Download PDF
Hamostaseologie
DOI: 10.1055/a-2687-9339
Original Article

The Timed Up and Go Test in Patients with Haemophilia: Assessing Reliability, Validity, and Predictive Variables

Authors

  • F. Tomschi

    1   Department of Sports Medicine, University of Wuppertal, Wuppertal, Germany

  • M. Brühl

    1   Department of Sports Medicine, University of Wuppertal, Wuppertal, Germany
    2   Department of Orthopaedics and Trauma Surgery, University of Bonn, Bonn, Germany

  • A. Schmidt

    1   Department of Sports Medicine, University of Wuppertal, Wuppertal, Germany

  • J. Wiese

    1   Department of Sports Medicine, University of Wuppertal, Wuppertal, Germany

  • A. Lorenz

    1   Department of Sports Medicine, University of Wuppertal, Wuppertal, Germany

  • T. Hilberg

    1   Department of Sports Medicine, University of Wuppertal, Wuppertal, Germany

Funding This study has received no specific grants from commercial, public, or non-profit entities and is totally self-funded.
Further Information
Also available at
 
 PDF Download Permissions and Reprints

Abstract

Background

The Timed Up and Go (TUG) test is frequently used to assess patients' functional mobility. However, its psychometric characteristics in patients with haemophilia (PwH) are unknown. This study's primary aim was to determine the validity, reliability, standard error of measurement (SEM), and minimal detectable change (MDC) of the TUG in PwH. The secondary aim was to determine predictors for the TUG time.

Methods

A total of 40 PwH were included. Test-retest reliability was assessed by the same rater at two time points and inter-rater reliability was assessed by two raters. Construct validity was tested via correlation analyses between the TUG and the haemophilia joint health score (HJHS), the short physical performance battery (SPPB), the HEP-Test-questionnaire, and the Haemophilia Activity List (HAL). SEM and MDC were calculated. Multiple linear regression analyses with several patient-specific predictors were performed.

Results

Test-retest and inter-rater reliability analyses revealed excellent ICCs of 0.990 (95% CI: 0.972–0.995) and 0.929 (95% CI: 0.870–0.962), respectively. The SEM and MDC of the TUG were 0.34 and 1.52 seconds, respectively. Large correlations (r > 0.5) were observed between the TUG and the HJHS, SPPB, HEP-Test-Q, and HAL. Regression analysis revealed the HJHS as the sole significant predictor, with the full model explaining 37.0% of the variance in TUG performance.

Conclusion

In PwH, the TUG is a reliable test possessing an excellent test-retest and inter-rater reliability, while showing a high validity. TUG times can mainly be predicted by HJHS. The TUG can therefore be considered a suitable tool to evaluate mobility in adult PwH.


 

Keywords

rare disease - functional mobility - motor performance - physical testing

Summary

This study evaluated the psychometric properties of the Timed Up and Go (TUG) test in patients with haemophilia (PwH). A total of 40 PwH were assessed to determine TUG reliability, validity, standard error of measurement (SEM), and minimal detectable change (MDC). Test-retest and inter-rater reliability were excellent (ICC = 0.990 and 0.929). The SEM was 0.34 seconds, and the MDC was 1.52 seconds. Validity was confirmed by large correlations between TUG times and the HJHS, SPPB, HEP-Test-Q, and HAL. Regression analysis identified joint health (HJHS) as the sole significant predictor, with the full model explaining 37.0% of the variance in TUG performance. The TUG is thus a valid and reliable tool to assess functional mobility in PwH.


