The temporal-spatial encoding of acupuncture effects in the brain

Abstract

Background: Functional acupoint specificity is crucial to the clinical efficacy of acupuncture treatment, such as pain relief. Whether acupuncture needling at a peripheral acupoint produces distinct patterns of brain responses remains controversial.

Results: This fMRI study employed the complex network analysis (CNA) to test the hypothesis that acupuncture stimulation at an acupoint correspondingly induced activity changes in one or more intrinsic or resting-state brain networks. Built upon the sustained effect of acupuncture and its time-varying characteristics, we constructed a dynamic encoding system with the hub anchored at the posterior cingulate cortex and precuneus (PCC/pC). We found that needling at two visual acupoints (GB37 and BL60) and a non-visual acupoint (KI8) induced a spatially converging brain response, which overlapped at the PCC/pC. We also found distinct neural modulations during and after acupoint stimulation. During this period, the PCC/pC interacted with a visual resting-state network in different patterns. Furthermore, there was a delayed functional correspondence between the intrinsic visual network and manipulation over the visual acupoints (i.e., GB37 or BL60), but not the non-visual acupoint (KI8) via the PCC/pC, implicating a specific temporal-spatial encoding/decoding mechanism underlying the post-effect of acupuncture.

Conclusions: This study provided an integrated view exploring the functional specificity of acupuncture in which both the needling sensation and the following neural cascades may contribute to the overall effect of acupuncture through dynamic reconfiguration of complex neural networks.fMRI, acupoints, posterior cingulate cortex, precuneus, temporal-spatial encoding, resting-state networks.

Figure 1

Experimental paradigm and the schematic of the current research frame. There are three scan sessions: a pre-acupuncture resting state (baseline) scan lasting 6 minutes; a needling stimulation scan lasting 6 minutes with a single trail electroacupuncture (EA) stimulation; and a post-needling scan lasting 12 minutes. Datasets RUN1, RUN31 and RUN32 are truncated from the pre-acupuncture resting state and the post-needling scan (see the Methods for details). This paper focuses on the the delayed post effects of acupuncture, which refers to the post-needling scan. Therefore, datasets RUN1, RUN31 and RUN32 are used for further analysis.

Figure 2

Results of psychophysical Deqi sensations. A. The intensity of the reported sensations measured by an average score (with standard error bars) on a scale from 0 denoting no sensation to 10 denoting an unbearable sensation. The average stimulus intensities were approximately similar during acupuncture at GB37, BL60 and KI8. No statistical differences were found among groups with regards to each sensation (p < 0.05). B. The percentage of subjects who reported having experienced the given sensation (at least one subject experienced the seven sensations listed). Numbness, fullness, dull pain, heaviness and soreness were found no statistical differences for all three groups (p < 0.05).

Figure 3

Results of conjunction analysis based on the DCT results. The masks of group-level DCT results of all three groups, i.e. GB37, BL60 and KI 8, were used to generate this map. The only overlapped region located in the posterior cingulate cortex and precuneus (PCC/pC) (p < 0.001, corrected, cluster size > 20 voxels, see the Methods for details).

Figure 4

Contrasting results between post-acupuncture functional connectivity and pre-acupuncture functional connectivity for three groups. The voxel-wised functional connectivity analysis was conducted using the PCC/pC as a seed point. The paired t-test compared the sustained effects with the baseline, i.e, RUN 31 v.s. RUN1 and RUN 32 v.s. RUN1) and the results were thresholded at P < 0.005 (uncorrected) and the cluster size > 5 voxels. Warm (cold) colors indicated the increase (decrease) in functional connectivity.

Figure 5

Illustration of variations of modulating effect of acupuncture over different acupoints. Comparisons of the functional connectivity were conducted between the visual acupoint group (GB37 or BL60) and the non-visual acupoint group (KI 8) during the late phase of acupuncture (Run 3): (A) GB37 vs. KI 8; (B) BL60 vs. KI8. The two sample t-test were used to determine significant between-group differences (P < 0.005 with a cluster size of 5 voxels). (C) The averaged time courses of fMRI BOLD signals in the occipital lobe, where showed significant between-group differences (circles on A and B). The temporal responses revealed almost opposite response patterns after stimulations over visual acupoint and the non-visual acupoint, i.e. GB37 vs. KI 8 and BL60 vs. KI8. The correlation coefficient between GB37 and KI 8 was -0.35 (P < 0.05, Bonferroni correction) and that of BL60 vs. KI8 was -0.18 (P < 0.05, Bonferroni correction).

Figure 6

The results from graphic theory-based CNA (see Methods), showing the topological and functional maps of the visual RSN (consisting of the yellow nodes) and its dynamic interactions with the PCC/pC (the red node) as a system encoding/decoding hub with respect to three different acupoints (two visual acupoints: GB37 and BL60, and a non-visual acupoint KI 8). The red lines referred to the significant (p < 0.05, see the Methods for details) positive functional connections among the visual function-related network, and the blue lines the significant negative connections.