The CSI, short for Central Sensitization Inventory, has 25 items for a score out of 100 and a screening threshold at 40. What exactly it measures is debated, and that debate changes how to use it. The interpreter is right below.
The threshold of 40 comes from a single study, in 121 patients and 129 non-patients, and it was built to screen, not to measure a mechanism. The detail is below.
Figures taken from the studies listed at the end of this page, every value carrying its source where it is written. The questionnaire itself is not reproduced.
What the CSI measures
The CSI was published in 2012 to screen for patients with what its authors call a central sensitivity syndrome: fibromyalgia, chronic fatigue syndrome, irritable bowel syndrome and a few other pictures with no identified organic cause.
It has 25 items and yields a score out of 100. Its measurement qualities are good from the original publication onwards: test-retest reliability at 0.82 and internal consistency at 0.88. The short form published in 2026, across 7,862 participants from 23 countries, brings the instrument down to seven items without losing much, with an internal consistency of 0.85.
One detail of that original publication is worth keeping, because it announces everything else. Factor analysis yields four factors, and its authors describe all of them as somatic and emotional. The content of the questionnaire is made of reported symptoms: fatigue, sleep, tension, digestive trouble, sensitivity to smells and to light.
The threshold of 40, and what it separates
The threshold of 40 out of 100 comes from a single study, published in 2013 in 121 patients of a pain centre and 129 non-patients.
Its performance there is respectable: area under the curve 0.86, sensitivity 81 %, specificity 75 %. In other words, out of four patients with a central sensitivity syndrome, the threshold picks up three; out of four subjects without one, it classifies three correctly.
The word its authors use is screening, not diagnosis. A score above 40 signals that it is worth looking; it establishes nothing on its own. The 2020 multicountry review, across 2,620 subjects from eight countries, indeed proposes reasoning in three severity levels rather than in crossing a single threshold.
Symptoms or mechanism: the literature is split
This is the point that decides how the CSI is used, and the literature does not answer it with one voice.
On one side, the 2024 meta-analysis, across 33 studies, 3,314 subjects and 154 effect sizes, finds significant correlations between the CSI and all five quantitative sensory testing modalities examined. The strongest concern pain thresholds, in particular the pressure pain threshold, where half the effect sizes are medium to large.
On the other, the 2025 study run in 77 patients with painful temporomandibular disorder and 101 controls finds no correlation with temporal summation nor with conditioned pain modulation, while anxiety, depression, catastrophising and sleep alone explain 68.9 % of the variation in scores. Its authors call for caution in that population. The reasonable reading is therefore this one: the CSI describes a symptomatic experience that often accompanies sensitisation without reducing to it, and a high score reads with the psychosocial context, never against it.
| What is measured | Value | Sample |
|---|---|---|
| Screening threshold | 40 / 100 | 121 patients, 129 non-patients |
| Sensitivity of that threshold | 81 % | same groups |
| Specificity of that threshold | 75 % | same groups |
| Area under the curve | 0.86 | same groups |
| Test-retest reliability | 0.82 | original publication |
| Internal consistency, 25 items | 0.88 to 0.93 | original, then 23 countries |
| Internal consistency, short form | 0.85 | 7,862 participants |
| Variance explained by psychosocial factors | 68.9 % | 178 subjects, temporomandibular disorder |
No validated French version
This is a practical limit, and a sharp one.
The CSI has been translated and validated in German, Finnish, Italian, Persian, Korean, Chinese, Arabic, Polish, Turkish, Russian and Nepali. No validated French version is indexed in PubMed as of 31 August 2026, neither under that name, nor under a French phrase, nor by a team from a French-speaking country.
What that implies in practice: the French translations in circulation have not shown that they keep the measurement properties quoted above, and the threshold of 40 has not been tested on them. Using the CSI in French remains possible to open a conversation; drawing a figure from it for the notes calls for saying so.
What this page does not do
It does not reproduce the 25 items, nor those of the short form. Nor does it offer a French translation: none exists that is validated and indexed, and manufacturing one would put into circulation a questionnaire whose measurement properties would not be established. The tool at the top of the page reads a score obtained elsewhere, and recalls where the threshold used to place it comes from.
Three administration pitfalls
Reading a score above 40 as a diagnosis
The threshold was built to screen, with 75 % specificity. One subject in four without a central sensitivity syndrome crosses it anyway: the score opens a question, it does not close one.
Presenting it to the patient as a measure of the nervous system
The items ask about fatigue, sleep, digestion, sensitivity to smells. Announcing that the figure measures the state of a central nervous system goes beyond what the content of the questionnaire allows, and what the patient takes away from it can weigh heavily.
