The reliability of upper limb rotation test in adolescent male basketball players

Abstract

Background

The Upper Limb Rotation Test is an upper limb performance assessment specific to overhead throwing that requires shoulder control and stability, while incorporating the entire kinetic chain. The purpose of this study is to investigate the reliability of the Upper Limb Rotation Test (ULRT) in adolescent male basketball players.

Methods

This study was conducted with 51 male basketball players at Hanibal Sports Club. It is an intra-rater reliability study. The Upper Limb Rotation Test was administered twice under the same conditions with a one-week interval. Intra-rater reliability was quantified using ICC (two-way mixed-effects, absolute agreement), Bland–Altman analysis, the standard error of measurement (SEM), and the 95% minimal detectable change (MDC).

Results

The intra-rater reliability of the test was found to be ICC = 0.93 for the dominant extremity and ICC = 0.96 for the nondominant extremity (p < 0.001). The SEM value was found to be dominant extremity 1.38, nondominant extremity 0.89, while the MDC value was found to be dominant extremity 3.82, nondominant extremity 2.46. Bland–Altman plots showed homoscedastic scatter with a small negative bias (retest slightly higher) and few points marginally outside the 95% limits of agreement.

Conclusions

Among adolescent male basketball players, the ULRT is a clinician-friendly, easily administered, and practical test for assessing upper-extremity rotation in a kinetic-chain position, and it can be reliably applied in this population.

Trial registration

This study is prospectively registered at NCT06269601 (clinicaltrials.gov) (Registration Date: 27/02/2024).

Introduction

Basketball is a globally popular team sport characterized by high-intensity, physical gameplay that requires a combination of strength, power, speed, balance, coordination, and dribbling skills [14, 22]. The increasing participation of adolescent athletes in sports, combined with early specialization in a single sport that involves intensive and repetitive activities, contributes to a higher incidence of overuse injuries [2, 31]. Approximately 30% of sports-related injuries in young overhead athletes have been reported to occur in the shoulder region [31].

The importance of upper extremity assessments has been emphasized as part of the periodic health evaluations of young athletes [1, 23, 26]. Physical performance tests play a crucial role in this evaluation by providing an objective assessment of the functional performance of the upper extremities in young athletes [4, 5, 27]. Physical performance tests are also frequently used to assess athletic performance and to evaluate outcomes following rehabilitation [4, 5, 8, 27]. The simplicity of administration, low equipment demands, and established validity and reliability of functional performance tests provide significant benefits for performance evaluation and assessment [13].

Although the literature contains a wide range of physical performance tests designed to assess the lower extremities, the number of physical performance tests available for evaluating the upper extremities remains limited [32]. The most used tests for assessing upper extremity performance include the Closed Kinetic Chain Upper Extremity Stability Test (CKCUEST), the Seated Medicine Ball Throw Test (SMBT), and the Upper Quarter Y-Balance Test [17, 28, 33]. Although the physical performance tests available in the literature for assessing upper extremity performance evaluate upper extremity functionality, they do not fully account for the specific requirements of overhead throwing, such as the combination of open kinetic chain and closed kinetic chain movements, trunk rotation, and the 90°/90° shoulder position [30].

Decleve et al. developed the Upper Limb Rotation Test (ULRT) to provide an assessment specifically tailored to the demands of overhead throwing [7]. The Upper Limb Rotation Test requires shoulder motor control and stability while bearing body weight. Initiating the test in a modified plank position allows for the inclusion of the entire kinetic chain in the assessment. Additionally, the test evaluates the shoulder in a position specific to overhead throwing, with 90° of abduction and 90° of external rotation [7]. We hypothesized that the Upper Limb Rotation Test, which allows for the assessment of the shoulder in a position specific to overhead throwing, is a reliable tool in adolescent male basketball players. The purpose of this study is to investigate the reliability of the Upper Limb Rotation Test in adolescent male basketball players.

Establishing the intra-rater reliability and absolute measurement error (SEM, MDC95) of the ULRT in adolescent male basketball players provides the psychometric foundation to use this test for performance monitoring and clinical decision-making in youth basketball.

