Cross-Cultural Adaptation and Validation for Central Aspects of Pain in the Knee Questionnaire: Based on Chinese Patients Undergoing Total Knee Arthroplasty for Knee Osteoarthritis

Background

Knee osteoarthritis (KOA) is a prevalent age-related degenerative joint disorder, primarily characterized by knee pain and functional impairment.¹ Its pathogenesis involves a complex interplay of peripheral pathological changes, which include articular cartilage degradation, synovial inflammation, subchondral bone sclerosis, and osteophyte formation.² These changes initially trigger pain by activating peripheral nociceptors.³ Epidemiological studies have reported that between 2008 and 2017, the average crude prevalence of KOA among Chinese adults aged 55 years and older is approximately 13.2%.⁴ Although total knee arthroplasty (TKA) is the primary treatment for end-stage KOA and has proven efficacy in restoring joint function, approximately 10–20% of patients continue to experience persistent pain and overall dissatisfaction postoperatively, largely due to unexplained postoperative pain.^5–7 This poor outcome is largely attributed to limitations of traditional treatments: most interventions, such as nonsteroidal anti-inflammatory drugs, physical therapy, or TKA itself, focus on resolving peripheral pathology while neglecting central nervous system (CNS) remodeling that sustains pain.⁸ In older patients, chronic knee pain not only results in diminished extensor strength, impaired walking ability, and reduced quality of life, but also markedly undermines satisfaction with surgical outcomes.^9–11 With the acceleration of population aging, the effective management of knee pain in older patients has emerged as a pressing clinical challenge.

In recent years, the concept of central sensitization (CS) has gained increasing recognition, and clearly distinguishing CS from peripheral sensitization is critical for guiding targeted treatment.¹² Peripheral sensitization refers to increased responsiveness of peripheral nociceptors to noxious or sub-noxious stimuli, triggered by local inflammation, tissue damage, or chemical mediators such as prostaglandins in KOA joints.¹³ It is a transient, localized process that resolves when peripheral pathology is controlled. In contrast, CS is defined as heightened responsiveness of nociceptive neurons in the CNS (such as spinal dorsal horn neurons and supraspinal brain regions) to normal or sub-threshold stimuli.^12,14,15 It arises from plastic changes in the CNS induced by long-term peripheral pain input, including activation of N-methyl-D-aspartate (NMDA) receptors, hyperactivation of glial cells, and reduced descending pain inhibition.^16–18 Importantly, CS has been identified as a major risk factor for persistent pain and dissatisfaction after TKA, as it perpetuates pain even when peripheral pathology is addressed by surgery.^19,20 Interventions targeting CS, such as cognitive behavioral therapy and certain neuromodulators, can enhance treatment responses. These interventions complement traditional peripheral-focused strategies and provide more effective relief of knee pain.^20,21

Current methods for assessing CS include quantitative sensory testing (QST), electroencephalography, functional magnetic resonance imaging (fMRI), and self-report questionnaires.^12,22,23 QST quantifies sensory thresholds, such as heat and pressure pain thresholds, to reflect CNS excitability, but it is susceptible to subjective bias and operator variability, whereas fMRI visualizes CS-related brain activation, such as in the insula and anterior cingulate cortex, but is costly and not easily accessible in clinical settings.^24,25 Among self-reported tools, the Central Sensitization Inventory (CSI) is the most widely used; it can predict pain outcomes associated with central mechanisms and evaluate treatment responses.²⁶ However, the CSI lacks disease specificity: it does not account for knee-specific pain manifestations, such as weight-bearing-related central pain and post-TKA knee hypersensitivity.²⁷ Thus, it may not adequately capture knee-specific central mechanisms in older patients with knee pain.

To address this gap, Akin-Akinyosone et al developed and validated the Central Aspects of Pain in the Knee (CAP-Knee) questionnaire.²⁸ This questionnaire consists of 8 items covering dimensions closely linked to central mechanisms, including neuropathic-like pain, fatigue, cognitive impact, catastrophizing, anxiety, sleep disturbance, depression, and pain distribution.²⁸ Evidence shows that higher total CAP-Knee scores are positively associated with greater pain intensity and poorer functional outcomes in KOA patients.^28,29 As a concise, knee-specific tool, the CAP-Knee holds promise for screening and stratifying patients with knee pain in clinical practice. This could further facilitate the development of individualized treatment strategies, such as prioritizing CS-targeted interventions for patients with high CAP-Knee scores.

