Skip to main content
NCBI home page
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2022 Mar 5.
Published in final edited form as: J Biomech. 2021 Jan 23;117:110266. doi: 10.1016/j.jbiomech.2021.110266

Age-related differences in humerothoracic, scapulothoracic, and glenohumeral kinematics during elevation and rotation motions

Christopher W Kolz 1,2, Hema J Sulkar 1,2, Klevis Aliaj 1,2, Robert Z Tashjian 1, Peter N Chalmers 1, Yuqing Qiu 4, Yue Zhang 4, K Bo Foreman 1,3, Andrew E Anderson 1,2,3, Heath B Henninger 1,2
PMCID: PMC7924070  NIHMSID: NIHMS1668034  PMID: 33517243

Abstract

Age affects gross shoulder range of motion (ROM), but biomechanical changes over a lifetime are typically only characterized for the humerothoracic joint. Suitable age-related baselines for the scapulothoracic and glenohumeral contributions to humerothoracic motion are needed to advance understanding of shoulder injuries and pathology. Notably, biomechanical comparisons between younger or older populations may obscure detected differences in underlying shoulder motion. Herein, biplane fluoroscopy and skin-marker motion analysis quantified humerothoracic, scapulothoracic, and glenohumeral motion during 3 static poses (resting neutral, internal rotation to L4-L5, and internal rotation to maximum reach) and 2 dynamic activities (scapular plane abduction and external rotation in adduction). Orientations during static poses and rotations during active ROM were compared between subjects <35 years and >45 years of age (N=10 subjects per group). Numerous age-related kinematic differences were measured, ranging 5–25°, where variations in scapular orientation and motion were consistently observed. These disparities are on par with or exceed mean clinically important differences and standard error of measurement of clinical ROM, which indicates that high resolution techniques and appropriately matched controls are required to avoid confounding results of studies that investigate shoulder kinematics. Understanding these dissimilarities will help clinicians manage expectations and treatment protocols where indications and prevalence between age groups tend to differ. Where possible, it is advised to select age-matched control cohorts when studying the kinematics of shoulder injury, pathology, or surgical/physical therapy interventions to ensure clinically important differences are not overlooked.

Keywords: shoulder, age, rotation, abduction, kinematics

Introduction

Shoulder injuries and pathologies differentiate by age. Shoulder dislocation and acromioclavicular joint separations typically occur in patients 20–40 years of age (Chillemi et al., 2013; Zacchilli and Owens, 2010). Degenerative rotator cuff pathology is uncommon in patients under 50 (Lazarides et al., 2015; Minagawa et al., 2013), and glenohumeral (GH) osteoarthritis is uncommon in patients under 60 (Thomas et al., 2016). Shoulder arthroplasty, both anatomic (TSA) and reverse (rTSA), is not commonly performed in patients under 55. Data suggests that age is a significant factor related to functional outcomes and revision rate following rotator cuff repair and shoulder replacement surgery (Anakwenze et al., 2017; Lazarides et al., 2015). Clinical management in younger patients is met with higher expectations due to greater activity levels, frequent return to sport, and the need for more durable outcomes (Henn et al., 2011; Lazarides et al., 2015). Therefore, understanding the relationship between age and shoulder motion helps clinicians better manage expectations of functional range of motion (ROM) and ultimately patient satisfaction.

Age affects shoulder function (Barnes et al., 2001; Murgia et al., 2018). Humerothoracic (HT) ROM decreases with age for abduction and external rotation, but increases with age for internal rotation (Barnes et al., 2001). A stable scapula, capable of pain free range of motion in coordination with the humerus, is the basis for shoulder ROM (Paine and Voight, 2013), therefore abnormal scapular motion impairs shoulder function (Kibler et al., 2013). HT motion is the sum of GH and scapulothoracic (ST) contributions, which are each challenging to accurately measure since 3D movement of the scapula and humerus beneath soft tissue requires advanced imaging or invasive techniques to precisely quantify (Bourne et al., 2007; Tempelaere et al., 2016). Therefore, age-related scapular and glenohumeral contributions to shoulder motion remain understudied, yet quantification is necessary to better understand their effects on shoulder injury, pathology, or clinical interventions.

Recruitment of subjects with healthy shoulders over age 45 is challenging due to a higher prevalence of pathology and past injury (Tempelhof et al., 1999; Thomas et al., 2016). Thus, many studies have used healthy control populations under 35 years old for comparisons to clinical populations over 45 years of age (Bey et al., 2011; Hebert et al., 2002). Mismatched age groups may confound the conclusions of these studies, so it is important to establish the effect of age on ST and GH motion to determine if age-matched populations are necessary.

The purpose of this study was to compare healthy HT, ST, and GH kinematics in subjects <35 and >45 years of age. We hypothesized that the >45 population would exhibit (1) less HT ROM during scapular plane abduction and external rotation and (2) greater HT ROM during internal rotation, and (3) ST and GH kinematics would differ by age.

Methods

Participants

After obtaining informed consent (IRB 71782), ten healthy subjects <35 years of age and ten healthy subjects >45 years of age participated in a study of shoulder motion. These age groups were motivated by the relative distribution of shoulder injury and pathology described previously. Subjects recruited from the University of Utah were in good health (no preconditions on activity level), BMI ≤30, with equal distribution of sex per group. Subjects were screened for gross shoulder abnormalities and any history of shoulder surgery or pain. After pre-screening, anteroposterior and axillary radiographs were obtained to examine parameters like joint congruency and spacing. Magnetic resonance imaging followed standard shoulder protocols for our institution (e.g. axial PD FS; coronal T1, T2 FS; sagittal T1, T2 FS, oblique T2 FS; 320×320 or 384×384 matrix; 3.5/3.8 mm or 5.0/5.5 mm slices; no contrast; Magnetom Avanto_Fit 1.5T, Siemens, Malvern, PA). A fellowship-trained, board-certified orthopedic shoulder surgeon (PNC or RZT) reviewed radiographic and magnetic resonance imaging for signs of pathology. Subjects determined to have asymptomatic pathology (e.g. rotator cuff tears, osteoarthritis) were excluded from further study.”

