Central sensitization in psoriatic arthritis: prevalence,clinical correlates,and association with therapeutic refractoriness in an unselected cohort excluding fibromyalgia

Abstract

Background:

Central sensitization (CS) contributes to persistent pain in psoriatic arthritis (PsA), yet its prevalence, independent predictors, and association with therapeutic refractoriness remain incompletely characterized across disease phenotypes.

Objectives:

To evaluate CS prevalence and clinical correlates in unselected PsA patients excluding fibromyalgia, with emphasis on predictors, multidimensional pain phenotypes, and the association between CS and therapeutic refractoriness.

Design:

Cross-sectional observational study.

Methods:

A total of 228 consecutive PsA patients were recruited from April to November 2025 at a single tertiary center. Patients with fibromyalgia (2016 modified ACR criteria, systematically screened) were excluded. CS was assessed using the Central Sensitization Inventory (CSI ⩾40). Disease activity, functionality, pain phenotypes (PAIN DETECT questionnaire (PDQ)), sleep quality (Pittsburgh Sleep Quality Index), depression and anxiety (Hospital Anxiety and Depression Scale), and physical activity (International Physical Activity Questionnaire) and therapeutic refractoriness were systematically measured. Binary and linear regression analyses identified independent predictors of CS.

Results:

CS was present in 35.1% of patients. CS was associated with higher disease activity indices (Clinical Disease Activity Index for Psoriatic Arthritis: 14 vs 9.8, p = 0.001; Ankylosing Spondylitis Disease Activity Score with C-reactive protein: 4.8 vs 1.8, p = 0.007), greater functional disability (Health Assessment Questionnaire—Disability Index: 0.8 vs 0.2, p = 0.001), increased neuropathic pain (55.6% vs 8.2%, p = 0.001), and notably higher prevalence of therapeutic refractoriness (61.5% vs 6.8%, p = 0.003). Independent predictors of CS presence were depression severity (HADS-D: odds ratio (OR) = 1.32, p = 0.01), poor sleep quality (Pittsburgh Sleep Quality Index: OR = 1.21, p = 0.001), and PDQ (OR = 1.07, p = 0.03). Depression, sleep quality, and fatigue explained 47% of CS variance in linear regression.

Conclusion:

Approximately one-third of unselected PsA patients experience CS, which independently contributes to therapeutic refractoriness despite inflammatory control. Psychosocial factors—particularly sleep dysfunction and depression—emerge as robust predictors of CS. These findings support systematic CS screening in PsA patients with inadequate symptomatic response and highlight the need for integrated biopsychosocial interventions addressing both inflammatory and non-inflammatory pain mechanisms.

Plain Language Summary

Central sensitization (pain hypersensitivity) is common in psoriatic arthritis and linked to poor treatment response, sleep problems, and depression

Why was this study done?

Psoriatic arthritis (PsA) is a condition that causes joint pain, stiffness, and swelling. Many patients continue to feel significant pain even when medication has successfully treated the swelling and inflammation. This type of pain may be caused by “central sensitization,” a condition where the nervous system becomes overly sensitive and amplifies pain signals. We wanted to find out how common this is in PsA patients, what factors (like sleep or mood) might be related to it, and if it makes treatments seem less effective.

What did the researchers do?

We studied 228 patients with PsA in a hospital setting. We specifically excluded patients who already had a diagnosis of fibromyalgia to ensure we were looking at pain related to PsA. We used questionnaires to measure central sensitization, pain levels, depression, anxiety, sleep quality, fatigue, and physical activity. We also checked their medical records to see if their treatments were working.

What did the researchers find?

We found that about one in three patients (35.1%) had central sensitization. These patients had higher pain scores and more difficulty with daily activities, even though their blood tests showed similar levels of inflammation to those without this condition. Importantly, patients with central sensitization were much more likely to be classified as having “difficult-to-treat” disease (refractory), often failing multiple biological medications. We also found that poor sleep, depression, and fatigue were strong predictors of having this type of pain.

What do the findings mean?

This study suggests that for many patients with PsA, persistent pain is caused by the nervous system rather than active inflammation. Doctors should check for central sensitization before switching strong medications, as these patients might benefit more from treatments that address sleep, mood, and pain sensitivity rather than just more anti-inflammatory drugs.

Keywords

central sensitization depression drug resistance nociplastic pain psoriatic arthritis sleep quality

Introduction

Psoriatic arthritis (PsA) is a chronic, systemic, inflammatory musculoskeletal disease affecting approximately 30% of patients with psoriasis. It is characterized by heterogeneous clinical presentations encompassing peripheral arthritis, enthesitis, axial involvement, and cutaneous/nail manifestations, typically developing in genetically predisposed individuals upon environmental trigger exposure.^1–3 Despite the availability of multiple therapeutic options and implementation of “treat-to-target” strategies, sustained remission rates remain limited at 15%–35% in routine clinical practice.^4,5 In most cases, persistent clinical activity is attributable to chronic non-inflammatory pain, a phenomenon that merits distinct mechanistic investigation.

Central sensitization (CS), defined as amplification of neural signaling within the central nervous system that elicits pain hypersensitivity, has emerged as a significant pathophysiological mechanism in various chronic musculoskeletal and rheumatic diseases.^6–8 CS is characterized by disproportionate pain responses to both painful and non-painful stimuli, with prominent hyperalgesia and allodynia, resulting from altered central neural plasticity, including dysregulation of pain-modulating neurotransmitter systems (serotonin, noradrenaline, dopamine), dysfunction of descending inhibitory pain pathways, and spinal microglial activation.^9–11

Regarding CS as a modulator of disease phenotype and therapeutic response, in rheumatic diseases including rheumatoid arthritis, axial spondyloarthritis, and PsA, chronic peripheral inflammation can trigger neuroplastic central nervous system changes facilitating CS development.¹² Persistent nociceptive input from inflamed joints promotes progressive sensitization of spinal nociceptive dorsal horn neurons, resulting in amplification and central pain signal processing.⁶ Notably, this central amplification may operate bidirectionally: CS can result from sustained inflammatory signaling, and conversely, established CS may perpetuate apparent therapeutic refractoriness through pain amplification mechanisms independent of peripheral inflammatory control.^6,10

CS in PsA has been increasingly recognized as a contributor to persistent pain, with reported prevalence ranging from 43% to 65% across recent cross-sectional studies. While previous investigations by Salaffi et al.,¹³ Kaya et al.,¹⁴ and Alp et al.¹⁵ have advanced the understanding of CS in PsA, several significant knowledge gaps remain. First, the relative contribution of psychosocial factors (particularly sleep quality and mood disturbances) as independent predictors of CS severity deserves investigation, given emerging evidence linking these dimensions to central pain processing in other conditions. Second, the specific association between CS and therapeutic refractoriness, as well as the temporal and mechanistic nature of this relationship, has not been systematically characterized. Third, comprehensive evaluation of CS across the spectrum of PsA disease phenotypes, including systematic assessment within mixed peripheral–axial presentations, remains limited.

