Much progress has been made in twin research since our last special issue on twin registries (Hur, Y.-M., & Craig, J. M. (2013). Twin Research and Human Genetics, 16, 1–12.). This special issue provides an update on the state of twin family registries around the world. This issue includes 61 papers on twin family registries from 25 countries, of which 3 describe consortia based on collaborations of several twin family registries. The articles included in this issue discuss the establishment and maintenance of twin registries, recruitment strategies, methods of zygosity assessment, research aims and major findings from twin family cohorts, as well as other important topics related to twin studies. The papers amount to approximately 1.3 million monozygotic, dizygotic twins and higher order multiples and their family members who participate in twin studies around the world. Nine new twin family registries have been established across the world since our last issue, which demonstrates that twin registers are increasingly important in studies of the determinants and correlates of complex traits from disease susceptibility to healthy development.

Higher health literacy is associated with higher cognitive function and better health. Despite its wide use in medical research, no study has investigated the genetic contributions to health literacy. Using 5783 English Longitudinal Study of Ageing (ELSA) participants (mean age = 65.49, SD = 9.55) who had genotyping data and had completed a health literacy test at wave 2 (2004–2005), we carried out a genome-wide association study (GWAS) of health literacy. We estimated the proportion of variance in health literacy explained by all common single nucleotide polymorphisms (SNPs). Polygenic profile scores were calculated using summary statistics from GWAS of 21 cognitive and health measures. Logistic regression was used to test whether polygenic scores for cognitive and health-related traits were associated with having adequate, compared to limited, health literacy. No SNPs achieved genome-wide significance for association with health literacy. The proportion of variance in health literacy accounted for by common SNPs was 8.5% (SE = 7.2%). Greater odds of having adequate health literacy were associated with a 1 standard deviation higher polygenic score for general cognitive ability [OR = 1.34, 95% CI (1.26, 1.42)], verbal-numerical reasoning [OR = 1.30, 95% CI (1.23, 1.39)], and years of schooling [OR = 1.29, 95% CI (1.21, 1.36)]. Reduced odds of having adequate health literacy were associated with higher polygenic profiles for poorer self-rated health [OR = 0.92, 95% CI (0.87, 0.98)] and schizophrenia [OR = 0.91, 95% CI (0.85, 0.96)). The well-documented associations between health literacy, cognitive function and health may partly be due to shared genetic etiology. Larger studies are required to obtain accurate estimates of SNP-based heritability and to discover specific health literacy-associated genetic variants.

Type 2 diabetes (T2D) is a chronic disease that disproportionately affects Indigenous Australians. We have previously reported the localization of a novel T2D locus by linkage analysis to chromosome 2q24 in a large admixed Indigenous Australian pedigree (Busfield et al. (2002). American Journal of Human Genetics, 70, 349–357). Here we describe fine mapping of this region in this pedigree, with the identification of SNPs showing strong association with T2D: rs3845724 (diabetes p = 7 × 10−4), rs4668106 (diabetes p = 9 × 10−4) and rs529002 (plasma glucose p = 3 × 10−4). These associations were successfully replicated in an independent collection of Indigenous Australian T2D cases and controls. These SNPs all lie within the gene encoding ceramide synthase 6 (CERS6) and thus may regulate ceramide synthesis.

Twin registries have developed as a valuable resource for the study of many aspects of disease and society over the years in many different countries. A number of these registries include large numbers of twins with data collected at varying information levels for twin cohorts over the past several decades. More recent expansion of twin datasets has allowed for the collection of genetic data, together with many other levels of ‘omic’ information along with multiple demographic, physiological, health outcomes and other measures typically used in epidemiologic research. Other twin data sources outside these registries reflect research interests in particular aspects of disease or specific phenotypic assessment. Twin registries have the potential to play a key role in many aspects of the artificial intelligence/machine learning-driven projects of the future and will continue to keep adapting to the changing research landscape.

