Study Links a Shrinking Cholinergic Hub to Slower Brain-Wide Signals and Spreading Tau in Early Alzheimer's

A new brain-imaging analysis of ADNI participants finds that a shrinking cluster of cholinergic neurons deep in the forebrain tracks with weaker infra-slow brain activity and with cortical tau spreading further as disease stage advances, an observational finding that stops short of proving cause.

A study published in Nature Communications on September 12, 2026 reports that degeneration of the nucleus basalis of Meynert (NbM), a small cluster of cholinergic neurons at the base of the forebrain, is associated with weaker infra-slow global brain activity and with more cortical tau in early Alzheimer's disease. The researchers, Yutong Mao, Baizhou Pan and Xiao Liu of Pennsylvania State University, analyzed resting-state fMRI and flortaucipir PET tau scans from 75 participants in the Alzheimer's Disease Neuroimaging Initiative (ADNI-3), average age 73, none of whom carried a clinical Alzheimer's diagnosis at enrollment; the group included 37 cognitively normal adults, three with subjective memory concerns and 35 with mild cognitive impairment, each scanned at least twice, nine months or more apart.

The team found that a smaller NbM was tied to weaker gBOLD activity, the fMRI measure of spontaneous, brain-wide slow waves, and that the two changes overlapped spatially with where cortical tau had accumulated. This pattern first appeared in early Braak-stage regions among cognitively unimpaired participants who already had abnormal cerebrospinal fluid markers of amyloid and tau, and extended into more advanced Braak regions among those with mild cognitive impairment, tracking the textbook order in which tau is thought to spread across the cortex. In a longitudinal comparison across the repeat visits, reductions in gBOLD activity linked to NbM decline came before, and statistically predicted, later increases in cortical tau, a temporal pattern the authors describe as consistent with gBOLD activity mediating the link between subcortical cholinergic loss and cortical tau buildup.

The authors are careful to frame this as an association uncovered through statistical mediation modeling in existing scan data, not a demonstrated mechanism confirmed by manipulating brain activity in people. The cholinergic system's vulnerability early in Alzheimer's disease is well established from prior autopsy and imaging work, and infra-slow global brain activity has previously been tied to tau pathology and to cerebrospinal fluid clearance in separate lines of research; what this paper adds is a specific statistical link tying the three together across two clinical stages in the same cohort. Whether reduced global brain activity is a cause of tau spread, a downstream marker of the same underlying degeneration, or something that could in principle be restored are questions the paper does not settle.

The findings will need replication in larger and more diverse cohorts, and confirmation through methods that can test the sequence directly, before this pathway can be considered established rather than a promising statistical association.