Abstract
Background:
Adaptation studies have evaluated ankle instability (AI) across various cultures and sports.
Purpose/Hypothesis:
The Cumberland Ankle Instability Tool (CAIT) is a valid patient-reported outcome measure used to assess the presence and severity of AI. The current study aimed to adapt the CAIT to the Turkish language, test its psychometric properties, and determine cutoff scores in a Turkish population with AI compared with observations in their healthy counterparts.
Study Design:
Cohort study (Diagnosis); Level of evidence, 3.
Methods:
To evaluate the Turkish version of the CAIT (CAIT-TR), data obtained from 236 recreationally active participants with and without AI were examined. Test-retest reliability, internal consistency, cutoff score, and ceiling and floor effects were assessed.
Results:
The test-retest reliability of the CAIT-TR was excellent, with an intraclass correlation coefficient of 0.968 and 0.978 for the right and left ankles, respectively. Cronbach α values for the CAIT-TR were calculated as 0.855 and 0.852 for the right and left ankles, respectively. The discriminative ability and cutoff score of the CAIT-TR were determined by defining the receiver operating characteristic curve and calculating the Youden index for the left and right ankles (0.579 and 0.551, respectively), with a cutoff score of 23.5 and threshold of ≤23.
Conclusion:
Our study demonstrates that the CAIT-TR is a valid and reliable tool for assessing the presence and severity of AI in recreationally active Turkish-speaking individuals. These data should help in providing a practical assessment tool for care in this population.
Registration:
NCT05977660 (ClinicalTrials.gov identifier).
Keywords
With the increasing number of recreational sports participants, the frequency of musculoskeletal injuries caused by physical activity has also risen. 8 The foot and ankle are the most commonly injured body parts after the knee in both the general population and athletes. Ankle sprain, the most common ankle injury, accounts for >80% of all ankle injuries. 5 Ankle sprains have a high recurrence rate, and approximately 30% of individuals with a history of ankle sprains develop chronic ankle instability (CAI). 26 CAI occurs within 1 year after the initial injury in 40.5% of individuals who have sustained their first ankle sprain, 5 and the incidence of CAI can reach up to 70% shortly thereafter. 7
Patient-reported outcome measures (PROMs) convert qualitative attributes—such as anxiety, pain, or social functioning—that cannot be directly observed or are impractical to measure directly (eg, performance of daily activities)—into quantitative scores that allow for mathematical analysis, including patient perceptions of specific aspects of one’s own health.3,17 Translating and adapting PROMs from one language to another is necessary when a local language–adapted version of a specific PROM does not exist for clinical or academic research.
Although at ≥17 PROMs are commonly used to evaluate patients with ankle instability (AI) in the literature, only 3 were developed with direct input from patients: the Cumberland Ankle Instability Tool (CAIT), Lower Extremity Functional Scale, and Foot and Ankle Ability Measure. Of these, the CAIT demonstrates the best internal consistency, content validity, and psychometric properties. It can be used both to diagnose instability and as an assessment tool. 10 The International Ankle Consortium has recommended its use for assessing the presence and severity of CAI. 20 However, the fact that CAIT’s original language is English limits its applicability for AI assessments in many non–English speaking countries and communities worldwide. Thus, it is essential to promote the use of this measurement tool across different languages and cultures. CAIT has been translated from English into numerous languages, including Spanish,2,24 Brazilian Portuguese, 4 Korean, 15 Japanese, 18 Dutch, 29 Persian, 9 French, 6 Cantonese, 12 Taiwanese, 19 Arabic, 16 Thai,13,32 Greek, 28 Urdu, 14 and Polish. 23 Additionally, a digital version of the CAIT is available. 25
Given the rising demand for PROM scales that can be widely used in both clinical and scientific research, this study aimed to adapt CAIT into Turkish (CAIT-TR) and Turkish culture, as well as to examine its psychometric properties in recreationally active participants with and without AI. We hypothesized that recreationally active participants with CAI would have significantly higher scores on the CAIT-TR than their non-CAI counterparts, with the calculated cutoff score indicating the presence or absence of CAI.
