Prescription medication and brain health

expernal supervisor

Bernd Taschler, Nuffield Department of Clinical Neurosciences

Background

Prescription drug use becomes nearly universal in later life, yet the effects of many commonly prescribed medications on brain health remain largely unknown. While some classes of drugs, such as anticholinergics, have been linked to cognitive decline, the broader causal impact of medications prescribed for other medical conditions is still poorly understood. Understanding these relationships is critical to avoid iatrogenic harm and support evidence-based risk/benefit decision-making in clinical care, particularly as we move toward more personalised approaches to medicine. 

This project leverages the UK Biobank (UKB), a rich resource offering extensive data on medication use, including self-reported baseline and brain imaging information, along with linked primary care records that provide detailed prescription data. These datasets enable the investigation of longitudinal trends in medication prescription and their potential long-term effects on brain health. More than 6,000 different drugs were self-reported by participants at baseline, creating a unique opportunity to explore these complex relationships at scale.

The goal of this project is to use extensive data from the UK Biobank imaging study (N>60,000) to explore how prescribed medications impact brain health. This high-risk/high-reward project could identify medications given for different indications which harm brain health, with substantial implications for clinical care.

Depending on the student’s interests, possible topics could include: using data-driven methods to quantify treatment-related brain effects, emulating target trials with observational data, incorporating genetics/multi-omics in a Mendelian randomisation framework to infer causal effects, using advanced AI/ML methods to investigate and predict relationships between prescription drug use and brain health. 

RESEARCH EXPERIENCE, RESEARCH METHODS AND TRAINING

Specific training will be provided in-house in neuroimaging analysis, statistics, Mendelian randomisation, scientific writing and presentation, with specific training courses where relevant.

The student will be joining strong interdisciplinary research teams based in the Big Data Institute with expertise in clinical medicine, population health, brain imaging analysis, neuroimaging statistics, and advanced machine learning. 

PROSPECTIVE  STUDENT

The ideal candidate will have good statistical software and programming skills (e.g. in R), a Master's degree in either statistics or epidemiology, and an interest in using large datasets, statistical/ML methods and genetics to answer clinically-relevant questions.