Weston, Matthew; Coleman, Neil J.; Spears, Iain R.
The core refers to the musculature of the shoulder stabilizers, trunk, and the upper leg muscles (22). A major role of the core musculature is to provide dynamic stiffness for the central joints of the body and particularly the spinal joints. The tension created by coordinated core muscular actions induces controllable stiffness of the spine via axial compression (24) and the ability to perform this is often called core stability (21). It is commonly held that a stable core will increase the efficiency of movement (30). Accordingly, there is often an assumption in the sports sciences that core exercises lead to performance improvements. Consequently, core muscles are the target of many strength and conditioning programs (12,29).
During the classic golf swing, the shoulder–hip complex can reach more than 45° rotation. During the modern swing in which emphasis is placed on creating axial rotations of the upper torso with respect to the pelvis, these trunk rotations can be even higher (10). The amount of relative rotation is strongly correlated with both ball velocity (27) and playing standard (36) and is considered by many as a desirable feature of golf swing kinematics. During this swing, the gluteus maximus has been found to be heavily involved in hip stabilization, and the erector spinae muscle group is involved with counteracting gravity (26,35). The abdominal muscles are also very active during the forward stages of the golf swing (28,35). Core training, made up of a series of exercises targeting a range of muscles including the abdominals, hip abductors/adductors, hip flexors, and lumbar spine extensors, is often advocated for golfers (23). To date, there have been several studies that have demonstrated improved golf swing performance as a result of exercise programs that comprise both core and swing-specific exercises. However, their experimental designs are such that it is difficult to isolate the benefits of the core exercises. The benefits observed could be due predominantly to the core exercises or, as adversaries of core training would argue, due to the swing-specific nature of the exercises (21). As a case in point, the effect of core training and resisted golf swings on golf swing parameters were quantified and considerable improvements were found (23), but it is not possible to determine whether the benefits were acquired due to the loads on the core or due to the swing-specific movements of the upper limb.
In a sport that is played by more than 25 million people worldwide, lower back pain accounts for 25%–76% of all golf-related injuries (11). The biomechanics of the swing is widely considered to be a major source of the problem (10). The large rotations of the trunk relative to the pelvis bring the vertebrae close to their extremes of motion, resulting in the stretching of the surrounding viscoelastic soft tissues. Although this process is believed to contribute to the power of the swing, it is also suggested to contribute to spinal deterioration (10). The swing also involves considerable lateral tilt, which is suggested to contribute to long-term problems in the intervertebral disks, particularly when combined with lumbar flexion (10). On the basis of the accumulated load theory (20), it is possible that swing-specific exercises come with a long-term cost, and thus reducing the swing-specific component of a training program may be one strategy to reduce the risk of back problems. A potential alternative core training, which has formed a component of these previously successful golf training interventions (23), is not without its critics (21) and is not proven to provide performance benefits to golfers (29). The aim of this study is to quantify the effect of an 8-wk isolated core training program on selected ball and club parameters during the golf swing and also the variability, a measure of swing consistency, of these measures.
Thirty-six male golfers (180.8 ± 6.8 cm, 89 ± 15 kg, 47 ± 12 yr) participated in the study. The participants were members of Dinsdale Spa Golf Club (Darlington, UK), and all held official club handicaps of ranging standards (Table 1). Participants were recruited through the completion of a questionnaire within the professional shop. All participants completed an informed consent form, and ethical clearance was granted by the Teesside University Ethics Committee. Participants were randomized 1:1 to the exercise intervention group (n = 18) or to the control group (n = 18). Those assigned to the exercise group completed a medical questionnaire and then completed an 8-wk core training exercise program. Both groups were instructed to continue with their normal levels of physical activity that included playing golf.
