Predictive Value of Electroacupuncture for Cognitive Outcomes in Mild Cognitive Impairment: A Retrospective Cohort Study

Background Mild cognitive impairment (MCI) is a high-risk stage preceding Alzheimer's disease. Readily available predictors for cognitive outcomes following electroacupuncture (EA) remain limited. Objective To identify potential predictors of cognitive response to EA in MCI patients. Methods This retrospective study included 85 MCI patients who received standardized EA at a Chinese hospital (December 2023–December 2024). The 6-month change in Montreal Cognitive Assessment (MoCA) score served as the efficacy endpoint; patients were grouped into effective group (MoCA rise ≥2 points, n = 52) versus ineffective group ( n = 33). Baseline demographics, clinical characteristics, EA parameters, and cognitive scores (MoCA, Mini-Mental State Examination [MMSE]) were collected. Univariate and multivariate logistic regression identified independent associated factors. A scoring model was developed and assessed by receiver operating characteristic analysis. Results The effective group showed lower cerebrovascular comorbidity ( p = .002), a higher proportion receiving EA ≥3 sessions/week, longer total treatment course, and more total sessions ( p < .05). Greater MoCA and MMSE gains were also observed ( p < .001). Independent associated factors included baseline MoCA, EA frequency <3 sessions/week, cerebrovascular comorbidity, and total treatment course (all p < .05). A predictive scoring system was constructed: 1 point per 1-point higher baseline MoCA, 1 point per additional treatment week, 5 points for ≥3 sessions/week, and 7 points for no cerebrovascular comorbidity. This model yielded an area under the curve of .677 (95% CI: 0.551–0.804) for favorable response, with an optimal cutoff of 27.5 points. Conclusion Response is associated with baseline MoCA, comorbidity, and EA dose. The model remains exploratory and warrants independent validation.