 

Introduction

Haemophilia is an inherited coagulation disorder caused by a deficiency or dysfunction of specific clotting factors, resulting in an increased propensity for bleeding episodes occurring mostly in joints and more seldomly in muscles. Haemophilia A and B affect approximately 1 in 5,000 and 1 in 30,000 male births, respectively, with a global prevalence estimated at over 400,000 individuals.[1] In patients with haemophilia (PwH), joint bleeds (i.e., haemarthroses) are commonly present with symptoms such as pain, joint swelling, and reduced mobility. When such bleeding episodes occur repeatedly, they not only increase the likelihood of future haemarthroses but also trigger degenerative changes within the joint. Over time, this can lead to the development of haemophilic arthropathy (HA), a chronic joint condition marked by progressive damage and dysfunction.[2] [3] The joints most commonly affected are the ankles, knees, and elbows,[4] and these impairments can compromise daily activities and independence, especially in aging patients with cumulative joint damage.[2] Clinical functional performance tests are to be seen as a supportive tool in the management and treatment of PwH.[5] These assessments provide crucial insights into the patients' everyday mobility capacities and mobility safety.[6] In the context of haemophilia care, the haemophilia joint health score (HJHS) stands out as the most frequently used assessment that is validated and reliable and specifically designed to evaluate joint health in adult PwH.[7] Despite its specific utility and wide application, the HJHS primarily focuses on joint-specific assessments and does not measure functional movement capabilities in a metric manner, but only by subjectively scoring gait parameters. Therefore, it was recently postulated that there is, on the one hand, a need to identify performance-based instruments capable of quantifying functional mobility, as these instruments can serve as essential tools to complement the existing subjective self-reported functional questionaries. On the other hand, a crucial recommendation was made to assess the psychometric properties of different clinical functional movement tests.[8] In this context, the Four-Square Step Test (FSST), an instrument primarily assessing dynamic balance, was recently validated in PwH, further supporting the relevance of such performance-based measures in this population.[9]

Preliminary work by Bladen et al have identified the Timed Up and Go (TUG) as a practical and useful test in adults as ranked by clinicians.[8] Yet, it is crucial to acknowledge that further psychometric evaluation is required before the TUG can entirely be recommended in this specific patient cohort. The TUG, known for its simplicity and effectiveness,[10] [11] [12] can be serve as a tool to assess functional mobility capabilities in PwH. Indeed, the most commonly used test for evaluating functional mobility and fall risk and lower limb strength in the clinic is the TUG.[13] In the context of haemophilia research, this test is frequently used,[9] [14] [15] and a recent study showed that the TUG is a feasible and safe test in adult PwH across a wide age range (age: 32–79 years) and across a wide joint status range (HJHS: 14–64 points).[2] To date, uncertainty remains regarding the TUG's ability to accurately measure functional mobility (validity) and its consistency across different assessments and raters (reliability) in adult PwH.

Based on these considerations, this study aims to conduct a psychometric evaluation of the TUG in adult PwH, while the secondary objective is to examine potential patient-specific characteristics that predict the TUG time.

Therefore, the following hypotheses are stated: (1) The TUG is a valid measure of functional mobility in adult PwH, demonstrating a moderate to large correlation with the HJHS and other established objective clinical tests and subjective measures measuring the same construct. (2) The TUG possesses a high reliability within a single rater as well as between different raters. (3) TUG times can be predicted by age, annualized bleeding rate (ABR), and HJHS.


 

Methods

General Study Design

This study was prospectively registered in the German Clinical Trials Register (DRKS00034103). All patients provided their written informed consent according to the Declaration of Helsinki, and the study protocol was approved by the local ethics committee (University of Wuppertal, SK/AE 240326). The TUG was administered by two raters (A, B), while rater A performed the TUG two times (test-retest reliability) and rater B one time (inter-rater reliability). Rater B also performed the SPPB testing. The order of the performed tests was randomized. Randomization was conducted by an independent researcher not involved in the study, using a pre-generated randomization list. In addition, the HJHS was administered to evaluate clinical joint status, and PwH completed standardized questionnaires. A detailed description of these assessments is provided below. The GRRAS checklist for reporting of studies for reliability and agreement was employed for this study.[16]


 

Participants

The sample size was determined using an a priori G*Power analysis, assuming a large standardized correlation (effect size = 0.5) between the TUG and the SPPB, as observed in a previous study.[2] Using an α-error probability of 0.05 and a power of 0.90 in an exact test-correlation bivariate two-tailed model, the minimum sample size required was 37 participants. Male adults aged 18 years or older with a diagnosis of mild, moderate, or severe haemophilia A or B were eligible for inclusion. Individuals with bleeding disorders unrelated to haemophilia, those who had undergone invasive joint therapies (e.g., radiosynoviorthesis) or surgeries (e.g., total endoprosthesis or arthrodesis) within the 6 months prior to the examination, or those who had taken acute analgesics outside their regular treatment regimen within 12 hours before the examination were excluded from participation. Additionally, a bleeding episode within the 2 weeks prior to the examination also led to exclusion.


