Key words
arm health - injury prevention - range of motion - strength
Introduction
Youth throwing athletes often participate in their sport while experiencing pain
[1]
[2]. As pain
may be a precursor to acute and/or chronic injury [1], throwing-related upper extremity pain has understandably received
considerable attention. Prior research has identified factors such as height,
weight/mass, joint range of motion (ROM), throwing workload, balance, and
biomechanics as important contributors to upper extremity pain [3]
[4]
[5]
[6]
[7]. However, an exploration of risk factors associated
with pain across throwing sports is still lacking.
The throwing skill is fundamental to baseball and softball athletes and requires
precise timing and sequencing of body segments. If motion progresses efficiently
from the legs, through the trunk, to the throwing arm and hand before ball release
[8], the throwing athlete will then be able to
deliver a ball at high speeds without added shear stress on the throwing shoulder
girdle and elbow [9]. Moreover, any restrictions
along this linked segmental chain, i. e. at the hips, can negatively impact
performance [10].
In baseball, decreased hip internal rotation (IR) of the stride leg (front leg) has
been associated with elbow and shoulder pain in youth (age: 9–12) players
[11]. Further, for adolescent baseball players,
limited hip flexion of the stride (front) and stance (back) leg, as well as
decreased hip IR ROM (seated 90°) of each leg, was found in a group with
pain compared to those who were pain-free [12].
Hamano et al. reported decreased hip external rotation (ER) ROM of the stance leg in
high school players who experienced a shoulder or elbow injury in the past 3 years
[13], and the authors later reported restrictions
in bilateral hip flexion ROM and stance leg hip IR in younger players (age:
7–14) when comparing upper limb injured and non-injured groups [13]. A study by Li et al. reported that professional
players who suffered an in-season injury to the hip region had a decreased total hip
arc ROM (IR+ER) on both sides compared to non-injured players. Although
there are some initial indications in baseball, not all studies have found clinical
differences in hip rotational ROM measures between players with and without upper
extremity pain [14]
[15].
In softball, however, the relationship between hip ROM and pain is less clear. As hip
function has been shown to play an important role in throwing mechanics and
performance [16]
[17]
[18]
[19], numerous studies have been conducted in collegiate [20]
[21], high school
[21] and youth [22] populations to gain further understanding. However, these studies
have only provided descriptive data for hip joint ROM. Whether there are differences
in hip ROM between groups of individuals with or without pain has not been fully
established. To date, only findings of decreased stance hip ER ROM in a group of
collegiate level softball pitchers experiencing upper extremity pain have been
reported [23]. Therefore, a comparison of hip ROM
measures between different pain status groups is needed.
An alternative way to assess hip function is via strength testing. Harding et al.
demonstrated that hip IR and abduction strength were correlated with years of
competitive play and number of months played in the last year for baseball pitchers
[24]; while Zeppieri et. al. reported decreases
in hip abduction strength in baseball pitchers over the course of a season [25]. Furthermore, stance hip extension strength has
been positively correlated with hip-shoulder separation angle during the pitch, and
both stride and stance hip abduction strength has been positively correlated with
elbow valgus torque [26]. Increased rotational forces
created by increased hip strength measures may be a risk factor for injury, if the
result is increased forces on the upper extremity [27]. An investigation into the effects of hip ROM on the kinetic chain
reported an association between increased drive hip ER isometric strength and
increased net energy outflow among throwing-arm segments [19].
As literature suggests, factors pertaining to the kinetic chain or kinematic
sequencing when throwing can lead to injury [28]
[29]. To date, several studies have compared functional
differences between baseball and softball, and predictors of injury [30]
[31]
[32]. However, limited data exist evaluating group
differences between pain status (upper extremity: pain/no pain), and sport
(baseball and softball), for hip rotational ROM and isometric strength. Therefore,
the purpose of this study is to investigate stance and stride hip range of motion
and isometric strength in youth baseball and softball athletes. Additionally, the
authors further sought to explore whether there were any significant differences in
stance and stride hip IR and ER ROM and isometric strength measures based on pain
status and sport. We hypothesized lower hip ROM and greater strength in baseball and
softball players experiencing pain. In regard to sport, we hypothesized that
softball players would have greater hip ROM than baseball players but less hip
strength.