Notably, no prior study has simultaneously addressed the independent contribution of CS to therapeutic refractoriness, the mediating role of psychosocial factors in this association, and the consistency of these findings across PsA phenotypes in a single fibromyalgia-free cohort.

Therefore, this study aimed to comprehensively evaluate CS as a key modulator of symptoms and therapeutic response in PsA. We recruited an unselected cohort of 228 patients with PsA while excluding those with established fibromyalgia diagnosis. Specific objectives were to:

Identify robust independent psychosocial predictors of CS, with particular emphasis on the integrated sleep-depression-fatigue triad as key determinants of central pain sensitization mechanisms.

Evaluate the independent association between CS and therapeutic refractoriness, while acknowledging that this relationship may be bidirectional: CS operating both as a consequence of persistent inflammatory signaling and/or as a perpetuating factor of inadequate therapeutic response.

Comprehensively characterize multidimensional pain phenotypes (nociceptive, neuropathic, and nociplastic components) in relation to CS presence and severity, demonstrating that CS contributes independently to pain despite inflammatory control.

Characterize CS prevalence and severity in an unselected PsA population and provide exploratory evaluation of consistency of CS associations across different disease phenotypes (peripheral, mixed, and axial presentations).

Methods

This cross-sectional observational study was conducted at the University Hospital of Salamanca (CAUSA), Salamanca, Spain, from April 7, 2025, to November 30, 2025. Consecutive patients were recruited who met the following inclusion criteria:

(a) Diagnosed with PsA according to CASPAR criteria¹⁶

(b) Age ⩾18 years

The exclusion criteria were: (a) an established diagnosis of fibromyalgia or fulfillment of the 2016 modified ACR fibromyalgia diagnostic criteria at the time of enrollment¹⁷; (b) known neurological disorders potentially affecting pain perception (e.g., diagnosed peripheral neuropathy, multiple sclerosis, spinal cord injury); and (c) significant cognitive impairment or inability to understand and complete the self-administered questionnaires.

All patients underwent systematic screening for fibromyalgia using the 2016 modified ACR criteria at the time of enrollment, regardless of clinical suspicion.¹⁷

The process is summarized in the flowchart (Figure 1).

Figure 1.

Flowchart of the patient recruitment process.

The study protocol was approved by the Ethics Committee of the Hospital Universitario de Salamanca (CEIm Code: 2024 12 1779—TFG, dated April 2, 2025). The study was conducted in accordance with the principles of the Declaration of Helsinki.¹⁸

Measured variables

CS assessment

CS was measured using the Central Sensitization Inventory (CSI), a self-administered 25-item questionnaire.¹⁹ The CSI consists of two parts: Part A contains 25 questions evaluated on a 5-point Likert scale (0 = “never” to 4 = “always”), with a total score range of 0–100. CS was classified as present when CSI ⩾40 points.¹⁹ Part B of the CSI investigated the presence of 10 central sensitivity syndromes (restless leg syndrome, chronic fatigue syndrome, fibromyalgia, temporomandibular joint dysfunction, migraines, irritable bowel syndrome, multiple chemical sensitivities, cervical trauma, panic attacks/anxiety, and depression). CS severity levels were categorized as: subclinical (0–29), mild (30–39), moderate (40–49), severe (50–59), and extreme (⩾60).²⁰

Therapeutic refractoriness

For the purposes of this study, therapeutic refractoriness was defined as the failure of two or more biologic disease-modifying antirheumatic drugs (bDMARD) or targeted synthetic disease-modifying antirheumatic drugs (tsDMARD) with different mechanisms of action, in accordance with current GRAPPA recommendations.^21,22

Baseline variables

Age (years), sex (male/female), disease duration (years), disease phenotype: stratified as peripheral (pain and swelling in peripheral joints), axial (inflammatory back pain with radiographic sacroiliitis grade II or higher or syndesmophytes),²³ or mixed (meeting both criteria), smoking status (current smoker, former smoker, never smoker), treatment with conventional synthetic disease-modifying antirheumatic drugs (csDMARD), bDMARD, or tsDMARD, number of affected entheses: measured by the modified Maastricht Ankylosing Spondylitis Enthesitis Scoring (mMASES) method,²⁴ presence of current or past dactylitis (yes/no), skin disease severity was assessed using the Psoriasis Area and Severity Index (PASI), which evaluates the extent and severity of psoriatic lesions (erythema, induration, and desquamation) across four body regions (head, trunk, upper extremities, and lower extremities). Scores range from 0 to 72, with higher scores indicating greater severity,²⁵ physical activity: measured by the International Physical Activity Questionnaire (IPAQ),²⁶ recorded in metabolic equivalent of tasks (METs). The IPAQ assesses physical activity based on three characteristics: intensity (low, moderate, vigorous), frequency (days per week), and duration (minutes per day). METs were estimated by multiplying the MET score for the type of activity (3.3 for low, 4 for moderate, 8 for vigorous) by the number of days per week and minutes spent doing the activity per day.

Disease activity, functionality, impact, and type of pain

In patients with peripheral involvement, disease activity was measured using the Clinical Disease Activity Index for Psoriatic Arthritis (cDAPSA).²⁷ This is a composite index for disease activity specific for PsA. As it is a clinical index, in this case, C-reactive protein (CRP) is not taken into account, unlike in DAPSA. It is calculated by summing the tender joint count (TJC; 0–68), swollen joint count (SJC; 0–66), the patient global assessment of disease activity score (between 0 and 10 on a numerical rating scale (NRS)), and pain NRS score (0–10). CRP level (mg/dl) was also measured.

