University of Southern California Dissertations and Theses

Lower extremity control and dynamics during the golf swing

(USC Thesis Other)

Lower extremity control and dynamics during the golf swing

PDF

Download a page range

Download transcript

Copy asset link

Request this asset

Transcript (if available)

Content Page 1 of 106

Lower Extremity Control and Dynamics
during the Golf Swing

Travis J. Peterson

Dissertation Chair: Jill L. McNitt-Gray

Dissertation Committee Members: David D’Argenio, Philip Requejo, Rand R. Wilcox

Degree Conferred: Doctor of Philosophy

Conferring Program: Biomedical Engineering

Conferring Body: Faculty of the USC Graduate School

University of Southern California

Defense Date: May 31, 2017

Degree Conferral Date: August 2017

Page 2 of 106

Dedication
This dissertation is dedicated to my wife, Mandy. You have always encouraged me to pursue my
dreams, and I could not have accomplished this goal without your love and support.
Page 3 of 106

Table of Contents

: Introduction .............................................................................................................. 6
Mechanics of the Golf Swing ................................................................................................... 7
Regulation Within and Between Clubs ..................................................................................... 9
Proposed Set of Experiments .................................................................................................. 9
References ............................................................................................................................ 10
: Specific Aims and Hypotheses ...............................................................................11
Specific Aim 1: Angular Impulse and Balance Regulation during the Golf Swing ................... 11
Specific Aim 2: Angular Impulse and Balance Control while Increasing Golf Shot Distance
within a Club .......................................................................................................................... 11
Specific Aim 3: Angular Impulse and Balance Control during the Golf Swing with the 9-iron
and 6-iron .............................................................................................................................. 12
Specific Aim 4: Angular Impulse and Balance Control with Modified Address Positions
within a Club .......................................................................................................................... 12
Specific Aim 5: Lower Extremity Joint Control during the Golf Swing ..................................... 13
Specific Aim 6: Regulation of Muscular Control while Regulating Golf Shot Distance
between Clubs ....................................................................................................................... 13
References ............................................................................................................................ 15
: Experimental Design ..............................................................................................16
Experimental Procedures ...................................................................................................... 16
Inclusion Criteria .................................................................................................................... 16
Exclusion Criteria .................................................................................................................. 16
Overview of Participant Tasks ............................................................................................... 16
Calibration Tasks ...............................................................................................................16
Experimental Tasks ...........................................................................................................16
Data Collection ...................................................................................................................... 17
Kinematics .........................................................................................................................17
Kinetics ..............................................................................................................................19
Lab Setup and Coordinate System ........................................................................................ 20
Muscle Recruitment and EMG ............................................................................................... 20
Data Processing and Analysis ............................................................................................... 20
Kinematics .........................................................................................................................20
Kinetics ..............................................................................................................................21
Muscle Recruitment ...........................................................................................................23
Page 4 of 106

Statistics ................................................................................................................................ 23
References ............................................................................................................................ 25
: Angular Impulse and Balance Regulation during the Golf Swing ............................27
Introduction ........................................................................................................................... 27
Methods ................................................................................................................................ 29
Results .................................................................................................................................. 30
Discussion ............................................................................................................................. 37
References ............................................................................................................................ 40
: Angular Impulse and Balance Control while Increasing Golf Shot Distance
within a Club .............................................................................................................................43
Introduction ........................................................................................................................... 43
Materials and Methods .......................................................................................................... 43
Results .................................................................................................................................. 45
Discussion ............................................................................................................................. 47
References ............................................................................................................................ 49
: Angular Impulse and Balance Regulation during the Golf Swing with the
9-iron and 6-iron ........................................................................................................................51
Introduction ........................................................................................................................... 51
Methods ................................................................................................................................ 51
Results .................................................................................................................................. 53
Discussion ............................................................................................................................. 57
References ............................................................................................................................ 58
: Regulation of Linear and Angular Impulse with Modified Address Positions
during the Golf Swing ................................................................................................................60
Introduction ........................................................................................................................... 60
Methods ................................................................................................................................ 61
Results .................................................................................................................................. 62
Discussion ............................................................................................................................. 71
References ............................................................................................................................ 74
: Coordination of lower extremity multi-joint control strategies during the golf swing 76
Introduction ........................................................................................................................... 76
Methods ................................................................................................................................ 77
Results .................................................................................................................................. 79
Discussion ............................................................................................................................. 84
References ............................................................................................................................ 86
Page 5 of 106

: Between-Club Differences in Lower Extremity Muscle Activation while Regulating
Reaction Forces during the Golf Swing .....................................................................................89
Introduction ........................................................................................................................... 89
Materials and Methods .......................................................................................................... 90
Results .................................................................................................................................. 91
Discussion ............................................................................................................................. 96
References ............................................................................................................................ 99
Appendix A - Force Plate Documentation ................................................................................ 102
References .......................................................................................................................... 105

Page 6 of 106

: Introduction

The game of golf is enjoyed by 24.1 million players in the United States alone [1]. Players of all
ages and abilities enjoy the game. As a lifetime sport, the objectives and outcomes of the game
may depend on the population [2]. Highly skilled players are often seeking maximum performance,
where players with balance impairments (aging populations, persons with lower limb amputations,
etc.) may be playing for the health and rehabilitation benefits that the game provides. The golf
swing itself is an athletic movement that must satisfy multiple mechanical objectives in order to
be completed successfully. This lends itself to the study of how to produce maximum performance
of the golf swing as well as understanding how the task could be used as an alternative
intervention to overcome balance-related impairments in a fun and social environment.
Creating rotation during the golf swing is influenced by the nervous system, musculoskeletal
system and the interaction of reaction forces with the foot-surface interface. Rotation with limited
translation while both feet are in contact with the ground involves satisfying multiple, and
sometimes, competing mechanical objectives at both the whole body and segment level. From a
whole body perspective, the ground reaction forces at the foot-surface interface interact with the
center of mass trajectory to satisfy the net linear and angular impulse requirements specific to the
task [3–5]. At the segment level, the forces must interact with the segment orientation and motion
to allow for multi-joint control. Failure to meet any of these mechanical objectives could result in
failure to perform the golf swing effectively as players hit golf shots of variable distances and
address positions.
During the course of play, players have their choice of clubs with different length and club face
characteristics (e.g. loft, lie angle) to regulate the trajectory and distance of the golf shot. Players
will also encounter different terrain they have to hit from. For example, players typically hit the
driver for maximum distance off of a smooth, even tee box. Iron shots may be hit from varying
distances away from the green, where the terrain may be uneven. Often, the distance required
for each shot may not be the player’s normal shot distance with a certain club, but also not so
different from their preferred distance to necessitate a change in club. These conditions will
ultimately have an effect on how players coordinate action of the rear and target legs to create
reaction forces at the foot-surface interface to generate rotation for a successful golf shot.
Understanding how individuals satisfy the mechanical objectives during the golf swing with
varying conditions could enlighten the ability to develop interventions and aid skill acquisition to
increase player performance [6].
The proposed series of research studies aims to determine how golf players regulate rotation on
a whole body and segment level with a variety of distance and address conditions. Linear and
angular impulse generation, as well as muscular activation and joint kinetics will be measured to
further our understanding of the control and dynamics of the golf swing. These measures will be
tested across the group and within an individual by using a within-subject design to identify
individual player strategies across the conditions.

Page 7 of 106

Mechanics of the Golf Swing

The golf swing can be broken down into five phases: the backswing, transition, early downswing,
late downswing, and the follow-through. During the backswing, the player-club system rotates
away from the target to setup the initial conditions of the downswing. Throughout transition, the
player initiates rotation of the system toward the target while limiting translation to maintain
balance. In the downswing, players continue the rotation initialized in transition. During late
downswing the mechanical objective is to begin stopping rotation and translation. The mechanical
objective during the follow-through is to halt the rotation and translation toward the target after
ball contact, and maintain balance over the base of support through the end of the swing.
Rotational effects created over the interval of interest covering the transition and early downswing
can be measured through the angular impulse. Angular impulse is the integral of the moment-
time curve, where the moment is created from the interaction of the reaction forces and the center
of mass trajectory from the rear and target legs. During the golf swing, the moment about a vertical
axis through the center of mass is calculated as the cross product between the position vector
from the CM in the transverse plane to the point of RFh application and the RFh (Figure 1.1).
Combining the position vector and the angle between the RFh and position vector, the effective
moment arm for the RFh can be calculated. During the interval of transition and early downswing,
the RFhs redirect to create angular impulse toward the target [7,8].

The reaction forces that create rotation also create the linear impulse generated during the golf
swing. Linear impulse is the translational effect of the RFh during the interval of interest. Previous
research in the golf swing has found that linear impulse in the transverse plane is directed toward
the target and effectively mitigated in the anterior-posterior direction [8]. Due to the rotational and
translational effects the RFh creates, the RFhs generated from both the rear and target leg need
to be coordinated for successful completion of the task.
RFh
R

Target
θ
R

θ
T

r
R

r
T

RFh
T

CM
Figure 1.1: Application of reaction forces of rear and target
legs during the downswing of a typical golf shot.
Page 8 of 106

RFh
R

Target
θ
R

θ
T

r
R

r
T

RFh
T

CM
RFh
R

Target
θ
R

θ
T

r
R

r
T

RFh
T

CM
RFh
R

Target
RFh
T

CM
RFh
R

Target
θ
R

θ
T

RFh
T

CM
Example solutions to increase
Angular Impulse:

𝐴𝑛𝑔𝑢𝑙𝑎𝑟 𝐼𝑚𝑝𝑢𝑙𝑠𝑒 = ( 𝑟 𝐶𝑀
sin 𝜃 𝑅𝐹 ℎ )𝑑𝑡
𝑡 2
𝑡 1

Increase 𝑹𝑭𝒉

• Need to control linear impulse in both
directions to maintain balance

Increase 𝒔𝒊𝒏 𝜽

• Need to control anterior-posterior
impulse to maintain balance
• May lose CM trajectory parallel toward
the target

Increase 𝒓 𝑪𝑴

• Often increasing for one leg, but
decreasing in the other.

*Note that changes in sin 𝜃 and 𝑟 𝐶𝑀

contribute to changes in effective moment
arm for each RFh. Increasing duration of time
the moment is applied is another method to
increase angular impulse.
Figure 1.2: Demonstration of multiple strategies to increase angular impulse and their implications.

Page 9 of 106

Regulation Within and Between Clubs

As shot distance or address position of the golf swing is altered, the mechanical demands of the
task must also be regulated. The golf player has multiple options to regulate angular impulse from
both the rear and target legs. These options include: magnitude of the RFh, orientation of the RFh
(RFh-angle), the position vector between the point of force application to the CM, and the duration
of time the moment is applied (Figure 1.2). In previous studies on elite golf players, players were
found to reduce the RFh with limited modification of the RFh-angle while reducing golf shot
distance within a 6-iron [8]. Coordination of rear and target leg RFh in the anterior and posterior
directions, respectively, achieved the rotational demands while allowing linear impulse to remain
similar across tasks and to maintain balance over the base of support.
An anteriorly directed RFh is expected to be derived from activation of muscles on the posterior
of the leg (hamstrings), while a posterior RFh is expected to be derived from activation of muscles
on the anterior of the leg (quadriceps). This is consistent to other research conducted during the
golf swing [9,10]. Our lab has also found the reaction force to be consistent with muscle activation
in a variety of other tasks performed by highly skilled athletes [3,5,11]. As players increased golf
shot distance between the pitching wedge and 4-iron, Marta and colleagues also found increases
in muscle activation of some lower extremity muscles in a mid-handicap population [10].
The kinematic context combined with forces at the joint level allows us to understand the
mechanical demands placed on each joint. Resultant net joint moments have been revealed to
be consistent with the orientation of the ground reaction force relative to the orientation of the
lower extremity [12–14]. Looking at the leg as a system, we can determine the multi-joint control
by accounting for the moments and forces of all the joints of the leg (ankle, knee and hip). The
multi-joint control of the leg has been discovered to be influenced by the regulation of whole body
dynamics to meet the mechanical demands of the task [3,5,11,15]. More generally, we can begin
to understand the how each leg contributes to the coordination of both legs in creating rotation
and limiting translation during the golf swing. This level of understanding will really allow for the
personalization of training and performance interventions by utilizing information about the
individual to guide the preparation plan [6].

Proposed Set of Experiments

This body of work aims at determining how individuals satisfy the whole-body and segment level
mechanical objectives during the golf swing while regulating golf shot distance within and between
clubs with varying initial address positions. A within-subject design will further elucidate how
individuals accomplish the tasks and will assist in the creation of interventions to prepare players
and increase performance [6]. Studying the golf swing with varying clubs, distances, and address
positions provides an opportunity to understand the regulation of impulse and balance. The
varying conditions are well practiced, goal-directed tasks that require regulation of whole-body
and segment level mechanical objectives to complete the task successfully.

Page 10 of 106

References

[1] National Golf Foundation, 2015, Summary of Golf Participation.
[2] Vandervoort, A. A., Lindsay, D. M., Lynn, S. K., and Noffal, G. J., 2012, “Golf is a Physical
Activity for a Lifetime,” Int. J. Golf Sci., 1, pp. 54–69.
[3] McNitt-Gray, J. L., Hester, D. M., Mathiyakom, W., and Munkasy, B. a, 2001, “Mechanical
demand and multijoint control during landing depend on orientation of the body segments
relative to the reaction force.,” J. Biomech., 34(11), pp. 1471–82.
[4] Peterson, T. J., Wilcox, R. R., and McNitt-Gray, J. L., 2016, “Angular impulse and balance
regulation during the golf swing,” J. Appl. Biomech., 32(4), pp. 342–349.
[5] Mathiyakom, W., McNitt-Gray, J. L., and Wilcox, R., 2006, “Lower extremity control and
dynamics during backward angular impulse generation in forward translating tasks.,” J.
Biomech., 39(6), pp. 990–1000.
[6] McNitt-Gray, J. L., Sand, K., Ramos, C., Peterson, T. J., Held, L., and Brown, K., 2015,
“Using technology and engineering to facilitate skill acquisition and improvements in
performance,” Proc. Inst. Mech. Eng. Part P J. Sport. Eng. Technol., 229(2), pp. 103–115.
[7] Barrentine, S. W., Fleisig, G. S., and Johnson, H., 1994, “Ground Reaction Forces and
Torques of Professional and Amateur Golfers,” Sci. Golf II Proc. 1994 World Sci. Congr.
Golf, 1(1), pp. 33–39.
[8] McNitt-Gray, J. L., Munaretto, J., Zaferiou, A., Requejo, P. S., and Flashner, H., 2013,
“Regulation of reaction forces during the golf swing,” Sport. Biomech., 12(2), pp. 121–131.
[9] Bechler, J. R., Jobe, F. W., Pink, M., Perry, J., and Ruwe, P. A., 1995, “Electromyographic
analysis of the hip and knee during the golf swing,” Clin J Sport Med, 5(3), pp. 162–166.
[10] Marta, S., Silva, L., Vaz, J. R., Castro, M. A., Reinaldo, G., and Pezarat-Correia, P., 2016,
“Electromyographic analysis of lower limb muscles during the golf swing performed with
three different clubs,” J. Sports Sci., 34(8), pp. 713–720.
[11] Mathiyakom, W., McNitt-Gray, J. L., and Wilcox, R., 2006, “Lower extremity control and
dynamics during backward angular impulse generation in backward translating tasks.,” Exp.
Brain Res., 169(3), pp. 377–88.
[12] Goh, H.-T., Sullivan, K. J., Gordon, J., Wulf, G., and Winstein, C. J., 2012, “Dual-task
practice enhances motor learning: a preliminary investigation,” Exp. Brain Res., 222(3), pp.
201–210.
[13] Macpherson, J. M., 1988, “Strategies that simplify the control of quadrupedal stance. I.
Forces at the ground.,” J. Neurophysiol., 60(1), pp. 204–17.
[14] Ting, L. H., and Macpherson, J. M., 2004, “A Limited Set of Muscle Synergies for Force
Control During a Postural Task,” J. Neurophysiol., 93(1), pp. 609–613.
[15] Mathiyakom, W., McNitt-Gray, J. L., Requejo, P. S., and Costa, K., 2005, “Modifying center
of mass trajectory during sit-to-stand tasks redistributes the mechanical demand across
the lower extremity joints.,” Clin. Biomech. (Bristol, Avon), 20(1), pp. 105–11.

Page 11 of 106

: Specific Aims and Hypotheses

Specific Aim 1: Angular Impulse and Balance Regulation during the Golf Swing

Goal: To determine how individual players coordinate the rear and target legs to regulate linear
and angular impulse while maintaining balance during golf swings with a 6-Iron and Driver.
 Hypothesis 1: Net angular impulse by an individual will increase when swinging with a
driver compared to a 6-iron.
 Hypothesis 2: Increases in angular impulse generation will involve contributions from both
the rear and target legs.
 Hypothesis 3: Increases in each leg’s angular impulse will arise from increases in RFh
magnitude with minimal modification of the RFh-angle relative to the target line
 Hypothesis 4: Linear impulse will be regulated when swinging with both clubs to maintain
balance over the base of support.

These hypotheses will be tested by comparing linear and angular impulse during the interval of
positive net moment and their contributing factors while golf players (n = 11) perform four to six
normal swings with a driver and 6-iron. Contributing factors to linear and angular impulse are:
reaction force, resultant horizontal reaction force (RFh), RFh-angle, moment arm, and time. The
Sign Test (α = .05) will be used to determine between club differences across the group. Within-
player differences will also be tested by calculating p-values using Cliff’s analog of the Wilcoxon-
Mann-Whitney test and applying a modified, step-down Fisher-type method to control the
familywise error rate over multiple comparisons [1–5].

Specific Aim 2: Angular Impulse and Balance Control while Increasing Golf Shot
Distance within a Club

Goal: To determine how individual players coordinate the rear and target legs to regulate angular
impulse while maintaining balance and increasing golf shot distance within the Driver.
 Hypothesis 1: Net angular impulse by an individual will increase with increasing golf shot
distance.
 Hypothesis 2: Increases in angular impulse generation will involve contributions from both
rear and target legs.
 Hypothesis 3: Increases in each leg’s angular impulse will arise from increases in RFh
magnitude with minimal modification of the RFh-angle relative to the target line.
 Hypothesis 4: Linear impulse will be regulated when swinging with both clubs to maintain
balance over the base of support.

These hypotheses will be tested by comparing linear and angular impulse during the interval of
positive net moment and their contributing factors while golf players (n = 10) perform four to six
normal and reduced distance swings with a 6-iron. Contributing factors to linear and angular
Page 12 of 106

impulse are: reaction force, resultant horizontal reaction force (RFh), RFh-angle, moment arm,
and time. The Sign Test (α = .05) will be used to determine between club differences across the
group. Within-player differences will also be tested by calculating p-values using Cliff’s analog of
the Wilcoxon-Mann-Whitney test and applying a modified, step-down Fisher-type method to
control the familywise error rate over multiple comparisons [1–5].

Specific Aim 3: Angular Impulse and Balance Control during the Golf Swing with
the 9-iron and 6-iron

Goal: To determine how individual players coordinate the rear and target legs to regulate angular
impulse while maintaining balance and increasing shot distance between the 9-iron and 6-iron
 Hypothesis 1: Net angular impulse by an individual will increase with the 6-iron compared
to the 9-iron.
 Hypothesis 2: Increases in angular impulse generation will involve contributions from both
the rear and target legs.
 Hypothesis 3: Increases in each leg’s angular impulse will arise from increases in RFh
with minimal modification of the RFh-angle relative to the target line.
 Hypothesis 4: Linear impulse will be regulated when swinging with both clubs to maintain
balance over the base of support.

These hypotheses will be tested by comparing linear and angular impulse during the interval of
positive net moment and their contributing factors while golf players (n = 10) perform four to six
normal swings with a 9-iron and 6-iron. Contributing factors to linear and angular impulse are:
reaction force, resultant horizontal reaction force (RFh), RFh-angle, moment arm, and time. The
Sign Test (α = .05) will be used to determine between club differences across the group. Within-
player differences will also be tested by calculating p-values using Cliff’s analog of the Wilcoxon-
Mann-Whitney test and applying a modified, step-down Fisher-type method to control the
familywise error rate over multiple comparisons [1–5].

Specific Aim 4: Angular Impulse and Balance Control with Modified Address
Positions within a Club

Goal: To determine how individual players coordinate the rear and target legs to regulate angular
impulse while maintaining balance as the address position modifies the configuration of the lower
extremities.
 Hypothesis 1: Net angular impulse will decrease when swinging with a modified address
position compared to the normal address position.
 Hypothesis 2: Regulation of angular impulse will arise from regulation of the RFh in the
rear and/or target leg.
 Hypothesis 3: Net linear impulse perpendicular and parallel to the target will be
comparable across conditions.

Page 13 of 106

These hypotheses will be tested by comparing linear and angular impulse during the interval of
positive net moment and their contributing factors while golf players (n = 9) perform ten swings
with a 6-iron as the address position is modified to raise the rear or target leg by 0.175 m.
Contributing factors to linear and angular impulse are: reaction force, resultant horizontal reaction
force (RFh), RFh-angle, moment arm, and time. The Sign Test (α = .05) will be used to determine
between club differences across the group. Within player differences will also be tested by
calculating p-values using Cliff’s analog of the Wilcoxon-Mann-Whitney test and applying a
modified, step-down Fisher-type method to control the familywise error rate over multiple
comparisons [1–5].

Specific Aim 5: Lower Extremity Joint Control during the Golf Swing

Goal: To determine the mechanical demand and multi-joint control of the rear and target legs
during the golf swing.
 Hypothesis 1: Rear and target leg 3D support moment will increase during swings with the
driver compared to the 6-iron.
 Hypothesis 2: Increases in 3D support moment will be due to increases in ankle, knee,
and hip net joint moments while maintaining relative contributions to 3D support moment.
 Hypothesis 3: Increases in 3D support moment would occur through increases in net joint
moment about an axis perpendicular to the respective leg plane to simplify control
strategies.

These hypotheses will be tested by comparing the net joint moments during the interval of interest
while golf players (n = 11) perform ten golf swings with the 6-iron and driver. Reaction forces will
be measured and quantified for each leg. Net joint moments will be calculated using inverse
dynamics. Functional joint centers will be used to define a leg plane and the axis perpendicular
to the leg plane [7]. The Sign Test (α = .05) will be used to determine between club differences
across the group. Within-player differences will also be tested by calculating p-values using Cliff’s
analog of the Wilcoxon-Mann-Whitney test and applying a modified, step-down Fisher-type
method to control the familywise error rate over multiple comparisons [1–5].

Specific Aim 6: Regulation of Muscular Control while Regulating Golf Shot
Distance between Clubs

Goal: To determine how individual players coordinate muscle activation in the rear and target legs
to regulate reaction force while regulating golf shot distance between the 6-iron and driver.
 Hypothesis 1: Increases in reaction force generation between clubs would correspond with
increases in activation of the involved muscles.
We expected control preferences of the individual player would be maintained within leg across
clubs, where reaction force generation by each leg would involve the same set of muscles,
Page 14 of 106

respectively. We also expected differences in muscle activation patterns would correspond to
observed differences in reaction force generation during the interval of interest.
This hypothesiswill be tested by comparing the electromyogram (EMG) of the lower extremity by
golf players (n = 12) as they performed 3-4 swings with the 6-iron and driver. Muscle activity will
be compared within player and will be normalized by the muscle activation achieved during
manual muscle tests [6]. Reaction force magnitude and direction will be measured and quantified
to compare with muscle activity to determine consistency between measures. The Sign Test (α
= .05) will be used to determine between club differences across the group.

Page 15 of 106

References

[1] Cliff, N., 1996, Ordinal Methods for Behavioral Data Analysis, Lawrence Erlbaum
Associates, Inc., Publishers, Mahwah, NJ.
[2] Neuhäuser, M., Lösch, C., and Jöckel, K.-H., 2007, “The Chen–Luo test in case of
heteroscedasticity,” Comput. Stat. Data Anal., 51(10), pp. 5055–5060.
[3] Hochberg, Y., 1988, “A sharper Bonferroni test for multiple tests of significance,”
Biometrika, 75, pp. 800–802.
[4] Hochberg, Y., and Tamhane, A. C., 1987, Multiple Comparison Procedures, John Wiley
& Sons, Inc., Hoboken, NJ, USA.
[5] Wilcox, R., and Clark, F., 2015, “Robust Multiple Comparisons Based on Combined
Probabilities From Independent Tests,” J. Data Sci., 13(1), pp. 1–11.
[6] Kendall, F., McCreary, E., Provance, P., Rodgers, M., and Romani, W., 2005, Muscles:
Testing and Function, with Posture and Pain, Lippincott Williams & Wilkins, Baltimore.
[7] Russell, I. M., Raina, S., Requejo, P. S., Wilcox, R. R., Mulroy, S., and McNitt-Gray, J. L.,
2015, “Modifications in Wheelchair Propulsion Technique with Speed,” Front. Bioeng. Biotechnol.,
3(October), pp. 1–11.

