Religious/spiritual beliefs and behaviors and health in the ALSPAC cohort (2021 - present)
Despite the centrality of religion to many people’s everyday lives, examining how religious/spiritual beliefs and behaviours (RSBB) shape human health (both mental and physical), and the factors shaping an individuals RSBBs, remains understudied. I am a Senior Research Associate on a large project led by Prof. Jean Golding which aims to explore these questions using over 30 years of data collected as part of the ALSPAC cohort.
Funding: John Templeton Foundation
Synthetic Electronic Health Record (EHR) data to support research (2024)
This project aims to explore the potential for synthetic data to facilitate research using synthetic EHR data, focusing specifically on NHS secondary care Hospital Episode Statistics. Due to the personal and sensitive nature of EHRs, the confidentiality of these data is protected by strict regulatory frameworks, and accessing EHRs can take months or even years; this remains a major barrier for researchers gaining timely access to real world data. Synthetic EHR data could assist researchers by allowing them to explore synthetic copies of these EHR data to develop code and analysis pipelines ahead of gaining access to the actual data. This small seedcorn project aims to develop a roadmap of the key requirements for synthesising EHRs and the needs of potential users of these data.
Funding: Jean Golding Institute Seedcorn Funding (University of Bristol)
Early-life access to green space and child cardiometabolic outcomes (2021 - 2023)
As part of the LifeCycle project, this study aims to explore how access to green space in early childhood is associated with later child cardiometabolic outcomes (i.e., BMI/obesity and blood pressure). Although some studies have suggested that access to green space is associated with lower rates of obesity and high blood pressure in children, the evidence is mixed. We aim to assess this question in two UK birth cohorts (ALSPAC) and Born in Bradford), using a structured life course approach. This might sound a bit intimidating, but it is just a method to explore how repeated exposures (in this case, access to green space) are associated with a later outcome - e.g., whether there is a critical period, sensitive period, accumulative of risk, etc.
Funding: EU Horizon 2020 Research and Innovation Programme
ALSPAC COVID-19 data collections (2020 - 2021)
While working at ALSPAC, I was involved in processing much of the COVID-19 questionnaire and antibody samples data collections. Descriptions of these data have been published as ‘data notes’ in the journal Wellcome Open Science, and are intended to help researchers using this data when conducting COVID-19 work in the ALSPAC cohort.
Funding: Wellcome Trust, Medical Research Council and University of Bristol
COVID-19, cardiovascular health and selection bias (2020 - present)
This project aims to investigate how selection bias (e.g., bias due to non-random selection into an analysis sample; also know as ‘collider bias’ - click here for a primer) may impact associations between COVID-19 infection/severity/death and cardiovascular outcomes. For instance, in ALSPAC we know that completion of COVID-19 questionnaires are a biased sample of all ALSPAC participants, which may cause bias in downstream analyses using this data. In addition to describing these selection pressures, this project also explores the magnitude of selection bias, and potential methods to overcome this bias (e.g., multiple imputation, inverse-probability weighting, etc.).
Funding: British Heart Foundation
Predicting common mental disorders from primary care records in ALSPAC (2019 - 2021)
The project aimed to explore common mental disorders (CMDs; depression and anxiety) in ALSPAC using both validated questionnaire data and primary care records over adolescence and yound adulthood (age 15-23). We aimed to: 1) explore how well GP records of CMDs predict questionnaire-diagnoses; 2) examine factors predicting identification in GP records, of those with a questionnaire-derived diagnosis; and 3) construct prediction models using all available primary care data to predict questionnaire-derived CMDs. Overall, we found that GPs frequently miss individuals with CMDs (sensitivities from 1.6 to 34.0%), although individuals with more severe symptoms are more likely to be detected by GPs. The final paper can be found here.
Funding: Medical Research Council
Cooperative dynamics among Agta hunter-gatherers (2013-2017)
My PhD was part of the Hunter-Gatherer Resilience Project (led my Prof Ruth Mace and Prof Andrea Migliano), which was a large, multi-site, project to explore how the world’s remaining forager populations are able to adapt to a changing world and display resilience in their way of life. My project specifically focused on the Filipino Agta, where I explored cooperative dynamics among this group from an evolutionary perspective (e.g., how much they cooperated, who they shared with, and why). To assess this, I developed two simple experimental measures of cooperation; one assessing who the Agta shared resources with, and another assessing who they took resources from. A few key findings are: i) levels of cooperation varied widely between camps (including one camp where no-one shared anything!), with levels of cooperation predicted predominantly by camp stability, with repeated interactions associated with higher levels of cooperation; ii) contrary to many theoretical models, we found that Agta were more likely to share with less cooperative camp-mates, displaying needs-based sharing, potentially as a system of long-term reciprocity; and iii) storytelling may help promote cooperation, as most stories told by hunter-gatherers were about social norms, Agta camps with more-skilled storytellers were associated with higher levels of cooperation, and skilled storytellers were preffered social partners and had increased reproductive success.
Funding: Leverhulme Trust