Activities

Five poses/activities were studied: (1) static resting neutral, (2) static internal rotation (IR) in adduction to L4-L5, (3) static IR in adduction to maximum reach (i.e. thoracic levels), (4) dynamic scapular plane abduction, and (5) dynamic external rotation (ER) in adduction (Figure 1). These were selected since scapular plane abduction is the most commonly studied shoulder motion (Krishnan et al., 2019), and reaching behind the back (i.e. functional IR, a combination of extension and IR) plays a critical role in executing tasks of daily living (Langer et al., 2012). Neutral was used to quantify the resting pose of the scapula and humerus relative to the torso.

Figure 1.

Figure 1.

Open in a new tab

Schematic of resting neutral (static) (A), internal rotation (IR) in adduction (static) (B), scapular plane abduction (active) (C), and external rotation (ER) in adduction (active) (D). Resting neutral was defined with the arm at the subject’s side, the elbow flexed 90°, the forearm pointing anteriorly, and hand in a thumb up position. IR in adduction was performed to the L4-L5 spinal level and to maximum reach up the subject’s spine, with the arm behind the subject’s back and the palm open and facing posteriorly. Scapular plane abduction was defined as arm elevation with the subject’s elbow fully extended and the hand in a thumb up position, approximately 30° anterior to the coronal plane. ER in adduction was axial rotation of the arm at the subject’s side with the elbow flexed 90° and the hand in a thumb up position, beginning from maximal internal rotation with the forearm on the abdomen up to peak ER. Subjects performed scapular plane abduction at approximately 90° per second of elevation and external rotation in adduction at approximately 45° per second of rotation. Performance speeds were selected to be physiologically relevant and similar to previously reported rates (Bey et al., 2006; Phadke et al., 2011). Subjects were allowed to practice each motion prior to data capture until they felt comfortable with the motion and the timing.

Motion Capture and Analysis

A custom high-speed biplane fluoroscopy system (Radiological Imaging Services, Hamburg, PA) was used to image in-vivo shoulder kinematics at 100 Hz. A ten-camera motion analysis system (Vicon, Centennial, CO) was temporally and spatially synchronized to measure body motion outside the fluoroscope field of view. Subjects practiced the pose/motion prior to data capture to become familiar with the instructions. To reduce exposure to ionizing radiation, a trial was only repeated if the subject shifted out of the field of view or image quality was insufficient to allow markerless tracking. If multiple trials were captured, the last/best trial was analyzed. A computed tomography (CT) scan of each subject’s right shoulder followed standard imaging protocols for our institution (e.g. 130 kVp, 170 mAs with CareDose; 512×512 matrix; 1.0 mm slices (no overlap); no contrast; SOMATOM Definition AS, Siemens, Malvern, PA) capturing the entire scapula and humerus.. The scapula and humerus were semi-automatically segmented (Mimics, Materialise, Plymouth, MI) to generate 3D reconstructions.

Model-based markerless tracking was used to semi-automatically align digitally reconstructed radiographs of the humerus and scapula with the corresponding fluoroscope images (Bey et al., 2006) (average Euclidian bias of 0.32±0.08 mm (pelvis) and 0.30±0.06 mm (femur) using the same system and software (Kapron et al., 2014)). Spatial and temporal calibration between marker and markerless systems allowed the torso (skin markers) to serve as the reference coordinate system to control for changes in posture and body habitus. Rotation matrices for HT, ST, and GH motions were calculated for each time point using subject-specific coordinate systems for the humerus, scapula, and torso (Figure 2). Each HT and GH rotation matrix was decomposed using the X-Z’-Y” sequence due to known issues numerically representing the underlying motion using the Y-X’-Y” sequence (Phadke et al., 2011). ST rotation matrices were decomposed using the Y-X’-Z” sequence (Wu et al., 2005). The glenoid center was used as the scapular origin (Kolz et al., 2020).

Figure 2.

Figure 2.

Open in a new tab

Local coordinate systems for the humerus and scapula. Humerus coordinate system: The origin was located at the centroid of a sphere fit to the articular surface of the humeral head. The vector from the midpoint of the elbow epicondyles to the humeral head center was defined as the humeral Y-axis (+superior). The normal to the plane defined by the humeral head center and the elbow epicondyles was the humeral X-axis (+anterior). The cross product of the humeral X- and Y-axes defined the humeral Z-axis (+lateral) (Wu et al., 2005). Scapula coordinate system: The origin was located at the glenoid center, defined by a circular fit of the inferior glenoid rim (De Wilde et al., 2010; Verstraeten et al., 2018). The vector from the trigonum spinae scapulae to the glenoid center was defined as the scapular Z-axis (+lateral). The normal to the plane defined by the glenoid center, inferior angle, and trigonum spinae scapulae was the scapular X-axis (+anterior). The cross product of the scapular Z- and X-axes defined the scapular Y-axis (+superior). Torso coordinate system (not shown): The origin was located coincident with the marker defining the incisura jugularis. The vector from the midpoint of the xiphoid process and T8 spinous process (obtained via anthropometric calculations) to the midpoint of the incisura jugularis and C7 spinous process defined the torso Y-axis (+superior). The normal to the plane defined by those four points was the torso Z-axis (+lateral). The cross product of the torso Z- and X-axes defined the torso X-axis (+anterior) (Wu et al., 2005). HT and GH rotations (X-Z’-Y”): Elevation (X), plane of elevation (Z’), and finally internal/external rotation (Y”). ST rotations (Y-X’-Z”): Protraction (Y), medial/lateral rotation (X’), and finally posterior tilt (Z”). Sign conventions for rotations followed the right-hand rule about the positive axes, but for more intuitive visual representation some were reversed from their coordinate system definitions (e.g. IR-/ER+, Depression-/Elevation+, and Medial Rotation-/Lateral Rotation+) as defined in the respective figure legends. This is a noted discrepancy between ISB and the International Shoulder Group definitions of elevation (Wu et al., 2005).

Scapulohumeral rhythm (SHR) was calculated two ways. First, as SHR=ΔGH/ΔST, where the change was calculated relative to the joint position at the start of the motion. Second, as SHR=δGH/δST, which was the instantaneous change relative to the prior increment of HT elevation. Since substantial reorientation can occur during dynamic motion, ROM was calculated as (max-min) across the motion, not (start-end). Therefore, ROM at times exceeded the difference between start and end orientations.

