Acyclic graphs to define models of relationships between human cytomegalovirus, cardiovascular disease and all-cause mortality in UK Biobank.

Human cytomegalovirus seropositivity has shown varying levels of association with cardiovascular disease and all-cause mortality in previous studies. This study uses updated data from the UK Biobank to test these associations. The association between human cytomegalovirus seropositivity and outcomes of incident cardiovascular disease, ischaemic heart disease, stroke and all-cause mortality was determined using multivariate Cox proportional-hazards models in 8740 UK Biobank participants aged 40-69 years. Two directed acyclic graphs depicting the hypothesized relationship between human cytomegalovirus infection (either in childhood or adulthood) and cardiovascular disease/all-cause mortality controlled the selection of biological and socioeconomic confounders to be adjusted for in each model. Human cytomegalovirus seropositivity was not significantly associated with cardiovascular disease, ischaemic heart disease, stroke or all-cause mortality when applying either the fully adjusted adulthood or childhood infection models. We found no significant association between human cytomegalovirus seropositivity and any of the measured outcomes. Further research is expected to include larger sample sizes, younger participants and more ethnically diverse cohorts.This article is part of the discussion meeting issue 'The indirect effects of cytomegalovirus infection: mechanisms and consequences'.