Page 16 of 106

: Experimental Design

Experimental Procedures

Highly-skilled amateur golf players were recruited (n ≥ 9) from the local university varsity teams
and surrounding southern California area for these studies. Each player gave their informed
consent in accordance with the University of Southern California Institutional Review Board for
Human Subjects. Players were given the opportunity to warm up and practice conditions in the
laboratory prior to completing data collection.
Inclusion Criteria

Players volunteering to participate in the study must have a golf handicap of +5 or lower [1].
Players must be free of injury at the time of data collection and be medically cleared to play.
Players must be able to complete up to 150 golf swings during the session (consistent with normal
practice sessions).
Exclusion Criteria

Players in this study must not have a history of neurologic deficits or other musculoskeletal
disorder. Players must not be currently using a prosthetic device on the lower extremity or require
the use of a balancing aid.
Overview of Participant Tasks

Calibration Tasks
Calibration tasks were necessary to establish a reference for subsequent trials during the data
collection. A static trial was taken with the player standing quietly on the force plates to calculate
the player’s body weight and to establish a mathematical relationship between the anatomical
and tracking markers used by the motion capture system. Furthermore, functional movement trials
were used to calculate the functional joint centers of the ankles, knees, hips and shoulders [2,3].
These tasks were performed using unloaded movements in multiple axes of joint motion to
capture the full movement of the each joint. If muscle recruitment measurements were captured
with electromyography, appropriate isometric manual muscle tests were performed to be used as
a muscle recruitment normalization for each individual during data processing [4].
Experimental Tasks
Players were asked to complete golf swings as they normally would on the golf course or during
practice. The natural shape of the individual players’ golf shot (i.e. fade, draw or straight) was not
controlled for. A target was placed downrange for players to aim at. Swings would be completed
using multiple clubs (9-iron, 6-iron, driver, TaylorMade-adidas). A distance goal was assigned for
each swing where players were asked to hit golf shots their normal, preferred distance for each
club. Players were also asked to perform swings while reducing shot distance ten yards (irons
and driver), or increasing shot distance by ten yards for each club (driver). A subset of players
were also asked to perform normal distance golf shots with a 6-iron where their lower extremities
were configured with either the target leg or rear leg raised to mimic realistic leg configurations
Page 17 of 106

during the course of play. A wooden box (0.4 x 0.6 x 0.175 m) will be affixed with bolts directly to
either the target or rear leg force plate.

Data Collection

Kinematics
Video
Two-dimensional video will be captured simultaneously in the frontal, sagittal and transverse
planes (60 Hz Panasonic, Secaucus, NY; 300 Hz, Casio, Dover, NJ, 300 Hz, JVC, Yokohama,
Japan) to determine body segment kinematics and confirm successful completion of each trial.
Motion Capture
Three dimensional kinematics of the players’ motion were captured using retroreflective markers
and a 16 camera (Natural Point, Optitrack, Corvallis, Oregon) system accompanied by Acquire3D
software (C-Motion, Germantown, Maryland). The camera system was calibrated prior to data
collection with point residuals < 0.7mm accuracy. To synchronize our kinematic and kinetic data,
a trigger signal was sent from the motion capture software via BNC cable to the data acquisition
card upon collection of three dimensional kinematics.
A custom, 97 marker set of anatomic and tracking retroreflective markers (1.2 cm diameter, B&L
Engineering, Santa Ana, California) were applied to the participant using adhesive hook and loops,
skin adhesive spray and self-adhering compressive tape. The data collection team applied the
markers and the experimenter confirmed proper marker placement (Figure 3.1).
Anatomic
At least three anatomic markers are required to anatomically define a reference system for each
segment. These markers will be applied directly to the skin over palpated bony landmarks. After
a static calibration is collected, the markers on the segment then may be removed and tracked
via the tracking markers. Anatomic markers can be removed due to interference with the
movement required. The anatomic markers that do not interfere may remain and may also be
used as tracking markers.
Page 18 of 106

Tracking
During motion trials, at least three, non-collinear markers per segment must be known to define
the reference system and orientation of the segment in space throughout the movement. These
tracking markers refer to the anatomic markers during the static calibration trial. Tracking markers
are used to create the reference system of a segment in a more reliable manner, often placing
markers in non-anatomic places with a lesser chance for marker movement. On the shanks and
thighs, we attached these markers to the segments via rigid cuffs and self-adhering compressive
tape. This methodology minimizes marker motion, vibration, and marker slip caused by relative
motion between the soft tissue and the underlying bony landmark if the marker was placed directly
on the skin. Each cuff was equipped with four retroreflective markers and placed on the segment
in agreement with suggestions from previous research [5]. Foot markers were applied directly to
the shoe or skin, pelvis markers were applied on top of the waistband and upper extremity markers
were placed either directly on the skin or on top of a tight-fitting shirt to reduce marker motion.
Figure 3.1: Anterior, posterior, lateral (left) and lateral (right) views of marker placement on players prior to data
collection.
Page 19 of 106

Kinetics
Ground Reaction Forces
Ground reaction forces were measured using two force plates (0.4 x 0.6 m, 1200 Hz, Kistler,
Amherst, Massachusetts), one under each foot. These data were amplified and converted from
analog to digital signals using custom data collection software in LabView (National Instruments,
Austin, Texas). The force plates were covered with a thin layer of flooring (Mondo, USA,
Conshohocken, Pennsylvania) to aid in slippage and mimic normal friction characteristics [6]. The
covers on the force plates were isolated from the surrounding surface so force was not transferred.
The properties of the flooring were chosen to maintain fidelity of the force signals while providing
the proper amount of friction to safely complete the task (Figure 3.2).

For trials requiring one of the legs to be raised, a custom wooden box (0.4 x 0.6 x 0.175 m) was
affixed with bolts directly to either the target or rear leg force plate. This box was tested prior to
use and was found to provide proper force transmission without attenuating the force signals
(Figure 3.3).

Figure 3.3: Custom box setup affixed directly to the force plate with bolts through pre-tapped holes on the force plate.

Figure 3.2: Force plates with their isolated coverings to provide adequate friction characteristics to safely complete the
task while maintaining fidelity of force transmission.
Page 20 of 106

Lab Setup and Coordinate System

The lab coordinate reference system was based on the force plate coordinate systems, but was
translated to its current origin location to match the kinematic and kinetic global coordinate
systems (Figure 3.4).
Muscle Recruitment and EMG

Activation of the lower extremity muscles were monitored using telemetered surface
electromyography (1200Hz, Konigsburg, Pasadena, CA). These data were amplified and
converted from analog to digital signals using custom data collection software in LabView
(National Instruments, Austin, Texas) to synchronize with the incoming force data. Skin over the
muscle belly was shaved and cleaned prior to placement of the surface electrodes. Dual
electrodes with a 1 cm inter-electrode distance (Noraxon, Scottsdale, Arizona) were placed over
the muscle belly, parallel to the muscle action.

Data Processing and Analysis

Kinematics
The processing of three dimensional kinematics involved multiple steps. After the data were
collected through Acquire3D (C-Motion, Germantown, MD, USA), the individual markers were
identified and output as type “.c3d” in AMASS (C-Motion). The “.c3d” files were imported into
Visual3D (C-Motion) to estimate functional joint centers. Marker data were then exported to “.txt”
file type for processing in Matlab (The Mathworks, Natick, Massachusetts).
After importing to Matlab, kinematic data were filtered and any gaps in the data were filled using
a cubic spline smoothing function. The smoothing factor “p” was chosen based on the work of
Jackson [7]. Because the function smooths by creating a cubic function of the data, subsequent
derivatives of all marker data were calculated to get marker velocity and acceleration data.
Figure 3.4: Motion capture camera positions relative to force plates (left) and lab based reference system relative to the
target line (right).
Page 21 of 106

Functional joint centers during the functional joint calibration movements were calculated for the
ankles, knees, hips and shoulders using custom Matlab code [8]. The initial guess of joint center
location was given from the output from Visual3D. This initial guess is necessary for the post-hoc
mathematical selection of an average/mode intersection of the instantaneous axes of rotation.
During the static calibration trial, the functional joint centers defined the endpoints of the
respective segments. At this time, a mapping from anatomic reference system to tracking
reference system was calculated (rotation matrix). By understanding this relationship, it is possible
to represent the anatomical landmarks and joint kinetics about anatomic reference segments even
if the anatomic landmarks are removed or lost during collection.
Segment inertial parameter estimates were used to calculate segment and total body center of
mass for use in whole-body and joint kinetics analysis [9]. This is necessary when the total body
center of mass (TBCM) is used to calculate whole-body linear and angular impulse. Additionally,
the segment center of mass will be utilized to calculate joint level moments.
The previously discussed cubic-spline method was utilized again to interpolate the kinematic data
(100 Hz) to match the sampling frequency of the kinetic data (1200 Hz). At this time no smoothing
occurs, and it is only interpolation of the data. Again, the cubic-spline method allows for derivatives
to be taken on polynomials to calculate velocity and acceleration. Angular velocity and
acceleration of each segment is then calculated using quaternion parameterization to reduce the
effects of rotation matrix based singularity points. With this information, joint kinetics were
calculated using inverse dynamics for the lower extremities to determine the forces and moments
experienced at each joint.
Kinetics
The magnitude and direction of the reaction forces for each leg were calculated from the force
plate manufacturer documentation (Appendix A). Resultant horizontal reaction force was
calculated for each leg as follows (Eq. 3.1):

𝑅𝐹 ℎ = √𝑅𝐹𝑥 2
+ 𝑅𝐹𝑦 2
2

Eq. 3.1

Subsequent linear impulses over the interval of interest were also calculated for each horizontal
direction (Eq. 3.2 and Eq. 3.3). Where this measure is the total effect of the force in that direction
over a period of time.

𝐿𝑖𝑛𝑒𝑎𝑟 𝐼𝑚𝑝𝑢𝑙𝑠𝑒 𝑋 − 𝑑𝑖𝑟𝑒𝑐𝑡𝑖𝑜𝑛 = 𝑅𝐹𝑥 𝑑𝑡 𝑡 2
𝑡 1

Eq. 3.2
𝐿𝑖𝑛𝑒𝑎𝑟 𝐼𝑚𝑝𝑢𝑙𝑠𝑒 𝑌 − 𝑑𝑖𝑟𝑒𝑐𝑡𝑖𝑜𝑛 = 𝑅𝐹𝑦 𝑑𝑡 𝑡 2
𝑡 1

Eq. 3.3

Page 22 of 106

The point of wrench application (PWA) was computed for each leg [10,11]. To account for the
horizontal forces present in the task, the PWA was used as an alternative to the center of pressure
(CP) [10–12]. In the absence of horizontal forces, the PWA becomes the CP. The following
equations (Eq. 3.4 and Eq. 3.5) outline the detail of those calculations [10,11]:

𝑋 𝑃𝑊𝐴 =
𝐹 𝑥 𝑀 𝑧 − 𝐹 𝑧 𝑀 𝑦 𝐹 2
−
𝐹 𝑥 2
𝑀 𝑌 − 𝐹 𝑥 𝐹 𝑦 𝑀 𝑥 𝐹 2
𝐹 𝑧
Eq. 3.4

𝑌 𝑃𝑊𝐴 =
𝐹 𝑧 𝑀 𝑥 − 𝐹 𝑥 𝑀 𝑧 𝐹 2
−
𝐹 𝑥 𝐹 𝑦 𝑀 𝑦 − 𝐹 𝑦 2
𝑀 𝑥 𝐹 2
𝐹 𝑧
Eq. 3.5

Where Mx, My, and Mz are the moments about each axis of the force plate and F represents the
vector magnitude of the force vector.
With knowledge of the initial TBCM position, the net linear impulse can be used to determine the
TBCM position through time [13]. Total body center of mass trajectories determined with this
method were found to be well aligned with the TBCM calculated using markers and kinematic
data accompanied with body segment parameters.
Combining the horizontal TBCM position and PWA from each foot, the moment from each leg
about a vertical axis passing from the TBCM were calculated (Eq. 3.6).
𝑀𝑜𝑚𝑒𝑛𝑡 𝑎𝑝𝑝𝑙𝑖𝑒𝑑 𝑏𝑦 𝑒𝑎𝑐 ℎ 𝑙𝑒𝑔 = 𝑟 𝐶𝑀
𝑥 𝑅𝐹 ℎ
Eq. 3.6

The free moment on each leg was taken into account to capture the total moment from each leg
(Eq. 3.7) [14].
𝑀𝑜𝑚𝑒𝑛𝑡 𝑜𝑓 𝑒𝑎𝑐 ℎ 𝑙𝑒𝑔 = 𝑟 𝐶𝑀
𝑥 𝑅𝐹 ℎ + 𝑀 𝑧
Eq. 3.7

To determine the contributing factors to the moment from each leg, the mathematically equivalent
form of the equation was used (Eq. 3.8).
𝑀𝑜𝑚𝑒𝑛𝑡 𝑜𝑓 𝑒𝑎𝑐 ℎ 𝑙𝑒𝑔 = 𝑟 𝐶𝑀
sin 𝜃 𝑅𝐹 ℎ + 𝑀 𝑧
Eq. 3.8

To determine the net rotational effects of the task, the net angular impulse during the interval of
interest was calculated as follows (Eq. 3.9):
𝑁𝑒𝑡 𝐴𝑛𝑔𝑢𝑙𝑎𝑟 𝐼𝑚𝑝𝑢𝑙𝑠𝑒 = 𝑀 𝑟𝑒𝑎𝑟 + 𝑀 𝑡𝑎𝑟𝑔𝑒𝑡 𝑑𝑡 𝑡 2
𝑡 1

Eq. 3.9

Page 23 of 106

Net joint moments can then be parsed based on anatomical reference systems (Figure 3.5A) [9].
Functional joint centers were also be used to define a leg plane and an axis perpendicular to the
leg plane similar to the arm plane defined by Russell and colleagues (Figure 3.5B) [15]. The axis
was used to determine how well aligned the reaction forces and moments were to the leg plane.
Muscle Recruitment
Electromyography data were filtered using a 4
th
order zero-phase, Butterworth band-pass filter
(10-400 Hz) [16]. These were quantified using root mean squared values in 20 ms average bins.
Muscle activation data were normalized to maximum binned values during isometric manual
muscle tests [4,17].
Statistics

Between condition differences across the group were determined using the Sign Test (α = 0.05).
The Sign Test was chosen due to the Student’s T-test inability to handle issues due to non-
normality, incorrect assumptions of variance, and unequal sample sizes between conditions [18].
Using the Sign test, with N = 10, α = 0.05, and a result of p = 0.05, a power between 1-β = 0.92386
and 1-β = 0.9885 is achieved [19]. A drastic drop in power to 1-β = 0.66342 occurs if the sample
size decreases to N = 8. Within-player differences between conditions were also tested to reveal
player specific strategies. The statistical methods used for within-player differences are a subset
of methods known as a two-state linear model. The probability for each variable of any condition
1 trial being less than any condition 2 trial was calculated within a player for each variable, where
each player served as their own control (R, open-source). Assuming local independence (i.e. no
order effect for trials within a condition), and that club conditions were independent (i.e. not directly
A) B)
Figure 3.5: Segment based anatomic reference systems of the lower extremity (A). Sagittal (left) and frontal (right)
plane views of the leg plane and its perpendicular axis.
Page 24 of 106

tied to each other) for each subject, p-values were calculated for each player using Cliff’s analog
of the Wilcoxon-Mann-Whitney test [20,21]. This method was chosen because it deals well with
small numbers of trials per condition [21]. A modified, step-down Fisher-type method was then
applied to control the familywise error rate (α = .05) over multiple comparisons where the level of
significance becomes α/k at each k
th
iteration [22–24].

The current statistics provide more flexibility by allowing heteroscedasticity across players. As the
number of trials increase per condition, Cliff’s method can achieve lower p-values if the null
hypothesis of is false. If the null hypothesis is true, the p-value should converge to 0.5 as the
sample size increases, assuming an asymptotically correct method is used. The modified, step-
down Fisher-type technique is dependent upon the distribution of p-values for each variable
measured because the significance level is adjusted at each step to compensate for multiple
comparisons [22–24]. Therefore, the presentation of within-player results provides a conservative
estimate of significant differences between club conditions.

Page 25 of 106

References

[1] Worsfold, P., Smith, N., and Dyson, R., 2008, “Low handicap golfers generate more torque
at the shoe-natural grass interface when using a driver,” J. Sport. Sci. Med., (7), pp. 408–
414.
[2] Leardini, A., Cappozzo, A., Catani, F., Toksvig-Larsen, S., Petitto, A., Sforza, V.,
Cassanelli, G., and Giannini, S., 1999, “Validation of a functional method for the estimation
of hip joint centre location,” J. Biomech., 32(1), pp. 99–103.
[3] Piazza, S. J., Okita, N., and Cavanagh, P. R., 2001, “Accuracy of the functional method of
hip joint center location: effects of limited motion and varied implementation.,” J. Biomech.,
34(7), pp. 967–73.
[4] Kendall, F., McCreary, E., Provance, P., Rodgers, M., and Romani, W., 2005, Muscles:
Testing and Function, with Posture and Pain, Lippincott Williams & Wilkins, Baltimore.
[5] Cappozzo, A., Cappello, A., Della Croce, U., and Pensalfini, F., 1997, “Surface-marker
cluster design criteria for 3-D bone movement reconstruction.,” IEEE Trans. Biomed. Eng.,
44(12), pp. 1165–74.
[6] Williams, K., and Cavanagh, P. R., 1983, “The mechanics of foot action during the golf
swing and implications for shoe design.,” Med. Sci. Sports Exerc., 15(3), pp. 247–255.
[7] Jackson, K. M., 1979, “Fitting of Mathematical Functions to Biomechanical Data,” IEEE
Trans. Biomed. Eng., BME-26(2), pp. 122–124.
[8] Schwartz, M. H., and Rozumalski, A., 2005, “A new method for estimating joint parameters
from motion data.,” J. Biomech., 38(1), pp. 107–16.
[9] de Leva, P., 1996, “Adjustments to Zatsiorsky-Seluyanov’s segment inertia parameters,”
J. Biomech., 29(9), pp. 1223–30.
[10] Shimba, T., 1996, “Consequences of Force Platform Studies,” Natl. Rehabiliation Res. Bull.
Japan, (17), pp. 17–23.
[11] Zatsiorsky, V. M., 2002, “Wrench Representation of the Ground Reaction Force,” Kinetics
of Human Motion, Human Kinetics Publishers, Champaign, IL, pp. 43–48.
[12] Cavanagh, P. R., Rodgers, M. M., and Iiboshi, A., 1987, “Pressure distribution under
symptom-free feet during barefoot standing.,” Foot Ankle, 7(5), pp. 262–276.
[13] King, D. L., and Zatsiorsky, V. M., 1997, “Extracting gravity line displacement from
stabilographic recordings,” Gait Posture, 6(1), pp. 27–38.
[14] Holden, J. P., and Cavanagh, P. R., 1991, “The free moment of ground reaction in distance
running and its changes with pronation.,” J. Biomech., 24(10), pp. 887–97.
[15] Russell, I. M., Raina, S., Requejo, P. S., Wilcox, R. R., Mulroy, S., and McNitt-Gray, J. L.,
2015, “Modifications in Wheelchair Propulsion Technique with Speed,” Front. Bioeng.
Biotechnol., 3(October), pp. 1–11.
[16] De Luca, C. J., Donald Gilmore, L., Kuznetsov, M., and Roy, S. H., 2010, “Filtering the
surface EMG signal: Movement artifact and baseline noise contamination,” J. Biomech.,
43(8), pp. 1573–1579.
Page 26 of 106

[17] Oddsson, L. I., Giphart, J. E., Buijs, R. J., Roy, S. H., Taylor, H. P., and De Luca, C. J.,
1997, “Development of new protocols and analysis procedures for the assessment of LBP
by surface EMG techniques.,” J. Rehabil. Res. Dev., 34(4), pp. 415–426.
[18] Wilcox, R., 2012, Modern Statistics for the Social and Behavioral Sciences: A Practical
Introduction, CRC Press Taylor & Francis Group, Boca Raton, FL.
[19] Dixon, W. J., 1953, “Power Functions of the Sign Test and Power Efficiency for Normal
Alternatives,” Ann. Math. Stat., 24(3), pp. 467–473.
[20] Cliff, N., 1996, Ordinal Methods for Behavioral Data Analysis, Lawrence Erlbaum
Associates, Inc., Publishers, Mahwah, NJ.
[21] Neuhäuser, M., Lösch, C., and Jöckel, K.-H., 2007, “The Chen–Luo test in case of
heteroscedasticity,” Comput. Stat. Data Anal., 51(10), pp. 5055–5060.
[22] Hochberg, Y., 1988, “A sharper Bonferroni test for multiple tests of significance,”
Biometrika, 75, pp. 800–802.
[23] Hochberg, Y., and Tamhane, A. C., 1987, Multiple Comparison Procedures, John Wiley &
Sons, Inc., Hoboken, NJ, USA.
[24] Wilcox, R., and Clark, F., 2015, “Robust Multiple Comparisons Based on Combined
Probabilities From Independent Tests,” J. Data Sci., 13(1), pp. 1–11.

Page 27 of 106

: Angular Impulse and Balance Regulation during
the Golf Swing

Introduction

Successful performance of the golf swing requires players to regulate the generation of linear and
angular impulse while maintaining balance. Players may choose golf clubs of different length and
clubface angles to control the trajectory of the ball during flight. As the required shot distance
increases (driver greater than 6-iron), linear and angular impulse generated at the foot-surface
interface is expected to be regulated to achieve the required clubhead speed [1–3] at ball contact.
Regulation of linear and angular impulse during well-practiced goal-directed movements involves
control of the reaction forces (RFs) generated during contact with the environment in relation to
the total body center of mass (CM) trajectory during the task [4]. Determining how an individual
player controls shot distance by regulating target and rear leg RFs in relation to the CM trajectory
during the swing can provide a more mechanistic basis for the design of interventions that aim to
improve performance, swing mechanics, and strategic decisions during play [1,5,6].

During the golf swing, angular impulse generation by the target and rear leg RFs (Figure 4.1) is
dependent on multiple factors. These factors include: the RFs applied to the body at the
foot/surface interface (center of pressure (CP)), the component of the position vector (from the
CM to the foot CP) perpendicular to the RF (moment arm), and the duration of time spent creating
a moment about the CM of the body-club system [5,7]. A posterior directed RF on the target leg
and an anterior directed RF applied on the rear leg have been associated with the moments
applied by the RFs about the vertical axis of the body-club system toward the target [1,5,7].
Vertical RFs [8,9] and CP patterns [10,11] have also been used to describe translation of the CM
in relation to the base of support where these measures affect the moment arm magnitudes of
the rear and target leg RFs. Previous research indicated skilled players consistently modified shot
distance with the same club by regulating the magnitude of the target and/or rear leg resultant
horizontal reaction forces (RFh), with minimal changes in RFh-angle relative to the target [1].
Specific strategies for controlling the RF were found to be unique for individual players [1,5,10–
13]. For example, lower handicap players have demonstrated shorter overall swing duration and
earlier timing of peak RFs when swinging with a 5-iron as compared to higher handicap golfers
[14]. Within-player increases in RFh have also been observed without significant modification in
swing duration when studying swings with longer golf clubs [5,15] and baseball bats [16,17].

Understanding how players coordinate their legs to produce the linear and angular impulse
required during the golf swings with different clubs is expected to improve training and facilitate
technique modification to enhance player preparation [6]. The identification of mechanisms that
individuals use to regulate linear and angular impulse at the whole body level has advanced our
understanding of control in a variety of tasks outside of golf [4,8,9,18–22]. For example,
populations with balance-related impairments often have limitations in their ability to generate
linear and angular impulse required in activities of daily living such as fall recovery [4,9,18], stair
descent [19], and gait termination [22]. In these populations, the inability to reliably regulate RFs
Page 28 of 106

generated with the lower limbs in relation to the CM trajectory contributes to the challenge of
regulating balance under different conditions [4,19,22]. Identifying strategies used by skilled golf
players to regulate impulse generation while maintaining balance during the swing is expected to
provide insights into potential solutions individuals can consider when preserving balance in other
weight bearing activities or overcoming balance-related impairments [23].

Figure 4.1: Top-down view of force plate setup and reference system for reaction force and angle
orientation (above). Resultant horizontal reaction force of four trials for rear and target legs of an exemplar
subject (mean ± SD, below) to display consistency of performance across trials (below). Ball contact occurs
at time = 0s, time of transition at t ≈ -0.3s.
Page 29 of 106

The aim of this study was to determine how skilled golf players regulate linear and angular impulse
they generate during golf swings with two different clubs. We hypothesized that net angular
impulse generated by an individual would increase when swinging with a driver compared to a 6-
iron. It was expected that observed increases in angular impulse generation between clubs would
involve contributions from both the rear and target legs. We also hypothesized the increase in
angular impulse generated by each leg would involve increases in RFh magnitude with minimal
modification of the RFh-angle relative to the target line thereby preserving the kinematic context
of RF generation between clubs. Because golfers remain upright through the follow through, we
anticipated that linear impulse would be regulated between clubs to maintain balance over the
base of support. We tested these hypotheses by measuring the RFs in relation to the CM
trajectory of the player-club system and comparing how linear and angular impulse generated by
target and rear legs as reflected by the RFh, RFh-angle, moment arm, and stance width.

Methods

Highly skilled players (n = 11; 5 female, 6 male, handicap < 5 [7], right handed) volunteered to
participate in accordance with the local institutional review board. The goal of the experimental
task was to hit a golf ball from their preferred address position toward a specific target downrange
using two different clubs (6-iron and driver, TaylorMade-adidas Golf) as they normally would on
the course. Each player was given adequate time to warm-up prior to hitting golf balls. Players
then performed four to six golf swings as they would normally for each club. The golf swings were
performed in blocks, starting with the 6-iron. Resultant horizontal force (RFh)-time curves of the
rear and target legs for an exemplar player display the level of consistency of RF generation
across four trials performed with each club (Figure 4.1).

Reaction forces (RFs) in three dimensions were measured at the turf-plate interface. Each foot
was supported by a single force plate (1200 Hz, Kistler, Amherst, NY, USA). A thin layer of artificial
turf was secured on top of each force plate to mimic surface characteristics [5]. The instant of
club-ball contact (t = 0s) was identified by using the signal from a microphone captured
simultaneously with RF data (National Instruments, Austin, Texas).

