Research Opportunities

Population-based cohorts have identified major modifiable risk factors for cardio-metabolic diseases, such as adiposity and physical activity, but the patterns and relevance of these factors varies greatly across populations, and previous evidence is predominantly from high-income countries. There is a high burden of cardio-metabolic diseases in South and Southeast Asian populations. However, the underlying mechanisms are yet to be fully elucidated, with previous evidence suggesting ethnically divergent body fat and muscle mass distribution to be a determining factor. Furthermore, physical activity has a complex relationship with body composition, and different patterns of physical activity between high- and low-/middle-income countries and between urban and rural areas might be an independent or explanatory factor in associations with cardio-metabolic diseases.

The objectives of this DPhil project may be to explore associations between different measures of body composition with objective measures of physical activity between populations and their individual and joint associations with cardio-metabolic diseases across different ethnicities, using data from different large-scale prospective studies. 

This project will use data from three large prospective studies: the Indian Study of Healthy Ageing (ISHA), the Malaysian Cohort, and the South and Southeast Asian participants of the UK Biobank. It will provide unique opportunity for novel insights into disease risks and aetiology to inform global non-communicable disease control and prevention efforts, and the student will have the chance to work collaboratively across the Global Populations Studies Group led by Prof Sarah Lewington, the Oxford Wearables group led by Prof Aiden Doherty, and the Oxford Centre for Diabetes, Endocrinology and Metabolism led by Prof Fredrik Karpe.

Infections pose a major risk to health globally. Antimicrobial resistance (AMR) threatens effective treatment of infection and healthcare associated infections (HCAIs) impact more than 10% of all hospital patients.

Advances in data availability and new artificial intelligence (AI) methods offer the chance to develop:
 

• More responsive, comprehensive, and automated HCAI/AMR surveillance generating better breadth and depth of intelligence to drive action and changes in practice to protect diverse populations at local, regional, and national levels.
• Predictive tools to improve care of individual patients and combat AMR.
• Methods, infrastructure and skills to optimally use rapidly-evolving electronic healthcare record and patient-contributed data, and emerging AI technologies. 

Several possible projects are available, including:
 

• Developing/testing automated electronic surveillance approaches for rapidly detecting changes in infections and identifying at-risk populations; and deploying these tools in hospitals and national systems
• Extending and piloting in hospitals predictions of personal AMR risk to optimise infection treatment, prevention and control, developing generalisable methods that can update over time/to new locations, and approaches for safely implementing them
• Pre-emptive surveillance, investigating which metrics of hospital processes (e.g. isolation/screening/diagnostic use/cleaning) are associated with HCAI/AMR to inform prevention

Large-scale electronic health record data can potentially answer a far greater number of questions about infection management than traditional epidemiological studies using questionnaires. Their volume and scale are continuously increasing as larger amounts of healthcare data are linked and de-identified for research. Examples and challenges include
 

• Can we identify a wider range of risk factors for infection to target interventions? The sheer number of factors that could be considered, many with substantial amounts of missing data, poses challenges to traditional epidemiological approaches. A recent novel statistical analysis approach called ‘doublethink’ has been proposed which could be applied to a range of microbiologically and/or syndromically defined infections to identify novel populations to target to reduce infection risks, and compared with other methods including machine learning.
• Can we work out how best to use diagnostic tests for infection, widely considered to be a key tool to improve antimicrobial stewardship, in the real-world? In what patient populations are they and should they be used, how often, and what are their ultimate effects on both antimicrobials prescribed and patient outcomes? 
   

Projects can exploit an existing large datawarehouse of de-identified individual patient data, the Infections in Oxfordshire Database. They would suit students interested in infections/antimicrobial usage and coding, who wish to gain experience in design of studies using healthcare records to answer real-world questions. There will be opportunities to learn how to manage and use large-scale electronic health record data, and apply a range of quantitative methods including novel causal epidemiological methods.

A population approach to disease prevention aims to shift the distributions of risk factors such as diet, physical activity and adiposity by changing the environments — economic, digital, physical and social — that influence everyone’s behaviour. To inform the next generation of environmental and policy strategies to prevent non-communicable diseases, we need to better understand the ways in which these interventions work, or could work, and the extent to which these are transferable between populations and contexts. This implies a need to improve our causal understanding of the disease prevention pathways linking a variety of environmental exposures via more proximal behavioural or metabolic outcomes to a variety of clinical outcomes.  