Statistical Analysis

A priori power analyses using two independent means were performed using G*Power (v 3.1.9.4, University of Düsseldorf, Germany). Mean differences were conservatively estimated between ages for abduction (10–15°) and ER (7.5°) using previously recorded changes in ROM (Barnes et al., 2001). Standard deviations were estimated as 7.5–10° (Barnes et al., 2001; Schwartz et al., 2014). To achieve 0.8 power a sample size of 9 subjects per group was required.

All demographics and static poses, including the start and end positions from dynamic motions, were compared between age groups using two tailed independent t-tests assuming equal variance. For comparisons between static poses within an age group (e.g. rest versus maximal IR), paired t-tests with Holm-Bonferroni correction for multiple comparisons were used. Dynamic rotations were interpolated at 0.1° increments and Statistical Parametric Mapping (SPM) was used to compare between age groups. SPM uses a time-domain dependent analysis with the advantage that it enables temporal comparisons to compute cluster-based p-values (Bangerter et al., 2019). A previous implementation of SPM using independent t-tests was adapted for our analysis (1d version 0.4, www.spm1d.org) (Pataky et al., 2015). Statistical significance (P≤0.050) and marginal significance (0.050<P≤0.100) used to detect potential regions of difference for consideration in future shoulder kinematics studies.

Results

Demographics

Ten healthy subjects <35 years of age (5M/5F; 26±3 yrs; 171.4±11.1 cm; 60.9±13.1 kg; 20.6±3.4 kg/m2) and ten healthy subjects >45 years of age (5M/5F; 58±7 yrs; 173.1±6.9 cm; 79.0±13.9 kg; 26.2±3.4 kg/m2) were recruited. Weight and BMI for individuals over 45 years were greater compared to the group under 35 years of age (P≤0.008). No differences were detected for height between the two groups (P=0.687). All subjects were right-hand dominant, and all subjects that met screening and inclusion criteria completed the entire study protocol. One subject (<35) was excluded from the analysis due to an incomplete marker set during post-processing calibration, preventing calculation of poses and orientations relative to the torso coordinate system. One subject (>45) was excluded from scapular plane abduction and a different subject (>45) was excluded from IR/ER in adduction due to poor image quality that prevented tracking of those trials.

Static Poses

For both age groups, IR poses had 20° or more HT internal rotation, elevation, and plane of elevation than resting neutral orientations (P≤0.001) (Figure 3 A-C). Subjects over 45 years of age experienced approximately 10° greater HT elevation during resting neutral, IR to L4-L5, and IR to maximum reach (P≤0.031) (Figure 3B). There were no differences by age for plane of elevation, but marginal significance was detected between ages in resting internal rotation (~10°, P=0.092) (Figure 3A).

Figure 3.

Figure 3.

Open in a new tab

Humerothoracic (A-C), scapulothoracic (D-F), and glenohumeral (G-I) orientations during static resting neutral, internal rotation (IR) (L4-L5), and IR (max) positions, grouped by age. Most differences in joint orientations were detected between the resting neutral and IR poses (A, B, C, E, F, G, I) but some differences between poses only arose within one age group (e.g. <35 G, H; >45 B, E, F, I). Age yielded clear differences in humerothoracic elevation (B), with marginal significance at rest for humerothoracic IR/ER (A). Data are presented relative to the torso coordinate system as mean±SD. Note that for more intuitive visual representation some sign conventions were reversed from their coordinate system definitions (e.g. IR-/ER+, Depression-/Elevation+, and Medial Rotation-/Lateral Rotation+).

For the ST joint, no differences were detected in scapular protraction by pose or age (Figure 3D). To the contrary, the scapula experienced approximately 10° more medial rotation (P≤0.001) (Figure 3E) and approximately 5° anterior tilt (P≤0.013) (Figure 3F) for IR poses compared to neutral for both age groups, with the exception of maximum IR in the <35 years group. While subjects >45 years tended toward higher mean protraction, tilt, and rotation than younger subjects in 8/9 cases, these differences only reached statistical significance for tilt (P≤0.057) (Figure 3F).

For the GH joint, internal/external rotation (Figure 3G) and plane of elevation (Figure 3I) inverted for IR poses compared to resting neutral. The L4-L5 and maximal IR poses differed for only 3/6 comparisons within an age group (P≤0.042) (Figure 3G-I). GH orientations did not differ by age. Individuals >45 years tended toward higher elevation (Figure 3H) but internal/external rotation and plane of elevation showed no consistent trends by age.

Dynamic Activities

Total ROM differed statistically for HT and GH IR/ER during scapular plane elevation, where the >45 years group achieved mean 22° more HT rotation and 13° more GH rotation than the <35 years group (P≤0.029) (Table 1). Since the HT and GH rotation end orientations did not differ by age during scapular plane abduction, the change in ROM arose from differences in starting rotation of the arm (P≤0.034). No other variations in ROM were detected, but 11/36 start and end orientations demonstrated at least marginal statistical significance (P≤0.089). The ST joint was responsible for 5/11 statistical differences detected, namely in scapular tilt and protraction. The HT and GH elevation and plane of elevation made up the remainder of differences in start and end positions.

Table 1.

Humerothoracic (HT), scapulothoracic (ST), and glenohumeral (GH) start and end orientations and ROM for scapular plane abduction and external rotation in adduction motions.

