Our research aims to uncover the lesser-known effects of childhood physical abuse on temporary changes in appetite and dietary habits. Previous studies in animal models have shown that exposure to physical trauma can lead to a temporary decrease in appetite, a finding that contrasts with the common belief and some human studies suggesting that abuse might lead to an increase in Body Mass Index (BMI) over time. Interestingly, our preliminary studies in older adult cohorts have also indicated a potential long-term increase in baseline BMI in individuals who experienced abuse during childhood. This project seeks to explore these dynamics in a child population using the rich data available from the ALSPAC cohort, focusing on immediate dietary responses and developmental outcomes following episodes of physical abuse.

Repair of adult tissues involves a complex interplay of several key cell lineages and inevitably leads to formation of a fibrotic collagenous scar, whereas embryonic tissues heal perfectly without any resulting scar deposition. This dramatic difference in repair efficiency between embryonic versus neonatal/adult tissues has been instrumental is guiding us towards potential causes of scarring. Indeed, we now believe that one major driver of scarring is the wound inflammatory response which doesn't initiate until a transition period in foetal development which, in turn, coincides with the developmental onset of tissue scarring. This insight has led us towards further mechanistic cell and molecular studies in model organisms, such as mouse and zebrafish, which help us better understand the scarring process and how one might modulate the wound inflammatory response in order to improve or prevent scarring.
Whilst these approaches, motivated by comparing embryonic versus adult healing, have been fruitful, it is clear that scarring is a complex, multifactorial response likely driven by a number of interacting mechanisms. We would like to use a conceptually similar comparative approach to gain further insights into the fundamental cell and molecular mechanisms of scarring by analysing differences in degree of scarring, not between embryo and adult, but rather across human adult populations since we know there is a range of “scarring phenotypes” from “minimal scarrer” to keloid scarring individuals. This use of human phenotypic variation in a population based, using an integrative approach: genetic variation (genetic association, genomewide association studies – GWAS), gene silencing (methylation, epigenetic association studies - EWAS), gene expression (transcriptomics, transcriptome-wide association studies - TWAS), has the potential not only to yield gene variant correlates of scarring, but also to point towards specific biological contributions to wound healing, both at causative and regulatory and levels.
The use of natural human experiments (e.g. Bacillus Calmette–Guérin (BCG) vaccination wound healing, Caesarean section (C-section) wound scarring and examples of human disease related fibroses) has never before been used for identification of scarring genes, even though the approach has proven to be powerful for discovering genes associated elsewhere with a wide variety of complex health outcomes (www.ebi.ac.uk/gwas/).
Fibrosis and scarring are closely related processes that involve the formation of excess fibrous connective tissue in organs or tissues, typically in response to injury or chronic inflammation. Fibrosis is defined as the accumulation of excess extracellular matrix components, particularly collagen, in and around damaged or inflamed tissues. Scarring is the end result of this process, where fibrous tissue replaces normal, functional tissue. In early stages, fibrosis may be reversible if the underlying cause is treated. However, advanced fibrosis and scarring are often permanent and irreversible. Fibrosis can occur in virtually any organ, including the lungs (pulmonary fibrosis), liver (cirrhosis), kidneys, heart, and skin. Fibrosis can be caused by various factors, including chronic inflammation, autoimmune diseases, infections, toxins, and radiation exposure. In some cases, the cause is unknown (idiopathic).
Furthermore, alongside a growing number of catalogues charting the results of human genetic association studies for health outcomes and intermediates there are tools able to consider (in frameworks of causal analysis) the existence of potentially causal and modifiable relationships between exposures of interest (e.g. inflammation, differential wound repair or scar) and health outcomes (e.g. wound healing, recovery, and disease).
Using human genetic data to help explore the potential of biological pathways contributing to health and disease in applied epidemiological designs is an approach that we have refined and developed and is an integrated approach to health research that has yielded important clinically relevant insights but has also indicated opportunities (e.g. associated signalling pathways for targeting) to unify basic science approaches with human population-based health data.
Understanding fibrosis and scarring is crucial for developing effective treatments for a wide range of chronic diseases. While advanced fibrosis remains a challenging medical problem, early intervention and new therapeutic approaches offer hope for improved outcomes.
In this study, we will use several strategies:
• We will investigate a more comprehensive number of variants in genes, using the topmed reference panel, given its diverse sample population from both Europeans and non-Europeans, with a total of 97,256 high-coverage genomes, to help identifying variants responsible for scarring and fibrosis.
• With a focus on those that are not that frequent in the population (frequency under 1%), we will investigate whole genome and exome sequencing variants, to help identifying variants responsible for scarring and fibrosis phenotypes.
• We will investigate regulation pathways of scarring and fibrosis phenotypes using the information provided by epigenetic and expression arrays.

