Measured Brain Perfusion Is Not the Same as an Inferred Therapeutic Mechanism

Reading an Exploratory ASL-MRI Study of Acupuncture in Depression

When reading neuroimaging studies, I often return to a simple question:

What was actually measured, and what was inferred from it?

This question may sound basic, but it is crucial. Brain imaging findings can be persuasive. Once we see language such as “brain activation,” “normalization,” or “enhanced prefrontal engagement,” it becomes easy to feel that a therapeutic mechanism has been demonstrated.

But a measured change in the brain is not always the same as evidence of clinical improvement, disease-specific normalization, or a treatment-specific mechanism.

This issue is not limited to acupuncture. It applies equally to psychotherapy, exercise, meditation, neuromodulation, placebo research, and many other interventions. Whenever physiological signals are translated into therapeutic narratives, the distance between measurement and interpretation matters.

What ASL-MRI Measures

I recently read an exploratory ASL-MRI study examining cerebral perfusion changes during and after acupuncture stimulation in patients with major depressive disorder.

ASL-MRI is a quantitative imaging method that can estimate cerebral blood flow, or CBF. In this context, the study was valuable because it attempted to characterize immediate neurovascular responses associated with acupuncture stimulation.

However, what was directly measured was specific:

• Regional CBF changes before, during, and after a single acupuncture session
• Differences in immediate perfusion response patterns between patients with MDD and healthy controls
• Neurovascular responses under a particular experimental condition

What was not directly measured was also important:

• Improvement in depressive symptoms
• Antidepressant efficacy
• A treatment-specific mechanism of acupuncture
• Disease-specific normalization of MDD-related brain dysfunction

This distinction matters.

The Distance Between Measurement and Interpretation

In exploratory neuroimaging studies, it is common to interpret observed changes in relation to known brain networks: the prefrontal cortex, limbic regions, emotion regulation, cognitive control, and so on.

That kind of interpretation is not inherently wrong. Without some functional interpretation, neuroimaging findings would be difficult to place in a meaningful scientific context.

The problem arises when the interpretation moves too quickly from:

observed perfusion change

to:

therapeutic mechanism

or:

disease-specific normalization

A single-session, exploratory CBF finding does not by itself show that symptoms improved, that a disorder-specific pathology was normalized, or that the observed response was specific to acupuncture.

To support those stronger claims, additional evidence would be needed: clinical outcome measures, sham or placebo controls, longitudinal assessment, sufficient sample size, and careful statistical correction.

Depression Is Not a Uniform Disease Entity

This issue is especially important in depression research.

Major depressive disorder is not a pathophysiologically uniform condition. Patients diagnosed with MDD can differ substantially in symptoms, risk factors, illness trajectories, treatment responses, and long-term outcomes.

A further complication is that “depression” itself is not a single, uniform biological entity. Major depressive disorder is a diagnostic category that can include markedly different clinical presentations and underlying mechanisms. For this reason, even if an immediate perfusion change is observed in a small group of patients diagnosed with MDD, it should not automatically be read as a reversal or normalization of a disease-specific brain abnormality.

The diagnostic label may be shared, but the underlying clinical and biological meanings may differ substantially across patients.

This does not make exploratory studies meaningless. Quite the opposite.

Exploratory studies are valuable precisely because they can generate hypotheses. But their value depends on making clear what has been observed and what remains hypothetical.

This Is Not Just an Acupuncture Problem

The issue is not whether acupuncture “works” or “does not work.”

The more general point is that neural responses to an intervention are not automatically equivalent to therapeutic mechanisms.

If acupuncture stimulation is associated with increased or decreased perfusion in certain brain regions, the direct conclusion is:

Under these experimental conditions, a change in regional cerebral blood flow was observed.

That is already a meaningful finding.

But it is a further step to say:

• Therefore, depressive symptoms improved.
• Therefore, the disorder was normalized.
• Therefore, acupuncture has a specific antidepressant mechanism.

Those may be hypotheses, but they require additional evidence.

The same caution applies to many other interventions. Brain imaging is powerful. That is exactly why the language used to describe it should be careful.

What Does “The Brain Changed” Actually Mean?

In public-facing science communication, phrases like “the brain changed” or “brain activity improved” are common.

But when reading a study, we should pause and ask:

What exactly was measured? Were clinical symptoms measured? Was there a control condition? Was intervention specificity tested? Was the effect immediate or sustained? How large was the sample? How uncertain are the statistical findings?

These questions change how we read the study.

The value of neuroimaging research does not lie only in striking images or strong explanatory language. It lies in the careful relationship between data, interpretation, and clinical meaning.

Separating Four Levels of Inference

One useful way to read such studies is to separate four levels.

First: measured signal. For example, ASL-MRI–derived regional cerebral blood flow.

Second: neurobiological interpretation. For example, how the observed change may relate to the prefrontal cortex, limbic system, or emotion regulation.

Third: clinical implication. For example, whether the change may be associated with symptom improvement or therapeutic benefit.

Fourth: intervention specificity. For example, whether the change is specific to acupuncture, or whether it may also reflect touch, expectation, attention, relaxation, sensory stimulation, or contextual factors.

These levels are related, but they are not the same.

An exploratory neuroimaging study may reasonably move from the first level to the second. But moving to the third or fourth level requires a different kind of evidence.

Careful Interpretation Strengthens Research

Pointing out this distinction is not a way of dismissing the study.

On the contrary, it helps preserve the value of exploratory research.

A finding such as “immediate CBF changes were observed during acupuncture stimulation” is meaningful on its own. It does not need to be overstated as proof of an antidepressant mechanism.

Clearer separation between measurement and inference would make such studies stronger, not weaker.

It would also help guide future research, including:

• Sham-controlled designs
• Integration of neuroimaging findings with clinical outcomes
• Longitudinal assessment
• Larger samples
• Consideration of symptom heterogeneity within MDD
• More cautious language about mechanism and therapeutic meaning

Exploratory research is not only about providing answers. It is also about clarifying what should be tested next.

Conclusion

When reading neuroimaging research, I think we should keep returning to three questions:

What was measured? What was interpreted? What remains an inference?

These are modest questions, but they matter.

This is especially true in fields such as acupuncture, where clinical experience, patient expectations, bodily sensation, cultural context, and practitioner–patient interaction can all shape the meaning of treatment.

Before saying that “the brain changed,” we should ask a more basic question:

What kind of change was actually measured?

References

Matsuura Y, Kikuchi T, Yamaguchi S, Yoshimasu H, Matsuda H, Okudaira T, Yasuno F, Sakai T, Tsuchiya K. An arterial spin-labeled magnetic resonance imaging study of brain activation in patients with major depressive disorder during acupuncture stimulation: An exploratory study. Psychiatry Research: Neuroimaging. 2026;360:112220. https://doi.org/10.1016/j.pscychresns.2026.112220

Lynall ME, McIntosh AM. The heterogeneity of depression. American Journal of Psychiatry. 2023;180:703–704. https://doi.org/10.1176/appi.ajp.20230574

Jiang X, Hu Y, Lin C, Li J, Dong Z, Yang X, Lu W, Zhang N, Xu H, Huang X, Kuang W, Gong Q. Unveiling the symptom structure of major depressive disorder and exploring its role in treatment response. Journal of Affective Disorders. 2026;392:120072. https://doi.org/10.1016/j.jad.2025.120072

Disclosure

This essay was developed from an unpublished correspondence-style commentary. I used ChatGPT to assist with language refinement and structural editing. I reviewed and edited the final text and take full responsibility for its content.