Materials and Methods
All testing procedures were approved by the University’s Institutional Review
Board. All consent and assent declarations were obtained prior to any testing.
Participants were recruited largely based on the lab’s promotion (emails and
media) and interaction with parents, players, and coaches among local communities
(primarily in the Southeast). Further, interested participants may obtain clinical
and biomechanical assessments. In this regard, non-random sampling techniques
(convenience/voluntary) were utilized. Study inclusion criteria required
participants to be under 18 years of age and actively competing on a team baseball
or softball roster at any position (pitcher, catcher, positional). Additionally,
participants had to be surgery free for the past six months. A total of 354 throwing
athletes [9 to 17 years old] were examined at random times throughout the year.
Participants reported to the laboratory, where physical assessments were
performed.
Hip ROM and Strength measurements
Hip rotational ROM was measured passively with the participant in a seated
position. The participant’s knees were flexed to 90° allowing
the legs to comfortably hang off the edge of the table with their hands resting
comfortably on the table to assist with trunk stabilization [16]
[33]
[34]
[35]. The hip
was in 90° of flexion and a towel was placed under the femur. A digital
inclinometer was aligned along the soft tissue contour of the tibia. The femur
was supported by the examiner to eliminate accessory motion. The hip was
passively rotated until capsular end-feel was achieved. At the point of a firm
capsular end-feel, while not permitting any accessory hip movement (hip hiking),
the ROM measurement was recorded [16]
[34]
[36].
Intra-rater reliability for ROM (ICC(3,k) of 0.92 to 0.95 for all
measurements) were reported. Hip total arc ROM was determined as the sum of
internal and external ROM respective to each hip. Hip rotational strength was
measured with the participant positioned as per the ROM measures. A handheld
dynamometer (Lafayette Instruments, Lafayette, IN) was placed three inches
proximal to the medial malleolus for ER and three inches proximal to the lateral
malleolus for IR. The participant was instructed to maximally
internally/externally rotate stance and stride limb hips against the
applied resistance of the examiner for 3 seconds. Two maximal effort
isometric “make” tests against the dynamometer were performed
and the peak value was taken for analysis. The investigator visually monitored
the participants for compensations at the trunk or hip. The stance leg hip
represented the side ipsilateral to the throwing arm and the stride leg hip
represented the side contralateral to the throwing arm.
Based on a health history questionnaire administered to players prior to
collection procedures, players were grouped into pain and no-pain groups based
on their self-reported responses to the question: “Do you currently
experience any pain/discomfort in your upper extremity, specifically
your throwing side?” A total of 166 baseball players (upper extremity
pain=23 [13.9%]; no upper extremity pain=143
[86.1%]), and 188 softball players (upper extremity pain=25
players [13.3%]; no upper extremity pain=163 [86.7%])
comprised the original set of data. Upper extremity pain groups consisted of
participants experiencing pain at any of the throwing-arm joints and/or
segments. Grouping assignments of participants experiencing upper extremity pain
was performed post collection.
Statistical analysis
With randomized controlled trials being the gold standard in testing causal
inference and based on the ethical nature of collecting measures on those with
UE pain, propensity score matching (PSM) in IBM SPSS (IBM Corp. Released 2021.
IBM SPSS Statistics for Windows, Version 28.0. Armonk, NY) was used to create
balance between the pain and no pain groups of players for each sport. This
removed potential confounders on hip ROM and strength measures, and pain
status.[37]
[38]
[39] In order to perform PSM,
participants with missing data were removed. This included eight baseball
players (pain=1; no pain=7) and one softball player (no pain).