In the case of patients with axial involvement, we used the Ankylosing Spondylitis Disease Activity Score with CRP (ASDAS-CRP).²⁸ In patients with mixed phenotype (meeting criteria for both peripheral and axial involvement), both cDAPSA and ASDAS-CRP were applied concurrently to capture disease activity across both domains, as no single validated composite index currently integrates peripheral and axial manifestations in PsA.

Functional assessment

Functional ability was measured using the Health Assessment Questionnaire—Disability Index (HAQ-DI) for peripheral involvement²⁹ and the Bath Ankylosing Spondylitis Functional Index (BASFI) for axial involvement.³⁰ The HAQ-DI is a 20-question, self-administered questionnaire measuring disability in eight functional categories.²⁹ Responses are scored 0–3, with final scores ranging 0–3, where higher scores indicate greater disability. The BASFI is a 10-item patient-reported measure of spinal mobility and physical function in spondyloarthropathies, scored 0–10 with higher scores indicating greater impairment.³⁰

Disease impact

Disease impact was assessed using the Psoriatic Arthritis Impact of Disease (PsAID-12) questionnaire,³¹ a 12-item patient-reported outcome measuring the impact of PsA on activities of daily living, symptoms, and emotional well-being.

The type of pain was assessed using the PAIN-DETECT questionnaire (PDQ), a self-administered instrument developed to identify neuropathic pain. It includes seven 5-point Likert scales (0 = never, 5 = very strongly) investigating qualitative characteristics of painful sensations (burning, tingling or prickling, pain to light touch, sudden pain attacks, cold or heat sensitivity, numbness, and pain triggered by slight pressure), plus 2 additional points for radiating pain indicated on a body manikin. A question investigates pain course pattern (scored −1 to 1). PDQ scores range from −1 to 38.³² Scores ⩽12 indicate likely nociceptive pain and scores ⩾19 suggest likely neuropathic pain. Intermediate scores (13–18) were classified as uncertain and excluded solely from the categorical analysis of pain phenotype (nociceptive vs neuropathic); however, all patients were included when painDETECT was analyzed as a continuous variable in correlations, between-group comparisons, and regression models.

Comorbidities

We assessed the emotional factors using the Hospital Anxiety and Depression Scale (HADS). The HADS is a 14-item scale designed to identify people with anxiety and depression among individuals with medical conditions. Scores range from 0 to 21 for each subscale (HADS-D for depression and HADS-A for anxiety) and can be classified into one of three categories: normal (0–7), borderline abnormal indicating a possible clinical disorder (8–10), and abnormal indicating a probable clinical disorder (11–21).³³

Fatigue was assessed with a Functional Assessment of Chronic Illness Therapy (FACIT) scale, specifically, the FACIT-fatigue scale, which has been validated for PsA³⁴ and consists of 13 items assessing self-reported fatigue and its impact on activities of daily living and functioning. Items are rated on a 5-point Likert-type scale from 0 to 4 yielding a total score between 0 and 52, higher scores indicating less fatigue.

Sleep quality was assessed with a specific tool for measuring sleep quality, the Pittsburgh Sleep Quality Index (PSQI).³⁵ Using this 19-item self-report instrument, patients assess their quality of sleep over the previous 30 days. The PSQI explores seven domains: subjective sleep quality, sleep latency, sleep duration, habitual sleep efficiency, sleep disturbances, use of sleep medication, and daytime dysfunction. Each domain is scored on a range between 0 and 3, and these scores are summed to obtain a global score ranging between 0 and 21 points, 0 reflecting no difficulties at all and 21 serious difficulties in all domains assessed. A PSQI score ⩾6 is considered to indicate poor sleep quality.

Obesity

Obesity was measured by body mass index (BMI).³⁶

Statistical analysis

Quantitative variables were expressed as means and standard deviations (SD) for normally distributed variables and as medians and interquartile ranges (IQR) for non-normally distributed variables. Qualitative variables were expressed as numbers and percentages (n/%). Normal distribution was assessed using the Shapiro–Wilk test.

Comparative analyses

Comparisons between two groups were performed using Student’s t-test for normally distributed quantitative variables and the Mann–Whitney U test for ordinal or non-normally distributed quantitative variables. Comparisons among more than two groups were performed using one-way ANOVA for normally distributed quantitative variables and the Kruskal–Wallis H test for ordinal or non-normally distributed quantitative variables.

Correlation analysis

Correlations between two quantitative variables were performed using Spearman’s rank correlation coefficient (rho), a non-parametric measure of monotonic correlation between variables.

Multivariate regression analysis

Binary logistic regression analysis was conducted with CS presence/absence as the dependent variable, and variables statistically significant in univariate analysis (p < 0.05) or previously described in the literature as associated with CS were included as independent variables.^13–15 Results are presented as odds ratios (OR) with 95% confidence intervals (95% CI). Model goodness-of-fit was assessed using the Nagelkerke R².³⁷

Linear regression analysis was conducted with CSI score (continuous dependent variable) and variables statistically significant in univariate analysis or described in the literature were included as independent variables. Results are presented as unstandardized beta (β) coefficients with 95% CIs. Model fit was assessed using adjusted R². Both models were carried out exclusively to explore the effect of independent variables on CSI and therefore their results are purely exploratory. Despite the exploratory nature of this study, a mediation analysis will be conducted to determine how psychosocial variables (depression, sleep dysfunction, and fatigue) may mediate the relationship between disease activity and CS. The bootstrap method has been used to estimate the 95% CIs given the distributions of the variables.

To analyze the reliability and stability of the logistic regression results, the sample size was calculated based on the events per variable, the number of predictor variables, and the expected proportion. If 10 events per variable, 10 independent variables, and an expected proportion of 0.5 are established, the sample size would be 200.³⁸

The statistical power for multiple regression was 89% for 10 independent variables, a significance level of 0.05, an effect size of 0.3, and a sample size of 228.

Statistical significance was set at p < 0.05.

Statistical analyses were conducted using SPSS version 28 (IBM Corp, Armonk, NY, USA) for Windows.³⁹ Statistical power of the models and mediation analyses were conducted using R packages “powermediation” v. 0.3.4 and mediation v. 4.5.1 (R Foundation for Statistical Computing, Vienna, Austria).