We recently reported an association of offspring educational attainment with polygenic risk scores (PRS) computed on parent’s non-transmitted alleles for educational attainment using the second GWAS meta-analysis article on educational attainment published by the Social Science Genetic Association Consortium. Here we test the replication of these findings using a more powerful PRS from the third GWAS meta-analysis article by the Consortium. Each of the key findings of our previous paper is replicated using this improved PRS (N = 2335 adolescent twins and their genotyped parents). The association of children’s attainment with their own PRS increased substantially with the standardized effect size, moving from β = 0.134, 95% CI = 0.079, 0.188 for EA2, to β = 0.223, 95% CI = 0.169, 0.278, p < .001, for EA3. Parent’s PRS again predicted the socioeconomic status (SES) they provided to their offspring and increased from β = 0.201, 95% CI = 0.147, 0.256 to β = 0.286, 95% CI = 0.239, 0.333. Importantly, the PRS for alleles not transmitted to their offspring — therefore acting via the parenting environment — was increased in effect size from β = 0.058, 95% CI = 0.003, 0.114 to β = 0.067, 95% CI = 0.012, 0.122, p = .016. As previously found, this non-transmitted genetic effect was fully accounted for by parental SES. The findings reinforce the conclusion that genetic effects of parenting are substantial, explain approximately one-third the magnitude of an individual’s own genetic inheritance and are mediated by parental socioeconomic competence.

A high prevalence of asthma has been documented among the inhabitants of Tristan da Cunha, an isolated island in the South Atlantic. The population derives from just 28 founders. We performed lung function testing, including methacholine inhalation challenge, allergen skin prick testing, and collected DNA from essentially all of the current island population (269 individuals), and genotyped a panel of 43 single-nucleotide polymorphisms (SNPs) reported as associated with asthma and atopy. We carried out a mixed-model association analysis using the known pedigree. There were 96 individuals diagnosed as asthmatic (36%), and heritability estimates were similar to those from nonisolated population samples (multifactorial threshold model, h2 = 48%). The first component from a genetic principal components analysis using the entire SNP panel was nonlinearly associated with asthma, with the maximum risk to those intermediate to reference (Human Genome Diversity Project) European and African samples means. The single most strongly associated SNP was rs2786098 (p = 5.5 × 10−5), known to regulate the gene DENND1B. This explained approximately one-third of the trait heritability, with an allelic odds ratio for the A allele of 2.6. Among A/A carriers, 10 out of 12 individuals were asthmatic. The rs2786098*A variant was initially reported to decrease the risk of childhood (atopic) asthma in European but slightly increase the risk in African-descended populations, and does significantly alter Th2 cell function. Despite an absence of overall association with this variant in recent asthma genome wide association studies meta-analyses, an effect may exist on the particular genetic background of the Tristan da Cunha population.

Large multigenerational cohort studies offer powerful ways to study the hereditary effects on various health outcomes. However, accounting for complex kinship relations in big data structures can be methodologically challenging. The traditional kinship model is computationally infeasible when considering thousands of individuals. In this article, we propose a computationally efficient alternative that employs fractional relatedness of family members through a series of founding members. The primary goal of this study is to investigate whether the effect of determinants on health outcome variables differs with and without accounting for family structure. We compare a fixed-effects model without familial effects with several variance components models that account for heritability and shared environment structure. Our secondary goal is to apply the fractional relatedness model in a realistic setting. Lifelines is a three-generation cohort study investigating the biological, behavioral, and environmental determinants of healthy aging. We analyzed a sample of 89,353 participants from 32,452 reconstructed families. Our primary conclusion is that the effect of determinants on health outcome variables does not differ with and without accounting for family structure. However, accounting for family structure through fractional relatedness allows for estimating heritability in a computationally efficient way, showing some interesting differences between physical and mental quality of life heritability. We have shown through simulations that the proposed fractional relatedness model performs better than the standard kinship model, not only in terms of computational time and convenience of fitting using standard functions in R, but also in terms of bias of heritability estimates and coverage.

Twin registries often take part in large collaborative projects and are major contributors to genome-wide association (GWA) meta-analysis studies. In this article, we describe genotyping of twin-family populations from Australia, the Midwestern USA (Avera Twin Register), the Netherlands (Netherlands Twin Register), as well as a sample of mothers of twins from Nigeria to assess the extent, if any, of genetic differences between them. Genotyping in all cohorts was done using a custom-designed Illumina Global Screening Array (GSA), optimized to improve imputation quality for population-specific GWA studies. We investigated the degree of genetic similarity between the populations using several measures of population variation with genotype data generated from the GSA. Visualization of principal component analysis (PCA) revealed that the Australian, Dutch and Midwestern American populations exhibit negligible interpopulation stratification when compared to each other, to a reference European population and to globally distant populations. Estimations of fixation indices (FST values) between the Australian, Midwestern American and Netherlands populations suggest minimal genetic differentiation compared to the estimates between each population and a genetically distinct cohort (i.e., samples from Nigeria genotyped on GSA). Thus, results from this study demonstrate that genotype data from the Australian, Dutch and Midwestern American twin-family populations can be reasonably combined for joint-genetic analysis.