<35 Years >45 Years P P P
Start End ROM Start End ROM Start End ROM
Scapular plane abduction HT IR/ER 13±6 85±14 72±17 -10±20 84±7 94±21 0.004 0.809 0.028
Elevation 6±6 157±12 151±15 10±4 155±6 145±5 0.102 0.712 0.266
Plane of elevation 7±10 23±9 25±8 2±6 20±7 28±10 0.237 0.452 0.560
ST Protraction 42±9 44±7 7±3 47±4 55±6 10±6 0.136 0.002 0.173
Lateral rotation −14±8 32±9 47±5 −14±10 36±9 50±4 0.887 0.330 0.138
Tilt −8±3 −3±4 5±3 −14±8 −10±7 5±3 0.041 0.023 0.973
GH IR/ER 40±9 53±7 18±6 23±19 50±7 31±16 0.034 0.393 0.029
Elevation 8±8 102±8 94±11 6±7 98±7 91±6 0.551 0.176 0.519
Plane of elevation 12±7 17±12 21±11 12±9 7±6 19±7 0.979 0.029 0.623
External rotation in adduction HT IR/ER −63±11 54±14 117±18 −64±8 48±9 113±8 0.803 0.323 0.521
Elevation 12±5 9±6 7±3 15±7 16±9 6±3 0.415 0.051 0.635
Plane of elevation 2±7 −1±5 7±3 8±8 −2±6 11±8 0.107 0.828 0.168
ST Protraction 45±8 36±9 10±5 48±5 37±8 11±8 0.380 0.903 0.626
Lateral rotation −13±8 −15±10 4±2 −13±9 −14±9 4±2 0.850 0.800 0.987
Tilt −7±4 −9±4 3±1 −14±7 −15±9 3±2 0.030 0.084 0.816
GH IR/ER −16±10 93±8 109±15 −11±7 91±12 102±9 0.288 0.690 0.275
Elevation 23±8 21±7 6±2 30±6 26±8 7±4 0.053 0.238 0.453
Plane of elevation 0±8 3±6 7±4 7±8 2±13 7±6 0.089 0.882 0.953
Open in a new tab

Reported as mean±SD degrees where bold italics denotes significance between age groups (P≤0.050). Italics denote marginal significance (0.050<P≤0.100). Since substantial reorientation can occur during a dynamic motion, ROM was calculated as the maximum range (max-min) across the entirety of the performed motion, not the difference between start and end orientations. At times this resulted in the ROM exceeding the difference between start and end orientations (e.g. GH plane of elevation during scapular plane abduction).

Dynamic motions were also evaluated across the HT ROM achievable by all subjects: 25° to 125° for abduction and −45° to 35° for IR/ER (Figure 4–7).During HT scapular plane abduction, both groups exhibited an increase in external rotation with humeral elevation, but the <35-year group was up to 20° higher than the >45 group across much of the ROM (Figure 4A). No differences were observed in plane of elevation between age groups in scapular plane abduction (Figure 4B). During external rotation in adduction, the >45-year group demonstrated approximately 5° more humeral elevation and 5° more anterior plane of elevation than the <35 group, though they only statistically differed over a fraction of the ROM (Figure 4C,D).

Figure 4.

Figure 4.

Open in a new tab

Humerothoracic rotations for scapular plane abduction (A, B) and internal/external rotation (IR/ER) in adduction (C, D). By age, statistical differences on the order of 10–20° were detected in IR/ER during scapular abduction (A), and approximately 5° for elevation and plane of elevation during IR/ER motions (C, D). Elevation during scapular plane abduction and IR/ER during ER in adduction are not shown because these act as the independent variables for the corresponding dynamic motions, resulting in a slope of 1, with no variance. Solid and dashed curves represent the mean and shaded regions represent the standard deviation of the respective populations. Note that for more intuitive visual representation some sign conventions were reversed from their coordinate system definitions (e.g. IR-/ER+, Depression-/Elevation+).

The >45-year group consistently had 10° more scapular protraction than the <35-year group during scapular plane abduction (Figure 5A). While the groups did not differ in medial/lateral ST rotation (Figure 5B), statistical differences arose for scapular anterior tilt, where the >45-year group was more anteriorly tilted by approximately 5° (Figure 5C). External rotation in adduction did not yield any qualitative or quantitative differences in ST protraction or rotation (Figure 5D,E), but the >45-year group was approximately 7° more anteriorly tilted across the ROM (Figure 5F).

Figure 5.

Figure 5.

Open in a new tab

Scapulothoracic rotations for scapular plane abduction (A-C) and internal/external rotation (IR/ER) in adduction (D-F). By age, statistical differences of approximately 10° were detected in protraction (A) and 5° in tilt (C) during scapular abduction. During IR/ER in adduction the scapulothoracic rotation was nearly identical between age groups (E), but there were differences of approximately 7° detected for scapula tilt across the range of motion (F).Solid and dashed curves represent the mean and shaded regions represent the standard deviation of the respective populations. Note that for more intuitive visual representation some sign conventions were reversed from their coordinate system definitions (e.g. Medial Rotation-/Lateral Rotation+).

Interestingly, the <35-year group tended toward higher mean GH rotations during abduction (Figure 6A-C), but this relationship inverted during external rotation in adduction (Figure 6D-F). Scapular plane abduction yielded no statistical differences in GH external rotation or elevation (Figure 6A,B), but up to 10° difference between groups for plane of elevation near terminal ROM (Figure 6C). The >45-year group had more external rotation and elevation, primarily during the internal rotation component of the IR/ER activity (Figure 6D,E). No differences were detected in plane of elevation during IR/ER in adduction (Figure 6F).

Figure 6.

Figure 6.

Open in a new tab

Glenohumeral rotations for scapular plane abduction (A-C) and internal/external rotation (IR/ER) in adduction (D-F). By age, statistical differences of approximately 10° were only detected in plane of elevation during scapular plane abduction (C). To the contrary, differences of 7–10° were detected for external rotation (D) and elevation (E) during IR/ER motions, with differences arising during the internal rotation component of the motion. Solid and dashed curves represent the mean and shaded regions represent the standard deviation of the respective populations. Note that for more intuitive visual representation some sign conventions were reversed from their coordinate system definitions (e.g. IR-/ER+, Depression-/Elevation+).

The HT plane of elevation was achieved as instructed, with a mean near 30° at 90° elevation, which was the instructed position (Figure 4B). At 90° HT elevation the GH plane of elevation showed means <5° but spanned over 30° (Figure 6C), demonstrating the underlying variance in ST and GH relationships relative to HT alignment.

There were no statistical differences in SHR (Figure 7), which was dependent on the calculation. When referenced to initial shoulder orientation, subjects peaked near 6:1 then rapidly decreased, stabilizing near 2:1 (Figure 7A). As an incremental change in GH and ST motion, the SHR was consistently near 2:1 (Figure 7B).

Figure 7.

