Structural covariance in the affective pain network in Alzheimer’s disease

Abstract

AbstractBackgroundPain consists of both sensory and affective networks. Literature exploring the effect of Alzheimer’s disease (AD) on pain is limited. Some evidence suggests that the affective networks are impacted earlier while the sensory networks are preserved until late in the course of disease (Cole 2006). To examine the preservation of networks underlying pain perception in AD, we examined the structural covariance of regions in the pain network. Using structural MRI (sMRI) data, we investigated structural covariance by modeling a network from the correlation of volume between regions. In structural covariance, path length represents the minimum number of steps that must be taken to travel between two nodes and may serve as a proxy for structural brain integrity. Characteristic path length averages together the path lengths for all nodes in a given a priori network to create a global network metric (Paldino 2017). We hypothesized that there would be alterations in altered characteristic and nodal path length in individuals with AD.MethodAs part of a larger study, cortical volume measurements were obtained for 78 participants (39 AD, 39 HC) via Ma‐CRUISE mutli‐atlas segmentation (Huo 2016). The affective pain network consisted of bilateral: anterior and posterior insula, amygdala, anterior cingulate cortex (ACC), and dorsolateral pre‐frontal cortex (dlPFC, specifically the superior frontal and medial frontal gyri, Woo 2017). The analysis was conducted in BRAPH using a undirected weighted graph with negative correlations set to zero. (Mijalkov 2017). Only characteristic and nodal path length were included in this analysis as these metrics are more robust for discriminating differences in AD (Mårtensson 2018).ResultNo significant differences were found for characteristic path length between groups (p = > 0.05). Additionally, no significant differences were found in path length for any of the nodes analyzed following correction with the Benjamini‐Hochberg procedure (Q = 0.05).ConclusionPreservation of characteristic and nodal path length metrics in the affective pain network provides preliminary evidence that this network may be preserved in mild to moderate AD. This clinically meaningful finding provides further evidence indicate that affective pain may be preserved longer than originally thought during the course of AD.