International Journal of Industrial Ergonomics 75 (2020) 102892

Contents lists available at ScienceDirect

International Journal of Industrial Ergonomics

journal homepage: http://www.elsevier.com/locate/ergon

Shoulder muscle activity in off-axis pushing and pulling tasks

Alison C. McDonald, Carmen Tsang, Kimberly A. Meszaros, Clark R. Dickerson *
Department of Kinesiology, University of Waterloo, Waterloo, ON, Canada

A R T I C L E I N F O A B S T R A C T

Keywords: Workspace design can often dictate the muscular efforts required to perform work, impacting injury risk. Within
Manual materials handling many environments, industrial workers often use sub-maximal forces in offset directions in to accomplish job
Shoulder muscle activity tasks. The purpose of this research was to develop methods to estimate shoulder muscle activation during seated,
Off-axis exertions
static, sub-maximal exertions in off-axis (non-cardinal) directions. Surface EMG signals were recorded from 14
upper extremity muscles in 20 right-handed university aged, right-handed males (age: 22 � 3 years, weight: 77.5
� 11.1 kg, height 179.0 � 7.0 cm) participated in this study. Each participant performed 60 submaximal ex­
ertions (40N) directed at 4 off-axis phase angles of 45� (45� , 135� , 225� , and 315� ) in 3 planes (frontal, sagittal,
and transverse) in 5 hand locations within a right handed reach envelope. The influence of hand location and
force direction on muscle activity was evaluated with a forced-entry stepwise regression model. The ability of
previously published on-axis prediction equations to predict muscle activity during these off-axis exertions was
also evaluated. Within each muscle, activity levels were affected by both hand location and three-dimensional
force direction and activation levels ranged from <1 to 37 %MVE. For each force direction there were 75 pre­
dictive equations selected and used, and the specific equation that best predicted activation depended on the
muscle, exertion direction and hand location evaluated. This work assists ergonomic workplace design to
minimize muscle demands during commonly performed off-axis exertions. These estimated demands can be
employed to improve workplace design to reduce workplace injuries and enhance worker productivity.

1. Introduction The physical layout of different workplaces can create postural

constraints and dictate the muscular efforts required to perform work,
Manual material handling (MMH) involves movement of materials which can modulate injury risk. In pushing and pulling tasks, shoulder
by hand through the means of carrying, lifting, lowering, pushing, and exposures relate to the posture and the direction and magnitude of force
pulling (Snook, 1978). Manual labor is a major responsibility of workers exerted by the hand(s). Workplace constraints limit the postures in­
in industrial settings. Although the predominant concern in scientific dividuals can operate in, thereby affecting force production capability
research on MMH has been on lifting, carrying, and lowering tasks; and subsequent shoulder loads during tasks. Shoulder postures also in­
many of these tasks have been replaced by pushing and pulling tasks (El fluence the efficiency of exertion (Lee, 2007). Antony and Keir (2010)
Ouaaid et al., 2018). Upper extremity injuries have been a consequence reported an 84% mean increase in muscle activity when the arm is raised
of the substitution and accounted for approximately one-fifth of the total from 30 to 90� . Foot position with respect to the load also influence
number of lost time claims in Ontario, Canada in 2017 (Workplace capacity during pushing and pulling exertions (Yu et al., 2018). Studies
Safety and Insurance Board [WSIB], 2017). These upper extremity in­ have evaluated both shoulder strength and muscle activation in on-axis
juries are often attributed to awkward postures, repetitive movement directions with varying locations and magnitudes of force (e.g., Alasim
and high forces in occupational manual materials handling. Ergonomic et al., 2019; Cudlip et al., 2016; Cudlip et al., 2018; Cudlip and Dick­
design principles can be employed to modify these risk factors, thereby erson, 2018; Daams, 1992; McDonald et al., 2012; McDonald et al.,
reducing workplace injuries, promoting worker safety and enhancing 2014; Nadon et al., 2016; Perdeaux et al., 2010), however off-axis ex­
worker productivity. The redesign of workstations through ergonomic ertions may be more common in the workplace, as there may be
adaptations (e.g., better working heights, lighter hand tools) can reduce biomechanical advantages to incorporating off-axis forces even in tasks
the postural load on the shoulders (Westgaard and Aaras, 1985). that have a primary axis of force demand (Fischer et al., 2013; Borgs

* Corresponding author. University of Waterloo, Department of Kinesiology, 200 University Avenue West, Waterloo, ON, Canada, N2L3G1.
E-mail address: [email protected] (C.R. Dickerson).

https://doi.org/10.1016/j.ergon.2019.102892
Received 26 April 2019; Received in revised form 11 November 2019; Accepted 23 November 2019
Available online 3 December 2019
0169-8141/© 2019 Elsevier B.V. All rights reserved.
A.C. McDonald et al. International Journal of Industrial Ergonomics 75 (2020) 102892