Figure 7.

Open in a new tab

Scapulohumeral rhythm (SHR) during scapular plane abduction was calculated as the ratio of GH elevation to ST lateral rotation. (A) Here the SHR (Δ) represents the change relative to the initial position of the subject. Early in the elevation the SHR was dominated by glenohumeral humeral motion relative to scapulothoracic, resulting in SHR upward of 6 in some subjects, but stabilized around a 2:1 ratio above approximately 50° of humeral elevation. (B) Here the SHR (δ) represents the instantaneous change relative to the prior increment of elevation. A 2:1 ratio was observed across the entire range of humeral elevation. No statistical differences were detected between sexes.

Discussion

The objective of this study was to quantify age-related differences in HT, ST, and GH kinematics to determine if mismatched selection of control populations could confound studies of shoulder motion. To meet this objective, the study measured shoulders in control subjects <35 and >45 years of age during three static poses and two dynamic motions using skin marker and biplane fluoroscopy motion capture systems. We hypothesized that the >45 population would exhibit (1) less HT ROM during scapular plane abduction and external rotation and (2) greater HT ROM during internal rotation, and (3) ST and GH kinematics would differ by age. Our first two hypotheses were unsupported, since only a few gross differences in HT ROM were detected, and none in the primary HT motions of scapular abduction and IR/ER in adduction. Our third hypothesis was generally supported since numerous statistical differences were detected in ST and GH kinematics by age. Scapulothoracic motion differed between groups in the range of 5–22° which is well above the accuracy of the biplane fluoroscopy system (Kapron et al., 2014), indicating they present as a function of age. Since nearly half of all measures demonstrated statistical differences between age groups, it is recommended that future studies of shoulder pathology and injury consider age-matching control and patient populations. With more laboratories using highly accurate motion analysis techniques like biplane fluoroscopy to examine shoulder motion, differences that were obscured using lower resolution techniques can no longer be overlooked.

Mean clinically important difference (MCID) defines the smallest difference of an intervention that a patient perceives as beneficial (Jaeschke et al., 1989). No MCIDs for ROM after rotator cuff repair exist, so a prior cuff repair study (Nazari et al., 2019) used values from shoulder arthroplasty where active abduction and external rotation following shoulder replacement have MCIDs of 7° and 3°, respectively (Simovitch et al., 2018). These were likely measured using a clinical goniometer, which exhibits standard errors of measurement of 5–8° (Muir et al., 2010), but is incapable of capturing accurate 3D shoulder motion. Mean differences in static poses and dynamic kinematics ranged from 5–22° (Table 1, Figures 3–6), which exceeds these MCIDs, including goniometer measurement error. Inappropriately aged control populations could inhibit the detection of clinically important differences or yield false positives.

Resting neutral orientations differed from IR behind the back for all but ST protraction (Figure 3). Differences were anticipated given the disparity in the poses, but scapular rotation and tilt, and GH rotation and plane of elevation, are the primary drivers for achieving IR positions. Lost IR due to pathology, injury, or surgical repair (Jain et al., 2013; Wirth et al., 2016) should focus on these variants to identify the source of the deficiency and potential approaches to mitigate the loss. Interestingly, age related differences of 5–10° only manifested in HT elevation and ST tilt. Perhaps higher BMI in the >45 group and consequent torso and arm girth required the humerus to deviate from the torso by a larger elevation angle, and decreased strength of the scapular stabilizing muscles allowed more anterior tilt. The origins of these differences should be studied to determine their effect on age-related shoulder kinematics.

Total ROM alone is likely not an adequate measure of in vivo shoulder function since it cannot readily differentiate the nuances of underlying ST and GH motion. Where prior studies noted differences by age (Barnes et al., 2001; Murray et al., 1985), our first two hypotheses did not support those observations. Only a few gross differences in HT ROM were detected, with none in the primary HT motions of scapular abduction and IR/ER (Table 1). This inconsistency may have arisen from measuring bone motion using biplane fluoroscopy versus lower resolution methods measuring body segments, that we expressed motions relative to a local torso coordinate system rather than global posture, or simply a lack of differences between groups. Regardless, most statistical differences arose in ST and GH joints which combine to generate HT motion. Here there were clear differences in ST and GH motion by age, on the order of 5–22°, for ST protraction and tilt (Figure 5) and GH rotation and elevation (Figure 6). In addition, the ST and GH poses differed more often than gross HT poses or ROM (Table 1). As clinical care and research focus more on deficiencies of ST motion and their effects on shoulder function, it will be critical to age match control cohorts so as to capture the inherent differences specific to ST and GH function.

Our SHR agrees with the generalized 2:1 ratio beyond ~50° HT elevation but is dependent on the calculation. As changes relative to the initial joint position (Δ, Figure 7A) the SHR near 6:1 illustrates the large contribution of GH versus ST motion in the early phase of abduction. When taken incrementally (δ, Figure 7B), both groups consistently achieved the 2:1 ratio across the ROM. While early SHR decay has been reported previously (Chung et al., 2019; Robert-Lachaine et al., 2016; Scibek and Carcia, 2012), others report nearly constant SHR (Inman et al., 1996; Ludewig et al., 2009; Matsuki et al., 2011). These seemingly contradictory reports are clearly affected by the calculation, but may also be affected by the data capture. The ST and GH measurements are often obscured by soft tissue artefact, which is avoided when using radiographic techniques like ours. Consistency in reporting is needed to reliably study pathology or procedures like rTSA. SHR decreases to ~1:1 after rTSA surgery, illustrating the increased demands on ST motion (de Toledo et al., 2012; Kwon et al., 2012; Walker et al., 2015). Given that rTSA patients are predominantly >55 years of age, along with our kinematics findings, comparisons to appropriately age matched cohorts are advised. Of note, SHR was essentially 2D, using only single axes of rotation for calculation of the proportion of GH to ST motion. Future work will examine if 3D SHR may overcome issues with identifying sensitive out of plane differences in motion (Lee et al., 2020).

