Acupuncture, Brain Imaging and Neuroscience Research

What brain imaging can tell us

Researchers use functional MRI (fMRI, which tracks blood-flow changes linked to brain activity) and PET scans (which can track chemicals binding in the brain) to see what happens in the brain during or after acupuncture. NCCIH says studies in animals and people, including imaging studies, have shown that acupuncture may affect nervous system function. [1] That is a measured change, not proof that it helps symptoms. This page reports what specific reviews and studies found. Evidence labels: promising but limited, mixed, preclinical, insufficient evidence. For the wider list of mechanisms, see How Acupuncture May Work.

The big fMRI review

A systematic review searched English, Chinese, Korean and Japanese databases and found 779 papers, of which 149 were described and 34 were pooled in meta-analysis. [2] Its findings:

• Brain response to acupuncture covered a broad network consistent with sensory, emotional and thinking processes, not just touch. [2]
• Real versus sham acupuncture differed in the middle cingulate, but other regions varied with how the analysis was done. [2]
• The authors called for more high-quality studies with more transparent methods, to improve consistency between studies. [2]

That makes the overall status mixed: many studies report changes, but the details do not line up well.

Resting-state connectivity

A second review covered 44 resting-state fMRI studies, in which people lie still and the scan measures how brain regions coordinate. [3] Nineteen of these studies, with 574 people in total, involved only healthy adult volunteers. [3] Compared with sham, real acupuncture could increase connectivity of the default mode and sensorimotor networks with pain-, emotion- and memory-related regions. [3] The authors "presumed" that connectivity is closely related to the mechanism; that is an interpretation, not a demonstrated cause. Healthy volunteers also differ from people with chronic pain, so results may not transfer.

PET: opioid receptors

A randomized PET study in people with fibromyalgia compared traditional and sham acupuncture over 4 weeks. [4] Real acupuncture was followed by increased opioid-receptor binding in several pain and sensory regions, while the sham group showed small reductions. [4] Long-term increases were linked to greater reductions in clinical pain. [4] This is the clearest human example in the sources we read, but it is one study in one condition.

Real needling, belief and imagination

• Expectancy. When people had high expectations, real and sham acupuncture produced equally strong reported pain relief, but the real treatment produced a greater drop in pain-related brain signal. [5] The authors read this as different mechanisms for acupuncture analgesia and expectancy-driven placebo analgesia. [5]
• Imagery. In a crossover study with 27 healthy volunteers, both real acupuncture and a video-guided imagery task raised pain threshold compared with their controls, and real acupuncture produced greater activation of the insula. [6] The study was in people without pain, so it cannot show benefit for chronic pain.

What imaging cannot say

• A scan shows a brain change; it does not show whether the person feels better in daily life.
• Many studies used only healthy volunteers. [3]
• Different analysis choices gave different results in the main fMRI review. [2]
• A brain difference between real and sham needling does not mean the difference is large enough to matter for a patient. For clinical results, see Does Acupuncture Work?.

Whether imaging can be used to choose who should receive acupuncture: proper evidence not found in the sources we read. We searched PubMed for acupuncture neuroimaging reviews and read the studies cited here.