Investigating associations of multiple common infections with brain structure and white matter lesions: findings from prospective neuroimaging studies

Abstract

AbstractBackgroundEvidence for a role of common infections in neurodegenerative and cerebrovascular diseases is conflicting, and the extent to which pathogen exposure associates with subclinical imaging markers of these diseases is unclear. We aimed to investigate associations between a panel of serological measures characterising exposure to common pathogens – selected based on their relevance to public health – and brain volume, hippocampal volume, and white matter lesions in two population‐based cohorts in the UK.MethodSerology and imaging data were available for 1371 participants across two studies: the UK Biobank (N = 935, mean age(SD)[years] = 55.3(7.7) at serology; 63.8(7.7) at imaging) and the Medical Research Council National Survey of Health and Development [N = 436, mean age(SD) [years] = 63.3(1.1) years at serology; 70.7(0.7) years at imaging). Serology measures included a) serostatuses of 17 pathogens, b) total proportion of 17 pathogens exposed to, and c) total proportion of 11 neurotropic pathogens exposed to (Table 1). Associations between serology measures and outcomes (brain volume, hippocampal volume, and white matter lesion volume) were tested using linear regression. Regression models included adjustments for total intracranial volume and clinic information (basic models), and age, sex, ethnicity, education, socioeconomic position and lifestyle factors (Figure 1). Findings across cohorts were meta‐analysed using random‐effects models. The present study will be expanded to include an additional cohort (Southall and Brent Revisited,N≤1200) to add more precision for identifying associations.ResultSeropositivity to seven pathogens associated with an outcome in basic models, and four in fully‐adjusted models: seropositivity to JC virus was associated with smaller brain volumes, Epstein‐Barr virus with larger brain volumes, Merkel cell virus with larger hippocampal volumes and fewer white matter lesions, and C.trachomatis with fewer white matter lesions (Figure 2). Neurotropic pathogen burden scores associated with smaller hippocampal volumes in basic models, but no associations were seen for either pathogen burden score in fully‐adjusted models.ConclusionWe found some evidence of both positive and negative associations between common infections and imaging markers of subclinical cerebrovascular disease and neurodegeneration, with no evidence of relationships with pathogen burden. This work highlights key relationships for further investigation, which may reveal novel pathways associated with brain structure and cerebrovascular pathology.