The remaining 158 baseball players (pain=22 [13.9%]; no
pain=136 [86.1%]) and 187 softball players (pain=25
[13.4%]; no pain=162 [86.6%]) were matched based on
their age, height, mass, and sport using a match tolerance of 0.2, sampling
without replacement. A total sample of 94 players was used for analysis, 44 of
which were baseball players (age: 13±2 years (range: 9–17
years), height: 165.2±13.0 cm, mass:
58.5±13.4 kg) and 50 who were softball players (age:
13±2 years (range: 10–17 years), height:
160.9±11.2 cm, mass: 62.7±17.9 kg). Participant
demographics according to sport and pain status can be found in [Table 1].
Table 1 Descriptive statistics.
|
Pain Status (n)
|
Height (cm)
|
Mass (kg)
|
Age (y)
|
|
BB
|
No (22)
|
165.8±14.5
|
59.7±14.7
|
13±2
|
|
Yes (22)
|
164.6±11.7
|
57.3±12.2
|
13±2
|
|
SB
|
No (25)
|
159.1±12.0
|
62.0±17.0
|
13±2
|
|
Yes (25)
|
162.7±10.3
|
63.4±19.1
|
14±2
|
BB=baseball; SB=softball.
Three separate two-way multivariate analysis of variance (MANOVA) tests were
performed in IBM SPSS (IBM Corp, Version 28.0. Armonk, NY) to determine
differences in 1) hip IR/ER ROM, 2) total arc HIP ROM and 3) hip
IR/ER strength measures based on pain status and sport. Total arc HIP
ROM was tested in a separate analysis due to less theoretical/empirical
support than other hip IR/ER ROM measures on upper extremity
pain/injury. There was a linear relationship between the dependent
variables, as assessed by scatterplots, and no evidence of multicollinearity, as
assessed by Pearson correlation (r<0.9). A total of 12 univariate
outliers were observed upon inspection of boxplots for ROM and strength data
that were grouped according to pain status and sport. The univariate outliers
were left in the analysis based on inspection of the data and no reason to
reject them.
All hip ROM measures grouped by pain status and sport data were normally
distributed as assessed by Shapiro-Wilk’s test (p>0.05),
except for stride hip IR ROM (p=.043) in the baseball no-pain
group, which showed negative skewness; stance hip ER ROM (p=.041)
in the softball pain group, which appeared platykurtic; and stance hip total arc
(p=.033) in the softball pain group, which showed negative
skewness. Hip strength measures grouped by pain status and sport were normally
distributed (p>0.05) apart from stride hip ER isometric strength
(p=.015) in the baseball pain group, which showed positive
skewness.
Results
A two-way MANOVA was run with two independent variables – pain status and
sport – and four dependent hip rotation ROM variables (seated stance leg
IR/ER and seated stride leg IR/ER) to understand whether there were
differences in throwing athletes’ hip ROM based on pain status and sport.
Descriptive data for hip ROM by sport and pain status can be found in [Table 2].
Table 2 Descriptive statistics for hip range of motion
(º) by sport & pain status.
|
Pain Status (n)
|
Stance Hip IR
|
Stance Hip ER
|
Stride Hip IR
|
Stride Hip ER
|
|
BB
|
No (22)
|
36.8±9.2
|
41.6±5.8
|
34.2±9.5
|
40.3±7.3
|
|
Yes (22)
|
35.2±9.1
|
43.9±8.2
|
35.8±6.6
|
45.6±8.7
|
|
SB
|
No (25)
|
38.7±11.3
|
45.0±8.5
|
36.0±10.4
|
47.2±8.3
|
|
Yes (25)
|
36.5±9.5
|
45.0±10.3
|
34.6±9.0
|
45.9±10.5
|
For hip ROM measures there was homogeneity of covariance, as assessed by
Box’s M test (Box’s M=38.948, p=.216). Equal
variance was assumed for all ROM measures via Levene’s Test of Homogeneity
of Variance (p>0.05), aside from stance hip ER ROM
(p<.05); though group sizes were similar, and adjustment to the alpha
level was not implemented based on the reported outcome. The interaction effect
between pain status and sport on the combined dependent hip ROM measures was not
statistically significant (F(4, 87)=2.089, p=0.089,
Wilks’ Λ=0.912, partial
η2+=0.088). The main effect of pain status on the
combined dependent hip ROM measures was not statistically significant (F(4,
87)=1.470, p=0.218, Wilks’ Λ=0.937,
partial η2= =0.063). The main effect of
sport on the combined dependent hip ROM measures was not statistically significant
(F(4, 87)=0.965, p=0.431, Wilks’
Λ=0.958, partial η 2=0.042). [Fig. 1] displays separate plots for each dependent
variable based on the two independent variables.