Sex can also affect shoulder biomechanics (Murgia et al., 2018; Nagamatsu et al., 2015). For completeness, we performed a secondary analysis by sex (see Supplement). The ten female (42±18 yrs; 165.5±4.0 cm; 61.4±12.8 kg; 22.3±4.3 kg/m2) and ten male subjects (42±18 yrs; 179.1±7.4 cm; 78.5±14.8 kg; 24.5±4.4 kg/m2) were re-pooled, and height and weight differed (P≤0.013). Static poses yielded some differences by sex, but inconsistently between poses (Figure S1). Sex did not affect ROM, and only a few statistical differences were detected for start and end positions, primarily between GH elevation during scapular plane abduction (Table S1). This appeared as a gross offset in start and end positions, since GH elevation ROM was very similar. Sex had only a minor effect on shoulder kinematics during motion. No HT or ST differences were detected by sex (Figure S2, S3). The GH elevation differed by 5–10° across scapular plane abduction (Figure S4B) and GH ER differed by a similar magnitude during IR/ER in adduction (Figure S4D). The trends in SHR were similar to the analysis by age, where again no significant differences were detected by sex (Figure S5).

This study is not without limitations. Although we captured two disparate age groups, there is likely a gradient of change that occurs over a human lifetime and as a function of changing body habitus. Though in good agreement with prior high resolution 3D data for HT, ST and GH shoulder motion (Ludewig et al., 2009), our data should not be considered definitive in magnitude, but rather evidence to support the need for matching of control cohorts where possible. It is recognized that substantial challenges exist in identifying healthy shoulders since prevalence of both full thickness rotator cuff tears and GH arthritis each individually exceed 30% in older individuals (Minagawa et al., 2013; Thomas et al., 2016). Second, variations in study population size may affect results. The present sample size was small compared to the 280 subjects who participated in (Barnes et al., 2001). However, that study was only able to investigate HT angles using a goniometer, which is prone to measurement error. The use of biplane fluoroscopy provided highly accurate kinematic data, but this technique is time- and resource-intensive and exposes individuals to ionizing radiation, limiting the number of participants feasible for analysis. This may have decreased the number of statistical differences detected, but our power analysis supported N=9 subjects per group to measure conservative effect sizes. The fact that numerous differences were detected herein, with deviations between age groups up to 22°, is sufficient evidence to warrant cohort matching for analyses of shoulder function. Prior studies report skin-marker motion analysis errors for the shoulder that exceed 10° (Cereatti et al., 2015), indicating that high resolution techniques may be required to discern clinically important differences below the levels perceptible using lower resolution measurements. This is especially important if ST and GH motion is critical to the research question under study. Finally, right shoulders were tested without external applied loads, but side, relative effort, and strength affect shoulder kinematics (Murgia et al., 2018; Schwartz et al., 2014). This is further evidence in support of appropriately matching control cohorts to mitigate factors that may obscure detection of changes in shoulder function by group.

In conclusion, this study compared individuals <35 and >45 years of age during 3 static poses and 2 dynamic activities, and quantified resultant kinematics of the HT, ST, and GH joints. These data provide valuable baselines for expected differences in shoulder motion by age, and detailed measurements of the ST and GH joints which were not previously reported. Data on IR of the arm behind the back were presented, highlighting areas of focus for individuals suffering from limited IR mobility. While scapular plane elevation and IR/ER in adduction ROM did not differ between groups, numerous differences in the ST and GH joints were uncovered. Protraction and tilt in the ST joint, and GH rotation and elevation, were the most likely sources of disparities between age groups and could provide targets for improving shoulder function in pathology, injury, and clinical care of the shoulder. These data highlight the importance of ST motion in overall shoulder motion, where differences up to 22° are present between younger and older individuals. These data support the need for age matching of control populations when studying shoulder (dys)function so as not to obscure variations solely based on inherent shoulder function of the reference population.

Supplementary Material

1

Acknowledgements

Research reported in this publication was supported by the National Institute of Arthritis and Musculoskeletal and Skin Diseases (NIAMS) of the National Institutes of Health under award number R01 AR067196 and by the University of Utah Population Health Research (PHR) Foundation, with funding in part from the National Center for Research Resources and the National Center for Advancing Translational Sciences, National Institutes of Health, through Grant UL1TR002538 (formerly 5UL1TR001067-05, 8UL1TR000105 and UL1RR025764). The research content herein is solely the responsibility of the authors and does not necessarily represent the official views of the National Institutes of Health or the University of Utah Population Health Research Foundation.