et al., 2019). Table 1

Industrial workers are more likely to use sub-maximal forces in offset Electrode Placement and maximum voluntary exertion (MVE) for each muscle
directions in established job tasks. The ability to evaluate the postural recorded (Boettcher et al., 2008; Chopp et al., 2010; Cram and Kasman, 1998;
load on the shoulder when force generation is in an off-axis direction can Kelly et al., 1996).
help determine what workstation configuration will be of less physical Muscle Electrode Placement MVE Test
demand for the worker. Often, in static pushing and pulling, individuals Upper Trapezius Along the ridge of the Lying prone with shoulder
preferred to exert a force with an angular deviation to generate the shoulder, slightly lateral to abducted at 90� , elbows
maximum manual force (Borgs et al., 2019). Individuals also voluntarily and one-half the distance extended, and with the
deviated from the true push-pull direction when exerting force to in­ between C7 and the acromion thumb pointed downward.
Middle Trapezius Horizontally next to the Lying prone with shoulders
crease force capacity (Seo and Armstrong, 2009). Subject-chosen pos­ medial border of the spine of abducted at 120� , elbow
tures rendered pulling forces, at elbow and shoulder heights, and to be the scapula extended, and with the
twice as large as fixed postures, which included the body positioned thumb pointed upwards.
perpendicular to the force measurement device (Daams, 1993). Work Participant pushed upward
against resistance.
has been done to quantify and to predict the differences in strength
Lower Trapezius Approximately 5 cm down Lying prone with shoulder
production during off-axis exertions (La Delfa et al., 2014; La Delfa and from the scapular spine on an abducted to 90� , elbow
Potvin, 2016, 2017). This work shows the importance of force direction oblique angle extended, elbow extended,
on workers capabilities but does not give indications of the muscular and with the thumb pointed
demands associated with different exertion directions. Specific muscle upwards.
Participant pushed upward
contributions to exertions are critical for task design, job rotation and
against resistance.
return to work accommodations (Raina and Dickerson, 2009). There­ Biceps Brachii Center of the muscle belly on Elbow flexed at 90� .
fore, it is practical to investigate the shoulder muscle demands in off-axis the dorsal aspect of the upper Participant pushes upward
force exertions. The relationship between an off-axis force and its two arm against resistance.
Triceps Approximately 2 cm lateral Lying supine with shoulder
relevant components has not been determined but establishing a cor­
from the midline of the arm, and elbow flexed at 90� .
relation can enable extrapolation of available information to estimate roughly 50% of the distance Participant extends elbow
forces in offset directions. between the acromion and the against resistance.
The purpose of the current research study was to evaluate shoulder olecranon
muscle activation during seated, static, sub-maximal exertions in off- Sternal Insertion of Approximately 2 cm out from Lying supine with shoulder
Pectoralis Major the axillary fold, horizontally abducted at 30� , elbow
axis (non-cardinal) directions. This study expands an electromyo­
on the chest wall flexed to 90� .
graphic shoulder spatial map series of investigations. The activation Participant pushed
levels from previously recorded cardinal (on-axis) forces (McDonald horizontally upward against
et al., 2012, 2014; Nadon et al., 2016) provided known on-axis com­ resistance.
Clavicular Approximately 2 cm below Seated with shoulder flexed
ponents and be used to predict the corresponding unknown off-axis
Insertion of the clavicle, medial to the at 90� and adducted.
exertions. This data was used to identify the postural and directional Pectoralis Major axillary fold, on the chest wall
dependency of hand force on muscle activity, using relationships be­ at an oblique angle toward the
tween on-axis and off-axis exertions. It was hypothesized that the in­ clavicle
termediate muscular response would be between the values achieved for Infraspinatus Approximately 4 cm below Lying on one side with
the spine of the scapula, on elbow flexed at 90� and
adjacent force directions.
the lateral aspect, over the shoulder externally rotated.
infrascapular fossa
2. Methods Supraspinatus Directly above the spine of the Lying on one side with
scapula, over the shoulder abducted at 5� ,
suprascapular fossa elbow extended, and with
2.1. Subjects
the thumb pointed upward.
Participant pushed upward
Twenty university aged, right-handed males (age: 22 � 3 years, against resistance.
weight: 77.5 � 11.1 kg, height 179.0 � 7.0 cm) participated. Partici­ Anterior Deltoid Anterior aspect of the arm, Seated with shoulder flexed
pants were excluded if they had a history of musculoskeletal pain or approximately 4 cm below the at 90� .
clavicle
disorders in their right shoulder within the past year. The University of
Middle Deltoid Approximately 3 cm below Seated with shoulder
Waterloo Office of Research Ethics approved this study and participants the acromion on the lateral abducted to 90� , elbow fully
read and signed informed consent forms before the study proceeded. aspect of the upper arm extended, and with the
thumb pointed forward.
2.2. Instrumentation Posterior Deltoid Approximately 2 cm below Lying prone with elbow
the lateral border of the spine extended, shoulder
of the scapula on an oblique abducted to 90� and
Surface electromyography (EMG) data were recorded from elec­ angle toward the arm externally rotated.
trodes overlying 14 muscles (anterior, middle and posterior deltoids,
biceps, triceps, infraspinatus, supraspinatus, upper, middle and lower
trapezius, serratus anterior, latissimus dorsi and the sternal and clavic­ 2.3. Experimental protocol
ular insertions of the pectoralis major) on the participants right side.
Skin was shaved and cleansed with ethanol prior to electrode placement Participants performed three sets of 14 muscle-specific maximum
to reduce impedance and Ag–AgCl bipolar surface electrodes, with a voluntary exertions (MVE) in postures chosen to elicit maximal muscle
fixed inter-electrode distance of 20 mm, were placed parallel to muscle activity per relevant muscle (Boettcher et al., 2008; Chopp et al., 2010;
fibers (Table 1). Raw EMG signals were sampled at 1500 Hz and band- Cram and Kasman, 1998; Kelly et al., 1996) (Table 1). Participants were
pass filtered (10–500 Hz) (Noraxon Telemyo 2400R T2, Scottsdale, instructed to slowly ramp up to their maximum exertion and maintain it
AZ, USA). Signals were differentially amplified with a common mode for 5 s. Researchers provided manual resistance and used verbal
rejection ratio of >100 dB at 60 Hz with an input impedance of 100MΩ encouragement to ensure participants exerted their maximal efforts
and converted using a 16-bit A/D card with a �10 V range. during the trials (McNair, 1996). A minimum rest period of 2 min was
given between each MVE to avoid fatigue (Chaffin, 1975; McDonald
et al., 2017).

2
A.C. McDonald et al. International Journal of Industrial Ergonomics 75 (2020) 102892

Each participant performed submaximal exertions directed at 4 off- intervals and evaluated based on the criterion of minimizing the dif­
axis phase angles of 45� (45� , 135� , 225� , and 315� ) in 3 planes (fron­ ference between the predicted muscle activity level and the level of
tal, sagittal, and transverse) in 5 hand locations (high left (HL) [-20, 30, muscle activation measured in this study. Our findings showed that in
60] cm, low left (LL) [-20, 30, 20] cm, neutral (NT) [20, 30, 20] m, high >62% of hand locations, force directions and muscles evaluated, using
right (HR) [40, 30, 60] cm, and low right (LR) [40, 30, 20] cm) one on-axis equation or the other, predicted better than any iteration of
(Fig. 1a). Each location was represented using X, Y, and Z co-ordinates the combined equations. Based on these findings, for each off-axis
and were located with respect to the centre of each participant’s torso exertion, muscle activity was predicted with the contributing on-axis
at umbilicus height. A Motoman HP50N robotic arm (West Carrollton, equation that minimized the difference between the predicted and
OH) was pre-programmed to place the handle at the specified hand lo­ measured off-axis muscle activity values. The equations used for each
cations for each trial (Fig. 1b). Each participant completed a total of 60 hand location, force direction and muscle combination are listed in
exertions that were 40N each, a force level which was selected to Tables 2–6.
coincide with prior experimental data (McDonald et al., 2012, 2014;
Nadon et al., 2016), in order to leverage these foundational data. This 2.5. Statistical analysis
level was previously selected to provide a balance between occupa­
tionally relevant force levels without being high enough to cause fatigue To evaluate the effect of hand location and force direction on muscle
in a multi-task experimental protocol. A MSA-6 force transducer (AMTI, activity, a force-entry stepwise regression model was created for each of
Watertown, MA) measured hand forces and a custom Labview program the 14 muscles and the total muscle activity value. For each muscle the
(National Instruments, TX) provided participants with visual feedback X-, Y-, and Z-axis force directions and the X- and Z-axis hand location
on polar plots (Fig. 2). Once the participant achieved the specified force were included in the regression models. The influence of hand location
direction they were asked to maintain the static exertion for 3 s, there and force direction on muscle activity was evaluated with an alpha level
was no time limit in achieving the correct exertion. Block randomization of p < 0.05.
for each plane was used for the order of trials and the combination of To evaluate the effectiveness of this predictive method, the differ­
hand location, force magnitude, and force direction. ence between measured individual value and predicted value were
calculated for each participant and these differences were compared to a
2.4. Data analysis value of zero using t-tests with an alpha level of p < 0.05. The equations
that predicted values that were not significantly different than zero,
EMG signals were demeaned by subtracting the average of each trial were considered to have predicted muscle activity well. All statistical
from each point and heart rate contamination was removed using a high analyses were completed with JMP software.
pass 4th order Butterworth filter (fc ¼ 30 Hz) (Drake and Callaghan,
2006). EMG data were full-wave rectified and low pass filtered using a 3. Results
4th order Butterworth filter fc ¼ 4 Hz. Data from the trial exertions were
normalized to the muscle specific MVEs of each muscle recorded. All 3.1. Mean muscle activity
trials were performed until the participant achieved the required phase
and hand force for 3 s, and the average EMG (aEMG) was calculated over Across the tasks evaluated (hand locations and force direction),
a 2 s window (1–2s) of the 3 s window. For each exertion, a total muscle muscle activation ranged from <1 to 37 %MVE. Within each muscle,
activity score was generated by calculating an average of all 14 muscles, activity levels were affected by both hand location and three-
this was a general score indicative of overall effort for each exertion. dimensional force direction. The X-axis hand location was the least
Another goal of this investigation was to develop equations to predict likely to have influence on muscle activity, and did not influence acti­
muscle activity across hand locations and force directions. Initially, 90 vation in 7/15 muscles examined (posterior deltoid, biceps, triceps,
predictive equations from previously published muscle activity predic­ infraspinatus, latissimus dorsi, upper trapezius, total muscle) (p < 0.05).
tion equations for on-axis exertions in the up-down (Nadon et al., 2016), The Z-Axis location was more influential and had a significant effect in
left-right (McDonald et al., 2014) and fore-aft (McDonald et al., 2012) all muscles except the triceps (p < 0.05). All force directions signifi­
directions were combined to create equations to predict muscle activa­ cantly influenced most muscles activation levels, excluding the upper
tion in the off-axis force directions. For each off-axis direction, the two trapezius in the X-axis direction, the posterior deltoid and sternal head
contributing on-axis predictions were iteratively combined in 5% of pectoralis major in the Y-axis direction, and the triceps in the Z-axis