We sought to investigate the risk of incident major depressive disorder (MDD) attributable to a range of sleep disorders in the Danish population. Data were obtained by linking longitudinal Danish population-based registers. A total of 65,739 individuals who had first onset of depression between 1995 and 2013 were selected as cases. For each case, a set of 20 controls of the same sex, birth month and year and who had not had depression by the date that the case was diagnosed were selected at random form the population (N = 1,307,580 in total). We examined whether there was an increased rate of prior sleep disorders in MDD cases compared to controls using conditional logistic regression. An increased risk of incident depression in cases was found for all sleep disorders analyzed. Highest incidence rate ratios (IRRs) were found for circadian rhythm disorders (IRR = 7.06 [2.78–17.91]) and insomnia of inorganic origin (IRR = 6.76 [4.37–10.46]). The lowest estimated IRR was for narcolepsy (IRR = 2.00 [1.26–3.17]). Those diagnosed with a sleep disorder in the last 6 months were at highest risk of developing depression compared to those with at least 1 year since diagnosis (3.10 vs. 2.36). Our results suggest that having any sleep disorder is a risk factor for incident depression. Depression screening should be considered for patients with sleep disorders, and where possible, long-term follow-up for mental health problems is advisable.

Disordered gambling (DG) is a rare but serious condition that results in considerable financial and interpersonal harms. Twin studies indicate that DG is heritable but are silent with respect to specific genes or pathways involved. Existing genomewide association studies (GWAS) of DG have been substantially underpowered. Larger GWAS of other psychiatric disorders now permit calculation of polygenic risk scores (PRSs) that reflect the aggregated effects of common genetic variants contributing risk for the target condition. The current study investigated whether gambling and DG are associated with PRSs for four psychiatric conditions found to be comorbid with DG in epidemiologic surveys: major depressive disorder (MDD), attention-deficit hyperactivity disorder (ADHD), bipolar disorder (BD) and schizophrenia (SCZ). Genotype data and survey responses were analyzed from the Wave IV assessment (conducted in 2008) of the National Longitudinal Study of Adolescent to Adult Health, a representative sample of adolescents recruited in 1994–1995 and followed into adulthood. Among participants classified as having European ancestry based on genetic analysis (N = 5215), 78.4% reported ever having gambled, and 1.3% reported lifetime DG. Polygenic risk for BD was associated with decreased odds of lifetime gambling, OR = 0.93 [0.87, 0.99], p = .045, pseudo-R2(%) = .12. The SCZ PRS was associated with increased odds of DG, OR = 1.54 [1.07, 2.21], p = .02, pseudo-R2(%) = .85. Polygenic risk scores for MDD and ADHD were not related to either gambling outcome. Investigating features common to both SCZ and DG might generate valuable clues about the genetically influenced liabilities to DG.

Twins Research Australia (TRA) is a community of twins and researchers working on health research to benefit everyone, including twins. TRA leads multidisciplinary research through the application of twin and family study designs, with the aim of sustaining long-term twin research that, both now and in the future, gives back to the community. This article summarizes TRA’s recent achievements and future directions, including new methodologies addressing causation, linkage to health, economic and educational administrative datasets and to geospatial data to provide insight into health and disease. We also explain how TRA’s knowledge translation and exchange activities are key to communicating the impact of twin studies to twins and the wider community. Building researcher capability, providing registry resources and partnering with all key stakeholders, particularly the participants, are important for how TRA is advancing twin research to improve health outcomes for society. TRA provides researchers with open access to its vibrant volunteer membership of twins, higher order multiples (multiples) and families who are willing to consider participation in research. Established four decades ago, this resource facilitates and supports research across multiple stages and a breadth of health domains.

The present study explores whether genetic factors explain variation in the levels of apostasy — defined as a disengagement from religious belief, identity and/or practice — in a US-based sample during the transition from adolescence to early adulthood. I posit that genetic factors at least partially explain the variance of three measures of apostasy: disengagement from religious institutions, cessation of prayer and religious disaffiliation. I argue that genetic factors associated with risk-taking behaviors, externalizing behaviors and/or correlates of apostasy may all influence the likelihood of becoming an apostate during the transition from adolescence to early adulthood in the USA. Results reveal that genetic factors explain approximately 34% of the variance in cessation of prayer and 75% of the variance in religious disaffiliation. However, genetic factors do not influence disengagement from religious institutions. This study advances our knowledge of the etiology of apostasy and highlights the need to incorporate genetic data into social scientific research.

