Preseason shoulder rotational isokinetic strength and shoulder injuries in volleyball players

Abstract

BACKGROUND:

Rotator cuff weakness is considered an important risk factor for shoulder injuries in volleyball.

OBJECTIVE:

To evaluate association of shoulder preseason strength status with shoulder injury occurrence in subsequent season.

METHODS:

Volleyball players ( $N=$ 181; 99 men) from Slovenian 1 ${}^{\text{st}}$ and 2 ${}^{\text{nd}}$ national league volunteered to participate in this prospective cohort study. Preseason isokinetic testing of the shoulder was conducted at 60 ${}^{\circ}$ /s in the concentric mode of contraction over a RoM of 60 ${}^{\circ}$ with five repetitions of internal (IR) and external (ER) rotation. During the subsequent season the players reported shoulders injuries through a weekly questionnaire.

RESULTS:

During the season we have registered 14 (7.7%) shoulder injuries (10 in men). All injuries affected the dominant shoulder. There was significant preseason weakness of ER and lower ER/IR strength ratio in players with shoulder injury. Normal strength ratio ER/IR was a significant protective factor (Exp (B) $=$ 0.217, 95% C.I. 0.058–0.811) for shoulder injury occurrence when controlled for sex and previous injury.

CONCLUSIONS:

The inclusion of systematic strengthening of the external rotators of the shoulder is necessary, especially for male volleyball players, as part of preventive measures for the prevention of shoulder injuries.

Keywords

Isokinetic testing assymetry injury risk muscle strength

1. Introduction

Volleyball is one of the most interesting team sports that is played worldwide at all proficiency levels [1]. The basic characteristics of volleyball are different fast and explosive movements, ranging from rapid direction changes in the playing field to numerous vertical jumps and rapid overhead arm accelerations (swings) in the attack and/or the defense phase of the game. Although some matches can last up to 3 hours, volleyball is predominantly anaerobic sport [2], and in order to play at competitive level individuals must have certain physical characteristics in addition to the technical and tactical knowledge of volleyball. One of such characteristics is the strength of the muscle groups that are necessary for the execution of basic volleyball movements: vertical jump and overhead arm swings.

Overhead arm swing is the movement encountered in many sports (e.g. baseball, handball, water polo, tennis, badminton, javelin throw) and in volleyball it is necessary for spiking and serving. For completion of overhead arm swing all four joints of shoulder region (acromioclavicular, sternoclavicular, glenohumeral and scapulothoracic joint) are participating in a coordinated manner. However, the greatest stress is placed on the shoulder internal and external rotators (rotator cuff) that are responsible for the execution (concentrically) and the control (eccentrically) of the spike and studies have indicated that rotator cuff overloading may be a primary etiology of tendon pathology [3]. Therefore, it is not surprising that rotator cuff muscle weakness as diagnosed with isokinetic testing is an important modifiable factor associated with shoulder injuries in volleyball [4]. Additionally, concentric strength of the dominant shoulder (e.g. measured at 60 ${}^{\circ}$ /s) is significantly correlated with spike velocity in volleyball, indicating its relationship with sport performance [5]. To sum up, shoulder rotator cuff strength has important implications for health and performance of volleyball players.

Epidemiological data shows that shoulder dysfunction is present in 8% to 20% of volleyball players and in almost 50% of the cases the nature of shoulder problems is such that it interferes with player’s sport performance [6], and these problems are responsible for the longest absence from the training and competition (6.2 weeks on average) [7]. More shoulder problems is reported by players at playing positions where spiking is more frequent and more intense (e.g. outside hitters, opposites, and middle blockers) [6], and this could be explained by differences in strength and strength ratios that were reported among different playing positions [8].

There were several cross sectional studies in which rotator cuff strength was compared between volleyball players with and without shoulder problems. Stickley [9] has reported that rotator cuff strength ratios were more related to shoulder injury prevalence than the absolute strength. Decreased dominant external rotation, higher deficit of dominant eccentric external rotation peak moments, and higher dominant rotator fatigability were correlated with previous shoulder pain/injury in a study which used mixed volleyball/handball population [10]. Finally, our research group has reported a findings from the largest volleyball cohort so far (99 men, 82 women) [11]. We found that male volleyball players with history of shoulder problems had a lower external-rotation/internal-rotation strength ratio of the dominant shoulder and that female volleyball players at a higher skill level were 2.5 times more likely to have abnormal strength ratios.