Following a score without watching what moves alongside
Two thirds of the variation in scores is explained, in at least one population, by anxiety, depression, catastrophising and sleep. A CSI that falls while sleep improves proves nothing beyond that.
Frequently asked questions
What does CSI stand for?
CSI stands for Central Sensitization Inventory. The questionnaire was published in 2012 by Mayer et al. to screen for central sensitivity syndromes, fibromyalgia among them.
How is the CSI scored?
It has 25 items each scored from 0 to 4, for a total score out of 100. A seven-item short form, published in 2026, offers an alternative validated across 23 countries.
What does a score above 40 mean?
Forty out of a hundred is the screening threshold, established with a sensitivity of 81 % and a specificity of 75 %. It signals that a central sensitivity syndrome is worth looking for, it does not establish one.
Does the CSI measure central sensitisation?
The question is open. A 2024 meta-analysis finds correlations with all five quantitative sensory testing modalities; a 2025 study in temporomandibular disorder finds none with two of them and explains 68.9 % of the variance by psychosocial factors.
Is there a validated French version?
No, none is indexed in PubMed to date, while a dozen other languages have one. The translations in circulation have not shown that they keep the properties of the original.
References
6 sources, PMIDs included
- Mayer TG, Neblett R, Cohen H, Howard KJ, Choi YH, Williams MJ, Perez Y, Gatchel RJ. The development and psychometric validation of the central sensitization inventory. Pain Pract 2012;12(4):276-85. PMID 21951710. The original publication. Test-retest reliability 0.82 and internal consistency 0.88. Factor analysis yields four factors, all reported as somatic and emotional, explaining 53.4 % of the variance. Patients with fibromyalgia score highest, the control population lowest.
- Neblett R, Cohen H, Choi Y, Hartzell MM, Williams M, Mayer TG, Gatchel RJ. The Central Sensitization Inventory (CSI): establishing clinically significant values for identifying central sensitivity syndromes in an outpatient chronic pain sample. J Pain 2013;14(5):438-45. PMID 23490634. The study that sets the threshold, in 121 patients of a pain centre and 129 non-patients. A score of 40 out of 100 best separates the two groups: area under the curve 0.86, sensitivity 81 %, specificity 75 %. The authors present the tool as a screening instrument.
- Cuesta-Vargas AI, Neblett R, Nijs J, Chiarotto A, Kregel J, van Wilgen CP, Pitance L, et coll. Establishing Central Sensitization-Related Symptom Severity Subgroups: A Multicountry Study Using the Central Sensitization Inventory. Pain Med 2020;21(10):2430-40. PMID 33118603. 2,620 subjects from eight countries, chronic pain patients and healthy subjects, split into two samples for cross-validation. Hierarchical clustering then latent profile analysis give the same answer: three severity levels, low, medium and high.
- Neblett R, Sanabria-Mazo JP, Luciano JV, Mirčić M, Čolović P, Bojanić M, Jeremić-Knežević M, Aleksandrić T, Knežević A. Is the Central Sensitization Inventory (CSI) associated with quantitative sensory testing (QST)? A systematic review and meta-analysis. Neurosci Biobehav Rev 2024;161:105612. PMID 38604015. 39 studies reviewed, 33 meta-analysed, 3,314 subjects and 154 effect sizes, across five quantitative sensory testing modalities. Significant correlations emerge for all five, the strongest with pain thresholds, in particular pressure pain threshold, where 51 % of effect sizes are medium to large.
- Salbego RS, Conti PCR, Soares FFC, Ferreira DMAO, Herreira-Ferreira M, de Lima-Netto BA, Costa YM, Bonjardim LR. Central sensitization inventory is associated with psychological functioning but not with psychophysical assessment of pain amplification. Eur J Pain 2025;29(2):e4713. PMID 39120067. 77 patients with painful temporomandibular disorder and 101 controls. The CSI correlates there with anxiety, depression, catastrophising, sleep and stress, which explain 68.9 % of the variation in its scores. It correlates with neither temporal summation nor conditioned pain modulation.
- Neblett R, Navarrete J, Knezevic A, Madi M, Caumo W, et coll. Development of a Central Sensitization Inventory short form using data from twenty-three countries. J Pain 2026;43:106256. PMID 41796622. 7,862 participants from 23 countries. A multi-step Rasch analysis cuts the 25 items down to seven, with an internal consistency of 0.85 against 0.93 for the long version. The CSI-7 tells fibromyalgia from healthy controls with an area under the curve of 0.98.
Page written by Anthony Baillon, physiotherapist, co-founder of Physio Learning. The questionnaire itself is by Mayer et al., 2012: this page documents and interprets it, it reproduces none of its items.