Methods

Study design and ethics

This research was designed to evaluate the intra-rater reliability of the ULRT using a two-session measurement design separated by seven days. This study was approved by the Ethics Committee of Sakarya University of Applied Sciences (approval date: 04.08.2023, decision no: 34/14) and was conducted by the Hanibal Basketball Sports Club. Since the athletes in the study were adolescents, written informed consent was obtained from their parents. Before data collection, the researchers explained to the participants the purpose of the study, that voluntary participation was essential, that anonymity was guaranteed, and that their data would be used only for this study. The participants were also allowed to withdraw from the study any time without stating any reason. The Declaration of Helsinki was adhered to throughout all phases of this study. This study is prospectively registered at NCT06269601 (clinicaltrials.gov) (Registration Date: 27.02.2024).

Participants

The participants consisted of adolescent basketball players aged between 14 and 18 years, recruited from Hanibal Sports Club. Demographic characteristics of the participants are presented in Table 1. To determine the sample size, a pilot study was conducted with a group of 20 participants, where a performance test was applied. The intra-class correlation coefficient (ICC) calculated from the pre- and post-test measurements on the non-dominant side was found to be 0.913. In a previous master’s thesis, the ICC for the same performance test was reported as 0.81 [9]. Using the "ICC.Sample.Size" package in R Studio (Version 4.3.0), the required sample size was calculated with a power (1-β) of 0.80 and an α level of 0.05 using the calculate ICC Sample Size function. As a result, the required sample size was determined to be 46 participants. However, considering the potential for participant dropout, a dropout rate of 10% was added. Therefore, the final required sample size was calculated as 51 participants. The inclusion criteria were between 14 and 18 years of age, having at least two years of basketball experience, and participating in training at least two days per week. The exclusion criteria had a history of any neurological or orthopedic injury. Demographic information (age, sports age, weight, height, body mass index), dominant extremity was recorded.

Table 1.

Demographic characteristics of the participants

Characteristics	Mean ± SD
Age (years)	15.96 (1.44)
Height (cm)	180.88 (5.56)
Body mass index (kg/m2)	20.57 (1.77)
Sport age (years)	6.23 (1.86)

Abbreviations SD standard deviation, cm centimeter, kg kilogram

ULRT Procedure

The validity and reliability of the ULRT were established in healthy adults by Decleve et al. (dominant extremity ICC = 0.76, nondominant extremity ICC = 0.78) [7]. The ULRT was administered in accordance with the instructions provided by Decleve et al. The participants participated in two assessment sessions conducted by the same investigator (the investigator was a PhD student in the Sports Physiotherapy program). The participants started the test in a modified push-up position (on elbows). Their backs were parallel to the floor, elbows flexed to 90°, feet shoulder-width apart, and arms positioned perpendicular to the floor. The participants were positioned next to a wall, ensuring that the shoulder, the epicondyle of the elbow, the greater trochanter of the femur, and the lateral malleolus of the ankle were in contact with the wall.

Participants were instructed to perform a trunk rotation combined with shoulder external rotation in a 90°−90° position (90° abduction and 90° external rotation) and to touch a vertically placed tape on the wall as quickly as possible (Fig. 1). A vertical tape was placed on the wall at a height corresponding to each participant’s full arm length in the 90° shoulder abduction and 90° external rotation position. This ensured that each participant reached a standardized height relative to their individual body dimensions. The purpose of the tape placement was to ensure that participants touched the wall in the correct 90°−90° shoulder position while rotating with the non-weight-bearing arm. The weight-bearing arm was positioned in line with the tape, while the opposite arm was instructed to reach and touch the wall in the 90°/90° shoulder position and then return to the starting position. Following the instructions and a demonstration, participants completed a familiarization trial consisting of 3 repetitions for each side. Verbal cues were provided when necessary. Finally, participants performed three 15-s test trials, with 45 s of rest between trials. A work-to-rest ratio of 1:3 was chosen, as this is considered optimal recovery time following short-duration, high-intensity tests [12]. The number of repetitions was recorded. The tested arm was defined as the arm maintaining the closed kinetic chain (CKC) position. The test was considered successfully completed if the participant maintained a flat back, kept the arm in the 90°−90° position, ensured the knees did not touch the floor, and kept the feet in the initial position throughout the test. The test was performed for both the dominant and non-dominant shoulders.

Fig. 1.