The magnitude (normalized by body weight) and direction of the resultant horizontal RF (RFh)
and the point of wrench application were computed for each leg [5,24]. Point of wrench application
was used as an alternative to the CP to account for the horizontal forces imposed by this activity
(in the absence of horizontal forces, point of wrench application becomes the CP) [24]. The CM
trajectory during the swing was calculated using the initial CM position (estimated as the net point
of wrench application at time of address) and the net linear impulse measured during the swing
based on methods developed by Zatsiorsky and King [25]. Moments about the vertical axis
passing through the CM were calculated as the cross-product of the RFh and position vector from
the CM to the point of wrench application summed with the free moment for each leg [26]. The
interval of interest began when the net moment toward the target became positive near transition
(between backswing and downswing) and ended near ball contact. Angular impulse generated
by each leg was calculated as the area under the moment-time curve during the interval of interest
and normalized by body mass. Linear impulse was calculated as the area under the force-time
Page 30 of 106

curves for both horizontal directions (parallel and perpendicular to the target line) during the same
interval and normalized by body mass. The RFh and linear impulse were directed parallel and
anteriorly perpendicular to the target line (Figure 4.1). The RFh-angle (°) was also expressed
relative to the target line at the time of peak RFh. Components of angular impulse (RFhs, RFh-
angles, and moment arms) were represented as the mean value of the 10 samples measured
before and after the maximum RFh.

Between club differences across the group were determined using the Sign Test (α = .05). Within-
player differences between clubs were also tested to reveal player specific strategies. The
statistical methods used for within-player differences are a subset of methods known as a two-
state linear model. The probability for each variable of any 6-iron trial being less than any driver
trial was calculated within a player for each variable, where each player served as their own
control (R, open-source). Assuming local independence (i.e. no order effect for trials within a
condition), and that club conditions were independent (i.e. not directly tied to each other) for each
subject, p-values were calculated for each player using Cliff’s analog of the Wilcoxon-Mann-
Whitney test [27,28]. This method was chosen because it deals well with small numbers of trials
per condition [28]. A modified, step-down Fisher-type method was then applied to control the
familywise error rate (α = .05) over multiple comparisons where the level of significance becomes
α/k at each k
th
iteration [29–31]. The current statistics provide more flexibility by allowing
heteroscedasticity across players. As the number of trials increase per condition, Cliff’s method
can achieve lower p-values. The modified, step-down Fisher-type technique is dependent upon
the distribution of p-values for each variable measured because the significance level is adjusted
at each step to compensate for multiple comparisons [29–31]. Therefore, the presentation of
within-player results provides a conservative estimate of significant differences between club
conditions.

Results

The net angular impulse generated by an individual significantly increased when swinging with a
driver compared to a 6-iron as a group (p = .001, Figure 4.1, Table 4.1) and individually for nine
of eleven players. Increases in net angular impulse with the driver were achieved by increases in
angular impulse from the rear (p < .001) and target legs (p = .001). Within-player analysis revealed
that four players increased rear angular impulse (Table 4.1) whereas six players increased target
leg angular impulse (Table 4.2) when swinging with the driver as compared to the 6-iron. No
significant differences in relative contribution for either leg to the angular impulse were observed
between 6-iron and driver.

Group comparisons indicate the increase in angular impulse generated by the target leg involved
significant increases in RFh magnitude (p < .001) and RFh-angle (p < .001), however no
significant changes were observed for the rear leg between 6-iron and driver. Within individuals,
target leg angular impulse generation with the driver was increased by increasing the RFh
magnitude for nine of eleven players (Table 4.2). Target leg RFh-angle became more parallel to
the target line with the driver for six of the players. Free moment contribution to the rear (Table
Page 31 of 106

4.1) and target leg angular impulse (Table 4.2) were not significantly different between the 6-iron
and driver as a group.

Net Angular
Impulse
(N∙m∙s∙kg
-1
)
Rear Angular
Impulse
(N∙m∙s∙kg
-1
)
Rear RFh
(%BW)
Rear Moment
Arm (m)
Rear RFh-Angle
(°)
Free Moment
Contribution
(%Rear Ang
Imp)
Player Club Mean SD Mean SD Mean SD Mean SD Mean SD Mean SD
1
6-iron 0.311 0.004 0.191 0.002 21.87 1.09 0.325 0.009 128.02 4.59 44.15 1.17
Driver 0.328 0.003 0.193 0.007 22.90 0.46 0.358 0.010 125.33 1.39
43.23 1.70
2
6-iron 0.206 0.011 0.147 0.010 19.56 1.62 0.205 0.045 129.30 11.62 82.12 5.11
Driver 0.247 0.008 0.169 0.006 19.28 0.49 0.113 0.084 163.48 31.56
76.23 4.34
3
6-iron 0.227 0.005 0.158 0.005 18.42 0.29 0.219 0.091 139.53 22.02 52.33 1.63
Driver 0.256 0.008 0.163 0.009 21.11 1.15 0.086 0.099 170.40 23.12 49.21 2.65
4
6-iron 0.275 0.007 0.124 0.010 24.11 0.77 0.144 0.015 141.62 8.37 52.46 5.80
Driver 0.308 0.005 0.148 0.009 23.50 0.45 0.183 0.015 136.25 6.99 43.60 3.31
5
6-iron 0.298 0.006 0.210 0.008 16.53 0.24 0.316 0.011 118.32 3.77 43.87 5.34
Driver 0.343 0.011 0.237 0.015 16.28 0.28 0.369 0.025 110.99 10.67 38.74 3.57
6
6-iron 0.361 0.008 0.195 0.017 23.68 1.01 0.207 0.058 86.54 1.33 48.03 9.94
Driver 0.410 0.004 0.287 0.017 25.42 1.25 0.398 0.036 82.70 3.10 32.87 3.63
7
6-iron 0.294 0.018 0.246 0.039 25.54 0.91 0.296 0.049 145.39 2.12 43.93 6.11
Driver 0.333 0.010 0.227 0.023 28.79 0.45 0.258 0.020 144.87 1.85
46.12 3.96
8
6-iron 0.307 0.008 0.144 0.003 18.89 0.44 0.287 0.024 111.01 1.58 50.54 2.13
Driver 0.373 0.006 0.159 0.005 20.09 2.15 0.108 0.107 165.90 31.03
49.52 1.73
9
6-iron 0.355 0.009 0.172 0.020 31.40 0.71 0.215 0.017 106.32 3.00 45.23 5.25
Driver 0.417 0.020 0.229 0.017 33.23 1.43 0.310 0.020 108.89 0.73 54.22 8.30
10
6-iron 0.250 0.007 0.166 0.004 24.93 0.60 0.142 0.020 138.78 3.39 67.81 12.18
Driver 0.328 0.003 0.179 0.007 28.45 0.74 0.179 0.002 131.75 4.33 44.97 0.68
11
6-iron 0.322 0.021 0.265 0.041 23.19 1.71 0.285 0.073 142.35 2.80 12.84 3.82
Driver 0.365 0.009 0.282 0.009 26.31 0.65 0.254 0.018 141.80 2.19 16.85 1.53
Increase 9 4 3 5 - 2
Decrease - - - - 1 2
No Change 2 7 8 6 10 7
Group
6-iron 0.291 0.049 0.184 0.044 22.56 4.16 0.240 0.065 126.11 18.64 49.39 16.95
Driver 0.337 0.054 0.207 0.049 24.12 4.92 0.238 0.112 134.76 26.95 45.05 14.44
p <0.001* 0.001* 0.052 0.52 0.52 0.23

Individual player differences based on Cliff's Analog of Wilcoxon-Mann-Whitney Test and modified Fisher

type method step-down technique

p Significant at α = .05 level when adjusted for multiple comparisons

Increase, decrease, no change based on within player analysis at α = .05 level

p* Group differences based on Sign Test at α = .05 level

RFh is the resultant horizontal reaction force

RFh-angle is the angle of the RFh relative to the target line

Table 4.1: Comparison of group and individual net angular impulse and rear leg angular impulse components
Page 32 of 106

Figure 4.2: Mean angular impulse (normalized by body mass) of each players’ golf swings with the 6-iron
(6I) and driver (D) during the interval of interest. Net angular impulse (above) as well as rear and target
leg angular impulse (below) for all players are displayed. Error bars represent ±1 standard deviation. Net
angular impulse is the summation of rear and target contributions for each club. Positive values were
plotted above and below zero for rear and target leg angular impulse (below). All significant differences
were denoted when tested at α = .05 level when adjusted for multiple comparisons. * Net angular impulse
significant difference between clubs, † rear angular impulse significant difference between clubs, ‡ target
angular impulse significant differences between clubs.
Page 33 of 106

Rear leg angular impulse generation between the 6-iron and driver was not seen to significantly
increase with increases in perpendicular distance between the CM and the point of force
application of the rear leg (moment arm length, Table 4.1). However, five of eleven players
displayed a significant increase in rear leg moment arm length. Stance width was seen to
significantly increase with the driver as a group (p < .001) and individually in nine of eleven players.
Differences in downswing duration between clubs were minimal and were not seen to be a major
contributing factor to observed changes in net angular impulse.

Stance Width
(m)
Target Angular
Impulse
(N∙m∙s∙kg
-1
)
Target RFh
(%BW)
Target
Moment Arm
(m)
Target RFh-
Angle (°)
Free Moment
Contribution
(%Target Ang
Imp)
Player Club Mean SD Mean SD Mean SD Mean SD Mean SD Mean SD
1
6-iron 0.470 0.011 0.120 0.004 32.70 0.74 0.214 0.011 279.44 4.23 27.30 1.90
Driver 0.507 0.003 0.135 0.006 39.18 0.77 0.204 0.010 291.65 1.53 26.84 1.25
2
6-iron 0.476 0.014 0.058 0.009 23.18 3.54 0.148 0.056 296.13 22.11 55.80 6.76
Driver 0.560 0.030 0.078 0.004 28.91 0.82 0.157 0.021 308.37 8.01 48.41 5.16
3
6-iron 0.434 0.011 0.069 0.005 23.07 0.90 0.135 0.003 289.70 1.89 44.91 4.16
Driver 0.538 0.010 0.094 0.006 29.93 1.00 0.109 0.010 311.14 1.41 52.52 4.74
4
6-iron 0.575 0.005 0.150 0.006 36.29 1.80 0.229 0.009 301.57 2.61 28.11 2.86
Driver 0.631 0.005 0.160 0.007 41.99 1.15 0.188 0.010 315.16 2.82 30.91 3.76
5
6-iron 0.402 0.021 0.088 0.006 14.41 0.31 0.182 0.011 260.74 1.45 47.25 2.57
Driver 0.513 0.010 0.106 0.012 17.09 0.66 0.163 0.020 264.87 4.79 47.02 6.89
6
6-iron 0.505 0.029 0.166 0.021 34.65 2.49 0.246 0.038 280.37 16.66 -1.76 3.11
Driver 0.658 0.012 0.123 0.019 40.45 1.33 0.089 0.041 320.99 12.85 -7.52 8.40
7
6-iron 0.463 0.010 0.047 0.022 33.85 0.86 0.088 0.039 300.72 0.97 98.89 52.60
Driver 0.527 0.007 0.106 0.013 35.85 0.60 0.139 0.015 308.65 0.63 45.49 6.85
8
6-iron 0.497 0.020 0.163 0.007 34.04 0.90 0.206 0.015 285.79 3.35 44.74 4.89
Driver 0.600 0.009 0.214 0.006 38.48 1.05 0.246 0.005 294.09 1.50 42.06 1.40
9
6-iron 0.516 0.008 0.182 0.017 32.65 0.67 0.266 0.013 271.59 2.22 37.44 5.64
Driver 0.566 0.009 0.188 0.033 37.24 1.46 0.114 0.076 319.68 21.49 31.88 3.18
10
6-iron 0.452 0.013 0.083 0.009 23.08 0.69 0.226 0.005 273.43 5.55 13.23 2.87
Driver 0.568 0.015 0.148 0.008 30.73 1.61 0.227 0.008 292.64 3.28 8.35 2.27
11
6-iron 0.569 0.018 0.057 0.020 21.17 0.95 0.031 0.053 314.52 1.86 6.94 20.35
Driver 0.639 0.013 0.083 0.006 23.50 1.09 0.054 0.029 318.66 9.24 -8.40 3.41
Increase 9 6 9 2 6 -
Decrease - 1 - 2 - 1
No Change 2 4 2 7 5 10
Group
6-iron 0.485 0.055 0.108 0.050 28.10 7.28 0.179 0.072 286.73 15.69 36.62 27.48
Driver 0.571 0.054 0.131 0.044 33.03 7.78 0.154 0.060 304.17 16.96 28.87 22.08
p <0.001* 0.001* <0.001* 0.97 <0.001* 0.066

Individual player differences based on Cliff's Analog of Wilcoxon-Mann-Whitney Test and modified Fisher

type method step-down technique

p
Significant at α = .05 level when adjusted for multiple comparisons

Increase, decrease, no change based on within player analysis at α = .05 level

p*
Group differences based on Sign Test at α = .05 level

RFh is the resultant horizontal reaction force

RFh-angle is the angle of the RFh relative to the target line

Table 4.2: Comparison of group and individual net stance width and target leg angular impulse components
Page 34 of 106

Net linear impulse perpendicular to target line remained near zero as a group (p = .55, Figure 4.3).
Rear and target leg impulse were anterior and posterior to the target line, respectively, when using
both clubs (Figure 4.3). Net linear impulse parallel to the target line decreased in magnitude with
the driver (p = .001, Figure 4.4). Rear leg linear impulse contributed to CM velocity towards the
target whereas target leg contributions acted in opposition (Figure 4.4).

Page 35 of 106

Figure 4.3: Mean linear impulse (normalized by body mass) perpendicular to the target line of each players’
golf swings with the 6-iron (6I) and driver (D) during the interval of interest. Net linear impulse (above) as
well as rear and target leg linear impulse (below) for all players are displayed. Error bars represent ±1
standard deviation. Net linear impulse is the summation of rear and target contributions for each club. All
significant differences were denoted when tested at α = .05 level when adjusted for multiple comparisons.
* Net linear impulse significant difference between clubs, † rear linear impulse significant difference
between clubs, ‡ target linear impulse significant differences between clubs.
Page 36 of 106

Figure 4.4: Mean linear impulse (normalized by body mass) parallel to the target line of each players’ golf
swings with the 6-iron (6I) and driver (D) during the interval of interest. Net linear impulse (above) as well
as rear and target leg linear impulse (below) for all players are displayed. Error bars represent ±1 standard
deviation. Net linear impulse is the summation of rear and target contributions for each club. All significant
differences were denoted when tested at α =.05 level when adjusted for multiple comparisons. * Net linear
impulse significant difference between clubs, † rear linear impulse significant difference between clubs, ‡
target linear impulse significant differences between clubs.
Page 37 of 106

Discussion

The skilled players in this study regulated angular impulse generation during the golf swing by
involving both the rear and target legs. Mechanisms used to regulate angular impulse generation
between clubs varied across players and involved coordination between the legs. Different
mechanisms for regulating angular impulse generated by each leg (e.g. modify RFh magnitude
and/or moment arm) are important to consider when designing interventions that aim to improve
player preparation and facilitate technique modifications. The determination of how skilled players
coordinate the linear and angular impulse generated by each leg during the golf swing may also
assist in the identification of effective strategies individuals can use to regulate RF in relation to
their CM trajectory while maintaining balance in other activities of daily life.

The results of this study are specific to highly skilled players with low handicaps and support
previous research that indicates RFs generated during golf swings with the same club are
consistent within-player and vary between players [1]. Within-player differences in RFs during
swings with the same club under different experimental conditions tend to be consistent for skilled
players and are relatively small in magnitude (e.g. regulation of shot distance with a 6-iron
produced differences of ~2-5 %BW) [1]. To minimize the effect of experimental conditions on the
results of this investigation, we chose to standardize the surface and club conditions comparable
to those used by players when practicing swings on a driving range. The artificial turf used to
cover the force plates was thin to provide practice-like friction conditions, while minimizing any
attenuation of force transmitted to the force-plates under each foot [1,5]. The kinetics-based CM
trajectory used in this study was used to improve resolution of the angular impulse computed, and
minimize propagation of error from segment kinematic based estimates of body segment
parameters [32,33]. The inclusion criteria for this exploratory study limited sample size and trials
per participant and as such the statistical analysis was chosen because it deals well with small
numbers of trials per condition. Increasing the number of trials per condition could strengthen the
findings. Incorporation of a within-player experimental design allowed for further insight as to how
individual players use their target and rear legs to regulate the net angular impulse generation
when they swing with each club.

As a group, the players in this study were found to increase the net angular impulse when using
the driver as compared to the 6-iron. This was achieved by coordinating increases in rear and
target leg angular impulse while generating the linear impulse required to maintain the CM over
the base of support. Individuals employed a variety of strategies to increase net angular impulse
through the modulation of multiple factors. During the interval of interest, the target leg RFh acted
posteriorly and away from the target while the rear leg RFh acted anteriorly and toward the target,
consistent with that reported during swings in golf and baseball batting [1,5,16,17]. Generation of
these RFhs allowed the legs to work together to generate linear impulse parallel to the target and
the net angular impulse needed to generate clubhead speed. Coincidentally, the RFhs generated
by the legs minimized translation of the CM perpendicular to the target line thereby preserving
balance over the base of support. The position of the CM relative to the point of RF application
contributed to variations in moment arm length that affected net angular impulse contributions
from both legs. Along with CM trajectory, the observed increase in stance width associated with
Page 38 of 106

the driver provided another means to alter the length of the moment arms. However, net linear
impulse (and subsequently, CM position change) parallel to the target decreased with the driver
while rear leg RFh and RFh-angle remained consistent within-player and between clubs. This
indicates that increases in stance width can contribute to larger moment arms and as a result,
contribute to increases in rear leg angular impulse.

The results of this study have identified how individual players regulate linear and angular impulse
when using a 6-iron as compared to a driver. While factors affecting angular impulse have been
studied previously (e.g. RFh regulation and CP patterns) [1,10,15,34], the mechanisms used by
each leg to increase angular impulse during the swing have not been delineated. Consistent with
prior research, duration of impulse generation was not a means by which angular impulse was
increased [1,5,15]. The percentage amount of angular impulse contributed by the rear and target
legs toward net angular impulse did not change between the 6-iron and driver. Similarly, the
percentage amount of angular impulse contributed by the free moment to either the rear or target
leg was not found to change across the group between the 6-iron and driver. This suggests the
percentage of angular impulse from each leg and free moment involvement toward net angular
impulse regulation remains consistent across clubs.

Rear leg angular impulse increased through a variety of factors. Rear leg RFh and RFh-angle
were generally maintained across clubs, consistent with that observed when individual players
regulate shot distance with a 6-iron [1]. The use of this mechanism to increase angular impulse
between clubs allows the rear leg to retain similar control patterns to regulate force across clubs.
Because of this, training the rear leg to generate RFh and RFh-angle consistently regardless of
the task or CM trajectory may benefit performance and provide the player with an opportunity to
increase force magnitude as another means to increase angular impulse.

The target leg contribution to increases in angular impulse generation between club conditions
mostly relied upon increases in RFh while generally maintaining or decreasing the target leg
moment arm. Because the target leg requires active control of force generation, this leg may be
predisposed toward mechanical load exposure. This may require a stronger focus on strength
and stability for the target leg to mitigate injury concern and increase performance. Research into
muscle activity patterns of the legs, physical examinations, and player injury histories has the
potential to further elucidate details of the actions of the legs to regulate impulse. By further
understanding the techniques golf players use, key insights into injury prevention and
performance could be discovered.

Knowledge of how players regulate impulse when using different clubs can assist when
personalizing interventions and modifying technique [6]. Because individuals employ unique
strategies to regulate impulse, players may benefit from feedback that is tailored to meet the
control objective of each leg. The strategies used by skilled golf players to regulate impulse
generation during the swing may also provide meaningful information to guide interventions used
by clinicians when assisting populations with balance-related impairments. Completion of the golf
swing requires the regulation of linear and angular impulse to achieve the task and simultaneously
maintain balance. Playing golf, similar to T’ai Chi [23,35], may serve as an alternative community-
Page 39 of 106

based rehabilitation intervention for maintaining and improving balance. Performing an engaging
goal-directed activity like a golf swing may provide those affected by balance-related impairments
with an opportunity to practice coordinating RF generation with both legs while maintaining
balance, which can become a difficult task for these populations [4,19,22].

In summary, the results of this study indicate that linear and angular impulse generated during
the downswing was coordinated between the legs when using the 6-iron and driver. Net angular
impulse was found to increase with the driver through contributions from the rear and target legs.
The skilled players in this study were found to use different mechanisms to regulate angular
impulse generation by each leg. In general, participants in this study modified target leg resultant
horizontal reaction force and orientation, while some participants in this study modified rear leg
moment arm increases in net angular impulse with the driver. This suggests the target leg may
benefit from a focus on strength and stability to effectively produce force through a range of
orientations to increase target leg angular impulse. Regardless of the angular impulse generation
mechanism used by an individual, net linear impulse was effectively mitigated perpendicular to
the target and directed parallel along the target line, all while the CM trajectory remained over the
base of support. These findings provide a basis for future work that investigates advantages of
different angular impulse generation mechanisms when encountering challenges experienced in
the course of play.
Page 40 of 106

References

[1] McNitt-Gray, J. L., Munaretto, J., Zaferiou, A., Requejo, P. S., and Flashner, H., 2013,
“Regulation of reaction forces during the golf swing,” Sport. Biomech., 12(2), pp. 121–131.
[2] Chu, Y., Sell, T. C., and Lephart, S. M., 2010, “The relationship between biomechanical
variables and driving performance during the golf swing.,” J. Sports Sci., 28(11), pp. 1251–9.
[3] Egret, C. I., Vincent, O., Weber, J., Dujardin, F. H., and Chollet, D., 2003, “Analysis of 3D
kinematics concerning three different clubs in golf swing,” Int. J. Sports Med., 24(6), pp. 465–469.
[4] Mathiyakom, W., McNitt-Gray, J. L., and Wilcox, R., 2006, “Lower extremity control and
dynamics during backward angular impulse generation in forward translating tasks.,” J. Biomech.,
39(6), pp. 990–1000.
[5] Williams, K., and Cavanagh, P. R., 1983, “The mechanics of foot action during the golf
swing and implications for shoe design.,” Med. Sci. Sports Exerc., 15(3), pp. 247–255.
[6] McNitt-Gray, J. L., Sand, K., Ramos, C., Peterson, T., Held, L., and Brown, K., 2015,
“Using technology and engineering to facilitate skill acquisition and improvements in performance,”
Proc. Inst. Mech. Eng. Part P J. Sport. Eng. Technol., 229(2), pp. 103–115.
[7] Worsfold, P., Smith, N., and Dyson, R., 2008, “Low handicap golfers generate more
torque at the shoe-natural grass interface when using a driver,” J. Sport. Sci. Med., (7), pp. 408–
414.
[8] Mathiyakom, W., and McNitt-Gray, J. L., 2008, “Regulation of angular impulse during fall
recovery,” J Rehabil Res Dev, 45(8), pp. 1237–1247.
[9] Pijnappels, M., Bobbert, M. F., and van Dieën, J. H., 2005, “Push-off reactions in recovery
after tripping discriminate young subjects, older non-fallers and older fallers.,” Gait Posture, 21(4),
pp. 388–94.
[10] Ball, K. A., and Best, R. J., 2007, “Different centre of pressure patterns within the golf
stroke I: Cluster analysis.,” J. Sports Sci., 25(7), pp. 757–70.
[11] Ball, K. A., and Best, R. J., 2007, “Different centre of pressure patterns within the golf
stroke II: group-based analysis.,” J. Sports Sci., 25(7), pp. 771–9.
[12] Gatt, C., Pavol, M., Parker, R., and Grabiner, M., 1998, “Three-dimensional knee joint
kinetics during a golf swing influences of skill level and footwear,” Am. J. Sports Med., (Idd).
[13] Mathiyakom, W., McNitt-Gray, J. L., and Wilcox, R., 2007, “Regulation of angular impulse
during two forward translating tasks.,” J. Appl. Biomech., 23(2), pp. 149–61.
[14] Queen, R., and Butler, R., 2013, “Difference in Peak Weight Transfer and Timing Based
on Golf Handicap,” J. Strength Conditiioning Res., 27(9), pp. 2481–2486.
[15] Nagao, N., and Sawada, Y., 1974, “A kinematic analysis of the golf swing by means of
fast motion picture in connection with racial difference.,” J. Sports Med. Phys. Fitness, 14(1), pp.
55–62.
Page 41 of 106

[16] Mason, B. R., 1987, “Ground Reaction Forces of Elite Australian Baseball Batters,”
Biomechanics XB, B. Jonsson, ed., Human Kinetics Publishers, Champaign, IL, pp. 749–752.
[17] Messier, S. P., and Owen, M. G., 1985, “The Mechanics of Batting: Analysis of Ground
Reaction Forces and Selected Lower Extremity Kinematics,” Res. Q. Exerc. Sport, 56(2), pp. 138–
143.
[18] Mathiyakom, W., McNitt-Gray, J. L., and Wilcox, R., 2006, “Lower extremity control and
dynamics during backward angular impulse generation in backward translating tasks.,” Exp. Brain
Res., 169(3), pp. 377–88.
[19] Bosse, I., Oberländer, K. D., Savelberg, H. H., Meijer, K., Brüggemann, G.-P., and
Karamanidis, K., 2012, “Dynamic stability control in younger and older adults during stair descent.,”
Hum. Mov. Sci., 31(6), pp. 1560–70.
[20] Glaister, B. C., Orendurff, M. S., Schoen, J. a, Bernatz, G. C., and Klute, G. K., 2008,
“Ground reaction forces and impulses during a transient turning maneuver.,” J. Biomech., 41(14),
pp. 3090–3.
[21] Jindrich, D. L., Besier, T. F., and Lloyd, D. G., 2006, “A hypothesis for the function of
braking forces during running turns.,” J. Biomech., 39(9), pp. 1611–20.
[22] Vrieling, A. H., van Keeken, H. G., Schoppen, T., Otten, E., Halbertsma, J. P. K., Hof, A.
L., and Postema, K., 2008, “Gait termination in lower limb amputees,” Gait Posture, 27(1), pp.
82–90.
[23] Tsang, W. W. N., and Hui-Chan, C. W. Y., 2004, “Effects of Exercise on Joint Sense and
Balance in Elderly Men: Tai Chi versus Golf,” Med. Sci. Sports Exerc., 36(4), pp. 658–667.
[24] Cavanagh, P. R., Rodgers, M. M., and Iiboshi, A., 1987, “Pressure distribution under
symptom-free feet during barefoot standing.,” Foot Ankle, 7(5), pp. 262–276.
[25] King, D. L., and Zatsiorsky, V. M., 1997, “Extracting gravity line displacement from
stabilographic recordings,” Gait Posture, 6(1), pp. 27–38.
[26] Holden, J. P., and Cavanagh, P. R., 1991, “The free moment of ground reaction in
distance running and its changes with pronation.,” J. Biomech., 24(10), pp. 887–97.
[27] Cliff, N., 1996, Ordinal Methods for Behavioral Data Analysis, Lawrence Erlbaum
Associates, Inc., Publishers, Mahwah, NJ.
[28] Neuhäuser, M., Lösch, C., and Jöckel, K.-H., 2007, “The Chen–Luo test in case of
heteroscedasticity,” Comput. Stat. Data Anal., 51(10), pp. 5055–5060.
[29] Hochberg, Y., 1988, “A sharper Bonferroni test for multiple tests of significance,”
Biometrika, 75, pp. 800–802.
[30] Hochberg, Y., and Tamhane, A. C., 1987, Multiple Comparison Procedures, John Wiley
& Sons, Inc., Hoboken, NJ, USA.
[31] Wilcox, R., and Clark, F., 2015, “Robust Multiple Comparisons Based on Combined
Probabilities From Independent Tests,” J. Data Sci., 13(1), pp. 1–11.
Page 42 of 106

[32] Lenzi, D., Cappello, A., and Chiari, L., 2003, “Influence of body segment parameters and
modeling assumptions on the estimate of center of mass trajectory,” J. Biomech., 36(9), pp. 1335–
1341.
[33] Zatsiorsky, V. M., and King, D. L., 1997, “An algorithm for determining gravity line location
from posturographic recordings,” J. Biomech., 31(2), pp. 161–164.
[34] Ball, K. A., and Best, R. J., 2011, “Golf styles and centre of pressure patterns when using
different golf clubs.,” J. Sports Sci., 29(6), pp. 587–90.
[35] Shih, J., 1997, “Basic Beijing Twenty-Four Forms of T’Ai Chi Exercise and Average
Velocity of Sway,” Percept. Mot. Skills, 84, pp. 287–290.