However, a simplified Chinese version of the CAP-Knee (CAP-Knee-CV) has not yet been developed. Therefore, the present study aims to translate and culturally adapt the CAP-Knee, and to evaluate its reliability, validity, and psychometric properties in Chinese KOA patients scheduled for TKA. The ultimate goal of this study is to lay a foundation for the clinical application of the CAP-Knee in China, particularly in its potential value for preoperative risk assessment, which helps identify patients at high risk of CS-related post-TKA pain, as well as for patient stratification in TKA.

Methods

Study Design and Ethical Approval

This prospective observational study was approved by the Medical Ethics Committee of Honghui Hospital, Xi’an Jiaotong University (Approval No. 202407008). All procedures conformed to the Declaration of Helsinki, and written informed consent was obtained from all participants prior to enrollment.³⁰ Participants were informed of their right to withdraw from the study at any time without penalty, and all data were de-identified to protect privacy.

Translation and Cross-Cultural Adaptation

The translation and cross-cultural adaptation of the CAP-Knee into simplified Chinese followed internationally recognized guidelines, including three stages: forward translation, back translation, and cognitive debriefing.^31,32 First, two native Chinese translators independently performed forward translations: one was an orthopedic surgeon proficient in English, and the other was a professional translator with medical translation experience. Their versions were synthesized under the supervision of a cross-cultural adaptation expert to form a preliminary consensus version. Next, two native English speakers without medical backgrounds independently back-translated the questionnaire to ensure semantic equivalence and identify potential ambiguities. Finally, a pilot test was conducted among 30 patients with end-stage KOA to assess comprehensibility, content integrity, and completion time. Based on participant feedback, revisions were made to generate the final CAP-Knee-CV (Figure S1).

Participants

From July 2024 to March 2025, patients with end-stage KOA scheduled for primary unilateral TKA were consecutively recruited from the Department of Knee Surgery, Honghui Hospital, Xi’an Jiaotong University. Inclusion criteria were: (a) age ≥18 years; (b) diagnosis of end-stage KOA with knee pain; (c) sufficient cognitive ability to complete questionnaires; (d) fluency in Mandarin; and (e) voluntary participation with signed informed consent. Exclusion criteria were: (a) inflammatory arthritis (eg, rheumatoid arthritis); (b) other conditions causing pain or impairing mobility (eg, hemiplegia, fracture, ligament injury, lower-limb vascular disease); (c) prior TKA; (d) comorbidities severely affecting daily life (eg, coronary heart disease, asthma, severe psychiatric disorders); and (e) inability to complete two CAP-Knee-CV assessments.

Demographic and clinical characteristics, including age, sex, BMI, Kellgren–Lawrence grade, education level, and living status, were obtained from the hospital medical records. All questionnaires were administered preoperatively under the supervision of trained interviewers. Missing items were completed with participants’ consent. To evaluate test–retest reliability, 50 patients were randomly selected to complete the CAP-Knee-CV again 7 days after the initial assessment, during which no additional intervention was performed.³³

Questionnaires

CAP-Knee-CV

The CAP-Knee consists of eight items covering central mechanisms related to knee pain: neuropathic-like pain, fatigue, cognitive-impact, catastrophizing, anxiety, sleep disturbance, depression, and pain distribution.^28,34–39 Items were scored as follows: “never” = 0, “sometimes” = 1, “often” = 2, and “always” = 2. To preserve measurement properties, the original four-level scoring was retained to avoid midpoint bias. Item 7 was reverse-scored (“never” = 2, “sometimes” = 1, “often” = 0, “always” = 0). For Item 8, pain distribution was scored as 0 if pain was localized to one knee and 2 if pain extended below the waist. All items were equally weighted, with a maximum score of 2 per item; total scores ranged from 0–16, with higher scores indicating greater central contributions to knee pain.