Upper limb rotation test position

Statistical analysis

The means and standard deviations were calculated for all dependent variables across participants. To assess the normality of the distribution for all measurements, the Kolmogorov–Smirnov test was initially performed. To evaluate the within-session reliability of the ULRT between trials on day 1 and day 2 intraclass correlation coefficients (ICC) two-way mixed model, with the absolute consent method and presentation of confidence interval 95% (CI 95%) were calculated. ICC values range from 0 to 1:1 = perfect reliability, 0.90–0.99 = very high reliability, 0.70–0.89 = high reliability, 0.50–0.69 = moderate reliability, 0.26–0.49 = low reliability, 0.00–0.25 = very low reliability or no reliability [29]. In addition, a Bland–Altman chart was plotted for each score from the ULRT to verify the absolute agreement between the first and second assessment, from the scatter plot between the difference of the two assessments and the average of the two evaluations. That way, it was possible to determine the error limits of agreement, outliers, and trends [3, 11]. To examine the absolute reliability of the ULRT, the standard error of measurement (SEM) and the minimal detectable change (MDC) were also calculated. The SEM was calculated using the formula SD x √(1-ICC), where SD represents the standard deviation of all participant scores [34]. The MDC was calculated using the formula SEM × 1.96 x √2 [34]. In addition to significance testing, effect sizes were calculated using Cohen’s d to evaluate the magnitude of observed differences. In this study, the calculated Cohen’s d values were 0.32 for the dominant extremity and 0.25 for the non-dominant extremity, both of which fall within the range of small effect sizes (d = 0.20–0.49) according to Cohen’s classification.

Results

The test–retest means and standard deviations of the ULRT are presented in Table 2. ULRT reliability analysis is summarized in Tables 3.

Table 2.

Mean and standard deviation scores for test (evaluation) and retest (revaluation) ULRT scores

ULRT scores (n = 51)	Test (X ± SD)	Retest (X ± SD)	Cohen d
Dominant extremity	12.37 (2.82)	13.24 (2.55)	0.32
Nondominant extremity	12.64 (2.35)	13.20 (2.21)	0.25

Abbreviations X mean, SD standard deviation

Table 3.

The Reliability Values of the Upper Limb Rotation Test

	ICC (95% CI)	SEM	MDC	p value
Dominant extremity	0.93 (0.88–0.96)	1.38	3.82	< 0.001*
Nondominant extremity	0.96 (0.93–0.98)	0.89	2.46	< 0.001*

Abbreviations CI confidence interval, ICC intra-rater correlation coefficient, SEM standard error measurement, MDC minimal detectable change

ULRT reliability

The intra-rater reliability of the test was found to be ICC = 0.93 for the dominant extremity and ICC = 0.96 for the non-dominant extremity (p < 0.001). The SEM value was found to be dominant extremity 1.38, nondominant extremity 0.89, while the MDC value was found to be dominant extremity 3.82, nondominant extremity 2.46. A lower SEM value indicates that the measurements are more reliable. The MDC represents the smallest detectable change between two measurements that can be identified by the instrument. Differences smaller than this value cannot be practically detected or considered statistically significant. In this study, the calculated Cohen’s d values were 0.32 for the dominant extremity and 0.25 for the non-dominant extremity, both of which fall within the range of small effect sizes (0.20–0.49). The small magnitude of effect indicates that there was no substantial difference between test and retest measurements, suggesting that the test produced consistent results across sessions. This finding reflects high internal consistency of the measurement process and implies that the learning effect was limited. Moreover, these results support the reusability of the test and indicate that the ULRT is a valid and reliable tool for repeated evaluations.

The agreement analysis between measurements

The agreement analysis between measurements of the scores (dominant extremity score, nondominant extremity score) can be observed in Figs. 2, 3 respectively.

Fig. 2.

Bland–Altman plots for intra rater test- retest dominant extremity score difference

Fig. 3.

Bland–Altman plots for intra rater test- retest nondominant extremity score difference

The Bland–Altman plots (Figs. 1 and 2) showed a uniform scatter of differences around the mean bias line with no obvious proportional bias (no funnel pattern), indicating homoscedasticity across the range of scores. A small negative bias was observed for both extremities (retest slightly higher than test), with only a few observations marginally outside the 95% limits of agreement, consistent with good absolute agreement and a potential learning effect.