Methods
Research Design
In the process of adapting the CAIT to the Turkish language and culture, forward and backward translation methods were used, taking into account the International Quality of Life Assessment standard. 1 In line with this method, the original version of the measurement tool was first translated into Turkish by 2 independent translators who were familiar with the Turkish language and culture. It was confirmed that the translators had no previous knowledge of the CAIT or current research. The resulting Turkish draft translation was evaluated by the translators and field experts (1 sports physician and 1 sports scientist (E.A., C.K.)). During the evaluation process, the most appropriate wording among the alternative expressions was selected by consensus. Two translators fluent in English then translated the final Turkish version of the measurement tool back into English. The meanings of the items were evaluated by comparing the Turkish translations with the original CAIT. To evaluate the validity of the translation, the resulting draft version was pretested by applying it to 20 participants. In line with participant feedback, a final version of the measurement tool was prepared to ensure clarity. Data were collected from the participants using the Turkish version of the questionnaire. Ethical approval for this study was obtained from Eskişehir Technical University. The study protocol was registered at clinicaltrials.gov. Before data collection, participants were informed in writing.
Participants
Participants consisted of 122 (51.7%) women and 114 (48.3%) men aged ≥18 years (mean ± SD, 25.81 ± 7.85). Recreationally active individuals were included in the study. Participants were considered recreationally active if they engaged in ≥20 minutes of physical activity 3 times per week. Individuals who had undergone lower-extremity surgery and those with lower-extremity injuries or musculoskeletal conditions unrelated to ankle sprains were excluded from the study. Additionally, individuals with balance disorders due to inner ear conditions were excluded. Participants were divided into 2 groups. The CAI group (n = 126) consisted of participants who had experienced ≥1 ankle sprain ≥2 months prior and had sought treatment previously or were undergoing treatment at the time of the study. These participants were required to have a history of “giving way,” recurrent sprains, and/or a “feeling of instability” in the injured ankle joint. They were also required to have a history of CAI with no pain, swelling, or physical activity limitations for ≥2 months before the study. Participants were evaluated individually by a sports medicine physician (E.A.) for the presence or absence of CAI in accordance with the specific diagnostic criteria of the International Ankle Consortium. 20 The control group (n = 110) included individuals who had no history of ankle sprains and were not seeking treatment for related issues. Participant characteristics are presented in Table 1.
Participant Characteristics a
Data are presented as n (%) or mean ± SD. CAIT, Cumberland Ankle Instability Tool.
Data Collection Tool
The questionnaire used for data collection consisted of 3 parts. The first part contained information on research ethics and voluntary participation. In the second part, the original questionnaire developed by Hiller et al 11 was adapted to the Turkish language and culture. This section consisted of 9 items. A score between 0 to 3, 0 to 4, or 0 to 5 could be obtained for each item using predefined options. The CAIT score was calculated as the sum of all item scores. Accordingly, the lowest possible total score was zero, and the highest was 30. A high CAIT score indicated greater ankle stability. The third part of the questionnaire included participant demographic data, including age, sex, height, weight, number of sprains, and time since last sprain.
Data Analysis
A test-retest analysis was used to evaluate the ability of the measure to produce consistent results in the same group at 2 different time points. For test-retest reliability, data were collected from a group of 36 individuals over a 21-day interval. To evaluate the relation between the 2 measurements, the intraclass correlation coefficients (ICCs) were calculated. Reliability was considered poor when the ICC was <0.40, moderate between 0.40 and 0.75, substantial between 0.76 and 0.90, and excellent >0.90. 2 The standard error of the mean was used to determine the level of difference between the 2 mean values, and the minimal detectable change (MDC) was calculated to detect the minimum score change beyond measurement error. The formulas “SD × √(1–ICC)” and “SEM × √2” were used to calculate SEM and MDC, respectively. 22 Cronbach α coefficient was used to evaluate the internal consistency of the measurement tool. To determine ceiling and floor effects, data from both groups were combined and analyzed. These effects were identified by calculating the percentage of participants who obtained the highest and lowest possible scores. To assess the discriminative power of the CAIT-TR for identifying individuals with AI, receiver operating characteristic (ROC) curve analysis was conducted. The Youden index was used to determine cutoff scores. The formula “sensitivity + (specificity–1)” was used to compute Youden index values. Potential cutoff scores were evaluated using positive likelihood ratios (LR+) and negative likelihood ratios (LR–). LR+ was calculated as “sensitivity / (1–specificity),” whereas LR– was calculated as “(1–sensitivity) / specificity.” 18
Results
Test-Retest Reliability
The pretest reliability of the CAIT-TR in 36 participants demonstrated of ICC of 0.968 (95% CI, 0.938-0.984; P < .001) and 0.978 (95% CI, 0.957-0.989; P < .001) for the right and left ankles, respectively, indicating excellent reliability. The mean first-test score for the right ankle was 18.47 ± 6.30, and the mean second-test score was 17.94 ± 5.58. The mean difference in CAIT scores for the right ankle was 0.53. The mean first-test score for the left ankle was 17.97 ± 5.68, and the mean second-test score was 17.03 ± 5.21. The mean difference in CAIT scores for the left ankle was 0.94. The CAIT SEM value was 1.06 and 0.81 for the right and left ankles, respectively, whereas the MDC value was 1.51 and 1.14 for the right and left ankles, respectively. All values are listed in Table 2. These findings confirm that the measurement tool has strong test-retest reliability.