Page 43 of 106

: Angular Impulse and Balance Control while
Increasing Golf Shot Distance within a Club

Introduction

The primary goal of golf is to place the ball in the hole with the fewest amount of strokes. By
increasing the distance with the tee shot, the player is left with a shorter distance to cover with
their remaining strokes. Increasing the distance with the driver may also assist players in using
fewer strokes per hole. Multiple studies have researched golf driving and its role in the game of
golf [1–3]. While not on the tee, golf players can use the varying length and loft characteristics of
the irons to regulate shot distance on the course. However, within the driver, the player must
regulate how they use their personal resources as they interface with the club to increase shot
distance. It has been shown that highly skilled players will regulate their linear and angular impulse
to control golf shot distance and maintain balance [4]. Understanding how individuals control golf
shot distance with the driver by regulating the reaction forces on the ground in relation to their
center of mass trajectory can provide insights toward player preparation and interventions for
player performance [4–6].
Previous research studying linear and angular impulse between the 6-iron and driver revealed
that players increased angular impulse (~0.046 N*m*s/kg) from both the rear and target legs while
maintaining balance over the base of support when hitting with the driver [4]. This primarily
occurred from increases in target leg RFh, RFh-angle, where some individuals displayed an
increase in rear leg moment arm as a method to increase angular impulse [4]. Within the 6-iron,
players have been found to decrease the RFh with the rear or target legs, while maintaining a
similar RFh-angle as they reduced shot distance [5]. These findings provide evidence that players
may regulate RFh and impulse measures as they increase golf shot distance within the driver.
This study aimed to determine how highly skilled golf players controlled linear and angular impulse
during the golf swing with a driver as they hit normally, and as they attempted to increase driver
shot distance. We hypothesized that angular impulse would increase with increasing distance as
a result of contributions from both the rear and target legs. Additionally, we hypothesized that
increases in rear leg and target leg resultant horizontal reaction force as well as target leg RFh-
angle more parallel to the target line would contribute to each leg’s contribution to angular impulse.
We also hypothesized that linear impulse would remain unchanged across conditions to maintain
balance over the base of support. We tested these hypotheses by measuring the reaction forces
and their interaction with the center of mass of the player and comparing the linear and angular
impulse regulated by the player through the components of RFh, RFh-angle, and moment arm.
Materials and Methods

Skilled players (n = 9, handicap < 5, 4 female, 5 male, 24.9 (10.7) years old) volunteered to
participate in accordance with the local institutional review board. Each player performed golf
shots towards a target placed behind a net with a driver (D, TaylorMade-adidas Golf, Carlsbad,
CA, USA). Players were asked to hit straight shots a distance they normally would on the course
(Normal) as well as attempting to hit the ball 10 yards further (+10). Golf shots were performed in
Page 44 of 106

blocks, starting with the Normal condition. Each player was given adequate time to complete their
normal warm-up routine prior to collection of golf swings.

Golf shots were initiated using the player’s preferred address position with each foot fully
supported by a force plate (1200 Hz, Kistler, Amherst, NY, USA). The magnitude (normalized by
body weight) and direction of the peak resultant horizontal reaction forces at the force plate
interface and point of wrench application (PWA) were computed for each leg [4,7,8]. The center
of mass trajectory during the golf swing was calculated by estimating an initial CM position as the
PWA at time of address, and driving the CM forward using the net linear impulse during the swing
[9,10].
Moments about the vertical axis passing through the CM were calculated as the cross-product of
the RFh and position vector from the CM to the PWA (moment arm) summed with the free moment
for each leg [4,11]. The interval of interest was defined as the phase between transition and ball
contact when the net moment toward the target was positive. Angular impulse was calculated as
the area under the moment curves for each leg during the interval of interest and was normalized
by body mass. Linear impulse was calculated as the area under the force-time curves for both
horizontal directions (parallel and perpendicular to the target line, Figure 5.1) during the interval
of interest and were normalized by body mass. Resultant horizontal reaction force and and RFh-
angle were also expressed relative to the target line. Components of angular impulse (RFhs, RFh-
angles, and moment arms) were represented as the mean value of the 10 samples measured
before and after the maximum RFh.
Between condition differences across the group were determined using the Sign Test (α = 0.05).
Within-player differences between conditions were also tested to further understand player-
specific strategies [4,12,13]. The probability for each variable of any Normal trial being less than
any +10 trial was calculated within a player for each variable, where each player served as their
own control (R, open-source). Assuming local independence (i.e., no order effect for trials within
a condition), and that club conditions were independent (i.e., not directly tied to each other) for
each subject, p-values were calculated for each player using Cliff’s analog of the Wilcoxon-Mann-
Whitney test [14,15]. A modified, step-down Fisher-type method was then applied to control the
familywise error rate (α = .05) over multiple comparisons where the level of significance becomes
α/k at each kth iteration [16–18].
Figure 5.1: Top-down view of force plate setup and reference system for reaction force and angle orientation
in the transverse plane.
Page 45 of 106

Results

The net angular impulse significantly increased across the group ~0.008 N*m*s/kg (Figure 5.3)
during the +10 yard condition compared to Normal (p = 0.004). Increases in net angular impulse
in the +10 yard condition were achieved by increases in target leg AI (p = 0.024). Although group
differences were revealed, individual results did not discover differences within a player for any
components of linear or angular impulse.

Figure 5.3A: Mean angular impulse (normalized by body weight)
of each players’ golf swings with the driver in the Normal (N) and
plus 10 yards (+10) conditions during the interval of interest. Net
angular impulse (above) as well as rear and target leg angular
impulse (below) for all players are displayed. Error bars
represent ±1 standard deviation. Positive values were plotted
above and below zero for rear and target leg angular impulse
(below).
Figure 5.3B: Mean peak target leg resultant horizontal
reaction force (RFh, top) and RFh-angle (bottom) of each
players’ golf swings with the driver in the normal (N) and
increased (+10) conditions (top). Error bars represent ±1
standard deviation.
Page 46 of 106

Target leg angular impulse significantly increased due to RFh magnitude (Figure 5.3, p = 0.002)
and RFh-angle (p = 0.004). Rear leg angular was not found to significantly change across
conditions (p = 0.47). However, rear leg RFh magnitude increased (p = 0.024) while rear leg free
moment contribution to rear leg AI significantly decreased during the +10 yard condition (p =
0.004). Rear leg RFh-angle was not found to significantly change between conditions (p = 0.96).
Neither the rear (p = 0.51) nor target (p = 0.18) leg changed the moment arm to regulate AI
generation.

Figure 5.5A: Mean linear impulse (normalized by mass)
perpendicular to the target line of each players’ golf swings
with the driver in the normal (D) and plus 10 yards (+10)
conditions during the interval of interest. Net linear impulse
(above) as well as rear and target leg linear impulse (below)
for all players are displayed. Error bars represent ±1
standard deviation. Significant differences across the group
were discovered in net linear impulse (p = .024) and target
leg linear impulse (p = .004). Net linear impulse is the
summation of rear and target leg contributions for each
club. All significant differences were denoted when tested
at α = 0.05 level when adjusted for multiple comparisons. *
Net linear impulse significant difference between
conditions, † rear linear impulse significant difference
between clubs, ‡ target linear impulse significant
differences between clubs.
Figure 5.5B: Mean linear impulse (normalized by mass)
parallel to the target line of each players’ golf swings with
the 6-iron (6I) and driver (D) during the interval of interest.
Net linear impulse (above) as well as rear and target leg
linear impulse (below) for all players are displayed. Error
bars represent ±1 standard deviation. Significant
differences across the group were discovered in net linear
impulse (p = .004). Net linear impulse is the summation of
rear and target contributions for each club. All significant
differences were denoted when tested at α = 0.05 level
when adjusted for multiple comparisons. * Net linear
impulse significant difference between clubs.
Page 47 of 106

Net linear impulse perpendicular to target line remained near zero, but increased in the +10
condition as a group (p = .024, Figure 5.5A). Rear (p = .51) and target leg (p = .004) linear impulse
were anterior and posterior to the target line, respectively, when using both clubs (Figure 5.5A).
Net linear impulse parallel toward the target line decreased in magnitude with the driver (p = .004,
Figure 5.5B). Rear leg linear impulse contributed to CM velocity towards the target whereas target
leg contributions acted in opposition (Figure 5.5B).

Discussion

Highly skilled golf players in this study regulated angular impulse within the driver through the
contributions of both the rear and target legs. Players coordinated the legs to generate angular
impulse through different mechanisms. For example, the target leg increased RFh and RFh-angle,
while the rear leg increased RFh, while decreasing free moment contributions. Linear impulse
regulation acted to decrease the magnitude of CM velocities in the +10 condition. The strategies
to regulate linear and angular impulse within the driver may provide insights into the ability of
players to control the interaction of their RFs with the CM trajectory to increase angular impulse
generation.
The results of this study are particular to highly skilled golf players. The measurement of reaction
forces necessitated the use of a laboratory environment, however the consistency and amplitude
of forces were consistent with previous research of similar populations [4,5,19]. Although players
performed limited trials per condition, the statistical methods used in this study have been chosen
to perform well under these circumstances [14–18].
Across the group, net angular impulse increased as the required shot distance increased within
the driver. Players amplified target leg angular impulse by increasing the target leg RFh
magnitude and RFh-angle to be more parallel to the target line in the +10 condition. These findings
are consistent with previous research on golf shot distance regulation between the 6-iron and
driver [4]. As found previously (Chapter 4), increases in peak RFh and RFh-angle suggests an
increase and in muscle activation and possible recruitment of additional muscles in the target leg
to meet the mechanical demand. The redirection of the RFh in the +10 condition may also alter
the imposed joint moments on the target leg that the player has to control.
No significant differences in rear leg angular impulse were discovered across conditions. This
occurred while the rear leg RFh increased in the +10 yard condition, which is aligned with previous
research [4,5]. While rear leg RFh increased, the rear leg free moment decreased. The reduction
of angular impulse generation from the rear leg free moment may have counteracted the effect of
increasing rear leg RFh on increasing angular impulse which lead to no differences in total rear
leg angular impulse across conditions. These results suggest that the regulation of the target leg
RFh and RFh-angle are key contributors to increasing angular impulse while increasing shot
distance with the driver. This may be a mechanism to simplify control of the net angular impulse
to only the target leg.
Net linear impulse in both directions decreased the magnitude of the CM velocity when hitting the
driver under the +10 condition. This is consistent with the increase in magnitude and reorientation
of the target leg reaction force to be more parallel and away from the target. Players may find that
generating reaction forces in this way allowed them to increase angular impulse while possibly
Page 48 of 106

decreasing the length of target leg moment arm through the orientation and smaller effect of linear
impulse to move the center of mass.
In this study, players were asked to hit a driver further than they normally would. While players
were able to complete the swings in the +10 condition successfully, the effective increases in net
angular impulse were relatively small in magnitude compared to increases in net angular impulse
between a 6-iron and driver [4]. The driver is the club designed to hit the furthest, where the 6-
iron is designed as a mid-distance club. As a result, players may be at the upper limits of their
ability to generate angular impulse. This supports the minimal increases in angular impulses
observed in this study when players attempted to increase driver shot distance. But how does net
angular impulse link to performance on the course? Once a player reaches the upper limits of golf
shot distance, limiting modifications to only target leg RFh and RFh-angle generation may have
advantages for golf shot consistency and accuracy. Allowing regulation of golf shot distance with
both legs, as observed in the 6-iron, may provide the player more options when encountering
different terrain on the course [4,5].
Highly skilled players with low handicaps were found to generate more net angular impulse when
increasing shot distance within the driver. The regulation of net angular impulse was achieved
primarily from the target leg through modifications in target leg RFh magnitude and orientation.
Increases in net angular impulse within the driver were found to be relatively small as compared
to modifications in angular between clubs (6-iron vs driver). Changes in RFh magnitudes and
orientations also may have decreased the net linear impulse magnitudes both perpendicular and
parallel to the target line. When using the driver, players may be reaching their upper limits of
their angular impulse generation ability. This study suggests players simplify their control
strategies with the driver to increase angular impulse, which may have advantages in regards in
golf driving performance.

Page 49 of 106

References

[1] Hume, P. a, Keogh, J., and Reid, D., 2005, “The role of biomechanics in maximising
distance and accuracy of golf shots.,” Sports Med., 35(5), pp. 429–49.
[2] Zhang, X., and Shan, G., 2013, “Where do golf driver swings go wrong? Factors
influencing driver swing consistency.,” Scand. J. Med. Sci. Sports, pp. 1–9.
[3] Nagao, N., and Sawada, Y., 1973, “A kinematic analysis in golf swing concerning driver
shot and No. 9 iron shot,” J. Sports Med. Phys. Fitness, 13(1), pp. 4–16.
[4] Peterson, T. J., Wilcox, R. R., and McNitt-Gray, J. L., 2016, “Angular impulse and balance
regulation during the golf swing,” J. Appl. Biomech., 32(4), pp. 342–349.
[5] McNitt-Gray, J. L., Munaretto, J., Zaferiou, A., Requejo, P. S., and Flashner, H., 2013,
“Regulation of reaction forces during the golf swing,” Sport. Biomech., 12(2), pp. 121–131.
[6] McNitt-Gray, J. L., Sand, K., Ramos, C., Peterson, T. J., Held, L., and Brown, K., 2015,
“Using technology and engineering to facilitate skill acquisition and improvements in
performance,” Proc. Inst. Mech. Eng. Part P J. Sport. Eng. Technol., 229(2), pp. 103–115.
[7] Shimba, T., 1996, “Consequences of Force Platform Studies,” Natl. Rehabiliation Res. Bull.
Japan, (17), pp. 17–23.
[8] Zatsiorsky, V. M., 2002, “Wrench Representation of the Ground Reaction Force,” Kinetics
of Human Motion, Human Kinetics Publishers, Champaign, IL, pp. 43–48.
[9] King, D. L., and Zatsiorsky, V. M., 1997, “Extracting gravity line displacement from
stabilographic recordings,” Gait Posture, 6(1), pp. 27–38.
[10] Lenzi, D., Cappello, A., and Chiari, L., 2003, “Influence of body segment parameters and
modeling assumptions on the estimate of center of mass trajectory,” J. Biomech., 36(9),
pp. 1335–1341.
[11] Holden, J. P., and Cavanagh, P. R., 1991, “The free moment of ground reaction in distance
running and its changes with pronation.,” J. Biomech., 24(10), pp. 887–97.
[12] Russell, I. M., Raina, S., Requejo, P. S., Wilcox, R. R., Mulroy, S., and McNitt-Gray, J. L.,
2015, “Modifications in Wheelchair Propulsion Technique with Speed,” Front. Bioeng.
Biotechnol., 3(October), pp. 1–11.
[13] Zaferiou, A. M., Wilcox, R. R., and McNitt-Gray, J. L., 2016, “Modification of Impulse
Generation during Pirouette Turns with Increased Rotational Demands,” J. Appl. Biomech.,
pp. 1–44.
[14] Cliff, N., 1996, Ordinal Methods for Behavioral Data Analysis, Lawrence Erlbaum
Associates, Inc., Publishers, Mahwah, NJ.
[15] Neuhäuser, M., Lösch, C., and Jöckel, K.-H., 2007, “The Chen–Luo test in case of
heteroscedasticity,” Comput. Stat. Data Anal., 51(10), pp. 5055–5060.
[16] Hochberg, Y., 1988, “A sharper Bonferroni test for multiple tests of significance,”
Biometrika, 75, pp. 800–802.
[17] Hochberg, Y., and Tamhane, A. C., 1987, Multiple Comparison Procedures, John Wiley &
Sons, Inc., Hoboken, NJ, USA.
Page 50 of 106

[18] Wilcox, R., and Clark, F., 2015, “Robust Multiple Comparisons Based on Combined
Probabilities From Independent Tests,” J. Data Sci., 13(1), pp. 1–11.
[19] Barrentine, S. W., Fleisig, G. S., and Johnson, H., 1994, “Ground Reaction Forces and
Torques of Professional and Amateur Golfers,” Sci. Golf II Proc. 1994 World Sci. Congr.
Golf, 1(1), pp. 33–39.

Page 51 of 106

: Angular Impulse and Balance Regulation during
the Golf Swing with the 9-iron and 6-iron

Introduction

The game of golf can be divided into three areas: driving, iron play, and putting. Driving the golf
ball off the tee is meant to minimize the distance left to the hole while positioning for the optimal
following shot. Iron play allows players to either set up future shots or to position the ball closest
to the pin to minimize the distance for the putt. In either case, regulating distance is vital within
the set of irons for optimal performance on the golf course. The set of irons provides a range of
clubs that vary in length, loft, and lie characteristics that allow players to produce golf shots of
varying distances and trajectories. Previous research has demonstrated that players will regulate
their linear and angular impulse to meet the mechanical demand as golf shot distance is
modulated [1–3]. Understanding how players regulate these measures between different irons
will provide insights and add to the knowledge of the entire set of clubs.

Many studies have evaluated the golf shot with a range of clubs (pitching wedge, 9-iron, 7-iron,
5-iron, 4-iron, 3-wood, driver), but these were largely based on kinematics or electromyography
[4–9]. Previous research has determined reaction forces decreased with reducing shot distance
within a 6-iron or driver [1,3] . Other groups have determined reaction forces increase as players
control linear and angular impulse between as they increase shot distance between irons and
drivers [2,10]. One study found vertical ground reaction forces to increase with shorter irons
compared to the driver [11]. However, they also found horizontal reaction forces to increase when
using the driver compared to either a 3-iron or 7-iron iron, but reported no differences between
irons [11]. There has also been research on muscular activation suggesting players may regulate
iron shots differently than driver shots [3,12]. The previous work suggests there is a need to
understand the regulation of reaction forces between clubs within the set of irons as players
regulate linear and angular impulse.

In this study, we aimed to determine how highly skilled golf players control linear and angular
impulse during the golf swing as they increased shot distance between the 9-iron and the 6-iron.
We hypothesized that net angular impulse would increase with the 6-iron due to contributions
from both the rear and target legs. We predicted that these increases would be due to increases
in RFh with minimal modification of the RFh-angle relative to the target line. We also hypothesized
that linear impulse would not change between conditions to preserve balance over the base of
support.

Methods

Skilled players (n = 10, handicap < 5) volunteered to participate in accordance with the local
institutional review board. Each player performed golf shots towards a target placed behind a net
with a 9-iron and a 6-iron (TaylorMade-adidas Golf, Carlsbad, CA, USA). Players were asked to
hit golf shots a distance they normally would on the course with both clubs. Golf shots were
Page 52 of 106

performed in blocks, with the starting condition randomized for each player. Each player was
given adequate time to complete their normal warm-up routine prior to collection of golf swings.

Golf shots were initiated using the player’s preferred address position with each foot fully
supported by a force plate (1200 Hz, Kistler, Amherst, NY, USA). The magnitude (normalized by
body weight) and direction of the peak resultant horizontal reaction forces at the force plate
interface and point of wrench application (PWA) were computed for each leg [2,13,14]. The center
of mass trajectory during the golf swing was calculated by estimating an initial CM position as the
PWA at time of address, and driving the CM forward using the net linear impulse during the swing
[15,16].
Moments about the vertical axis passing through the CM were calculated as the cross-product of
the RFh and position vector from the CM to the PWA (moment arm) summed with the free moment
for each leg [2,17]. The interval of interest was defined as the phase of transition through ball
contact when the net moment toward the target was positive. Angular impulse was calculated as
the area under the moment curves for each leg during the interval of interest and was normalized
by body mass [2]. Linear impulse was calculated as the area under the force-time curves for both
horizontal directions (parallel and perpendicular to the target line, Figure 6.1) during the interval
of interest and were normalized by body mass. Resultant horizontal reaction force and RFh-angle
were also expressed relative to the target line. Components of angular impulse (RFhs, RFh-
angles, and moment arms) were represented as the mean value of the 10 samples measured
before and after the maximum RFh.
Between condition differences across the group were determined using the Sign Test (α = 0.05).
Within-player differences between conditions were also tested to further understand player-
specific strategies [2,18,19]. The probability for each variable of any 9-iron trial being less than
any 6-iron trial was calculated within a player for each variable, where each player served as their
own control (R, open-source). Assuming local independence (i.e., no order effect for trials within
a condition), and that club conditions were independent (i.e., not directly tied to each other) for
each subject, p-values were calculated for each player using Cliff’s analog of the Wilcoxon-Mann-
Whitney test [20,21]. A modified, step-down Fisher-type method was then applied to control the
Figure 6.1: Top-down view of force plate setup and reference system for reaction force and angle orientation
in the transverse plane.
Page 53 of 106

familywise error rate (α = .05) over multiple comparisons where the level of significance becomes
α/k at each kth iteration [22–24].
Results

The net angular impulse significantly increased across the group 0.01 N*m*s/kg (Figure 6.2) with
the 6-iron compared to 9-iron (p = 0.022). This was not associated with increases in target (p =
0.088) or rear leg angular impulse (p = 0.088). Although group differences were revealed,
individual results did not discover differences within a player for any components of angular
impulse across clubs.
The net linear impulse across the group both perpendicular to (p = 0.088, Figure 6.3) and parallel
to the target line (p = 0.32, Figure 6.4) were not found to differ between the 9-iron and 6-iron.
Players remained near zero impulse perpendicular to the target, while impulse contributed to CM
velocity parallel toward the target.

Page 54 of 106

Figure 6.2: Mean angular impulse (normalized by body weight) of each players’ golf swings with the 9-iron (9I)
and 6-iron (6I) during the interval of interest. Net angular impulse (above) as well as rear and target leg angular
impulse (below) for all players are displayed. Net angular impulse increased with the 6-iron compared to the
9-iron across the group (p = 0.022). Error bars represent ±1 standard deviation. Net angular impulse is the
summation of rear and target contributions for each club. Positive values were plotted above and below zero
for rear and target leg angular impulse (below). Significant differences were tested at α = 0.05 level when
adjusted for multiple comparisons.
Page 55 of 106

Figure 6.3: Mean linear impulse (normalized by mass) perpendicular to the target line of each players’ golf
swings with the 9-iron (9I) and 6-iron (6I) during the interval of interest. Net linear impulse (above) as well as
rear and target leg linear impulse (below) for all players are displayed. Error bars represent ±1 standard
deviation. Net linear impulse is the summation of rear and target contributions for each club. No significant
differences were discovered when tested at α = 0.05 level and adjusted for multiple comparisons.
Page 56 of 106

Figure 6.4: Mean linear impulse (normalized by mass) parallel to the target line of each players’ golf swings
with the 9-iron (9I) and 6-iron (6I) during the interval of interest. Net linear impulse (above) as well as rear and
target leg linear impulse (below) for all players are displayed. Error bars represent ±1 standard deviation. Net
linear impulse is the summation of rear and target contributions for each club. All significant differences were
denoted when tested at α = 0.05 level when adjusted for multiple comparisons. * Net linear impulse significant
difference between clubs.
Page 57 of 106

Discussion

We expected increases in net angular impulse with the 6-iron as a result of large regulations in
target leg angular impulse and smaller increase of rear leg angular impulse. We found almost no
differences in RFh or RFh angle with either the rear or target leg, which is contrary to what we’ve
discovered previously [1–3]. Linear impulse remained near zero perpendicular to the target both
clubs, and had little change in linear impulse parallel to the target. The differences in full swings
between 9-iron and 6-iron are almost negligible. We expected to find slight regulation between
the 9-iron and 6-iron, but discovered that players perform full swings with the lower end of irons
in a similar way to simplify the control strategy. This provides some evidence to suggest that
players perform full swings with the irons and the driver in distinct manners.