To our knowledge there is only one prospective cohort study so far that investigated the true relationship between rotator cuff strength and shoulder problems in the sample of 66 volleyball player (34 men) [12]. The eccentric maximal strength developed by the internal and external rotators was found to represent a protective factor in volleyball players (respective odds ratios $=$ 0.946, $p=$ 0.01 and 0.94, $p=$ 0.05) and isokinetic shoulder testing was proposed for better identification of players at risk for shoulder problems. As prospective studies allow for better identification of true risk factors (e.g. unlike cross sectional studies in which we can only debate whether the findings are cause or consequence), we report herewith the findings from our study in which the large volleyball cohort described above [11] was prospectively followed over the period of one competitive season after the isokinetic shoulder testing was conducted during the preseason.

2. Methods

2.1 Participants

One hundred and eighty one ( $N=$ 181; 99 men) volleyball players from all Slovenian 1 ${}^{\text{st}}$ and 2 ${}^{\text{nd}}$ national league volunteered to participate in this prospective cohort study by signing a written informed consent. The study was organized with the support of the Slovenian Volleyball Federation that provided full participation of all clubs while this sample that met the inclusion criteria represented about 90% of the total volleyball population in Slovenia. The main inclusion criteria were: age $\geqslant$ 18 years, regular participation in volleyball training at least 3 times per week, no history of shoulder pain in the previous 3 months and absence of all general contraindications for isokinetic strength testing of the shoulder. National Medical Ethics Committee (no. 63/07/12) has approved the study.

2.2 Procedures

Each participant has completed an enrolment questionnaire about previous shoulder injuries, hand dominance, playing position, level of play, and hours of training. The D hand was defined as the hand used to serve or spike. Anthropometric measurements (body mass, body height, skinfolds) were performed using a stadiometer and scale (models 222 and 762, respectively; Seca Instruments Ltd, Hamburg, Germany) and Harpenden skinfold calipers (Holtain Ltd, Crosswell, Crymych, United Kingdom). From 7 skinfold measures, we calculated the body fat percentage using the Jacksons Pollock formula [13].

The same experienced examiner performed all the isokinetic testing procedures. Players from the same volleyball club were tested on the same day. A day prior to testing no practice was allowed. Each testing session started with a warm up consisting of 3 sets of 10 repetitions of shoulder internal and external rotation using elastic resistance band. All participants were given a detailed explanation about the testing procedure, which was also demonstrated on an independent subject not participating in the study prior to testing.

Testing was performed using Techno-Gym REV 9000 isokinetic dynamometer (TehcnoGym, SpA, Via G. Perticari 20, 47035 Gambet-Tola, Forli, Italy) in concentric-concentric mode at 60 ${}^{\circ}$ /s for shoulder external (ER) and internal (IR) rotators of the dominant (D) and non-dominant (ND) shoulder. Measurement was conducted in seated position with the arm at 90 ${}^{\circ}$ of abduction and the range of motion (RoM) extended from 90 ${}^{\circ}$ to 30 ${}^{\circ}$ of external rotation [14]. The participant was allowed to see the moment curve on the display and was given consistent verbal encouragement throughout the testing protocol. The details of the testing protocol are provided elsewhere [11]. The main outcome measure was the body weight normalized peak moment (PM) of the shoulder ER and IR in Newton-meters per kilogram bodyweight. The PM-based strength ratio: ER/IR was later calculated. Based on this ratio a new dichotomous variable (normal, abnormal) was introduced. An abnormal ER/IR ratio was defined as lying outside the range 0.60–0.75 according to the criteria set by Ellenbecker and Davies [15]. Additionally, a receiver operating curve (ROC) analysis was later performed to evaluate the ability of ER/IR ratio to discriminate between injured and uninjured players and find an optimal cut off value.

2.3 Statistical analyses

All data were analyzed using the IBM SPSS Software for Windows (version 25, SPSS Inc., Chicago, Illinois, USA). Categorical variables are displayed as numbers and percentages, and continuous variables are presented as means and standard deviations. All numeric variables were firstly checked for normality of distribution with Shapiro-Wilk’s test.