 

Timed Up and Go Test Procedure

The TUG test was conducted under standardized conditions as previously described for patients with knee osteoarthritis[12] and rheumatoid arthritis.[10] Participants were instructed to stand up from a chair with armrests with a seat height of 46 cm, walk 3 metres (m) comfortably and safely to a cone, turn around, return to the chair, and sit down again. The wording of the instructions was standardized. The time taken to complete the task was measured using a stopwatch with a precision of 1/100 seconds. Timing began with the word “go” after a 3-second countdown and was stopped when the participant was completely seated. Patients were allowed to use assistive devices during the task if necessary. The TUG tests were administered at two time points by the same rater (Rater A; F.T.) to evaluate test-retest reliability. The interval between tests was approximately 60 minutes,[17] which was chosen to avoid changes in the patient's condition (e.g., bleeding episode, different factor level, use of pain medication) that might potentially have occurred when longer time periods had been used. Further, given that haemophilia is a rare disease, study participants often had to travel significant distances to attend the examination. Therefore, conducting all assessments within a single visit was essential to minimize patient burden and ensure feasibility, while allowing enough time between tests to reduce potential fatigue or learning effects. Additionally, an independent second rater (Rater B; M.B.) conducted the TUG test in a separate room, without knowledge of outcomes measured by Rater A, to assess inter-rater reliability. Two TUG trials were performed by each patient under each condition, and the average time of the two recorded trials was used for data analysis. Both raters possessed several years of experience of performing the TUG in clinical and research settings and had received professional training.


 

Further Assessments

The HJHS v2.1 was employed for the clinical orthopaedic assessment of joint health and all HJHS examinations were performed by the same single experienced researcher (F.T.). This comprehensive evaluation focuses on the ankles, knees, and elbows and examines axial deformity, muscle atrophy, swelling, range of motion (measured with a goniometer for extension and flexion), crepitus during movement, joint pain, and muscle strength. Additionally, gait analysis was included as part of the assessment. The HJHS produces a total score ranging from 0, representing optimal joint health, to 124, indicating the worst functional and structural joint condition.[7]

Patients performed the SPPB, as a closed test battery, consisting of three components: a standing balance test (two-legged stand, semi-tandem stand, and tandem stand), a 4-m walk test performed at a normal pace, and a five-times sit-to-stand test, which measures the time required to rise from a chair without armrests five times consecutively as fast and safe as possible. The patient's arms are positioned on the chest. The total SPPB score ranges from 0 to 12, with a maximum of 4 points possible for each test. Higher scores indicate better functional ability. The SPPB has been validated as safe and feasible for use in PwH.[2] [18] Besides, metric results of the five-times sit-to-stand test (in seconds) and the gait speed measured during the 4-m walk test (in metres per second) were documented and used for further analyses. When patients were not able to perform the five-times sit-to-stand test within 30 seconds, 0 points were documented for the SPPB score. For the metric evaluation, these datasets were not considered in any analyses. The SPBB was conducted by a single experienced researcher (M.B.).

