The question the researchers started with is a basic one: when someone who already lives with chronic pain is given a controlled experimental pain stimulus, what does the brain do that is different from what it does in someone without chronic pain?

It is a cleaner question than it sounds. Most of what we know about pain processing comes from studies on people without chronic pain — healthy volunteers in controlled conditions, receiving brief stimuli, measuring the brain’s acute response. That baseline tells you something, but it does not tell you what happens when the nervous system has been carrying a persistent pain signal for months or years. Whether chronic pain changes how the brain processes new pain — not just during flares, but in response to any fresh stimulus — is a different question, and one that requires looking directly at a different population.

In a 2026 systematic review and meta-analysis published in The Journal of Pain, Al-Faraj, Hanel, and Valentini did exactly that. They gathered 48 whole-brain functional magnetic resonance imaging studies that had presented experimental pain stimuli — heat, pressure, cold — to participants with chronic primary pain diagnoses alongside pain-free controls, and subjected the imaging data to coordinate-based meta-analysis. The total sample came to 2,052 participants. Their aim was not to confirm that the brains of people with chronic pain look different. It was to determine, with the best available statistical methods, where those differences appear and what they tell us — and where the evidence is not yet strong enough to support a conclusion.

What Is Chronic Primary Pain, and Why Does the Distinction Matter?

The International Association for the Study of Pain introduced the chronic primary pain framework into the eleventh revision of the International Classification of Diseases in 2019. The category was created to recognize a group of conditions that had been poorly served by a classification logic that required chronic pain to be explainable by underlying tissue pathology. In primary pain conditions — which include fibromyalgia, primary headache disorders, migraine, irritable bowel syndrome without a confirmed intestinal cause, and several others — the pain itself is the condition. It is not a symptom of something else that has not yet been found.

This distinction matters for research design. If pain is being treated as a secondary phenomenon, studies tend to focus on the primary pathology. When pain is the phenomenon of interest, researchers can ask directly: what is the nervous system doing that sustains this signal?

Al-Faraj and colleagues focused their meta-analysis on this category specifically, drawing studies that examined participants whose diagnoses fell within the CPP framework. The conditions represented in their included studies were primarily migraine, fibromyalgia, and irritable bowel syndrome. This is not a uniform population — the pain experiences of a person with fibromyalgia and a person with migraine differ in quality, location, and temporal pattern. What the researchers were looking for was not what is the same about these conditions, but what pattern of brain activation is detectable across them when pain processing is directly challenged.

Three Regions That Activate Consistently Across Conditions

The meta-analysis produced a conjunction analysis of 39 within-subject experiments — that is, studies that measured brain responses to experimental pain within the same participants, rather than comparing groups. Across those experiments, three regions showed consistent activation: the dorsal anterior insula, the mid-cingulate gyrus, and the medial frontal gyrus.

Each of these regions has an established role in pain processing, but not the same role. The dorsal anterior insula is associated with interoception — the brain’s representation of the body’s internal state, including sensory dimensions of pain. It is active in the processing of pain intensity and the felt quality of painful experience. The mid-cingulate gyrus is involved in what researchers describe as the affective-motivational dimension of pain: the part that generates the sense of unpleasantness, drives behavioral responses, and — in healthy nociception — serves as a signal that prompts the organism to attend and respond. The medial frontal gyrus contributes to the cognitive evaluation of pain and its meaning in context.

The co-activation of these three regions across diverse conditions and study designs suggests that what the meta-analysis is detecting is not specific to migraine, or to fibromyalgia, or to any single experimental stimulus. It is a consistent fingerprint of the provoked-pain response in people living with chronic primary pain — a distributed response that spans sensory, affective, and evaluative dimensions.

What this does not establish is a mechanism. Finding that three regions activate is not the same as knowing why, or what sequence of events produces that activation, or whether the activation pattern causes chronification, follows from it, or both (Al-Faraj, Hanel & Valentini, The Journal of Pain, 2026).