Fig. 1 Estimated marginal means of stance and stride hip internal and
external range of motion. Note: Error bars are 95% confidence
intervals. BB=baseball; SB=softball; IR=internal
rotation; ER=external rotation.
A second two-way MANOVA was run with two independent variables – pain status
and sport – and two dependent hip total arc rotation variables (seated
stance leg total arc and seated stride leg total arc) to understand whether there
were differences in throwing athletes’ total hip arc ROM based on pain
status and sport. Descriptive data for total hip arc by sport and pain status can be
found in [Table 3].
Table 3 Descriptive statistics for hip total arc range of
motion (º) by sport & pain status.
|
Pain Status (n)
|
Stance Hip Total Arc
|
Stride Hip Total Arc
|
|
BB
|
No (22)
|
78.5±12.2
|
74.6±11.6
|
|
Yes (22)
|
79.1±15.1
|
81.4±14.0
|
|
SB
|
No (25)
|
83.7±14.6
|
83.2±12.9
|
|
Yes (25)
|
81.5±16.0
|
80.4±15.2
|
Note: BB=baseball; SB=softball.
For total arc hip ROM measures there was homogeneity of covariance, as assessed by
Box’s M test (Box’s M=8.243, p=0.543). Equal
variance was assumed for all ROM measures via Levene’s Test of Homogeneity
of Variance (p>0.05). The interaction effect between pain status and
sport on the combined dependent total arc hip ROM measures was statistically
significant (F(2, 89)=4.079, p=0.020, Wilks’
Λ=0.916, partial η2+=0.084).
To determine which outcomes differed based on the interaction, we used two factorial
ANOVAs. For stance hip total arc ROM there was no statistically significant
interaction between pain status and sport (F(1,90)=0.210,
p=0.648, partial η2=0.002. An analysis of
main effects revealed no statistically significant main effects for pain status
(F(1,90)=0.0.60, p=.807, partial
η2 =0.001) or sport
(F(1,90)=1.59, p=.210, partial
η2 =0.017) for stance hip total arc ROM. For
stride hip total arc ROM there was no statistically significant interaction between
pain status and sport (F(1,90)=2.942, p=0.09, partial
η2 =0.032. An analysis of main effects
revealed no statistically significant main effects for pain status
(F(1,90)=0.538, p=0.465, partial
η2 =0.006) or sport
(F(1,90)=1.876, p=0.174, partial
η2 =0.020) for stride hip total arc ROM.
[Fig. 2] displays separate plots for each
dependent variable based on the two independent variables.
Fig. 2 Estimated marginal means of stance and stride hip total arc
range of motion, pain status by sport. Note: Error bars are 95%
confidence intervals. BB=baseball; SB=softball; Hip Total
Arc=internal rotation+external rotation.
A third two-way MANOVA was run with the independent variables of pain status and
sport and four dependent hip rotation isometric strength variables (seated stance
leg IR/ER and seated stride leg IR/ER) to understand if there were differences in
throwing athletes’ hip rotation isometric strength measures based on pain
status and sport. Descriptive data for hip isometric strength by sport and pain
status can be found in [Table 4].
