Studying Chiropractic with Imaging: Another Dead Salmon?

Neuroimaging has a strange superpower: it can make almost any idea look expensive, sophisticated, and one grant cycle away from changing the world. Add a colorful brain map, sprinkle in a few arrows, and suddenly a modest pilot study starts strutting around like it owns the conference hall. That is part of what makes the question in this title so useful. When researchers study chiropractic with imaging, are they uncovering meaningful biology, or are they sometimes dressing up uncertainty in a very shiny lab coat?

The answer is not a cheap punch line, and it is not a lazy dismissal either. Chiropractic research, especially research on spinal manipulation, does include serious efforts to understand pain, movement, and nervous system responses. Some evidence suggests spinal manipulation can help certain people with musculoskeletal pain, particularly low back pain and some neck pain. But once imaging enters the picture, especially functional imaging, the standards need to get tougher, not softer. Otherwise, we risk repeating the lesson made famous by one of science’s most beloved cautionary props: the dead salmon.

What the Dead Salmon Actually Taught Science

The “dead salmon” story became famous because it was funny, but it stuck because it was devastatingly useful. Researchers scanned a dead Atlantic salmon with fMRI and found what looked like brain activity when loose statistical thresholds were used. No, the fish was not secretly contemplating human emotions from beyond the seafood aisle. The point was that neuroimaging can generate false positives if researchers do not properly correct for multiple comparisons. In plainer English, if you test enough tiny pieces of data, random noise will eventually cosplay as a discovery.

That lesson matters far beyond fish-based comedy. Functional MRI does not directly photograph thoughts. It measures changes linked to blood oxygenation and then relies on layers of modeling, thresholds, assumptions, and analysis choices. That is not a flaw; it is just how the method works. But it means researchers need disciplined statistics, clear hypotheses, transparent methods, appropriate controls, and enough participants to keep wishful thinking from hijacking the results.

So when a paper claims that spinal manipulation “changes the brain,” the proper response is not immediate applause or immediate mockery. It is a grown-up question: How do you know?

What Researchers Mean by “Imaging” in Chiropractic Studies

In this conversation, “imaging” can mean two very different things.

Diagnostic imaging in clinical care

This is the familiar stuff: X-rays, MRI, and CT scans of the spine. These tests can be valuable when a clinician is looking for fractures, severe neurologic compromise, cancer, infection, major trauma, surgical planning issues, or other red flags. MRI is especially helpful when persistent symptoms, worsening deficits, or specific structural concerns justify a closer look.

But here is the part that tends to ruin the glamour shot: most uncomplicated low back pain does not need immediate imaging. Evidence-based guidance in the United States has said this for years. If the patient has no major red flags, early routine imaging often adds cost, anxiety, and incidental findings without improving outcomes. Translation: sometimes the scariest thing on the scan is a perfectly ordinary age-related change that was never the true cause of the pain in the first place.

Research imaging in mechanism studies

This is where things get more intriguing and more vulnerable to hype. Researchers use tools such as fMRI, EEG, evoked potentials, spectroscopy, or source-localization methods to explore what happens in the nervous system before and after spinal manipulation. These studies are often trying to answer a reasonable question: if manual therapy changes pain, movement, or sensory processing, can we detect corresponding changes in the brain or central nervous system?

That is a fair research question. In fact, it is an important one. The trouble starts when a mechanism study is treated like a clinical verdict. Finding a short-term shift in a network signal, prefrontal response, or default mode pattern does not automatically prove durable healing, better long-term function, or superiority over other treatments. A fascinating signal is still not the same thing as a meaningful outcome.

What the Best Evidence Says About Spinal Manipulation

Let’s start with the least dramatic part, because it is also the most grounded. In the United States, spinal manipulation is not some fringe curiosity used by six mysterious people in mountain cabins. Millions of adults receive chiropractic care. Major evidence reviews and clinical summaries generally conclude that spinal manipulation may provide small to modest improvements for some musculoskeletal pain conditions, especially low back pain. For chronic low back pain, it lands among several non-drug options that may help, along with exercise and multidisciplinary approaches.

