A common sequence: the pain is widespread and persistent, the investigations are done, and they come back unremarkable. The implication people are left with — sometimes stated, more often not — is that an absence of findings means an absence of cause. That inference rests on an assumption worth examining, which is that pain should be proportional to tissue damage.

The Assumption

The intuitive model of pain is a signalling one. Damage occurs, nerves report it, the brain registers the report, and the intensity of the experience corresponds to the extent of the injury.

It works well for acute injury. It fails, repeatedly and in both directions, everywhere else — people with substantial joint degeneration who report little pain, and people with severe persistent pain and unremarkable imaging.

A model that mispredicts in both directions is not a slightly imperfect model. It is describing something other than what determines pain intensity.

The Concept That Exists to Explain the Gap

A 2026 review puts it in one sentence: central sensitisation explains the mismatch between structural damage or inflammation and pain intensity in chronic musculoskeletal disease (Journal of Pain Research, 2026).

The idea is that the pain system’s own gain has changed. The processing of signals in the central nervous system becomes amplified, so a given input produces a larger output than it once did.

That is not a metaphor for imagined pain. It is a description of altered processing in the pathway that produces the experience — and it makes specific, testable predictions.

What It Looks Like When Measured

The review lists four measurable manifestations.

Hyperalgesia — a painful stimulus producing more pain than it should. Allodynia — pain from something that should not be painful at all, such as light touch or clothing. Expanded receptive fields — the area from which pain can be triggered growing beyond the original site. And impaired endogenous pain inhibition — the body’s built-in dampening system working less effectively.

That fourth one is easy to skim past and is arguably the most important. There is a descending system whose job is to turn pain signalling down. If it is working less well, the experience intensifies without anything new happening at the tissue.

These are assessed with tools including the Widespread Pain Index and Symptom Severity Scale, and with quantitative sensory testing and algometry — measurement, not inference from a description.

The measurement point is worth dwelling on because of what it rules out. Quantitative sensory testing and algometry apply controlled stimuli and record thresholds; they do not ask someone to rate how they feel in general. A threshold that has shifted is an observation, not a report.

Not a Fibromyalgia-Only Story

The review states that central sensitisation defines the phenotype of fibromyalgia — but it also contributes to persistent pain in osteoarthritis, rheumatoid arthritis and psoriatic arthritis, in ways that may not correlate with inflammation or structural damage.

That matters for the argument. If the same mechanism operates in conditions with unambiguous visible pathology, then it is not a category invented for people whose scans are clear. It is a general feature of how persistent pain works, more prominent in some conditions than others.

What Imaging Does and Does Not Show

A separate 2026 systematic review, PROSPERO-registered and following PRISMA 2020, searched four databases for studies published between January 2010 and October 2025 comparing adults with fibromyalgia against healthy controls. Ninety-one studies qualified: 24 morphometry, 8 diffusion or structural connectivity, 32 resting-state fMRI and 40 task-based fMRI (European Journal of Pain, 2026).

The breadth of that search is part of the finding. Ninety-one studies across four distinct methodological families — measuring volume, white-matter structure, activity at rest, and activity during tasks — is close to the whole available literature for one condition over fifteen years.

Its most consistent morphometric result was reduced grey matter volume or cortical thickness in anterior cingulate and insular regions — both regions central to how pain is evaluated rather than merely detected.

Both regions are worth naming properly. The anterior cingulate and the insula are consistently implicated in the evaluative and interpretive side of pain — how much a signal matters, how threatening it is — rather than in detecting it. A structural difference concentrated there is consistent with a change in appraisal rather than in sensation, though a systematic review of cross-sectional comparisons cannot establish that.

The review is equally clear about the state of the evidence: neuroimaging indicates central nervous system involvement, yet findings remain inconsistent across methods. Ninety-one studies and the honest summary is still that the methods disagree.

What This Does Not Mean

It does not mean the pain is produced by the brain in the sense of being invented. Altered central processing is a physical change in a physical system. ‘Central’ describes where in the pathway, not whether it is real.

It does not identify a cause. Central sensitisation describes a state of the pain system. Why it develops in one person and not another is not answered by either review.

It is not diagnostic imaging. No scan diagnoses fibromyalgia. The imaging review’s own conclusion is that findings are inconsistent across methods, and group-level differences do not transfer to individual scans.

And it does not follow that any particular treatment works. Establishing a mechanism is not the same as establishing what changes it, and neither review is a treatment trial.

What it does support is the thing most worth saying: an unremarkable scan is not evidence that nothing is wrong. There is a described, measurable mechanism for pain that exceeds what the tissue explains — which is also why the body’s signals and their interpretation are worth separating.