Neurodevelopment is sensitive to genetic and pre/postnatal environmental influences. These effects are likely mediated by epigenetic factors, yet current knowledge is limited. Longitudinal twin studies can delineate the link between genetic and environmental factors, epigenetic state at birth and neurodevelopment later in childhood. Building upon our study of the Peri/postnatal Epigenetic Twin Study (PETS) from gestation to 6 years of age, here we describe the PETS 11-year follow-up in which we will use neuroimaging and cognitive testing to examine the relationship between early-life environment, epigenetics and neurocognitive outcomes in mid-childhood. Using a within-pair twin model, the primary aims are to (1) identify early-life epigenetic correlates of neurocognitive outcomes; (2) determine the developmental stability of epigenetic effects and (3) identify modifiable environmental risk factors. Secondary aims are to identify factors influencing gut microbiota between 6 and 11 years of age to investigate links between gut microbiota and neurodevelopmental outcomes in mid-childhood. Approximately 210 twin pairs will undergo an assessment at 11 years of age. This includes a direct child cognitive assessment, multimodal magnetic resonance imaging, biological sampling, anthropometric measurements and a range of questionnaires on health and development, behavior, dietary habits and sleeping patterns. Data from complementary data sources, including the National Assessment Program — Literacy and Numeracy and the Australian Early Development Census, will also be sought. Following on from our previous focus on relationships between growth, cardiovascular health and oral health, this next phase of PETS will significantly advance our understanding of the environmental interactions that shape the developing brain.

Approximately 12% of U.S. adults have type 2 diabetes (T2D). Diagnosed T2D is caused by a combination of genetic and environmental factors including age and lifestyle. In adults 45 years and older, the Discordant Twin (DISCOTWIN) consortium of twin registries from Europe and Australia showed a moderate-to-high contribution of genetic factors of T2D with a pooled heritability of 72%. The purpose of this study was to investigate the contributions of genetic and environmental factors of T2D in twins 45 years and older in a U.S. twin cohort (Washington State Twin Registry, WSTR) and compare the estimates to the DISCOTWIN consortium. We also compared these estimates with twins under the age of 45. Data were obtained from 2692 monozygotic (MZ) and same-sex dizygotic (DZ) twin pairs over 45 and 4217 twin pairs under 45 who responded to the question ‘Has a doctor ever diagnosed you with (type 2) diabetes?’ Twin similarity was analyzed using both tetrachoric correlations and structural equation modeling. Overall, 9.4% of MZ and 14.7% of DZ twins over the age of 45 were discordant for T2D in the WSTR, compared to 5.1% of MZ and 8% of DZ twins in the DISCOTWIN consortium. Unlike the DISCOTWIN consortium in which heritability was 72%, heritability was only 52% in the WSTR. In twins under the age of 45, heritability did not contribute to the variance in T2D. In a U.S. sample of adult twins, environmental factors appear to be increasingly important in the development of T2D.

Genetically informative research designs are becoming increasingly popular as a way to strengthen causal inference with their ability to control for genetic and shared environmental confounding. Co-twin control (CTC) models, a special case of these designs using twin samples, decompose the overall effect of exposure on outcome into a within- and between-twin-pair term. Ideally, the within-twin-pair term would serve as an estimate of the exposure effect controlling for genetic and shared environmental factors, but it is often confounded by factors not shared within a twin-pair. Previous simulation work has shown that if twins are less similar on an unmeasured confounder than they are on an exposure, the within-twin-pair estimate will be a biased estimate of the exposure effect, even more biased than the individual, unpaired estimate. The current study uses simulation and analytical derivations to show that while incorporating a covariate related to the nonshared confounder in CTC models always reduces bias in the within-pair estimate, it will be less biased than the individual estimate only in a narrow set of circumstances. The best case for bias reduction in the within-pair estimate occurs when the within-twin-pair correlation in exposure is less than the correlation in the confounder and the twin-pair correlation in the covariate is high. Additionally, the form of covariate inclusion is compared between adjustment for only one’s own covariate value and adjustment for the deviation of one’s own value from the covariate twin-pair mean. Results show that adjusting for the deviation from the twin-pair mean results in equal or reduced bias.

