Abstract / Summary
Abstract Background Frailty predicts adverse postoperative outcomes, but its association with electroencephalographic (EEG) features during remimazolam-induced loss of consciousness (LOC) in elderly patients is unknown. This pre-planned secondary analysis tested whether frailty, defined by the “FRAIL” (Fatigue, Resistance, Ambulation, Illness, Loss of weight) scale, is associated with frontal EEG features during awake baseline and at LOC. Methods We enrolled 49 elderly patients (≥ 65 years) undergoing surgery with remimazolam induction. Frailty was preoperatively assessed using the “FRAIL” scale; patients were categorized as non-frail (n = 25) or frailty-prone (n = 24). Frontal EEG was recorded during awake baseline and at LOC. We analyzed conventional band powers, quantitative EEG parameters, and aperiodic (offset and exponent) and periodic components via spectral decomposition. Mixed-design repeated-measures analysis of variance evaluated Group (frailty status) and State (awake/LOC) effects. Results No significant Group main effect or Group × State interaction was observed for any EEG metric (all P > 0.05). In contrast, significant State effects were found for most parameters (P < 0.001). Despite substantial differences in American Society of Anesthesiologists (ASA) physical status between the two groups (P < 0.001), their EEG characteristics remained indistinguishable. Conclusions Preoperative "FRAIL" scale scores did not predict differences in frontal EEG features between non‑frail and frailty‑prone elderly patients during awake baseline or remimazolam‑induced LOC. These findings suggest that somatic frailty and brain vulnerability are dissociable. Thus, perioperative brain risk prediction should rely on direct indicators of brain reserve, such as cognitive assessments, resting‑state EEG, or neuroimaging, rather than frailty scales alone. Registration: Chinese Clinical Trial Registry (ChiCTR2500113571; December 1, 2025).