Page 58 of 106

References

[1] McNitt-Gray, J. L., Munaretto, J., Zaferiou, A., Requejo, P. S., and Flashner, H., 2013,
“Regulation of reaction forces during the golf swing,” Sport. Biomech., 12(2), pp. 121–131.
[2] Peterson, T. J., Wilcox, R. R., and McNitt-Gray, J. L., 2016, “Angular impulse and balance
regulation during the golf swing,” J. Appl. Biomech., 32(4), pp. 342–349.
[3] Peterson, T. J., 2016, “Regulation of the Lower Extremities during the Golf Swing with
Different Clubs and Varying Address Positions,” University of Southern California.
[4] Egret, C. I., Vincent, O., Weber, J., Dujardin, F. H., and Chollet, D., 2003, “Analysis of 3D
kinematics concerning three different clubs in golf swing,” Int. J. Sports Med., 24(6), pp.
465–469.
[5] Healy, A., Moran, K. a, Dickson, J., Hurley, C., Smeaton, A. F., O’Connor, N. E., Kelly, P.,
Haahr, M., and Chockalingam, N., 2011, “Analysis of the 5 iron golf swing when hitting for
maximum distance.,” J. Sports Sci., 29(10), pp. 1079–88.
[6] Marta, S., Silva, L., Vaz, J. R., Castro, M. A., Reinaldo, G., and Pezarat-Correia, P., 2016,
“Electromyographic analysis of lower limb muscles during the golf swing performed with
three different clubs,” J. Sports Sci., 34(8), pp. 713–720.
[7] Abernethy, B., Neal, R., Moran, M., and Parker, A., 1990, “Expert-novice differences in
muscle activity during the golf swing,” Science and Golf: Proceedings of the World Scientifc
Congress of Golf, E.& F. Spon, ed., London, United Kingdom, pp. 54–60.
[8] Nagao, N., and Sawada, Y., 1973, “A kinematic analysis in golf swing concerning driver
shot and No. 9 iron shot,” J. Sports Med. Phys. Fitness, 13(1), pp. 4–16.
[9] Horton, J. F., Lindsay, D. M., and Macintosh, B. R., 2001, “Abdominal muscle activation of
elite male golfers with chronic low back pain.,” Med. Sci. Sports Exerc., 33(29), pp. 1647–
1654.
[10] Barrentine, S. W., Fleisig, G. S., and Johnson, H., 1994, “Ground Reaction Forces and
Torques of Professional and Amateur Golfers,” Sci. Golf II Proc. 1994 World Sci. Congr.
Golf, 1(1), pp. 33–39.
[11] Worsfold, P., Smith, N., Dyson, R., and Dyson, 2007, “A comparison of golf shoe designs
highlights greater ground reaction forces with shorter irons,” J. Sport. Sci. Med., (6), pp.
484–489.
[12] Farber, A. J., Smith, J. S., Kvitne, R. S., Mohr, K. J., and Shin, S. S., 2009,
“Electromyographic analysis of forearm muscles in professional and amateur golfers.,” Am.
J. Sports Med., 37(2), pp. 396–401.
[13] Shimba, T., 1996, “Consequences of Force Platform Studies,” Natl. Rehabiliation Res. Bull.
Japan, (17), pp. 17–23.
[14] Zatsiorsky, V. M., 2002, “Wrench Representation of the Ground Reaction Force,” Kinetics
of Human Motion, Human Kinetics Publishers, Champaign, IL, pp. 43–48.
[15] King, D. L., and Zatsiorsky, V. M., 1997, “Extracting gravity line displacement from
stabilographic recordings,” Gait Posture, 6(1), pp. 27–38.
[16] Lenzi, D., Cappello, A., and Chiari, L., 2003, “Influence of body segment parameters and
Page 59 of 106

modeling assumptions on the estimate of center of mass trajectory,” J. Biomech., 36(9),
pp. 1335–1341.
[17] Holden, J. P., and Cavanagh, P. R., 1991, “The free moment of ground reaction in distance
running and its changes with pronation.,” J. Biomech., 24(10), pp. 887–97.
[18] Russell, I. M., Raina, S., Requejo, P. S., Wilcox, R. R., Mulroy, S., and McNitt-Gray, J. L.,
2015, “Modifications in Wheelchair Propulsion Technique with Speed,” Front. Bioeng.
Biotechnol., 3(October), pp. 1–11.
[19] Zaferiou, A. M., Wilcox, R. R., and McNitt-Gray, J. L., 2016, “Modification of Impulse
Generation during Pirouette Turns with Increased Rotational Demands,” J. Appl. Biomech.,
pp. 1–44.
[20] Cliff, N., 1996, Ordinal Methods for Behavioral Data Analysis, Lawrence Erlbaum
Associates, Inc., Publishers, Mahwah, NJ.
[21] Neuhäuser, M., Lösch, C., and Jöckel, K.-H., 2007, “The Chen–Luo test in case of
heteroscedasticity,” Comput. Stat. Data Anal., 51(10), pp. 5055–5060.
[22] Hochberg, Y., 1988, “A sharper Bonferroni test for multiple tests of significance,”
Biometrika, 75, pp. 800–802.
[23] Hochberg, Y., and Tamhane, A. C., 1987, Multiple Comparison Procedures, John Wiley &
Sons, Inc., Hoboken, NJ, USA.
[24] Wilcox, R., and Clark, F., 2015, “Robust Multiple Comparisons Based on Combined
Probabilities From Independent Tests,” J. Data Sci., 13(1), pp. 1–11.

Page 60 of 106

: Regulation of Linear and Angular Impulse with
Modified Address Positions during the Golf Swing

Introduction
During the golf swing, players must coordinate the legs to regulate linear and angular impulse
generation to satisfy the whole-body mechanical objectives[1]. On the course, players also need
to modify their address position by changing their lower extremity configurations to accommodate
for varying terrain conditions (e.g. tee shots off level ground, shots played out of sand bunkers,
and shots completed on uneven terrain)[2]. With approximately 25% of golf shots occurring in
areas that are prone to uneven terrain, the ability to coordinate reaction force (RF) generation
relative to the center of mass (CM) in these situations plays a significant role in the success of
the golf player [3–5]. Determining how an individual player completes golf swings while modifying
address position by regulating RFs in relation to the CM may provide the opportunity to improve
performance as well as lower golf scores.
Previous research has found the net angular impulse to increase via increases in target leg
resultant horizontal reaction force (RFh) and direction, while the rear leg RFh is generally
maintained while increasing golf shot distance [6,7]. Highly skilled golf players were also found to
increase stance width when hitting with the driver as a possible method to increase moment arms
for the rear and target leg RFhs. Net linear impulse perpendicular to the target was found to be
maintained near zero, while net linear impulse parallel to the target was less toward the target as
players hit swings with a driver compared to a 6-iron. This coordination strategy implies a specific
role of each leg during the golf swing. The regulation of RFs as the task is modified has also been
seen in other rotational tasks such as dance turns [8,9] and turning while walking [10]. However,
these tasks are typically performed with a similar set of kinematic conditions that maintain similar
muscle lengths at which to create reaction forces (RFs) at the foot-surface interface.
Modifying the kinematic context within a task has been found to affect the generation of RFs (e.g.
fall recovery) [11]. During walking on uneven terrain, whole-body angular momentum and lower
extremity joint control were modified by forcing footfalls that were either inverted or everted [12].
Altering limb configurations has also been found to redistribute the mechanical load across the
lower extremity joints during the sit-to-stand task [13]. Similar changes in whole body and segment
kinetics with modified kinematic context have been discovered in turning while walking [14] and
amputee gait [15]. Based on these findings, we expect modifying the kinematic context of the legs
by raising either leg is expected to have an effect on how the legs regulate RFs to generate linear
and angular impulse at the whole body level.
This study aimed to determine how individual golf players coordinate the legs to regulate linear
and angular impulse as the lower extremity address position was modified to simulate conditions
that players face on the course. We hypothesized (1) net angular impulse will decrease when
swinging with a modified address position compared to the normal address position. We expected
to see increases in angular impulse from the leg in the normal address position as the angular
impulse from the leg in the modified position decreases. We also hypothesized (2) the regulation
of angular impulse will arise from regulation of the RFh in either leg. Finally, we hypothesized (3)
linear impulse will be comparable between the normal and modified address positions. We tested
these hypotheses by measuring the RFhs and their interaction with the CM while comparing how
Page 61 of 106

linear and angular impulse were regulated through changes in RFh, RFh-angle, and moment arm
length.
Methods
Skilled players (n = 9, handicap < 5) volunteered to participate in accordance with the local
institutional review board. Each player performed ten golf shots towards a target placed behind a
net with a 6-iron (TaylorMade-adidas Golf, Carlsbad, CA, USA) in the Normal (N), Target Leg Up
(Tup), and Rear Leg Up conditions (Rup). Players were asked to hit straight shots a distance they
normally would on the course. Golf shots were performed in blocks, with the starting condition
randomized for each player. Each player was given adequate time to complete their normal warm-
up routine prior to collection of golf swings.
Golf shots were initiated using the player’s preferred address position with each foot fully
supported by a force plate (1200 Hz, Kistler, Amherst, NY, USA). The magnitude (normalized by
body weight) and direction of the peak RFhs at the force plate interface, and point of wrench
application (PWA) were computed for each leg [1,16,17]. The center of mass (CM) trajectory
during the golf swing was calculated by estimating an initial CM position as the PWA at time of
address, and driving the CM forward using the net linear impulse during the swing [18,19]. During
trials where the address position was modified, a rigid wooden box (0.4 x 0.6 x 0.175 m) was
affixed directly to the force plate with bolts (Figure 7.1, below). Calculations for force, moment,
and PWA were adjusted to compensate for the height offset of the box.
Moments about the vertical axis passing through the CM were calculated as the cross-product of
the RFh and position vector from the CM to the PWA (moment arm) summed with the free moment
for each leg [1,20]. The interval of interest was defined as the phase of transition through ball
contact when the net moment toward the target was positive [1]. Angular impulse was calculated
as the area under the moment curves for each leg during the interval of interest and was
Figure 7.1: Top-down view of force plate setup and reference system for reaction force and angle orientation
in the transverse plane (above). Setup and dimensions of the box used to modify leg configurations as
attached via four bolts threaded into force plate 1 (below).
Page 62 of 106

normalized by body mass. Linear impulse was calculated as the area under the force-time curves
for both horizontal directions (parallel and perpendicular to the target line) during the interval of
interest and was normalized by body mass. Resultant horizontal reaction force and RFh-angle
were also expressed relative to the target line (Figure 7.1, above). Components of angular
impulse (RFhs, RFh-angles, and moment arms) were represented as the mean value of the 10
samples measured before and after the maximum RFh.
Multiple pairwise comparisons were made to detect changes between all conditions. Between
condition differences across the group were determined using the Sign Test (α = 0.05). Within-
player differences between conditions were also tested to further understand player-specific
strategies [1,8,21]. The probability for each variable of any trial of condition 1 being less than any
trial from condition 2 was calculated within a player for each variable, where each player served
as their own control (R, open-source). P-values were calculated for each player using Cliff’s
analog of the Wilcoxon-Mann-Whitney test [22,23]. A modified, step-down Fisher-type method
was then applied to control the familywise error rate (α = .05) over multiple comparisons where
the level of significance becomes α/k at each k
th
iteration [24–26].
Results
Individuals used player specific strategies to consistently respond to the modifications in address
position as they generated linear and angular impulse. The net angular impulse generated by
players was not significantly different between conditions across the group (Figure 7.2, Table 7.1).
Player specific differences in net angular impulse were discovered via the within-player statistical
design in some cases. Between modified leg conditions, target leg angular impulse increased
when hitting with the target leg up compared to the rear leg up across the group (p = 0.024, Figure
7.3 above, Table 7.2). Rear leg angular impulse was not significantly different between conditions
across the group, however individuals displayed player specific strategies (Figure 7.3 below,
Table 7.2).
Target leg angular impulse was regulated primarily through changes in RFh magnitude and
orientation. As compared to the rear leg up condition, target leg RFhs decreased in the normal (p
= 0.004) and target leg up conditions (p = 0.004) across the group. Coincidentally, target leg RFhs
were oriented more parallel away from the target in the rear leg up condition compared to normal
(p = 0.002) and target leg up conditions (p = 0.002). Rear leg angular impulse was generated with
RFhs oriented more parallel toward the target with the rear leg up (p = 0.004) and target leg up
(p = 0.004) as compared to the normal condition (Table 7.1). Across the group, stance width was
significantly greater than normal when modifying address positions with either the rear leg (p =
0.002) or target leg up (p = 0.002, Figure 7.4, Table 7.2). The duration of the interval of interest
was not significantly different between conditions across the group.
The net linear impulse perpendicular to the target was found to be near zero or directed slightly
posterior to the player and was significantly different between conditions (Figure 7.5 above).
Compared to normal, net linear impulse perpendicular to the target was significantly less posterior
with the rear leg up (p = 0.002). With the target leg up, net perpendicular linear impulse was
significantly more posterior compared to normal (p = 0.024). Between modified address positions,
net perpendicular linear impulse was significantly more posterior with the target leg up compared
to the rear leg up condition (p = 0.002). Players coordinated rear and target leg contributions to
maintain minimal linear impulse perpendicular to the target line (Figure 7.6).
Page 63 of 106

Net linear impulse parallel to the target line was toward the target, but was significantly less toward
the target in the target leg up condition compared to normal (p = 0.002) and rear leg up conditions
(p = 0.002, Figure 7.5 below). To decrease the amount of net linear impulse parallel toward the
target, players hitting with their target leg up either increased target leg impulse away from the
target, decreased rear leg linear impulse toward the target, or a combination of the two methods.
Compared to normal, target leg linear impulse away from the target significantly increased with
the rear leg up (p = 0.004) and target leg up (p = 0.002) across the group (Figure 7.7). Players
also displayed significant increases in target leg linear impulse away from the target with the target
leg up compared to rear leg up across the group (p = 0.004).

Figure 7.2: Net angular impulse (normalized by body mass) of each players’ golf swings in the Normal, Rear
leg up (Rup), and Target leg up (Tup) condition during the interval of interest. Net angular impulse is the
summation of rear and target contributions for swing. All significant differences were denoted when tested at
α = .05 level when adjusted for multiple comparisons. a Net angular impulse significant difference between
Normal and Rup, b net angular impulse significant difference between Normal and Tup, c net angular impulse
significant difference between Rup and Tup.
Page 64 of 106

Figure 7.3: Target (above) and Rear (below) leg angular impulse (normalized by body mass) of each players’
golf swings in the Normal, Rear leg up (Rup), and Target leg up (Tup) condition during the interval of interest.
All significant differences were denoted when tested at α = .05 level when adjusted for multiple comparisons.
a leg angular impulse significant difference between Normal and Rup, b leg angular impulse significant
difference between Normal and Tup, c leg angular impulse significant difference between Rup and Tup.
Page 65 of 106

Figure 7.4: Stance width at address of each players’ golf swings in the Normal, Rear leg up (Rup), and Target
leg up (Tup) condition. All significant differences were denoted when tested at α = .05 level when adjusted for
multiple comparisons. a Linear impulse significant difference between Normal and Rup, b linear impulse
significant difference between Normal and Tup, c linear impulse significant difference between Rup and Tup.
Page 66 of 106

Table 7.1: Comparison of group and individual net angular impulse and rear leg angular impulse components.

Net Angular
Impulse
(N∙m∙s∙kg
-1
)
Rear Angular
Impulse
(N∙m∙s∙kg
-1
)
Rear RFh
(%BW)
Rear Moment
Arm (m)
Rear RFh-
Angle (°)
Player Club Mean SD Mean SD Mean SD Mean SD Mean SD
1
Normal
0.224 0.005 0.134 0.008 0.232 0.004 0.175 0.020 139.2 2.0
Rup
0.219 0.007 0.120 0.007 0.189 0.011 0.202 0.015 144.4 7.2
Tup
0.218 0.005 0.127 0.006 0.236 0.008 0.109 0.009 152.1 3.2
2
Normal
0.205 0.010 0.124 0.008 0.179 0.011 0.110 0.030 129.4 9.2
Rup
0.205 0.012 0.114 0.009 0.185 0.009 0.181 0.089 138.4 20.4
Tup 0.201 0.009 0.132 0.007 0.202 0.016 0.038 0.029 173.5 15.2
3
Normal 0.279 0.004 0.159 0.005 0.226 0.004 0.138 0.011 143.2 1.9
Rup 0.274 0.006 0.177 0.007 0.188 0.017 0.369 0.049 127.1 7.7
Tup 0.250 0.010 0.128 0.013 0.178 0.016 0.133 0.024 140.6 5.1
4
Normal 0.241 0.005 0.096 0.008 0.239 0.007 0.132 0.014 139.3 3.8
Rup 0.235 0.007 0.131 0.006 0.224 0.011 0.259 0.025 140.3 3.6
Tup
0.235 0.009 0.098 0.008 0.207 0.009 0.107 0.017 150.7 7.1
5
Normal 0.290 0.005 0.118 0.008 0.237 0.008 0.198 0.028 118.6 10.4
Rup 0.265 0.007 0.095 0.010 0.241 0.016 0.172 0.034 152.0 6.9
Tup
0.291 0.006 0.115 0.006 0.217 0.012 0.135 0.043 142.3 14.6
6
Normal
0.272 0.013 0.124 0.018 0.237 0.009 0.133 0.024 143.5 9.0
Rup 0.233 0.005 0.113 0.009 0.213 0.018 0.121 0.068 155.7 15.4
Tup 0.231 0.007 0.106 0.011 0.225 0.009 0.107 0.016 146.3 3.9
7
Normal 0.271 0.006 0.130 0.010 0.210 0.007 0.181 0.051 134.7 13.5
Rup 0.244 0.008 0.108 0.007 0.168 0.018 0.195 0.153 150.4 33.2
Tup 0.267 0.009 0.115 0.010 0.199 0.013 0.086 0.031 175.5 16.1
8
Normal 0.317 0.006 0.130 0.007 0.185 0.005 0.247 0.028 109.3 1.4
Rup 0.329 0.016 0.116 0.008 0.257 0.031 0.037 0.033 176.9 5.6
Tup
0.337 0.007 0.124 0.008 0.177 0.005 0.196 0.049 129.9 12.3
9
Normal
0.308 0.009 0.158 0.021 0.260 0.010 0.226 0.036 136.6 5.5
Rup 0.228 0.013 0.088 0.019 0.244 0.027 0.139 0.080 150.3 15.6
Tup
0.254 0.021 0.108 0.023 0.267 0.021 0.107 0.050 157.7 11.1
Group
Normal
0.267 0.007 0.130 0.010 0.223 0.007 0.171 0.027 132.7 6.3
Rup
0.248 0.009 0.118 0.009 0.212 0.018 0.186 0.061 148.4 12.8
Tup 0.254 0.009 0.117 0.010 0.212 0.012 0.113 0.030 152.1 9.8
Group
p-
values
N v Rup 0.180 0.180 0.510 0.470 0.004*
N v Tup 0.180 0.180 0.510 0.002* 0.004*
Rup v Tup 0.470 0.150 0.470 0.004* 0.15

Individual player differences based on Cliff's Analog of Wilcoxon-Mann-Whitney Test and modified Fisher

type method step-down technique

p
N vs Rup Significant at α = .05 level when adjusted for multiple comparisons

p N vs Tup Significant at α = .05 level when adjusted for multiple comparisons

p Rup vs Tup Significant at α = .05 level when adjusted for multiple comparisons

p*
Group differences based on two-sample Student's T-test at α = .05
level

RFh is the resultant horizontal reaction force

RFh-angle is the angle of the RFh relative to the target line

N is Normal Condition

Rup is Rear Leg Up Condition

Tup is Target Leg Up Condition

Page 67 of 106

Table 7.2: Comparison of group and individual net angular impulse and target leg angular impulse components.

.

Stance Width
(m)
Target Angular
Impulse
(N∙m∙s∙kg
-1
)
Target RFh
(%BW)
Target
Moment Arm
(m)
Target RFh-
Angle (°)
Player Club Mean SD Mean SD Mean SD Mean SD Mean SD
1
Normal 0.388 0.007 0.089 0.007 0.229 0.009 0.200 0.021 299.00 2.72
Rup 0.512 0.018 0.100 0.005 0.186 0.008 0.242 0.072 284.28 21.72
Tup 0.443 0.012 0.090 0.007 0.237 0.009 0.012 0.072 358.91 1.83
2
Normal 0.465 0.012 0.081 0.013 0.224 0.026 0.257 0.056 295.23 16.44
Rup 0.572 0.040 0.092 0.009 0.205 0.019 0.250 0.020 270.82 4.96
Tup 0.626 0.020 0.070 0.011 0.240 0.022 0.105 0.020 344.48 23.28
3
Normal 0.478 0.009 0.121 0.005 0.279 0.014 0.246 0.015 271.80 3.36
Rup 0.708 0.030 0.097 0.007 0.217 0.014 0.239 0.011 266.99 3.09
Tup 0.530 0.027 0.122 0.013 0.220 0.014 0.114 0.011 338.49 22.66
4
Normal 0.507 0.007 0.145 0.006 0.259 0.008 0.182 0.033 316.77 8.63
Rup 0.647 0.016 0.104 0.008 0.210 0.014 0.133 0.038 317.19 14.61
Tup 0.685 0.013 0.138 0.010 0.342 0.020 0.085 0.038 346.22 8.61
5
Normal 0.482 0.012 0.172 0.005 0.356 0.011 0.276 0.012 288.28 2.36
Rup 0.676 0.017 0.170 0.006 0.255 0.011 0.295 0.011 265.80 3.62
Tup 0.549 0.009 0.175 0.008 0.304 0.011 0.306 0.011 300.05 4.53
6
Normal 0.434 0.008 0.148 0.010 0.334 0.009 0.223 0.018 304.47 2.25
Rup 0.540 0.017 0.120 0.007 0.259 0.018 0.219 0.020 303.25 9.00
Tup 0.541 0.013 0.124 0.008 0.246 0.014 0.259 0.020 312.85 3.88
7
Normal 0.503 0.014 0.141 0.007 0.254 0.015 0.222 0.022 296.64 7.49
Rup 0.636 0.028 0.136 0.006 0.169 0.031 0.277 0.026 274.32 7.19
Tup 0.601 0.022 0.152 0.013 0.236 0.021 0.260 0.026 314.76 6.20
8
Normal 0.482 0.012 0.187 0.010 0.335 0.009 0.255 0.018 286.01 2.02
Rup 0.741 0.042 0.213 0.022 0.281 0.014 0.263 0.026 292.45 14.77
Tup 0.688 0.045 0.213 0.012 0.326 0.035 0.054 0.026 348.30 4.39
9
Normal 0.512 0.005 0.150 0.019 0.326 0.009 0.239 0.042 305.98 2.06
Rup 0.664 0.023 0.140 0.018 0.230 0.029 0.246 0.050 302.07 20.42
Tup 0.645 0.034 0.146 0.021 0.323 0.085 0.236 0.050 324.62 13.36
Group
Normal 0.472 0.009 0.137 0.009 0.288 0.012 0.233 0.026 296.02 5.26
Rup 0.633 0.026 0.130 0.010 0.224 0.017 0.240 0.030 286.35 11.04
Tup 0.590 0.022 0.137 0.012 0.275 0.026 0.159 0.030 332.08 9.86
Group
p-
values
N v Rup 0.002* 0.510 0.002* 0.470 0.180
N v Tup 0.002* 0.470 0.510 0.510 0.002*
Rup v Tup 0.510 0.024* 0.004* 0.180 0.002*

Individual player differences based on Cliff's Analog of Wilcoxon-Mann-Whitney Test and modified Fisher

type method step-down technique

p
N vs Rup Significant at α = .05 level when adjusted for multiple comparisons

p N vs Tup Significant at α = .05 level when adjusted for multiple comparisons

p
Rup vs Tup Significant at α = .05 level when adjusted for multiple
comparisons

p*
Group differences based on two-sample Student's T-test at α = .05
level

RFh is the resultant horizontal reaction force

RFh-angle is the angle of the RFh relative to the target line

N is Normal Condition

Rup is Rear Leg Up Condition

Tup is Target Leg Up Condition

Page 68 of 106

Figure 7.5: Net linear impulse (normalized by body mass) perpendicular to the target (above) and parallel to
the target (below) of each players’ golf swings in the Normal, Rear leg up (Rup), and Target leg up (Tup)
condition during the interval of interest. Net linear impulse is the summation of rear and target contributions
for each swing. All significant differences were denoted when tested at α = .05 level when adjusted for multiple
comparisons. a Net linear impulse significant difference between Normal and Rup, b net linear impulse
significant difference between Normal and Tup, c net linear impulse significant difference between Rup and
Tup.
Page 69 of 106

Figure 7.6: Net linear impulse perpendicular to the target (normalized by body mass) from the target (above)
and rear (below) legs of each players’ golf swings in the Normal, Rear leg up (Rup), and Target leg up (Tup)
condition during the interval of interest. All significant differences were denoted when tested at α = .05 level
when adjusted for multiple comparisons. a Linear impulse significant difference between Normal and Rup,
b linear impulse significant difference between Normal and Tup, c linear impulse significant difference between
Rup and Tup.
Page 70 of 106

Figure 7.7: Net linear impulse parallel to the target (normalized by body mass) from the target (above) and
rear (below) legs of each players’ golf swings in the Normal, Rear leg up (Rup), and Target leg up (Tup)
condition during the interval of interest. All significant differences were denoted when tested at α = .05 level
when adjusted for multiple comparisons. a Linear impulse significant difference between Normal and Rup,
b linear impulse significant difference between Normal and Tup, c linear impulse significant difference between
Rup and Tup.
Page 71 of 106

