1. Introduction
Golf has become an increasingly popular sport worldwide and is viewed as an enjoyable and practical means of staying active throughout an individual’s lifetime. Previous studies have demonstrated that playing golf provides an adequate amount of physical activity to improve overall health and well-being, especially for elderly golfers whose physiological training threshold is lowered by age [
1,
2]. While golf can provide some benefits for general health and fitness, the sport also appears to have particular risks of injury that may significantly affect players’ enjoyment of the activity [
3].
Along with its popularity, the injury rate for golfers has indeed increased steadily over the years, with golf-related low-back pain (LBP) being the most common injury [
4,
5,
6]. There are several factors that may contribute to LBP, including poor swing mechanics and fatigue due to overuse [
7]. Rapidly swinging a club is a crucial part of the golf game. In order to create a potential advantage at the beginning of the competition, golfers tend to dedicate considerable practice time for swings each day to generate a fast and powerful swing. Poor swing mechanics combined with the overuse problem may ultimately increase the risk of LBP for golfers. Furthermore, along with improper swing mechanics, sub-optimal physical fitness could produce considerable or abnormal forces localized in the lumbar region [
8]. This can cause significant muscle spasms due to back muscle strain or spinal ligament sprain, which usually leads to the development of LBP.
To identify physical deficiencies that are critical to the golf swing and injury prevention, the world’s leading golf education organization, Titleist Performance Institute (TPI), has developed a golf-specific physical screening system similar to the Functional Movement Screen (FMSTM). The TPI has identified some inappropriate swing mechanics that they categorize as “swing faults” to help coaches and golfers better understand swing mechanics and improve their game.
The FMSTM is comprised of seven fundamental movement patterns (tests) that require a balance of mobility and stability [
9]. The overhead squat (OHS) is one such test that the FMSTM uses to assess bilateral, symmetrical, and functional mobility of the hips, knees, ankles, shoulders, and thoracic spine, as well as the stability and motor control of core musculature [
10]. The TPI also uses the OHS test as one of its movement screens to assess golfers’ strength, flexibility, and balance [
11]. An individual with restrictions on spine mobility, hip mobility, or core motor function may fail the OHS test.
Studies on OHS performance have identified some swing faults that are documented by TPI [
11]. One of the most common swing faults among amateur golfers is known as “loss of posture” [
12], where the golfer has changed the knee flexion angle, trunk flexion angle, or head position between their address posture and impact position [
11]. Another common swing fault is “slide,” which is an excessive lateral shift of the hips toward the target on the downswing. Gulgin and his colleagues found that golfers with low overhead squat ability were two to three times more likely to exhibit early hip extension, loss of posture, or slide during the swing in comparison to golfers who could correctly perform an OHS [
11]. They further suggested that common swing faults are linked to inconsistent ball striking and reduced performance [
11]. Speariett and Armstrong found that the overhead squat is one of the most difficult tests for amateur golfers to perform, so much so that participants who were unable to perform the overhead squat most commonly presented with loss of posture (90%) and slide (80%) [
12].
The mechanics of the spine during a golf swing in golfers with or without LBP has been well established. Compared to healthy golfers, those with LBP may generate more lateral bending accompanied by flexion of the spine during the downswing phase [
13,
14]. Fortunately, professional golfers possess the capability to minimize the recurrence of injuries through technical adjustments [
15,
16]. Grimshaw and Burden reported the successful elimination of golf-related LBP in professional golfers, partly from reducing the amount of trunk flexion and by adopting a side-bend during the downswing [
15]. The side-bend with trunk flexion can limit the amount of trunk rotation available during the golf swing and may apply more shear force to the spine, thus increasing the risk of injury [
17]. The physical requirements of the golf swing may be similar to that of the OHS. Likewise, limitations in the mobility of the hip and spine or weaknesses of the core muscles may cause golfers to compensate through the loss of posture and slide.
Golfers with LBP have shown lower hip and spine mobility and delaying core muscle activation compared to healthy golfers [
1,
18,
19]. Considering that both the golf swing and OHS demand normal function of the core, lower limb, and shoulder mobility in three-dimensional space, the FMS over-head squat which also evaluate core, lower limb, and shoulder mobility maybe a useful test that can assess all of these elements simultaneously. It may therefore be possible to prevent LBP by using the OHS as a test to assess players’ spinal biomechanics during the golf downswing. To the best of our knowledge, no study to date has investigated the impact of overhead squat abilities on lumbar spine flexion and lateral bending biomechanical variables of golf swing.
To understand the mechanisms of LBP, scientists have developed different methodologies that measure the lumbar joint loads during the golf swing. For example, Hosea and his colleagues found that lumbar spine shear force during the golf swing was 80 percent greater in amateurs than in professionals, where the compressive force for both groups was more than eight times body weight (BW) [
20]. However, the different methodologies used in these studies may have affected the results obtained. For example, Lim and his colleagues and Hosea found that shear loads were about 1.6 BW to 0.6 BW while the peak compressive loads were greater than 8 BW [
20,
21]. Although both studies found the lumbar joint loads to increase continuously during the downswing phase there was less agreement in the lumbar shear loads. To address these conflicting results, the current study used a computer modelling and simulation approach to calculate lumbar joint forces during the downswing.
