Search

Estimates from Mendelian randomization studies of unrelated individuals can be biased due to uncontrolled confounding from familial effects. Here we describe methods for within-family Mendelian randomization analyses and use simulation studies to show that family-based analyses can reduce such biases. We illustrate empirically how familial effects can affect estimates using data from 61,008 siblings from the Nord-Trøndelag Health Study and UK Biobank and replicated our findings using 222,368 siblings from 23andMe. Both Mendelian randomization estimates using unrelated individuals and within family methods reproduced established effects of lower BMI reducing risk of diabetes and high blood pressure. However, while Mendelian randomization estimates from samples of unrelated individuals suggested that taller height and lower BMI increase educational attainment, these effects were strongly attenuated in within-family Mendelian randomization analyses. Our findings indicate the necessity of controlling for population structure and familial effects in Mendelian randomization studies., Family-based study designs have been applied to resolve confounding by population stratification, dynastic effects and assortative mating in genetic association analyses. Here, Brumpton et al. describe theory and simulations for overcoming such biases in Mendelian randomization through within-family studies.

BACKGROUND Attention deficit hyperactivity disorder (ADHD) and autism spectrum disorder (ASD) are two highly heritable neurodevelopmental disorders. Several lines of evidence point towards the presence of shared genetic factors underlying ASD and ADHD. We conducted genomic analyses of common risk variants (i.e. single nucleotide polymorphisms, SNPs) shared by ASD and ADHD, and those specific to each disorder. METHODS With the summary data from two GWAS, one on ASD (N = 46,350) and another on ADHD (N = 55,374) individuals, we used genomic structural equation modelling and colocalization analysis to identify SNPs shared by ASD and ADHD and SNPs specific to each disorder. Functional genomic analyses were then conducted on shared and specific common genetic variants. Finally, we performed a bidirectional Mendelian randomization analysis to test whether the shared genetic risk between ASD and ADHD was interpretable in terms of reciprocal relationships between ASD and ADHD. RESULTS We found that 37.5% of the SNPs associated with ASD (at p

Now, more than a decade into the genome-wide association study (GWAS) era, continued progress in genomic discovery for psychiatric traits is easily taken for granted. The winning formula of large-scale collaborative science, uncompromising methodological rigor, and brute force sample size accumulation has become well established, and its regular empirical returns have become progressively less headline-making. However, we should not be complacent. As the results of the systematic review by Akingbuwa et al.1 in this issue of the Journal make clear, accumulating the necessary numbers of genotyped cases to identify the majority of common risk variants remains a daunting and distant prospect for many psychiatric traits relevant to children and adolescents. But more than this, the challenge of providing clinically and societally meaningful returns on the considerable investments of the GWAS era will require a continued and concerted research effort well into the future.

The study aims to (1) explore the contribution of direct and indirect genetic effects to both general and domain-specific early neurodevelopmental problems, (2) investigate to what extent those are captured in polygenic liabilities for specific neurodevelopmental conditions and related cognitive traits, and (3) test the specificity (general/specific) of the overall degree to which indirect and direct effects of these specific polygenic liabilities influence early neurodevelopmental problems.

Preregistration of analyses

Cognitive and academic performance appear to be highly heritable, especially with age, rising from 20–50% in childhood and adolescence to between 50–80% in adulthood (Bouchard & McGue, 1981; Haworth et al., 2010; Kovas et al., 2013). Although this appears to suggest that environmental factors play only a minor role in the development of cognitive and academic performance, there is good reason to believe, firstly, that parental genotype influences children not only directly but also through environmentally mediated pathways (via what has come to be known as ‘genetic nurture’; Kong et al. 2018), and, secondly, that the environment may have an amplifying effect on genetic influences through mechanisms of evocative gene-environment correlation (rGE; Dickens & Flynn, 2001). While there have been many recent studies of genetic nurture and educational attainment (Wang et al. 2020), there has been far less research on evocative rGE in educational development. Evocative rGE occurs when an individual’s genetically influenced characteristics systematically evoke responses from their environment, in turn, mediating or even amplifying genetic effects (Plomin, DeFries Loehlin, 1977; Scarr & McCartney, 1983). As these evoked environmental conditions are correlated with genetic influences, their effects may be masked by estimates of genetic effects. There is some evidence of evocative rGE in cognitive development and educational attainment. For example, there is evidence of evocative rGE in early verbal and non-verbal cognitive development from one study that used multivariate twin modelling (Tucker-Drob & Harden, 2012). There is also evidence of evocative rGE in educational attainment from a study examining the effects on parenting of children’s education polygenic scores (PGS), after statistically controlling for mothers’ polygenic scores (Wertz et al., 2019). However, the evidence base is small and, to our knowledge, evocative effects on parenting of educational attainment PGS have never been examined while controlling for PGS from both parents. We propose to do so using data from the Norwegian Mother, Father and Child Cohort Study (MoBa) sample. We also plan to examine whether early evocative effects on parenting may in turn influence children’s subsequent academic performance. Research suggests that early childhood language skills are early manifestations of genetic influences on later cognitive and academic performance (Austerberry et al., 2021; Verhoef et al., 2021) and may evoke differences in parenting that in turn influence children’s academic performance (Tucker-Drob & Harden, 2012). Guided by this evidence, we will also examine whether early language skills mediate any observed genetic effects on later academic performance, and whether early language skills are correlated with early parenting.

