The golf swing, performed to project the golf ball into flight, is a complex dynamic movement. The degree to which the force produced by the body musculature is transferred to the ball depends upon the reaction force from the ground against the feet (Carlsoo, 1967). Traction at the golfer's feet has traditionally been achieved by using relatively long 6 or 8 mm metal spikes (Figure 1), which fix into the outer sole of the shoe in order to indent or penetrate the turf of the golf course. However, traditional metal spikes compress and grip grass roots, grass or soil, with the likelihood of creating spike marks. The increased popularity of golf has resulted in more surface damage to courses and putting greens (Hammond and Baker, 2002), and caused the wearing of traditional metal spike golf shoes on many courses to be forbidden or restricted. Such bans resulted in the development of alternative spikes for golf shoes, such as shown in Figure 2. Alternative-spike traction is provided by surface protrusions (or cleats) that penetrate only several millimetres into the turf, and were thought by Slavin and Williams, 1995 to be associated with a higher probability of slip. Some shoemakers incorporate additional raised mouldings of various designs between the spike positions to increase outer shoe sole traction during the swing in order to reduce the chance of the foot slip. The modern alternative spike sole design relies on surface area and form to grip the ground rather than long spikes. With the potential for injury and performance impairment arising from an accidental slip, in such a fast, powerful movement as the golf swing, there is the need for realistic, ecologically relevant testing to reassure and inform players, coaches and designers of the performance characteristics of modern golf shoe designs.

In golf play, the better golfer is given a lower handicap which reflects their golfing ability. The expectations and demands of the low handicap golfer from footwear are likely to be different to those of a high handicap player. The low handicap player is likely to perform fewer golf swings but the needs for effective power generation, control and accuracy are paramount. The game of golf also involves the selection of a number of different types of club, designed to aid particular forms of ball flight, in order to reach a hole with the lowest number of shots. The club used for longest shots is the driver and it has the longest shaft and largest club head. The 3-iron and 7-iron are metal irons, which have different lofts to assist ball projectile flight. The 3-iron is often the longest iron carried for long fairway shots. The 7-iron has a shorter shaft and is used for medium to short approach shots. Knowledge of how these key game factors interact during the performance of the golf swing, particularly in relation to the torque generation at the modern golf shoe outer sole to natural turf interface, would aid in questions arising concerning golf shoe selection and design. Laboratory based human studies conducted on artificial turf surfaces have underpinned the scientific knowledge (Barrentine et al., 1994; Dillman and Lange, 1994; Hume et al., 2005; Williams and Cavanagh, 1983). This has inherent limitations because of the induced change in the mechanism of outer shoe sole spike penetration and compression. Williams and Sih, 1998, who investigated ground reaction forces when wearing different types of golf shoes on artificial turf, highlighted the need for shoe assessments to be made on a natural grass surface. Worsfold et al., 2007 have reported greater vertical ground reaction force in spiked golf shoes at the shoe-natural turf interface when irons were used in comparison to a driver. It was also noted that greater mediolateral forces occurred across each foot when the driver was used in comparison to the irons. Use of the longer driver club to achieve powerful distance shots involves the generation of more rotational force, and this is likely to impact and interact with maintenance of an adequate support base. Further research is required to confirm the relevance of rotational force (torque) at the shoe-natural grass interface during performance of the golf swing.

The aim of this research was to measure the maximal torque and torque generated at the feet when a driver and two shorter iron clubs were used to perform golf swings on natural grass turf. Also, to identify any differences in torque associated with golfing ability (handicap) or the traction provided by different golf shoe outer sole designs.

Shoes
A traditional golf shoe with 8 mm metal spikes, an alternative spike golf shoe, and a flat soled shoe were assessed. All shoes had leather uppers that were identical in the cases of the metal spike and alternative design shoes. The metal spike shoe (Figure 3) had an ethylene vinyl acetate (EVA) mid-sole, thermoplastic urethane outsole (Adidas Stripe Tournament) and was fitted with 7 Fast Twist™ 8 mm metal spikes. The alternative spike shoe (Figure 4) had an EVA mid-sole and dual density injection moulded polyurethane (TPU) outsole (Adidas Z-Traction Tour) fitted with 7 Fast Twist™ alternative Adidas 5 mm spikes. The flat-soled golf shoe (Figure 5) had an EVA mid-sole, Stilo adapted flat sole and was not fitted with any spikes to provide additional traction. All shoes were new for the research, and a range of sizes was available.