• We’ve developed a model which simulated how the health state of our population will evolve over the next two decades (https://realworlddatascience.net/case-studies/posts/2024/05/08/dpm.html).
• This is crucially underpinned by what has happened to the health state of our population in the recent data.
• Essentially, the model ‘baseline scenario’ extrapolates these recent health state changes.
• However, we only have three years of data to determine these health state changes (the key restriction here is the all-important primary care data – other data sources do go back further).
• We are interested to understand how the health state of our population (or ALSPAC cohort thereof) has changed over a longer time period.
• This could help us better shape the ‘baseline scenario’ by extrapolating behaviour from a longer time period (not just the last three years).
• Crucial to this is data on individual attributes over time, which we source from our primary care data (2020+).
• Specifically, we use the individual’s Cambridge Multimorbidity Score (CMS), published here: https://doi.org/10.1503/cmaj.190757.
• This is calculated from 37 chronic conditions, see here: https://www.cmaj.ca/content/cmaj/suppl/2020/01/28/192.5.E107.DC1/190757-....
• Would it be possible to calculate these over time for the ALSPAC cohort, and thus calculate the CMS score for each individual, going back as long as possible?
• Additionally, we’d like to understand (1) inward/outward migration to the cohort, re patient characteristics, and (2) any detectable/inferable delays in diagnoses, re patient characteristics.
• The outputs of this will help us better understand health state changes in our population, and so better inform the ‘baseline scenario’ in our abovementioned simulation model.

Understanding genetic architecture in quantitative traits and complex diseases has been pivotal in biomedical research. Over years of effort and technological advancements, GWASs and sequencing studies have revealed that the genetic architecture of complex traits often comprise of a range of genetic components, including common variant, rare inherited and de novo variants. However, there remains an understudied aspect: the genetic architecture at individuals-level. For instance, individuals in the tails of trait distribution can be predominantly driven by common polygenic variants or by rare/ de novo variants with substantial effects.

In 2022, 63% of the world population actively used social media, spending a daily average of 2 hours and 27 minutes on these platforms. Despite its widespread adoption, our understanding of social media's impact on health and well-being remains limited, mostly due to lack of data availability on social media use. This proposal has two objectives. The first aims to study the relationship between social media use and mental health across different stages of life, considering various factors that influence social media exposure. The second investigates whether genetic susceptibilities towards mental disorders explain patterns of social media use. The exploration of these objectives will contribute towards advancing our comprehension of social media.

Obesity is the consequence of sustained positive energy balance over time where excess energy intake is the main pathway leading to obesity. Eating behaviors play a significant role in regulating food intake. Commonly used questionnaires measuring eating behaviors exhibit high correlations with each other and have been shown to share an underlying factor varying in severity - uncontrolled eating. Uncontrolled eating is associated with obesity and BMI cross-sectionally and longitudinally. Behavioral genetic models provide evidence that uncontrolled eating is a heritable trait with heritability estimates comparable to those seen in other personality or behavioral traits. Heritable variances of BMI and uncontrolled eating constructs show significant overlap. The genetic overlap between uncontrolled eating and BMI suggests that self-control of food intake may serve as the key mediating mechanism through which genetic factors influence differences in body weight. However, causality in this relationship is not necessarily unidirectional, as higher BMI could also lead to increased levels of uncontrolled eating. Another common construct in eating behaviour is food restriction, or restraint. Together with uncontrolled eating, they explain vast majority of variance in eating behaviour traits. The aim of this study is to investigate the genetic basis of uncontrolled eating and restrictive eating, examine their genetic correlations with various other phenotypes, and assess the causal relationship between uncontrolled eating, restraint and BMI.

Adverse exposures and events during the period of early-life growth are increasingly recognized as important to later-life disease risk. For example, up to 50% of chronic obstructive pulmonary disease (COPD) has been linked to impaired lung growth early in life. A challenge to understanding and exploiting this knowledge is the lack of a simple method to quantify the cumulative impact of adverse early-life growth conditions among adults. This amendment proposes to validate a novel quantitative index of early-life growth conditions that can be deployed across the lifespan.

A recent genome-wide association study of 5.4M adults reported a saturated map of genetic variants associated with height that accounts for over 90% of trait heritability and explains up to 45% of trait variance.5 Since human height is determined in part by genetics and in part by early-life growth conditions, it follows that the difference between measured height and genotype-predicted height (height-GaP) may represent a quantitative and cumulative index of adverse early-life growth conditions. In support of this hypothesis, analysis of UKBiobank data demonstrated that a larger height-GaP deficit was associated with several retrospectively ascertained early-life factors known to adversely affect growth, as well as subsequent all-cause and respiratory mortality.

Retrospective assessment of early-life growth conditions limit the strength of inferences that can made from these UKBiobank observations and motivate this amendment to our current ALSPAC proposal, which is already examining a genotype-based index and lung function.

Social inequalities exist across a range of social and health outcomes.3,4 They are harmful to individuals and have been exacerbated by the Covid-19 pandemic and cost-of-living crisis.5,6 The UK is one of the most unequal developed nations,7 where inequalities cost an estimated £106bn a year8 and are a largely agreed national policy priority (e.g., All Our Health; Levelling Up). Previous research has largely focussed on social or genetic influences in isolation, ignoring one of two truths; that population level social and cultural factors impacts individual outcomes, and that biology has an impact on human behaviour. This project will address the urgent issue of social inequality by examining how (dis)advantage is transmitted from parents to children. It will draw on methods from population genetics to improve evidence and complement existing social scientific research into the formation of inequalities.

We know that health behaviours during pregnancy (e.g., diet and physical activity) can impact on health outcomes (e.g., gestational diabetes). However, less is known about the impact of varying health behaviours on mental health outcomes (e.g., perinatal depression), and how the behaviour change of (reducing or increasing) these behaviours may impact perinatal mental health. This project will use traditional and causal epidemiological methods to investigate the impact of alcohol, tobacco and physical activity during pregnancy on maternal mental health using two longitudinal birth cohorts, to highlight pregnancy timepoints to reduce harm.