Skip to navigation Skip to main content Skip to footer

Approved research

Mental health risk factors and outcomes: elucidating reciprocal causal relationships

Principal Investigator: Dr Jean-Baptiste Pingault
Approved Research ID: 29819
Approval date: July 17th 2018

Lay summary

Health Conditions: The project will focus on mental health, including substance use and related outcomes (e.g. hospital self-harm records). Aim: The aetiology of mental health can be characterised as a network of reciprocal relationships between (i) distinct mental health phenotypes (e.g. depression, substance abuse, and self-harm) and (ii) multiple risk factors (e.g. education, brain structure). Recent advances in genetic epidemiology enable stronger causal inference in observational studies. Building on these techniques, the project will aim to map out the network of causal relationships between those diverse mental health-related phenotypes and multiple relevant risk factors. Implications: Establishing the directionality, effect size, and causal status of reciprocal relationships between different mental health phenotypes and between those phenotypes and relevant risk factors, will provide useful information to clinicians for risk assessment. In addition, it will provide new insights to design more targeted public health interventions. Suitability of the UK Biobank. The project will require assessments of relevant mental health outcomes and risk factors, genome-wide data, and brain structure data. In addition, the methods that will be implemented require a very large sample. Therefore, the UK Biobank matches perfectly the project?s requirements. The present project will study the reciprocal relationships between mental health phenotypes (e.g. depression to self-harm) as well as the relationships between those phenotypes and a variety of risk factors. The research will first focus on predefined relationships of interest (examples in methodology). Building on this first step, we will then examine multiple risk factors and mental health outcomes simultaneously, mapping out the network of their reciprocal relationships. The methods, described below, are familiar to the PI and collaborators and required analyses will be run on a cluster at the host institution, which already hosts UK Biobank data. Full cohort.

Scope extension: Scope. Our current project aims to map out the network of causal relationships between mental health-related phenotypes and multiple relevant risk factors, using genetically informed causal inference methods. Here, we aim to extend this project in several ways. First, at the phenotypic level, we will extend the analyses we have conducted for mental health to phenome-wide analyses, including physical health phenotypes, lifestyles, cognitive and biological markers. Second, we plan to broaden the scope of our analyses to further explore pathways to disease. Examples include mediation analyses of polygenic effects on endpoint outcomes via intermediate phenotypes such as brain structure, brain function and resting state. Other examples include colocalization analyses or the use of UKB data in conjunction with other datasets such as expression datasets to better understand pathways to disease. Finally, we will use exome and sequencing data in addition to genome-wide data to better understand the aetiological role of rare variants in mental health and beyond.