Discussion
This study sought to understand how individuals coordinated linear and angular impulse
regulation during the golf swing under modified address positions. The highly skilled players in
this study coordinated angular impulse generation between the rear and target legs to generate
similar amounts of net angular impulse with the 6-iron on flat and uneven address positions.
Individuals used a variety of methods (e.g. increasing stance width, redirecting RFhs) to regulate
player specific changes in leg angular impulse. Net linear impulse was found to differ across tasks,
suggesting that players may have found it more difficult to handle the modifications in address
position in these measures. Understanding how players adjust their strategies to generate linear
and angular impulse under modified address positions is important to consider when preparing
players to manage golf shots on uneven terrain that they are faced with on the course.
The results of this study are specific to highly skilled golf players with low handicaps that
consistently generate RFs under different swing conditions [1,6,27]. Standardization of the club
and surface conditions were comparable to when players hit when on a driving range and when
hitting from uneven lies on the golf course. The height of the modified address positions elicited
large effect sizes and were reported by the participants to be comparable to what they typically
encounter on challenging courses. The kinetics-based CM trajectory utilized in this study was
improved resolution of the angular impulse computed, and minimized propagation of error from
segment kinematic based estimates of body segment parameters [19,28]. The statistical analysis
allowed for a within-player experimental design along with group differences and provided insights
into how individuals regulate whole-body mechanics when they swung a 6-iron under modified
address conditions.
The net angular impulse of players in this study did not differ from normal when hitting golf shots
from a modified address position. However, individuals displayed player specific responses to the
challenge induced by hitting off uneven terrain. Despite changes in lower extremity configuration,
individuals successfully coordinated rear and target leg angular impulse to generate the net
angular impulse on uneven surfaces. Modifying the height of the legs relative to the ball is a
challenge that golf players regularly face on the course, and this group of highly skilled players
was able to negotiate angular impulse levels that were similar to normal, flat ground conditions
hitting with the 6-iron [1]. When players regulated net angular impulse of across modified address
positions, the differences were on the order of angular impulse regulation differences between
the 6-iron and driver [1]. With the intent of hitting the 6-iron the same distance, individual players
were able to generate similar amounts of net angular impulse at the whole-body level regardless
of initial address position.
Individuals used a variety of methods to generate the rear and target leg angular impulse for each
condition. For example, when hitting with the rear leg up compared to normal, player 3 increased
their rear leg angular impulse through a larger rear leg moment arm that compensated for a
smaller rear leg RFh. This strategy may allow players to maintain or increase levels of rear leg
angular impulse when the rear leg RFh generation is altered as the rear leg is elevated. At the
same time, player 3 decreased their target leg angular impulse through decreases in peak RFh.
Coordinating increases in rear leg angular impulse with corresponding decreases in target leg
angular impulse assists players in generating a similar amount of net angular impulse with a
modified address position as compared to normal.
While hitting with modified address positions, players increased stance width and redirected RFhs
to be more parallel to the target line. Other studies have found players to increase stance width
Page 72 of 106

when hitting with the driver to a similar width as when modifying address position, and noted more
parallel RFhs in the target leg [1]. Increasing stance width may be a mechanism to help maintain
balance when raising one leg and an opportunity to increase the length of moment arms when
players were unable to redirect RFhs. The variety of combinations of factors players chose to
regulate net linear and angular impulse highlights the need for understanding individual player
strategies to successfully complete the golf swing under modified address positions.
Interestingly, the net linear impulse was different between conditions. Similar to the net angular
impulse, players coordinated rear and target leg linear impulse contributions to generate the net
linear impulse both perpendicular and parallel to the target. Although there were significant
differences between conditions, the amount of linear impulse in either direction for all conditions
was similar to regulations in linear impulse generation when increasing shot distance between
the 6-iron and driver [1].
Net linear impulse perpendicular to the target was generally maintained near zero or slightly
anterior to the player. The most anterior net linear impulse occurred when swinging with the target
leg up. This was likely due to target leg RFhs being more parallel with the target leg up, which did
not allow the target leg to provide the posteriorly oriented RFhs to reduce the amount of anterior
net linear impulse. Players likely needed to account for these changes in anterior net linear
impulse when hitting with the target leg up in order to effectively complete the swing.
Net linear impulse parallel to the target was directed toward the target in all golf shot conditions.
With the rear leg up, players had more net linear impulse toward the target than with the target
leg up. These findings are consistent with larger target leg RFhs pointed away from the target in
the target leg up condition. Having the target leg elevated likely limited the ability of players to
transfer support of body weight to the target leg. This potential limitation could present a challenge
for players who make dramatic shifts in support from the rear to target leg.
Understanding how players regulate impulse when hitting with modified address positions will
benefit individual players by providing information to personalize interventions that may modify
technique and training. Elevating the rear or target legs contributed to different approaches from
individuals, suggesting that training under each condition would be beneficial toward discovering
effective solutions to manage net linear and angular impulse. Modifying address positions by
elevating either leg is just a subset of the many address positions players may encounter during
play, and it remains to be seen how other address positions (e.g. anterior or posterior slopes)
may elicit changes in whole-body linear and angular impulse generation.
Modifying the address position not only changes the current leg orientation, but also has
implications toward muscle activity and joint loading. When changing leg height or widening the
stance at address, the muscle lengths will change which could alter the ability of the muscles to
produce similar forces. Other studies have discovered changes in elite golf player muscle
activation and RFh magnitude and orientation while modifying address position by increasing
stance width when hitting with a driver [29]. The results of this study suggest that muscle activation
may also be regulated as players redirected RFhs with both legs while hitting under modified leg
height conditions. Due to the change in limb orientation, the mechanical demand on the leg joints
will also be different [30]. Under these more flexed leg conditions, players may be subjected to
larger net joint moments. The increased joint loading may not be a problem at the time, but
eventually could become an issue through repeated excessive joint loading. Players may need to
change their swing strategy to stave off potential future issues. All these things need to be factored
Page 73 of 106

in along with the individual’s strengths and weaknesses to best equip the player to perform under
these altered address conditions.
The results of this study indicate net linear and angular impulse were coordinated between the
legs when modifying address position by elevating the rear or target leg. Net angular impulse did
not change across conditions, which may indicate that players match their net angular impulse
generation to each club and distance, regardless of address position. Although the group did not
only displayed changes target leg angular impulse, individuals increased stance width and
redirected RFhs to be more parallel to the target line while generating rear and target leg angular
impulse under modified address conditions. Changing the address position did have an effect on
net linear impulse generation, where raising the target leg increased the amount of posterior linear
impulse while decreasing the amount of linear impulse toward the target. Understanding how
players generate linear and angular impulse while hitting a 6-iron under modified address
positions is important to consider when developing interventions to train players to successfully
complete the swing under these conditions. These findings also provide a foundation to
investigate the muscular demands and coordination of multi-joint control under the variety of
uneven terrain conditions encountered on the course.

Page 74 of 106

References
[1] Peterson, T. J., Wilcox, R. R., and McNitt-Gray, J. L., 2016, “Angular impulse and balance
regulation during the golf swing,” J. Appl. Biomech., 32(4), pp. 342–349.
[2] Langdown, B. L., Bridge, M. W., and Li, F., 2013, “Address Position Variability in Golfers
of Differing Skill Level,” Int. J. Golf Sci., 2, pp. 142–151.
[3] Wiseman, F., and Chatterjee, S., 2006, “Comprehensive Analysis of Golf Performance on
the PGA Tour: 1990-2004,” Percept. Mot. Skills, (102), pp. 109–117.
[4] Finley, P. S., and Halsey, J. J., 2004, “Determinants of PGA Tour Success: An
Examination of Relationships Among Performance, Scoring, and Earnings,” Percept. Mot.
Skills, (98), pp. 1100–1106.
[5] Dorsel, T. N., and Rotunda, R. J., 2001, “Low Scores, Top 10 Finishes, and Big Money:
An Analysis of Professional Golf Association Tour Statistics and How these Relate to
Overall Performance,” Percept. Mot. Skills, (92), pp. 575–585.
[6] McNitt-Gray, J. L., Munaretto, J., Zaferiou, A., Requejo, P. S., and Flashner, H., 2013,
“Regulation of reaction forces during the golf swing,” Sport. Biomech., 12(2), pp. 121–131.
[7] Peterson, T. J., 2016, “Regulation of the Lower Extremities during the Golf Swing with
Different Clubs and Varying Address Positions,” University of Southern California.
[8] Zaferiou, A. M., Wilcox, R. R., and McNitt-Gray, J. L., 2016, “Modification of Impulse
Generation during Pirouette Turns with Increased Rotational Demands,” J. Appl. Biomech.,
pp. 1–44.
[9] Zaferiou, A. M., Wilcox, R. R., and McNitt-Gray, J. L., 2016, “Modification of impulse
generation during piqué turns with increased rotational demands,” Hum. Mov. Sci., 47, pp.
220–230.
[10] Dixon, P. C., Stebbins, J., Theologis, T., and Zavatsky, A. B., 2014, “Ground reaction
forces and lower-limb joint kinetics of turning gait in typically developing children.,” J.
Biomech., pp. 1–8.
[11] Mathiyakom, W., and McNitt-Gray, J. L., 2008, “Regulation of angular impulse during fall
recovery,” J Rehabil Res Dev, 45(8), pp. 1237–1247.
[12] Yeates, K. H., Segal, A. D., Neptune, R. R., and Klute, G. K., 2016, “Balance and Recovery
on Coronally-Uneven and Unpredictable Terrain,” J. Biomech., pp. 1–7.
[13] Mathiyakom, W., McNitt-Gray, J. L., Requejo, P. S., and Costa, K., 2005, “Modifying center
of mass trajectory during sit-to-stand tasks redistributes the mechanical demand across
the lower extremity joints.,” Clin. Biomech. (Bristol, Avon), 20(1), pp. 105–11.
[14] Orendurff, M. S., Segal, A. D., Berge, J. S., Flick, K. C., Spanier, D., and Klute, G. K., 2006,
“The kinematics and kinetics of turning: limb asymmetries associated with walking a circular
path.,” Gait Posture, 23(1), pp. 106–11.
[15] Segal, A. D., Orendurff, M. S., Czerniecki, J., Schoen, J., and Klute, G. K., 2011,
“Comparison of transtibial amputee and non-amputee biomechanics during a common
turning task.,” Gait Posture, 33(1), pp. 41–47.
[16] Shimba, T., 1996, “Consequences of Force Platform Studies,” Natl. Rehabiliation Res. Bull.
Japan, (17), pp. 17–23.
Page 75 of 106

[17] Zatsiorsky, V. M., 2002, “Wrench Representation of the Ground Reaction Force,” Kinetics
of Human Motion, Human Kinetics Publishers, Champaign, IL, pp. 43–48.
[18] King, D. L., and Zatsiorsky, V. M., 1997, “Extracting gravity line displacement from
stabilographic recordings,” Gait Posture, 6(1), pp. 27–38.
[19] Lenzi, D., Cappello, A., and Chiari, L., 2003, “Influence of body segment parameters and
modeling assumptions on the estimate of center of mass trajectory,” J. Biomech., 36(9),
pp. 1335–1341.
[20] Holden, J. P., and Cavanagh, P. R., 1991, “The free moment of ground reaction in distance
running and its changes with pronation.,” J. Biomech., 24(10), pp. 887–97.
[21] Russell, I. M., Raina, S., Requejo, P. S., Wilcox, R. R., Mulroy, S., and McNitt-Gray, J. L.,
2015, “Modifications in Wheelchair Propulsion Technique with Speed,” Front. Bioeng.
Biotechnol., 3(October), pp. 1–11.
[22] Cliff, N., 1996, Ordinal Methods for Behavioral Data Analysis, Lawrence Erlbaum
Associates, Inc., Publishers, Mahwah, NJ.
[23] Neuhäuser, M., Lösch, C., and Jöckel, K.-H., 2007, “The Chen–Luo test in case of
heteroscedasticity,” Comput. Stat. Data Anal., 51(10), pp. 5055–5060.
[24] Hochberg, Y., 1988, “A sharper Bonferroni test for multiple tests of significance,”
Biometrika, 75, pp. 800–802.
[25] Hochberg, Y., and Tamhane, A. C., 1987, Multiple Comparison Procedures, John Wiley &
Sons, Inc., Hoboken, NJ, USA.
[26] Wilcox, R., and Clark, F., 2015, “Robust Multiple Comparisons Based on Combined
Probabilities From Independent Tests,” J. Data Sci., 13(1), pp. 1–11.
[27] Barrentine, S. W., Fleisig, G. S., and Johnson, H., 1994, “Ground Reaction Forces and
Torques of Professional and Amateur Golfers,” Sci. Golf II Proc. 1994 World Sci. Congr.
Golf, 1(1), pp. 33–39.
[28] Zatsiorsky, V. M., and King, D. L., 1997, “An algorithm for determining gravity line location
from posturographic recordings,” J. Biomech., 31(2), pp. 161–164.
[29] Peterson, T. J., 2017, “Chapter 9: Regulation of Muscular Control while Regulating Golf
Shot Distance Between Clubs,” Lower Extremity Control and Dynamics during the Golf
Swing (Dissertation).
[30] Peterson, T. J., 2017, “Chapter 8: Lower Extremity Joint Control during the Golf Swing,”
Lower Extremity Control and Dynamics during the Golf Swing (Dissertation).

Page 76 of 106

: Coordination of lower extremity multi-joint control
strategies during the golf swing

Introduction
During the golf swing, players need to coordinate multi-joint control of the lower extremities to
satisfy the mechanical objectives at the whole-body level. At the whole-body level, players must
regulate reaction forces (RFs) of the rear and target legs in relation to the center of mass (CM) to
create rotation and limit translation to satisfy the net linear and angular impulse requirements of
the task. At the limb level, the RFs generated in relation to lower extremity segments must also
be effectively coordinated to facilitate multi-joint control of both legs. Determining how individuals
satisfy the mechanical objectives of a task under different conditions has been effective in
understanding subject-specific coordination strategies in other well-practiced, goal directed tasks
such as dance turns and diving [1–3]. Understanding how an individual coordinates RF generation
between legs as well as multi-joint control of each leg, provides players and coaches a
mechanistic foundation to design interventions that may facilitate improvements in player
performance [4].
When skilled players increase golf shot distance, the net angular impulse generated during the
swing has been found to increase with the driver compared to the 6-iron by coordinating
contributions from the rear and target legs [5]. Individuals used player-specific strategies, yet
observed increases in net angular impulse with the driver were associated with target leg
contributions and were attributed to increases in target leg resultant horizontal RF. When
regulating shot distance within a 6-iron, players were also found to increase the RFs from the
target and/or rear legs [6]. These reported increases in RFs and net angular impulse generation
associated with increases in golf shot distance suggest that the mechanical demand imposed on
the ankle, knee, and hip of the target and rear legs will also likely increase with increases in shot
distance [5,6].
The regulation of RFs in relation to lower extremity segments is known to affect multi-joint control
of the leg [7]. The ankle, knee, and hip net joint moments (NJMs) controlling the limb are known
to be sensitive to both the magnitude and orientation of the RF relative to the lower segments and
the NJMs at the adjacent joints [1,7–11]. Modifying the limb configurations in relation to the RF
has also been associated with the redistribution of the mechanical load imposed on the lower
extremity joints [11]. We expect that increases in RF when swinging with a driver as compared to
a 6-iron will contribute to increases in ankle, knee, and hip NJMs of the target and rear legs [5].
Determining how players coordinate the multi-joint control of the rear and target legs is important
to understanding how individuals satisfy whole-body mechanical objectives when regulating golf
shot distance between clubs. Previously, studies on joint kinetics during the golf swing have
focused on control of the wrists, back, and lower extremities [12–18]. Greater external rotator
moments were reported in the rear leg versus the target leg in middle handicap players swinging
a driver [15]. Changing the orientation of the target leg was discovered to contribute to reported
differences in target leg, frontal plane knee moments for highly skilled players swinging a 5-iron
[16]. These studies highlight how reorientation of the leg relative to the RF could also affect multi-
joint control of the limb [7].
Page 77 of 106

Understanding how individual players orient their legs when regulating golf shot distance within
and between clubs can provide insight as to player-specific multi-joint control preferences [6,16].
Alignment of the RF with the leg plane acts to distribute the ankle, knee, and hip NJMs controlling
the limb about an axis perpendicular to the leg plane. Aligning RFs in this manner may also
simplify the multi-joint control of each leg by primarily taxing larger muscle groups [1]. This also
may allow players to reposition the legs (e.g. changes in address position due to stance width,
hip rotation, terrain, etc.) in global space while maintaining a similar multi-joint control strategy
across clubs. Similar analyses of forces and moments in relation to a plane have previously been
performed in upper extremity wheelchair propulsion and lower extremity dance turns to determine
multi-joint control of the entire limb [1,19].
This study investigated how skilled golf players control the rear and target legs when using a 6-
iron and driver on flat terrain. We hypothesized that (1) rear and target leg 3D support moments
would increase with the driver as compared to the 6-iron during the downswing at a time when
whole body moments about the CM were being created. To achieve increases in leg 3D support
moment, we hypothesized that (2) increases in 3D support moment would arise due to increases
in NJMs at the ankle, knee, and hip while maintaining relative contributions to the 3D support
moment. We expected that players would attempt to simplify their control strategies to take
advantage of the larger muscle groups during swings with greater 3D support moments and
hypothesized that (3) increases in 3D support moment would involve increases in ankle, knee,
and hip NJM primarily about an axis perpendicular to the leg plane. We tested these hypotheses
comparing the lower extremity NJMs in the rear and target legs for individual players during
swings with a 6-iron and a driver.
Methods
Highly-skilled golf players (n = 10, 5 female, 5 male, handicap < 5 [20]) volunteered to participate
in accordance with the local institutional review board. Players performed 10 full swings each with
a 6-iron and a driver (TaylorMade-adidas Golf, Carlsbad, CA, USA) toward a target downrange
as they normally would during the course of play. The golf swings were performed in blocks,
starting with the 6-iron.
Reaction forces at the foot-surface interface were measured as players initiated golf swings from
their preferred address positions with each foot supported by a force plate (1200 Hz, Kistler,
Amherst, NY). The point of wrench application (PWA) was computed for each leg [5,21,22].
Segment kinematics were collected with a 16-camera motion capture system (100 Hz, Natural
Point OptiTrack, Corvallis, OR) using Acquire3D software (C-Motion, Germantown, MD) and a
custom, 97 retroreflective marker set. Kinematic data were filtered and any gaps in the data were
filled using a cubic spline smoothing function where the degree of smoothing was determined by
Jackson’s method [23]. The kinematic data were simultaneously interpolated to synchronize with
kinetic data during the filtering process.
Segment reference systems and body segment parameters were defined according to research
by de Leva [24]. Functional joint centers of the ankle, knee, and hip were defined using functional
movements [25]. Inverse dynamics were calculated to determine net joint forces and NJMs using
custom Matlab software (The Mathworks, Natick, MA). Joint moments were expressed about each
segment’s anatomical axes [24].
Page 78 of 106

To determine the amount of control each leg had to regulate, the 3D support moment was
calculated as the sum of squares of the resultant NJMs (Figure 8.1 A). The 3D support moment
is a three dimensional analog to the two dimensional support moment proposed by Winter [26].
The cross product of vectors representing the thigh (hip to knee) and shank (knee to ankle) was
computed to create a leg plane (Figure 8.1 B) , defined by its perpendicular axis [1,19]. This axis
was used to determine what percentage of the 3D support moment and NJMs acted about an
axis perpendicular to the leg plane as a measure of control. NJMs and 3D support moments were
represented as RMS values of 20ms bins. Calculated variables were averaged over the interval
of interest which is defined as the 0.1s period surrounding the peak target leg moment about the
CM in the transverse plane [5].
Between club difference in NJMs and 3D support moments were compared across the group and
within a player (R, open-source). Group differences were determined using the Sign test (α =
0.05). To determine within player differences between clubs, each player served as their own
control by calculating the probability of any 6-iron trial being less than any driver trial for each
variable. P-values were calculated for each player using Cliff’s analog of the Wilcoxon-Mann-
Whitney test [27,28]. A modified, step-down Fisher-type method was then applied to control the
familywise error rate (α = 0.05) over multiple comparisons as the level of significance is calculated
at the k
th
iteration [29–31].

Figure 8.1: A) Resultant net joint moments (NJMs) at the ankle, knee, and hip for the rear (left) and target
(right) legs represented as circles at each joint (circles get larger with increasing magnitude) during the
interval of interest. The 3D support moment is a measure of control for the leg, and is defined as the sum
of squares of the ankle, knee, and hip NJMs of each leg. B) The leg plane was created as the cross product
of vectors passing through the ankle, knee, and hip joint centers as a way to further understand the multi-
joint control of the individual. The leg plane is defined by the vector perpendicular to it and travels with the
leg as the leg is reoriented in space.
A B
Page 79 of 106

Results
The target leg 3D support moment significantly increased when swinging with a driver compared
with a 6-iron as a group (p = 0.001) and individually for nine of ten players (Figure 8.2, Table 8.1).
Increases in target leg 3D support moment with the driver were achieved via increases in target
leg ankle (p = 0.013), knee (p = 0.001), and hip (p = 0.002) NJMs across the group (Figure 8.3,
Table 8.1). Within-player analysis of significant changes in target leg NJM system control revealed
three players regulated control at one joint, three players regulated control at two joints, while
three players regulated control at all three joints.
The relative contributions of the target leg ankle, knee, and hip NJMs to the 3D support moment
were not significantly different between swings with the 6-iron and driver across the group. Within-
player analysis of significant changes in target leg relative contributions to 3D support moment
revealed four players regulated control at one joint, one player regulated control at two joints,
while another player regulated control at all three joints.
Overall target leg 3D support moment about an axis perpendicular to the leg plane increased with
the driver across the group (p = 0.001), and individually for nine of ten players (Figure 8.4, Table
8.1). Increases in target leg 3D support moment about an axis perpendicular to the leg plane with
the driver resulted from significant increases in ankle (p = 0.002), knee (p = 0.001), and hip (p =
0.002) NJM about the perpendicular axis across the group. Within-player analysis of significant
changes in target leg NJM about an axis perpendicular to the leg plane revealed three players
regulated control at one joint, three players regulated control at two joints, while three players
regulated control at all three joints.

Figure 8.2: 3D support moment for one trial of a 6-iron overlaid on top of one trial of a driver for an exemplar player in
both the target (left) and rear (right) legs. Notice that a rise in 2D moment about the center of mass (CM) is accompanied
by a rise in 3D support moment for both legs and clubs. Across the group, players increased the amount of target leg
3D support moment (p = 0.001) when hitting golf shots with the driver compared to the 6-iron. Vertical lines represent
the interval of interest.
Page 80 of 106

The rear leg 3D support moment was not significantly different when comparing swings of the 6-
iron with the driver across the group (p = 0.35, Figure 8.2, Table 8.2). Individually, six of ten
players significantly decreased rear leg 3D support moment with the driver as compared to the 6-
iron (Table 8.2). Rear leg knee NJM magnitude significantly decreased with the driver across the
group (p = 0.001). Within-player analysis of significant changes in rear leg NJM system control
revealed three players regulated control at one joint, three players regulated control at two joints,
and two players regulated control at all three joints.
When swinging with the driver, the relative contribution to the 3D support moment of the rear leg
ankle NJM increased (p = 0.002) while relative contribution of the knee NJM decreased across
the group (p = 0.001, Table 2). Within-player analysis of significant changes in rear leg relative
contributions to 3D support moment revealed two players regulated control at one joint, seven
players regulated control at two joints, while one player regulated control at all three joints.
Rear leg NJM about an axis perpendicular to the leg plane was not significantly different between
clubs as a group (p = 0. 35, Table 2), but did significantly decrease with the driver for six of ten
players. Rear leg knee NJM about an axis perpendicular to the leg plane significantly decreased
with the driver across the group (p = 0.001). Within-player analysis of significant changes in rear
leg NJM control about an axis perpendicular to the leg plane revealed three players regulated
control at one joint, four players regulated control at two joints, and three players regulated control
at all three joints.