To maximize ecological validity, we conducted the testing sessions in the naturalistic setting of the participants’ golf club using a portable launch monitor (Vector Pro Launch Monitor VPR200; AccuSport, Inc., Winston-Salem, NC). The system captures multiple exposures of the ball immediately after impact. The frames are used to calculate the linear and the angular components of balls launch velocity and also the tangential distance of the center of gravity of the club head relative to the point of impact. The software outputs various parameters associated with the golf swing performance. It was chosen to report club-head speed, ball backspin, and ball sidespin. The spin rates are calculated from the images. The club-head speed is estimated using a four-step process based on the law of conservation of momentum. Specifically, the tangential component of momentum is calculated from the calculated ball spin rates and known inertial characteristics of the ball. The normal component of momentum of the club head is derived from the linear momentum of the ball and combined with the tangential components to determine the velocity vector (and therefore speed) of the club head. However, on the basis of the fundamental laws of physics, there are nonetheless two assumptions being made in the estimates of club-head speed; that is, the coefficient of restitution between the ball and the club and the inertial characteristics of the club are known. Unfortunately, these parameters are not given; thus, to minimize the effect of these assumptions on the estimates, individuals used the same club for all testing sessions, and the same ball was used throughout. The participants performed the golf shots on a level practice mat using their own five iron. During each testing session, club-head speed and ball spin rates were collected for 10 shots, and the mean of the participants’ 10 shots was chosen as the summary measure of performance for each test. Approximately 20 min after completion of the swing trials, an isometric flexor endurance (25) was performed. Briefly, this test involves placing the upper body against a purpose-built support at an angle of 60° with respect to the horizontal. The support was then removed, and the participant attempted to maintain a static position for as long as possible. Isometric flexor endurance is measured by the time elapsed before the trunk falls lower than 60°. Furthermore, given the similarity of this test and some of the exercises performed during the intervention, it is expected that changes in performance reflect adherence to the exercise program. In total, the participants completed three testing sessions, two at baseline (separated by 7 d) and one within 4 d after the end of the exercise intervention (8 wk later). Performance recorded during the second pretest trial acted as the pretest score for each of the participants.
The reliability of the golf club-head speed, ball backspin, ball sidespin, and endurance test was examined using a test–retest experimental design before the intervention. All testing was performed at approximately the same time of day, and all 36 participants completed three testing sessions. Reliability was quantified using the test–retest correlation coefficient and the typical error (18). All four performance measures demonstrated good test–retest reliability with test–retest correlation coefficients of 0.84 (90% confidence limits [CL] = ±0.09), 0.72 (±0.14), 0.68 (±0.16), and 0.78 (±0.12) for club-head speed, backspin, sidespin, and endurance test, respectively. The values for test–retest reliability were slightly higher than those found for a driver club in recreational players (23), which is most probably due to different clubs being used. In this study, typical error (%) was 4.4% (±1.0%), 12.2% (±2.7%), and 30.3% (±7.2%) for club-head speed, backspin, and sidespin, respectively. The reliability of the flexor endurance was slightly less than previously reported (25) with our typical error being 32.8% (±7.8%). The typical error data provide valuable information for the interpretation of performance changes on an individual level.
A plethora of core strengthening and stability exercises has been reported in the literature. The eight core exercises to be used for this intervention were chosen on the basis of simplicity, avoidance of lateral bending of the vertebral column, and not requiring additional equipment. The exercise program was deliberately dissimilar from the golf swing but was designed to activate similar muscles groups to those involved in the swing. The eight core exercises chosen were double-leg squat, bent-leg curl up, superman, supine bridge, prone bridge, quadruped, lunge, and side bridge. The first three of which were adapted from previous EMG studies. The double-leg squat was performed with the feet shoulder width apart and with neutral spinal alignment (13). The bent leg curl up was performed with the arms folded across the chest, and head, shoulders, and upper back were raised off the floor (1). The superman exercise was performed in a prone position with neutral spine alignment, and the arms and the legs were fully extended and held above the floor (5). These three exercises have been shown to elicit high levels of muscle activity (i.e., >60%) in the multifidus, external obliques, and longissimus (1,5,13) and above the threshold for inducing gains in core strength (33) while in the other muscles of the core eliciting levels (10%–25% MVIC) for inducing gains in core stability (33). The latter five exercises are described in detail elsewhere (7). Supine bridge, side bridge, prone bridge, quadruped, and standing lunge were all performed with the spine in neutral alignment. The side bridge exercise has been suggested to strengthen the gluteus medius and the abdominal external oblique muscles, and the quadruped arm/lower extremity lift exercise may help strengthen the gluteus maximus muscle. The lunge elicits high levels of activity (>45% maximum voluntary isometric contraction [MVIC]) in the vastus medialis. The side bridge produces activity greater than 60% MVIC in the external oblique (13) and greater than 45% MVIC in the gluteus medius (7). The quadruped elicits high levels of activity (>45% MVIC) in the gluteus maximus. The lunge produces EMG levels greater than 45% MVIC in the vastus medialis. All the other exercises produce EMG levels less than 45% MVIC and are considered to be beneficial for training endurance or stabilization (7). The exercises were performed slowly through the range of motion with a 10-s hold. They are considered to be low risk in terms of the exerting rapid loads on the viscoelastic soft tissues of the spine or causing lateral bending. To minimize the learning effects, we provided all participants a familiarization session. During the 8-wk intervention period, the eight core exercises were repeated three times a week. Functional progression was incorporated after 4 wk by adding additional limb movements and lengthening the duration of the holding position from 10 to 15 s. The additional limb movements were as follows: arms raised during double-leg squat, heel touch during bent leg curl up, contralateral arms and leg raises during superman, leg raises during supine bridge, hip extension to prone bridge, contralateral arms and leg raises during quadruped, slow rotation of trunk when in lunge position, and hip abduction during side bridge.
