Reliability of P3 Event-Related Potential During Working Memory Across the Spectrum of Cognitive Aging

Devos, Hannes; Burns, Jeffrey M.; Liao, Ke; Ahmadnezhad, Pedram; Mahnken, Jonathan D.; Brooks, William M.; Gustafson, Kathleen · 2020 · Crossref

DOI: 10.3389/fnagi.2020.566391

archive: archived pipeline: cataloged verified

Get this paper ↗ (DOI — opens at the source; we link to it, we don't host it)

Summary

This study addresses the lack of established test-retest reliability for event-related potentials (ERPs) in older adults across a spectrum of cognitive abilities, ranging from healthy aging to Alzheimer’s disease (AD). While ERPs offer high temporal resolution for studying cognitive aging, the stability of these measures in populations with heterogeneous cognitive profiles has not been previously characterized. The research aimed to determine if P3 ERP metrics are stable enough to distinguish biological changes from measurement error in older individuals, and to assess how age, cognitive status, amyloid burden, and task difficulty influence this reliability. The researchers conducted a test-retest reliability study involving 39 older adults (mean age 74.1 years) recruited from the University of Kansas Alzheimer’s Disease Center. Participants were categorized into three groups: cognitively normal with non-elevated beta-amyloid (Aβ−, n=15), cognitively normal with elevated beta-amyloid (Aβ+, n=16), and those with mild cognitive impairment or AD (MCI/AD, n=8). All participants underwent EEG recording while performing an n-back working memory task at three difficulty levels (0-back, 1-back, and 2-back) during a baseline session and a follow-up session approximately two weeks later. P3 peak amplitude and latency were measured at predetermined frontal and central channels, with Fz serving as the primary outcome variable. Reliability was assessed using intraclass correlation coefficients (ICC), and generalized linear mixed models were used to evaluate the impact of demographic and clinical variables on ERP stability. The results indicated that P3 peak amplitude demonstrated excellent reliability for the 0-back (ICC = 0.82) and 1-back (ICC = 0.87) tasks, but only fair reliability for the more demanding 2-back task (ICC = 0.53). In contrast, P3 peak latency showed substantially lower reliability, ranging from fair (ICC = 0.54 for 0-back) to poor (ICC = 0.17 for 2-back). Behavioral performance measures, including accuracy and response times, exhibited excellent reliability across all task levels. Statistical analysis revealed that age, disease group, and task difficulty did not significantly confound the stability of P3 amplitude or latency. However, Montreal Cognitive Assessment (MOCA) scores showed a trend toward negatively correlating with P3 amplitude stability (p = 0.07), suggesting that lower cognitive function may be associated with greater variability in ERP responses. The study concludes that P3 peak amplitude is a stable electrophysiological measure for assessing working memory processes in older adults, particularly at lower cognitive loads, whereas P3 latency is less reliable. The findings support the use of P3 amplitude as a biomarker for studying normal and pathological cognitive aging, including preclinical AD, provided that cognitive status is considered. The authors suggest that while the 2-week interval was sufficient to wash out practice effects, future research should explore longer-term stability and the sensitivity of ERPs to subtle neurobiological changes over time.

Provenance

The full processing record for this entry. Every stage of this paper's journey through the pipeline is logged — what ran, with which tool and model, how many attempts it took, and when it last completed.

Summary generated by qwen3.6-27b-prismaquant on 2026-06-26; verification: verified.

Topics

Ranked by relevance to this paper. Hover a topic for its definition.

• cognitive capacity variation