Test-Retest Values of the Turkish Version of CAIT a
ICC, intraclass correlation coefficient; MDC, minimal detectable change.
Internal Consistency
In the reliability analysis (n = 236), Cronbach α value for the Turkish version of the CAIT was 0.855 and 0.852 for the right and left ankles, respectively. These values indicate that the measurement tool had good internal consistency for both ankles.
Ceiling and Floor Effects
When the CAIT scores of the 236 participants were examined, 25 (10.6%) participants achieved the maximal score for the left ankle, whereas 19 (8.1%) did so for the right ankle. No participants obtained the minimal score. Based on these results, ceiling and floor effects were not observed.21,27
Discriminative Ability
To evaluate the discriminative ability of the Turkish version of the CAIT, data from 236 participants were analyzed, an ROC curve was constructed (Figure 1), and Youden index values were calculated.

Receiver operating characteristic (ROC) curve of left and right Cumberland Ankle Instability Tool scores.
As shown in Table 3, the maximal Youden index values were 0.579 and 0.551 for the left and right ankles, respectively. The cutoff score for distinguishing individuals with and without AI was 23.5. Thus, individuals with CAIT scores of ≤23 were classified into the AI group. As a result of the analysis, the following values were obtained. Right ankle: sensitivity, 0.836; specificity, 0.714; area under the ROC curve, 0.810; LR+, 2.93; and LR−, 0.229. Left ankle: sensitivity, 0.809; specificity, 0.770; area under the ROC curve, 0.834; LR+, 3.52; and LR−, 0.248.
Sensitivity, Specificity, Youden Index, LR+ and LR– Values at Potential CAIT Cutoff Scores a
CAIT, Cumberland Ankle Instability Tool; Sn, sensitivity; Sp, specificity; LR+, positive likelihood; LR–, negative likelihood; NA, not applicable.
Discussion
The major findings of this study demonstrated that the CAIT-TR exhibited good internal consistency and excellent test-retest reliability for both ankles, with no evidence of ceiling or floor effects. For the Turkish-speaking individuals, the CAIT-TR demonstrated a cutoff score of 23.5, with scores ≤23 indicating AI. These findings support the validity and reliability of the CAIT-TR as a tool for assessing the presence and severity of AI in recreationally active Turkish-speaking individuals.
The rise in recreational sports participation has led to a notable increase in musculoskeletal injuries, particularly ankle sprains.5,8 Ankle sprains are highly recurrent and often lead to mechanical and/or functional instability, resulting in CAI. 6 Mechanical AI results from physical changes such as pathological laxity and degenerative joint disease and is clinically assessed through physical examination methods such as manual testing and stress radiography.6,18 In contrast, functional AI arises from neuromuscular alterations and is associated with deficits in neuromuscular control, proprioception, and muscle strength. 18 Unlike mechanical instability, functional instability is more difficult to assess and is typically evaluated using PROMs.6,18 CAIT is a validated PROM used to identify and quantify the severity of functional AI. 11 This study was conducted to investigate the reliability and validity of the Turkish version of the CAIT questionnaire in identifying and evaluating patients with CAI. The English CAIT was translated into Turkish and then back-translated into English to ensure accuracy. The meaning of all items was preserved, and no modifications were applied. Based on the results of this study, the CAIT is valid for use in the Turkish population for the evaluation of AI.
The internal consistency of PROMs is assessed by calculating the Cronbach alpha coefficient. A high internal consistency value for the CAIT represents a high degree of intercorrelation among all items. Cronbach α for the original version of the CAIT was reported as 0.83. 11 In the Turkish version, this value was approximately equal to that of the original version. 11 Previous studies validating the CAIT have reported internal consistency coefficients ranging from 0.77 to 0.92.2,4,9,11,24,29 When the Cronbach α values for the right and left ankles obtained in this study are compared with those reported in other cross-cultural adaptation studies, the Brazilian-Portuguese version (0.86-0.88) 4 and the Korean version (0.92-0.90) 15 demonstrated slightly higher values, whereas the Persian version showed lower internal consistency (0.81-0.79) 9 for the right and left ankle. This study focused on recreationally active individuals, whereas differences in sample composition in other studies, including general population, university students, or elite athletes, may explain the variability in findings.4,9,15 The test-retest reliability results for the Turkish version of the CAIT were excellent for both the right and the left ankles.
