The polygenic hazard score mediates the association between plasma neurofilament light chain and brain morphometry in dementia spectrum

Abstract

Introduction

Blood-based biomarkers such as plasma neurofilament light chain (pNfL) are crucial biomarkers for Alzheimer's disease (AD). Additionally, neuroimaging techniques such as tensor-based morphometry (TBM), which identify structural changes in the brain, can provide valuable insights into AD pathophysiology. However, the role of genetics in linking the blood based biomarkers and imaging findings has not been well understood. Therefore, we aimed to investigate whether the polygenic hazard score (PHS), affects the association between neurofibrillary tangles and neuritis plaques and brain imaging findings.

Methods

Using the Alzheimer's Disease Neuroimaging Initiative (ADNI) database, we enrolled all participants for whom a complete dataset of pNfL, PHS, and TBM was available. Using Python, we analyzed the associations between pNfL levels and the TBM data of 567 participants incluidng 152 cognitively normal individuals, 309 participants with mild cognitive impairment (MCI), and 106 patients with AD. We used a mediation analysis to identify the effect of PHS in how pNfL is associated with TBM measures.

Results

We found a negative correlation between the accelerated TBM measure and NfL levels in both the MCI and AD groups. The pNfL concentration predicted both accelerated statistical and anatomical TMB measures in patients with MCI. Furthermore, PHS mediatedthe association between statistical TBM measures and NfL levels in AD patients, to the extent that the significant association between NfL and TBM measures disappeared after accounting for PHS.

Conclusion

We showed that although pNfL can predict the cognitiee decline and imaging findings in AD, this effect is mediated by the PHS. Therefore, PHS should be considered when investigating AD biomarkers and their corresponding imaging findings.