Our research programme offers the scope for a variety of PhD projects using the methods of epidemiology, natural experimental evaluation, qualitative causal process observation or evidence synthesis, singly or in combination, to investigate the mechanisms by which interventions can more effectively and equitably shift population dietary and physical activity patterns. Potential lines of inquiry include:  
•    Mechanism-focused systematic reviews of potential intervention strategies, e.g. using EBM+ or similar methods  
•    Epidemiological analyses to clarify causal pathways for interventions and novel intervention targets in the food or transport environments  
•    Investigating causal mechanisms in intervention studies in the food or transport environments.
•    Interactions between food and activity environment exposures 
•    Using routinely available georeferenced data on environmental changes and linking with well characterised cohorts to conduct statistical and spatial analyses

Supervisor(s): Jean Adams, Louise Foley, David Ogilvie, Jenna Panter, Martin White  

Programme: Population Health Interventions  

Traditional epidemic models attempt to fit the spread of infectious disease across populations. However, populations are ultimately an abstraction. At a more fundamental level diseases spread between individuals, not populations. This reflection motivates the network approach. Network approaches in epidemiology help to sharpen our intuitions. For example, they have shown that, in principle, heterogeneity can sustain a disease that would on average go extinct, or asymmetric transmission can alter the effectiveness of interventions. Unfortunately, applying network epidemic modelling directly to data is a challenge since we don't know the relevant contact network, or even how to measure it.   This project will develop computational and mathematical methods for fitting networked epidemic models to empirical epidemiologic data in the regime of high uncertainty. Even when data are of low quality, it may be possible to extract meaningful insights. We may never know exactly who was in contact with whom, or who was infected first, but we may still be able to determine the mode of transmission. And, in the end, this latter question is the scientific question of interest. Algorithms will be derived and validated to lay foundations for data driven approaches.  This project has potential applications for disease control and optimization of interventions. We will focus on whether contact traces – noisy and incomplete at the best of times – contain sufficient signal for answering basic research questions, so that future interventions are optimized. If they do, how do we extract this signal? If not, what modifications would make contact traces more informative without harming their immediate goals?

Globally, HIV and schistosomiasis are leading causes of death due to infectious diseases. Despite available interventions, the infections remain uncontrolled in low-income settings causing acute and chronic morbidities. Intestinal schistosomiasis is caused by a parasitic blood fluke, most commonly of the species Schistosoma mansoni, and is predominantly found in sub-Saharan Africa. Chronic infections lead to advanced disease including liver fibrosis, portal hypertension, upper gastrointestinal tract bleeding, and severe anaemia. In the context of coinfections, severe clinical outcomes including death may be likely due to immune failure, interactions related to general fibrosis, and responses to starting antiretroviral therapy. In this project, you will have the opportunity to work with cutting-edge statistical and big data approaches alongside state-of-the art immunology to examine disease progression in the context of schistosome and HIV coinfections in arguably some of the poorest settings worldwide.

The group of Associate Prof. Chami studies schistosomiasis evaluating transmission, clinical outcomes, and treatment strategies, especially for liver fibrosis, in the SchistoTrack Cohort with the Uganda Ministry of Health. This Cohort is the largest individual-based cohort tracking individuals prospectively in the context of schistosomiasis. At Oxford, students can get exposure to computational, big data approaches to clinical epidemiology and field experience in global health research.

The group of Dr. Sereti studies HIV immune pathogenesis with a focus on inflammatory complications related to HIV and coinfections. Studies on biomarkers and how they may assist in identifying early people with HIV who may develop inflammatory and other adverse complications is currently an active area of investigation in the lab as they can also inform disease pathogenesis and new targeted interventions.

At the NIH, students can get experience in immunology research (wet lab) with optional exposure to complicated cases within a clinical setting.

In this project you will use state-of-the-art viral sequencing data, combined with epidemiological data and mathematical modeling, to create an integrated understanding of HIV transmission. HIV places an enormous burden on global health. Implementing treatment and interventions can save millions of lives, but to do this effectively requires us to be able to predict the outcome of interventions, and to be able to accurately assess how well they are working once implemented. For HIV, these efforts are hampered by long durations of infections, and rapid within-host viral evolution during infection, meaning the virus an individual is infected with is unlikely to be the same as any viruses they go on to transmit.
 
For this project, you will identify individuals enrolled in the Rakai Community Cohort Project, based in Uganda, who are part of possible transmission chains, and for whom multiple blood samples are available throughout infection and at the time of transmission. These samples will be sequenced using state-of-the-art technology developed at the University of Oxford enabling the sequencing of thousands of whole virus genomes per sample, without the need to break the viral genomes into short fragments (whole-haplotype deep sequencing). Using this data, you will comprehensively characterize viral diversity during infection and at the point of transmission.

Key questions you will tackle are:
 
- Do ‘founder-like’ viruses (similar to those that initiated infection) persist during chronic infection?
- Is there a consistent pattern of evolution towards population consensus virus?
- Are ‘founder-like’ viruses, or ‘consensus-like’ viruses more likely to be transmitted?
- Does the transmission of drug-resistant virus depend on the history of the transmitting partner?

*This project is available for the 2021 Oxford-NIH Pilot Programme*