The CAIT is a validated discriminative questionnaire capable of detecting AI, demonstrating high test-retest reliability and acceptable concurrent validity in comparison with the Lower Extremity Functional Scale. 11 Test-retest analysis was used to assess the ability of a measurement to produce consistent results at 2 different time points within the same group. 30 The test-retest reliability and internal consistency of the Turkish version of the CAIT were similar to those of the original English version (ICC2,1 = 0.96) 31 and previous adaptations of the CAIT.4,6,15,24
The CAIT generates a total score between 0 and 30, with higher scores reflecting greater ankle stability. 11 While the original validation study identified a cutoff score of ≤27 for detecting instability, a recalibrated analysis of the English version recommends the use of ≤25 as a more appropriate threshold.11,31 The discriminative ability and cutoff score of the CAIT-TR were determined by analyzing the ROC curve and calculating the Youden index (0.579 and 0.551 for the left and right ankles, respectively) with a cutoff score of 23.5. The ideal cutoff point in the Turkish population was lower than the English cutoff (≤27), 11 slightly lower than the recalculated English cutoff (≤25), 31 and equal to the French cutoff (≤23). 6 The cutoff score of the Turkish version was also comparable with those of other versions of the CAIT.18,28 This difference in cutoff scores may be due to the distinct characteristics of the study populations. As an example, the Dutch version, which was administered to patients hospitalized in an orthopaedic outpatient setting, identified a cutoff score of 11.5. 29 Unlike previous studies that recruited athletes, university students, or general population samples, the current study focused on recreationally active individuals, who are at increased risk for ankle injuries.4,9,15 In Turkish-speaking populations, a cutoff value of ≤23 can be used to assess AI in recreationally active individuals.
The SEM was used to calculate the minimum difference between 2 measurements that can be considered clinically significant. These values were similar to those of the Dutch version (0.82) 29 but lower than those reported in other versions of the CAIT.6,9,15,30 The MDC represents the smallest score change that can be considered a true change, unaffected by measurement error. In this study, MDC values were calculated as 1.51 and 1.14 for the right and left ankles, respectively, whereas previous studies reported a wide range of values (2.2, 29 4.8, 30 and 5.6-6.5 9 ). Researchers and clinicians should consider observed change scores to be indicative of a clinically meaningful difference in instability only when they exceed the MDC value. Changes below this threshold should be interpreted as falling within the limits of measurement error and therefore disregarded in clinical interpretation.
Ceiling and floor effects were defined as present when >15% of participants reached the scale’s maximal or minimal score, suggesting reduced sensitivity at the extremes of functional performance. 27 The original English version of the CAIT did not provide data on ceiling and floor effects. 11 Similar to most versions of the CAIT, no ceiling or floor effects were observed in this study.4,6,24,28,30
The CAIT-TR can be considered a valid and reliable self-reported instrument with demonstrated psychometric properties for the assessment of AI. For clinicians, it offers real-time and objective data regarding patients’ AI, enabling the assessment of recurrent sprain risk, and the implementation of effective patient monitoring. For researchers, it may facilitate the selection of more homogeneous control and instability groups, potentially enhancing the internal validity and consistency of study results. CAIT-TR serves as a valuable tool for the assessment of AI in both clinical practice and academic research.
Limitations
As a limitation of this study, it can be noted that the inclusion of only recreationally active individuals may result in variations in the identified cutoff values when applied to elite athletes, individuals assessed before and after treatment, or different age groups. Future research should be conducted to determine cutoff scores in specific age groups and different sports. If these values differ across groups, using appropriate cutoff scores for each group may improve clinical assessment. As a second limitation, there was some loss of participants due to exclusion criteria and loss to follow-up. However, the sample size remained sufficient for statistical analysis.
Conclusion
Our study demonstrates that the CAIT-TR is a valid and reliable tool for assessing the presence and severity of AI in recreationally active Turkish-speaking individuals. The cutoff score for identifying CAI was determined to be ≤23. These data should help in providing a practical assessment tool for care in recreationally active individuals.