That is the sober version. Not miracle cure. Not fraud wrapped in an anatomy poster. Just one option among others, with benefits that tend to be modest, variable, and context-dependent.

At the same time, the evidence is much weaker when claims drift beyond musculoskeletal care into broad “brain optimization” territory. Reviews looking at spinal manipulation and “brain function” have found scattered signals, inconsistent findings, mixed quality, and uncertain clinical relevance. In other words, the research landscape is not empty, but it is far from a victory parade.

Why Imaging-Based Chiropractic Research Is So Tempting

There are good reasons researchers keep returning to imaging. Pain is not just a local tissue event. Chronic pain involves sensory processing, expectation, emotion, attention, learning, and behavior. The brain participates in all of that. If spinal manipulation affects pain, muscle activation, fear of movement, or sensory integration, it is entirely plausible that some neural correlates could show up in imaging or electrophysiologic measures.

And to be fair, some pilot studies have reported changes after manipulation in regions linked to attention, pain processing, or sensorimotor integration. Others have described shifts in prefrontal activity or default mode network behavior. These findings are interesting. They may help shape future hypotheses. They may eventually improve how manual therapies are studied.

But “interesting” is not the same as “settled,” and “hypothesis-generating” is not the same as “clinically proven.” This is where the dead salmon paddles back into frame, dripping with statistical judgment.

Where the Trouble Starts: Five Ways Imaging Can Oversell the Case

1. Small samples make big claims look bigger than they are

Many chiropractic imaging studies are small. Small studies are useful for exploration, but they are fragile. They are more likely to produce unstable estimates, exaggerated effects, and results that fail to replicate. A bright blob on a scan may be real, but in a small sample it may also be a very expensive coincidence.

2. Sham controls are hard, and weak controls can flatter the treatment

Manual therapy is notoriously difficult to study with airtight sham procedures. Participants often feel whether they received something active, and expectation can influence pain, attention, and even neural measures. Some authors in this field openly acknowledge that better sham conditions would strengthen the interpretation of their imaging results. That is refreshingly honest, and it is exactly the point.

3. A neural change is not automatically a health benefit

A different signal in the prefrontal cortex or default mode network may sound impressive, but the real question is whether the patient functions better, hurts less over time, uses less medication, sleeps better, or returns to daily activities. If the imaging changes do not clearly connect to outcomes people actually care about, the scan risks becoming decorative science.

4. Multiple comparisons can turn noise into theater

Neuroimaging produces mountains of data. Without rigorous correction and careful reporting, chance findings can masquerade as biology. This is the heart of the dead salmon warning. The danger is not that all imaging is meaningless. The danger is that colorful images can seduce readers into forgetting how many analytic choices sit between raw data and confident conclusions.

5. Clinical imaging and research imaging are often confused

Patients may hear that a study used MRI or fMRI and assume it proves a clinician should order more scans. Not so. Diagnostic imaging and research imaging serve different purposes. A mechanistic study about brain response after manipulation does not justify routine early imaging for uncomplicated back pain. Those are separate questions, and mixing them together creates more fog than insight.

So, Is It Another Dead Salmon?

Sometimes yes, but not always.

If a paper uses imaging to generate a careful, modest hypothesis about how spinal manipulation might influence pain processing, that is not a dead salmon. That is normal scientific exploration. If the study is transparent about limitations, uses appropriate controls, avoids grandiose language, and does not pretend a neural signal equals a cure, then it belongs in the conversation.

But if a tiny study with soft controls and flashy scans is used to imply that chiropractic care has now been proven to “rewire the brain,” “restore neurological function,” or validate sweeping claims far beyond the data, then yes, the fish starts smelling familiar.

