Disruption of concordance among spontaneous measures of intrinsic brain connectivity in Alzheimer’s disease

Abstract

AbstractBackgroundIn the last years, several voxel‐wise data‐driven metrics gained attention to study resting‐state connectivity in healthy aging and various pathological states. More recently, a new measure assessing the interaction between the commonly used metrics has been proposed and made widely available via the DPABI toolbox: the so‐called concordance index (CI). This index reflects the coupling of different metrics. The value of this new measure in Alzheimer's disease (AD) research is yet unclear.MethodHealthy young controls (YC, n=26), healthy senior controls (SC, n=29), and patients diagnosed with AD (n=35) underwent 3T resting‐state fMRI. First, the fractional amplitude of low frequency fluctuations (fALFF), regional homogeneity (ReHo), and degree centrality (DC) were calculated. Next, the CI among these measures was determined for each time window and each brain voxel across time by using Kendall's W. Differences in the CI between groups were examined on a whole brain, voxel‐wise, and network‐level using masks of seven of the most prominent intrinsic brain networks, with age, sex, head motion and gray matter volume as covariates.ResultThe whole‐brain CI was significantly lower in AD patients as compared to YC and SC, which is in line with results from analyzing the single metrics. Reductions of the CI in AD was consistently observed in all networks investigated except for the limbic network. In SC compared to YC, no network‐level CI differences were detectable. On the voxel level, AD patients compared to SC showed lower CI in the right middle temporal gyrus, and SC compared to YC revealed lower CI in the left middle frontal gyrus. Analyzing the single metrics for their own, we observed a lower fALFF in the right angular gyrus in AD compared to SC, but no differences were detected for ReHo and DC.ConclusionThe newly proposed CI based on resting‐state connectivity may serve as a more sensitive imaging marker in AD than single metrics. The data suggest that this index distinguishes well between neurodegeneration and healthy aging.