This study is reported in accordance with the Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) guidelines. The completed STROBE checklist is provided as File S4.⁴⁰

Results

In our cohort, 35.1% of patients presented with CS. The mean CSI was 34.6 (SD: 14.3). The distribution of the different states of CS is shown in Figure 2.

Figure 2.

Distribution of central sensitization states.

Relationship of CS with therapeutic refractoriness and baseline variables

Patients with CS demonstrated a significantly higher prevalence of therapeutic refractoriness (61.5% vs 6.8%, p = 0.003). Women had a higher frequency of CS than men. A sex-stratified descriptive analysis of CS prevalence, disease activity, pain phenotypes, psychosocial factors, and functional outcomes is provided in Table S1. In addition, patients with CS had more enthesitis and performed less physical activity. The rest of the results related to baseline variables are shown in Table 1. Of note, only three patients (1.3%) presented with axial-predominant disease, precluding any meaningful statistical comparison for this subgroup. Accordingly, all data pertaining to axial-predominant PsA should be considered purely descriptive.

Table 1.

Therapeutic refractoriness and baseline characteristics according to central sensitization status.

	Total (228)	No CS (148)	CS (80)	p
Age (median/IQR)	56 (13)	56 (12)	54.5 (14)	0.4
Sex (M/F) (N)	139/89	99/49	40/40	0.01
Refractory (yes/no)	26/202	10/138	16/64	0.003
Disease duration (years) (median/IQR)	9 (9)	8.5 (9)	9 (8)	0.9
Disease form (peripheral/mixed/axial)	197/28/3	129/17/2	68/11/1	0.9
Smoking status (S/Ex/N)	60/99/69	36/67/45	24/32/24	0.6
csDMARDs (yes/no)	144/84	94/54	50/30	0.8
Methotrexate	110	71	39	0.6
Leflunomide	9	7	2
Sulfasalazine	25	16	9
bDMARDs or tsDMARDs (yes/no)	73/155	41/107	32/48	0.05
iTNFα	50	32	18	0.3
iIL17	10	3	7
iIL12/23	5	2	3
iIL23	4	2	2
iJAKs	3	1	2
Apremilast	1	1	0
Enthesitis (median/IQR)	0 (2)	0 (2)	2 (4)	0.01
Dactylitis (yes/no)	49/179	31/117	18/62	0.7
PASI (median/IQR)	0.9 (2.4)	0.7 (2.2)	1.1 (3.5)	0.1
Physical activity (METs; median/IQR)	2613.0 (6492.2)	2940.5 (7644)	1386.2 (5812)	0.03

bDMARD, biologic disease-modifying antirheumatic drug; CS, central sensitization; csDMARD, conventional synthetic disease-modifying antirheumatic drug; DMARD, disease-modifying antirheumatic drug; IQR, interquartile range; METs, metabolic equivalent of task; mMASES, modified Maastricht Ankylosing Spondylitis Enthesitis Score; PASI, Psoriasis Area Severity Index; tsDMARD, targeted synthetic disease-modifying antirheumatic drug.

In the correlation analysis, the CSI score correlated with the number of affected entheses (rho = 0.26, p < 0.001) and physical activity (rho = −0.15, p = 0.03). We found no correlation with age (rho = −0.1, p = 0.1), disease duration (rho = 0.02, p = 0.6), or PASI (rho = 0.1, p = 0.1).

Relationship of CS with activity, functionality, impact, and PDQ

Patients with CS had more disease activity in the peripheral forms related to pain, activity, and the number of tender joints. Similar results were found in patients with axial manifestations. In the peripheral forms, we also found differences in functionality. Finally, the impact of the disease and neuropathic pain was greater in the group of patients with CS. The results are shown in Table 2.

Table 2.

Activity, functionality, disease impact, and PDQ.

	Total	No CS	CS	P
cDAPSA (median/IQR)	11(10)	9.8(10.4)	14(8)	0.001
TJC (median/IQR)	1(2)	1(2)	2(2)	0.004
SJC (median/IQR)	1(1)	1(1)	2(1)	0.1
Pain VAS (median/IQR)	4(4)	3(4)	5(3)	0.001
Activity VAS (median/IQR)	4(4)	4(5)	5(3)	0.001
CRP (mg/dl) (median/IQR)	0.2(0.4)	0.2(0.4)	0.2(0.3)	0.1
ASDAS-CRP (mean/SD)^b	2.3(1.3)	1.8(0.7)	4.8(3.3)	0.007
HAQ-DI (median/IQR)^a	0.5(1)	0.2(0.6)	0.8(0.7)	0.001
BASFI (mean/SD)^b	2.1(0.8)	3.5(2.8)	4.2(2.9)	0.3
PSAID (median/IQR)	3.1(3.3)	1.9(2.8)	4.5(2.4)	0.001
PAIN DETECT (median/IQR)	8(10)	6.0(8)	13(11)	0.001
PAIN DETECT (Nociceptive/Neuropathic)	30/158	10/122	20/36	0.001

In peripheral and mixed forms (n = 225).

In axial and mixed forms (n = 31).

ASDAS-CRP, Ankylosing Spondylitis Disease Activity Score with C-reactive protein; BASFI, Bath Ankylosing Spondylitis Functional Index; cDAPSA, clinical Disease Activity Index for Psoriatic Arthritis; CRP, C-reactive protein; HAQ-DI, Health Assessment Questionnaire-Disability Index; IQR, interquartile range; PDQ, PAIN-DETECT questionnaire; PsAID, Psoriatic Arthritis Impact of Disease; SD, standard deviation; SJC, swollen joint count; TJC, tender joint count; VAS, visual analogue scale.

The CSI score correlated with cDAPSA (rho = 0.4, p < 0.001), and within its components, it was related to Pain VAS (rho = 0.3, p < 0.001), Activity VAS (rho = 0.3, p < 0.001), and TJC (rho = 0.2, p < 0.001). It also correlated with ASDAS-CRP (rho = 0.5, p < 0.001), HAQ (rho = 0.4, p < 0.001), PsAID (rho = 0.5, p < 0.001), and PD (rho = 0.4, p < 0.001). On the other hand, we found no correlation with CRP (rho = −0.05, p = 0.4), SJC (rho = 0.1; p = 0.06), or BASFI (rho = 0.3, p = 0.09).

