Elsje van Bergen

Autism is diagnosed much more often in men than in women. Is this the result of biological differences, or male-centric science and healthcare with diagnostic 'blind spots'? With SCANNER, scientists will investigate both explanations and develop sensorimotor diagnostic tools to improve gender-sensitive care for people with autism.

Poor educational achievement and ADHD in childhood are a major risk factors for poor health and low income in adulthood. Poor educational outcomes tend to run in families, generating a cycle of inequality. Children whose parents talk and read less to them, tend to struggle in school. This parent-child association has been interpreted as a causal effect of parenting, but parents provide their children not only with a rearing environment, but also genes. We can only discover causal effects of the home learning environment (e.g., parent-child interactions, reading storybooks, household chaos) if we control for genetics or run an RCT. Our overarching aim is to discover which aspects of the home learning environment causally impact children’s risk of ADHD and learning difficulties. GenEd will use a variety of innovative observational and experimental designs. We will employ existing rich datasets and will collect novel data to triangulate three intergenerational designs. Using an online-learning platform, we will assess parent and child literacy and numeracy in 3000 twin families. In an RCT, 2000 of these families will get access to the online-learning platform to boost the home learning environment. GenEd will: 1. Include gene-environment interplay in studying children’s ADHD-symptoms, motivation, and literacy and numeracy skills 2. Discover which features of the home learning environment have a causal impact 3. Test the effectiveness of offering families an online-learning platform to practice literacy and numeracy GenEd ’s interdisciplinairity (representing education and genetics), combined with breakthroughs in statistical modelling, gene-finding work, and educational technology, make us uniquely suited to lead this timely project. GenEd has the ground-breaking opportunity to unravel the mechanisms underlying familial educational disadvantage. This will inform policy on how to target malleable causal mechanisms to ensure that all children can learn and thrive.

Parents of lower socioeconomic status (SES) not only tend to have more health problems, but also tend to have offspring with more health problems. What remains a question is whether this association reflects a causal effect of SES on health, or reflects genetic or environmental confounding. Causality is typically studied by randomized controlled trials, where confounds are random over conditions and easily controlled for. However, it is not feasible to randomly allocate people to intervention conditions in which SES is altered, so to study causal influences of SES on health we rely on natural experiments. We propose to apply two natural experiments to study intergenerational causality. In these studies genetics serves as a research tool. Using this tool we will study effects of different SES indicators (e.g. educational attainment (EA), household income, and social deprivation) on physical and mental health outcomes. In the first design, referred to as the twin discordance design, we will study genetically-identical (i.e., monozygotic or MZ) twins who differ in (so are discordant for) the SES indicator. Does the twin member who is of lower SES compared to his/her co-twin also suffer poorer health? If so, this strengthens the evidence for a causal relation between SES and health, because the association is controlled for genetic and childhood confounders, as these are identical in MZ twins. In the second design, the intergenerational (Mendelian randomization, or MR) design, we study the effects of parental and offspring SES on offspring health outcomes. We leverage the fact that each offspring within a nuclear family inherits a mixture of genetic variants from both parents, in some cases inheriting more SES increasing (vs. decreasing) genetic variants, rendering his or her genetic predisposition for SES above (vs. below) that of his/her parents. As genotypes which predispose to higher or lower SES are inherited randomly, they form the perfect basis for natural experiments. The genetic predispositions are utilized as instruments to identify the ’true’ effect of SES on health. This intergenerational design enables us to study the effect of both parental and offspring SES on health. At the etiological level, individual differences in SES can be explained by the interplay of genetic and environmental differences. Consider for example EA, important influences on EA are IQ and personality, which in turn are influenced by genes and environmental factors. Hence, the complex multifactorial construct of SES is also influenced, but not determined, by genes. That is, differences among people in SES are partly due to genetic differences, like a genetic predisposition to ease-of-learning and perseverance. The twin discordance and intergenerational MR design will be employed to test causal effects of aspects of SES on health, in particular cardio-metabolic (e.g., myocardial infarction and type II diabetes) and mental health (e.g., depression and ADHD). These outcomes are prevalent and impose a heavy burden on affected individuals, families, and society at large. As it also conceivable that health problems lead to lower SES, we test for the presence of causality in either direction. To realize this project we will combine three rich and unique research facilities: the Netherlands Twin Register (NTR), Statistics Netherlands (CBS), and the Geoscience and Health Cohort Consortium (GECCO). NTR includes thousands of twin families with longitudinal health and genetic data and very good national coverage (see: Figure 1 Together these resources provide wide variety of SES indicators, which, if added to the models separately, can shed light on the mechanism(s) of how SES influences health. In addition, CBS (diagnosis and hospital data) and NTR (questionnaire and self-report) provide a wealth of health outcome data. Replication of findings will be performed in international twin registers with access to similar data. The project can implicate and address the causality of the association between SES and health outcomes. Malleable behaviors and exposures that are causally implicated provide evidence-based targets for future interventions to improve health outcomes. Behaviors and exposures that are associated with mental or cardio-metabolic health, but not because of causal mechanisms, can be discarded as intervention targets, saving valuable time and resources. Consider, for example, the finding that a failure to complete secondary education causally predisposes poor health outcomes, but only in those with a lower parental SES. The ability to easily identify those at risk for poor health outcomes, allows for the selective application of preventative programs (e.g. adult education), and risk screening (e.g. population screening for certain disease at earlier ages dependent on educational background). Thus, findings will improve health-economic modeling of the consequences of social inequality, and thus shape effective policy.