The primary purpose of this study was to determine differences in lumbar spine kinematics and joints loads during the downswing between golfers who execute a proper OHS and those who do not. A secondary aim was to investigate whether the ability of a golfer to perform the OHS test is related to their golf swing performance. We hypothesized that golfers who can perform the overhead squat properly would produce smaller lumbar spine’s joint loads, joint angular displacements, and joint angular velocities on L1-L2, L2-L3, L3-L4, L4-L5, L5-S1 joints during the downswing, and hence better performance compared to golfers who could not complete the OHS test.
2. Materials and Methods
2.1. Participants
Twenty-one right-handed golfers aged 18 to 30 years volunteered to participate in this study. All participants were free of any musculoskeletal injuries or disease that would have prevented them from performing their normal golf swing motion or impeded their ability to participate in the overhead deep squat screen. The study was conducted with ethics approval from the Human Research Ethics Committee of the local institution, and participants provided their written informed consent prior to commencement of testing.
2.2. Experimental protocol
All experiments were conducted in the Biomechanics Laboratory at National Taiwan Sports University. On arrival, each participant was informed of the study’s purpose and the experimental protocol. Testing was divided into two parts. Each golfer’s overhead squat performance was first measured using the FMSTM kit, after which a biomechanical evaluation of the golf swing was performed using three-dimensional video motion capture techniques.
2.3. Overhead Squat Test
The verbal test instruction of the overhead squat test was based on the description given by Cook [
9], where one certified FMSTM instructor executes the test. each participant wore their own sneakers during the test, and they assumed the starting position by placing their feet approximately shoulder-width apart and aligned in the sagittal plane. They then gripped a dowel while maintaining their elbows bent at 90° with the dowel overhead. Next, the dowel was pressed overhead by abducting both shoulders and extending the elbows. Participants were then instructed to squat as far as possible while maintaining both heels in contact with the floor and keeping the dowel directly overhead. Each participant was allowed up to three trials to perform the test successfully. Scoring criteria for the OHS (see
Table 1) were used to divide all participants into a high scoring group (HS: 3 points,
Figure 1a) and a low scoring group (LS: 2 points or 1 point,
Figure 1b&
Figure 1c) for further analysis.
2.4. Kinematic and Kinetic Data Collection
A 3D motion analysis system (Motion Analysis Corporation, Santa Rosa, USA) with 11 Eagle Digital high-speed cameras sampling at 250 Hz was used to record kinematic data as each participant performed a golf swing using a driver. Forty-nine retro-reflective markers (diameter, 10-12 mm) were placed on anatomical landmarks bilaterally – front of the head, rear head, cervical 7, thoracic 10, acromion, upper arm, elbow lateral, radius, ulna, third metacarpophalangeal joint, anterior superior iliac spine, posterior superior iliac spine, thigh, knee, shank, ankle, ankle med, toe, and heel (
Figure 2). Two force plates (Kistler 9281CA; Kistler Instrument AG, Winterthur, Switzerland), each sampling at 2000 Hz, were used to collect ground reaction force data from the lead and trailing legs during the swing. The force data are synchronized with the motion analysis system. The kinematic and ground force data were filtered using a fourth-order low-pass Butterworth filter at a cut-off frequency of 14 Hz. The kinematic and kinetic data were used as input to a musculoskeletal modeling pipeline available in OpenSim to calculate lumbar spine kinematics and joint loading [
22].
For the lower lumbar region of the longissimus thoracic, a pair of wireless sEMG sensors (Trigno, Delsys Inc., Natick, MA, USA) were placed on the interspace between L1 and L2 on both sides. The EMG and kinematic data were synchronized using a video camera connected to the EMG system, which recorded the golf swing and displayed the images in real time along with the EMG signals in the Delsys EMGworks software (Delsys Inc., Natick, MA, USA). Electromyographic data were filtered (six-pole Butterworth and bandpass filtered 25-500 Hz) and full-wave rectified using signal processing software (EMGworks Analysis software).
Each participant was given 5 minutes to warm up prior to data collection. The participant was then instructed to stand on the force plates and perform a maximal swing using the driver. Data were collected for 5 trials. The TrackMan (TrackMan IIIe, Vedbaek, Denmark) Doppler radar system was placed behind the ball striking area to measure the speed of the ball after impact.
2.5. Computer simulation
The full-body lumbar spine model (FBLS,
https://simtk.org/home/fullbodylumbar) comprising 21 segments, 30 degrees of freedom, and 324 musculotendon actuators was used to simulate each golf swing [
23]. Before the motion capture, the marker setting of FBLS model was modified for scale, so the model and motion capture data could be matched to fit. All data were converted to a useable format, the generic musculoskeletal model was scaled to match each participant’s body anthropometry [
22]. The inverse kinematics routine in OpenSim was then used to minimize differences between the positions of skin markers on the participants and virtual markers on the model. This procedure was undertaken in order to achieve a dynamically consistent set of kinematics and kinetics that best matched the experimentally collected data [
22].