Relationship of CS with comorbidities

Patients with CS had a higher score on the HADS-A, HADS-D, and PSQI questionnaire, as well as more fatigue. We found no relationship with BMI. The results are shown in Table 3.

Table 3.

Comorbidities.

	Total	No CS	CS	p
HADS-A (median/IQR)	5(5)	4(4)	7.1(6)	0.001
HADS-D (median/IQR)	3(6)	2(3)	6.1(5)	0.001
PSQI (median/IQR)	7(8)	5(3)	11.1(7)	0.001
FACIT-F (median/IQR)	39(16)	43(12)	32(13)	0.001
BMI (kg/m²) (median/IQR)	26.1(5.8)	26(5.3)	27.2(6.6)	0.7

BMI, body mass index; CS, central sensitization; FACIT-F, Functional Assessment of Chronic Illness—Fatigue; HADS-A and HADS-D, Hospital Anxiety and Depression Scale Anxiety and Depression subscales, respectively; IQR, interquartile range; PSQI, Pittsburgh Sleep Quality Index.

We found a correlation between the CSI score and HADS-A (rho = 0.4, p < 0.001), HADS-D (rho = 0.4, p < 0.001), and the PSQI (rho = 0.4, p < 0.001), as well as more fatigue (rho = −0.5, p < 0.001). There was no correlation with BMI (rho = 0.01, p = 0.1).

In the binary logistic regression analysis using the presence of CS as the dependent variable (present/absent) and sex, enthesitis, Pain VAS, TJC, PD, HADS-A, HADS-D, PSQI, FACIT-F, and PASI as independent variables, the following results were obtained: PD (p = 0.03, OR = 1.07 (95% CI: 1.00–1.14)), HADS-D (p = 0.01, OR = 1.32 (95% CI: 1.06–1.65)), and PSQI (p < 0.001, OR = 1.21 (95% CI: 1.09–1.36)). The Nagelkerke R² value of the model was 0.55, and the Hosmer–Lemeshow test was not statistically significant (Chi-square test = 10.824, df = 8, p = 0.212). The sensitivity and specificity of the model were 62.7% and 86.4%, respectively, and the accuracy was 78.8%. Figure 3 shows a forest plot with the results of the logistic model.

Figure 3.

Forest plot of odds ratios and their 95% confidence intervals estimated in the logistic model.

In the linear regression analysis (Supplemental Material) using the CSI value as the dependent variable and sex, enthesitis, Pain VAS, TJC, PD, HADS-A, HADS-D, PSQI, FACIT-F, and PASI as independent variables, the following results were obtained: HADS-D (p = 0.04, β = 0.94 (95% CI:0.13–1.75)), PSQI (p = 0.002, β = 0.71 (95% CI: 0.27–1.15)), and FACIT-F (p = 0.01, β = −0.34 (95% CI: −0.61 to −0.76)). Adjusted R² = 0.37. The rest of the variables: sex (p = 0.4), enthesitis (p = 0.5), Pain VAS (p = 0.1), TJC (p = 0.6), HADS-A (p = 0.1), and PASI (p = 0.1).

Mediation analyses indicated that the effect of disease activity on CS was partially mediated by depression, fatigue, and sleep deficiency. The percentages of the degree of mediation for each of the psychosocial variables were 28% for depression, 40% for fatigue, and 30% for sleep (see the results of the mediation models in the Supplemental Material, Table S2). Figure 4 shows the path diagrams for the three variables.

Figure 4.

Path diagrams showing the mediation of depression (a), sleep quality (b), and fatigue (c) in the association between disease activity (cDAPSA) and CS. Each path shows the model estimator and its 95% confidence intervals.

Discussion

This study provides a comprehensive evaluation of CS as a key modulator of symptoms and therapeutic response in an unselected, consecutively recruited cohort of 228 patients with PsA, with distinctive evaluation of patients with both peripheral and mixed disease manifestations. The overarching finding is that CS is independently associated with symptomatic burden and apparent therapeutic refractoriness, with psychosocial mechanisms playing a potential mediating role that is amenable to intervention, highlighting the importance of integrated biopsychosocial approaches in PsA management. However, the associations detected should be interpreted with great caution without attributing them a causal nature, given the type of sample and the purely exploratory nature of the models developed.

Emotional factors emerged as robust and independent predictors of CS in this cohort. In multivariate logistic regression analysis, depression severity (HADS-D) represented an independent risk factor for CS presence (OR = 1.32, p = 0.01), while poor sleep quality (PSQI) was the strongest predictor of CS severity (OR = 1.21, p = 0.001). Fatigue (FACIT-F: β = −0.34, p = 0.01) demonstrated a direct association with CS presence. Critically, these psychosocial factors—sleep dysfunction, depression, and fatigue—represent a functionally integrated triad rather than independent phenomena.^6,9–11,41

Sleep disruption impairs descending pain inhibitory pathway function through effects on serotonergic and noradrenergic signaling, predisposing to or perpetuating depressive symptoms and enhanced pain sensitivity.⁶ Depression further disrupts sleep architecture through HPA axis dysregulation and altered serotonergic signaling, while both depression and sleep dysfunction enhance fatigue perception through effects on central monoamine systems.^9–11 This sleep-mood-pain-fatigue axis likely operates bidirectionally: chronic inflammatory disease drives emotional distress and sleep disruption through both inflammatory (TNF-α, IL-6 effects on CNS) and non-inflammatory mechanisms, which in turn perpetuate or amplify CS through alterations in descending pain inhibition and neurotransmitter dysregulation.^6,9,10

Depression, sleep quality, and fatigue together explained 47% of CS variance in linear regression models (Adjusted R²: 0.37), exceeding the explanatory power of objective inflammatory markers. This finding emphasizes the paramount importance of emotional assessment and psychological intervention in PsA patients with suspected or established CS. Targeted interventions addressing sleep dysfunction, depression, and fatigue may be associated with reductions in CS severity.