Subjective physical functioning was evaluated using the validated and disease-specific questionnaires HAL and HEP-Test-Q, in which higher scores reflect higher physical functioning.[19] [20]


 

Statistics

Statistical analyses were performed using the IBM© SPSS 29 software (Armonk, NY, USA) and data are presented as median (25%-quartile, 75%-quartile) [min-max] unless otherwise indicated. Normal distribution testing revealed slight deviations with regard to kurtosis. However, due to the robust nature of ANOVA-based ICC analyses and the sufficiently large sample size, the primary analyses are performed using the original data, unless otherwise indicated. Further, this approach enables an adequate and easy interpretation of untransformed results.[21]

To evaluate convergent construct validity, TUG scores were compared with the results obtained from the measures of HJHS, SPPB, HAL, and the HEP-Test-Q using Spearman's rho (rs ). The magnitude of the correlations can be interpreted with 0.1, 0.3, and 0.5 representing small, moderate, and large correlations, respectively.[22]

Test-retest and inter-rater reliability were analyzed using intraclass correlation coefficients (ICC(3,1)) with absolute agreement and the presentation of single measures.[23] To test the robustness of these results, the TUG data were further logarithmically transformed (log 10 transformation) to achieve a normal-like distribution, as previously done.17 Bland-Altman plots were generated to provide a visual representation of agreement. The standard error of measurement (SEM) was calculated using the formula SEM = SD with SD being the SD of the first TUG test performed and r being the ICC of the test-retest calculation. The minimum detectable change (MDC) was calculated using the formula MDC = 1.96 .

A multiple linear regression analysis was conducted to identify the most significant patient-specific predictors of the TUG time (rater A, first test). The chosen predictors included HJHS,[2] BMI,[24] age,[25] ABR,[26] haemophilia type,[27] and haemophilia severity[28] as they were shown to have an influence on the musculoskeletal system influencing functional mobility. Predictors were included using the enter method. The regression model was assessed using the adjusted R2 to evaluate the proportion of variance explained, while the ANOVA test determined the overall model significance. To assess the assumption of homoscedasticity, the Breusch-Pagan test was conducted with no need to adjust the analysis. Multicollinearity was examined using variance inflation factor (VIF) values, ensuring no significant collinearity among predictors (VIF < 10).


 

 

Results

The study included 40 PwH and anthropometric and disease-specific data are presented in [Table 1]. All PwH received prophylaxis therapy, except for two patients who received on-demand therapy (one mild and one moderate patient). TUG times required during the different conditions are presented in [Table 2]. In the study, 13 (32.5%) PwH required >10.0 seconds to complete the TUG (Rater A, Test 1). The results of the other assessments are as follows: SPPB total: 10.0 points (8.0, 11.0) [3.0–12.0]; SPPB five-times sit to stand: 2.0 points (0.3, 3.0) [0.0–4.0]; SPPB 4-m walk: 4.0 points (4.0, 4.0) [2.0–4.0] and 1.1 m/s (1.0, 1.2) [0.5–1.5]; SPPB balance: 4.0 (3.3, 4.0) [1.0–4.0]; HAL: 81 points (66, 94) [18–100]; and HEP-Test-Q: 62 points (48, 78) [20–96]. Time needed for the SPPB five-times sit to stand was 14.4 seconds (11.9, 15.7) [10.4–21.4], yet 10 patients were not able to conduct this test within the set cut-off time of 30 seconds. No adverse events of any kind (e.g., bleeding episode, accidental fall, dizziness) occurred as part of the TUG and SPPB.

Table 1

Anthropometric and disease-specific data of PwH (N = 40)

Age (years)

59 (50, 63) [21–77]

Body mass (kg)

79 (73.5, 86.8) [54.0–115.0]

Height (cm)

177.5 (174.0, 184.0) [163.0–191.0]

Body mass Index (kg/m2)

24.9 (23.6, 26.8) [20.3–31.5]

Haemophilia type ( n )

A = 32, B = 8

Haemophilia severity ( n )

Severe = 32, moderate = 7, mild = 1

Therapy ( n )

Prophylaxis = 38, on-demand = 2

HJHS (score points)

34 (21, 51) [1–75]

ABR

0 (0, 1.5) [0–7]

Abbreviations: ABR, annualized bleeding rate; HJHS, Haemophilia Joint Health Score.


Note: Data presented as median (25%-quartile, 75%-quartile) [min-max] unless otherwise marked.