The aim of this study is to examine whether alcohol, tobacco and caffeine use in pregnancy might be causally associated with ADHD in offspring and use negative control and Mendelian Randomization methods to disentangle potential causal intrauterine effects.

Background: Parents and children resemble each other in ADHD behaviors. A key theoretical postulate from the evolutionary life history theory is that children use information from their environment (e.g., predictability and resource availability) and develop traits that are optimized for future success in that environment. Life history theory therefore expects that variation in children’s ADHD behaviors reflects evolved calibration to the developmental context that parents provide. Alternatively, children may resemble their parents not because of the provided environments, but because of their transmitted genes.Methods: We employed a large children-of-twins-and-siblings sample (N=22,350 parents & 11,566 8-year-old children) of the Norwegian Mother, Father and Child Cohort Study (MoBa). This enables disentangling intergenerational influences via parental genes and parental behavior (i.e., genetic and environmental transmission, respectively). Fathers reported on their symptoms (or absence thereof) and mothers on their own and their child’s symptoms.Results: Child ADHD behaviors correlated 0.24 with ADHD behaviors in mothers and .10 with those in fathers. These correlations were largely due to genetic transmission; genetic transmission was five times larger than environmental transmission. Variation in children’s ADHD behaviors was largely explained by heritability (57%), with small effects of parental ADHD behaviors (2% environmental transmission), and gene-environment correlation (3%). The remainder was due to unique environmental influences and noise.Conclusion: The intergenerational transmission of ADHD behaviors is primarily due to genetic transmission, with little evidence for parental ADHD behaviors causing children’s ADHD behaviors. This contradicts the life history theory. The child-specific variability in ADHD behaviors is due, in equal amounts, to genetic factors that influence children’s but not parents’ ADHD behaviors, and to environmental influences and noise that make siblings unique from one another. We conclude that ADHD is not the outcome of adaptive behavior to the (family) environment, but rather a neurodevelopmental disorder with a strong genetic basis.

BackgroundThe Norwegian Mother, Father, and Child Cohort Study (MoBa) is a population-based pregnancy cohort, which includes approximately 114,500 children, 95,200 mothers, and 75,200 fathers.Genotyping of MoBa has been conducted through multiple research projects, spanning several years; using varying selection criteria, genotyping arrays, and genotyping centres. MoBa contains numerous interrelated families, which necessitated the implementation of a family-based quality control (QC) pipeline that verifies and accounts for diverse types of relatedness.MethodsThe MoBaPsychGen pipeline, comprising pre-imputation QC, phasing, imputation, and post-imputation QC, was developed based on current best-practice protocols and implemented to account for the complex structure of the MoBa genotype data. The pipeline includes QC on both single nucleotide polymorphism (SNP) and individual level. Phasing and imputation were performed using the publicly available Haplotype Reference Consortium release 1.1 panel as a reference. Information from the Medical Birth Registry of Norway and MoBa questionnaires were used to identify biological sex, year of birth, reported parent-offspring (PO) relationships, and multiple births (only available in the offspring generation).ResultsIn total, 207,569 unique individuals (90% of the unique individuals included in the study) and 6,981,748 SNPs passed the MoBaPsychGen pipeline. The relatedness checks performed throughout the pipeline allowed identification of within-generation and across-generation first-degree, second-degree, and third-degree relatives. The individuals passing post-imputation QC comprised 64,471 families ranging in size from singletons to 84 unique individuals (singletons are included as families as other family members may not have been genotyped, imputed, or passed post-imputation QC). The relationships identified include 287 monozygotic twin pairs, 22,884 full siblings, 117,004 PO pairs, 23,299 second-degree relative pairs, and 10,828 third-degree relative pairs.DiscussionMoBa contains a highly complex relatedness structure, with a variety of family structures including singletons, PO duos, full (mother, father, child) PO trios, nuclear families, blended families, and extended families. The availability of robustly quality-controlled genetic data for such a large cohort with a unique extended family structure will allow many novel research questions to be addressed. Furthermore, the MoBaPsychGen pipeline has potential utility in similar cohorts.

