Abstract
BACKGROUND: With the intensification of population aging, the incidence of mild cognitive impairment (MCI) continues to rise, and the risk of developing dementia is high, making it a global public health challenge. Traditional Chinese medicine non-pharmaceutical interventions (TCM-NPIs) is an important means of treating MCI. However, in clinical practice, there are various therapies, and there is no clear optimal plan yet. This study aims to evaluate the intervention effects of different TCM-NPIs on the cognitive function of MCI patients through network meta-analysis (NMA) and rank their effectiveness. METHODS: The system retrieved randomized controlled trials (RCTs) related to TCM-NPIs in the treatment of MCI from 7 databases, with the search time frame from database inception to December 26, 2025. Bayesian NMA was performed on the included articles using R 4.4.3 software. RESULTS: A total of 76 RCTs were included, involving 4,895 MCI patients and 9 types of TCM-NPIs. The NMA results showed that in terms of improving the Mini-Mental State Examination (MMSE) score effect, physical and mental exercise (MBE), massage (TN), and acupoint catgut embedding (ACE) ranked the top three. Among them, MBE (MD = 2.18, 95% CrI: 0.86 to 3.55) had the best effect, with a SUCRA of 82.56%. Regarding improvements in Montreal Cognitive Assessment (MoCA) scores, TEAS, TN, and ACE ranked in the top three, with TEAS (MD = 3.31, 95% CrI: 1.87 to 4.75) showing the best effect and a SUCRA of 88.16%. Subgroup analysis and meta-regression indicated that intervention duration and the MCI population did not have a significant impact on cognitive outcomes. In the analysis of internal difference characteristics of intervention measures, the method of moxibustion had a moderating effect on MMSE scores (p < 0.05), with moxibustion devices showing better efficacy (MD = 1.42, 95% CI: 0.96 to 1.87). TCM-NPIs demonstrated high safety, with no serious adverse events observed. CONCLUSION: Existing evidence suggests that different types of TCM-NPIs have varying therapeutic effects in improving cognitive function. Among them, MBE shows the most significant in the improvement of MMSE scores, and TEAS has the best effect in the improvement of MoCA scores. SYSTEMATIC REVIEW REGISTRATION: https://www.crd.york.ac.uk/PROSPERO/view/CRD420251248307, PROSPERO registration number: CRD420251248307.