Background
Antidepressants are among the most commonly prescribed medications for depression, yet many patients experience variable or incomplete treatment responses. Identifying why individuals respond differently remains a major challenge in mental health research and clinical care. Current evidence is often limited by short follow-up periods, narrow clinical data, and insufficient integration of biological and environmental factors. Advances in health data science and genomics now provide an opportunity to better understand the determinants of treatment outcomes at scale.
Novelty & Importance
This project will take a comprehensive, data-driven approach to investigate antidepressant treatment trajectories using linked electronic health records (EHRs) and genetic data. By combining primary care data from the Clinical Practice Research Datalink (CPRD) with the UK Biobank, the study will analyse large-scale longitudinal data enriched with genetic, clinical, and environmental information. Natural language processing (NLP) techniques will unlock insights from unstructured text, while secondary care EHRs will be used to validate findings—clarifying whether treatment switching reflects lack of efficacy, side effects, or adherence issues. This integrated, multimodal approach goes beyond existing studies, enabling more precise modelling of individual differences in antidepressant response.
Aims & Objectives
The project aims to identify the clinical, environmental, and genetic factors that shape antidepressant treatment outcomes. Specifically, it will:
1. Build a comprehensive longitudinal dataset of antidepressant prescribing trajectories linked to genetic and phenotypic data.
2. Develop algorithmic and NLP pipelines (e.g., large language models) to measure treatment response, resistance, and switching behaviours.
3. Validate findings using secondary care records to better understand the drivers of switching and nonresponse.
4. Apply advanced causal inference methods to identify factors that directly influence outcomes.
This research will generate new evidence to support more personalised and effective antidepressant prescribing, improving care for people living with depression.

