Comparative efficacy of non-invasive brain stimulation on cognition function in patients with mild cognitive impairment: A systematic review and network meta-analysis.

Background: Mild cognitive impairment (MCI) is a critical time window for implementing prevention strategies to attenuate or delay cognitive decline. Non-invasive brain stimulation (NIBS) techniques are promising non-pharmacological therapies for improving the cognitive function of MCI, but it is unclear which type of NIBS protocol is most effective. This study aimed to compare and rank the beneficial effect of different NIBS methods/protocols on cognitive function and examine the acceptability of NIBS in patients with MCI.

Methods: Electronic search of PubMed, Cochrane Library, EMBASE, China National Knowledge Infrastructure, Wanfang Database, and Chongqing VIP Database up to November 2023. Patients with diagnosis of MCI were included. The primary outcomes were acceptability and pre-post treatment changes in global cognitive function, and the secondary outcomes were specific cognitive domains (language and executive function). All network meta‑analysis procedures were performed under the frequentist model. A protocol for this systematic review was registered in PROSPERO (Registration number: CRD42023441448).

Results: A network meta-analysis was conducted on 19 eligible RCTs consisting of 599 subjects. Compared with the sham stimulation, Repetitive Transcranial Magnetic Stimulation over the Bilateral dorsolateral prefrontal cortex (rTMS-F3F4) showed the strongest improvement in global cognitive function in MCI patients (SMD =1.52[95 %CIs =0.49-2.56]), followed by rTMS over the left dorsolateral prefrontal cortex (rTMS-F3) (SMD =1.25[95 %CIs =0.57-1.93]); Moreover, rTMS-F3F4 showed more significant efficacy in language function (SMD =0.96[95 %CIs = 0.20-1.72]); No statistically significant differences were found among the other cognitive domains. Compared with the rTMS-F4, rTMS-F3F4 showed a stronger improvement in global cognitive function in MCI patients (SMD =1.80[95 %CIs =0.02-3.59]). Similar results were obtained in subgroup analyses of cognitive function. All the methods were well-tolerated with an acceptable safety profile.

Conclusion: The present findings provide evidence of the benefits of NIBS, especially TMS stimulating the bilateral dorsolateral prefrontal cortex, for the beneficial effect on cognitive and language function in patients with MCI. However, because few studies were available for inclusion, additional well-designed, large-scale RCTs are warranted to support exploring longer-term dynamic effects.