CSI

The CSI comprises Parts A and B; only Part A was used in this study.²⁶ It includes 25 items scored on a 0 (“never”) to 4 (“always”) Likert scale, yielding a total score of 0–100. Higher scores indicate more pronounced features of central sensitization. The reliability and validity of the CSI have been previously validated in Chinese KOA populations.⁴⁰

Pain Intensity While Walking

Pain intensity during walking was assessed using an 11-point Numerical Rating Scale (NRS, 0–10), where 0 represents no pain and 10 represents the worst imaginable pain.⁴¹ This tool has been widely validated as a simple and reliable pain measure.⁴²

The Western Ontario and McMaster Universities Osteoarthritis Index (WOMAC)

The WOMAC comprises 24 items divided into three subscales: pain (5 items), stiffness (2 items), and physical function (17 items).⁴³ Each item is scored from 0 to 4, with higher scores indicating more severe pain, stiffness, and functional limitations. The simplified Chinese version has been validated in TKA patients.⁴⁴

The 5-Level EQ-5D (EQ-5D-5L)

The EQ-5D-5L was used to assess health-related quality of life (HRQOL) across five domains: mobility, self-care, usual activities, pain/discomfort, and anxiety/depression.⁴⁵ Each dimension is rated on a five-level scale (no, slight, moderate, severe, or extremely severe difficulties). The Chinese version of the EQ-5D-5L was used, with established validity in KOA patients.⁴⁶

Sample Size Determination

To calculate the required sample size, we adhered to established guidelines for factor analysis, which recommend having between 4 and 10 participants for each item, as well as a total sample size exceeding 100 participants. Based on these criteria, the sample size for this study was set to be at least 100 participants.^47,48

Statistical Analysis

Descriptive statistics were used to summarize demographic and clinical characteristics. Continuous variables were presented as mean ± standard deviation (SD), and categorical variables as percentages. The Kolmogorov–Smirnov test was used to examine the distribution of CAP-Knee-CV, CSI, NRS, WOMAC, and EQ-5D-5L scores. All analyses were performed using SPSS version 26.0 (IBM, Chicago, IL), with a two-tailed significance threshold of p < 0.05.

Psychometric Evaluation

Structural Validity

Exploratory factor analysis (EFA) was conducted to examine the underlying factor structure of the CAP-Knee-CV. Suitability for factor analysis was assessed using the Kaiser–Meyer–Olkin (KMO) test and Bartlett’s test of sphericity. Factors were extracted using maximum likelihood estimation with promax rotation. Factors with eigenvalues >1.0 were retained, and loadings ≥0.40 were considered significant.⁴⁹ A two-factor model was hypothesized for the CAP-Knee-CV, based on observed structural variability in cross-culturally adapted pain questionnaires. For example, the Japanese adaptation of the CAP-Knee exhibits a three-factor structure, which differs from the original English version’s unidimensional framework.^28,29

Criterion Validity

Spearman correlation coefficients were used to assess relationships between CAP-Knee-CV and CSI, NRS, WOMAC, and EQ-5D-5L. Correlation strength was classified as follows: >0.81 = very strong; 0.61–0.80 = strong; 0.41–0.60 = moderate; 0.21–0.40 = weak; <0.20 = negligible.⁵⁰ Based on content analysis, we hypothesized that CAP-Knee-CV would show moderate correlations with these measures.

Internal Consistency and Test–Retest Reliability

Internal consistency was assessed using Cronbach’s α (unacceptable <0.70, acceptable 0.70–0.79, good 0.80–0.89, excellent ≥0.90).⁵¹ Test–retest reliability was evaluated using the ICC: moderate (0.50–0.75), good (0.75–0.90), and excellent (>0.90).⁵² Following psychometric guidelines, a 2–14 days interval is recommended to minimize clinical change and recall bias; a 7-days interval was chosen in this study.³³

Feasibility

Feasibility was assessed by recording participant difficulties, missing responses, and average completion time of the CAP-Knee-CV.

Results

Participants

A total of 120 patients completed the baseline survey, including 76 women (63.3%). The mean age was 65.2 ± 5.5 years, and the mean BMI was 25.7 ± 4.0 kg/m². Most patients were classified as Kellgren–Lawrence grade IV (63.3%). Detailed demographic and clinical characteristics, including education level and living status, are presented in Table 1. The mean CAP-Knee-CV total score was 8.4 ± 3.9.

Table 1 Demographic Data of the Sample Group Population (n = 120)

To further assess the statistical power of the sample size for factor structure validation, we conducted a post-hoc power analysis using G*Power 3.1 software. The analysis was based on two hypothesized models: a two-factor model and the original unidimensional model. The parameters for the analysis were set as follows: a medium effect size (f = 0.3), a significance level of α = 0.05, and the actual sample size of 120 participants.⁵³ The number of predictors was set at 2 for the two-factor model and 1 for the unidimensional model. The results indicated that the statistical power for both the two-factor model and the unidimensional model was 0.99, suggesting that the sample size of 120 participants is sufficient to reliably detect the true effects of both factor structures. This power analysis supports the adequacy of the sample size for factor structure validation and subsequent reliability and validity assessments.

