In fibromyalgia, the capacity to make a plan and follow it through is not a moral failing or a question of effort. It is a measurable change in higher-order executive function — visible on neuropsychological tests, on the EEG, on the fMRI — and it travels with the pain.
"Planning deficit disorder" is not a clinical diagnosis. The closest peer-reviewed framing is impairment in high-order executive function — specifically the capacity to organise, sequence, and execute goal-directed action — observed in fibromyalgia syndrome (FMS) and grouped clinically under the umbrella of dysexecutive symptoms1. Patients describe it as "fibrofog." Neuropsychologists test it with the Zoo Map Test, the Key Search Test, and the Tower of Hanoi12.
The distinction matters. "Brain fog" sounds atmospheric, vague, easy to wave off. Planning and problem-solving impairment is a specific, reproducible deficit on standardised tasks that have been used to evaluate the same domain in stroke, Parkinson's disease, and mild cognitive impairment. That fibromyalgia patients underperform on these instruments — with effect sizes in the moderate-to-large range — is no longer in serious dispute158.
A useful way to hold this: the symptom is not a fog rolling across the road. It is the blueprint itself, drawn carefully and then dissolving at the edges before the building can be built.
The clinical literature has a phrase for what this feels like from the inside: the knowing–doing gap — the felt severance between full understanding of what needs to happen and an inability to assemble the steps to make it happen. When this pattern dominates other cognitive functions, the named syndrome is Dysexecutive Syndrome (DES), a term introduced by neuropsychologist Alan Baddeley in the late 1980s to distinguish impaired planning, sequencing, and behavioural control from the older "frontal lobe syndrome" label5. Baddeley's point was that the failing element is a function, not just a region. DES is the umbrella. Selective plan-generation deficit is one of its sharpest sub-patterns — and the one fibromyalgia produces most cleanly.
Cognitive psychology has converged on a hierarchical model of executive function. Adele Diamond's framework, building on Miyake and colleagues, identifies three core executive components — and a layer of high-order functions that rest on them1.
In a 2025 study, thirty women with fibromyalgia and thirty matched healthy controls completed a battery of these tests1. The fibromyalgia group scored significantly lower on the harder, less-structured versions of each high-order task — the Zoo Map Test (planning) and the five-disk Tower of Hanoi (sequential problem-solving). Effect sizes were moderate to large (η²p between 0.08 and 0.11), with the pattern holding even after controlling for anxiety and depression for several tasks.
Crucially, the simple components predicted the high-order failures. Worse mental flexibility predicted worse Key Search performance. Worse interference control predicted worse Zoo Map 2 performance. Worse working-memory updating predicted more moves on the Tower of Hanoi1. The plan does not fail because the patient is unmotivated. The plan fails because the substrate that builds plans is degraded at three levels at once.
This is not a marginal observation about a minority of patients. A 2025 case–control study found measurable cognitive dysfunction in 72.3% of fibromyalgia patients versus 5.3% of healthy controls16. Cognitive impairment is not a side note to the condition. It is a core feature.
The lived report runs like this: once the plan exists, I can carry it out; it is the making of the plan that I cannot do. That distinction is not vague self-criticism. It maps onto a sharp anatomical split that cognitive neuroscience has been making for thirty years.
Cognitive planning — the upstream act of constructing a future sequence of steps, holding it in working memory, checking that the steps cohere, deciding when the plan is "ready" — runs primarily in the dorsolateral prefrontal cortex, with supporting contributions from the cerebellum and globus pallidus19. It is generative and effortful, and it happens before any action begins. Plan execution — running the steps once they exist — is handed off to a different network entirely: the supplementary motor area, the motor cortex, and basal-ganglia procedural circuits19. fMRI studies show these two systems are spatially and temporally separable.
This matters because fibromyalgia's neural footprint — reduced dlPFC activity on fMRI, depleted prefrontal dopamine, structural microstructural loss in dlPFC and orbitofrontal cortex — sits squarely on the upstream system318. The downstream execution machinery is largely preserved. A person with fibromyalgia who is handed a finished plan can usually carry it out. The thing they cannot reliably do is generate the plan in the first place. Independent neuropsychological work on related conditions confirms the split: patients can score within normal ranges on plan-execution tasks while showing profound deficits in plan-generation knowledge20.
The clinical implication is precise. The symptom is not "I have lost the ability to think." It is "the module that builds plans has been throttled, while every other cognitive capability I have is sitting there intact, waiting for an executable plan to act on." The blueprint metaphor still holds — but it is sharper than first drawn. The blueprint is missing not because the building is impossible to build. The blueprint is missing because the room where blueprints are drawn has had its lights dimmed.
The dorsolateral prefrontal cortex, the dorsal anterior cingulate, the anterior insula — these are the regions activated when a brain holds a plan in mind, suppresses a competing impulse, or switches strategy. They are also the regions activated when a brain processes pain, especially pain that has been re-evaluated cognitively (catastrophised, anticipated, attended to)36.
EEG studies of fibromyalgia patients performing cognitive control tasks show reduced frontal brain electrical activity during interference suppression, and that reduction correlates with subjective cognitive complaints6. Functional MRI shows lower activation in the ventrolateral prefrontal cortex, thalamus, and inferior parietal cortex compared with healthy controls during cognitive tasks34. Structural imaging adds another layer: gray-matter volume reductions in the dorsolateral PFC, dorsal anterior cingulate, and orbitofrontal cortex have been reported in fibromyalgia cohorts7.
Real-world data complete the picture. In chronic-pain patients, executive function performance predicts how much pain interferes with daily life — independent of pain intensity itself10. The deficit is not abstract; it shows up at the kitchen counter, at the workplace, at the moment a plan needs to be enacted.