Footnotes
Appendix
Turkish Version of Cumberland Ankle Instability Tool (CAIT-TR) a
HER BİR soruda ayak bileklerinizi EN İYİ tanımlayan SADECE BİR ifadeyi işaretleyiniz
| SOL | SAĞ | PUAN | ||
|---|---|---|---|---|
| 1 | Ayak bileğimde ağrı hissederim | |||
| Hiçbir zaman | □ | □ | 5 | |
| Spor yaparken | □ | □ | 4 | |
| Düz olmayan zeminlerde koşarken | □ | □ | 3 | |
| Düz zeminlerde koşarken | □ | □ | 2 | |
| Düz olmayan zeminlerde yürürken | □ | □ | 1 | |
| Düz zeminlerde yürürken | □ | □ | 0 | |
| 2 | Ayak bileğimin STABİL OLMADIĞINI hissederim | |||
| Hiçbir zaman | □ | □ | 4 | |
| Spor yaptığım bazı zamanlar (Her zaman değil) | □ | □ | 3 | |
| Spor yaparken sıklıkla (Her zaman) | □ | □ | 2 | |
| Günlük aktiviteler sırasında bazı zamanlar | □ | □ | 1 | |
| Günlük aktiviteler sırasında sıklıkla | □ | □ | 0 | |
| 3 | KESKİN dönüşler yaparken ayak bileğimin STABİL OLMADIĞINI hissederim | |||
| Hiçbir zaman | □ | □ | 3 | |
| Koşarken bazen | □ | □ | 2 | |
| Koşarken sıklıkla | □ | □ | 1 | |
| Yürürken | □ | □ | 0 | |
| 4 | Merdivenlerden inerken ayak bileğimin STABİL OLMADIĞINI hissederim | |||
| Hiçbir zaman | □ | □ | 3 | |
| Hızlı indiğim zaman | □ | □ | 2 | |
| Ara sıra | □ | □ | 1 | |
| Her zaman | □ | □ | 0 | |
| 5 | TEK ayak üzerinde durduğumda ayak bileğimin STABİL OLMADIĞINI hissederim | |||
| Hiçbir zaman | □ | □ | 2 | |
| Parmak ucunda durduğumda | □ | □ | 1 | |
| Düz bastığım zaman | □ | □ | 0 | |
| 6 | Ayak bileğimin STABİL OLMADIĞINI hissederim | |||
| Hiçbir zaman | □ | □ | 3 | |
| Sağa ve sola doğru tek ayakla sıçradığımda | □ | □ | 2 | |
| Tek ayak sıçradığımda | □ | □ | 1 | |
| Sıçradığım zaman | □ | □ | 0 | |
| 7 | Ayak bileğimin STABİL OLMADIĞINI hissederim | |||
| Hiçbir zaman | □ | □ | 4 | |
| Düz olmayan zeminlerde koştuğumda | □ | □ | 3 | |
| Düz olmayan zeminlerde jog attığımda | □ | □ | 2 | |
| Düz olmayan zeminlerde yürüdüğümde | □ | □ | 1 | |
| Düz bir zeminde yürüdüğümde | □ | □ | 0 | |
| 8 | GENEL OLARAK ayak bileğim dönmeye/burkulmaya başladığında bunu durdurabilirim | |||
| Anında | □ | □ | 3 | |
| Çoğunlukla | □ | □ | 2 | |
| Ara sıra | □ | □ | 1 | |
| Hiçbir zaman | □ | □ | 0 | |
| Ayak bileğim hiç dönmedi | □ | □ | 3 | |
| 9 | GENEL OLARAK ayak bileğim döndüğünde/burkulduğunda “normale” dönme süresi | |||
| Neredeyse anında | □ | □ | 3 | |
| Bir günden kısa sürer | □ | □ | 2 | |
| 1-2 gün sürer | □ | □ | 1 | |
| 2 günden fazla sürer | □ | □ | 0 | |
| Ayak bileğim hiç dönmedi | □ | □ | 3 | |
Puanlama ölçeği sağdadır. Puanlama sistemi katılımcı versiyonunda görünmemektedir.
Final revision submitted April 7, 2025; accepted April 30, 2025.
One or more of the authors has declared the following potential conflict of interest or source of funding: Eskişehir Technical University, Scientific Research Projects Commission, supported this study under the project No: 22ADP299. AOSSM checks author disclosures against the Open Payments Database (OPD). AOSSM has not conducted an independent investigation on the OPD and disclaims any liability or responsibility relating thereto.
Ethical approval for this study was obtained from the Board of Scientific Research and Publishing Ethics of Eskişehir Technical University on April 28, 2022 (Decision No. 9/1; Document Verification Code: BSRKY1ECUU).
