Introduction

The brain imaging component of UK Biobank (UKB) aims to scan 100,000 participants using multi-modal MRI^1,2. Participants have been deeply phenotyped and genotyped, and consent to long-term access to their health records¹. We have previously described our work on developing a quantitative susceptibility mapping (QSM) pipeline for the susceptibility-weighted MRI data in UKB, and the subsequent genetic associations with magnetic susceptibility (χ) measures (referred to as imaging-derived-phenotypes, IDPs)³. In this abstract, we extend this work to describe our findings from the phenotypic univariate association analyses between a broad range of UKB non-imaging phenotypes and QSM (and T2*) IDPs.

Methods

Susceptibility-weighted MRI (gradient echo) data from the current data release of 35,885 UKB subjects (53.11% female), aged 45-82y (mean=64.04y, SD=7.5y) at time of imaging, were processed using our QSM pipeline³. Median χ (referenced to CSF) and T2* were extracted from 8 subcortical structures (Fig. 1), and χ/T2* of white matter hyperintensities (WMH) relative to normal-appearing white matter was measured.

To perform the phenotypic analysis, we carried out univariate (pairwise) association analyses between 17,485 UKB non-imaging phenotypes and QSM/T2* IDPs (Fig. 1,2), where Pearson correlation r and P values were calculated. Here we refer to all non-imaging phenotypes as “phenotypes”, to distinguish them from IDPs. Phenotypes are grouped into 17 categories including early life factors (e.g., birth weight), lifestyle (e.g., diet, alcohol consumption), physical/body measures (e.g., BMI, blood assays), cognitive test scores (e.g., numeric memory), health outcomes (e.g., clinical diagnosis) and mental health variables (e.g., major depression).

P values were used to identify the strongest associations, with correlation r value serving as a measure of effect size. The full set of correlations was corrected for multiple comparisons using Bonferroni correction (-log₁₀P_uncorrected>7.10). Phenotypes identified to have a significant association with QSM IDPs were further investigated using voxel-wise (univariate Pearson) correlation maps.

Results and Discussion

We identified statistically-significant associations of 251 phenotypes with QSM IDPs, and 224 phenotypes with T2* IDPs. We compared the strength of QSM and T2* associations. Associations in some phenotype categories (e.g., alcohol consumption) are more specific to QSM IDPs (Fig. 3a) whereas other categories (e.g., cardiac) are more specific to T2* IDPs (Fig. 3b). However, the majority of phenotype categories show a mixed pattern of associations, including both common and distinct associations (e.g., blood assays) (Fig. 3c)

Here we describe a few example associations in four phenotype categories in detail. An overview of these categories is given in Table 1.

Blood assays

As shown in Fig. 1, the strongest associations are found in the blood assays category. They are haemoglobin-related phenotypes, which show positive correlations with QSM IDPs. This positive correlation is consistent with χ’s established relationship with iron, as haemoglobin-related blood measures are used clinically as a marker for a subject’s iron level⁴. The voxel-wise correlation map for the strongest correlation (mean corpuscular haemoglobin) is shown in Fig. 4a.

Health outcomes

In this category, the strongest association is found between the QSM WMH IDP and multiple sclerosis (self-reported) (Fig. 2), consistent with previous literature⁵. Its voxel-wise correlation map (Fig. 4b) reveals spatially contiguous negative associations in sub-regions of thalamus (including the pulvinar nucleus and lateral geniculate nucleus), as well as focal white matter regions such as the optic radiations. Previous literature has reported structural damage of the thalamic lateral geniculate nucleus in multiple sclerosis patients, reflecting potential damage of the visual pathway in multiple sclerosis^6,7.

Food and drink

The strongest associations in this category relate to tea consumption, which correlates negatively with χ. Although no direct link between tea intake and χ measures have been described previously, polyphenols in both green and black tea have been reported as brain-permeable, natural iron chelators that have demonstrated neuroprotective effects^8,9. The voxel-wise correlation map with tea intake is shown in Fig. 4c. Its spatial pattern of correlations is in line with a previous rodent study in which black tea extract reduced oxidative stress levels in the substantia nigra and striatum⁸.

Alcohol consumption

Alcohol consumption correlates more strongly with QSM IDPs than T2* IDPs in all cases (Fig. 3a). Voxel-wise correlation map with frequency of consuming six or more units of alcohol (Fig. 4d) reveals spatially contiguous positive associations in the putamen and sub-regions of substantia nigra and caudate, but also negative associations in sub-regions of the thalamus. These results recapitulate a previous study finding higher χ in the putamen, caudate and substantia nigra in subjects with alcohol use disorder, which has been linked to abnormal body iron accumulation¹⁰. χ in the thalamus correlated with phenotypes relating to the quantity of alcohol consumption. The thalamus is involved in the frontocerebellar circuit and Papez circuit, which are particularly affected by alcohol consumption¹¹. Although no previous studies have linked χ in the thalamus with alcohol use disorder, neuroimaging studies have reported reductions in thalamic volume and connectivity in alcohol use disorder patients¹¹.

Conclusion

Our results demonstrate that QSM and T2* contribute complementary information, as reflected in distinct patterns of associations. This QSM resource provides an opportunity to investigate χ contrast in previously unexplored territory, which might lead to advances in application of QSM in neuroscience. It is thus expected to be a valuable addition to the UKB neuroimaging resource.