Urbanicity and psychosis- are cities bad for mental health? A large-scale data linkage study using electronic health records

Lead Supervisor
Dr Jayati Das-Munshi
Senior Lecturer / Clinician ScientistInstitute of Psychiatry, Psychology and Neuroscience (IoPPN), King’s College London

Professor Craig Morgan, Dr Margaret Heslin
Health Service and Population Research Department, King’s College London

Project Details

Mitochondria are involved in a wide range of fundamental cellular processes, from cellular energy production to thermogenesis, lipid biosynthesis and cell death. So wide ranging are their functions, that diseases associated with mitochondria manifest across almost all tissues, often affecting multiple organs simultaneously. Mitochondria have their own genome that codes for a small number of proteins that form core components of the electron transport chain, thus modulating key metabolic processes. However, proteins encoded in the nuclear genome control many of the processes occurring in mitochondria, thus making interactions between the nuclear and mitochondrial genomes key to health. 

A growing body of evidence indicates that strong social risk factors, particularly related to urban built environments, may be associated with an increased risk of psychosis. Urban areas may be “riskier” as they may represent areas of higher deprivation, a heightened risk of social isolation, reduced social support and may also be less cohesive and socially fragmented. These social risk factors may represent environmental insults to the brain in susceptible individuals, however to date, it is unclear which of these aspects play a role in the heightened risk of psychosis in urban areas. 

This project will use a large dataset with clinical data from South London & Maudsley Trust  linked to individual-level data from UK census 2011 (a rich source of social and demographic information) to determine social predictors in the urban built environment (household poverty, social isolation, family support, neighbourhood social fragmentation, ethnicity/ migration status) with the onset of psychosis. This is the first data linkage of its kind in the UK, bring together rich clinical data from one of the largest mental healthcare providers in Europe, with detailed individual-level social data. The catchment area of the study serves an ethnically diverse population of 1.3 million people in an urban part of southeast London. The project will allow the student to assess the social risks for the onset of psychosis in a cohort of approximately 20,000 people linked to up to 5 population controls. This will be the first time that a project of this kind has been undertaken using UK data and we anticipate rich insights informing our understanding relating to the social determinants of psychosis. 

The project would suit a candidate with a strong quantitative background (e.g. statistics / epidemiology) wishing to develop expertise within data sciences, focusing on the social determinants of severe mental illness. We would welcome applications from students with a grounding in quantitative social sciences willing to work within an interdisciplinary capacity. The studentship will include analytical methods training as well as a rich programme of interdisciplinary and transferable skills training provided through the DTP and the newly established Centre for Society and Mental Health. 

Brief timeline:
Year 1: Undertake systematic reviews, gain approvals, data cleaning / commence analyses, upgrade to PhD from MPhil at 9 months; 
Year 2: Conduct analyses; training, present findings at relevant conferences/ dissemination; 
Year 3: Finalise analyses for publication and PhD submission.


All ethical approvals are in place (NHS Research Ethics Committee approvals) and approvals from National Statistician’s Data Ethics Advisory Committee (NSDEC) (for ONS data) and  Clinical Record Interactive Search (CRIS) oversight committee (for clinical data from South London and Maudsley (SLAM) Trust) to access and analyse linked data are in place. The student will be supported by supervisors to gain ‘approved researcher status’ to access the linked data. Linked data will be accessible from Office for National Statistics (ONS) secure research sites in central London and unlinked CRIS data will be accessible from within the SLAM firewall. The student will be supported to gain all approvals in order to be able to access the data and we do not anticipate any delays. 


severe mental illness; data linkage; social determinants; urbanicity; psychosis