A particularly novel observation in this study is the strong association between CS and therapeutic refractoriness. Patients with CS demonstrated a significantly higher prevalence of inadequate response to two or more different biologic or tsDMARDs (61.5% with CS vs 6.8% without CS, p = 0.003). This finding reveals a significant association between CS and poor therapeutic response and identifies a previously underrecognized mechanism contributing to the substantial proportion of PsA patients failing to achieve remission despite conventional and biologic therapy.^4,5

However, the temporal and causal nature of the CS-refractoriness association requires careful interpretation. This cross-sectional observation likely reflects bidirectional mechanisms rather than simple unidirectional causality.^6,10 First, persistent inflammatory signaling may coexist with CS: inadequate inflammatory control despite multiple therapeutic trials may perpetuate nociceptive input that progressively sensitizes central nervous system pathways, establishing CS that amplifies pain perception independently of remaining inflammatory burden.^6,10,12,13

Second, established CS may be associated with apparent therapeutic refractoriness: once CS becomes established through chronic nociceptive input, the resulting pain amplification may perpetuate apparent treatment failure through subjective disease activity measures and patient-reported outcomes, even if underlying inflammatory activity is successfully suppressed.^6,10 These mechanisms are not mutually exclusive and likely operate in a complementary fashion. Prospective investigation with objective inflammatory markers and longitudinal tracking of CS evolution in relation to therapeutic interventions will be essential to clarify these temporal relationships and establish causal pathways.

The multidimensional characterization of pain phenotypes revealed important associations between CS and neuropathic pain components. Patients with CS demonstrated higher PD scores (median 13 vs 6, p = 0.001) and a higher proportion met criteria for likely neuropathic pain (55.6% vs 8.2%, p = 0.001).^6–8 These findings underscore the complexity of pain in PsA and highlight the inadequacy of single-dimensional pain assessment.

The relationship between CS and disease activity demonstrated that patients with CS exhibited significantly higher clinical disease activity indices despite similar inflammatory markers. Specifically, CS was associated with elevated cDAPSA scores (median 14 vs 9.8, p = 0.001) in patients with peripheral involvement and markedly higher ASDAS-CRP scores (median 4.8 vs 1.8, p = 0.007) in those with axial disease manifestations. These associations remained statistically significant even after accounting for traditional inflammatory markers (CRP: median 0.2 mg/dL (IQR 0.4) vs 0.2 mg/dL (IQR 0.3) in CS+ and CS− groups, respectively; p = 0.1), suggesting that CS was independently associated with higher disease activity scores, potentially reflecting the influence of subjective pain components rather than amplification of objective inflammatory burden.^6,7 The association between cDAPSA and CS is mediated by psychosocial variables that were statistically significant in the regression model. Fatigue has been the predominant mediator in the impact of CS and, to a lesser extent, depression and poor sleep quality.

This dissociation between inflammatory markers and pain severity in patients with CS provides clinical evidence that inflammatory control alone may not fully account for pain in patients with CS. A need exists for multimodal pain management strategies addressing both inflammatory and non-inflammatory mechanisms.^6–8 The functional disability burden attributable to CS was substantial. Patients with CS demonstrated significantly worse HAQ-DI scores (median 0.8 vs 0.2, p = 0.001) in peripheral arthritis and a trend toward higher BASFI scores in axial involvement. The magnitude of functional impairment substantially exceeded what would be expected from objective inflammatory markers alone, indicating a marked discordance between inflammatory control and functional outcomes.^6–8

A key finding is that 35.1% of the PsA cohort exhibited CS (CSI ⩾40), a prevalence lower than the 45.2% reported by Salaffi et al.¹³ in a similarly designed study also excluding fibromyalgia patients. The difference likely reflects demographic and clinical variations between cohorts, particularly the lower proportion of women in our population (39.0% vs 60%), as sex-based differences in pain sensitivity and CS manifestation are well-documented across chronic pain conditions.^6–8,42

This study systematically compared CS associations across distinct disease phenotypes. The very low number of axial-predominant patients (n = 3) prevents any statistical inference or generalizable conclusions regarding CS in this phenotype. Therefore, all observations related to axial-predominant PsA in this study are purely descriptive and should not be extrapolated beyond this cohort. Larger multicenter studies with adequate representation of axial PsA are needed to address this knowledge gap. CS prevalence and associations with disease activity were consistent across peripheral and mixed phenotypes, suggesting relatively homogeneous CS manifestations across PsA presentations.¹² However, these exploratory observations regarding axial disease should be interpreted with substantial caution given the limited prevalence of axial-predominant presentations in this population. This represents an important knowledge gap meriting larger prospective studies specifically focused on axial PsA, given that axial involvement occurs in approximately 20%–30% of PsA patients and may involve distinct inflammatory and neurobiological mechanisms.^1,12

The association between enthesitis and CS (rho = 0.26, p < 0.001) aligns with findings by Sariyildiz et al.,⁴³ who identified enthesitis (MASES) as an independent predictor of CS in axSpA (OR = 2.47, p = 0.03), alongside disease activity and anxiety. These findings suggest that multi-site entheseal inflammation may act as a sustained nociceptive driver promoting CS across the spondyloarthritis spectrum, although the limited number of axial-predominant patients in our cohort (n = 3) warrants confirmation in larger dedicated studies.

Physical activity levels (IPAQ) showed an inverse correlation with CSI scores (rho = −0.15, p = 0.03), with patients with CS reporting lower activity levels. However, this cross-sectional association does not allow causal inference, and the direction of the relationship remains uncertain. Reverse causality is equally plausible: CS-related pain hypersensitivity and fatigue may directly limit exercise tolerance, leading to reduced physical activity rather than low activity driving CS. Additionally, this relationship may be confounded by depression and sleep dysfunction, both of which independently reduce motivation for physical activity. Prospective longitudinal studies incorporating objective accelerometry-based activity monitoring are needed to disentangle the directionality of this association and determine whether exercise interventions can modify CS severity in PsA.

Limitations

The cross-sectional design precludes the determination of causal relationships between CS and other variables assessed, with particular importance for clarifying whether CS represents a consequence of persistent inflammation, a perpetuating factor of inadequate treatment response, or both.^6,10 The exclusion of patients with fibromyalgia, while methodologically appropriate for isolating PsA-related CS mechanisms,¹⁸ limits generalizability to unselected clinical populations where comorbid fibromyalgia is prevalent.⁴¹

Data on axial disease manifestations should be interpreted as exploratory given the small number of axial-predominant patients (n = 3), though mixed peripheral–axial patients (n = 28) provide preliminary insights requiring confirmation in larger dedicated axial PsA cohorts.¹² Reliance on validated patient-reported outcome instruments (CSI, PD, PSQI, HADS) introduces potential for reporting bias, though these instruments demonstrate robust psychometric properties.^20,33,34,36

Furthermore, although no significant association was found between CS and objective inflammatory markers (CRP, SJC), these measures have limited sensitivity for detecting subclinical or residual inflammation. The incorporation of more precise inflammatory assessments—such as musculoskeletal ultrasound with power Doppler or calprotectin—would have strengthened the interpretation of whether elevated composite disease activity scores in patients with CS reflect nociplastic amplification rather than uncontrolled inflammation and should be considered in future studies.