Recent research has suggested that across Western developed societies, the influence of genetics on educational outcomes is relatively constant. However, the degree to which family environment matters varies, such that countries with high levels of intergenerational mobility have weaker associations of family background. Research in this vein has relied on twin-based estimates, which involve variance decomposition, so direct assessment of the association of genes and environments is not possible. In the present study, we approach the question by directly measuring the impact of child genotype, parental genetic nurture, and parental realized education on educational achievement in primary and secondary school. We deploy data from a social democratic context (Norway) and contrast our findings with those derived from more liberal welfare state contexts. Results point to genetics only confounding the relationship between parent status and offspring achievement to a small degree. Genetic nurture associations are similar to those in other societies. We find no, or very small, gene–environment interactions and parent–child genotype interactions with respect to test scores. In sum, in a Scandinavian welfare state context, both genetic and environmental associations are of similar magnitude as in societies with less-robust efforts to mitigate the influence of family background. publishedVersion

Evidence linking parental diagnoses of inflammatory bowel disease (IBD) with offspring autism is inconclusive. We conducted four complementary studies to investigate associations between parental diagnoses of IBD and offspring autism and elucidate their underlying aetiology. (1) Nationwide population-based cohort study using Swedish registers to examine associations between parental IBD diagnoses and autism diagnoses in offspring, (2) Linkage disequilibrium (LD)-score regression to estimate the genetic correlation between the phenotypes. (3) Polygenic risk score (PRS) analyses in the Avon Longitudinal Study of Parents and Children (ALSPAC) to investigate associations between maternal genetic liability to IBD and autism factor mean score in offspring. (4) Two-sample Mendelian randomization (MR) to assess bidirectional causal links between genetic liability to IBD and autism. We found evidence of an association between parental IBD diagnoses and offspring autism (maternal: OR= 1.32; 95% CI: 1.25 to 1.40; p

Neurodevelopmental disorders are widely acknowledged to be complex and multifactorial in origin, but this is rarely reflected in the approaches used to study them. We reflect on the 2021 Annual Research review and its introduction of a new conceptual framework designed to make the complexity of early neurodevelopment more empirically tractable. We evaluate the review authors’ justification, explanation, and guidance for implementation of their framework in the context of their stated goals and highlight key assumptions that support its conceptual validity. Finally, we offer a genetic epidemiological perspective on potential applications, suggesting ways in which genomic data can be used to elucidate causal mechanistic processes within the AMEND framework.

Indirect genetic effects from relatives may result in misleading quantifications of heritability, but can also be of interest in their own right. In this paper we propose Trio-GCTA, a model for separating direct and indirect genetic effects when genome-wide single nucleotide polymorphism data have been collected from parent-offspring trios. The model is applicable to phenotypes obtained from any of the family members. We discuss appropriate parameter interpretations and apply the method to three exemplar phenotypes: offspring birth weight, maternal relationship satisfaction, and paternal body-mass index, using real data from the Norwegian Mother, Father and Child Cohort Study (MoBa).

BackgroundPoint estimation in Mendelian randomization (MR), an instrumental variable model, usually requires strong homogeneity assumptions beyond the core instrumental conditions. Bounding, which does not require homogeneity assumptions, is infrequently applied in MR.ObjectiveWe aimed to demonstrate computing nonparametric bounds for the causal risk difference derived from multiple proposed instruments in an MR study where effect heterogeneity is expected,MethodsUsing data from the Norwegian Mother, Father, and Child Cohort Study and Avon Longitudinal Study of Parents and Children (n=4457, 6216) to study the average causal effect of maternal pregnancy alcohol use on offspring attention deficit hyperactivity disorder symptoms, we proposed 11 maternal SNPs as instruments. We computed bounds assuming subsets of SNPs were jointly valid instruments, for all combinations of SNPs where the MR model was not falsified.ResultsThe MR assumptions were violated for all sets with more than 4 SNPs in one cohort and for all sets with more than 2 SNPs in the other. Bounds assuming one SNP was an individually valid instrument barely improved on assumption-free bounds. Bounds tightened as more SNPs were assumed to be jointly valid instruments, and occasionally identified directions of effect, though bounds from different sets varied.ConclusionsOur results suggest that, when proposing multiple instruments, bounds can contextualize plausible magnitudes and directions of effects. Computing bounds over multiple assumption sets underscores the importance of evaluating the assumptions of MR models.SynopsisStudy questionDo nonparametric bounds provide useful information in the context of MR studies of prenatal exposures with multiple proposed genetic instruments?What’s already knownPoint estimation in MR typically requires strong, unverifiable homogeneity assumptions beyond the core MR assumptions. Bounds, which do not require homogeneity assumptions, are rarely applied in MR.What this study addsWe computed bounds on the average causal effect of alcohol consumption during pregnancy on offspring ADHD symptoms in two European cohorts, proposing 11 genetic variants as instruments. Our results suggest that, when proposing multiple instruments, bounds can contextualize plausible magnitudes and directions of effects.