Analyses were initially performed separately for sex. Preseason isokinetic findings of the D and ND shoulder were compared between injured and uninjured players using a one-way analysis of variance, while the differences between dominant and non-dominant shoulder were evaluated using paired $t$ -test. A chi square test and binary logistic regression were used to calculate relative risk and odd ratio for shoulder injuries using chosen predictors.

3. Results

Basic characteristics of the volleyball players that participated in the study are presented in Table 1.

Table 1
Main characteristics of the volleyball players

	Men ( $N=$ 99)		Women ( $N=$ 82)
	Number	%	Number	%
Playing level
1st Division	54	54.5%	51	62.2%
2nd Division	45	45.5%	31	37.8%
Playing position
Opposite	17	17.2%	17	20.7%
Blocker	26	26.3%	19	23.2%
Left	10	10.1%	12	14.6%
Setter	19	19.2%	13	15.9%
Receiver	27	27.3%	21	25.6%
Hand dominance
Right	93	93.9%	77	93.9%
Left	6	6.1%	5	6.1%
Previous shoulder injury
Yes	9	9.1%	6	7.3%
	Mean	Std. Dev.	Mean	Std. Dev.
Age (years)	23	5	22	4
Body mass (kg)	84	10	68	9
Body height (cm)	189	7	175	7
Percentage body fat (%)	10.7	3.1	21.6	3.8
Hour of training/season (h)	413	141	311	141
Hour of play/season (h)	57	21	58	22

During the one competitive season we have registered 14 shoulder injuries (10 in male and 4 in female volleyball players). All injuries affected the dominant shoulder. In male players there were 8 new injuries and 2 re-injuries, while in female players there were 2 new injuries and 2 re-injuries. Previous injury was significantly related with shoulder injury incidence only in females ( $p=$ 0.01). However, there was no sex related difference in the incidence of shoulder injuries ( $\chi^{2}=$ 1.714, $p=$ 0.190).

Table 2

External-rotator and internal-rotator strength and strength ratios of the dominant (D) and nondominant (ND) shoulders (60 ${}^{\circ}$ /s) in volleyball players by sex and shoulder injury status

			Shoulder injury during season
Variable	Side	Normalized strength	MALES				FEMALES
			Yes		No		Yes		No
			Mean $\pm$ Std. Dev.		Mean $\pm$ Std. Dev.		Mean $\pm$ Std. Dev.		Mean $\pm$ Std. Dev.
Internal rotation	D	Nm/kg	0.95 ${}^{\text{a}}$	$\pm$ 0.13	0.87 ${}^{\text{a}}$	$\pm$ 0.13	0.56	$\pm$ 0.01	0.58 ${}^{\text{a}}$	$\pm$ 0.11
	ND	Nm/kg	0.82	$\pm$ 0.10	0.81	$\pm$ 0.12	0.45	$\pm$ 0.11	0.55	$\pm$ 0.09
External rotation	D	Nm/kg	0.45 ${}^{\text{b}}$	$\pm$ 0.14	0.54 ${}^{\text{a}}$	$\pm$ 0.11	0.37	$\pm$ 0.06	0.42 ${}^{\text{a}}$	$\pm$ 0.08
	ND	Nm/kg	0.49	$\pm$ 0.06	0.51	$\pm$ 0.09	0.36	$\pm$ 0.03	0.38	$\pm$ 0.07
Strength ratio	D		0.47 ${}^{\text{a,b}}$	$\pm$ 0.11	0.62	$\pm$ 0.13	0.67	$\pm$ 0.11	0.74	$\pm$ 0.15
	ND		0.60	$\pm$ 0.09	0.64	$\pm$ 0.11	0.85	$\pm$ 0.35	0.71	$\pm$ 0.14

D – dominant shoulder; ND – non-dominant shoulder; a – significant difference between D and ND shoulder (please note that in subgroup of injured players this is comparison of healthy and injured shoulder); b – significant differences between injured and uninjured shoulder.