The single-center nature of recruitment may introduce selection bias and limit generalizability, though consecutive recruitment of eligible patients minimizes systematic selection artifact.

However, this study has significant strengths. The sample size of 228 patients represents a substantial cohort for this condition, particularly given the strict exclusion of established fibromyalgia, allowing for a cleaner assessment of CS intrinsic to PsA.¹⁷

This is one of the first studies to comprehensively evaluate physical activity levels using the IPAQ in relation to CS in PsA, providing novel insights into the interplay between lifestyle factors and pain sensitization.⁴⁴ The study systematically compared these variables across distinct disease phenotypes (peripheral and mixed), offering a more granular understanding of CS manifestations in PsA.¹² Additionally, the comprehensive psychosocial evaluation with integrated assessment of the sleep-depression-fatigue triad provides mechanistic insight into central pain processing in PsA.^6,9–11

Future research should address several key directions. First, prospective longitudinal studies are needed to clarify the temporal relationship between CS and therapeutic refractoriness, determining whether CS precedes, follows, or co-evolves with treatment failure. Second, the incorporation of quantitative sensory testing would provide objective neurophysiological confirmation of CS beyond self-reported measures such as the CSI. Third, given the robust association between the sleep-depression-fatigue triad and CS, clinical trials evaluating targeted interventions such as cognitive behavioral therapy are warranted.

Conclusion

This study demonstrates that CS affects approximately one-third of unselected PsA patients and independently associates with therapeutic refractoriness despite similar inflammatory markers. CS represents pain amplification disproportionate to objective inflammatory burden, indicating a distinct non-inflammatory mechanism contributing to persistent symptoms. The relationship between CS and therapeutic refractoriness may reflect coexisting processes: persistent inflammatory signaling co-occurring with CS, and established CS being associated with apparent treatment failure through pain amplification mechanisms independent of peripheral inflammatory control.

Psychosocial factors—particularly sleep dysfunction, depression, and fatigue—emerge as robust independent predictors of CS severity, with sleep and mood disturbances together explaining nearly half the variance in CS presence. These factors appear to operate as an integrated triad affecting central pain processing mechanisms.

These findings support systematic CS screening in PsA patients with inadequate symptomatic response despite adequate inflammatory control, integrated biopsychosocial interventions addressing both inflammatory and non-inflammatory pain mechanisms with particular emphasis on sleep optimization and mood management, and prospective investigation to clarify temporal relationships between CS and therapeutic refractoriness and to evaluate intervention efficacy in modifying CS severity in PsA.

Supplemental Material

sj-docx-1-tab-10.1177_1759720X261453292 – Supplemental material for Central sensitization in psoriatic arthritis: prevalence, clinical correlates, and association with therapeutic refractoriness in an unselected cohort excluding fibromyalgia

Supplemental material, sj-docx-1-tab-10.1177_1759720X261453292 for Central sensitization in psoriatic arthritis: prevalence, clinical correlates, and association with therapeutic refractoriness in an unselected cohort excluding fibromyalgia by Carolina Chacón, Esther Toledano, Rubén Queiró, Javier Martín-Vallejo, Roberto Díaz-Peña, Daniel Martín, Cristina Hidalgo Calleja, Maria Dolores Sánchez, Pilar Sánchez-Conde, Inés Llamas and Carlos Montilla in Therapeutic Advances in Musculoskeletal Disease

Footnotes

Acknowledgements

None.

Declarations

ORCID iDs

Esther Toledano

Rubén Queiró

Cristina Hidalgo Calleja

Carlos Montilla

Supplemental material

Supplemental material for this article is available online.

References

Ritchlin

Colbert

Gladman

DD.

Psoriatic arthritis. N Engl J Med 2017; 376(10): 957–970.

FitzGerald

Ogdie

Chandran

, et al. Psoriatic arthritis. Nat Rev Dis Primers 2021; 7(1): 59.

Ogdie

Coates

Gladman

DD.

Treatment guidelines in psoriatic arthritis. Rheumatology (Oxford) 2020; 59(Suppl. 1): i37–i46.

Gossec

Baraliakos

Kerschbaumer

, et al. EULAR recommendations for the management of psoriatic arthritis with pharmacological therapies: 2019 update. Ann Rheum Dis 2020; 79(6): 700–712.

Coates

Fransen

Helliwell

PS.

Defining minimal disease activity in psoriatic arthritis: a proposed objective target for treatment. Ann Rheum Dis 2010; 69(1): 48–53.

Woolf

CJ.

Central sensitization: implications for the diagnosis and treatment of pain. Pain 2011; 152(3 Suppl.): S2–S15.

Nijs

George

Clauw

, et al. Central sensitisation in chronic pain conditions: latest discoveries and their potential for precision medicine. Lancet Rheumatol 2021; 3(5): e383–e392.

Kosek

Clauw

Nijs

, et al. Chronic nociplastic pain affecting the musculoskeletal system: clinical criteria and grading system. Pain 2021; 162(11): 2629–2634.

Latremoliere

Woolf

CJ.

Central sensitization: a generator of pain hypersensitivity by central neural plasticity. J Pain 2009; 10(9): 895–926.

10.

Ossipov

Dussor

Porreca

Central modulation of pain. J Clin Invest 2010; 120(11): 3779–3787.

11.

Chamessian

Zhang

YQ.

Pain regulation by non-neuronal cells and inflammation. Science 2016; 354(6312): 572–577.

12.

Kilic

Ozgocmen

, et al. Impact of central sensitization on clinical parameters in patients with rheumatoid arthritis. North Clin Istanb 2024; 11(2): 140–146.

13.

Salaffi

Farah

Bianchi

, et al. Central sensitization in psoriatic arthritis: relationship with composite measures of disease activity, functional disability, and health-related quality of life. J Rheumatol 2024; 51(2): 144–149.