Where Patients and Controls Diverge: The Same Stimulus, Different Priorities

When Al-Faraj and colleagues moved from the within-subject analysis to a comparison between participants with chronic primary pain and pain-free controls, a clearer picture of the differences emerged.

People with chronic primary pain showed stronger activation in the anterior insula relative to controls. People without chronic pain showed stronger activation in the mid-cingulate gyrus relative to those with chronic pain.

This divergence is worth dwelling on. Both regions are active in both groups — the between-subject contrast is about relative weight, not presence or absence. But the difference in emphasis is functionally meaningful. Greater interoceptive engagement in the chronic pain group suggests a nervous system that is particularly attentive to the body’s internal signals during a painful event — one that may be tracking the sensory dimensions of pain more intensively. The stronger mid-cingulate engagement in controls suggests a nervous system that is allocating more resource to the motivational response component — the alarm, the urgency, the behavioral signal to move or withdraw.

One interpretation is that in chronic pain, the system has reorganized its priorities: more continuous interoceptive monitoring, less acute alarm response, because the alarm has already been running. This is a plausible interpretation of the imaging pattern, but it remains interpretive. Functional MRI measures blood-oxygen-level-dependent signal as a proxy for neural activity; the translation from signal differences to mechanistic differences involves assumptions that the researchers themselves flag as unresolved.

The meta-analysis also noted that existing heterogeneity in study design — different pain stimuli, different acquisition protocols, varying levels of statistical threshold — limits how confidently any single conclusion can be drawn. The authors conclude that methodologically harmonized studies are needed before the CPP framework can function as a clinical diagnostic tool rather than a research classification (Al-Faraj, Hanel & Valentini, The Journal of Pain, 2026).

Structural Changes: How Chronic Pain May Reshape Cortical Architecture

The imaging findings from Al-Faraj and colleagues describe what the brain does during a pain event. A separate set of studies published in 2026 describe what chronic pain appears to do to the brain over time.

A large individual-participant-data meta-analysis by Loke and colleagues, published in Brain Communications (2026), examined cortical structure across five chronic pain conditions, combining data from 401 individuals with chronic pain and 245 pain-free controls. Rather than relying on a single structural metric, the researchers used multiple simultaneous measures — cortical thickness, surface area, and intrinsic curvature — to build a more complete picture of whether and how cortical architecture differs in people with chronic pain.

The findings were widespread. Cortical thinning appeared in temporo-occipital regions. Volumetric reductions were detected in several areas. Most strikingly, intrinsic curvature — the folding and shape of the cortical surface — was increased across 49 of 68 cortical regions examined. The researchers characterized the overall pattern as chronic pain functioning as a disorder of large-scale cortical reorganization: not a focal change in one pain-processing area, but a distributed shift across the cortex (Loke et al., Brain Communications, 2026).

A smaller but methodologically distinct study by Häuselmann and colleagues, published in NeuroImage: Clinical (2026), examined 30 participants meeting criteria for chronic primary pain alongside 30 matched controls. Using multimodal MRI, they found increased gyrification in left prefrontal areas and reduced sulcal depth in right frontal, orbitofrontal, and temporo-occipital regions. They also found evidence of disrupted relationships between the stress system and brain architecture in the CPP group — a finding that connects the structural changes to the known role of chronic stress in pain chronification (Häuselmann et al., NeuroImage: Clinical, 2026).

Neither of these structural studies can establish causal direction. Brain structure changes with experience across many contexts — learning, stress, illness, medication. Whether the structural differences observed in chronic pain populations preceded the pain, developed alongside it, or were produced by it cannot be determined from cross-sectional MRI data. Both studies also worked with specific samples and diagnostic categories that may not represent the full range of chronic primary pain. The pattern they describe is consistent across multiple methods and research groups; what it means mechanistically remains an open question.

Fibromyalgia's Neuroimaging Signature: What 91 Studies Found

Fibromyalgia presents a particular challenge for neuroimaging research. It is among the most studied chronic primary pain conditions in terms of imaging, and it is also the one where findings have been most inconsistent across studies — different imaging protocols, different patient groups, different definitions, different statistical thresholds.