Data are presented as the mean ± SD. Before analysis, all outcome measures were log transformed and then back transformed to obtain the percent difference, with uncertainty of the estimates expressed as 90% CL between the posttest and the pretest. This is the appropriate method for quantifying changes in athletic performance (17). We used mixed effects linear modeling (IBM SPSS version 21.0, Armonk, NY) to analyze the effect of the core stability training intervention on our four outcome measures. This method allows for and quantifies (as an SD) individual differences in response to the intervention, which are frequently highly variable. An ANCOVA method was adopted to compare the two groups, with the pretest score as a covariate to control for chance imbalance in our measures between the control and the intervention groups at baseline (34). Effects were evaluated for practical significance by prespecifying 0.2 between-subject SD as the smallest worthwhile effect (4). Inference was then based on the disposition of the confidence interval for the mean difference to the smallest worthwhile effect; the probability (percentage chances) that the true population difference between trials was substantially beneficial, harmful (>0.2 SD), or trivial was calculated as per the magnitude-based inference approach (2). These percentage chances were qualified via probabilistic terms and assigned using the following scale: <0.5%, most unlikely; 0.5%–5%, very unlikely; 5%–25%, unlikely; 25%–75%, possibly; 75%–95%, likely; 95%–99.5%, very likely; and >99.5%, most likely (16,17). Magnitude-based inferences were then categorized as clinical for all four outcome measures. The default probabilities for declaring an effect clinically beneficial are <0.5% (most unlikely) for harm and >25% (possibly) for benefit (16). A clinically unclear effect is therefore possibly beneficial (>25%) with an unacceptable risk of harm (>0.5%) (17). To evaluate the effectiveness of the core stability intervention on the variability of the participants’ selected swing parameters, we calculated the within-subject coefficients of variation (CV, %) for each performance measure (before and after). These data were then analyzed using the same mixed linear model previously described. In the absence of a known sampling distribution for a difference in SD, the 90% CL values for the mean differences were constructed using a bias corrected accelerated bootstrapping technique of 2000 samples with replacement from the original data. To interpret the magnitude of a CV, we doubled and assessed the adjusted between-group differences in CV against a scale of 0.2 (small), 0.6 (moderate), and 1.2 (large) of the between-subject SD of the pretest for each variable (31,17). Relations between the participants’ performance test scores and golfing handicap were examined using a Pearson’s product moment correlation, with 90% CL also presented. The following scale of magnitudes (17) was used to interpret the correlation coefficients as follows: <0.1, trivial; 0.1–0.3, small; 0.3–0.5, moderate; 0.5–0.7, large; 0.7–0.9, very large; and >0.9, nearly perfect. All reliability measures were calculated using a custom-made spreadsheet (18). Inferences were based on uncertainty in standardized magnitudes of effects.
Descriptive data for both study groups are displayed in Table 1. A large negative correlation was observed between golf handicap and preintervention club-head speed (r = −0.61; 90% CL ±0.18). A moderate negative correlation was observed between golf handicap and backspin (r = −0.41; ±0.20), with a small positive correlation between golf handicap and sidespin (r = 0.20; ±0.27). A trivial negative correlation was observed between golf handicap and core endurance test performance (r = −0.07; ±0.21).