Discussion

This study aimed to assess the reliability of the ULRT in adolescent male basketball players. Although the relative coefficients demonstrate excellent reliability values, the agreement analysis still exhibits fluctuations, which may be associated with individual differences and adaptation to the test. In summary, the findings of this study indicate that the ULRT demonstrates excellent reliability for assessing upper extremity rotation in a kinetic chain position among adolescent male basketball players. A learning effect may have occurred during the one-week interval between the test and retest sessions, as athletes may have become more familiar with the test procedure during the second administration compared to the first, potentially enhancing test–retest consistency. Additionally, adolescent athletes may demonstrate better adaptation to the test protocols due to their ongoing neuromuscular development, increased proprioceptive awareness, high motor learning capacity, and strong motivation for performance evaluation, particularly within the context of overhead sports. Consistent with our a priori hypothesis, the ULRT demonstrated excellent intra-rater reliability in adolescent male basketball players, supporting its use as a reliable tool for performance monitoring in this population.

In their study evaluating the test–retest reliability of the ULRT test in healthy adults, Decleve et al. reported that the test demonstrated high reliability within this population (ICC = 0.76 for the dominant extremity and ICC = 0.78 for the nondominant extremity) [7]. In our test–retest reliability study conducted with overhead athletes, including basketball, handball, and volleyball players, the ULRT test demonstrated high reliability in this athletic population (ICC = 0.80 for dominant extremity, ICC = 0.81 for nondominant extremity) [10]. The findings of the present study are consistent with previous research examining the test–retest reliability of the ULRT in other populations. Moreover, the results of this study demonstrated excellent reliability, further supporting the strong psychometric properties of the ULRT in this population. However, when relative reliability analysis (ICC) is used in reliability studies, the reported reliability values may be overestimated [6]. Therefore, incorporating methods that assess data variance and agreement between measurements is recommended as a complementary approach in reliability research. In this study, in addition to ICC values, Bland–Altman plots were also utilized. The use of Bland–Altman plots allowed for a more detailed examination of data variance and agreement models between measurements, aiming for a more comprehensive interpretation of the results. The horizontal line shown in the Bland–Altman plots for both the dominant and non-dominant extremities represent the mean difference between the test and retest measurements. In both plots, the proximity of the difference to zero indicates a low level of systematic error. The upper and lower dashed lines represent the limits of agreement within ± 1.96 SD. For both the dominant and non-dominant extremities, most of the data points fall within these limits,however, some points appear to lie outside. Notably, in the non-dominant extremity, the lower limit is slightly shifted downward compared to the dominant extremity, which may indicate greater measurement discrepancies in some individuals. Although the data for both the dominant and non-dominant extremities appear largely homogeneous, a few outliers at the upper and lower extremes are noticeable. This may indicate that some individuals exhibit greater variability between test and retest measurements. A comprehensive evaluation of the Bland–Altman plots for both the dominant and non-dominant extremities suggests that repeatability is generally good, with most data points falling within the limits of agreement. However, the presence of outliers in some athletes may indicate potential issues with measurement consistency. Additionally, greater variability can be observed in lower scores, which may be attributed to differences in athletes' performance levels. The differences between test–retest evaluations can be explained by increased familiarity and adaptation to the test after the initial trial. Despite the observed differences between test–retest assessments, both relative and absolute reliability, as well as agreement analysis, indicated that the test demonstrated high reliability.

SEM and MDC values assist clinicians in interpreting individual changes in performance. Clinically, the SEM reflects how much a score might vary due to random measurement error, even if the athlete's true performance has not changed. The MDC represents the smallest change that can be confidently considered real. If a test score exceeds the MDC, it likely indicates a true improvement or decline, rather than normal test variability. The SEM and MDC values obtained for the dominant and nondominant extremities in the present study provide important insights into the absolute reliability of the ULRT. SEM represents the within-subject measurement error, while MDC defines the smallest detectable change that exceeds this error and can therefore be considered a meaningful change [21]. In the present study, the SEM was calculated as 1.38 for the dominant extremity and 0.89 for the nondominant extremity, indicating lower measurement error and higher measurement precision for the nondominant side.

Similarly, the MDC values were 3.82 for the dominant extremity and 2.46 for the nondominant extremity, suggesting that smaller changes in ULRT performance can be considered meaningful for the nondominant extremity, whereas larger changes are required to be considered meaningful for the dominant extremity. This finding may be explained by the stabilizing role of the nondominant limb, which is particularly relevant in overhead athletes, as the nondominant extremity often plays a crucial role in postural control and proprioceptive feedback during throwing and overhead movements [17].