Figure 8.3: Ankle, knee, and hip net joint moment (NJM) contributions to the 3D support moment for one trial of a 6-
iron for an exemplar player in both the target (left) and rear (right) legs. As a group, players increased the magnitude
of target leg ankle (p = 0.001), knee (p = 0.001), and hip (p = 0.001) NJMs while decreasing the rear leg knee NJM (p
= 0.001) when swinging with the driver compared to the 6-iron. Target leg relative contributions remained similar across
clubs while rear leg ankle NJM relative contribution to 3D support moment increased (p = 0.002) and rear leg knee
NJM relative contribution decreased with the driver (p = 0.001). In general, target leg knee NJM and rear leg hip NJM
comprised the majority of the support moment for their respective legs during the interval of interest for both clubs.
Vertical lines represent the interval of interest.
Page 81 of 106

Figure 8.4: 3D support moment for one trial of a 6-iron (left) and driver (right) for an exemplar player for the target leg.
A majority of the 3D support moment is comprised of NJMs about an axis perpendicular to the target leg plane with
both clubs. As a group, players increased 3D support moment about an axis perpendicular to the leg plane in the target
leg (p = 0.001). Vertical lines represent the interval of interest.
Page 82 of 106

Table 8.1: Comparison of group and individual contributions to target leg 3D Support Moment.
3D Support Moment Ankle Knee Hip
Player Club Total
About
perp
axis
Not
about
perp
axis
NJM
About
perp
axis
Not
about
perp
axis
NJM
About
perp
axis
Not
about
perp
axis
NJM
About
perp
axis
Not
about
perp
axis
1
6-iron 2.340 1.916 0.423 0.235 0.176 0.059 0.924 0.839 0.085 1.181 0.902 0.279
Driver 2.738 2.377 0.361 0.258 0.186 0.072 1.045 0.974 0.071 1.435 1.217 0.218
2
6-iron 2.487 2.173 0.314 0.303 0.242 0.062 1.428 1.320 0.108 0.755 0.611 0.144
Driver 3.388 3.034 0.354 0.535 0.445 0.090 1.760 1.642 0.117 1.093 0.946 0.147
3
6-iron 1.613 1.200 0.413 0.352 0.346 0.006 0.799 0.698 0.101 0.462 0.157 0.305
Driver 2.310 1.715 0.595 0.332 0.321 0.011 1.223 1.080 0.143 0.755 0.314 0.441
4
6-iron 4.605 3.906 0.699 0.565 0.243 0.322 2.962 2.760 0.202 1.078 0.902 0.175
Driver 5.574 4.699 0.875 0.509 0.258 0.251 3.865 3.560 0.305 1.200 0.881 0.318
5
6-iron 3.824 2.990 0.834 0.727 0.681 0.046 1.986 1.728 0.258 1.111 0.581 0.530
Driver 5.594 4.656 0.938 0.914 0.828 0.086 3.205 2.887 0.317 1.475 0.940 0.535
6
6-iron 4.688 3.691 0.997 0.881 0.671 0.211 2.019 1.750 0.269 1.788 1.271 0.517
Driver 6.470 5.422 1.048 1.115 0.903 0.212 2.976 2.660 0.316 2.379 1.859 0.520
7
6-iron 3.876 3.616 0.261 0.520 0.447 0.074 2.269 2.173 0.096 1.087 0.996 0.091
Driver 4.517 4.278 0.239 0.628 0.517 0.111 2.778 2.693 0.086 1.111 1.069 0.043
8
6-iron 4.163 3.543 0.620 1.188 0.762 0.426 2.272 2.127 0.145 0.704 0.654 0.049
Driver 4.989 4.212 0.777 1.523 0.998 0.525 2.763 2.557 0.206 0.703 0.657 0.046
9
6-iron 4.826 4.194 0.632 0.847 0.687 0.160 3.021 2.838 0.183 0.958 0.669 0.288
Driver 5.735 5.132 0.604 1.054 0.791 0.262 3.577 3.410 0.167 1.104 0.930 0.174
10
6-iron 3.324 2.613 0.711 0.597 0.522 0.075 2.034 1.800 0.234 0.693 0.291 0.402
Driver 6.462 4.422 2.041 1.020 0.905 0.115 3.521 2.863 0.658 1.921 0.654 1.267
Group
6-iron 3.575 2.984 0.590 0.622 0.478 0.144 1.971 1.803 0.168 0.982 0.703 0.278
Driver 4.778 3.995 0.783 0.789 0.615 0.174 2.671 2.433 0.239 1.318 0.947 0.371
p-
value
0.001 0.001 0.001 0.000 0.000 0.087 0.001 0.001 0.001 0.001 0.000 0.001
All units are expressed as (N*m/kg)
2

Abbreviations: NJM = net joint moment, perp = perpendicular

Note: Individual player differences based on Cliff's Analog of Wilcoxon-Mann-Whitney Test and

modified Fisher-type method step-down technique.

### Significant at α = 0.05 when adjusted for multiple comparisons

p-value Significant group differences based on Sign Test at α = 0.05

Page 83 of 106

Table 8.2: Comparison of group and individual contributions to rear leg 3D Support Moment.
3D Support Moment Ankle Knee Hip
Player Club Total
About
perp
axis
Not
about
perp
axis
NJM
About
perp
axis
Not
about
perp
axis
NJM
About
perp
axis
Not
about
perp
axis
NJM
About
perp
axis
Not
about
perp
axis
1
6-iron 3.971 3.146 0.826 0.216 0.163 0.053 0.879 0.699 0.179 2.876 2.283 0.594
Driver 3.311 2.536 0.775 0.187 0.133 0.053 0.662 0.496 0.166 2.462 1.907 0.556
2
6-iron 3.405 3.308 0.097 0.252 0.236 0.016 0.540 0.513 0.027 2.613 2.559 0.054
Driver 3.750 3.631 0.119 0.725 0.699 0.026 0.460 0.426 0.033 2.565 2.506 0.059
3
6-iron 2.422 2.371 0.051 0.173 0.157 0.016 0.448 0.442 0.006 1.801 1.772 0.029
Driver 2.055 1.896 0.160 0.146 0.118 0.028 0.246 0.218 0.027 1.664 1.559 0.104
4
6-iron 4.373 3.643 0.730 0.281 0.265 0.015 0.912 0.738 0.174 3.180 2.639 0.540
Driver 5.571 5.282 0.289 0.532 0.505 0.027 0.736 0.672 0.063 4.303 4.105 0.198
5
6-iron 4.189 3.414 0.775 0.416 0.402 0.014 0.624 0.434 0.190 3.149 2.578 0.571
Driver 3.569 2.964 0.605 0.547 0.527 0.020 0.340 0.207 0.133 2.681 2.229 0.452
6
6-iron 5.812 5.576 0.236 0.521 0.451 0.070 0.946 0.907 0.040 4.344 4.217 0.127
Driver 4.797 4.547 0.250 0.521 0.438 0.083 0.580 0.522 0.058 3.696 3.587 0.109
7
6-iron 4.736 3.778 0.958 0.473 0.406 0.067 0.872 0.659 0.213 3.392 2.713 0.678
Driver 4.219 3.322 0.897 0.453 0.403 0.050 0.659 0.446 0.213 3.107 2.473 0.634
8
6-iron 4.800 4.458 0.342 0.439 0.327 0.112 0.937 0.835 0.102 3.424 3.296 0.127
Driver 5.272 5.015 0.257 0.486 0.374 0.111 0.873 0.807 0.066 3.913 3.833 0.079
9
6-iron 3.384 3.226 0.158 0.471 0.400 0.071 0.579 0.548 0.030 2.335 2.277 0.057
Driver 3.327 3.106 0.222 0.598 0.522 0.075 0.475 0.431 0.044 2.255 2.152 0.103
10
6-iron 6.663 5.439 1.224 0.378 0.327 0.051 1.153 0.909 0.244 5.132 4.203 0.929
Driver 4.717 3.931 0.786 0.264 0.224 0.040 0.678 0.529 0.150 3.774 3.179 0.596
Group
6-iron 4.376 3.836 0.540 0.362 0.313 0.048 0.789 0.668 0.121 3.225 2.854 0.371
Driver 4.059 3.623 0.436 0.446 0.394 0.051 0.571 0.476 0.095 3.042 2.753 0.289
p-
value
0.001 0.000 0.001 0.006 0.002 0.029 0.001 0.001 0.002 0.000 0.001 0.001
All units are expressed as (N*m/kg)
2

Abbreviations: NJM = net joint moment, perp = perpendicular

Note: Individual player differences based on Cliff's Analog of Wilcoxon-Mann-Whitney Test and

modified Fisher-type method step-down technique.

### Significant at α = 0.05 when adjusted for multiple comparisons

p-value Significant group differences based on Sign Test at α = 0.05

Page 84 of 106

Discussion
This study sought to understand how skilled golf players control the target and rear legs when
hitting golf shots with a 6-iron and driver. When swinging with the driver, target leg ankle, knee,
and hip NJMs increased as compared to the 6-iron. While player-specific solutions were evident,
the relative contribution of the leg ankle, knee, and hip NJMs to the target leg 3D support moment
were maintained across the group. Multi-joint control strategies used to control the target and rear
legs were found to be different, yet the majority of the 3D support moments were produced by
NJMs about an axis perpendicular to each leg plane. These results emphasize the importance of
recognizing how an individual coordinates RF generation between legs when using different clubs,
and can serve as a mechanistic foundation for the design of interventions that facilitate
improvements in player performance.
The results of this study are specific to highly skilled players with low handicaps. The consistency
of target and rear leg force-time characteristics in this study support previous research that
indicates golf swings with the same club and condition are consistent within-player and vary
between players [5,6,32,33]. To minimize the effect of experimental conditions on the results of
this investigation, we chose to standardize the surface and club conditions comparable to those
used by players when practicing swings on a driving range. The force plate coverings were thin
to provide practice-like friction conditions, while minimizing any attenuation of force transmitted to
the force-plates under each foot [6,32]. The inclusion criteria for this exploratory study limited
sample size and trials per participant. As a result the statistical approach incorporates within-
player comparisons to provide deeper insights into individual player strategies to regulate the
multi-joint control with each club.
The magnitude of the 3D support moment of each leg changed throughout the swing to satisfy
the whole-body mechanical objectives of the task. At transition, players began to increase the
NJMs in both rear and target legs to increase the 3D support moment during the interval of interest.
The mechanical loading of the legs coincided with the timing of increases in whole-body rotation
about the CM discovered previously [5,33]. Following the interval of interest, the rear and target
leg 3D support moments decreased with decreases in whole-body rotation about the CM as
players began to slow the rotation nearing ball contact [5,6]. Coincidentally, the 3D support
moment and NJMs of the rear leg may assist players to translate the CM toward the target as the
target leg extends near ball contact.
Between-club modifications of multi-joint control occurred primarily in the target leg, however both
the rear and target legs displayed player specific modifications in ankle, knee, and hip NJMs. With
the driver, target leg 3D support moment increased due to the increase in NJMs across the ankle,
knee, and hip. In contrast, the rear leg 3D support moment decreased in a majority of players as
a result of decreases in knee NJM. Amplification in target leg NJMs with the driver at each joint is
consistent with increases in target leg RF at a similar orientation relative to the leg. This finding
supports the idea that players used a similar target leg multi-joint control strategy with both clubs,
choosing to amplify the NJM magnitudes without changing the relative contribution to target leg
3D support moment. Decreasing rear leg knee NJM with the driver suggests a decrease in RF or
a change in relative orientation of the limb and reaction force. Reorientation of the limb would be
consistent with increasing the stance width when hitting with the driver [5].
Players may have simplified multi-joint control strategies by keeping the majority (>80%) of their
3D support moment about an axis perpendicular to the plane with both legs and across club
conditions. Generating RFs that are aligned with the leg plane to maintain NJMs primarily about
Page 85 of 106

an axis perpendicular to the leg plane may serve as a method of promoting advantageous NJMs.
This is analogous to aligning the RFs with the leg or leg plane during cutting and dance turning
movements [1,34]. Large RFs misaligned with the leg or NJMs not about an axis perpendicular to
the leg plane may be indications that the leg is in a disadvantageous position, and modifications
in technique may be necessary. There may also be benefits to coordinating the 3D support
moments in this manner to allow muscle groups of larger cross-sectional area to regulate the RFs
[1,35,36].
The relative contribution of the ankle, knee, and hip NJMs to the 3D support moment provides
insights into how the multi-joint control strategy creates the action of the leg. During the interval
of interest, the target leg 3D support moment relied primarily on knee NJM and secondarily on hip
NJM, whereas the rear leg relied primarily on hip NJM and secondarily on knee NJM. Because
the NJMs were primarily about an axis perpendicular to the leg plane, these findings suggest that
the target leg will require the action of the uni- and bi-articular muscles that cross the front of the
knee and hip joint, while the rear leg will require uni- and bi-articular muscles that cross the back
of the knee and hip joint. The activation of the bi-articular muscles would allow the larger muscle
groups of the respective legs to provide control at multiple joints [1,35,36].
This study advanced our understanding of how highly skilled golf players satisfy the mechanical
objectives at the whole-body level by coordinating RF generation between legs and multi-joint
control within leg when hitting golf shots of varying distances. Although individuals used player
specific strategies, target leg NJMs were amplified when swinging with the driver, while rear leg
NJMs were generally maintained. Although multi-joint control strategies were different between
legs, the majority of 3D support moment was generated about an axis perpendicular to the plane
with both legs, which may simplify control strategies by allowing larger muscle groups to
coordinate the action. Understanding the relative NJM distribution within a leg, along with its
kinematic context, provides insights as to the challenge imposed on the uni- and bi-articular
muscles contributing to the lower extremity joint kinetics of the target and rear legs during the golf
swing. These results highlight the need for player- and leg- specific interventions to improve
regulation of shot distance with different clubs.
Page 86 of 106

References
[1] Zaferiou, A. M., Flashner, H., Wilcox, R. R., and McNitt-Gray, J. L., 2017, “Lower extremity
control during turns initiated with and without hip external rotation,” J. Biomech., 52, pp.
130–139.
[2] Mathiyakom, W., McNitt-Gray, J. L., and Wilcox, R., 2006, “Lower extremity control and
dynamics during backward angular impulse generation in backward translating tasks.,” Exp.
Brain Res., 169(3), pp. 377–88.
[3] Mathiyakom, W., McNitt-Gray, J. L., and Wilcox, R., 2006, “Lower extremity control and
dynamics during backward angular impulse generation in forward translating tasks.,” J.
Biomech., 39(6), pp. 990–1000.
[4] McNitt-Gray, J. L., Sand, K., Ramos, C., Peterson, T. J., Held, L., and Brown, K., 2015,
“Using technology and engineering to facilitate skill acquisition and improvements in
performance,” Proc. Inst. Mech. Eng. Part P J. Sport. Eng. Technol., 229(2), pp. 103–115.
[5] Peterson, T. J., Wilcox, R. R., and McNitt-Gray, J. L., 2016, “Angular impulse and balance
regulation during the golf swing,” J. Appl. Biomech., 32(4), pp. 342–349.
[6] McNitt-Gray, J. L., Munaretto, J., Zaferiou, A., Requejo, P. S., and Flashner, H., 2013,
“Regulation of reaction forces during the golf swing,” Sport. Biomech., 12(2), pp. 121–131.
[7] McNitt-Gray, J. L., Hester, D. M., Mathiyakom, W., and Munkasy, B. a, 2001, “Mechanical
demand and multijoint control during landing depend on orientation of the body segments
relative to the reaction force.,” J. Biomech., 34(11), pp. 1471–82.
[8] Goh, H.-T., Sullivan, K. J., Gordon, J., Wulf, G., and Winstein, C. J., 2012, “Dual-task
practice enhances motor learning: a preliminary investigation,” Exp. Brain Res., 222(3), pp.
201–210.
[9] Macpherson, J. M., 1988, “Strategies that simplify the control of quadrupedal stance. I.
Forces at the ground.,” J. Neurophysiol., 60(1), pp. 204–17.
[10] Ting, L. H., and Macpherson, J. M., 2004, “A Limited Set of Muscle Synergies for Force
Control During a Postural Task,” J. Neurophysiol., 93(1), pp. 609–613.
[11] Mathiyakom, W., McNitt-Gray, J. L., Requejo, P. S., and Costa, K., 2005, “Modifying center
of mass trajectory during sit-to-stand tasks redistributes the mechanical demand across
the lower extremity joints.,” Clin. Biomech. (Bristol, Avon), 20(1), pp. 105–11.
[12] Vaughan, C. L., 1981, “A three-dimensional analysis of the forces and torques applied by
a golfer during the downswing.pdf,” Biomech. VII-B, pp. 325–331.
[13] Neal, R., and Wilson, B. D., 1982, “3D Kinematics and Kinetics of the Golf Swing,” Int. J.
Sport Biomech., 1, pp. 221–233.
[14] Donatelli, R., Dimond, D., and Holland, M., 2012, “Sport-specific biomechanics of spinal
injuries in the athlete (throwing athletes, rotational sports, and contact-collision sports).,”
Clin. Sports Med., 31(3), pp. 381–96.
[15] Foxworth, J. L., Millar, A. L., Long, B. L., Way, M., Vellucci, M. W., and Vogler, J. D., 2013,
“Hip joint torques during the golf swing of young and senior healthy males.,” J. Orthop.
Sports Phys. Ther., 43(9), pp. 660–5.
[16] Lynn, S. K., and Noffal, G. J., 2010, “Frontal plane knee moments in golf: Effect of target
Page 87 of 106

side foot position at address,” J. Sport. Sci. Med., 9(2), pp. 275–281.
[17] Ward, C. E., 2015, “Hip Moments in the Golf Swing: Relationship to Overhead Deep Squat
& Early Extension,” California State University, Fullerton.
[18] Gatt, C., Pavol, M., Parker, R., and Grabiner, M., 1998, “Three-dimensional knee joint
kinetics during a golf swing influences of skill level and footwear,” Am. J. Sports Med., (Idd).
[19] Russell, I. M., Raina, S., Requejo, P. S., Wilcox, R. R., Mulroy, S., and McNitt-Gray, J. L.,
2015, “Modifications in Wheelchair Propulsion Technique with Speed,” Front. Bioeng.
Biotechnol., 3(October), pp. 1–11.
[20] Worsfold, P., Smith, N., and Dyson, R., 2008, “Low handicap golfers generate more torque
at the shoe-natural grass interface when using a driver,” J. Sport. Sci. Med., (7), pp. 408–
414.
[21] Shimba, T., 1996, “Consequences of Force Platform Studies,” Natl. Rehabiliation Res. Bull.
Japan, (17), pp. 17–23.
[22] Zatsiorsky, V. M., 2002, “Wrench Representation of the Ground Reaction Force,” Kinetics
of Human Motion, Human Kinetics Publishers, Champaign, IL, pp. 43–48.
[23] Jackson, K. M., 1979, “Fitting of Mathematical Functions to Biomechanical Data,” IEEE
Trans. Biomed. Eng., BME-26(2), pp. 122–124.
[24] de Leva, P., 1996, “Adjustments to Zatsiorsky-Seluyanov’s segment inertia parameters,”
J. Biomech., 29(9), pp. 1223–30.
[25] Schwartz, M. H., and Rozumalski, A., 2005, “A new method for estimating joint parameters
from motion data.,” J. Biomech., 38(1), pp. 107–16.
[26] Winter, D. A., 1980, “Overall Principle of Lower Limb Spport During Stance Phase of
Gait.pdf.crdownload,” J. Biomech., 13, pp. 923–927.
[27] Cliff, N., 1996, Ordinal Methods for Behavioral Data Analysis, Lawrence Erlbaum
Associates, Inc., Publishers, Mahwah, NJ.
[28] Neuhäuser, M., Lösch, C., and Jöckel, K.-H., 2007, “The Chen–Luo test in case of
heteroscedasticity,” Comput. Stat. Data Anal., 51(10), pp. 5055–5060.
[29] Hochberg, Y., and Tamhane, A. C., 1987, Multiple Comparison Procedures, John Wiley &
Sons, Inc., Hoboken, NJ, USA.
[30] Hochberg, Y., 1988, “A sharper Bonferroni test for multiple tests of significance,”
Biometrika, 75, pp. 800–802.
[31] Wilcox, R., and Clark, F., 2015, “Robust Multiple Comparisons Based on Combined
Probabilities From Independent Tests,” J. Data Sci., 13(1), pp. 1–11.
[32] Williams, K., and Cavanagh, P. R., 1983, “The mechanics of foot action during the golf
swing and implications for shoe design.,” Med. Sci. Sports Exerc., 15(3), pp. 247–255.
[33] Barrentine, S. W., Fleisig, G. S., and Johnson, H., 1994, “Ground Reaction Forces and
Torques of Professional and Amateur Golfers,” Sci. Golf II Proc. 1994 World Sci. Congr.
Golf, 1(1), pp. 33–39.
[34] Qiao, M., Brown, B., and Jindrich, D., 2014, “Compensations for increased rotational
Page 88 of 106

inertia during human cutting turns,” J. Exp. …, (217), pp. 432–443.
[35] Lieber, R. L., 2002, Skeletal muscle structure, function & plasticity: the physiological basis
of rehabilitation, Lippincott Williams & Wilkins, Philadelphia.
[36] Lieber, R. L., and Ward, S. R., 2011, “Skeletal muscle design to meet functional demands,”
Phil. Trans. R. Soc. B, 366(1570), pp. 1466–1476.

Page 89 of 106

: Between-Club Differences in Lower Extremity
Muscle Activation while Regulating Reaction Forces during
the Golf Swing

Introduction
To successfully complete the golf swing, the
player must generate reaction forces (RFs) at
the foot-surface interface to create rotation of
the body-club system toward the target [1–3].
Highly skilled golf players have been found to
consistently coordinate and regulate RF
generation of the system of both leg when
increasing golf shot distance both within and
between clubs [2,3]. The consistency of RF
regulation suggests that players precisely
coordinate the activation of the lower extremity
muscles to control their foot-surface
interactions (Figure 9.1).Understanding how
highly skilled players control the muscles of the
legs to regulate RF generation provides insight
into player performance and the control
preferences of individual players [4].
Research on the muscle activation of the lower
limbs [5,6], trunk [7–13], shoulder [7,14–17],
and forearm [18,19] has been conducted to
understand neuromuscular control involved in
golf swing performance. These studies
focused primarily on relating muscle activation
patterns to swing kinematics. Bechler and
associates discovered the rear leg hip
extensors (gluteus maximus, biceps femoris and semimembranosus) and target leg adductor
magnus and vastus lateralis were active during forward swing as the legs worked in conjunction
to rotate the pelvis [5]. They hypothesized that peak activity in target leg hip extensors was
associated with maintenance of knee flexion in the target leg during the acceleration phase [5].
Marta and colleagues observed increases of muscle activation during the forward swing and
acceleration phases in the target leg semitendinosus, right vastus medialis, rectus femoris,
gastrocnemius and tibialis anterior when hitting with a 4-iron compared to a pitching wedge [6].
These muscle activation patterns appear to be consistent with RFs measured during golf swings
[2,3,20] and in other athletic tasks [21–23].
Skilled golf players have been found to effectively coordinate RF generation between the target
and rear legs during the golf swing to satisfy the mechanical objectives of the golf swing [2,3].
Figure 9.1: Resultant horizontal reaction force of 3 trials for
rear and target legs of player 12 (mean ± SD) to display
consistency of performance across trials. Ball contact
occurs at t = 0s, time of transition at t ≈ 0.3s. Images below
represent view from underneath player at 0.1s intervals
throughout the downswing. Rear leg (left) reaction forces
(RFs) are represented in blue, target leg (right) RFs are
represented in green. Notice how RFs redirect near
transition to create rotation about the center of mass.
Page 90 of 106

During the downswing, players have been found to redirect and amplify target and rear leg
resultant horizontal reaction forces (RFhs) to generate net angular impulse needed to perform the
task [3]. The rear leg is involved in generating a RFh directed anterior and toward the target
whereas the target leg is involved in generating a RFh directed posterior and away from the target,
both of which help to create rotation about the center of mass [2,3,20]. Previous studies have
revealed that highly skilled players regulate rear or target leg RFhs while maintaining RFh-angle
relative to the target line when regulating golf shot distance with the same club (e.g. 6-iron) [2].
Increasing RFh magnitude while maintaining orientation suggests that selective activation of the
same muscle groups could serve as an effective RFh regulation strategy. When regulating golf
shot distance between a 6-iron and driver, skilled players increased RFh magnitude yet directed
the target RFh more parallel to the target [3]. The change RFh orientation when increasing RFh
magnitude between swings with different clubs suggests that different muscles may be involved
when a player uses different clubs.
The orientation of the RF relative to lower extremity segments is known to affect the target and
rear leg joint kinetics during the golf swing [24]. Modifications in the contribution of the ankle, knee,
and hip net joint moments (NJMs) to the target and rear leg 3D support moments are expected to
affect the muscles involved in the generation of the NJMs required during the swing. We expect
that target leg knee extensor muscles will be involved in generating knee extensor NJMs that are
the primary contributors to the target leg 3D support moment during the swing [24,25]. Similarly,
we expect that rear leg hip extensor muscles will be involved in generating the hip extensor NJMs
that are the primary contributors to the rear leg 3D support moment during the swing [24,26].
The aim of this study was to determine how highly skilled golf players coordinate lower extremity
muscle activation in the rear and target legs to regulate RFs generated during swings with a 6-
iron and driver. We expected control preferences of the individual player would be maintained
within leg across clubs, where RF generation by each leg would involve the same set of muscles,
respectively. We hypothesized the increases in RF generation between clubs would correspond
with increases in activation of the involved muscles. Lastly, we expected differences in muscle
activation patterns would correspond to observed differences in RF generation during the interval
of interest. We tested this hypothesis by comparing the muscle activation and RFs of the lower
extremities while performing golf swings with the 6-iron and driver.
Materials and Methods
Skilled golfers (n = 12; handicap < 5, right handed) volunteered and were provided informed
consent prior to enrolling in this study in accordance with local Institutional Review Board.
Subjects were on average (standard deviation) 28.6 (14.2) years old, 1.76 (0.12) m in height, and
76.3 (15.4) kg in body mass. Each player performed 3-4 swings using a driver (D) and a 6-iron
(6I, Taylor Made-adidas golf), where each player was instructed to hit a full swing with each club
as they normally would on the golf course. Placement of the force plates allowed players to use
their preferred address position to hit the golf ball toward the target.
Three dimensional RFs at the artificial turf-plate interface were quantified during each swing with
each foot being supported by a single force plate (Kistler, 1200 Hz) [1]. Ball contact was
synchronized at the time of club-ball contact using a microphone signal collected simultaneously
with the RF-time data (t = 0s at ball contact, National Instruments, Austin, Texas). Activation of
Page 91 of 106

lower extremity muscles were monitored via telemetered electromyography (EMG) using surface
electrodes (1x1cm
2
Konigsberg, Pasadena, CA). The EMG and force signals were captured
simultaneously through a BNC connection between the EMG system and force collection
computer.
The muscle activation root mean square values were filtered with a zero-lag, fourth-order
recursive Butterworth filter (10-350 Hz), full wave rectified, and integrated into 20 ms bins [27].
Muscle activations were normalized by a sustained level of contraction during a manual muscle
test conducted to isolate the individual muscle [28]. Muscle activations were measured bilaterally
for: gluteus maximus (GMax), gluteus medius (Gmed), biceps femoris (BF),
semimembranosus/semitendinosus (SM/ST), rectus femoris (RF), vastus lateralis (VL), adductors
(Add), and tensor fasciae latae (TFL).
To capture the muscle activity associated with the rise of RFh, the interval of interest began when
target leg RFh reached 25% of its peak value and ended when it reached 75% of its peak value.
The interval of interest typically coincided with the transition from backswing to downswing and
continued through early downswing. The level of muscle activation was expressed as an
integration of bins occurring during the interval of interest to measure the amount of activity
produced by the muscle. Between club differences in integrated muscle activity and RFs across
the group were determined using the Sign Test (α = 0.05).
Results
Individual players activated the same set of muscles when swinging the 6-iron and driver. The set
of muscles involved in RF regulation were player-specific and muscle activation patterns
corresponded with RF generation during the swing. The primary muscles involved in RF
generation during early downswing included: target leg rectus femoris and vastus lateralis, as well
as rear leg gluteus maximus and biceps femoris (Figure 9.2). Additionally, most players activated
target leg adductors along with rear leg semimembranosus/semitendinosus and gluteus medius.
Target leg peak resultant horizontal RF and vertical RF at peak RFh significantly increased with
the driver compared to the 6-iron (p = < 0.001, Table 9.1). Rear leg peak RFs did not reveal
significant changes. The rate of increase in the resultant horizontal and vertical RFs for both legs
was comparable between clubs. Target leg RFhs reached 25% of peak value to create rotation
about the center of mass at approximately ~0.3s prior to ball contact. Across the group, players
significantly increased the duration of the interval of interest with the driver (p = 0.007) as
compared to the 6-iron.
Observed changes in target leg muscle activation corresponded with changes in target leg RFs
when using the driver compared to the 6-iron. Muscle activation of target leg adductors (p = 0.004),
rectus femoris (p = 0.004), vastus lateralis (p = 0.03), gluteus medius (p < 0.001), and
semimembranosus/semitendinosus (p < 0.001) significantly increased the amount of integrated
muscle activity with the driver compared to the 6-iron across the group while target leg RFs
increased (Figure 9.3 & Figure 9.4). The rear leg biceps femoris (p = 0.004), gluteus maximus (p
= 0.004), gluteus medius (p = 0.004), semimembranosus/semitendinosus (p = 0.004), adductors
(p = 0.001), and tensor fasciae latae (p = 0.004) significantly increased their integrated muscle
Page 92 of 106

activity with the driver compared to the 6-iron without increasing rear leg RFs across the group.
Figure 9.3 & Figure 9.4 highlight differences in muscle activation patterns for two different players.
Figure 9.2: Typical muscle activation and reaction forces of the rear and target legs for player 7 during one swing with
a driver. Note that rear leg hip extensor muscles and target leg hip flexor and knee extensor muscles are active during
the interval of interest as the resultant horizontal force (RFh) increases. Vertical lines denote the interval of interest
starting at a time of 25% of peak RFh and ending at 75% of peak RFh. Time = 0s is ball contact. GMax = gluteus
maximus, GMed = gluteus medius, Add = adductors, TFL = tensor fasciae latae, RF = rectus femoris, VL = vastus
lateralis, BF = biceps femoris, SM/ST = semimembranosus/semitendinosus, RFv = vertical reaction force , RFh =
resultant horizontal reaction force, RFv = vertical reaction force.
Page 93 of 106

Table 9.1: Comparison of Target and Rear leg horizontal and vertical reaction forces.