Table 2

Timed Up and Go test (TUG) times under three different conditions

Rater A, Test 1

Rater A, Test 2

Rater B, Test 1

TUG [seconds]

Median (25%-quartile, 75%-quartile) [min-max]

9.11 (8.19, 10.90) [5.42–21.69]

9.03 (8.01, 10.50) [5.66–20.52]

9.50 (8.49, 11.35) [5.81–18.54]

Mean ± SD

10.16 ± 3.43

9.92 ± 3.28

10.22 ± 2.96

Test-retest and inter-rater reliability analyses reveal excellent ICCs of 0.990 (95% CI: 0.972–0.995) and 0.929 (95% CI: 0.870–0.962), respectively. The SEM and MDC of the TUG are 0.34 and 1.52 seconds, respectively. Test-retest and inter-rater reliability analyses using log10 transformed TUG data reveal very similar ICCs of 0.985 (95% CI: 0.966–0.993) and 0.921 (95% CI: 0.857–0.957), respectively. Results of the correlation analyses for the convergent construct validity are presented in [Fig. 1]. [Fig. 2A, B] depict the Bland-Altman plots with biases (mean differences) and 95% limits of agreement (LoA) for the intra- and inter-rater tests.

Zoom
Fig. 1 Correlogram of correlation analyses between the Timed Up and Go (TUG) and further measures. Results are presented as rs values. The darker the boxes the stronger the correlation (see bar right to the correlogram). HAL, Haemophilia Activity List; HEP, HEP-Test-Q; HJHS, haemophilia joint health score; SPPB, short physical performance battery.
Zoom
Fig. 2 Bland-Altman plots of the difference between two readings of the Timed Up and Go time (seconds) by the same rater (A, intra-rater) and between two raters (B, inter-rater) with mean differences (bias) and upper and lower levels of agreement (LoA) of differences between the two readings.

In the light of the multiple linear regression analysis, the regression model including HJHS score, BMI, ABR, haemophilia type, haemophilia severity, and age accounted for 37.0% (adjusted R2 = 0.370) of the variance in TUG performance (R = 0.68, unadjusted R2 = 0.467; SEE = 2.73 seconds F(6,33) = 4.825, p = 0.001). Only the HJHS score was detected as a statistically significant predictor (p = 0.013). Additionally, VIF values were below 2, indicating no multicollinearity issues. [Table 3] presents the individual regression coefficients. The regression analysis was confirmed by the regression analysis using log10 transformed TUG data (see [Supplementary Material]).

Table 3

Regression coefficients from the multiple linear regression model predicting Timed Up and Go (TUG) time

Predictor

B

SE

Beta

t

p-value

Tolerance

VIF

Constant

0.693

0.078

−

8.832

<0.001

−

−

ABR

−0.001

0.002

−0.094

−0.711

0.482

0.920

1.087

Type

−1.068

1.161

−0.126

−0.920

0.364

0.860

1.162

Severity

0.587

1.003

0.082

0.585

0.563

0.822

1.216

Age

0.074

0.050

0.245

1.501

0.143

0.608

1.645

BMI

0.144

0.169

0.110

0.850

0.401

0.966

1.035

HJHS

0.078

0.030

0.454

2.617

0.013

0.537

1.863

Abbreviations: ABR, annualized bleeding rate; B, unstandardized regression coefficient; Beta, standardized regression coefficient; HJHS, haemophilia joint health score; p-value, probability value; SE, standard error of B; t, t-statistic; Tolerance, collinearity statistic; VIF, variance inflation factor.


Note: The model included HJHS score, ABR, BMI, haemophilia type, haemophilia severity, and age as predictors, explaining 37.0% (adjusted R2 = 0.37) of the variance in TUG performance (unadjusted R2 = 0.467). Only the HJHS score emerged as a statistically significant predictor (p = 0.013).



