Results of RSN Analysis
Group maps of the best fitting independent components for the DMN and SMN demonstrated consistently robust spatial distribution relative to their respective templates, both before and after verum () and sham (data not shown) acupuncture. The best fitting component for the DMN demonstrated resting connectivity within the inferior parietal lobule, posterior cingulate, and medial areas of the inferior, middle and superior frontal gyri, as well as the precuneus. The best fitting component for the SMN involved bilateral primary somatosensory and motor cortices, secondary somatosensory cortex and the supplementary motor area (SMA).
When comparing DMN connectivity after ACUP with that before, there was increased connectivity of this network to a number of areas (, ) including the amygdala, hippocampal formation, periaqueductal gray (PAG), substantia nigra (SN), middle temporal gyrus (MTG, ~BA 21), supplementary motor area (SMA), and anterior cingulate (ACC, ~BA 24), posterior parietal (~BA 7), and primary visual (V1) cortices. There were no areas of decreased connectivity following ACUP. Conversely, after SHAM, the DMN demonstrated increased connectivity with only the temporo-occipital cortex (~BA 37/39) and decreased connectivity with areas of the middle (~BA 21) and inferior temporal (~BA 20) gyri (, ). This “decreased” connectivity resulted from a greater z-score before SHAM than after (2 rightmost columns, )
Changes in functional connectivity of the DMN following ACUP and SHAM (After – Before)
Brain regions in the DMN and SMN modulated by both verum and sham acupuncture.
After ACUP, the SMN demonstrated increased connectivity with the anterior cingulate cortex (~BA 32/24), pre-SMA (~BA 8/6), and cerebellum (simplex, lateral hemispheric zone 1, VI l) (, ). Conversely, after SHAM, the SMN demonstrated decreased connectivity with the dorsolateral prefrontal cortex (~BA 8, greater before versus after, , ).
The changes in functional connectivity of the SMN for ACUP and SHAM (After – Before)
RSN Correlation with Acupuncture-induced ANS modulation
In order to explore the relationship between changes in brain functional connectivity and acupuncture-induced ANS modulation, we calculated the inter-subject Pearson correlation coefficient between the change in HRV metrics (LFu, HFu) during acupuncture minus the pre-stimulus rest and the change in z-statistic in ROIs taken from significant regions from the paired t-test above for both DMN and SMN. The metrics demonstrated consistent correlation with increasing hippocampal formation connectivity to the DMN following ACUP (LFu: r = −0.73, HFu: r = 0.68, p<0.01, ). This result suggests that increased parasympathetic and decreased sympathetic modulation during ACUP is associated with increased post-ACUP DMN connectivity with the hippocampal formation. In addition, the change in post-ACUP DMN connectivity with the MTG also demonstrated a negative correlation with LFu (r = −0.70, p<0.01). These data passed the Grubb's test for outliers (p<0.05). No other significant correlations were found for either ACUP or SHAM, though it should be remembered that correlations were tested only for regions significant in the fMRI group analysis (paired t-test). Thus, while regions such as the insula (particularly the right anterior insula) are associated with the central autonomic network and have demonstrated brain response correlated with HRV [19
], the insula was not a region which demonstrated either increased or decreased connectivity to either the DMN or SMN following acupuncture and was not included in this analysis.
Changes in connectivity of the DMN following Acupuncture are correlated with HRV
Results of Psychophysical Analysis
A statistical analysis found that MASS Index, a measure of deqi intensity, was greater for ACUP compared to SHAM stimulation (paired t-test, p<0.01). Furthermore, there was a greater intensity (paired t-test, p<0.05) of spreading for ACUP (3.6±2.4) compared to SHAM (1.7±1.5). Interestingly, SHAM was associated with greater relaxation (−1.2±2.4, μ±σ) compared to ACUP (0.2±2.3, μ±σ). There was no significant difference in the intensity of sharp pain (ACUP: 3.7±2.7; SHAM: 4.9±2.6) or throbbing (ACUP: 3.9±1.9; SHAM: 2.6±2.4). An omnibus test (Fisher's combined probability test) found that ACUP and SHAM also differed in the commonality of sensations, p(n=14) < 0.001. Differences also existed in the prevalence of specific deqi sensations elicited (). The prevalence of “aching” (ACUP: 86.7% of subjects, SHAM: 40.0%, p<0.01), “fullness” (ACUP: 46.7%, SHAM: 13.3%, p<0.05), and “dull pain” (ACUP: 93.3%, SHAM: 40.0%, p<0.005) was found to be greater for ACUP (uncorrected for multiple comparisons due to significant omnibus test). It should also be noted that while sharp pain was reported for both conditions, when debriefed, most subjects who reported sharp pain described this sensation as only transient, occurring typically at the start of a stimulation block.
Results of psychophysical analysis
We also wanted to explore the relationship between changes in brain functional connectivity and the intensity of sensations evoked by acupuncture. We calculated the inter-subject Pearson correlation coefficient between the intensity of acupuncture-evoked sensations and the change in z-statistic in ROIs taken from the paired t-test above for both DMN and SMN. We found a trend for increasing ACUP evoked “soreness” correlating with increasing SMN connectivity with pre-SMA (r = 0.65, p=0.081, Bonferroni corrected). This result is consistent with the hypothesis that changes in resting SMN connectivity, a network containing brain regions subserving the sensory-discriminative aspects of pain, are related to the sensations evoked by acupuncture stimulation. No significant correlations were found for DMN ROIs, nor for any ROIs from the sham acupuncture paired t-test.
For the rest runs following ACUP and SHAM, sensation intensity scores were collected for 11 subjects. Of these, four subjects reported feeling any sensation at some point during the rest run immediately following ACUP (0.9±1.3, overall μ±σ) and two following SHAM (0.4±0.8). Ultimately, there was no significant difference between mean sensation intensity in the rest runs following ACUP versus those following SHAM (p>0.1).