Dr Samuel Leighton
- Affiliate (School of Health & Wellbeing)
email:
Samuel.Leighton@gla.systa-s.com
Clarice Pears Building, 90 Byres Road, Glasgow, G12 8TB
Biography
Dr Samuel Leighton is a Consultant General Adult Psychiatrist at Leverndale Hospital, NHS Greater Glasgow & Clyde, specialising in psychosis and severe mental illness. He holds membership of both the Royal College of Psychiatrists (MRCPsych) and the Royal College of Physicians (MRCP(UK)), reflecting his dual training and commitment to excellence in both medicine and psychiatry. His clinical practice spans early intervention in psychosis and the management of complex mental and physical comorbidity.
Dr Leighton is recognised for advancing clinical prediction modelling, machine learning, and causal inference in psychiatry. His research career began with creating digital tools, including creating NHS Greater Glasgow & Clyde’s "GP Antibiotics" mobile app, and developing expertise in statistical programming. He has since contributed a series of high‑impact first‑author publications across major journals such as Molecular Psychiatry, The British Journal of Psychiatry, The Lancet Digital Health, Schizophrenia Bulletin Open, and Journal of Neurology, Neurosurgery & Psychiatry. He is a strong supporter of open-access research and always publishes all his analytical code online.
His early work demonstrated that chemokine CXCL8 distinguishes between depressed and non‑depressed individuals. This was followed by the first external validation study predicting outcomes in first‑episode psychosis, which generated international collaboration. Working with colleagues across Edinburgh, Birmingham, and Copenhagen, he helped develop validated prediction models for multiple domains of psychosis recovery. His Chief Scientist Office–funded Clinical Academic PhD Fellowship further produced leading studies on prediction in psychosis, methodological appraisal of existing models, and the association between delirium and later dementia risk.
Dr Leighton’s work has influenced Scottish Government policy on Early Intervention in Psychosis outcome measures in pathfinder sites across Scotland, and he continues to contribute nationally through Healthcare Improvement Scotland. He co‑authored a book chapter on AI‑driven clinical decision support systems and is an active contributor in the emerging field of causal prediction modelling, with perspective pieces and methodological articles shaping debate around actionability in precision psychiatry. Locally he works closely with Dr Fani Deligianni in the School of Computing Science. He also collaborates widely across the UK and internationally including with Dr Rajeev Krishnadas' group at the University of Cambridge and the Causality in Healthcare AI Hub led by Prof Sotos Tsaftaris at the University of Edinburgh. He regularly supervises doctoral, medical, and undergraduate research students.
Alongside his academic and clinical commitments, Dr Leighton co‑founded Ben Èideann Ltd, producers of internationally distributed Kosher Scotch whisky, reflecting his wider interest in technology, entrepreneurship, and the practical translation of ideas into real‑world impact. He also serves as a Director of the Scottish Jewish Archives Centre, demonstrating a longstanding commitment to historical research, heritage preservation, and community engagement.
Research interests
Causal prediction and precision psychiatry
Developing fully actionable clinical prediction models that identify intervenable causal factors to support personalised treatment decisions in psychosis and severe mental illness. This includes methodological innovation using causal mediation, counterfactual prediction, and modern evaluation frameworks.
Psychosis-related cardiometabolic risk
Investigating mechanisms and prevention of antipsychotic-induced obesity, insulin resistance, and inflammation, and developing tools to target interventions to individuals at highest risk at the right time.
Machine learning and clinical prediction modelling
Designing, validating, and critically evaluating multivariable prediction models for outcomes such as remission, recovery, and quality of life in first‑episode psychosis. Particular interests include methodological rigour, reproducibility, and external validation across diverse cohorts.
Target Trial Emulation and real-world evidence
Applying causal methods such as Target Trial Emulation to routinely collected healthcare data to examine real-world treatment effects relevant to psychiatry, especially for cardiometabolic interventions in psychosis.
Digital health and medical AI implementation
Translating research into clinical tools, including development of mobile apps and contributions to medical‑device–regulated prediction systems. Strong commitment to open science, transparent modelling, and safe deployment of clinical decision support software.
Big‑data psychiatry and long‑term outcomes
Using electronic health records and population-scale datasets to investigate trajectories of severe mental illness, including cognitive decline and dementia risk following delirium and psychosis.
Research culture, teaching, and mentorship
Supporting the research community through supervision of doctoral and undergraduate students, mentoring early‑career clinicians, delivering workshops on prediction modelling and machine learning, and peer reviewing for leading journals.
Grants
Grants and Awards listed are those received whilst working with the University of Glasgow.
- CSO clinical training grant - psychotic disorders
Chief Scientist Office
2019 - 2022
Supervision
I welcome enquiries from prospective MSc and doctoral students interested in projects on risk prediction and causal machine learning in severe mental illness. My current research focuses on obesity and cardiometabolic disease in psychosis, and I am keen to support students wishing to develop skills in prediction modelling, causal inference, and the application of advanced analytical methods to clinically meaningful problems in psychiatry.