The Wellcome Trust Longitudinal Population Studies (LPS) Strategy encourages integration of data from a variety of resources, including commercial data on shopping habits; and states that research methods are needed to underpin efficient linkages of multiple datasets. Our project addresses this by building on core Avon Longitudinal Study of Parents and Children transactional data linkage program, seeking to develop a methodology for efficient integration of transactional data into LPS research.

This 14 months project will create a methodology, a software package, guidelines and three proof-of-concept studies to the use of linked transactional data in LPS. The guidelines will include the description of the full cycle of data linkage from workshops and surveys with participants about feasibly and acceptability of such data linkage, through consent campaigns, data linkage and a framework of working with linked dataset. As a part of the project we will create a software package that can integrate the data into LPS and make it usable for LPS researchers. Further, we will run exploratory and dissemination workshops with LPS researchers to first investigate utility of data linkage and second provide them with the tools to set up transactional data linkages and conduct research with transactional data. The framework of using linked transactional data with LPS developed in this research project can be rolled out to be used in other cohort studies (e.g., through Cohort and Longitudinal Studies Enhancement Resources, CLOSER), making possible a harmonised longitudinal population studies approach for studying public health with transactional data.

The project will examine the longitudinal associations of worry and anxiety disorders with anorexia nervosa (AN), and seeks to address which of worry and anxiety disorders is most important in explaining later AN development. The project will supplement analyses completed in another sample, with a view to informing theories of AN development and prevention interventions.

Childhood recurrent abdominal pain (RAP) is very common, affecting around 10% of children, and highly co-morbid with anxiety. The link between anxiety and eating disorders is well established, and clinically I have noticed that many adolescents with eating disorders suffered from RAP in childhood. A theory underpinning childhood RAP is that children develop a fear of normal somatic gut sensations, and similarly one of the core features of eating disorders, such as anorexia nervosa and bulimia, is a fear of somatic sensations. I would like to analyse the ALSPAC dataset to assess whether childhood RAP predicts later eating disorder cognitions and behaviours.

It is highly likely that there is a broad and reproducible footprint of variation in body composition (lean and fat mass) on the circulating metabolome. Technological developments are now such to allow a broad characterisation of circulating metabolome using an approach called mass spectrometry which can allow for an analysis of the cell derived products of metabolic processes by exposures of interest. Here the exposure of interest is body mass index (BMI) - or weight given height.

The problem with looking at the direct relationship between BMI and measures of the metabolome is that BMI is related to many other biological and non-biological things. Therefore, when one analyses the relationship between BMI measured directly and health outcomes (like metabolites), then it is the case that BMI and those other things are jointly assessed as they are correlated. To avoid this, it is possible to conduct randomised controlled trials which shuffle up these correlations before assessing the impact of a risk factor - however you cannot randomise to BMI. Therefore, by taking genetically predicted BMI, it is possible to avoid this problem and to assess the impact of BMI variation on metabolite measures in ALSPAC.

We will do this by using the ALSPAC genetic data to generate a score for the burden of genetic contributions to higher and lower BMI, then select samples from storage for metabolite measurement at a company called Metabolon.

The main analysis of the data generated by this experiment will be focused on characterising the impact of BMI on the human metabolome. In principle, the two groups generated in this sample (high and low BMI as prescribed by being at the ends of a distribution of BMI genetic risk) will be compared.

Across the developed world, employment has declined in middle-wage `routine’ jobs and increased strongly in jobs requiring a degree. Recent evidence from the U.S. and Sweden indicates this change is associated with an increase in the demand for jobs requiring social skills. Over the last 30 years the UK labour market has witnessed a well-documented increase in the supply of graduates while the occupational structure has shifted towards managerial jobs. However, there is no strong evidence as yet that these changes have benefitted workers with higher social abilities.
In this project we will investigate the labour market returns to social skills in the context of the UK. Thanks to the availability of long-running cohort studies, the UK offers an opportunity to examine unique data on individual’s characteristics, such as cognitive, socio-emotional, and physical abilities. ALSPAC is particularly suitable to our purposes as it also collects genetic information which can be used to carry out a Mendelian randomization research design to obtain causal estimates of the impact of social skills on wages or employment.
The changing labour market is at the centre of some of the most important social policy debates worldwide, especially those focused on the effect of artificial intelligence on the demand for different types of skills. We will contribute to these debates by providing novel evidence of the relationship between different types of skills and labour market outcomes.

