The Role of Viruses in the Development of Islet Autoimmunity and Gestational Diabetes Mellitus

Viral infections are potential candidate triggers of islet autoimmunity (IA) and gestational diabetes mellitus (GDM), particularly in at risk individuals. Despite reported associations, molecular evidence remains limited. This thesis investigates the role of the respiratory viral infections in the development of IA in children, within the Environmental Determinant of Islet Autoimmunity (ENDIA) cohort study. It also examines the association between gut viral infections and the development of GDM in pregnant mothers as part of the Microbiome Understanding in Maternity Study (MUMS). For the ENDIA study, 54 cases (children with IA) and 161 matched controls (1:3 ratio) were recruited. The selection of cases was based on seropositivity to more than one islet specific autoantibody in two or more consecutive study samples or the onset of T1D as of 30 June 2020. Biospecimens were collected at each trimester during pregnancy, two visits within two weeks after birth, three-monthly visits until 24 months of age, and six-monthly thereafter until the development of T1D. For the MUMS nested case-control study, 16 cases (mothers with GDM) and 48 matched controls (1:3 ratio) were recruited. Stool samples were collected from mothers during the three trimesters of pregnancy to analyse the association between the gut virome and the development of GDM. Nucleic acids were extracted from longitudinal throat swabs (n=1448; ENDIA cohort) and stool samples (n=187; MUMS cohort), followed by NGS library preparation. Library fragments containing viral sequences were enriched through virome capture and sequenced on the Illumina NovaSeq 6000 platform. Taxonomic classification of viral reads was achieved using the Chan Zuckerberg Infectious Disease (CZ-ID) open-source platform. For statistical analysis, univariate, multivariate and longitudinal methods were employed. Univariate logistic regression analysis of individual viral taxa showed no statistically significant association with either IA or GDM (p > 0.05). Multivariate analysis showed no separate or distinct clustering patterns based on grouping (cases versus controls) and referring to non-significant differences in virome composition (PERMANOVA, p-value > 0.05). Alpha diversity (Simpson and Shannon) analysis suggested stable virome diversity between cases and controls (Wilcoxon rank test, p-value > 0.05). Overall, these findings did not demonstrate significant associations between respiratory viral infections and the development of IA or gut viral infections and GDM.