The present paper focuses on G×E in cognitive ability where the environmental variable is an unmeasured environmental factor that is uncorrelated in family members. We examined this type of G×E in the GHCA-database (Haworth et al., Behav Genet 39:359-370, 2009), which comprises data of 14 different cognition studies from four different countries including participants of different ages. Results indicate that for younger participants (4-13 years), the strength of E decreases across the additive genetic factor A, but that this effect reverts for older participants (17-34 years). However, a clear and general conclusion about the presence of a genuine G×E is hampered by differences between the individual studies with respect to environmental and genetic influences on cognitive ability.

The Heteroscedastic Graded Response Model with a Skewed Latent Trait: Testing Statistical and Substantive Hypotheses Related to Skewed Item Category Functions

The Graded Response Model (GRM; Samejima, Estimation of ability using a response pattern of graded scores, Psychometric Monograph No. 17, Richmond, VA: The Psychometric Society, 1969) can be derived by assuming a linear regression of a continuous variable, Z, on the trait, θ, to underlie the ordinal item scores (Takane & de Leeuw in Psychometrika, 52:393-408, 1987). Traditionally, a normal distribution is specified for Z implying homoscedastic error variances and a normally distributed θ. In this paper, we present the Heteroscedastic GRM with Skewed Latent Trait, which extends the traditional GRM by incorporation of heteroscedastic error variances and a skew-normal latent trait. An appealing property of the extended GRM is that it includes the traditional GRM as a special case. This enables specific tests on the normality assumption of Z. We show how violations of normality in Z can lead to asymmetrical category response functions. The ability to test this normality assumption is beneficial from both a statistical and substantive perspective. In a simulation study, we show the viability of the model and investigate the specificity of the effects. We apply the model to a dataset on affect and a dataset on alexithymia.