Acupuncture can mobilize endogenous opioid system and produce analgesia. However

there is some variability of the analgesic effect between individuals

which is related with cholecystokin level of the subject. In the present study

we try to use functional magnetic resonance imaging (fMRI) to address the question by which neuropathways CCK influences acupuncture induced analgesia. Two different frequencies of transcutaneous electric acupoint stimulation (TEAS) were administered to normal human body. TEAS was used instead of traditional acupuncture for its similar analgesic effect and convenience of parameter adjustment. Our previous studies have indicated that low and high frequency TEAS generated effects through different neuropathways

in which some nuclei of midbrain

thalamus and hypothalamus played an important role. 25 healthy volunteers were randomly divided into two groups to receive low and high frequency TEAS respectively. Data from functional MRI scanning (EPI sequence

TR=3000 ms

TE=45 ms

Flip angle: 90°) were collected at the same time of stimulation. Basal and after TEAS pain thresholds were measured by radiant heat withdrawal test 1～3 days before fMRI examination and changes of pain threshold were calculated as the index of analgesic effect of TEAS. Functional data were processed with cross correlation of timecourse and stimulation curve after registration

normalization and detrending. The averaged signal intensity of every interested region was linear regressed according to the change of pain threshold. We found that in low frequency TEAS group

activation intensity of contralateral primary and supplementary motor areas (MI and SMA)

bilateral secondary somatosensory area (SⅡ)

contralateral thalamus and anterior cingulate cortex (BA 24)

ipsilateral superior temporal gyrus

insula had a significant linear correlation with the change of pain threshold. Furthermore

the signal intensity of bilateral hippocampus and the change of pain threshold had a negative linear correlation. While in the areas of primary somatosensory area (SⅠ) and inferior parietal lobule (BA 40)

such linear correlation did not exist. In the case of high frequency TEAS

the results were similar to those mentioned above

that is

the corresponding somatosensory areas and the connective cortex were all activated. While the relevant motor related areas were seldom activated. Parts of the limbic system such as bilateral amygdala

perigenual anterior cingulate cortex

nucleus accumbence

and premedial frontal cortex

were inhibited

among which signals of amygdala and nucleus accumbence were negatively related with the analgesic effect. Our results suggested that different frequencies of TEAS stimulation activated different brain areas. Some brain areas are specifically involved in the TEAS analgesia process. They may play an important role in the acupuncture analgesic neuropathways.