 

Discussion

This study highlights the TUG as a reliable and valid tool for assessing functional mobility in PwH. The obtained results support hypothesis (1), with an excellent test-retest reliability (ICC of 0.990) and a similarly high inter-rater reliability (ICC of 0.929) observed. These values demonstrate excellent consistency when the test is administered by the same or different raters, underlining the robustness of the TUG test for clinical and research applications. Results further support hypothesis (2) as the TUG was largely correlated with established constructs assessing functional mobility of PwH, such as the HJHS, SPPB (total score), gait speed in the 4-m walk test, HEP-Test-Q, and HAL (see [Fig. 1]). Hypothesis (3) was only partially supported. Although the regression model explained 37.0% of the variance in TUG performance, the only significant predictor was the HJHS score. In contrast, age and ABR did not significantly contribute to the model. In particular, these values of the reliability test are consistent with previous studies in populations with musculoskeletal conditions where the TUG test demonstrated high reliability, such as in patients with rheumatoid arthritis (ICC 0.95–0.98)[10] and knee osteoarthritis (ICC 0.96–0.97).[12]

The TUG can be described as highly valid in PwH, as large correlations are observed between the TUG times and further objective clinical tests measuring the same construct. Large correlations are observed with the HJHS, total SPPB score, and 4-m walk time. The TUG has been reported to correlate with the SPPB total score in patients with total knee arthroplasty (r = −0.78),[29] which is a similar correlation as observed herein. Low to moderate correlations are found between TUG times and the five-times sit-to-stand time. Although patients are asked to perform the TUG at a safe and comfortable gait speed, they are instructed to complete the five-times sit-to-stand test as fast as possible. Hence, these two tests assess different physical performance categories, as explosive lower limb strength is required to achieve fast times on the five-times sit-to-stand test.[30] Due to the different movement demands of the TUG and the five-times sit-to-stand test, greater deviations are observed between the TUG times and the five-times sit-to-stand test. Besides, datasets of patients unable to perform the five-times sit-to-stand test within 30 seconds were excluded from the analysis using the metric outcome. In contrast, large correlations are observed when the SPPB five-times sit-to-stand score (range: 0–4) is used.

Moreover, large correlations are observed between the TUG and subjective measures of physical performance capacity (HAL, HEP-Test-Q), indicating that the TUG not only reflects objective mobility impairments, but also aligns well with patient-reported functional limitations. This aligns with previous research in PwH[2] and suggests that the TUG is a comprehensive measure of functional mobility in PwH, capturing both the physical and perceived impact of mobility limitations.

Our findings indicate that the overall regression model, which included HJHS, BMI, ABR, haemophilia type, haemophilia severity, and age, explained 37.0% (adjusted R2 = 0.37, unadjusted R2 = 0.467) of the variance in TUG performance. These results are supported by the regression model predicting the TUG time using the log10 transformed TUG data as the dependent variable, where 43.3% (adjusted R2 = 0.433, unadjusted R2 = 0.520) of the variance in TUG performance were explained. Among these variables, only the HJHS score emerged as a statistically significant predictor (in both models). This reinforces the relevance of the HJHS as the overarching predictor of mobility impairment in this population, appearing to outweigh other factors such as age and ABR in its association with TUG performance. Interestingly, ABR was not a significant predictor of TUG time, likely reflecting advancements in new factor replacement therapies and non-factor therapies, which have significantly reduced ABR to nearly zero in many patients.[31] [32] [33] As a result, ABR may no longer adequately capture individual functional impairments, as even patients with minimal or no bleeding histories (e.g., within the recent years) can exhibit haemophilic arthropathy due to previous or not detected subclinical bleedings.[34] This shift highlights the need for alternative parameters beyond ABR to assess mobility outcomes in haemophilia care.[2] [8]

The median time required to complete the TUG test in the present study was 9.03 to 9.50 seconds (mean: 9.92–10.22 seconds), which is comparable to results reported in other studies involving adults with musculoskeletal diseases. For example, Alghadir et al reported a mean time of 10.88 seconds in patients with knee osteoarthritis.[12] Similarly, in patients with long-standing rheumatoid arthritis a mean TUG time of 11.1 was reported.[10] Notably, Podsiadlo and Richardson, in their foundational study of the TUG test in frail elderly individuals, reported a mean time of 10.0 seconds[35] and results presented herein reveal that 13 out of 40 PwH (32.5%) required >10 seconds to complete the TUG. It should be noted that the cut-off value for PwH has not been scientifically confirmed and several studies have been conducted to establish normative values for the TUG test across different age groups and patient populations with potential cut-off values ranging from 7.7 to 14.5 seconds.[13] [35] [36] [37] [38] [39]