Footnotes

Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

References

  1. Anakwenze OA, Yehyawi T, Dillon MT, Paxton E, Navarro R, Singh A, 2017. Effect of Age on Outcomes of Shoulder Arthroplasty. Perm J 21, 16–056. [DOI] [PMC free article] [PubMed] [Google Scholar]
  2. Bangerter C, Romkes J, Lorenzetti S, Krieg AH, Hasler CC, Brunner R, Schmid S, 2019. What are the biomechanical consequences of a structural leg length discrepancy on the adolescent spine during walking? Gait Posture 68, 506–513. [DOI] [PubMed] [Google Scholar]
  3. Barnes CJ, Van Steyn SJ, Fischer RA, 2001. The effects of age, sex, and shoulder dominance on range of motion of the shoulder. J Shoulder Elbow Surg 10, 242–246. [DOI] [PubMed] [Google Scholar]
  4. Bey MJ, Peltz CD, Ciarelli K, Kline SK, Divine GW, van Holsbeeck M, Muh S, Kolowich PA, Lock TR, Moutzouros V, 2011. In vivo shoulder function after surgical repair of a torn rotator cuff: glenohumeral joint mechanics, shoulder strength, clinical outcomes, and their interaction. Am J Sports Med 39, 2117–2129. [DOI] [PMC free article] [PubMed] [Google Scholar]
  5. Bey MJ, Zauel R, Brock SK, Tashman S, 2006. Validation of a new model-based tracking technique for measuring three-dimensional, in vivo glenohumeral joint kinematics. J Biomech Eng 128, 604–609. [DOI] [PMC free article] [PubMed] [Google Scholar]
  6. Bourne DA, Choo AM, Regan WD, MacIntyre DL, Oxland TR, 2007. Three-dimensional rotation of the scapula during functional movements: an in vivo study in healthy volunteers. J Shoulder Elbow Surg 16, 150–162. [DOI] [PubMed] [Google Scholar]
  7. Cereatti A, Rosso C, Nazarian A, DeAngelis JP, Ramappa AJ, Della Croce U, 2015. Scapular Motion Tracking Using Acromion Skin Marker Cluster: In Vitro Accuracy Assessment. Journal of Medical and Biological Engineering 35, ePub. [Google Scholar]
  8. Chillemi C, Franceschini V, Dei Giudici L, Alibardi A, Salate Santone F, Ramos Alday LJ, Osimani M, 2013. Epidemiology of isolated acromioclavicular joint dislocation. Emerg Med Int 2013, 171609. [DOI] [PMC free article] [PubMed] [Google Scholar]
  9. Chung H, Kim D, Banks SA, Son J, Kim Y, On M, Yeom J, 2019. Evaluation of three-dimensional in vivo scapular kinematics and scapulohumeral rhythm between shoulders with a clavicle hook plate and contralateral healthy shoulders. Int Orthop 43, 379–386. [DOI] [PubMed] [Google Scholar]
  10. de Toledo JM, Loss JF, Janssen TW, van der Scheer JW, Alta TD, Willems WJ, Veeger DH, 2012. Kinematic evaluation of patients with total and reverse shoulder arthroplasty during rehabilitation exercises with different loads. Clin Biomech (Bristol, Avon) 27, 793–800. [DOI] [PubMed] [Google Scholar]
  11. Hebert LJ, Moffet H, McFadyen BJ, Dionne CE, 2002. Scapular behavior in shoulder impingement syndrome. Arch Phys Med Rehabil 83, 60–69. [DOI] [PubMed] [Google Scholar]
  12. Henn RF 3rd, Ghomrawi H, Rutledge JR, Mazumdar M, Mancuso CA, Marx RG, 2011. Preoperative patient expectations of total shoulder arthroplasty. J Bone Joint Surg Am 93, 2110–2115. [DOI] [PubMed] [Google Scholar]
  13. Inman VT, Saunders JB, Abbott LC, 1996. Observations of the function of the shoulder joint. 1944. Clin Orthop Relat Res, 3–12. [DOI] [PubMed] [Google Scholar]
  14. Jaeschke R, Singer J, Guyatt GH, 1989. Measurement of health status. Ascertaining the minimal clinically important difference. Control Clin Trials 10, 407–415. [DOI] [PubMed] [Google Scholar]
  15. Jain NB, Wilcox RB 3rd, Katz JN, Higgins LD, 2013. Clinical examination of the rotator cuff. PM R 5, 45–56. [DOI] [PMC free article] [PubMed] [Google Scholar]
  16. Kapron AL, Aoki SK, Peters CL, Maas SA, Bey MJ, Zauel R, Anderson AE, 2014. Accuracy and feasibility of dual fluoroscopy and model-based tracking to quantify in vivo hip kinematics during clinical exams. J Appl Biomech 30, 461–470. [DOI] [PMC free article] [PubMed] [Google Scholar]
  17. Kibler WB, Ludewig PM, McClure PW, Michener LA, Bak K, Sciascia AD, 2013. Clinical implications of scapular dyskinesis in shoulder injury: the 2013 consensus statement from the ‘Scapular Summit’. Br J Sports Med 47, 877–885. [DOI] [PubMed] [Google Scholar]
  18. Kolz CW, Sulkar HJ, Aliaj K, Tashjian RZ, Chalmers PN, Qiu Y, Zhang Y, Foreman KB, Anderson AE, Henninger HB, 2020. Reliable interpretation of scapular kinematics depends on coordinate system definition. Gait Posture 81, 183–190. [DOI] [PMC free article] [PubMed] [Google Scholar]
  19. Krishnan R, Bjorsell N, Gutierrez-Farewik EM, Smith C, 2019. A survey of human shoulder functional kinematic representations. Med Biol Eng Comput 57, 339–367. [DOI] [PMC free article] [PubMed] [Google Scholar]
  20. Kwon YW, Pinto VJ, Yoon J, Frankle MA, Dunning PE, Sheikhzadeh A, 2012. Kinematic analysis of dynamic shoulder motion in patients with reverse total shoulder arthroplasty. J Shoulder Elbow Surg 21, 1184–1190. [DOI] [PubMed] [Google Scholar]
  21. Langer JS, Sueoka SS, Wang AA, 2012. The importance of shoulder external rotation in activities of daily living: improving outcomes in traumatic brachial plexus palsy. J Hand Surg Am 37, 1430–1436. [DOI] [PubMed] [Google Scholar]
  22. Lazarides AL, Alentorn-Geli E, Choi JH, Stuart JJ, Lo IK, Garrigues GE, Taylor DC, 2015. Rotator cuff tears in young patients: a different disease than rotator cuff tears in elderly patients. J Shoulder Elbow Surg 24, 1834–1843. [DOI] [PubMed] [Google Scholar]
  23. Lee B, Kim D, Jang Y, Jin H, 2020. Three-dimensional in vivo scapular kinematics and scapulohumeral rhythm: a comparison between active and passive motion. J Shoulder Elbow Surg 29, 185–194. [DOI] [PubMed] [Google Scholar]
  24. Ludewig PM, Phadke V, Braman JP, Hassett DR, Cieminski CJ, LaPrade RF, 2009. Motion of the shoulder complex during multiplanar humeral elevation. J Bone Joint Surg Am 91, 378–389. [DOI] [PMC free article] [PubMed] [Google Scholar]