Glycosuria is a common disorder of pregnancy characterised by the presence of glucose in urine. Glycosuria may be caused by an increase in blood glucose such that the renal tubules are overwhelmed and complete reabsorption of presented glucose is not possible, a benign lowering of the renal threshold, or inhibition of renal tubule reabsorption 1,2. Primarily a result of a lower renal threshold, which is a consequence of increased renal blood flow, glycosuria during pregnancy may be a product of gestational diabetes and shows evidence as an indicator for later life adversity such as offspring risk of obesity 1,3.

Genes predisposing to type 2 diabetes have been associated with gestational diabetes 4 and there is growing evidence that impaired glucose regulation in pregnancy, including glycosuria, is associated with negative metabolic outcomes for offspring such as non-alcoholic fatty liver disease 5. Patients who develop gestational diabetes are at increased risk of gestational diabetes in future pregnancy, as well as being at risk of developing type 2 diabetes over the life-course 4. The presence of glycosuria may indicate future adverse outcomes in pregnancy and the life-course. To identify genetic variants associated with such a phenotype and to provide greater understanding of disease development, progression, and related complications we will set out to conduct a genome-wide association study (GWAS) of glycosuria during pregnancy, and investigate the genetic overlap between identified loci and multiple diabetic-related traits.

1. Lawlor, D. A. et al. Association of existing diabetes, gestational diabetes and glycosuria in pregnancy with macrosomia and offspring body mass index, waist and fat mass in later childhood: findings from a prospective pregnancy cohort. Diabetologia 53, 89-97, doi:10.1007/s00125-009-1560-z (2010).
2. Ferrannini, E. Learning From Glycosuria. Diabetes 60, 695-696, doi:10.2337/db10-1667 (2011).
3. Patel, S. et al. Associations of Gestational Diabetes, Existing Diabetes, and Glycosuria With Offspring Obesity and Cardiometabolic Outcomes. Diabetes Care 35, 63-71, doi:10.2337/dc11-1633 (2012).

Despite increasingly gender egalitarian ideals, remarkable gender gaps persist in children’s early development and subsequent education, labour market and family choices. The question as to what extent biological factors, such as (prenatal) sex hormones, moderate or set a limit to social influences has received growing attention over the past decade. Yet our understanding of these relationships remains patchy and much remains to be explored. This project takes an interdisciplinary approach and integrates theoretical perspectives from sociology, social and clinical biopsychology and neuroscience. To-date, the number of empirical studies, which considered at least one measure of sex hormones and at least one set of socialisation factors or applied other innovative designs to shed light on the interaction of hormonal and socialisation influences, are limited (e.g., Berenbaum, Bryk and Beltz 2012; Davis and Risman 2015; Hines et al. 2002; Udry 2000). Drawing on the Avon Longitudinal Study of Parents and Children (ALSPAC), we will investigate how prenatal sex hormones and gendered parental environments and role modelling influence children’s interests in school subjects at different ages and the gender (a)typicality of their occupational aspirations and actual choices during adolescence and young adulthood. We extend the current literature by using prospective reports by parents and children over longer periods from early childhood to young adulthood. The findings will contribute to a better understanding as to what extent malleable environmental conditions may be more or less effective in altering gender differences in children’s identities, academic and occupational aspirations and choices.

The health of pregnant mothers can influence the health of her offspring across their entire lifespan. In high income countries, as many as 1 of every 10 mothers experiences diabetes during pregnancy. The impact of exposure to diabetes during pregnancy on the heart health of offspring is poorly understood. We recently found that individuals exposed to diabetes in the womb were 2-3 times more likely to develop high blood pressure and heart disease in their teens and early 20s, compared to individuals not exposed to diabetes in the womb. We want to determine if this is something we can detect in childhood or adolescence and possible reasons why maternal diabetes could predispose children and adolescents to high blood pressure and heart disease later in life.

Genes help determine whether an individual is left- or right-handed, but the genes that contribute to handedness are mostly unknown. The goal of our study is to identify genetic factors that determine handedness by comparing genetic information from left- and right-handed individuals. We also plan to compare genetic data for handedness to data from other traits and diseases to determine whether handedness is correlated with later risk of disease.