Table 4 Descriptive statistics for hip isometric strength (N)
by sport & pain status.
|
Pain Status (N)
|
Stance Hip IR
|
Stance Hip ER
|
Stride Hip IR
|
Stride Hip ER
|
|
BB
|
No (22)
|
129.0±56.8
|
119.0±42.7
|
132.3±49.4
|
118.6±50.6
|
|
Yes (22)
|
129.1±34.4
|
125.6±33.0
|
140.8±43.9
|
110.5±34.9
|
|
SB
|
No (25)
|
132.8±53.6
|
110.0±36.3
|
134.3±48.1
|
113.3±32.9
|
|
Yes (25)
|
116.4±42.9
|
102.9±36.5
|
117.2±42.7
|
101.3±33.1
|
Equal variance was assumed for all ROM measures as assessed by Levene’s Test
of Homogeneity of Variance (p>0.05) except for stride hip ER strength
(p<0.05), though group sizes were again similar, adjustment of
the alpha level was not performed based on the reported outcome. The interaction
effect between pain status and sport on the combined dependent hip strength measures
was not statistically significant (F(4, 87)=0.971,
p=0.428, Wilks’ Λ=0.957, partial
η2 =0.043). The main effect of pain status on
the combined dependent hip strength measures was not statistically significant
(F(4, 87)=1.195, p=0.319, Wilks’
Λ=0.948, partial η2 =0.052).
Similarly, the main effect of sport on the combined dependent hip strength measures
was not statistically significant (F(4, 87)=1.806,
p=0.135, Wilks’ Λ=0.923, partial
η2 = 0.077). [Fig. 3] displays separate plots for each dependent
variable based on the two independent variables.
Fig. 3 Estimated marginal means of stance and stride hip internal and
external isometric strength. Note: Error bars are 95% confidence
intervals. BB=baseball; SB=softball; IR=internal
rotation; ER=external rotation.
Discussion
We hypothesized lower hip ROM and greater strength in baseball and softball players
experiencing pain. Further, we also hypothesized that softball players would have
greater hip ROM than baseball players but less hip strength. The current study
revealed that there were no significant interactions based on pain status and sport
on hip IR/ER ROM measures or significant main effects for pain status or sport on
hip measures included in this study when groups were matched for age, height, and
mass, and sport. Our hypothesis was not supported by these outcomes which expected
hip characteristics to differ between pain status and sport, or at least by pain
status or sport alone.
This is the first study to compare hip characteristics of youth throwing
athletes’ reporting upper extremity pain across throwing sports. A single
cross-sectional study has already compared passive hip ROM across multiple sports or
sex [40], finding that females had greater hip IR ROM
than males, independent of sport. Interestingly, ROM were largely similar across
sports, except for cross country athletes whose hip ROM differed from
baseball/softball, tennis, basketball, and soccer athletes. While the
current study’s data agree with the trend of females (softball) having
increased mean hip ROM values than men (baseball) – regardless of pain
status – no significant differences were found between sports in our
study.
The current data also revealed no difference in hip IR ROM for the stride leg in
youth baseball throwing athletes reporting pain (pain: 35.8±6.6º vs.
no pain, 34.2±9.5º). These findings differ from previously reported
findings of internal rotation of the stride hip when flexed 90 degrees between pain
and pain-free groups [11]. Sekiguchi and colleagues
[11] found that youth baseball players (9 to 12
years old) with elbow or shoulder pain had significant restrictions in hip IR ROM of
the stride leg compared to those without pain (35.8º [pain] vs. 43.7
º [no pain],p=0.022). Saito et al. [12] observed stride leg hip IR values of 36.4º in baseball
athletes reporting elbow pain vs. 46.9º in players without pain.
Nevertheless, the results of this study do replicate non-significant differences
between pain and pain-free groups in drive leg hip IR/ER ROM and stride leg
hip ER ROM [11].
It is possible these differences are due to differing ROM measurement techniques used
for hip IR and ER between studies. Where hip IR ROM testing in the aforementioned
studies was done in a supine position with the hip flexed to 90 degrees – a
method that has been shown to have higher inter-rater and intra-rater reliabilities
[41] – hip rotational ROM and strength in
our study was measured passively with the participant in a seated position, similar
to Martin et al., who evaluated youth softball athletes’ hip ROM in a seated
position and found no difference in hip ROM or strength between athletes reporting
upper extremity pain vs. no pain [42]. It may,
therefore, be necessary to conduct more research to clarify the relationship of hip
measures and upper extremity pain in throwing athletes measured seated vs. supine
(or prone).