Using newly available polygenic scores for educational attainment and cognitive ability, this paper investigates the possible presence and causes of a negative association between IQ and fertility in the Wisconsin Longitudinal Study sample, an issue that Retherford and Sewell first addressed 30 years ago. The effect of the polygenic score on the sample’s reproductive characteristics was indirect: a latent cognitive ability measure, comprised of both educational attainment and IQ, wholly mediated the relationship. Age at first birth mediated the negative effect of cognitive ability on sample fertility, which had a direct (positive) effect on the number of grandchildren. Significantly greater impacts of cognitive ability on the sample’s fertility characteristics were found among the female subsample. This indicates that, in this sample, having a genetic disposition toward higher cognitive ability does not directly reduce number of offspring; instead, higher cognitive ability is a risk factor for prolonging reproductive debut, which, especially for women, reduces the fertility window and, thus, the number of children and grandchildren that can be produced. By estimating the effect of the sample’s reproductive characteristics on the strength of polygenic selection, it was found that the genetic variance component of IQ should be declining at a rate between −.208 (95% CI [−.020, −.383]) and −.424 (95% CI [−.041, −.766]) points per decade, depending on whether GCTA-GREML or classical behavior genetic estimates of IQ heritability are used to correct for ‘missing’ heritability.

Determining (1) the direction of causation and (2) the size of causal effects between two constructs is a central challenge of the scientific study of humans. In the early 1990s, researchers in behavioral genetics invented what was termed the direction of causation (DoC) model to address exactly these two concerns. The model claims that for any two traits whose mode of inheritance is sufficiently different, the direction of causation can be ascertained using a sufficiently large genetically informative sample. Using a series of simulation studies, we demonstrate a major challenge to the DoC model, namely that it is extremely sensitive to even tiny amounts of non-shared confounding. Even under ideal conditions for the DoC model (a large sample, N = 10,000), a large causal relationship (e.g., a causal correlation of .50) with very different modes of inheritance between the two traits (e.g., a pure AE model for one trait and a pure CE model for another trait) and a modest degree (correlation of .10) of non-shared confounding between the two traits results in the choice of the wrong causal models and estimating the wrong causal effects.

We aimed to study the eating behavioral traits that associate with body mass index (BMI) among BMI-discordant twin pairs. This cross-sectional study examined self-reported eating behaviors in 134 healthy young adult twin pairs (57 monozygotic [MZ] and 77 same-sex dizygotic [DZ]), of whom 29 MZ and 46 DZ pairs were BMI discordant (BMI difference ≥ 3 kg/m2). In both MZ and DZ BMI-discordant pairs, the heavier co-twins reported being less capable of regulating their food intake optimally than their leaner co-twins, mainly due to ‘frequent overeating’. Furthermore, the heavier co-twins reported augmented ‘disinhibited eating’, ‘binge-eating scores’ and ‘body dissatisfaction’. The twins agreed more frequently that the heavier co-twins (rather than the leaner co-twins) ate more food in general, and more fatty food, in particular. No significant behavioral differences emerged in BMI-concordant twin pairs. Overeating — measured by ‘frequent overeating’, ‘disinhibited eating’ and ‘binge-eating score’ — was the main behavioral trait associated with higher BMI, independent of genotype and shared environment.

Structural equation modeling (SEM) is an important research tool, both for path-based model specification (common in the social sciences) and also for matrix-based models (in heavy use in behavior genetics). We developed umx to give more immediate access, relatively concise syntax and helpful defaults for users in these two broad disciplines. umx supports development, modification and comparison of models, as well as both graphical and tabular outputs. The second major focus of umx, behavior genetic models, is supported via functions implementing standard multigroup twin models. These functions support raw and covariance data, including joint ordinal data, and give solutions for ACE models, including support for covariates, common- and independent-pathway models, and gene × environment interaction models. A tutorial site and question forum are also available.

The East Flanders Prospective Twin Survey (EFPTS) is a registry of multiple births in the province of East Flanders, Belgium. Since its start in 1964, over 10,000 twin-pairs have been registered. EFPTS has several unique features: it is population-based and prospective, with the possibility of long-term follow-up; the twins (and higher order multiple births) are recruited at birth; basic perinatal data are recorded; chorion type and zygosity are established; since 1969, placental biopsies have been taken and frozen at –20°C for future research. Since its origin, the EFPTS has included placental data and allows differentiation of three subtypes of monozygotic twins based on the time of the initial zygotic division: the dichorionic–diamniotic pairs (early, with splitting before the fourth day after fertilization), the monochorionic–diamniotic pairs (intermediate, splitting between the fourth- and the seventh-day postfertilization) and the monochorionic–monoamniotic pairs (late, splitting after the eighth day postfertilization). Studies can be initiated taking into account primary biases, those originating ‘in utero’. Such studies could throw new light on the consequences of early embryological events and the gene–environment interactions as far as periconceptional and intrauterine environment are concerned.