Fig. 1. (a) Red circles depict the 5 hand locations in [X, Y, Z] axes: high left (HL) [-20, 30, 60] cm, low left (LL) [-20, 30, 20] cm, neutral (NT) [20, 30, 20] cm, high
right (HR) [40, 30, 60] cm, and low right (LR) [40, 30, 20] cm (b) A motorman HP% robotic arm was used to position the MSA-6 force transducer in the 5 pre-
programmed hand locations. (For interpretation of the references to colour in this figure legend, the reader is referred to the Web version of this article.)

3
A.C. McDonald et al. International Journal of Industrial Ergonomics 75 (2020) 102892

Fig. 2. Hand forces (magnitude, direction) were measured with an MSA-6 force transducer (AMTI, Watertown, MA) and a custom Labview program (National
Instruments, TX) provided participants with visual feedback on polar plots.

Table 2
The average difference � standard deviation between the predicted and measured muscle activity in the High-Left (HL) hand location, for 15 muscles (total, anterior
deltoid (Adel), middle deltoid (Mdel), posterior deltoid (Pdel), biceps (Bi), triceps (Tri), infraspinatus (Infra), supraspinatus (Supra), clavicular (PecC) and sternal
(PecS) heads of pectoralis major, latissimus dorsi (Lats), serratus anterior (Sert) and lower (Ltrap), middle (Mtrap) and upper (Utrap) trapezius) in the left-down (LD),
left-back (LB), forward-left (FL), up-left (UL), down-back (DB), down-forward (DF), up-back (UB), up-forward (UF), down-right (DR), back-right (BR), forward-right
(FR) and up-right (UR) force directions. For each muscle and direction combination the on-axis equation used to predict activation is indicated in bold font. Exertion
conditions where the difference between the predicted and measured muscle activity values is not significantly different than 0 are shaded in grey, these were
considered to predict well.
HL LD LB FL UL DB DF UB UF DR BR FR UR

Total Down 12 Back Forward Up Back Forward Up 145 � Forward Right Back Forward Up 128 �
� 25 4�53 29 � 20 76.�49 77 � 34 32 � 24 63 19 � 32 49 � 20 33 � 7 � 28 58
39
Adel Left Back Left Left Back Forward Back Forward Right Back Right Right
6�3 5 � 10 6�3 10 � 11 6�5 3�3 13 � 14 5�5 0�2 1 � 11 2�3 14 � 11
Mdel Left Back Left Left Back Forward Back Forward Right Back Forward Right
2�2 3�7 1�2 6�7 4�3 1�2 6 � 10 3�4 1�2 1�6 3�4 8�8
Pdel Left Left Left Left Back Forward Back Forward Right Right Forward Up 9 � 11
0�1 1�1 0�1 1�1 6�0 0�3 2�4 0�4 0�3 0�4 6�7
Bi Down 2 � 6 Back Forward Up 2 � 4 Back Forward Back Forward Right Right Right Up 5 � 10
8�7 1�2 10 � 7 0�3 9�9 1�4 1�3 3�7 2�2
Tri Left Left Left Left Back Forward Back Forward Down Back Forward Up 1 � 7
2�2 1�2 4�4 1�2 5�1 3�4 3�2 6�6 6�3 4�2 4�6
Infra Left Back Left Up 0 � 7 Back Forward Up 11 � Up6�11 Down Back Forward Up 16 �
0�2 1�5 1�3 5�2 4�4 12 3�2 7�9 5�7 11
Supra Down 0 � 5 Back Left Up 7 � 4 Back Forward Up 12 � 8 Forward Right Back Right Up 12 � 9
1�6 0�2 4�3 2.�3 1�3 1�3 1.�4 1�2
PecC Left Back Left Up 16 � Down 26 � Forward Up 20 � Forward Right Back Right Up 2�7
3�8 10 � 17 7�7 15 10 1�4 14 0�4 1�3 9�10 0�1
PecS Left Left Left Left Back Forward Back Forward Right Right Right Right
0�5 0�6 7�2 5�4 7�8 3�2 10 � 2 4�1 0�2 1�1 0�1 1�1
Lats Left Left Left Left Back Forward Back Forward Right Right Right Up 0 � 4
2�1 0�2 3�1 0�2 13 � 1 3�3 10 � 3 2�2 1�2 1�2 3�2
Sert Left Back Left Up 20 � Back Forward Back Up 16 � Right Back Right Up 28 �
1�4 9�9 45 � 3 10 3�6 0�2 34 � 14 10 1�3 4�7 8�8 18
L Trap Down 7�3 Back Left Up 1 � 4 Back Forward Up 8 � 6 Forward Right Back Forward Up 16 � 8
2�5 5�3 2�4 7�5 2�6 4�4 6�5 3�7
M Down 2�4 Back Left Left Back Forward Back Up 7 � 3 Right Right Forward Right
Trap 2�4 1�2 3�3 5�2 5�5 1�5 1�4 2�2 4�4 10 � 10
U Trap Left Back Left Up 10 � 7 Back Forward Up 15 � Up 4 � 5 Right Back Right Up 11 � 9
3�4 4�8 4�5 1�2 11 1�2 0�6 0�2