A 2026 systematic review by Flutt and colleagues, published in the European Journal of Pain, attempted to map the most reproducible findings by gathering 91 neuroimaging studies examining adults with fibromyalgia. The studies included morphometric work, diffusion imaging, resting-state functional MRI, and task-based functional MRI — which means the review was synthesizing multiple types of evidence about what the brain looks like, how its regions connect at rest, and how it responds during pain and other tasks.

Across the structural literature, the most consistent finding was reduced grey matter volume in three regions: the anterior cingulate cortex, the insular cortex, and the prefrontal and orbitofrontal cortices. These are regions associated with pain affect, interoceptive processing, and top-down regulation — which is to say, the cortical machinery involved in evaluating and modulating pain signals rather than merely transmitting them.

The functional literature pointed toward a more specific mechanism. Flutt and colleagues identified impaired top-down pain control as the most robust finding across methods. The specific pathway implicated was the connection between the rostral anterior cingulate cortex and the periaqueductal grey and brainstem — a descending modulation pathway that normally suppresses or attenuates incoming pain signals. When this pathway functions well, the brain can dampen pain signals from below. In fibromyalgia, the evidence from multiple study types suggested this dampening is impaired: the prefrontal engagement was dysregulated, and the descending modulation pathway showed weakened connectivity.

This is consistent with the nociceptive sensitization hypothesis for fibromyalgia — the idea that the central nervous system has become more rather than less responsive to pain signals over time. The neuroimaging pattern Flutt and colleagues describe is not a single-study finding; it appears across structural, resting-state, and task-based work, which adds weight to it. What it does not establish is the direction of causation or the precise molecular mechanisms involved (Flutt et al., European Journal of Pain, 2026).

What the Research Does Not Yet Establish

The pattern described across these four studies is consistent and comes from multiple methodologies: a coordinate-based fMRI meta-analysis, an individual-participant-data structural meta-analysis, a multimodal MRI case-control study, and a systematic review of 91 neuroimaging papers. That degree of convergence across methods and research groups is meaningful evidence that something real is being measured. It does not, by itself, resolve the most important questions.

On causation: every structural imaging study included here is cross-sectional. They measured participants at one point in time. Whether the brain changes documented in people with chronic pain developed before the pain began, emerged alongside it as it chronified, or are a consequence of years of living with persistent pain — including the stress, sleep disruption, and behavioral changes that accompany it — cannot be determined from any of these designs. Animal models and longitudinal studies are making progress on this question, but the human imaging literature cannot yet answer it.

On individual prediction: every figure in these studies is a group-level average. The divergent insula-versus-mid-cingulate pattern between chronic pain patients and controls describes what differs on average across thousands of participants. It does not describe what any individual’s scan will show, and it is not, at present, a diagnostic test. Al-Faraj and colleagues are explicit about this in their conclusion: the CPP framework needs methodologically harmonized studies before functional imaging can move from a research instrument to a clinical one.

On fibromyalgia specifically: the Flutt et al. review synthesized 91 studies, but those studies varied widely in methodology, sample selection, and statistical approach. The most consistent findings describe a direction — reduced grey matter in affective-evaluative regions, impaired top-down modulation — but the precise magnitude of these changes, and the degree to which they vary across fibromyalgia subtypes or treatment histories, has not been reliably established.

On the group boundary: the studies distinguish between people who meet criteria for chronic primary pain diagnoses and those who do not. Many people who experience chronic pain symptoms do not meet formal diagnostic criteria, and some who meet criteria have atypical presentations. The imaging findings speak to defined clinical populations; extrapolating them to everyone who experiences persistent or recurrent pain requires caution.

The value of this research is not that it tells a person with chronic pain what is happening in their brain right now. The value is that it demonstrates, with increasing methodological precision, that chronic pain is a nervous system phenomenon — one with measurable, distributed, multi-regional correlates — and that the assumption that pain must have a localized peripheral cause has not survived contact with the imaging data.