Moderation of persistent homology-based functional connectivity on cognitive decline in AD

1579 

Abstract Submission 

Seonjoo Lee¹, Pei Liu², Xi Zhu³, Yaakov Stern⁴

¹Columbia University Irving Medical Center, Columbia University, and The New York, New York, NY, ²Columbia University, New York, NY, ³New York State Psychiatric Institute, New York, NY, ⁴Columbia University Irving Medical Center, NEW YORK, NY

Seonjoo Lee  
Columbia University Irving Medical Center, Columbia University, and The New York
New York, NY

Pei Liu  
Columbia University
New York, NY

Xi Zhu  
New York State Psychiatric Institute
New York, NY

Yaakov Stern  
Columbia University Irving Medical Center
NEW YORK, NY

Network analysis has been widely used to understand the complex interactions and organization of the human brain. Resting-state functional networks, often studied for cognition and aging, are analyzed through network analysis with modular assumption. However, it is more plausible to assume a core-periphery or rich club structure accounts for brain functions where the hubs are tightly interconnected to allow for integrated processing [1]. To address this, we introduced persistent homology-based functional connectivity (PHFC) indices, including backbone strength (BS), backbone dispersion (BD), and cycle strength (CS), to quantify integrated processing patterns. BS reflects overall functional integration, BD indicates differences in critical information flow, and low CS suggests strong information flow through the backbone rather than additional cycles to study cognitive aging [2]. Using large public data, our study investigates the role of PHFC indices in Alzheimer's disease pathology, revealing their potential advantages beyond traditional measures.

Sample included 503 adults (mean age 74.53 [age range 56-96], 50.5% female) from the ADNI. Amyloid positivity was measured using av45 PET SUVR with cut point 1.1. Resting-state functional magnetic resonance imaging (rs-fMRI) data was obtained from 3T scanners. The time series data were extracted using Power's atlas, functional connectivity was computed using Pearson's correlation coefficients, and the PHFC indices were calculated as implemented in the PHFconn R package (https://github.com/hyunnamryu/PHfconn). We performed separate mixed effect regression for each executive function and episodic memory composite scores as dependent variables. Each mixed effect model included PH-based functional connectivity indices, years from baseline, amyloid positivity and their two-way and three-way interactions as fixed effects, and subject intercepts and years from baseline slopes as random effects. We also included baseline age, sex, and education as covariates for all models.

There was a significant three-way interaction between BD, days from baseline, and amyloid positivity (=-0.05, CI=-0.09 to -0.02, p=.005) for executive function. For episodic memory, significant three-way interactions were found for BD (=-0.04, CI=-0.08 to -0.01, p=.008), BS (=-0.04, CI=-0.08 to -0.01, p=.006), and CS (=-0.03, CI=-0.06 to -0.00, p=.034) with days from baseline and amyloid positivity. These results indicate that the patterns of information integration in functional connectivity moderate the effect of amyloid burden on the speed of cognitive decline. Specifically, in the amyloid positivity group, higher BD is associated rapid cognitive decline, while lower BD is associated with faster decline in the absence of amyloid.

By employing the PHFC indices, we have uncover intricate patterns of information integration with the functional networks. These patterns of information integration patterns play a crucial role in moderating the relationship between Alzheimer's disease related amyloid burden on cognitive decline.

Neurodegenerative/ Late Life (eg. Parkinson’s, Alzheimer’s) ²

Connectivity (eg. functional, effective, structural) ¹

Cognition

Data analysis

FUNCTIONAL MRI