Sublexical compensatory reorganization of Chinese character recognition in Alzheimer's disease

This study investigated the cognitive-functional shift in visual character recognition across the cognitive decline trajectory, ranging from healthy aging through Mild Cognitive Impairment (MCI) to Mild Alzheimer's Disease (AD). Specifically, it was hypothesized that as global semantic networks degrade, the reliance on rapid holistic processing decreases, prompting a compensatory shift toward sublexical features.

Using a controlled unmasked priming lexical decision paradigm, this study evaluated behavioral performance across three distinct participant groups (N = 30 per group): healthy controls (HC), individuals with MCI, and patients with Mild AD. The experimental stimuli consisted of 80 Chinese phonograms matched across key psycholinguistic properties. This study utilized mixed-design Analyses of Variance (ANOVAs) alongside item-level hierarchical multiple regression (HMR) models to isolate the specific drivers of reaction times, assessing the predictive power of visual complexity, whole-character frequency, semantic transparency, and phonetic regularity.

The analytical models revealed distinctly divergent psycholinguistic trajectories. The HC group demonstrated highly efficient holistic processing, with performance primarily driven by whole-character frequency. Conversely, the MCI group exhibited a unique transitional cognitive profile, characterized by a simultaneous reliance on both holistic whole-word access and analytic sublexical features. Crucially, in the Mild AD tgroup, the holistic processing advantage was entirely abolished. This functional disintegration triggered a compensatory reorganization toward bottom-up sublexical properties, with patients relying heavily on semantic transparency and phonetic regularity to achieve semantic access.

These findings demonstrate a linguistic Compensation-Related Utilization of Neural Circuits Hypothesis (CRUNCH) effect within the degenerating brain. As top-down lexical-semantic hubs deteriorate, reflected behaviorally in this paradigm by the loss of holistic processing advantages, the cognitive system actively compensates by upregulating bottom-up, sublexical pathways. Furthermore, the transitional sublexical dependency observed during the MCI stage highlights the translational potential of psycholinguistic variables as sensitive, noninvasive behavioral biomarkers for the early screening and monitoring of dementia.