Introduction: Current advice to improve fertility includes reducing alcohol and caffeine consumption, achieving healthy weight-range, and stopping smoking. Advice is informed by observational evidence, which is often biased by confounding. Methods: This study uses data from the Norwegian Mother, Father and Child Cohort Study (MoBa) and the Medical Birth Registry of Norway. First, we analysed associations between health behaviours prior to pregnancy (alcohol and caffeine consumption, body-mass index (BMI) and smoking) and multiple indicators of fertility (including number of children, time to conception, and miscarriage) (n=83,128 women, 67,555 men), adjusting for birthyear, education and attention deficit and hyperactive-impulsive (ADHD) traits. Second, we used individual-level Mendelian randomisation (MR) to explore possible causal effects of health behaviours on fertility outcomes (n=27,216 women, 26,131 men). Finally, we performed summary-level MR for available outcomes (n=91,462-1,232,091) and conducted multi-variable MR to control for education and ADHD liability. Results: In observational analysis, higher BMI and smoking (and to a lesser extent caffeine) were predominantly associated with reduced fertility outcomes. Unexpectedly, higher alcohol consumption was associated with predominantly improved fertility outcomes. There was little evidence from individual-level MR analyses, except smoking and higher BMI were associated with younger age at first birth in women (mean difference in years, per SD increase in genetic score; smoking: -2.65 (95%CI: -3.57, -1.73); BMI: -0.11 (95%CI: -0.16, -0.08)) and men (smoking: -2.82 (95%CI: -4.07, -1.58); BMI: -0.17 (95%CI: -0.23, -0.11)). These results were replicated in the summary-level MR analysis, however effects attenuated after adjusting for education and ADHD liability. Conclusions: Most observational evidence for associations between health behaviours and fertility was not supported by MR analyses, suggesting possible residual confounding. Evidence from MR analyses supported an effect of smoking and higher BMI on younger age of first birth, but multivariable MR suggested this might be explained by underlying liability to ADHD and low educational attainment.

BackgroundMood and anxiety disorders, often emerging during adolescence, account for a large share of the global burden of disability. Prospectively assessed premorbid early signs and trajectories can provide useful insights for early detection and development of these disorders.MethodsUsing the health registry linked Norwegian Mother, Father and Child Cohort Study (MoBa) of 110,367 children, we here examine cross-sectional and longitudinal association between temperamental traits, emotional and behavioral problems in childhood (0.5-8 years) and diagnosis of mood or anxiety (emotional) disorders in adolescence (10-18 years). We included birth year and sex, retrieved from the Medical Birth Registry of Norway, as covariates in all analyses.ResultsLogistic regression analyses showed consistent and increasing associations between childhood negative emotionality, behavioral and emotional problems and adolescent diagnosis of emotional disorders, present from 6 months of age (negative emotionality) and with similar magnitude of association for the associated traits. Latent profile analysis incorporating latent growth models identified five developmental profiles of emotional and behavioral problems. A profile of early increasing behavioral and emotional problems with combined symptoms at 8 years (1.3% of sample) was the profile most strongly associated with emotional disorders in adolescence (OR vs. reference: 5.00, 95% CI: 3.73-6.30).ConclusionsWe found a consistent and increasing association between negative emotionality, behavioral and emotional problems in early to middle childhood and mood and anxiety disorders in adolescence. A developmental profile coherent with early and increasing disruptive mood dysregulation across childhood was most predictive of adolescent emotional disorders. Our results highlight the importance of early emotional dysregulation and childhood as a formative period in the development of adolescent mood and anxiety disorders, supporting a potential for prevention and early intervention initiatives.