Procedure
Golfers used their own driver, 3-iron and 7-iron. Five shots were performed with each club from a rubber Astroturf mat, into which a tee could be inserted. Titleist white DT golf balls were used and testing took place in an outdoor field setting in good sunny weather. Golfers were asked to play straight shots as normal with each club, and not to draw or fade the ball. The outcome of the each golf shot, whether straight (±8° approximately) or miss-hit was noted. A random testing order was used for both shoe and club type. Golfers adopted their natural golf swing stance with one foot on each force platform. The ground reaction forces of the right and left feet were measured simultaneously on two Kistler 9851 force platforms. The platforms were covered in a natural grass turf surface, similar to that found on a teeing off area on a golf course. The 30 mm deep turf was adhered using a thin slip of clay to smooth plastic plates (Janaway and Dyson, 2000; Smith et al., 2002; 2004; 2006). The plates were screwed onto the top of each force platform and a 35 mm vertical offset applied to account for the plate and turf depth. Force data were passed to two Kistler 9865 amplifiers, which were connected to an Amplicon 12-bit analogue to digital converter. Kistler Bio-Ware 3.1 software controlled data sampling at 1000 Hz and recording of the data to hard disk for storage and subsequent analysis. A 200Hz Peak Performance Technologies camera was placed in front of the golfer to capture whole body and club movement with recording on a Panasonic AG-MD830 video recorder. Posterior lower leg and foot movements were captured with a JVC Compact VHS GR-FX 12EK 50Hz video camcorder to aid in subsequent qualitative analysis to preserve data integrity.

Data analysis
Separate analyses were carried out for torque at the front foot and back foot. In the case of a right handed golfer the front foot would be the left foot and the back foot the right foot. This investigation of the ability of the shoe designs to enable torque generation at the shoe-turf interface during the changes in direction of the swinging club in the frontal plane (from anticlockwise backswing to clockwise downswing and follow-through) revealed associated directional changes in torque generation at the feet (Figure 6). The ability to maintain torque generation at the feet, while the swinging club changes direction from backswing to downswing and follow-through, is critical to the golf swing action and the avoidance of possible foot slip, injury or performance impairment. Hence in this analysis the amount of torque generated (from maximum to minimum torque values to reflect the directional change in torque at the foot interface) was measured in addition to the maximal torque (Tzmax).

Data were checked for its integrity and plotted; a Mauchley's test was used to test sphericity, homogeneity of variance was tested using a Levene's test and a Kolmogorov-Smirnov test was used to assess the normality of the distribution of scores. A three-way ANOVA with repeated measures (Club (driver, 3-iron, 7-iron)*Handicap (low, medium, high) *Shoe (metal, alternative, flat)) was used to identify any significant main effects, and significant differences were identified using a Post Hoc Tukey HSD test (alpha<0.05).

The Mauchley's test of sphericity; Levene's test of homogeneity and Kolmogorov-Smirnov test of normality all revealed non significant findings. Straight shots were achieved in 89%, 71% and 46% of cases by the low, medium and high handicap players respectively. The golf shot outcomes were similar for the low, medium and high handicappers in both the metal spike shoe (87%; 76%; 54% respectively) and in the golf shoe with alternative spikes (85%; 74%; 54% respectively).

The Tzmax recorded (Figure 7) at the back foot in the spiked golf shoes when using the driver and 3-iron was around 7 Nm and slightly less with the shorter 7-iron (6 Nm) for the 24 golfers. At the front foot mean Tzmax in both spiked shoes was nearly 20 Nm when using a driver, but this was only slightly more than when the irons were used (17-18 Nm). For neither back or front foot were the Tzmaz values statistically different when the three types of club were used, and no significant difference was identified for handicap group. Overall the front foot Tzmax on grass turf was 2.7 times greater (16.6-19.7 Nm) than the back foot Tzmax (5.7-7.8 Nm).

Figure 8 illustrates that for all 24 golfers the torque generated at the front foot, whichever shoe was worn, was similar (38-43 Nm) when the three different clubs were used. However, for each club the front foot torque generated was much greater than the back foot torque which ranged from 10.7-16.0 Nm. Post hoc analysis indicated that for the driver, the metal spike shoe torque generation at the back foot was significantly more than when the flat sole shoe was worn (p < 0.01, Figure 8). This supports the concept that metal spikes provide additional traction.

Consideration of handicap level in relation to torque generation (Table 1), and statistical analysis indicated that handicap did not moderate front foot torque generation. However a trend was evident at the back foot. For each handicap group the torque generated with the driver was greater than the torque generated if the 3-iron or 7-iron was used. Post-hoc analyses revealed that back foot torque generation with the driver by low handicap golfers was greater than torque generated by either the medium or high handicap players, with no difference in torque generated wearing different shoes (Table 1).

• Shoe to natural turf torque generation is an important component in performing a golf swing with a driver club.
• Torque at the shoe to natural turf interface was similar at the front foot when using a driver, 3-iron and 7-iron with Tzmax (17-19 Nm approx) and torque generation in the entire backswing and downswing of 40 Nm.
• Torque at the back foot was less than at the front foot when using the driver, 3-iron and 7-iron; Tzmax was 6-7 Nm, and torque generation 10-16 Nm with a trend to be greater when the driver was used.
• Low handicap golfers generated significantly more torque at the back foot than the medium or high handicappers (P<0.05) when using a driver.
• The metal spike shoe on natural turf allowed significantly more torque generation at the back foot than a flat-soled golf shoe when using a driver. Results have implications for golf shoe design.