The smarter conclusion is not that chiropractic imaging studies are nonsense. It is that they must clear a higher bar than marketing language usually admits. In this area, the prettiness of the picture should be the least persuasive part of the evidence.

What Responsible Research Would Look Like

Better chiropractic imaging research would not need dramatic headlines. It would need larger samples, preregistered methods, stronger sham comparisons, transparent statistical correction, replication across labs, and outcome measures that matter to patients. It would also clearly separate three questions that too often get blended into one smoothie of confusion:

First, does spinal manipulation help this condition in real people? Second, if it helps, how large and durable is the benefit? Third, what biological mechanisms might help explain that effect?

Those are related questions, but they are not interchangeable. Mechanism studies are valuable. They just should not be allowed to skip ahead and declare clinical victory before the clinical evidence catches up.

Bottom Line

Studying chiropractic with imaging is not inherently misguided. In fact, it can be useful. The nervous system matters in pain, and it makes sense to investigate how manual therapies interact with brain and sensory processing. But imaging is a tool, not a truth serum. It can clarify a question, or it can magnify a weak one until it looks profound.

The dead salmon remains the perfect mascot for this debate because it reminds us to stay humble. A scan can be technically impressive and still scientifically flimsy. A treatment can have some real-world value and still be oversold by mechanism claims. And a field can move forward only when it stops asking whether the image looks convincing and starts asking whether the evidence actually is.

So, is chiropractic imaging research another dead salmon? Not by default. But whenever a bright brain image is asked to do the heavy lifting for weak evidence, the fish is absolutely back on the table.

Experiences from the Lab, Clinic, and Reading Room

One of the most revealing experiences around this topic is how differently the same image feels depending on where you are standing. To a patient in pain, a scan can feel like validation. It says, “See? Something is real here.” That emotional effect matters. Pain is exhausting, and many people spend months trying to convince others that they are not exaggerating. When an MRI or brain image enters the conversation, it can seem like the argument is finally over. The picture becomes proof, even when the actual interpretation is more cautious than the patient realizes.

Clinicians often experience the opposite problem. They know images can be helpful, but they also know how often they complicate simple cases. A scan can reveal age-related changes, disc bulges, narrowing, or asymmetries that look dramatic on paper and barely matter in practice. Then begins the familiar ritual: the patient stares at the report, the clinician explains that not every abnormality is meaningful, and everyone in the room silently wishes the word “abnormal” had never been printed in such intimidating font. In that moment, imaging feels less like clarity and more like a very expensive source of panic.

Researchers have their own version of this experience. Early in a project, a new imaging result can feel thrilling. A region lights up, a connectivity pattern shifts, a graph behaves nicely, and the temptation is immediate: maybe this is the mechanism. Maybe this is the missing piece. Then the slower emotions arrive. Was the sample too small? Was the control condition convincing? Were the statistical thresholds conservative enough? Did the signal survive correction? Does the effect replicate? Can anyone honestly connect it to a real clinical improvement? This is the less glamorous side of science, where enthusiasm gets audited.

Students reading this literature often go through a mini-education in intellectual humility. At first, the colorful figures are persuasive. Later, the methods section starts hitting back. Suddenly the decisive-looking paper turns out to be a pilot trial with a tiny sample, a short follow-up, and language that sprints far ahead of the actual data. It is a useful kind of disappointment. You learn that skepticism is not cynicism. It is quality control.

Perhaps the most grounded experience of all belongs to careful practitioners, including some chiropractors, who work with patients every day and know that pain care is messy. They understand that one treatment rarely explains everything, that improvement can be gradual, and that the patient’s story matters as much as the machine. For them, imaging is neither hero nor villain. It is simply one tool that should answer a specific question. Used well, it informs care. Used badly, it distracts from care.

That may be the most useful lived lesson in this entire debate. The real world is not anti-imaging or pro-imaging. It is anti-confusion. People want honest explanations, sensible treatment, and fewer claims that sound revolutionary right up until replication arrives with a fire extinguisher.