The adjusted effect of the core training intervention (Table 2) was a likely small beneficial effect on golf club-head speed, with the SD of the individual responses being 1.7% (±4.3%). The effect on backspin and sidespin was unclear, with the SD of the individual responses being −8.2% (±17%) and −32% (±47%), respectively. There was a very likely small beneficial effect (possibly moderate) of the intervention on endurance test performance. The SD of the individual responses was −17% (±42%). Core training also made the golf swing more consistent (Table 3), as evidenced by a moderate decrease in percent variability of club-head speed and a small decrease in variability of backspin. There was a small increase in variability of sidespin.
Despite core training being fundamental to many exercise programs, very little is known about its isolated effect on sports performance. The aim of this study was to quantify the effect of an 8-wk isolated core training program on golf club-head speed and ball spin parameters. This is the first study to quantify these effects in golf. We observed a likely small beneficial effect of our exercise intervention on club-head speed. The exercise intervention also had a very likely small beneficial effect on core endurance. A further effect of the exercise intervention was a moderate to small reduction in the variability of the participants’ repeated club-head speed and backspin, with a small increase the variability of sidespin. Although the findings apply specifically to golf, it is possible that these benefits could transfer to other sports that require substantial asymmetrical movements of the spine (e.g., tennis, hockey).
Exercise interventions have previously been shown to increase club-head speed in the golf swing and these improvements in speed range from 0.5% to 6.3% (6,8,23,32). In this study, the improvement in club-head speed was 3.6%, which is comparable with those mentioned earlier. Furthermore, we observed a large association between golf club-head speed and golfing ability, as determined by the golfers’ handicap, lending support to the validity of club-head speed as a measure of golf swing performance. We found unclear effects on other measures of the golfing swing performance, namely, backspin and sidespin. However, these measures demonstrated only moderate and small associations with golfing ability. Although our exercise intervention was successful in improving core endurance, the translation of this to golf performance may be questionable given the trivial association between performance on this test and golfing handicap.
We also examined the effect of the exercise intervention on the variability, as determined by the CV of the participant’s 10 golf swings, of our performance measures. Postintervention decreases were observed for the variability of club-head speed, backspin, with an increase in sidespin variability. These results could lend some support for not just a cleaner strike of the ball but also that it is hit with greater consistency. Although the coefficients of variation for backspin and sidespin were high, suggesting that this variable may not be a stable indicator of physical performance (14), it is the relative change in this variable that is of importance in the present study for determining the effect of our intervention on the consistency of the repeated golf swings. Most coaches would agree that consistency is a desirable feature of the golf swing (19).
To reiterate, the core exercises used in the study were mostly isometric with the spine in a neutrally aligned orientation. Thus, the exercises are expected to place a lower and more evenly distributed stress on the spinal column than swing-specific exercises. On the basis of skill acquisition theory, the degree of transfer of these training effects is likely to be much lower (21). From this perspective the benefits shown in this study are somewhat unexpected. Specifically, the core training program is an effective strategy to improve some parameters associated with the golf swing. Unfortunately, the experiment was not designed to establish a causal relationship; thus, the benefits could be due to a range of factors (e.g., reduced co-contractions, improvements in neural pathways, or strength). Nonetheless, it is suggested that these benefits are achieved with a reduced cost to the spine when compared with swing-specific exercises. Such suggestions are somewhat controversial, given rising concerns related to core training and back pain (21) and also findings of a recent meta-analysis (29), which found only minimal performance benefits across a range of sports.
