The abnormality occurs at molecular, cellular as well as network levels in patients with Alzheimer’s disease (AD) prior to diagnosis. Most previous connectivity studies were conducted at 1 out of 3 (local, meso and global) scales in subjects covering only part of the entire AD spectrum (subjective cognitive decline, SCD; amnestic mild cognitive impairment, aMCI; and then fully manifest AD). Data interpretation within the framework of disease progression is therefore difficult. The current study included 3 age- and sex-matched cohorts: SCD ( n = 32), aMCI ( n = 37) and fully-established AD ( n = 30). A group of healthy elderly subjects ( n = 40) were included as a normal control (NC). Network connectivity was examined at the local (degree centrality), meso [subgraph centrality (SC)], and global (eigenvector and page-rank centralities) levels. As compared to NC, SCD subjects had isolated decrease of SC in primary (somatomotor and visual) networks. aMCI subjects had decreased centralities at all three scales in associative (frontoparietal control, dorsal attention, limbic and default) networks. AD subjects had increased centrality at the global scale in all seven networks. There was a positive association between Montreal Cognitive Assessment (MoCA) scores and DC in the frontoparietal control network in SCD, a negative relationship between Mini-Mental State Examination (MMSE) scores and EC in the somatomotor network in AD. These findings suggest that the primary network is impaired as early as in SCD. Impairment in the associative network also starts at the local level at this stage and may contribute to the cognitive decline. As associative network impairment extends from local to meso and global scales in aMCI, compensatory mechanisms in the primary network are activated.

The progression of Alzheimer’s Disease (AD) has been proposed to comprise three stages, subjective cognitive decline (SCD), mild cognitive impairment (MCI), and AD. Was brain dynamics across the three stages smooth? Was there a critical transition? How could we characterize and study functional criticality of human brain? Based on dynamical characteristics of critical transition from nonlinear dynamics, we proposed a vertex-wise Index of Functional Criticality (vIFC) of fMRI time series in this study. Using 42 SCD, 67 amnestic MCI (aMCI), 34 AD patients as well as their age-, sex-, years of education-matched 54 NC, our new method vIFC successfully detected significant patient-normal differences for SCD and aMCI, as well as significant negative correlates of vIFC in the right middle temporal gyrus with total scores of Montreal Cognitive Assessment (MoCA) in SCD. In comparison, standard deviation of fMRI time series only detected significant differences between AD patients and normal controls. As an index of functional criticality of human brain derived from nonlinear dynamics, vIFC could serve as a sensitive neuroimaging marker for future studies; considering much more vIFC impairments in aMCI compared to SCD and AD, our study indicated aMCI as a critical stage across AD progression.

No abstract

Understanding the critical features of the human brain at multiple time scales is vital for both normal development and disease research. A recently proposed method, the vertex-wise Index of Functional Criticality (vIFC) based on fMRI, has been testified as a sensitive neuroimaging marker to characterize critical transitions of human brain dynamics during Alzheimer’s disease progression. However, it remains unclear whether vIFC in healthy brains is associated with neuropsychological and neurophysiological measurements. Using the Nathan Kline Institute/Rockland lifespan cross-sectional datasets and openfMRI single participant longitudinal datasets, we found consistent spatial patterns of vIFC across the entire cortical mantle: the inferior parietal and the precuneus exhibited high vIFC. On a time scale of years, we observed that vIFC increased with age in the left ventral posterior cingulate gyrus. On a time scale of days and weeks, vIFC demonstrated the capacity to identify a link between anxiety and pulse. These results showed that vIFC can serve as a useful neuroimaging marker for detecting physiological, behavioral, and neurodevelopmental transitions. Based on the criticality theory in nonlinear dynamics, the current vIFC study sheds new light on human brain studies from a nonlinear perspective and opens potential new avenues for normal and abnormal human brain studies.