Phenotypic and genetic associations of quantitative magnetic susceptibility in UK Biobank brain imaging

Abstract

A key aim of epidemiological neuroscience studies is the identification of novel markers to assess brain health and evaluate the efficacy of therapeutic interventions. A broad range of tissue constituents associated with healthy brain functions and diseases (e.g., iron, myelin and calcium) are characterised by unique magnetic properties, motivating the development of measurements sensitive to these effects. Quantitative susceptibility mapping (QSM) is an emerging MRI technique which enables in vivo quantification of tissue magnetic susceptibility, providing the opportunity to identify such markers. In this study, we developed a robust QSM modelling pipeline and evaluated the potential of QSM in a cohort of 35,885 subjects (ages 45-82), alongside extensive genetic and phenotypic associations as part of UK Biobank. Our analysis identified a rich set of distinctive phenotypic and genetic associations with QSM. This included phenotypic associations with body iron, diet, alcohol and diseases, and genetic associations related to a broad range of biological functions including iron, calcium homeostasis, myelination, extracellular matrix. Importantly, we found that QSM and T2* have unique patterns of associations, demonstrating the added value in including QSM IDPs in the UK Biobank brain imaging resource, with considerable promise to identify novel in vivo brain health markers.

Impact of moderate alcohol consumption on brain iron and cognition: observational and genetic analyses

“…In contrast, we observed associations in those drinking just >7 units per week. A recent phenomewide association study of quantitative susceptibility in the same dataset reported significant associations in basal ganglia regions with higher frequency binge-drinking [10]. Regional heterogeneity in iron concentrations is well described although the aetiology is not understood [43].…”

“…QSM depends on phase images, which were obtained from individual coil channels, combined, masked and unwrapped. Magnetic susceptibility (χ) was calculated using a QSM pipeline including background field removal, dipole inversion and CSF referencing as described elsewhere [10]. Median χ values (in parts-per-billion) across voxels within each region were calculated, resulting in 16 QSM IDPs.…”

“…Genetic associations with serum markers of iron homeostasis (serum iron, ferritin, transferrin saturation (of iron-binding sites of transferrin occupied by iron), and total iron binding capacity) were obtained from the largest published GWAS to date (meta-analysis of deCODE, INTERVAL and Danish Blood Donor Study) [31]. Genetic associations with brain T2* and susceptibility (IV-outcomes) were obtained from the largest GWAS of brain imaging phenotypes [10, 32]. All genetic associations were based on GWAS of European ancestry samples.…”

“…Whilst neurological sequelae of inherited iron overload disorders have long been recognised [6], higher brain iron has now been implicated in the pathophysiology of Alzheimer’s and Parkinson’s diseases [7, 8]. Intriguingly, not only does the clinical profile of alcohol-related dementia overlap with these disorders, but recent observational evidence suggests heavy alcohol use may associate with iron accumulation in the brain [9, 10].…”

Reliability of multi-modal MRI-derived brain phenotypes for multi-site assessment of neuropsychiatric complications of SARS-CoV-2 infection

“…To enable qualitative neurological assessment of individual patients the median phase and magnitude data were processed to provide maps highlighting features indicative of abnormal iron deposition, e.g., due to microbleeds. Quantitative susceptibility mapping (QSM) was also performed, using the phase data and a recently developed UKB pipeline (Wang et al, 2021) . Susceptibility maps were generated using the iLSQR algorithm (Li et al, 2015), with susceptibility values reported relative to the susceptibility measured in CSF.…”