Table 3

Multivariate logistic regression model for shoulder injury

Variables in the equation	B	df	Sig.	Exp (B)	Lower	Upper
					95% C.I. for Exp (B)
Sex	$-$ 0.220	1	0.740	0.802	0.218	2.951
Previous shoulder injury	$-$ 1.014	1	0.149	0.363	0.092	1.439
Strength ratio ER/IR of the dominant shoulder	$-$ 1.526	1	0.023	0.217	0.058	0.811
Constant	3.533	1	0.000	34.225

Preseason isokinetic strength findings of the shoulder (Table 2) in male players with subsequent shoulder injury indicate significant preseason weakness of ER (0.45 vs. 0.54; $F=$ 6.06, $p=$ 0.016) and lower ER/IR strength ratio (0.47 vs. 0.62; $F=$ 12.98, $p=$ 0.001) in comparison with players without the injury. No such findings were found in female players. The comparison of D and ND shoulder finding has shown that in uninjured players there were significant differences in IR and ER, with no differences in the strength ratios. On the other hand, in the injured male players, IR of the D shoulder was significantly stronger compared to its ND counterpart, resulting in significantly lower ER/IR strength ratio in the D shoulder.

There was a significant association between a decrease in the ER/IR ratio and shoulder injuries in the male players ( $\chi^{2}(1)=$ 8.53, $p=$ 0.005), but we have not noticed such association in the female players ( $\chi^{2}(1)=$ 0.26, $p=$ 0.506). The relative risk for shoulder injuries in male volleyball players with decreased strength ratio was 2.18 (95% C.I. 1.57–3.01) and was 13 times higher than in the group with normal strength ratio 0.171 (95% C.I. 0.03–1.10). When controlled for previous injury, our data shows that this is true only for new injuries ( $\chi^{2}(1)=$ 7.37, $p=$ 0.009).

In the logistic regression model (Table 3) we have used sex (men, women), previous injury (yes/no) and strength ratio ER/IR of the dominant shoulder (normal, abnormal) to predict injury occurrence. The model was statistically significant ( $\chi^{2}(3)=$ 11.79, $p=$ 0.008), and normal strength ratio ER/IR was a significant protective factor (Exp (B) $=$ 0.217, 95% C.I. 0.058–0.811) for shoulder injury occurrence when controlled for sex and previous injury. Finally, ROC analysis (Fig. 1) showed a significant accuracy of dominant shoulder strength ratio ER/IR in discriminating between healthy and injured players but only in men (AUC 0.832, 95% CI 0.694 to 0.970, $p<$ 0.0001). The ER/IR cut-off level associated with belonging to healthy players group is 0.5379 (where sensitivity is 75%, and false positive rate is 10%).

Figure 1.

The receiver operating curve for classifying healthy players using the strength ratio between external and internal rotators of the dominant shoulder.

4. Discussion

The main findings of our study have shown that dominant IR and ER muscles in uninjured male and female volleyball players, are significantly stronger compared to those of the non-dominant shoulder, resulting in symmetrical ER/IR strength ratio of both shoulders. However, in the injured male volleyball players significantly stronger IR of dominant shoulder, was not followed by concomitantly stronger ER, resulting in a significantly lower ER/IR ratio of the dominant shoulder during preseason testing. Additionally, the logistic regression model indicates that the normal strength ratio represents an important protective factor for shoulder injuries in volleyball when controlled for sex and previous shoulder injury. When preseason results were compared between injured and uninjured players, it was shown that injured players had significantly weaker external rotators ( $p=$ 0.016) and consequently significantly lower ER/IR strength ratio ( $p=$ 0.001).

Compared to the only prospective study about rotator cuff strength and shoulder injury so far [12] which concluded that the eccentric maximal strength developed by the IR and ER groups was found to represent a protective factor in the volleyball players (respective odds ratios $=$ 0.946, $p=$ 0.01 and 0.94, $p=$ 0.05), we may add that a concentric ER/IR strength ratio above 0.60 is also a protective factor in volleyball players (odd ratio 0.217 [0.058–0.811]; $p=$ 0.023). Additionally, we have showed that cut off point of ER/IR strength ratio in men is 0.54, indicating that previously reported normal range, was slightly higher then the statistically calculated one. This means that we can expect internal rotators dominance in healthy volleyball players that results in slightly lower ER/IR ratio than one proposed by Ellenbecker and Davies [15]. The fact that these players are without shoulder injury despite such ER/IR ratio could probably be explained by multifactorial adaptations to shoulder loading during career. There may be other neuro-muscular factors that contribute to the stability and health status of the shoulder such as the dynamic control ratio (ERecc/IRcon), rate of force development, joint position sense and plyometric capacities. Our finding are in concordance with Stickley [9] who reported that rotator cuff strength ratios are more related to shoulder injury prevalence than the absolute strength. It is important to notice that our results suggest that the ER/IR ratio is important in new shoulder injuries (e.g. in players without history of shoulder problems), which means that rotator cuff strength could indeed be the primary etiological factor in the development of shoulder problems in volleyball. Although, the lack of eccentric testing in our study could be considered as limitation, we believe that when dealing with large group of participants it is better to conduct testing at low concentric velocity (such as 60 ${}^{\circ}$ /s) as compliance is high, and it is not time consuming and is not related with some reproducibility issues like eccentric testing [16]. Single velocity concentric testing could be considered for screening and more thorough test with additional velocities and/or contraction types could be reserved for subjects in which the screening process highlights some shoulder problems (e.g. weakness of external rotators as was shown in our study).