direction (p > 0.05). As an example, mean activation in the three regions anterior deltoid muscle (F10 ¼ 117.74) is dependent on all of the location
of the deltoid muscle ranged from <1 to 29 %MVE (anterior (<1–29 % axes and all direction axes (p < 0.05) except for the Z axis direction (p >
MVE), middle (<1–14 %MVE), posterior (<1–17 %MVE) regions) 0.05). Activity in the sternal head of the pectoralis major (F10 ¼ 57.01)
(Fig. 3). The influence of hand location and exertion direction on muscle and the posterior deltoid (F10 ¼ 39.08) muscles are dependent on hand
activity levels is dependent on the muscle examined. Muscle activity in location in the Z axis and exertion directions in the X and Z axes (p <
the middle deltoid (F10 ¼ 75.33), supraspinatus (F10 ¼ 106.76), serratus 0.05). In the infraspinatus muscle (F10 ¼ 80.84), activity level is
anterior (F10 ¼ 134.46), the clavicular head of the pectoralis major (F10 dependent on exertion location in the Z axis and direction in the X and Y
¼ 84.65) and the lower (F10 ¼ 130.01) and middle portions of the axes and in the triceps muscle (F10 ¼ 29.28) it is dependent only on
trapezius (F10 ¼ 93.82) is influenced by exertion direction in the X, Y exertion direction in the X and Y axes (p < 0.05). Total muscle activity is
and Z axes and by hand location in the X and Z axes (p < 0.05). Muscle dependent on exertion direction in the X, Y and Z axes and by hand
activity in the biceps (F10 ¼ 30.86), latissimus dorsi (F10 ¼ 62.14) and location in the Z axes (p < 0.05, F10 ¼ 124.48).
upper trapezius (F10 ¼ 147.11) is also dependent on all of these factors
(p < 0.05) except for location in the X axis (p > 0.05). Activity in the

4
A.C. McDonald et al. International Journal of Industrial Ergonomics 75 (2020) 102892

Table 3
The average difference � standard deviation between the predicted and measured muscle activity in the High-Right (HR) hand location, for 15 muscles in the 12 force
directions. For each muscle and direction combination the on-axis equation used to predict activation is indicated in bold font. Exertion conditions where the difference
between the predicted and measured muscle activity values is not significantly different than 0 are shaded in grey, these were considered to predict well (see Table 2 for
description of abbreviations).
HR LD LB FL UL DB DF UB UF DR BR FR UR

Total Down 34 � Back Forward Up 94 � Back Forward Up 159 � Forward Down 27 � Back Forward Up 105 �
24 29 � 6 � 39 67 73 � 51 � 24 70 2 � 32 24 31 � 45 � 26 47
33 23 65
Adel Left Back Forward Left Back Forward Back Forward Right Right Right Up 5 � 2
2�2 6�3 3�4 10 � 8 3�1 5�2 8�5 4�4 0�1 1�4 2�2
Mdel Left Back Forward Up 14 � 7 Back Forward Up 16 � 10 Up 9 � 4 Right Back Right Up 9 � 6
1�3 0�6 3�4 5�3 2�3 0�3 7 � 12 1�3
Pdel Left Left Left Left Back Forward Back Forward Right Back Right Right
1�2 1�2 1�2 1�2 10 � 3 3�2 7�7 3�1 1�3 2 � 11 4�1 1�4
Bi Left Back Forward Up 9 � 9 Back Forward Back Forward Right Right Right Up 2 � 4
1�6 2�5 5�5 5�3 2�3 2�5 0�3 1�2 4�5 1�2
Tri Left Left Left Left Back Forward Back Forward Down 1 � 6 Back Forward Up 2 � 6
1�2 0�1 1�1 1�2 32 � 2 1�3 2�3 4�1 5�7 1�5
Infra Left Left Left Up 5 � 7 Back Forward Up 22 � 15 Forward Down 1 � 6 Back Right Up 19 � 15
0�3 6�4 2�4 10 � 6 5�2 0�5 3 � 10 1�4
Supra Left Left Forward Up 7 � 7 Back Forward Up 17 � 12 Forward Right Back Right Up 8 � 6
1�4 6�5 1�5 10 � 3 3�3 2�4 3�2 0 � 11 2�3
PecC Left Left Forward Up 1 � 12 Back Forward Back Up 8 � 9 Right Right Right Up 0 � 1
0�6 1�7 7�8 4�1 0�5 0�7 0�1 1�1 1�2
PecS Left Left Left Left Back Forward Back Forward Right Right Right Right
2�2 4�1 0�3 3�1 6�0 2�2 5�1 5�1 0�0 0�1 0�1 1�1
Lats Left Left Left Left Back Forward Back Forward Down 3 � 5 Right Right Up 0 � 3
2�5 2�1 3�1 1�2 12 � 3 4�2 12 � 3 3�2 5�4 0�4
Sert Left Back Forward Up 18 � Back Forward Up 26 � 12 Forward Right Back Forward Up 20 � 10
0�3 7�5 8�7 10 3�3 2�2 12 � 6 2�3 5�7 1�3
L Trap Left Back Forward Up 8 � 7 Back Forward Up 23 � 14 Forward Down 2 � 6 Back Right Up 22 � 12
2�5 5�5 0�7 6�7 5�5 2�6 11 � 2�5
12
M Left Left Left Up 5 � 8 Back Forward Up 18 � 12 Forward Right Back Right Up 11 � 7
Trap 3�6 8�5 5�6 9�5 5�6 2�4 1�5 5 � 11 4�
U Trap Left Back Forward Up 10 � 9 Back Forward Up 17 � 12 Up 6 � 5 Right Back Right Up 8 � 7
1�4 2�7 4�5 5�3 2�2 1�2 1�8 1�3