Target Leg Rear Leg

Peak RFh (BW) RFv at Peak RFh (BW) Peak RFh (BW) RFv at Peak RFh (BW)

6-iron Driver 6-iron Driver 6-iron Driver 6-iron Driver
Player Mean SD Mean SD Mean SD Mean SD Mean SD Mean SD Mean SD Mean SD
1 0.29 0.02 0.32 0.02 0.72 0.04 0.92 0.09 0.25 0.01 0.25 0.01 0.22 0.05 0.18 0.03
2 0.47 0.01 0.47 0.02 1.49 0.04 1.40 0.09 0.31 0.01 0.33 0.01 0.41 0.05 0.41 0.09
3 0.35 0.03 0.40 0.00 0.76 0.07 0.91 0.01 0.24 0.01 0.23 0.01 0.46 0.04 0.42 0.05
4 0.17 0.01 0.21 0.02 0.61 0.01 0.71 0.02 0.22 0.01 0.25 0.01 0.51 0.04 0.49 0.02
5 0.30 0.01 0.30 0.01 0.98 0.06 1.03 0.03 0.16 0.00 0.14 0.00 0.16 0.01 0.21 0.02
6 0.35 0.01 0.41 0.02 1.05 0.13 1.99 0.05 0.31 0.01 0.29 0.03 0.26 0.13 0.92 0.10
7 0.23 0.01 0.27 0.02 0.34 0.01 0.41 0.02 0.20 0.00 0.21 0.01 0.45 0.01 0.45 0.05
8 0.43 0.03 0.49 0.02 1.54 0.28 1.72 0.09 0.28 0.01 0.32 0.01 0.32 0.06 0.36 0.03
9 0.21 0.01 0.27 0.01 0.86 0.20 1.05 0.02 0.17 0.02 0.18 0.01 0.38 0.03 0.49 0.02
10 0.17 0.00 0.21 0.01 0.72 0.11 0.87 0.08 0.19 0.00 0.19 0.00 0.46 0.07 0.39 0.06
11 0.31 0.01 0.35 0.01 0.90 0.01 1.04 0.01 0.19 0.00 0.19 0.00 0.22 0.02 0.27 0.01
12 0.26 0.01 0.34 0.01 1.02 0.02 1.14 0.07 0.21 0.00 0.23 0.01 0.39 0.02 0.37 0.01
Group 0.29 0.01 0.34 0.01 0.92 0.08 1.10 0.05 0.23 0.01 0.23 0.01 0.35 0.04 0.41 0.04
p < 0.001* < 0.001* 0.37 0.78
* Group differences as determined by the Sign Test ( α = 0.05)

Page 94 of 106

Figure 9.3: Comparison of muscle activation and reaction forces of the rear (left) and target (right) legs for player 10
during one swing with both the 6-iron (-) and driver (+). Notice increased integrated muscle activity of target leg hip
flexors and knee extensors (RF and VL) during the interval of interest as resultant horizontal reaction force (RFh)
increases with the driver. Images below represent view from underneath player at 0.1s intervals throughout the swing.
Vertical lines denote the interval of interest starting at a time of 25% of peak RFh and ending at 75% of peak RFh. Time
= 0s is ball contact. Rear leg (left) reaction forces are represented in blue, target leg (right) reaction forces are
represented in green. GMax = gluteus maximus, GMed = gluteus medius, Add = adductors, TFL = tensor fasciae latae,
RF = rectus femoris, VL = vastus lateralis, BF = biceps femoris, SM/ST = semimembranosus/semitendinosus, RFv =
vertical reaction force , RFh = resultant horizontal reaction force, RFv = vertical reaction force.
Page 95 of 106

Figure 9.4: Comparison of muscle activation and reaction forces of the rear (left) and target (right) legs for player 1
during one swing with both the 6-iron (-) and driver (+). Notice increased integrated muscle activity of target leg knee
extensors (VL) as well as rear leg hip extensors (GMax, SM/ST, and BF) during the interval of interest as resultant
horizontal reaction force (RFh) increases with the driver. Images below represent view from underneath player at 0.1s
intervals throughout the swing. Vertical lines denote the interval of interest starting at a time of 25% of peak RFh and
ending at 75% of peak RFh. Time = 0s is ball contact. Rear leg (left) reaction forces are represented in blue, target leg
(right) reaction forces are represented in green. GMax = gluteus maximus, GMed = gluteus medius, Add = adductors,
TFL = tensor fasciae latae, RF = rectus femoris, VL = vastus lateralis, BF = biceps femoris, SM/ST =
semimembranosus/semitendinosus, RFv = vertical reaction force , RFh = resultant horizontal reaction force, RFv =
vertical reaction force.
Page 96 of 106

Discussion
This study aimed to determine how skilled golf players coordinate lower extremity muscle
activation to regulate RFs generated during the golf swing. In general, players activated the same
muscles sets in each leg when hitting with both clubs. Activation of muscles on the anterior of the
target leg was consistent with posterior directed RFs, and active muscles on the posterior of the
rear leg were consistent with anterior directed RFs that were coordinated between legs to create
rotation about the center of mass. For the highly skilled players in this study, increases in target
leg muscle activation corresponded to increases in target leg RFs when using the driver compared
to the 6-iron. Understanding how individuals coordinate muscular activity of the lower extremities
while they regulate RF generation between clubs could lead to the identification of effective
strategies players can use to increase golf shot distance.
The insights gained in this study are specific to highly skilled golf players with low handicaps. The
consistency of RFs generated within a club by individuals is aligned with previous research
(McNitt-Gray et al. 2013; Peterson et al. 2016). A laboratory environment was necessary to
complete some aspects of the data collection, but we standardized clubs and conditions to what
players would normally use when practicing on a driving range. The inclusion criteria for this study
limited the sample size and trials per condition, however the statistical methods chosen aimed to
mitigate these effects by using each player as their own control.
When increasing shot distance with the driver compared to the 6-iron, players increased target
leg RFs. Rear leg RFs were not found to be significantly different between clubs. Increases in
target leg RFs corresponded with increases in the level of target leg muscle activation during
swings with the driver as compared to the 6-iron. An increase in the duration of the interval of
interest associated with the rise of RFh potentially contributed to observed increases in the level
of muscle activity observed with the driver compared to the 6-iron. In addition, the multimodal
activation of agonistic and antagonistic muscles acting to control each leg, involves control
priorities other than just regulation of RFh magnitude and direction. Between-player differences
in muscle activation patterns provide insights as to how individuals choose to accomplish the golf
swing when using different clubs.
Within-player comparisons of lower extremity muscle activation patterns indicate that individual
players executed the golf swing similarly by using the same set of muscles across clubs. As found
previously, the skilled players in this study generated target and rear leg RFhs to create rotation
about the center of mass [2,3,20]. Analysis of the muscle activations corresponding to target and
rear leg RFh generation indicates that individual players activated the same set of muscles when
hitting with both clubs. This action required coordinated muscle activation between legs to
produce the RFhs necessary for each task. During the interval of interest associated with the rise
in RFh, muscles on the anterior of the target leg (rectus femoris and vastus lateralis) were
activated to produce a “push” on the ground thereby generating a RFh posterior and away from
the target. Simultaneously, the muscles on the posterior of the rear leg (gluteus maximus,
semimembranosus/semitendinosus, and biceps femoris) were activated to create a “pull” on the
on the ground thereby generating a RFh anterior and toward the target. Increases in the level of
target leg muscle activation corresponded with increases in target leg RFh magnitudes.
Page 97 of 106

Rotation and translation of the body towards the target during the downswing requires coordinated
generation of RFs between legs as well as effective multjoint control of the ankle, knee, and hip
of each leg [3]. Players in this study used a combination of mono- and bi-articular muscles to
coordinate control of the target and rear legs during the golf swing. The bi-articular rectus femoris
of the target leg was activated, which extended the knee while also flexing the hip. Target leg
knee extension was assisted by the mono-articulatar vastus lateralis, while target leg adductors
were activated to adduct the leg. The rear leg activated gluteus maximus, biceps femoris and
semimembranosus/semitendinosus to extend the hip. Gluteus medius was also active to assist in
rear leg hip abduction. Additionally, the gluteus muscles may aid in hip stabilization while
producing their primary actions. The hamstrings muscles may have implications toward rear leg
knee flexion while also stabilizing the knee joint through their counteracting rotational functions at
the knee. These findings are consistent with those discovered in other studies [5,6]. Co-
contraction of synergistic and antagonistic muscles may assist players in effectively creating joint
action while maintaining joint stability. These actions are important in rotating the body towards
the target while also translating the center of mass toward the target during the downswing [3].
The activation of muscles crossing the hips has consequences on trunk position as well as whole-
body dynamics and lower extremity multijoint control. As trunk position changes, the total body
center of mass and its interaction with the RFs at the ground has implications for linear and
angular impulse generation at the whole body level [21,29]. Because the trunk is comprised of
approximately 50% of body mass [30], the position of the trunk during the swing is critical when
controlling the RF relative to the total body center of mass. Activation of muscles crossing a joint
contributes to the net joint moment controlling the adjacent segments. As the trunk changes
position relative to the thigh, the muscles across the hip change length. It has been shown that
muscles tend to have a preferred length at which to produce force [31]. Maintaining these muscle
lengths across golf shots with different clubs may allow players to consistently produce the
necessary RFs to achieve mechanical objectives at the whole body level while effectively
controlling the target and rear legs. Trunk position may also affect how trunk stability is achieved
through activation of muscles intrinsic to the trunk (core muscles). Further investigation is needed
to elucidate the how control of the lower extremities affects control of the trunk during the golf
swing [13,32].
Activation of the mono- and bi-articular muscles on the anterior of the target leg and posterior of
the rear leg corresponds with the generation of the RFs needed to create rotation and translation
during the golf swing as well as the 3D support moments of the target and rear legs [24]. When
accounting for the electromechanical delay ( 20-60 ms) [33,34], the muscle activation patterns of
the target and rear legs found in this study are consistent with the ankle, knee, and hip net joint
moments of the target and rear legs necessary to complete the golf swing [24]. During the late
downswing golf players have been found to rely primarily on target leg knee net joint moment and
rear leg hip net joint moment, and secondarily on target leg hip net joint moment and rear leg
knee net joint moment [24]. These results are consistent with activation of the muscles anterior to
the target leg and posterior to the rear leg found in this study.
In general, muscular activity of the lower extremities increased with the driver compared with the
6-iron. The increases in muscular activation observed in swings with the driver involved activation
of the same set of target and rear leg muscles when hitting with the 6-iron. The scaling of target
Page 98 of 106

and rear leg muscle activation between clubs as a method of regulating RF magnitude and
direction likely benefits multi-joint control of each leg. Although this study focused on the
neuromuscular control of the hip and knee, neuromuscular control of the lower extremities in
relation to the trunk requires further investigation. Knowledge of player-specific modifications in
lower extremity muscle activation when regulating RFs during golf swings with different clubs
provides a basis of understanding an individual’s muscular control while regulating golf shot
distance, and will aid in designing interventions that take advantage of a player’s strengths to
enhance performance.

Page 99 of 106

References
[1] Williams, K., and Cavanagh, P. R., 1983, “The mechanics of foot action during the golf
swing and implications for shoe design.,” Med. Sci. Sports Exerc., 15(3), pp. 247–255.
[2] McNitt-Gray, J. L., Munaretto, J., Zaferiou, A., Requejo, P. S., and Flashner, H., 2013,
“Regulation of reaction forces during the golf swing,” Sport. Biomech., 12(2), pp. 121–131.
[3] Peterson, T. J., Wilcox, R. R., and McNitt-Gray, J. L., 2016, “Angular impulse and balance
regulation during the golf swing,” J. Appl. Biomech., 32(4), pp. 342–349.
[4] McNitt-Gray, J. L., Sand, K., Ramos, C., Peterson, T. J., Held, L., and Brown, K., 2015,
“Using technology and engineering to facilitate skill acquisition and improvements in
performance,” Proc. Inst. Mech. Eng. Part P J. Sport. Eng. Technol., 229(2), pp. 103–115.
[5] Bechler, J. R., Jobe, F. W., Pink, M., Perry, J., and Ruwe, P. A., 1995, “Electromyographic
analysis of the hip and knee during the golf swing,” Clin J Sport Med, 5(3), pp. 162–166.
[6] Marta, S., Silva, L., Vaz, J. R., Castro, M. A., Reinaldo, G., and Pezarat-Correia, P., 2016,
“Electromyographic analysis of lower limb muscles during the golf swing performed with
three different clubs,” J. Sports Sci., 34(8), pp. 713–720.
[7] Pink, M., Perry, J., and Jobe, F. W., 1993, “Electromyographic analysis of the trunk in
golfers,” Am J Sport. Med, 21(3), pp. 385–388.
[8] Watkins, R. G., Uppal, G. S., Perry, J., Pink, M., and Dinsay, J. M., 1996, “Dynamic
electromyographic analysis of trunk musculature in professional golfers,” Am J Sport. Med,
24(4), pp. 535–538.
[9] Cole, M. H., and Grimshaw, P. N., 2008, “Electromyography of the trunk and abdominal
muscles in golfers with and without low back pain,” J. Sci. Med. Sport, 11(2), pp. 174–181.
[10] Cole, M. H., and Grimshaw, P. N., 2008, “Trunk muscle onset and cessation in golfers with
and without low back pain,” J. Biomech., 41(13), pp. 2829–2833.
[11] Bulbulian, R., Ball, K. A., and Seaman, D. R., 2001, “The short golf backswing: Effects on
performance and spinal health implications,” J. Manipulative Physiol. Ther., 24(9), pp. 569–
575.
[12] Horton, J. F., Lindsay, D. M., and Macintosh, B. R., 2001, “Abdominal muscle activation of
elite male golfers with chronic low back pain.,” Med. Sci. Sports Exerc., 33(29), pp. 1647–
1654.
[13] Silva, L., Marta, S., Vaz, J., Fernandes, O., Castro, M. A., and Pezarat-Correia, P., 2013,
“Trunk muscle activation during golf swing: Baseline and threshold.,” J. Electromyogr.
Kinesiol., 23(5), pp. 1174–82.
[14] Jobe, F. W., Perry, J., and Pink, M., 1989, “Electromyographic shoulder activity in men
and women professional golfers.,” Am. J. Sports Med., 17(6), pp. 782–7.
[15] Kao, J. T., Pink, M., Jobe, F. W., and Perry, J., 1995, “Electromyographic analysis of the
Page 100 of 106

scapular muscles during a golf swing.,” Am. J. Sports Med., 23(1), pp. 19–23.
[16] Pink, M., Jobe, F. W., and Perry, J., 1990, “Electromyographic analysis of the shoulder
during the golf swing.,” Am. J. Sports Med., 18(2), pp. 137–140.
[17] Jobe, F. W., Moynes, D. R., and Antonelli, D. J., 1986, “Rotator cuff function during a golf
swing.,” Am. J. Sports Med., 14(5), pp. 388–392.
[18] Farber, A. J., Smith, J. S., Kvitne, R. S., Mohr, K. J., and Shin, S. S., 2009,
“Electromyographic analysis of forearm muscles in professional and amateur golfers.,” Am.
J. Sports Med., 37(2), pp. 396–401.
[19] Glazebrook, M. A., Curwin, S., Islam, M. N., Kozey, J., and Stanish, W. D., 1994, “Medial
epicondylitis. An electromyographic analysis and an investigation of intervention
strategies.,” Am. J. Sports Med., 22(5), pp. 674–679.
[20] Barrentine, S. W., Fleisig, G. S., and Johnson, H., 1994, “Ground Reaction Forces and
Torques of Professional and Amateur Golfers,” Sci. Golf II Proc. 1994 World Sci. Congr.
Golf, 1(1), pp. 33–39.
[21] Mathiyakom, W., McNitt-Gray, J. L., and Wilcox, R., 2006, “Lower extremity control and
dynamics during backward angular impulse generation in backward translating tasks.,” Exp.
Brain Res., 169(3), pp. 377–88.
[22] McNitt-Gray, J. L., 2000, “Subject specific coordination of two- and one-joint muscles
during landings suggests multiple control criteria.,” Motor Control, 4, pp. 84–88.
[23] McNitt-Gray, J. L., 2000, “Musculoskeletal loading during landing,” Biomech. Sport, (Idd).
[24] Peterson, T. J., 2017, “Chapter 8: Lower Extremity Joint Control during the Golf Swing,”
Lower Extremity Control and Dynamics during the Golf Swing (Dissertation).
[25] Gatt, C., Pavol, M., Parker, R., and Grabiner, M., 1998, “Three-dimensional knee joint
kinetics during a golf swing influences of skill level and footwear,” Am. J. Sports Med., (Idd).
[26] Foxworth, J. L., Millar, A. L., Long, B. L., Way, M., Vellucci, M. W., and Vogler, J. D., 2013,
“Hip joint torques during the golf swing of young and senior healthy males.,” J. Orthop.
Sports Phys. Ther., 43(9), pp. 660–5.
[27] de Luca, C., 1997, “The use of surface electromyography in biomechanics,” J. Appl.
Biomech., (13), pp. 135–163.
[28] Kendall, F., McCreary, E., Provance, P., Rodgers, M., and Romani, W., 2005, Muscles:
Testing and Function, with Posture and Pain, Lippincott Williams & Wilkins, Baltimore.
[29] Mathiyakom, W., McNitt-Gray, J. L., and Wilcox, R., 2006, “Lower extremity control and
dynamics during backward angular impulse generation in forward translating tasks.,” J.
Biomech., 39(6), pp. 990–1000.
[30] de Leva, P., 1996, “Adjustments to Zatsiorsky-Seluyanov’s segment inertia parameters,”
J. Biomech., 29(9), pp. 1223–30.
Page 101 of 106

[31] Herzog, W., Guimaraes, a C., Anton, M. G., and Carter-Erdman, K. a, 1991, “Moment-
length relations of rectus femoris muscles of speed skaters/cyclists and runners.,” Med.
Sci. Sports Exerc., 23(11), pp. 1289–1296.
[32] Marta, S., Silva, L., Castro, M. A., Pezarat-Correia, P., and Cabri, J., 2012,
“Electromyography variables during the golf swing: a literature review.,” J. Electromyogr.
Kinesiol., 22(6), pp. 803–13.
[33] McNitt-Gray, J. L., Hester, D. M., Mathiyakom, W., and Munkasy, B. a, 2001, “Mechanical
demand and multijoint control during landing depend on orientation of the body segments
relative to the reaction force.,” J. Biomech., 34(11), pp. 1471–82.
[34] Bach, T. M., Chapman, A. E., and Calvert, T. W., 1983, “Mechanical resonance of the
human body during voluntary oscillations about the ankle joint.,” J. Biomech., 16(1), pp.
85–90.

Page 102 of 106

Appendix A - Force Plate Documentation

Force, moment and center of pressure calculations based on manufacturer documentation [1].
Page 103 of 106

Page 104 of 106

Page 105 of 106

Page 106 of 106

References

[1] Vaughan, C. L., 1999, Kistler Force Plate Formulae.

Abstract (if available)

Linked assets

University of Southern California Dissertations and Theses

Conceptually similar

PDF

Regulation of linear and angular impulse generation: implications for athletic performance

PDF

Identifying injury risk, improving performance, and facilitating learning using an integrated biomechanics informatics system (IBIS)

PDF

Upper extremity control and dynamics during manual wheelchair propulsion at different speeds and wheelchair configurations

PDF

Control and dynamics of turning tasks with different rotation and translation requirements

PDF

Using nonlinear feedback control to model human landing mechanics

PDF

Control and dynamics during horizontal impulse generation

PDF

Understanding reactive balance control strategies in non-disabled and post-stroke gait

PDF

Knowledge‐driven decision support for assessing radiation therapy dose constraints

PDF

Using biomechanics to understand how an individual accomplishes a task in a variety of contexts

PDF

Pathomechanics of femoroacetabular impingement syndrome: utilizing subject-specific modeling approaches to investigate the influence of hip joint morphology and neuromuscular control

PDF

Engineering scalable two- and three-dimensional striated muscle microtissues for human disease modeling

PDF

Provable reinforcement learning for constrained and multi-agent control systems

PDF

Development of an integrated biomechanics informatics system (IBIS) with knowledge discovery and decision support tools based on imaging informatics methodology

PDF

The many ways of growing old: aging, activity patterns, and function in modern foragers

PDF

Modeling and vibration analysis of wheelchair-users

PDF

New approaches in modeling and control of dynamical systems

PDF

Investigation of preclinical testing methods for total ankle replacements

PDF

Control of two-wheel mobile platform with application to power wheelchairs

PDF

Functional brain correlates for premovement planning and compensatory adjustments in rapid aimed movement

PDF

The effects of jump strategy modification on energetics during dance jumps

Asset Metadata

Creator Peterson, Travis Jeffrey (author)

Core Title Lower extremity control and dynamics during the golf swing

School Viterbi School of Engineering

Degree Doctor of Philosophy

Degree Program Biomedical Engineering

Degree Conferral Date 2017-08

Publication Date 07/24/2017

Defense Date 05/31/2017

Publisher University of Southern California. Libraries (digital)

Tag angular impulse,biomechanics,golf,linear impulse,multi-joint control,neuromuscular control,OAI-PMH Harvest,reaction force

Format theses (aat)

Language English

Contributor Electronically uploaded by the author (provenance)

Advisor McNitt-Gray, Jill L. (committee chair), D'Argenio, David (committee member), Requejo, Philip (committee member), Wilcox, Rand R. (committee member)

Creator Email travis.peterson.3@gmail.com,travispe@usc.edu

Permanent Link (DOI) https://doi.org/10.25549/usctheses-oUC11256035

Unique identifier UC11256035

Identifier etd-PetersonTr-5602.pdf (filename)

Legacy Identifier etd-PetersonTr-5602

Dmrecord 411005

Document Type Dissertation

Format theses (aat)

Rights Peterson, Travis Jeffrey

Internet Media Type application/pdf

Type texts

Source University of Southern California (contributing entity), University of Southern California Dissertations and Theses (collection)

Access Conditions The author retains rights to his/her dissertation, thesis or other graduate work according to U.S. copyright law. Electronic access is being provided by the USC Libraries in agreement with the author, as the original true and official version of the work, but does not grant the reader permission to use the work if the desired use is covered by copyright. It is the author, as rights holder, who must provide use permission if such use is covered by copyright. The original signature page accompanying the original submission of the work to the USC Libraries is retained by the USC Libraries and a copy of it may be obtained by authorized requesters contacting the repository e-mail address given.

Repository Name University of Southern California Digital Library

Repository Location USC Digital Library, University of Southern California, University Park Campus MC 2810, 3434 South Grand Avenue, 2nd Floor, Los Angeles, California 90089-2810, USA

Repository Email cisadmin@lib.usc.edu