Structural Validity

The KMO value was 0.827, and Bartlett’s test of sphericity was significant (p < 0.001), indicating suitability for factor analysis. EFA with scree plot inspection (Figure 1) supported a two-factor model. The eigenvalues of Factor 1 and Factor 2 were 3.9 and 1.4, accounting for 33.3% and 32.8% of the variance, respectively, with a cumulative variance contribution of 66.1%. Factor loadings are shown in Table 2. Factor 1 comprised items 1, 5, 6, 7, and 8, representing patients’ emotional and physiological responses to knee pain, and was labeled the “psychological–physiological response factor”. Factor 2 included items 2, 3, and 4, reflecting cognitive impairment and fatigue, and was labeled the “cognitive–fatigue factor”.

Table 2 Factors in Factor Analysis and Varimax Rotated Factor Matrix of the CAP-Knee-CV

Figure 1 Scree plot of exploratory factor analysis.

Criterion Validity

Correlations between the CAP-Knee-CV and other instruments are summarized in Table 3. The CAP-Knee-CV total score showed strong correlations with the CSI total score (r = 0.610, p < 0.001) and with NRS (r = 0.697, p < 0.001). Moderate correlations were observed with the WOMAC pain subscale (r = 0.526, p < 0.001) and total WOMAC score (r = 0.530, p < 0.001), as well as with the EQ-5D-5L total score (r = 0.510, p < 0.001).

Table 3 Spearman Correlation Coefficients of CAP-Knee-CV, WOAMC, CSI, NRS and EQ-5D-5L

Internal Consistency and Test–Retest Reliability

Reliability results are presented in Table 4. The CAP-Knee-CV demonstrated good internal consistency (Cronbach’s α = 0.843) and excellent test–retest reliability (ICC = 0.957, 95% CI: 0.924–0.976). All ICC values exceeded 0.90, confirming stable responses over 7 days.

Table 4 Internal Consistency and Test–Retest Reliability of the CAP-Knee-CV

Feasibility

All participants successfully completed the CAP-Knee-CV, with no missing data. The mean completion time was 3.6 ± 0.9 minutes, and no substantial difficulties were reported.

Discussion

This study aimed to develop and validate a simplified Chinese version of the CAP-Knee and to evaluate its psychometric properties in patients with KOA undergoing TKA. The findings demonstrated that the CAP-Knee-CV exhibited a clear two-factor structure, good internal consistency, excellent test–retest reliability, and significant correlations with pain-related outcomes (CSI, NRS, WOMAC, and EQ-5D-5L), indicating its clinical applicability in the Chinese patient population.

A key finding of our study was the two-factor structure of the CAP-Knee-CV, which diverges from the unidimensional model of the original English version.²⁸ In our Chinese cohort, items clustered into two distinct domains: one reflecting emotional and physiological responses to pain (neuropathic-like pain, anxiety, depression, sleep disturbance, pain distribution) and the other indicating cognitive impairment and fatigue (fatigue, cognitive impact, catastrophizing). This structural difference likely reflects cultural variations in how pain-related symptoms are perceived and categorized.⁵⁴ Consistent with cross-cultural pain research, Chinese populations often adopt a holistic “mind-body” framework that integrates emotional and physical experiences, whereas Western samples may conceptualize these domains as separate.⁵⁵ This interpretation is further supported by comparisons with the Japanese version of the CAP-Knee, which exhibited a three-factor structure, with “sleep disturbance” and “fatigue” as independent dimensions, and some items failing to load adequately.²⁹ Such cross-cultural variability underscores that direct translation of patient-reported outcome measures is insufficient. Rigorous cultural adaptation is necessary to ensure measurement equivalence and validity in diverse populations. For instance, Chinese participants in our study frequently described sleep disturbance as a consequence of pain or anxiety, justifying its inclusion in the “psychological–physiological response factor”.⁵⁶ In contrast, Japanese patients in the prior validation study reported sleep disturbance as a standalone symptom.²⁹