Despite the noticeable differences between men and women, we have not been able to show that the occurrence of shoulder joint injuries varies between the sexes, although the absolute number of injured male volleyball players was 2.5 times higher than that of female volleyball players. This is in agreement with previous findings that excluded sex as an independent risk factor for shoulder injuries in volleyball [6].

The difference in strength and strength ratios findings between the sexes could otherwise be explained by (a) the number of hand swings during a single point (in women the point lasts longer, and there is usually more ball exchange between the sides, which also means the need for more swings in women compared to men) [17]; (b) there could be differences in the amount and specificity of strength training between the sexes (Table 1; 413 vs 311 h of training per year in men vs women, respectively), and (c) there could be sex-related differences in the biomechanical execution of the spike. For example, during maximum shoulder external rotation, potential energy is also stored in trunk rotation [6] and female volleyball players could be weaker in this segment and therefore rely more on the shoulder girdle than on the torso rotators in the desire for the best possible quality of offensive spikes.

This last hypothesis could be partly supported by the fact that the values of the ER/IR ratio actually differed between the sexes. In uninjured volleyball players these values were between 0.60 and 0.65, and in uninjured female players between 0.70 and 0.78. This was also shown by the Forthomme et al. [12] where ER/IR strength ratio for men and women was 0.68 and 0.80, respectively. Even studies on adolescent volleyball players [18] have shown that strength ratio is higher in girls then in boys. Compared with values from previous research, the classical references [19, 20] appear to be sufficient for a protective effect in men, while somewhat more recent guidelines [15], appear to be more suitable for women. The higher strength ratio in females actually supports our hypothesis that women could be relying more on shoulder girdle for spiking than on the trunk rotation.

5. Conclusion

Based on our findings, we conclude that the inclusion of systematic strengthening of the external rotators of the shoulder is necessary, especially for male volleyball players, as part of measures for the prevention of shoulder injuries. While strong internal rotators are necessary for optimal performance, strong external rotators are obviously necessary for optimal health of the shoulder in volleyball. Strength and conditioning coaches should therefore pay more attention to match strong internal rotators with strong external rotators.

Author contributions

CONCEPTION: VH, ED, TS

PERFORMANCE OF WORK: PP, AH

INTERPRETATION OR ANALYSIS OF DATA: VH, TS

PREPARATION OF THE MANUSCRIPT: VH

REVISION FOR IMPORTANT INTELLECTUAL CONTENT: VH, ED, PP, AH, TS

SUPERVISION: ED

Ethical considerations

National Medical Ethics Committee (no. 63/07/12) has approved the study on July 31 ${}^{\text{st}}$ , 2012. All participants signed a written informed consent.

Funding

The study was supported by Slovenian Research Agency, grant number P5-0147.

Footnotes

Acknowledgments

The authors have no acknowledgments.

Conflict of interest

All authors declare no conflict of interest. Given his role as an Editorial Board Member, Vedran Hadžić had no involvement nor access to information regarding the peer review of this article.

References

Reeser

Verhagen

Briner

Askeland

Bahr

. Strategies for the prevention of volleyball related injuries. Br J Sports Med. 2006; 40(7): 594-600. discussion 599-600.

Popadic Gacesa

Barak

Grujic

. Maximal anaerobic power test in athletes of different sport disciplines. J Strength Cond Res. 2009; 23(3): 751-5.

