Folate and other B vitamins are one-carbon micronutrients essential for a range of biological processes such as DNA replication, repair and gene expression. Folate, for instance, plays a key role in fetal development and can significantly reduce the risk of neural tube defects (NTD). Folate is available in small quantities in many foods however it is recommended that women supplement with folic acid (synthetic form of folate) daily during pregnancy. The association between insufficient folate and NTD has led to several countries recommending the fortification of foods such as flour and cereal grains with folic acid. In addition to a putative role of maternal folate deficiency in fetal development, blood folate levels have been linked, albeit inconsistently, to risk of various cancers in population-based studies. It is believed that one mechanism through which folate may influence subsequent adverse developmental outcomes and cancer risk in early life is via differential DNA methylation (a form of gene regulation) in blood. Consequently, our proposed research aims to further interrogate an effect of one-carbon micronutrient levels on various cancer outcomes, in addition to examining whether a putative effect of one-carbon micronutrient levels on both adverse developemental outcomes and cancer may operate through epigenetic mechanisms in early life. To perform these analyses we will use genome-wide association data derived from previous analyses performed in ALSPAC (B2104) to 1) estimate the causal effect of genetically-proxied blood folate on various adult cancers (using previously published summary GWAS data for these cancers), and 2) to examine the effect of genetic risk scores for various one-carbon micronutrients on DNA methylation profiles in children.

Little is known regarding adolescent cognitive functioning before individuals develop schizophrenia (SZ). Based on neurocognitive measures of impulsivity, we have found that nonpsychotic adolescent first-degree biological relatives of SZ patients (FDR) are more impulsive than adolescents with no SZ family history. Our study further indicates that structural and functional abnormalities in prefrontal brain regions confer increased impulsivity in FDR. These findings have led us to postulate that abnormalities in frontal-striatal interactions that mediate reward processing may underlie increased impulsivity among individuals at-risk for SZ. This ALSPAC proposal will extend our research by studying adolescent impulsivity in individuals before they are diagnosed with SZ. This project is significant because it has high potential to advance current understanding regarding SZ susceptibility. It will lead to new knowledge regarding the role of prefontal brain circuits during adolescence in mediating reward processing and impulsive behaviors prior to the clinical manifestations of SZ.

There are 750 fasting blood samples that were collected from ALSPAC participants aged 7-8. Measuring metabolites levels in these samples will help us understand how metabolic changes in the body relate to various exposures and outcomes. Metabolite levels were previously quantified for individuals aged 7-8, but these metabolite levels were measured in non-fasting samples. Thus measuring these new samples will help us understand if there are any differences between metabolite levels when fasting and not fasting and how to better analyse our current data.

Background

A flourishing and prosperous society is dependent on healthy, educated, and resilient young people with the life skills to become thriving adults. The factors that shape long-term health and wellbeing exert their influence from an early stage. Ensuring a thriving adult population in the future depends on providing the necessary support and conditions for young people in the here and now. The influence of social determinants on long-term health outcomes is largely accepted and understood by those working in health and social policy. However, the implications of this are less well understood by decision makers across other sectors and the wider public. The social determinants approach to health emphasises the importance of creating the conditions that promote good long-term health outcomes across the life course and acting early in life to increase people’s ability to build the foundations needed to thrive.

We know early socioeconomic status (SES) is linked to later health outcomes (Warren, 2009). But we do not know much about how this works in the 12-24 age period, nor what happens when we focus directly on the assets young people are accumulating. Some of the more traditional supports and safety nets (such as a buoyant youth labour market and access to welfare benefits) are not as available to today’s young people compared to previous generations. There is a sense that some of them may be transitioning into adulthood with fewer ‘assets’ or ‘resources’ to help them weather the storms as they age, and that this will be related to their later health outcomes as older adults. Thus, the rationale for this study developed as follows:

• The Health Foundation took as its starting point the work of Naomi Eisenstadt on Life Chances of Young People in Scotland (2017) and defined the foundations of a healthy life as:
 The potential to engage in good quality work
 Access to secure, affordable homes in flourishing communities
 A network of stable, supportive relationships and good social and emotional
wellbeing, and
 Established habits that promote and maintain good health.

• An engagement exercise then worked with young people themselves to identify the assets which they felt had contributed to their current situation across the 4 foundations. These assets were defined as:

 Skills and qualifications - “how right my skills are for the career I want”;
 Personal connections - “the confidence and connections to navigate the adult world”;
 Financial and practical support - “having the support to achieve what I want from life”,
 Emotional support - “people I can lean on emotionally

• The state of health and health behaviours in early adulthood (mid 20s) is likely to be related to later health outcomes. If poor habits and health outcomes are already emerging by this age, they are likely to persist.

• It seems possible that at least some of today’s young people have the odds stacked against them in terms of their experiences 12-24, compared with (a) their current peers, and (b) previous generations.

• If we knew more about how to support young people aged 12-24, helping them to enter the mid-20s in a state of positive wellbeing, then we can work to protect their future health outcomes.

Rather than focusing on parental SES, this study will be unique in looking at the assets accumulated by the young people themselves that they then take on with them into the next life stage. However we will examine whether there are interactions between parental SES and young people’s assets, enabling us to identify whether assets (when present) are differentially protective for young people from differing SES families.

We will measure levels of thousands of proteins and peptides present in biological samples collected from ALSPAC participants at 3 or more ages using a novel methodolgy (SWATH Mass spectrometry). This will enable us to identify biomarkers which change as individuals develop and biomarkers that show differences between individuals. This data will be analysed in association with the comprehensive data available from ALSPAC participants, including genetic, phenotypic, environmental exposure and outcome data to look for reasons for differences. This analysis could identify biomarkers associated with particular outcomes and how these change during development or why different individuals have different outcomes.

