Longitudinal Changes in Hippocampal-Default Mode Network Connectivity and Cognition During Electroacupuncture Intervention in Type 2 Diabetes-Associated Cognitive Decline: A Pilot Resting-State fMRI Study.

Abstract

Type 2 diabetes mellitus-associated cognitive dysfunction (T2DACD) is strongly linked to disruption of cognition-supporting brain networks. Electroacupuncture (EA) has increasingly been examined as a nonpharmacological intervention targeting cognitive impairment; however, the longitudinal neurobiological patterns accompanying EA in T2DACD have not been fully characterized. In this exploratory pilot study, we evaluated alterations in resting-state functional connectivity (RSFC) across hippocampal subregions and default mode network (DMN) nodes in individuals with T2DACD. For this exploratory, open-label, non-randomized, uncontrolled pilot investigation, we enrolled 20 individuals with T2DACD and 26 healthy controls (HCs). Longitudinal analyses included 15 participants who had usable post-treatment data. T2DACD was defined according to clinical type 2 diabetes mellitus criteria, and a Montreal Cognitive Assessment (MoCA) score <26. Patients underwent an 8-week Adjust Zang-fu and Arouse Spirit electroacupuncture (AZASE) intervention. Cognitive, emotional, metabolic, and seed-based resting-state functional magnetic resonance imaging (rs-fMRI) outcomes were evaluated. Clinical longitudinal analyses were carried out according to protocol, whereas imaging analyses were limited to participants with usable scans after treatment. At baseline, individuals with T2DACD showed lower MoCA scores than HCs and had poorer executive and working-memory outcomes. Among the 15 participants with complete longitudinal neuroimaging data, MoCA scores increased from 23.27 ± 1.71 to 26.07 ± 1.53 over the intervention period. Within-group alterations were observed for Flanker-conflict reaction time and accuracy, Stroop-conflict reaction time, 1-back reaction time, Self-Rating Anxiety Scale score, and Hamilton Depression Rating Scale score. Omnibus group analyses detected significant effects for the left anterior hippocampus-left angular gyrus connection and the left posterior hippocampus-right middle occipital gyrus connection. Post-treatment scans demonstrated reduced RSFC in several hippocampal-DMN and intra-DMN connections. The association between change in left anterior hippocampus-left angular gyrus RSFC and change in Stroop-conflict performance was no longer statistically significant after Bonferroni correction. For neuroimaging outcomes, cluster-level family-wise error (FWE) correction (p < 0.05) was applied to control for multiple comparisons; clinical outcomes are reported descriptively as hypothesis-generating findings. In this exploratory pilot study, we identified preliminary alterations in cognitive outcomes and hippocampal-DMN connections over an 8-week EA intervention in individuals with T2DACD. In view of the small sample size and the open-label, non-randomized, uncontrolled design, the findings should be interpreted as hypothesis-generating observations rather than definitive evidence of a specific causal treatment effect. Larger controlled studies will be needed to determine whether the observed neural and behavioral alterations are reproducible and clinically meaningful in practice. No: ChiCTR2000040268, https://www.chictr.org.cn/hvshowproject.html?id=76145&v=1.3.

In the evidence

This study is part of the evidence base for a synthesis in the library. Here is how each one recorded it.

• State of the evidence: Default mode network ↕Mixed

  After 8 weeks of electroacupuncture, MoCA scores increased and several hippocampal-DMN and intra-DMN resting-state connections decreased, presented as preliminary hypothesis-generating observations.

  Synthesized October 19, 2026

Comparable studies

Other non-randomized and open-label trials on default mode network, most cited first.