LR3-LI4 needling side and configuration are associated with phase-specific cortical activation and network organization: a functional near-infrared spectroscopy study in healthy adults

Background Whether dominant-side, non-dominant-side, or bilateral acupuncture produces comparable cortical activation remains unresolved. This study compared phase-specific cortical hemodynamic responses across these configurations at LR3-LI4 using functional near-infrared spectroscopy (fNIRS). Methods In total, 53 healthy right-handed adults were randomized to dominant-side acupuncture (DS, n = 13), bilateral (BI, n = 15), non-dominant-side stimulation (NDS, n = 14), or sham (SH, n = 11). Cortical activity was measured across bilateral prefrontal, motor, and somatosensory regions during baseline, needle retention, and post-removal sessions. Functional connectivity and graph-theoretic metrics were analyzed as complementary network-level outcomes. Results During retention, DS showed greater left dorsolateral prefrontal cortex (DLPFC) and frontopolar area oxygenated hemoglobin (FPA HbO) activation than BI (DLPFC: mean difference = 0.110, 95% CI 0.036 to 0.184, p = 0.005; FPA: 0.146, 95% CI 0.044 to 0.248, p = 0.007) and SH (DLPFC: 0.070, p = 0.040; FPA: 0.114, p = 0.034). NDS showed limited prefrontal effects but greater right S1 activation than BI (mean difference = 0.179, 95% CI 0.079 to 0.279, p < 0.001). BI exhibited biphasic somatosensory responses, with initial right S1 deactivation during retention ( t = −2.944, p = 0.010) followed by post-removal rebound ( t = 3.188, p = 0.006). Functional connectivity analysis suggested configuration-specific associative patterns: DS showed dynamic frontoparietal-somatosensory coupling, BI showed delayed frontoparietal-motor connectivity enhancement, and NDS showed delayed cross-hemispheric sensorimotor coupling. All active groups showed enhanced bilateral FPA and right DLPFC-FPA connectivity compared with SH post-removal ( F = 3.207, p = 0.031; F = 3.189, p = 0.031). Graph analysis further indicated phase-related network reorganization, with significant effects for global efficiency (FDR p = 0.021) and clustering coefficient (FDR p = 0.024). Conclusion In healthy right-handed adults, LR3-LI4 needling configuration was associated with distinct phase-specific cortical response patterns. DS was associated with greater retention-phase prefrontal HbO activation than BI; NDS showed a more somatosensory-dominant pattern, and BI showed attenuated retention-phase responses followed by post-removal network changes. These findings provide short-term neurophysiological evidence that the needling side may influence cortical responses, but they do not establish clinical superiority. Patient-based studies are needed to determine whether these cortical signatures predict therapeutic outcomes.