Working memory — the capacity to hold a plan in mind while executing each step — depends on a dopamine-mediated pathway connecting the hippocampus to the prefrontal cortex. The hippocampus consolidates context; the PFC manipulates it. Dopamine is what keeps the line open3.
Pathophysiology reviews report that dopamine is reduced by approximately 30% in fibromyalgia patients, with the most pronounced loss in hippocampal regions3. Prolonged stress exposure elevates glucocorticoids, which combined with excessive NMDA-receptor activation contributes to hippocampal dysfunction and, over time, hippocampal atrophy3. The result is what the literature calls neural noise — a signal-to-noise problem in which the cognitive substrate becomes harder to read out, even when the underlying machinery is intact.
Imaging studies have specifically documented working-memory deficits associated with an altered frontoparietal memory network in fibromyalgia patients9. The deficit is not psychological resistance to thinking. It is a network operating with less of the neuromodulator it needs to hold its own outputs steady.
Microglia are not bystanders. When activated by sustained nociceptive input, infection, or stress signalling — including via Toll-like receptor 4 (TLR4) — they release pro-inflammatory cytokines, substance P, nitric oxide, and excitatory amino acids13. The downstream consequences are stereotyped: cognitive impairment, fatigue, sleep disruption, mood disturbance, amplified pain. Sickness behaviour, made chronic.
This is no longer inferred from drug response or animal models. A 2025 review summarising recent hybrid MR/PET imaging studies in fibromyalgia patients reports widespread microglial activation measurable in vivo — indexed by uptake of a TSPO radioligand that binds preferentially to activated microglia and astrocytes — in regions overlapping with the cognitive networks already described. Increased binding in parietal grey matter correlates directly with cognitive symptoms, pain intensity, and reduced quality of life17. A separate diffusion-kurtosis imaging cohort found reduced microstructural integrity in bilateral dorsolateral prefrontal cortex and orbitofrontal cortex in FM patients, with elevated amyloid-β as an additional inflammatory marker18. The picture is no longer "the cells are probably activated." It is "the activation is visible on a scan, and it tracks the symptoms."
Mohapatra and colleagues performed bulk RNA sequencing on peripheral blood mononuclear cells from ninety-six fibromyalgia patients and ninety-three healthy controls15. The cohort did not separate cleanly from controls along a single inflammatory axis. Instead, the patients clustered into distinct transcriptomic subgroups, each with a characteristic dominant pathway. That finding has implications well beyond classification: if four patients carry four different molecular signatures, then a clinical trial that averages across them may report a null result even when each subgroup responds to a different mechanism.
Independent analytic work using a covariance-geometry approach — the Dynamical State Index (DSI), defined as the ratio of the two largest eigenvalues of the cohort gene-expression covariance matrix — confirms this pattern. Healthy cohorts present anisotropic geometry, with one biological axis dominating. Fibromyalgia cohorts present compressed, near-spherical geometry: no single program is in charge. Within the patient cloud, four sub-clusters emerge that align with the Mohapatra signatures15. The biological interpretation is that fibromyalgia is not driven by one inflammatory axis but by a loss of coordination — immune static where there ought to be immune signal.
The clinical consequence of this is that "fibromyalgia" — as it appears on a chart — is a coarse label over at least four molecular states. Trial outcomes that look modest in aggregate may be hiding strong responses in a single subgroup. Cognitive symptoms may track more strongly with some subgroups than others. The field has not yet mapped this in detail, but the geometry says it should.
Naltrexone, at the high doses used to treat opioid dependence, is a μ-opioid receptor antagonist. At low doses — typically 1.5 to 6 mg daily — it does something different. It crosses the blood–brain barrier and acts as an antagonist at Toll-like receptor 4 (TLR4) on microglia1314. TLR4 is the receptor that activates microglial inflammation in response to danger signals, including those generated by sustained nociceptive input. Antagonising it damps the cascade that produces the cytokines, substance P, nitric oxide, and excitatory amino acids responsible for cognitive impairment, fatigue, and amplified pain.
Systematic reviews and meta-analyses of LDN in fibromyalgia consistently report modest but statistically significant reductions in pain. A 2023 PMC systematic review and a 2024 meta-analysis with trial sequential analysis both found a mean pain-score reduction in the order of −0.86 points (95% CI −1.20 to −0.51, p<0.001) versus placebo, with favourable safety profiles11. A 2023 Lancet Rheumatology randomised controlled trial of 6 mg once-daily naltrexone in women with fibromyalgia added the largest single RCT to the evidence base, contributing to but not transforming the overall signal12.
On cognitive symptoms specifically, the evidence is thinner. The 2023 systematic review notes that results indicate LDN might improve memory complaints associated with fibromyalgia and recommends future trials investigate this directly11. Case reports and open-label series describe improvements in "cognitive impairment" and "waking unrefreshed" alongside pain reduction11, but these are not the same level of evidence as a powered RCT with cognition as a primary endpoint. That trial has not yet been done.
There is one more thread worth pulling. If fibromyalgia is genuinely heterogeneous at the molecular level — four subgroups, distinct dominant pathways, different geometries — then the LDN response may not be uniform across the cohort. The subgroup carrying the strongest microglial / neuroinflammatory signal is the one in which the TLR4-antagonism mechanism should bite hardest. The subgroup driven primarily by ECM dysregulation or HERV reactivation may respond differently or not at all. Stratified trials, not averaged trials, are what the geometry asks for.
The plan does not hold because the substrate that builds plans is being asked to share a finite resource with a pain signal that does not turn off.
Naming it precisely is not the same as fixing it. But it is the first thing a person who has been told for years that they are simply not trying hard enough is owed.