3.2. Predictive effects effectively predicted, in 67% of hand locations (Table 3). The supra­
spinatus and posterior deltoids are the most often effectively predicted
Each muscle, hand location and force direction was predicted using muscles in the low left and right locations respectively (47%) (Tables 4
equations from previously published on-axis muscle activity prediction and 5). In the neutral hand position the clavicular head of the pectoralis
equations (McDonald et al., 2012, 2014; Nadon et al., 2016). The spe­ major is predicted effectively in 53% of force directions (Table 6).
cific equation required to best predict muscle activity depended on the
muscle, exertion direction and hand location (Tables 2–6). For each 4. Discussion
force direction there were 75 predictive equations selected and utilized.
In three of the four exertions force directions with a leftward force The purpose of this investigation was to identify the postural and
component, the leftward equation was more frequently used than the directional dependency of hand force on muscle activity for 14 upper
other directions (number of leftward equations (/75): left-down (LD), extremity muscles during seated, static, sub-maximal, 40N, off-axis ex­
60; left-back (LB), 49; left-up (LU), 48; left-forward (LF) 35). The ertions within the right-handed reach envelope. Across tasks, muscle
rightward direction equation was more frequently used in three of the activation ranged from <1 to 37 %MVE and the activation levels
four exertions with a rightward force component (number of rightward depended on both hand location and force direction. Each off-axis
equations (/75): right-down (RD), 53; right-forward (RF), 44; right-up exertion was a combination of two equal cardinal components and
(RU), 40; right-back (RB), 35). Muscle activity in the remaining four muscle activity during these off-axis exertions was sufficiently predicted
force directions were most often best predicted with the forward with previously published on-axis prediction equations from one of the
(up-forward (UF), 60; down-forward (DB), 60) and backward two force components of each off-axis exertion.
(up-backward (UB), 58; down-backward (DB), 57) prediction equations. Hand location significantly influenced levels of muscle activation
The difference between the participants measured muscle activity across the shoulder complex, dependent on the individual muscles
and predicted muscle activity values for each exertion were compared to examined. The up/down hand location (Z axis) influenced activation for
a value of zero. Predictions that were not significantly different than all muscles except the triceps brachii muscle. The fixed fore/aft position
0 were considered to be the most effective predictors. Across all hand may have limited the changes in triceps line of action throughout the
locations, the high left and high right locations had the most equations range of examined postures. In prior investigations of both static and
considered to be effective, 44% in each location (Tables 2–6). The low dynamic pushing and pulling efforts, elevating the limb also modified
left, low right and neutral positions had effective equations for 33–37% muscle demands, increasing demand with increased elevation (Brook­
of exertions (Tables 2–6). The muscles with the most number of effec­ ham et al., 2010; McFarland et al., 2018). There was more variability in
tively predicting equations across force directions depends on hand the influence of the left/right (X axis) position on specific muscle acti­
location. In the high left location, the posterior deltoid and biceps vation, affecting just over half the muscles examined (anterior deltoid,
muscles are effectively predicted in 53% of the force direction (Table 2). middle deltoid, supraspinatus, pectoralis major (sternal, clavicular),
In the high right hand location, the triceps muscle is most often serratus anterior, trapezius (lower, middle)). Muscle segments are

5
A.C. McDonald et al. International Journal of Industrial Ergonomics 75 (2020) 102892

Table 4
The average difference � standard deviation between the predicted and measured muscle activity in the Low-Left (LL) hand location, for 15 muscles in the 12 force
directions. For each muscle and direction combination the on-axis equation used to predict activation is indicated in bold font. Exertion conditions where the difference
between the predicted and measured muscle activity values is not significantly different than 0 are shaded in grey, these were considered to predict well (see Table 2 for
description of abbreviations).
LL LD LB FL UL DB DF UB UF DR BR FR UR

Total Left Back Forward Up 57 � Back Forward Back Forward Down 47 � Right Forward Up 44 �
37 � 15 1�3 30 � 12 32 74 � 16 28 � 14 39 � 11 � 26 48 34 � 3�20 22
31 12
Adel Left Left Left Left Back Forward Back Forward Down 2�3 Right Forward Right
5�2 1�1 1�2 5�5 5�1 4�2 3�4 4�3 2�2 1�3 8�4
Mdel Left Left Forward Left Back Forward Back Forward Right Right Forward Right
1�1 0�0 1�1 3�3 3�1 2�3 2�1 2�2 1�7 3�0 3�4 1�2
Pdel Left Left Left Left Back Forward Back Forward Right Right Forward Right
0�1 7�5 0�1 0�1 5�1 1�3 5�0 0�3 4�8 3�1 3�4 2�2
Bi Down Back Forward Up 3�4 Back Forward Back Forward Right Right Right Right
1�2 4�2 1�3 11 � 4 2�2 5�6 1�3 2�2 2�4 1�2 2�3
Tri Left Left Left Left Back Forward Back Forward Down 9�4 Back Forward Up 2�1
5�3 0�1 2�2 0�1 1�4 1�5 5�1 7�2 6�1 5�3
Infra Left Left Left Left Back Forward Back Up 10 � Down 1�5 Back Forward Up 3 � 5
0�2 1�1 4�3 3�3 3�1 5�4 2�2 10 3�1 3�8
Supra Down 2 � Back Forward Left Back Forward Back Forward Down 2�7 Right Forward Right
1 6 � 10 4�1 2�1 1�1 2�2 1�2 2�2 0�1 1�2 1�2
PecC Left Left Forward Up 8 � 16 Down 17 � Forward Back Up 1�8 Right Right Right Up 2�7
17 � 6 5 � 14 4�6 6 3�2 0 � 11 0�1 6�4 1�1
PecS Left Left Forward Left Back Forward Back Forward Right Right Right Right
8�4 1�2 9�4 0 � 12 9�4 4�1 4�11 2�2 0�0 1�1 0�1 1�2
Lats Left Left Forward Left Back Forward Back Forward Right Right Right Right
2�1 2�5 2�1 1�2 1�2 3�2 15 � 1 3�2 3�3 1�2 1�2 1�1
Sert Down Back Left Left Back Forward Back Forward Right Back Forward Right
2�3 4�1 1�2 3�4 4�2 0�1 0�3 6�4 2�1 4�2 3�3 4�3
L Trap Down Left Left Left Down Forward Back Up 11 � 5 Down 2 � 4 Right Forward Right
6�1 2�1 6�2 6�3 5�2 7�3 8�4 4�3 3�5 9�5
M Left Back Left Left Back Forward Back Up 8 � 3 Down 0 � 9 Right Forward Right
Trap 2�2 1�2 2�1 3�1 1�2 2�3 1�3 2�1 1�5 1�2
U Trap Down Back Forward Left Back Forward Back Forward Down 2�5 Back Forward Right
3�2 41 � 1�1 3�2 2�1 1�1 0�2 2�3 1�1 0�2 3�2
28