The CAP-Knee-CV’s reliability metrics are comparable to those of other international versions. Its Cronbach’s α (0.843) exceeds that of the original English version (α = 0.74) and is consistent with the Japanese adaptation (α = 0.86), indicating strong internal consistency.²⁸ The excellent test–retest reliability (ICC = 0.957) aligns with the original study (ICC = 0.91), confirming the tool’s ability to stably capture CS features over time.²⁸ These results indicate that the CAP-Knee-CV is a reliable instrument that can stably capture CS syndrome over time. Regarding criterion validity, the CAP-Knee-CV demonstrated strong positive correlations with the CSI (r = 0.610, p < 0.001), a gold-standard measure of general CS, confirming its convergence with core CS constructs. It also showed strong correlations with walking pain (NRS; r = 0.697, p < 0.001) and moderate correlations with KOA-specific function (WOMAC; total score: r = 0.530, p < 0.001) and HRQOL (EQ-5D-5L; r = 0.510, p < 0.001). These associations are consistent with the well-established link between CS and worse pain intensity, functional limitations, and HRQOL in KOA, further validating the CAP-Knee-CV as a clinically relevant measure of central pain mechanisms.^19,20

Notably, the CAP-Knee-CV addresses a critical gap in personalized perioperative care by identifying patients at risk for iatrogenic modulation of central pain pathways, a concern highlighted in recent research on KOA.²⁸ Iatrogenic inhibition, or the unintended alteration of central pain processing due to medical interventions, significantly contributes to persistent post-TKA pain.^57,58 For example, overreliance on opioid analgesics during the perioperative period can exacerbate CS through NMDA receptor activation, while inadequate preoperative pain management prolongs nociceptive input, further amplifying CS.^59–61 Preoperative CAP-Knee-CV assessment can stratify patients based on CS severity. Patients with high scores (indicating prominent central mechanisms) may benefit from multimodal perioperative analgesia, such as femoral nerve blocks combined with cognitive behavioral therapy, to mitigate iatrogenic exacerbation of CS, while patients with low scores may respond adequately to standard peripheral-focused care.^58,60 This approach has the potential to reduce the 10–20% rate of persistent post-TKA pain linked to unaddressed central mechanisms, highlighting the CAP-Knee-CV’s relevance for clinical practice.⁵⁷

Feasibility is another strength of the CAP-Knee-CV. With an average completion time of 3.6 ± 0.9 minutes and a 100% completion rate, it outperforms longer CS assessment tools, such as the CSI, which requires approximately 6.4 minutes to complete.⁴⁰ This is a critical advantage in time-constrained clinical settings, such as preoperative TKA clinics.

This study has several limitations. First, we did not incorporate objective measures of CS, such as QST for pressure pain thresholds or fMRI to assess CS-related brain activation, such as in the insula or anterior cingulate cortex. Future studies integrating QST or fMRI would enable examination of concordance between the CAP-Knee-CV and objective CS biomarkers, further strengthening its validity. Second, recruitment from a single center in northwestern China may limit generalizability to patients from other regions with distinct cultural or socioeconomic profiles. Multicenter studies with geographically diverse samples are needed to confirm the CAP-Knee-CV’s performance across broader Chinese KOA populations. Third, our sample focused on moderate-to-severe KOA patients scheduled for TKA, and the tool’s validity in earlier-stage KOA or non-surgical patients remains untested. Expanding validation to these groups would confirm the CAP-Knee-CV’s utility across the full KOA disease spectrum. Finally, we did not evaluate the CAP-Knee-CV’s predictive validity for iatrogenic outcomes, such as opioid use or persistent post-TKA pain. Longitudinal studies tracking these endpoints would further establish the tool’s clinical value for risk stratification.

Conclusion

The CAP-Knee-CV demonstrates strong reliability and validity as a tool for assessing CS in Chinese patients with KOA awaiting TKA. It reveals a distinctive two-factor structure, consisting of the “psychological–physiological response factor” and “cognitive–fatigue factor”, which contrasts with the unidimensional structure of the original English version and the three-factor structure in the Japanese version. This underscores the importance of cultural adaptation in cross-cultural pain assessment. With robust psychometric properties and a short completion time of approximately 3.6 minutes, the CAP-Knee-CV is well-suited for use in preoperative TKA settings, such as screening for high-risk CS patients to guide personalized analgesia strategies. Future research should integrate objective measures like QST and conduct multicenter validations to broaden its applicability, while also exploring its potential in cross-cultural CS research.