Impairment of Auditory Integration in Subjective Cognitive Decline, Amnestic Mild Cognitive Impairment, and Alzheimer’s Disease
Changming Wang,Zhibin Wang,Lei Liu,Pengcheng Yang,Beijia Xie,Xinrui Shi,Jianping Jia,Liang Li,Yi Tang
DOI: https://doi.org/10.1002/alz.050625
2021-01-01
Abstract:AbstractBackgroundLocalization of sound sources in a noisy environment or ‘auditory scene analysis’ depends on central auditory processing (CAP) that is vulnerable in cognitive impairment. Here, we investigated the behavioral and electrophysiological characteristics of auditory integration, as a particular function in CAP, in relation to neuropsychological functions in SCD, aMCI, and AD.MethodWe introduced a binaural integration paradigm to measure sound‐localization‐related features (identification and retention of acoustic signals) by detecting break in interaural correlation (BIC) and primitive auditory memory (PAM) in wideband noises. We used a 128‐channel EEG to record the event‐related‐potential (peak amplitude, peak latency, and inter‐trial phase coherence [ITPC]) for the paradigm.ResultA total of 121 participants (mean age 65.9 years) were included. Significant differences in BIC detection and PAM were revealed between groups adjusting for age and gender (BIC: F = 20.2, P < 0.001; PAM: F = 24.6, P < 0.001). Post‐hoc analysis showed that BIC detection and PAM were significantly higher in NC or SCD groups compared with aMCI or AD groups. Further adjusting for neuropsychological scores, the significances were not changed, suggesting that auditory integration impairment was partially independent from top‐down cognitive functions. In EEG experiment, the most prominent differences in EEG metrics among the four groups included the ITPC of N1 at left parietooccipital region (F = 3.85, P = 0.012) and of P2 at left parietooccipital (F = 3.89, P = 0.011) and frontocentral regions (F = 8.09, P < 0.001). Notably, there was a linear decrease for the ITPC of N1 and a U‐shape‐curve change for the ITPC of P2 among the four groups, indicating that EEG patterns are differentially modulated depending on disease severity. For the discrimination of AD vs NC and MCI vs NC, the AUC was 0.92 and 0.85 for BIC detection; 0.94 and 0.90 for PAM, respectively.ConclusionAuditory integration, as a partially independent phenotype, is impaired in aMCI and AD populations. A U‐shape‐curve change of P2 with the increasing disease severity was found, indicating a compensation on auditory scene analysis. Results support binaural integration paradigm as a biomarker for screening cognitive decline and identifying auditory integration deficit.