influenced by their line of action and moment arms, in relation to the muscle fatigue, reducing force generating capacity and cascading into
intended movement (Brown et al., 2007). In this investigation, hand additional changes in activation pattern (Ebaugh et al., 2005, 2006;
position influenced posture, changing lines of action, which partly Endo et al., 2001; Halder et al., 2000; McQuade et al., 1998; Picco et al,
explain the diverse muscle responses across conditions as the external 2010). The combined consideration of the hand location and force di­
force, joint axes, and moment arms interact in the generation of mo­ rection is an essential component of a workplace design strategy to
ments, affecting muscular demand and level of activation. In comparison mitigate shoulder muscle demands.
to previous work examining 40N on-axis exertions in the up/down and Equations capable of predicting approximate muscle demands
left/right directions, the individual muscle efforts to complete these throughout the range of possible force directions are important for
off-axis exertions (1–37 %MVE) were greater, with activations ranging workplace design goals of limiting muscle activation and it is an
from 1.6 to 29.5 %MVE in the up/down axis and 1–27 %MVE in the essential expansion to the previous on-axis work that generated pre­
left/right axis. Similarly, hand location affects muscle activation during dictive equations for on-axis exertions (McDonald et al., 2012, 2014;
40 N exertions in overhead postures (Cudlip et al., 2016). Combined Nadon et al., 2016). These individual on-axis equations were able to
with the previous work examining on-axis demands, these findings predict activation during the off-axis exertions more effectively than
support the fundamental consideration of hand location for lowering iteratively combining them to generate new equations. This can likely be
muscular demands during off-axis exertions in the workplace. explained by individual muscle functions and lines of actions dictating
Similar to hand location, the influence of force direction varied the dominant contributions to specific actions within their predominant
across individual muscles. The X axis component (left/right) influenced functions. For example, the middle deltoid muscle is a strong arm
all muscles except the upper trapezius and the Y axis component (fore/ abductor and the majority of postures and exertions (72%) with a
aft) affected all muscles except the posterior deltoid and the sternal head rightward (abducting) component, the dominant predictive equation
of the pectoralis major. Comparable to the Z axis (up/down) component was in the rightward direction. Another example of this occurs for the
of the hand location, the up/down component of the force direction pectoralis major. The sternal head of the pectoralis major is a strong
affected all muscles except the triceps brachii. Picco et al., 2010 adductor of the arm, and this is reflected in the leftward equation being
observed exertion direction to have the greatest effect on muscle activity the dominant component in predicting this muscles activation in 96% of
and influence on hand force capacity. The maximal force capable was exertions that contained a leftward force component. The clavicular
greatest in a vertical downward direction, followed by upward then head largely contributes to humeral flexion, and this function is re­
medial and lateral exertions (Picco et al., 2010). Due to the anatomical flected in the upward equation being predominant in the up-left force
arrangement of the shoulder complex and its wide range of functions, direction in 4/5 postures. Using these predictive equations to estimate
muscle activation levels have multiple mechanical effects. Muscle acti­ muscle demands during commonly performed off-axis exertions can give
vation levels influence scapular position, sub-acromial space width, workplace designers insight into the best hand position and force di­
range of motion and force production capabilities. Working at higher rection combinations to control shoulder demands.
levels of muscle activation during repetitive work tasks can induce

6
A.C. McDonald et al. International Journal of Industrial Ergonomics 75 (2020) 102892

Table 5
The average difference � standard deviation between the predicted and measured muscle activity in the Low-Right (LR) hand location, for 15 muscles in the 12 force
directions. For each muscle and direction combination the on-axis equation used to predict activation is indicated in bold font. Exertion conditions where the difference
between the predicted and measured muscle activity values is not significantly different than 0 are shaded in grey, these were considered to predict well (see Table 2 for
description of abbreviations).
LR LD LB FL UL DB DF UB UF DR BR FR UR

Total Left Left Forward Left Back Forward Up 83.6 � Forward Down 19 � Back Forward Right
45 � 19 38 � 12 7�21 43 � 13 38 � 24 39 � 12 13.4 14 � 26 32 34 � 42 10 � 21 120 � 44
Adel Left Left Forward Left Back Forward Back Forward Right Right Forward Right
1�1 0�0 1�3 4�2 0�0 5�1 0.3 � 0.3 1�3 1�1 0�1 2�4 3�5
Mdel Left Left Forward Left Back Forward Back Forward Right Back Right Right
1�1 1�0 0�2 3�2 4�2 1�1 3.1 � 1.6 6�4 2�3 5�5 0�3 3�7
Pdel Left Left Left Left Back Forward Back Forward Right Back Right Right
0�1 1�1 0�0 0�1 9�5 1�1 10.7 � 2.0 0�1 0�4 1 � 10 5�2 0�5
Bi Down 1 � 2 Left Forward Up 2 � 4 Back Forward Back Up 5 � 5 Right Right Right Up 5 � 6
1�1 1�4 3�1 4�2 0.6 � 2.5 3�2 2�3 2�3
Tri Left Left Left Left Back Forward Back Forward Down 4�7 Back Forward Up 1 � 3
4�3 3�2 1�1 1�1 2�7 1�2 11.4 � 1.2 1�1 1�8 1�1
Infra Left Left Left Left Back Forward Up 9.7 � 4.1 Forward Down 5 � 6 Back Forward Right
0�1 0�1 1�2 0�1 5�3 4�2 2�4 11 � 8 5�6 12 � 8
Supra Left Left Forward Left Back Forward Back Forward Right Back Right Right
3�0 1�0 2�2 2�1 3�1 3�1 0.4 � 1.8 1�2 2�2 5�4 4�2 6�6
PecC Left Left Forward Up 4 � 5 Back Forward Back Up 5 � 6 Right Right Right Right
1�5 2�2 5�6 2�1 1�2 2.6 � 0.6 1�1 1�1 1�1 1�1
PecS Left Left Left Left Back Forward Back Forward Right Right Right Right
0�8 5�3 3 � 11 5�4 6�1 3�3 6.8 � 0.4 6�2 0�1 0�1 0�1 0�1
Lats Left Left Left Left Back Forward Back Forward Right Back Right Right
0�5 3�5 7�2 8�1 5�9 4�2 19.4 � 3.6 4�1 8�5 9�7 1�1 1�2
Sert Down Left Forward Left Back Forward Back Forward Right Right Forward Right
4�3 3�2 5�4 4�3 2�2 2�2 0.0 � 0.6 6�3 2�1 2�1 4�3 7�3
L Trap Down Back Forward Left Back Forward Back Forward Down 3 � 5 Back Forward Right
5�1 4�1 6�3 10 � 3 1�2 8�3 3.0 � 3.6 1�6 12 � 7 2�6 15 � 8
M Trap Left Left Left Left Back Forward Back Forward Right Back Right Right
1�2 1�1 1�1 1�2 4�6 4�2 0.1 � 4.5 1�4 0�7 12 � 10 2�4 9�9
U Trap Down Left Forward Left Back Forward Back Forward Right Right Forward Right
2�0 2�0 0�3 3�2 0�1 1�0 0.6 � 0.9 3�3 0�1 2�3 2�2 5�5

Table 6
The average difference � standard deviation between the predicted and measured muscle activity in the Neutral Position (NP) hand location, for 15 muscles in the 12
force directions. For each muscle and direction combination the on-axis equation used to predict activation is indicated in bold font. Exertion conditions where the
difference between the predicted and measured muscle activity values is not significantly different than 0 are shaded in grey, these were considered to predict well (see
Table 2 for description of abbreviations).
NP LD LB FL UL DB DF UB UF DR BR FR UR