  25. Matsuki K, Matsuki KO, Mu S, Yamaguchi S, Ochiai N, Sasho T, Sugaya H, Toyone T, Wada Y, Takahashi K, Banks SA, 2011. In vivo 3-dimensional analysis of scapular kinematics: comparison of dominant and nondominant shoulders. J Shoulder Elbow Surg 20, 659–665. [DOI] [PubMed] [Google Scholar]
  26. Minagawa H, Yamamoto N, Abe H, Fukuda M, Seki N, Kikuchi K, Kijima H, Itoi E, 2013. Prevalence of symptomatic and asymptomatic rotator cuff tears in the general population: From mass-screening in one village. J Orthop 10, 8–12. [DOI] [PMC free article] [PubMed] [Google Scholar]
  27. Muir SW, Corea CL, Beaupre L, 2010. Evaluating change in clinical status: reliability and measures of agreement for the assessment of glenohumeral range of motion. N Am J Sports Phys Ther 5, 98–110. [PMC free article] [PubMed] [Google Scholar]
  28. Murgia A, Hortobagyi T, Wijnen A, Bruin L, Diercks R, Dekker R, 2018. Effects of age and sex on shoulder biomechanics and relative effort during functional tasks. J Biomech 81, 132–139. [DOI] [PubMed] [Google Scholar]
  29. Murray MP, Gore DR, Gardner GM, Mollinger LA, 1985. Shoulder motion and muscle strength of normal men and women in two age groups. Clin Orthop Relat Res, 268–273. [PubMed] [Google Scholar]
  30. Nagamatsu T, Kai Y, Gotoh M, Madokoro K, Shiba N, 2015. Effects of sex differences on scapular motion during arm elevation. SICOT J 1, 9. [DOI] [PMC free article] [PubMed] [Google Scholar]
  31. Nazari G, MacDermid JC, Bryant D, Dewan N, Athwal GS, 2019. Effects of arthroscopic vs. mini-open rotator cuff repair on function, pain & range of motion. A systematic review and meta-analysis. PLoS One 14, e0222953. [DOI] [PMC free article] [PubMed] [Google Scholar]
  32. Paine R, Voight ML, 2013. The role of the scapula. Int J Sports Phys Ther 8, 617–629. [PMC free article] [PubMed] [Google Scholar]
  33. Pataky TC, Robinson MA, Vanrenterghem J, 2013. Vector field statistical analysis of kinematic and force trajectories. J Biomech 46, 2394–2401. [DOI] [PubMed] [Google Scholar]
  34. Pataky TC, Vanrenterghem J, Robinson MA, 2015. Zero- vs. one-dimensional, parametric vs. non-parametric, and confidence interval vs. hypothesis testing procedures in one-dimensional biomechanical trajectory analysis. J Biomech 48, 1277–1285. [DOI] [PubMed] [Google Scholar]
  35. Phadke V, Braman JP, LaPrade RF, Ludewig PM, 2011. Comparison of glenohumeral motion using different rotation sequences. J Biomech 44, 700–705. [DOI] [PMC free article] [PubMed] [Google Scholar]
  36. Robert-Lachaine X, Allard P, Godbout V, Tetreault P, Begon M, 2016. Scapulohumeral rhythm relative to active range of motion in patients with symptomatic rotator cuff tears. J Shoulder Elbow Surg 25, 1616–1622. [DOI] [PubMed] [Google Scholar]
  37. Schwartz DG, Cottrell BJ, Teusink MJ, Clark RE, Downes KL, Tannenbaum RS, Frankle MA, 2014. Factors that predict postoperative motion in patients treated with reverse shoulder arthroplasty. J Shoulder Elbow Surg 23, 1289–1295. [DOI] [PubMed] [Google Scholar]
  38. Scibek JS, Carcia CR, 2012. Assessment of scapulohumeral rhythm for scapular plane shoulder elevation using a modified digital inclinometer. World J Orthop 3, 87–94. [DOI] [PMC free article] [PubMed] [Google Scholar]
  39. Simovitch R, Flurin PH, Wright T, Zuckerman JD, Roche CP, 2018. Quantifying success after total shoulder arthroplasty: the minimal clinically important difference. J Shoulder Elbow Surg 27, 298–305. [DOI] [PubMed] [Google Scholar]
  40. Tempelaere C, Pierrart J, Lefevre-Colau MM, Vuillemin V, Cuenod CA, Hansen U, Mir O, Skalli W, Gregory T, 2016. Dynamic Three-Dimensional Shoulder Mri during Active Motion for Investigation of Rotator Cuff Diseases. PLoS One 11, e0158563. [DOI] [PMC free article] [PubMed] [Google Scholar]
  41. Tempelhof S, Rupp S, Seil R, 1999. Age-related prevalence of rotator cuff tears in asymptomatic shoulders. J Shoulder Elbow Surg 8, 296–299. [DOI] [PubMed] [Google Scholar]
  42. Thomas M, Bidwai A, Rangan A, Rees JL, Brownson P, Tennent D, Connor C, Kulkarni R, 2016. Glenohumeral osteoarthritis. Shoulder Elbow 8, 203–214. [DOI] [PMC free article] [PubMed] [Google Scholar]
  43. Walker D, Matsuki K, Struk AM, Wright TW, Banks SA, 2015. Scapulohumeral rhythm in shoulders with reverse shoulder arthroplasty. J Shoulder Elbow Surg 24, 1129–1134. [DOI] [PubMed] [Google Scholar]
  44. Warmenhoven J, Harrison A, Robinson MA, Vanrenterghem J, Bargary N, Smith R, Cobley S, Draper C, Donnelly C, Pataky T, 2018. A force profile analysis comparison between functional data analysis, statistical parametric mapping and statistical non-parametric mapping in on-water single sculling. J Sci Med Sport 21, 1100–1105. [DOI] [PubMed] [Google Scholar]
  45. Wirth B, Kolling C, Schwyzer HK, Flury M, Audige L, 2016. Risk of insufficient internal rotation after bilateral reverse shoulder arthroplasty: clinical and patient-reported outcome in 57 patients. J Shoulder Elbow Surg 25, 1146–1154. [DOI] [PubMed] [Google Scholar]
  46. Wu G, van der Helm FC, Veeger HE, Makhsous M, Van Roy P, Anglin C, Nagels J, Karduna AR, McQuade K, Wang X, Werner FW, Buchholz B, International Society of B, 2005. ISB recommendation on definitions of joint coordinate systems of various joints for the reporting of human joint motion--Part II: shoulder, elbow, wrist and hand. J Biomech 38, 981–992. [DOI] [PubMed] [Google Scholar]
  47. Zacchilli MA, Owens BD, 2010. Epidemiology of shoulder dislocations presenting to emergency departments in the United States. J Bone Joint Surg Am 92, 542–549. [DOI] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

1
NIHMS1668034-supplement-1.docx (2.4MB, docx)

RESOURCES