The human gut microbiome, a huge community of microorganisms residing within our digestive system that interact to aid digestion, protect against pathogens and create essential metabolites, has been suggested to play a role in health and disease. However, much of the literature (arising from animal or observational human studies) attempting to assess this relationship has so far been inconsistent and unable to prove cause-and-effect relationships. Whilst traditional epidemiological studies have gone some way in assessing the association between the gut microbiome and health, these study designs are small and limited in their ability to draw causal inference due to the existence of other factors associated with both the gut microbiome and health (“confounding”), the possibility of reverse causation (where an association is actually due to a disease causing variation in the gut microbiome rather than the reverse) and many forms of bias. Thus, these limitations can distort or skew our observed association and lead us to draw false conclusions. Additionally, randomized controlled trials (RCTs), the best and most effective way of assessing causality in epidemiological studies, are not feasible, expensive and likely unethical within a healthy human population. I propose to address this question – does the gut microbiome cause adverse health and disease?

With the UK Research and Innovation (UKRI) Future Leaders Fellowship (FLF), I will combine my expertise in the design, analysis and interpretation of causal inference methods (such as Mendelian randomization (MR)) and population health sciences with microbiome health to address the fundamental problem in the study of the gut microbiome in human health – causality. Analogous to nature’s RCT, MR is a method that uses randomly allocated genetic variation within a population (identified by genome-wide association studies (GWAS)) as a proxy for a “confounded” exposure (i.e., the gut microbiome) to improve causal inference in an association between that exposure and an outcome (e.g., colorectal cancer). In my current role, I lead several projects and am involved in international collaborations, upon which I will build in this fellowship. Specifically, within a working group of six individuals, I am involved in a GWAS of the gut microbiome in the Flemish Gut Flora Project (FGFP; N~3000) and a GWAS of dietary composition within the Social Sciences Genetics Association Consortium.

With the established collaborations, data resources and world-renowned expertise in epidemiology and causal inference methods available to me at the Medical Research Council Integrative Epidemiology Unit (MRC-IEU), the UKRI FLF will enable this interdisciplinary, collaborative, data-driven and large-scale research with the aim of challenging the impact of the gut microbiome on important health outcomes (specifically, colorectal cancer, type 2 diabetes and inflammatory bowel disease) alongside one of the key exposures dominating the literature as a regulator of the gut microbiome - dietary intake and composition. To achieve this aim, I have identified a novel set of deliverable objectives to complete within the first four years of the UKRI FLF scheme: (1) collate data and results from the GWAS of dietary composition, the gut microbiome and outcomes of interest; (OB2) use these results in MR analyses to understand the causal role played by diet on the gut microbiome and of the gut microbiome on health; (OB3) perform targeted follow-up analyses with comprehensive measures of the microbiome and the product of their biological processes within the gut. I will be the first to unify epidemiological and genetic data with causal inference methods and gut microbiome research, areas in which I have been building experience, but which have yet to be brought together. The gut microbiome has great potential to have clinical importance to improve health given well-directed therapeutic development.

Impulsivity is a multidimensional psychological trait that has been linked to addictive disorders and other mental
health conditions. However, most of the studies to date have used cross-sectional designs. Our current ALSPAC project added impulsivity phenotypes to the cohort at one time point and this proposal is to add the same measures to a second time point. This will allow us to examine impulsivity's causal/consequential role as a risk mechanism.

Congenital heart diseases (CHDs) are the most common type of birth anomaly and a major source of childhood death and illness. Studies that have included both mothers and children have shown that CHD in the child is more likely to occur in those with mothers who are older, smoke during pregnancy, weigh more and have other pregnancy problems such as increased blood pressure and blood sugar levels.

Few studies have explored the extent to which these associations are causal for CHDs. To our knowledge there are no large-scale studies, with healthy control groups that have used new analysis methods to explore whether these associations are causal.

While it is widely recognised that a woman's lifestyle and experiences during pregnancy can influence the development and health of her child, remarkably little robust, scientific research supports this link. This limited evidence-base means that much of the current public health advice given to pregnant women is inconsistent, confusing and may potentially even be harmful. Additionally, there is emerging evidence that a man's behaviour can influence the health of their unborn children, but very little public health advice is currently offered to fathers-to-be, and the scientific evidence is even more scant.

This work aims to better understand how both maternal and paternal lifestyles in the prenatal period influence offspring health and social outcomes. Findings will highlight whether interventions to improve child health are best targeted at mothers, fathers or both parents.

The project makes use of detailed and diverse data that has already been collected through several large birth cohort studies from the UK, the Netherlands, Norway and Denmark; allowing access to rich data on hundreds of thousands of people.

Using these data, I will identify whether maternal and/or paternal smoking, alcohol intake, fat/sugar intake, caffeine and physical activity in the prenatal period are correlated with a range of childhood outcomes including birth weight, cleft lip/palate, body mass index, IQ and educational attainment.

I will then apply several state-of-the-art statistical methods to infer whether these parental health behaviours actually cause these offspring outcomes, or whether the observed correlations arise because of other "confounding" factors (e.g. socioeconomic position).

Where it looks like the relationship is causal, I will use several statistical methods to infer whether the outcome is caused by maternal behaviour, paternal behaviour, or a combination of the two.

Finally, I will explore the biological basis to these causal relationships by studying blood levels of DNA methylation (epigenetics), gene expression and metabolites in the children.

Findings from this work could help improve childhood health by providing better evidence about how the lifestyles of mothers and fathers shape childhood health and whether public health advice would be most effective if aimed at mothers, fathers or both parents.