To investigate the primary aim of this study, results of inverse kinematics were used to derive the lumbar joint angle during impact in the sagittal and frontal plane, peak angular velocity, and angular displacement during the downswing phase in the sagittal and frontal plan.
Next, static optimization (SO) was performed to resolve the net joint moments into individual muscle forces at each instant in time. Finally, the joint reaction analysis tool was used to calculate the internal vertebral joint loads [
24]. Lumbar spinal loading was calculated by solving the dynamical equations of motion with the input of muscle forces, gravity, and inertia. Moreover, to attenuate the noise contained within the raw marker data, a filtering process was applied during static optimization, using a low-pass sixth-order Butterworth digital filter at a cut-off frequency of 14 Hz, which was determined based on residual analysis. All loads reported for a given vertebra were those acting upon it from the inferior vertebra. For example, the L5-S1 loads reported are those from S1 acting on L5. The force was calculated using the Newton’s 2nd Law:
where
is the force applied by the S1 vertebra to the L5 vertebra,
is the matrix of inertial properties of the L5 vertebra,
is a vector of angular and linear accelerations of the L5 vertebra,
is the force applied by the L4 vertebra to the L5 vertebra, and
and
are muscle forces and gravitational forces acting on the L5 vertebra. The L5-S1 compressive force was calculated as the component of
parallel to the longitudinal axis of the L5 vertebra, which was used for all subsequent analyses. The L5-S1 shear force was calculated the same way, but parallel to the anteroposterior axis of the L5 [
24]. Peak shear and compressive forces acting at each lumbar spine joint, specifically, L1-L2, L2-L3, L3-L4, L4-5, and L5-S1, were calculated and used in the statistical analyses described below.
Finally, model simulations were validated by comparing the muscle activations calculated in the model against EMG data measured during the golf swing. The EMG data were normalized by the peak activation measured during the swing phase and was compared to simulated muscle activations, which were defined between 0 and 1. We compared the average activation of the longissimus thoracic muscle of 4 subjects to the corresponding EMG (
Figure 3).
2.6. Statistical analyses
Descriptive statistics were analyzed to assess means and standard deviations between the low scoring group (LS) and high scoring group (HS). A Pearson’s chi-square test was used to compare gender distributions, and an independent t-test was used to determine significant differences in demographic and performance data between different groups. An independent t-test was used to examine differences in all lumbar kinematics and joint loads during the downswing phase of the golf swing between the LS and HS groups. The Pearson correlation is used to tested between the measured EMG and simulated activation level through time series for each muscle to validate the model. Statistical significance was set at p < 0.05, and SPSS 20.0 (SPSS, Chicago, IL) statistical software was used for all data analysis.
5. Conclusions
We presented the difference in lumbar spine kinematics and joint loads during the golf downswing between golfers with different overhead squat abilities. Golfers with better performance in the overhead squat test demonstrated significantly greater angular extension displacement in the sagittal plane, lower lumbar extension angular velocity, and lower L4-S1 joint shear force compared to golfers with lower performance in the overhead squat test. Due to the requirements of performing the overhead squat, better performance in this test also reflects an advantage in hip and spine flexibility and core stability, which are associated with swing mechanics and risk of LBP. The study’s findings therefore suggest that the overhead squat test can be a useful index in assessing the lumbar kinematics and joint loading patterns during the downswing, and provides a training guide reference to reduce the risk of a golf-related low back injury.
There are several limitations of the current study. First, while different club types could affect lumbar swing kinematics [
34,
43] and hence the loads applied to the lumbar spine during the downswing phase. Further study is needed to investigate the relationship between the overhead squat ability and lumbar loads during the downswing when an iron club is used. Second, the highest L4-L5 shear loads were found after ball impact [
21]. However, the joint loads are also applied to the lumbar spine during the follow-through, and the end of the follow-through is also considered a critical element related to LBP in the golf swing [
44]. Future work should examine lumbar kinematic and kinetic variables in the follow-through phase of the golf swing in order to extend the impact of the overhead squat ability on the lumbar joint loading.
Third, the current study only examined lumbar kinematics and joint loads in the sagittal and frontal planes. Axial rotation of the lumbar spine is associated with ball speed, although an over-rotated lumbar spine might result in excess loads that relate to LBP [
45,
46]. Rapid spinal rotation during the golf swing, combined with physical limitations, may play a role in golf-specific injuries [
47]. Hence, additional research may prove beneficial for LBP prevention by investigating the difference in lumbar rotation kinematics between golfers with different overhead squat abilities. Finally, participants in the current study were healthy golfers free from any injuries that may have prevented them from performing the golf swing. To better understand the chronic impact of overhead squat ability on the risk of LBP during the golf swing, the next logical step is to investigate the difference in low back injury rates between HS and LS over time. In this way, the effect of the overhead squat test on LBP prevention could be studied.