Total Left Left Forward Up 49 � 17 Back Forward Back Forward Down 42 � 15 Back Forward Up 79 � 23
31 � 13 31 � 13 20 � 17 63 � 10 30 � 18 42 � 17 3�19 27 � 17 6�21
Adel Left Left Left Left Back Forward Back Forward Right Back Right Right
2�1 2�1 1�2 3�3 2�0 5�1 2�3 1�3 1�1 1�1 2�2 3�3
Mdel Left Left Forward Left Back Forward Back Forward Right Right Right Right
2�2 1�2 0�1 2�2 3�1 1�1 3�1 2�3 1�3 0�3 2�1 1�2
Pdel Left Left Left Left Back Forward Back Forward Right Right Right Right
0�1 0�1 0�1 1�0 7�1 0�1 9�1 1�1 1�2 0�2 3�2 3�2
Bi Left Left Forward Up 1�3 Back Forward Back Forward Right Right Right Up 0 � 5
0�2 2�3 0�2 7�2 1�2 1�4 2�2 2�2 1�3 0�3
Tri Left Left Left Left Back Forward Back Forward Down 5�4 Back Forward Up 1�1
4�2 0�1 0�1 0�1 2�2 3�4 5�1 4�1 3�2 1�4
Infra Left Left Left Up 1�1 Back Forward Back Forward Down 2 � 3 Back Forward Up 11 � 8
1�1 0�1 1�2 7�2 5�2 4�2 1�3 1�5 4�6
Supra Left Left Forward Left Back Forward Up 4 � 2 Forward Right Right Right Right
1�1 2�2 1�2 3�2 4�1 2�2 0�2 2�2 1�2 1�2 2�2
PecC Left Left Left Left Back Forward Back Up 3 � 6 Right Right Right Right
4�3 3�3 1�6 1�6 6�1 2�2 2�4 0�1 0�1 1�1 1�2
PecS Left Left Left Left Back Forward Back Forward Right Right Right Right
5�2 6�2 3�3 5�2 7�1 4�1 7�1 4�1 0�1 0�0 0�0 0�1
Lats Left Left Left Left Back Forward Back Forward Right Right Right Right
2�3 2�2 5�2 5�1 11 � 4 3�3 19 � 2 4�1 7�4 6�4 0�2 0�2
Sert Left Left Left Left Back Forward Back Forward Right Back Right Right
1�2 0�1 3�3 3�2 1�1 1�1 0�1 5�2 1�1 1�1 1�2 3�3
L Trap Left Back Left Left Back Forward Back Up 9 � 4 Right Back Right Right
1�2 3�3 3�3 6�4 3�2 8�2 5�6 1�4 6�5 3�4 8�6
M Trap Left Left Left Up 2 � 2 Back Forward Back Forward Right Right Right Up 11 � 5
1�2 1�2 2�3 6�3 3�4 4�3 1�3 0�5 3�5 2�4
U Trap Left Back Forward Left Back Forward Back Forward Right Back Right Right
1�1 0�2 0�2 4�2 1�1 1�1 2�3 2�2 1�1 0�2 1�2 3�3

7
A.C. McDonald et al. International Journal of Industrial Ergonomics 75 (2020) 102892

Fig. 3. Average muscle activity (%MVE) for the anterior (black, ADEL), middle (grey, MDEL) and posterior (white, PDEL) deltoids during exertions in 12 directions
(back-left (BL), back-right (BR), down-back (DB), down-forward (DF), down-left (DL), down-right (DR), forward-left (FL), forward-right (FR), up-back (UB), up-
forward (UF), up-left (UL), up-right (UR)), and in 5 hand locations (a) high-left, (b) high-right, (c) lower-left, (d) lower-right and (e) neutral position.

4.1. Limitations postures used both within and between individuals. In alignment with
the previously published on-axis studies, the protocol did not dictate the
Limitations of this investigation should be considered when applying arm postures used during the trials as five locations were fixed to an
these results. This study used a convenience sample of twenty right- absolute coordinate system. The participants were instructed to sit up­
handed males within a university setting. The participants were be­ right with the midline of their body aligned with the robot. Although this
tween the ages of 19 and 35 so the data cannot easily be generalized to likely contributed to variability, it is more realistic to a workplace sce­
other populations. There was also variability in self-selected body nario where workers often employ different postures and movements to