14.

Kaya

Tecer

Kılıç

, et al. Impact of central sensitization on clinical and functional aspects of psoriatic arthritis. Medicina (Kaunas) 2024; 60(9): 1449.

15.

Alp

Kurut Aysin

Cinakli

, et al. Prevalence of central sensitization and neuropathic pain in patients with psoriatic arthritis: a cross-sectional study. Arch Rheumatol 2025; 40(1): 87–97.

16.

Taylor

Gladman

Helliwell

, et al CASPAR Study Group. Classification criteria for psoriatic arthritis: development of new criteria from a large international study. Arthritis Rheum 2006; 54(8): 2665–2673.

17.

Wolfe

Clauw

Fitzcharles

, et al. 2016 Revisions to the 2010/2011 fibromyalgia diagnostic criteria. Semin Arthritis Rheum 2016; 46(3): 319–329.

18.

World Medical Association. Declaration of Helsinki: ethical principles for medical research involving human subjects. JAMA 2013; 310(20): 2191–2194.

19.

Mayer

Neblett

Cohen

, et al. The development and psychometric validation of the Central Sensitization Inventory. Pain Pract 2012; 12(4): 276–285.

20.

Neblett

Hartzell

Mayer

, et al. Establishing clinically relevant severity levels for the Central Sensitization Inventory. Pain Pract 2017; 17(2): 166–175.

21.

Marzo-Ortega

Harrison

Fragoulis

, et al. EULAR points to consider and consensus definitions for difficult-to-manage and treatment-refractory psoriatic arthritis. Ann Rheum Dis 2026; 85(1): 61–74.

22.

Coates

Soriano

Corp

, et al. Group for Research and Assessment of Psoriasis and Psoriatic Arthritis (GRAPPA): updated treatment recommendations for psoriatic arthritis 2021. Nat Rev Rheumatol 2022; 18(8): 465–479.

23.

Van der Linden

Valkenburg

Cats

. Evaluation of diagnostic criteria for ankylosing spondylitis. A proposal for modification of the New York criteria. Arthritis Rheum 1984; 27: 361–368.

24.

Heuft-Dorenbosch

Spoorenberg

van Tubergen

, et al. Assessment of enthesitis in ankylosing spondylitis. Ann Rheum Dis 2003; 62(2): 127–132.

25.

Fredriksson

Pettersson

Severe psoriasis—oral therapy with a new retinoid. Dermatologica 1978; 157(4): 238–244.

26.

Craig

Marshall

Sjöström

, et al. International Physical Activity Questionnaire: 12-country reliability and validity. Med Sci Sports Exerc 2003; 35(8): 1381–1395.

27.

Schoels

Aletaha

Alasti

, et al. Disease activity in psoriatic arthritis (PsA): defining remission and treatment success using the DAPSA score. Ann Rheum Dis 2016; 75(5): 811–818.

28.

Lukas

Landewé

Sieper

, et al. Development of an ASAS-endorsed disease activity score (ASDAS) in patients with ankylosing spondylitis. Ann Rheum Dis 2009; 68(1): 18–24.

29.

Fries

Spitz

Kraines

, et al. Measurement of patient outcome in arthritis. Arthritis Rheum 1980; 23(2): 137–145.

30.

Calin

Nakache

Gueguen

, et al. Defining disease activity in ankylosing spondylitis: is a combination of variables (Bath Ankylosing Spondylitis Disease Activity Index) an appropriate instrument? Rheumatology (Oxford) 1999; 38(9): 878–882.

31.

Gossec

de Wit

Kiltz

, et al. A patient-derived and patient-reported outcome measure for assessing psoriatic arthritis: elaboration and preliminary validation of the Psoriatic Arthritis Impact of Disease (PsAID) questionnaire, a 13-country EULAR initiative. Ann Rheum Dis 2014; 73(6): 1012–1019.

32.

Freynhagen

Baron

Gockel

, et al. PainDETECT: a new screening questionnaire to identify neuropathic components in patients with back pain. Curr Med Res Opin 2006; 22(10): 1911–1920.

33.

Zigmond

Snaith

RP.

The Hospital Anxiety and Depression Scale. Acta Psychiatr Scand 1983; 67(6): 361–370.

34.

Chandran

Bhella

Schentag

, et al. Functional Assessment of Chronic Illness Therapy-Fatigue Scale is valid in patients with psoriatic arthritis. Ann Rheum Dis 2007; 66(7): 936–939.

35.

Buysse

Reynolds

3rd Monk

, et al. The Pittsburgh Sleep Quality Index: a new instrument for psychiatric practice and research. Psychiatry Res 1989; 28(2): 193–213.

36.

WHO. Obesity: preventing and managing the global epidemic. Report of a WHO consultation. World Health Organ Tech Rep Ser 2000; 894: i–xii, 1–253.

37.

Nagelkerke

NJD

. A note on a general definition of the coefficient of determination. Biometrika 1991; 78(3): 691–692.

38.

Peduzzi

Concato

Kemper

, et al. A simulation study of the number of events per variable in logistic regression analysis. J Clin Epidemiol 1996; 49(12): 1373–1379.

39.

Field

Discovering statistics using IBM SPSS statistics. 5th ed. London: Sage Publications, 2017.

40.

Von Elm

Altman

Egger

, et al.; STROBE Initiative. The Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) statement: guidelines for reporting observational studies. Lancet 2007; 370(9596): 1453–1457.

41.

Haack

Simpson

Sethna

, et al. Sleep deficiency and chronic pain: potential underlying mechanisms and clinical implications. Neuropsychopharmacology 2020; 45(1): 205–216.

42.

Häuser

Ablin

Fitzcharles

, et al. Fibromyalgia. Nat Rev Dis Primers 2015; 1: 15022.

43.

Sariyildiz

Coskun Benlidayi

Turk

, et al. Biopsychosocial factors should be considered when evaluating central sensitization in axial spondyloarthritis. Rheumatol Int 2023; 43(5): 923–932.

44.

Rice

Nijs

Kosek

, et al. Exercise-induced hypoalgesia in pain-free and chronic pain populations: state of the art and future directions. J Pain 2019; 20(11): 1249–1266.

Supplementary Material

Please find the following supplemental material available below.

For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.

For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.

0.00 MB

0.04 MB