 

Limitations

One limitation is the use of a manual stopwatch to record time, which may introduce minor human error but at the same time depicts a real-world scenario in clinical practice. Future studies might consider automated timing systems such as light barriers to enhance measurement precision. It is also important to note that although the SPPB is widely used in older adults and other musculoskeletal populations, it has not yet been formally validated in PwH, which could affect the interpretability of findings. However, in the absence of a gold standard for functional mobility tests in PwH, this approach is commonly used, and the TUG has previously been validated against the SPPB in other populations.[11] A further limitation of this study is that it evaluated the statistically calculated MDC but the absence of an established minimal clinically important difference (MCID) value for the TUG in PwH limits the clinical interpretation of meaningful change. Future studies employing anchor-based methods are required to define MCID thresholds for this population. The sample size calculation was based on a standardized large expected correlation (r = 0.5) derived from our previous findings, which may have underestimated the required sample size. Future studies might consider more conservative estimates (e.g., r = 0.3), as recommended in recent methodological literature.[40] The same accounts for the evaluation of reliability. Here, the sample size can be considered as being slightly below the optimal thresholds recommended for reliability studies (e.g., 45–55 participants).[41]


 

Conclusion

The psychometric evaluation of the TUG in adult PwH presented herein provides the scientific basis for its use as a clinical assessment tool for functional mobility in this specific patient cohort revealing an excellent test-retest and inter-rater reliability. Besides, a high validity was observed compared with both objective and subjective measures. The regression model accounted for 37.0% of the variance in TUG performance, identifying the HJHS as the only significant predictor, while age and ABR showed no meaningful contribution. The TUG is a fast, feasible, and psychometrically sound tool for assessing mobility and fall risk and may be particularly useful in older PwH or PwH with existing joint involvement where identifying deteriorating mobility and fall risk might require intervention. However, to ensure accurate and safe administration, which is particularly important in individuals with advanced joint impairments, it should be conducted by trained healthcare professionals.


 

 

Conflict of Interest

The authors declare that they have no conflict of interest.

Acknowledgements

The authors would like to thank the patients for their participation in this study.

Ethics Approval and Consent to Participate

The study was approved by the ethics committee of the University of Wuppertal (SK/AE 240326). All patients provided written informed consent.


Consent to Participate

All patients provided written informed consent to participate in this study.


Availability of Data and Materials

The datasets of the current study are available from the corresponding author on reasonable request.


Authors' Contributions

F.T.: had the idea of this study, designed the study, wrote the first draft of the manuscript, conducted statistical analysis, performed measurements; M.B.: conducted measurements; M.B., A.S., A.L., J.W., and T.H.: revised the manuscript and provided expertise; T.H.: supervised the study.


Supplementary Material


Address for correspondence

Dr. Fabian Tomschi
Department of Sports Medicine, University of Wuppertal
Moritzstr. 14, 42117 Wuppertal
Germany   

Publication History

Received: 16 May 2025

Accepted: 21 August 2025

Article published online:
13 October 2025

© 2025. Thieme. All rights reserved.

Georg Thieme Verlag KG
Oswald-Hesse-Straße 50, 70469 Stuttgart, Germany


  Permissions and Reprints
Zoom
Fig. 1 Correlogram of correlation analyses between the Timed Up and Go (TUG) and further measures. Results are presented as rs values. The darker the boxes the stronger the correlation (see bar right to the correlogram). HAL, Haemophilia Activity List; HEP, HEP-Test-Q; HJHS, haemophilia joint health score; SPPB, short physical performance battery.
Zoom
Fig. 2 Bland-Altman plots of the difference between two readings of the Timed Up and Go time (seconds) by the same rater (A, intra-rater) and between two raters (B, inter-rater) with mean differences (bias) and upper and lower levels of agreement (LoA) of differences between the two readings.