There are several limitations to our approach in terms of addressing the stated aim: first, in terms of experimental control, and second, in terms of the interpretative value of the results. With regard to the former, neither the quantity nor the quality of the exercises undertaken by the participants were directly monitored or controlled. Golf is an individual pursuit, and finding suitable times to hold regular organized exercise sessions was extremely difficult. Training logs were also considered, but these are subjective and known to be prone to participant bias (3). In terms of the quality of the exercise, limitations were also present. Specifically, training adaptations are known to occur when the level of muscle activity is between 10% and 25% maximum voluntary contraction (33), yet the techniques available to quantify muscle activity (i.e., quantify the quality of the exercise) are susceptible to issues of variability. Overcoming such issues is time consuming; thus, monitoring the quality of the exercises was not practical in this study. As a pragmatic alternative to monitoring the quantity and quality of the exercise, we measured what is considered to be an obvious training adaptation of the exercises that is core endurance. Specifically, the change in core endurance is expected to serve as a valid, albeit indirect, summary measure of the quality and quantity of the exercise being undertaken by the participants. The clear and substantial benefits in terms of the core endurance would indicate that adherence was good. The second limitation was with regard to the interpretative value of the results. It is recognized that the performance measures chosen do not necessarily translate into lower golfing scores, although we did find a large association between the estimates of golf club-head speed and golfing handicap, a direct measure of golfing ability. Club-head speed quantifies the movements at the distal end of the kinetic chain and is thus a summary measure of the biomechanical events (e.g., trunk lateral flexion, wrist flexion speed), leading up to the instant of ball impact. It is recognized that although this approach simplifies the task of statistical analysis, it also makes it difficult to isolate and discriminate the underlying causes of the changes in the swing in terms of biomechanical parameters. In addition, given the complexities of the biomechanics of the core, it would be naive to assume that all the exercises included in this intervention have contributed evenly, if at all. Our experiment was not designed to establish a causal relationship but rather to highlight whether core training in isolation is worth pursuing in terms of golf swing performance. Thus, taken together, there are several limitations and simplifications that reduce the experimental control and restrict the interpretative values of the results. Nonetheless, the experimental design is considered sufficiently robust to enable us to isolate the effects of core training on some measures of performance and thus address the stated aim.
The results of our study demonstrated a likely small beneficial effect of our practical and safe core training exercise intervention on club-head speed. As far as the authors are aware, this is the first study to quantify the effect of an isolated core training program on performance measures in golf. Furthermore, the exercise intervention also had a very likely small beneficial effect on the golfers’ core endurance, as measured by an isometric flexor endurance test. A further effect of our exercise intervention was a reduction in the variability of some measures of golfers’ repeated club-head speed and backspin, indicating a more consistent golf swing.
The authors acknowledge Gordon Cattrell, head professional at Middlesbrough Golf Club, for the loan of testing equipment. No funds were received for this work.
The authors have no conflict of interest to disclose related to this study.
The results of the present study do not constitute endorsement by the American College of Sports Medicine.
1. Axler CT, McGill SM. Low back loads over a variety of abdominal exercises: searching for the safest abdominal challenge. Med Sci Sports Exerc
. 1997; 29 (6): 804–11.
2. Batterham AM, Hopkins WG. Making meaningful inferences about magnitudes. Int J Sports Physiol Perform
. 2006; 1 (1): 50–7.
3. Borresen J, Lambert IM. The quantification of training load the training response and the effect on performance. Sports Med
. 2009; 39 (9): 779–95.
4. Cohen J. Statistical Power Analysis for the Behavioral Sciences
. 2nd ed. Hillsdale (NJ): Lawrence Erlbaum Associates; 1988. p. 567.
5. Comfort P, Pearson SJ, Mather D. An electromyographical comparison of trunk muscle activity during isometric trunk and dynamic strengthening exercises. J Strength Cond Res
. 2011; 25 (1): 149–54.
6. Doan BK, Newton RU, Kwon YH, Kraemer WJ. Effects of physical conditioning on intercollegiate golfer performance. J Strength Cond Res
. 2006; 20 (1): 62–72.
7. Ekstrom RA, Donatelli RA, Carp KC. Electromyographic analysis of core trunk, hip, and thigh muscles during 9 rehabilitation exercises. J Orthop Sports Phys Ther
. 2007; 37 (12): 754–62.
8. Fletcher IM, Hartwell M. Effect of an 8-week combined weights and plyometrics training program on golf drive performance. J Strength Cond Res
. 2004; 18 (1): 59–62.
9. Fradkin AJ, Sherman CA, Finch CF. How well does club head speed correlate with golf handicaps? J Sci Med Sport. 2004; 7 (4): 465–72.
10. Gluck GS, Bendo JA, Spivak JM. The lumbar spine and lower back pain in golf: a literature review of swing biomechanics and injury prevention. Spine J
. 2008; 8 (5): 778–88.
11. Gosheger G, Liem D, Ludwig K, Greshake O, Winkelmann W. Injuries and overuse syndromes in golf. Am J Sports Med
. 2003; 31 (3): 438–43.
12. Hibbs AE, Thompson KG, French D, Wrigley A, Spears I. Optimizing performance by improving core stability and core strength. Sports Med
. 2008; 38 (12): 995–1008.