In this context, the SEM and MDC values reported in the present study offer a useful reference for distinguishing between real performance changes and variations attributable to measurement error during repeated assessments [16, 21]. For example, if a change in ULRT performance following an intervention does not exceed the MDC value, this change should not be interpreted as a clinically meaningful improvement but rather as a reflection of normal measurement variability. This highlights the importance of incorporating SEM and MDC thresholds when interpreting longitudinal changes in ULRT performance, particularly in the context of rehabilitation or performance monitoring in overhead athletes.

De Oliveira et al. evaluated the test–retest reliability of the Closed Kinetic Chain Upper Extremity Stability Test (CKCUEST) in adolescents and reported moderate to high reliability in this population ICC = 0.68 for average touches score, ICC = 0.68 for normalized score, ICC = 0.87 for power score) [6]. Decleve et al. also investigated the test–retest reliability of the Modified Closed Kinetic Chain Upper Extremity Stability Test (MCKCUEST) in adolescent volleyball and basketball players and reported good to excellent reliability in this athletic population (ICC = 0.93) [8]. De Oliveira et al. reported moderate reliability for the CKCUEST (ICC = 0.68) [6], while Decleve et al. reported very high reliability (ICC = 0.93) [8]. This discrepancy may be attributed to the fact that Decleve et al. investigated the reliability of the Modified Closed Kinetic Chain Upper Extremity Stability Test rather than the original CKCUEST. In the CKCUEST, the hands are positioned at a standardized distance of 91.44 cm apart [19], whereas in the MCKCUEST, the hands are placed at shoulder width [15]. The modified hand positioning, which adjusts the starting position to match the individual’s shoulder width, may facilitate better performance and improve test feasibility across participants with varying anthropometric characteristics [8]. As a result, the reliability values reported for the MCKCUEST may be higher.

Overhead movement is a dynamic activity that requires high velocity, extensive shoulder range of motion, and full-body involvement through the kinetic chain [25]. During overhead movement, the coordinated activation of multiple muscle groups and joints is required to ensure efficient force transmission through the kinetic chain [18, 24]. The trunk serves as a critical intermediary within the kinetic chain, facilitating the transfer of energy and force from the lower extremities to the upper extremities. Additionally, the rotational movement of the trunk enhances the force generated during overhead motion [20]. Considering the role of the kinetic chain in overhead throwing, its routine assessment may be important. The ULRT, CKCUEST, and UQYBT assess performance within the kinetic chain [10, 15, 17]. The Upper Limb Rotation Test (ULRT) incorporates both trunk rotation and throwing position of the shoulder into the assessment of the kinetic chain [7].

Strengths

The identification of ULRT as a reliable measurement tool in adolescent male basketball players is significant for its inclusion in the routine physical performance assessments of adolescent athletes. This would allow for the evaluation of various parameters influencing overhead throwing in a kinetic chain position, such as trunk rotation, trunk stabilization, and overhead throw-specific positioning.

Limitations

This study included only adolescent male basketball players. Due to anatomical and physiological differences, the reliability of ULRT in adolescent female basketball players remains unknown. Future studies should consider evaluating the validity and reliability of ULRT in adolescent female basketball players.

Conclusions

The ULRT is a clinician-friendly, easily administered, and practical test for assessing upper extremity rotation in a kinetic chain position. It can be reliably used to evaluate performance in adolescent male basketball players.

Authors’ contributions

Birgul D. Gultekinler: Conceptualization, Methodology, Data curation, Investigation, Writing – review & editing. Burak Bayram Kaya: Data curation. Ertuğrul Gelen: Review & editing. Volga Bayrakcı Tunay: Conceptualization, review & editing.

Funding

This research did not receive any specific grant from funding agencies in the public, commercial, or not- for-profit sectors.

Data availability

The datasets used and/or analyzed during the current study are available from the corresponding author on reasonable request.

Declarations

Ethics approval and consent to participate

This study received ethical approval from the Ethics Committee of Sakarya University of Applied Sciences (Approval date: 04.08.2023, Approval number: 34/14). Since the athletes in the study were adolescents, written informed consent was obtained from their parents. Before data collection, the researchers explained to the participants the purpose of the study, that voluntary participation was essential, that anonymity was guaranteed, and that their data would be used only for this study. Informed consent was taken from all participants. The participants were also allowed to withdraw from the study any time without stating any reason. The Declaration of Helsinki was adhered to throughout all phases of this study.

Consent for publication

Written informed consent was obtained from all participants (and their legal guardians) for the publication of their personal or clinical details and any accompanying images in this study.

Competing interests

The authors declare no competing interests.