Klich

Pietraszewski

Zago

Galli

Lovecchio

Kawczynski

. Ultrasonographic and myotonometric evaluation of the shoulder girdle after an isokinetic muscle fatigue protocol. J Sport Rehabil. 2020; 29(8): 1047-52.

Tooth

Gofflot

Schwartz

Croisier

Beaudart

Bruyere

, et al. Risk factors of overuse shoulder injuries in overhead athletes: a systematic review. Sports Health. 2020; 12(5): 478-87.

Forthomme

Croisier

Ciccarone

Crielaard

Cloes

. Factors correlated with volleyball spike velocity. Am J Sports Med. 2005; 33(10): 1513-9.

Reeser

Joy

Porucznik

Berg

Colliver

Willick

. Risk factors for volleyball-related shoulder pain and dysfunction. PM R. 2010; 2(1): 27-36.

Verhagen

Van der Beek

Bouter

Bahr

Van Mechelen

. A one season prospective cohort study of volleyball injuries. British Journal of Sports Medicine. 2004; 38(4): 477-81.

Kim

Park

Kuo

Park

. Isokinetic performance of shoulder external and internal rotators of professional volleyball athletes by different positions. Sci Rep. 2020; 10(1): 8706.

Stickley

Hetzler

Freemyer

Kimura

. Isokinetic peak torque ratios and shoulder injury history in adolescent female volleyball athletes. J Athl Train. 2008; 43(6): 571-7.

10.

Tonin

Strazar

Burger

Vidmar

. Adaptive changes in the dominant shoulders of female professional overhead athletes: mutual association and relation to shoulder injury. Int J Rehabil Res. 2013; 36(3): 228-35.

11.

Hadzic

Sattler

Veselko

Markovic

Dervisevic

. Strength asymmetry of the shoulders in elite volleyball players. J Athl Train. 2014; 49(3): 338-44.

12.

Forthomme

Wieczorek

Frisch

Crielaard

Croisier

. Shoulder pain among high-level volleyball players and preseason features. Med Sci Sports Exerc. 2013; 45(10): 1852-60.

13.

Pollock

Laughridge

Coleman

Linnerud

Jackson

. Prediction of body density in young and middle-aged women. J Appl Physiol. 1975; 38(4): 745-9.

14.

Hadzic

Ursej

Kalc

Dervisevic

. Reproducibility of shoulder short range of motion in isokinetic and isometric strength testing. Journal of Exercise Science & Fitness. 2012; 10(2): 83-9.

15.

Ellenbecker

Davies

. The application of isokinetics in testing and rehabilitation of the shoulder complex. J Athl Train. 2000; 35(3): 338-50.

16.

Malerba

Adam

Harris

Krebs

. Reliability of dynamic and isometric testing of shoulder external and internal rotators. J Orthop Sports Phys Ther. 1993; 18(4): 543-52.

17.

Kountouris

Drikos

Aggelonidis

Laios

Kyprianou

. Evidence for differences in men’s and women’s volleyball games based on skills effectiveness in four consecutive olympic tournaments. Comprehensive Psychology. 2015; 4: 30.50.CP.4.9.

18.

Wang

H-K

Juang

L-G

Lin

J-J

Wang

T-G

Jan

M-H

. Isokinetic performance and shoulder mobility in Taiwanese elite junior volleyball players. Isokinetics and Exercise Science. 2004; 12: 135-41.

19.

Ivey

, Jr. Calhoun

Rusche

Bierschenk

. Isokinetic testing of shoulder strength: normal values. Arch Phys Med Rehabil. 1985; 66(6): 384-6.

20.

Shklar

Dvir

. Isokinetic strength relationships in shoulder muscles. Clin Biomech (Bristol, Avon). 1995; 10(7): 369-73.

Preseason shoulder rotational isokinetic strength and shoulder injuries in volleyball players

Abstract

BACKGROUND:

OBJECTIVE:

METHODS:

RESULTS:

CONCLUSIONS:

Keywords

1. Introduction

2. Methods

2.1 Participants

2.2 Procedures

2.3 Statistical analyses

3. Results

Table 1 Main characteristics of the volleyball players

5. Conclusion

Author contributions

Ethical considerations

Funding

Footnotes

Acknowledgments

Conflict of interest

References

Table 1
Main characteristics of the volleyball players