13. Hibbs AE, Thompson KG, French DN, Hodgson D, Spears IR. Peak and average rectified EMG measures: which method of data reduction should be used for assessing core training exercises? J Electromyogr Kinesiol
. 2011; 21 (1): 102–11.
14. Hopkins WG, Hawley JA, Burke L. Design and analysis of research on sport performance enhancement. Med Sci Sports Exerc
. 1999; 31 (3): 472–85.
15. Hopkins WG. Measure of reliability in sports medicine and science. Sports Med
. 2000; 30 (1): 1–15.
16. Hopkins WG. A spreadsheet for deriving a confidence interval, mechanistic inference and clinical inference from a P
. 2007; 11: 16–20.
17. Hopkins WG, Marshall SW, Batterham AM, Hanin J. Progressive statistics for studies in sports medicine and exercise science. Med Sci Sports Exerc
. 2009; 41 (1): 3–13.
18. Hopkins WG. Reliability from consecutive pairs of trials [Excel spreadsheet]. In: A New View of Statistics
. Available from: Internet Society for Sport Science, sportsci.org/resource/stats/xrely.xls. 2000.
19. Hume PA, Keogh J, Reid D. The role of biomechanics in maximising distance and accuracy of golf shots. Sports Med
. 2005; 35 (5): 429–49.
20. Kumar S. Theories of musculoskeletal injury causation. Ergonomics
. 2001; 44 (1): 17–47.
21. Lederman E. The myth of core stability. J Bodyw Mov Ther
. 2010; 14 (1): 84–98.
22. Lehman GJ. Resistance training for performance and injury prevention in golf. J Can Chiropr Assoc
. 2006; 50 (1): 27–42.
23. Lephart SM, Smoliga JM, Myers JB, Sell TC, Tsai YS. An eight-week golf-specific exercise program improves physical characteristics swing mechanics and golf performance in recreational golfers. J Strength Cond Res
. 2007; 21 (3): 860–9.
24. McGill SM. Low back exercises: evidence for improving exercise regimens. Phys Ther
. 1998; 78 (7): 754–65.
25. McGill SM, Childs A, Liebenson C. Endurance times for low back stabilisation exercises: clinical targets for testing and training from a normal database. Arch Phys Med Rehabil
. 1999; 80 (8): 941–4.
26. McHardy A, Pollard H. Muscle activity during the golf swing. B J Sports Med
. 2005; 39 (11): 799–804.
27. Myers J, Lephart S, Tsai YS, Sell T, Smoliga J, Jolly J. The role of upper torso and pelvis rotation in driving performance during the golf swing. J Sports Sci
. 2008; 26 (2): 181–8.
28. Pink M, Perry J, Jobe FW. Electromyographic analysis of the trunk in golfers. Am J Sports Med
. 1993; 21 (3): 385–8.
29. Reed CA, Ford KR, Myer GD, Hewett TE. The effects of isolated and integrated core stability training on athletic performance measures: a systematic review. Sports Med
. 2012; 42 (8): 697–706.
30. Shinkle J, Nesser TW, Demchak TJ, McMannus DJ. Effect of core strength on the measure of power in the extremities. J Strength Cond Res
. 2012; 26 (2): 373–80.
31. Smith B, Hopkins WG. Variability and predictability of finals times of elite rowers. Med Sci Sports Exerc
. 2011; 43 (11): 2155–60.
32. Thompson CJ, Cobb KM, Blackwell J. Functional training improves club head speed and functional fitness in older golfers. J Strength Cond Res
. 2007; 21 (1): 131–7.
33. Vezina MJ, Hubley-Kozey CL. Muscle activation in therapeutic exercises to improve trunk stability. Arch Phys Med Rehabil
. 2000; 81 (10): 1370–9.
34. Vickers AJ, Altman DG. Statistics notes: analysing controlled trials with baseline and follow up measurements. BMJ
. 2001; 323 (7321): 1123–4.
35. Watkins RG, Uppal GS, Perry J, Pink M, Dinsay JM. Dynamic electromyographic analysis of trunk musculature in professional golfers. Am J Sports Med
. 1996; 24 (4): 535–8.
36. Zheng N, Barrentine SW, Fleisig GS, Andrews JR. Kinematic analysis of swing in pro and amateur golfers. Int J Sports Med
. 2008; 29 (6): 487–93.