8
A.C. McDonald et al. International Journal of Industrial Ergonomics 75 (2020) 102892

complete required tasks. The third aspect to consider is the limited range Daams, B.J., 1992. Comparison of moments exerted round one or more joints in the arm.
Contemporary ergonomics. In: Proceedings of the Ergonomics Society’s Annual
of force type, magnitude and direction evaluated. Only one force level
Conference, vol. 36, pp. 104–109.
(40N) was included; this aligned with previously published on-axis Daams, B.J., 1993 Apr. Static force exertion in postures with different degrees of
literature, but caution should be taken when extended these findings freedom. Ergonomics 36 (4), 397–406.
to other force magnitudes, despite evidence of proportionate scaling of Drake, J.D., Callaghan, J.P., 2006. Elimination of electrocardiogram contamination from
electromyogram signals: an evaluation of currently used removal techniques.
muscular demands at occupationally important forces (20–60N) (Mes­ J. Electromyogr. Kinesiol. 16 (2), 175–187.
zaros et al., 2018). Ebaugh, D.D., McClure, P.W., Karduna, A.R., 2006. Scapulothorasic and glenohumeral
kinematics following an external rotation fatigue protocol. J. Orthop. Sport. Phys.
Ther. 16, 224–235.
5. Conclusions Ebaugh, D.D., McClure, P.W., Karduna, A.R., 2005. Three-dimensional scapulothoracic
motion during active and passive arm elevation. Clin. Biomech. 20, 700–709.
Fundamental insights into directional dependency of muscular de­ El Ouaaid, Z., Shirazi-Adl, A., Plamondon, A., 2018. Trunk response and stability in
standing under sagittal-symmetric pull-push forces at different orientation,
mands during off-axis exertions were achieved. Average total and spe­ elevations and magnitudes. J. Biomech. 70, 166–174.
cific shoulder muscle activities depended on both hand locations and Endo, K., Ikata, T., Katoh, S., Takeda, Y., 2001. Radiographic assessment of scapular
force direction, in a complex manner that varied by muscle. Muscle rotational tilt in chronic shoulder impingement syndrome. J. Orthop. Sci. 6, 3–10.
Fischer, S.L., Picco, B.R., Wells, R.P., Dickerson, C.R., 2013. The roles of whole body
activity was predictable for submaximal (40N), seated exertions in off- balance, shoe-floor friction, and joint strength during maximum exertions: searching
axis directions from previously published prediction equations for one for the “Weakest Link”. J. Appl. Biomech. 29 (1), 1–11.
of the two adjacent on-axis force components. The consideration of Halder, A.M., Itoi, E., An, K., 2000. Anatomy and biomechanics of the shoulder. Orthop.
Clin. N. Am. 31, 159–176.
shoulder anatomy, especially muscular line of action, informs muscular
Kelly, B.T., Kadrmas, W.R., Kirkendall, D.T., Speer, K.P., 1996. Optimal normalization
strategy. Muscular demands of right-hand pushing and pulling tasks in tests for shoulder muscle activation: an electromyographic study. J. Orthop. Sport.
all three planes can be estimated to improve workstation designs, lower Phys. Ther. 14, 647–653.
muscular demands and reduce musculoskeletal injury risk. La Delfa, N.J., Freeman, C.C., Petruzzi, C., Potvin, J.R., 2014. Equations to predict female
manual arm strength based on hand location relative to the shoulder. Ergonomics 57
(2), 254–261.
Acknowledgements La Delfa, N.J., Potvin, J.R., 2016. Multidirectional manual arm strength and its
relationship with resultant moment and arm posture. Ergonomics 59 (12),
1625–1636.
This project was partially funded through an NSERC Discovery Grant La Delfa, N.J., Potvin, J.R., 2017. The ‘arm force field’ method to predict manual arm
held by Dr. Clark Dickerson, 311895-2011. Equipment used in the grant strength based on only hand location and force direction. Appl. Ergon. 59, 410–421.
was funded through combined support from the Canada Foundation for Lee, T., 2007. Pushing strengths under restricted space. Hum. Factors Ergon. Manuf. 17,
95–102.
Innovation and the Ontario Research Fund. Dr. Dickerson is also McDonald, A., Picco, B., Belbeck, A., Chow, A.Y., Dickerson, C.R., 2012. Spatial
partially supported as the Canada Research Chair in Shoulder Mechanics dependency of shoulder muscle demands in horizontal pushing and pulling. Appl.
by NSERC. Ergon. 43, 971–978.
McDonald, A.C., Brenneman, E.C., Cudlip, A.C., Dickerson, C.R., 2014. The spatial
dependency of shoulder muscle demands for seated lateral hand force exertions.
Appendix A. Supplementary data J. Appl. Biomech. 30, 1–11.
McDonald, A.C., Sonne, M.W.L., Peter, J., Keir, P.J., 2017. Optimized maximum
voluntary exertion protocol for normalizing shoulder muscle activity. Int. Biomech.
Supplementary data to this article can be found online at https://doi.
4, 9–16.
org/10.1016/j.ergon.2019.102892. McFarland, D.C., Poppo, M.N., McCain, E.M., Saul, K.R., 2018. Spatial dependency of
shoulder muscle demand during dynamic unimanual and bimanual pushing and
References pulling. Appl. Ergon. 73, 199–205.
McNair, P.J., 1996. Verbal encouragement: effects on maximum effort voluntary muscle
action. Br. J. Sports Med. 30 (3), 243–245.
Alasim, H.N., Nimbarte, A.D., Jaridi, M., 2019. Impact of pulling direction and Meszaros, K.A., Vidt, M.E., Dickerson, C.R., 2018. CR Dickerson The effects of hand force
magnitude of force exertion on the activation of shoulder muscles international. variation on shoulder muscle activation during submaximal exertions. Int. J. Occup.
J. Ind. Ergonom. 69, 14–22. Saf. Ergon. 24 (1), 100–110.
Antony, N.T., Keir, P.J., 2010. Effects of posture, movement and hand load on shoulder McQuade, K.J., Dawson, J., Smidt, G.L., 1998. Scapulothoracic muscle fatigue associated
muscle activity. J. Electromyogr. Kinesiol. 20 (2), 191–198. with alterations in scapulohumeral rhythm kinematics during maximum resistive
Boettcher, C.E., Ginn, K.A., Cathers, I., 2008. Standard maximum isometric voluntary shoulder elevation. J. Orthop. Sport. Phys. Ther. 28, 74–80.
contraction tests for normalizing shoulder muscle EMG. J. Orthop. Sport. Phys. Ther. Nadon, A.L., Vidt, M.E., Chow, A.Y., Dickerson, C.R., 2016. The spatial dependency of
26, 1591–1597. shoulder muscular demands during upward and downward exertions. Ergonomics 59
Borgs, S.P., La Delfa, N.J., Dickerson, C.R., 2019. An evaluation of off-axis manual forces (10), 1294–1306.
and upper extremity joint moments during unilateral pushing and pulling. Perdeaux, K.M., Fischer, S.L., Dickerson, C.R., 2010. Isometric strength and
Ergonomics 62 (1), 52–64. corresponding muscle activation as a function of hand location and force direction:
Brookham, R.L., Wong, J.M., Dickerson, C.R., 2010. Upper limb posture and submaximal resolving the role of individual muscles on strength capacity. Occup. Ergon. 9,
hand tasks influence shoulder muscle activity. Int. J. Ind. Ergon. 40, 337–344. 141–151.
Brown, J., Wickham, J., McAndrew, D., Xu-Feng, H., 2007. Muscles within muscles: Picco, B.R., Fischer, S.L., Dickerson, C.R., 2010. Quantifying scapula orientation and its
coordination of 19 muscle segments within three shoulder muscles during isometric influence on maximal hand force capability and shoulder muscle activity. Clin.
motor tasks. J. Electromyogr. Kinesiol. 17, 57–73. Biomech.(Bristol, Avon) 25, 29–36.
Chaffin, D.B., 1975. Ergonomics guide for the assessment of human static strength. Am. Raina, S., Dickerson, C.R., 2009. The influence of job rotation and task order on muscle
Ind. Hyg. Assoc. J. 36 (7), 505–511. fatigue: a detoid example. Work 34, 205–213.
Chopp, J.N., Fischer, S.L., Dickerson, C.R., 2010. On the feasibility of obtaining multiple Seo, N.J., Armstrong, T.J., 2009. Biomechanical analysis for handle stability during
muscular maximal voluntary excitation levels from test exertions: a shoulder maximum push and pull exertions. Ergonomics 52 (12), 1568–1575.
example. J. Electromyogr. Kinesiol. 20 (5), 896–902. Snook, S.H., 1978. The design of manual handling tasks. Ergonomics 21, 963–985.
Cudlip, A.C., Meszaros, K.A., Dickerson, C.R., 2016. The influence of hand location and Westgaard, R.H., Aaras, A., 1985. The effect of improved workplace design on the
force direction on shoulder muscular activity in females during nonsagittal development of work-related musculo-skeletal illnesses. Appl. Ergon. 16 (2), 91–97.
multidirectional overhead exertions. Hum. Factors 58 (1), 120–139. By the Numbers: WSIB Statistical Report, 2017. Workplace Safety and Insurance Board
Cudlip, A.C., Holmes, M.W.R., Callaghan, J.P., Dickerson, C.R., 2018. The effects of [WSIB].
shoulder abduction angle and wrist angle on upper extremity muscle activity in Yu, D., Xu, X., Lin, J., 2018. Impact of posture choice on one-handed pull strength
unilateral right handed push/pull tasks. Int. J. Ind. Ergon. 64, 102–107. variations at low, waist, and overhead pulling heights. Int. J. Ind. Ergon. 64,
Cudlip, A.C., Dickerson, C.R., 2018. Female maximal push/pull strength capabilities by 226–234.
humeral abduction angle in bilateral exertions. Appl. Ergon. 70, 136–141.
Cram, J.R., Kasman, G.S., 1998. Introduction to Surface Electromyography. Aspen
Publications, Inc, Maryland.