The repetitive nature of throwing sports can result in hip ROM and strength
adaptations [25]
[43]
[44]. Chan et al. described decreases
in hip ROM and increases in shoulder ROM for a professional baseball population over
the course of season [44], while others have reported
decreases in collegiate baseball players’ stance and stride hip passive ROM,
and isometric hip abduction strength in the post-season compared to pre-season [25]
[43]. Similarly,
bilateral decreases of passive hip ROM, and hip abduction strength, were found
across a competitive season of softball. Considering these findings and prior work
regarding hip tightness [45]
[46] limited pre-season hip ROM may be considered as a risk factor for
shoulder/elbow pain [45]
[46]. We did not evaluate for the effect of pre-season
vs. in-season timing on hip measures in our study and agree this should be explored
further. While not significant at the univariate level of testing, the means for
stride hip total arc ROM in softball players without pain
(M±SD=83.2±12.9) were relatively higher than baseball
players without pain (M±SD=74.6±11.6). These trends in
stride hip total arc ROM may be clinically meaningful in athletes without pain as
changes in stride hip total arc ROM may occur over the course of season(s),
potentially related to arm-pain.
Recent data in both baseball and softball suggest a relationship between hip strength
and injury risk in throwing athletes. Albiero et al. [26] reported that elbow varus torque was found to be strongly correlated
with both stride and stance hip strength in adolescent baseball pitchers.
Additionally, it has been documented that collegiate pitchers reporting upper
extremity pain present with decreased hip ER ROM on the throwing side (stance) and
reductions in hip IR (throwing side) isometric strength and hip ER (glove side)
isometric strength [23]. It is unclear why the
current study did not reveal a relationship between pain and isometric hip strength.
Hip strength increases with age and it is possible the relationship between strength
and upper extremity pain does not reach significance until athletes reach older
adolescence.[47]
Considered a strength to the study, propensity score matching was employed to obtain
age, height, weight, and sport matched-control groups. Control matching has been
used in prior work [48]
[49] to account for potential confounders that may be associated with risk
factors for pain/injury [7]
[50]
[51]. As the
intention of this study was to investigate the presence of upper extremity pain and
sport on hip characteristics, we controlled for age and anthropometric parameters.
The pain group comprised athletes experiencing upper extremity pain at the time of
testing, which allowed us to compare the presence of pain directly to those without
pain. However, pain scales may be a viable method for examining further links
between pain and hip characteristics.
Due to the small sample size (N=44 baseball and N=50 softball) we
were unable to compare pitchers vs. positional players in our analysis. Though,
other work has not identified differences in hip ROM between softball pitchers and
positional players [22] or at least players who
identified their primary position as a pitcher or positional player. We feel a
larger study is needed to determine differences in hip measures that might have been
masked by combining positional players, with potentially reduced annual workload,
relative to pitchers. Secondly, data were captured in a cross-sectional study
design, yet a longitudinal assessment with serial measures may provide a better
understanding of the interplay between pain and hip characteristics over time in
youth baseball and softball athletes. Lastly, we did not evaluate the differences in
other known risk factors for injury like throwing biomechanics, workload, etc. on
upper extremity pain in our analysis as this model was utilized to understand how
these clinical measures may be useful for interventions directed at injury
prevention.
Prior research in baseball and softball indicates various modifiable and
non-modifiable intrinsic and extrinsic risk factors for upper extremity injuries.
When controlling for age, height and body mass, rotational hip measures were similar
between baseball and softball athletes. Moreover, there was no relationship between
sport, hip ROM, hip strength, and pain in our model. More research is needed to
determine the interrelatedness between hip ROM, hip strength, and upper extremity
pain in a larger cohort of throwing athletes. Further, identifying sex- and
sport-specific differences in hip ROM in throwing athletes with upper extremity pain
in baseball and softball may be useful for player assessment and injury
prevention.
