Using Genetic Methods to Investigate Causal Pathways Linking Early Life Stress to Mental and Physical Health Problems and their Co-occurrence

: (Alternative Format Thesis)

Abstract

A substantial body of research links early life stress (ELS) to the development of mental and physical health problems. However, more research is needed to establish ELS as a causal risk factor for the co-occurrence of mental and physical health problems, and to identify the underlying mediating and moderating mechanisms. Previous studies have primarily investigated these associations using conventional observational epidemiology methods that are susceptible to unmeasured confounding and reverse causation, reflecting a need for alternative study designs. Hence, the primary aim of the present thesis was to investigate causal pathways linking ELS to mental and physical health problems, as well as their co-occurrence, using genetically informed methods.

By employing genetic methods, the present thesis sought to overcome limitations associated with conventional observational epidemiology designs. A total of five studies were conducted. Study 1 assessed whether ELS is a causal risk factor for multimorbidity between depression, coronary artery disease and type 2 diabetes using two-sample Mendelian randomisation (MR). Study 2 examined the underlying biological pathways linking ELS to multimorbidity using two-step, multivariable MR. Study 3 explored psychological stress reactivity as a moderator of ELS to depression association using longitudinal population-based data. Study 4 investigated whether a combined phenotype of loneliness and social isolation was a causal mediator of ELS to schizophrenia relationship using genomic structural equation modelling and MR. Study 5 evaluated DNA methylation as an alternative biomarker for brain-based disorders by conducting a systematic review of neuroimaging epigenetic research.

Overall, studies provided evidence for a causal effect of ELS on mental and physical health problems, including their co-occurrence. Three biomarkers were identified as potential causal mediators of the association between ELS and multimorbidity: glycated haemoglobin, triglycerides, and high-density lipoprotein (HDL) cholesterol. Additionally, stress reactivity was identified as a moderator of ELS to depression association. However, loneliness/social isolation did not seem to causally mediate the association between ELS and schizophrenia. The systematic review provided evidence for an association between DNA methylation patterns and brain structure and function, highlighting DNA methylation as another plausible pathway via which ELS may affect health.

Taken together, these findings (i) demonstrate the need to prevent or reduce ELS exposure, as doing so may decrease the risk of poor health outcomes; (ii) point to glycated haemoglobin, triglyceride, HDL cholesterol, and stress reactivity as potential screening or intervention targets; and (iii) emphasise the need for concurrent management of mental and physical health problems. Future studies may explore other potential pathways linking ELS to health, such as ageing, the microbiome, DNA methylation, and structural or functional brain alterations. Moreover, future research may assess the robustness of these findings using genetic data for different forms of ELS and multimorbidity.

Date of Award	26 Jun 2024
Original language	English
Awarding Institution	University of Bath
Supervisor	Esther Walton (Supervisor), Graeme Fairchild (Supervisor) & Andrew Ward (Supervisor)