Metabolomic profiling in small vessel disease identifies multiple associations with disease severity

Abstract

Cerebral small vessel disease (SVD) is a major cause of vascular cognitive impairment and dementia. There are few treatments, largely reflecting limited understanding of the underlying pathophysiology. Metabolomics can be used to identify novel risk factors in order to better understand pathogenesis and to predict disease progression and severity.

We analysed data from 624 patients with symptomatic cerebral SVD from two prospective cohort studies. Serum samples were collected at baseline and patients underwent MRI scans and cognitive testing at regular intervals with up to 14 years of follow-up. Using ultra-performance liquid chromatography mass spectrometry and nuclear magnetic resonance spectroscopy, we obtained metabolic and lipidomic profiles from 369 annotated metabolites and 54,764 unannotated features and examined their association with respect to disease severity, assessed using MRI SVD markers, cognition, and future risk of all-cause dementia.

Over 100 annotated metabolites were significantly associated with SVD imaging markers, cognition, and progression to dementia. Decreased levels of multiple glycerophospholipids, sphingolipids, and sterol lipids were associated with increased SVD load as evidenced by higher white matter hyperintensities (WMH) volume, lower mean diffusivity normalised peak height (MDNPH), greater brain atrophy, and impaired cognition. Higher levels of several amino acids and nucleotides were associated with higher WMH volume, greater atrophy, and lower MDNPH. Lower baseline levels of carnitines and creatinine were associated with higher annualised change in peak width of skeletonised mean diffusivity (PSMD), and several metabolites, including lower levels of valine, caffeine, and VLDL analytes, were associated with future dementia incidence. Additionally, we identified 1,362 unannotated features associated with lower MDNPH and 2,474 unannotated features associated with increased WMH volume.

Our results show multiple distinct metabolic signatures that are associated with imaging markers of SVD, cognition, and conversion to dementia. Further research should assess causality and the use of metabolomic screening to improve the ability to predict future disease severity and dementia risk in SVD. The metabolomic profiles may also provide novel insights into disease pathogenesis and help identify novel treatment approaches.