Increased shoe sole hardness results in compensatory changes in the utilized coefficient of friction during walking

Gait & Posture 30 (2009) 303–306

Contents lists available at ScienceDirect

Gait & Posture journal homepage: www.elsevier.com/locate/gaitpost

Increased shoe sole hardness results in compensatory changes in the utilized coefficient of friction during walking Yi-Ju Tsai a, Christopher M. Powers b,* a b

Department of Recreation and Health Care Management, Chia Nan University of Pharmacy and Science, 60, Erh-Jen Road., Sec. 1, Jen-Te, Tainan, Taiwan Division of Biokinesiology and Physical Therapy, University of Southern California, 1540 E. Alcazar St., CHP-155, Los Angeles, CA 90089-9006, United States

A R T I C L E I N F O

A B S T R A C T

Article history: Received 30 June 2008 Received in revised form 21 May 2009 Accepted 25 May 2009

Based on mechanical testing, harder soled shoes have been shown to provide less available friction than soft soled shoes. Whether or not humans adjust their utilized coefficient of friction (COFu) and gait kinematics to accommodate the decrease in available friction while wearing hard soled shoes is not known. Fifty-six young adults participated in this study. Ground reaction forces, full body kinematics, stride characteristics and subjective perception of footwear slipperiness were recorded under both hard and soft soled shoe conditions. Paired t-tests were used to identify the differences between two shoes conditions. Results indicated that the peak COFu was significantly less when wearing the hard soled shoes compared to when wearing the soft soled shoes (0.23 vs. 0.26, P < 0.001). The decrease in peak COFu was the result of a decrease in the resultant shear forces at the time of peak COFu as no difference in the vertical ground reaction forces was observed. When wearing hard soled shoes, subjects demonstrated decreased total body center of mass (COM) acceleration prior to and immediately following initial contact, decreased walking velocity, shortened stride length, and reduced ankle dorsiflexion angle at initial contact. Taken together, we believe that these gait modifications represent behavioral adaptations to wearing shoes that are perceived to be more slippery. ß 2009 Elsevier B.V. All rights reserved.

Keywords: Gait Utilized coefficient of friction Center of mass Footwear Sole hardness

1. Introduction Slips have been recognized as a significant cause of falls and are one of the most common causes of occupational accidents [1,2]. Slip and fall injuries are a serious problem to both industry and society due to the enormous financial costs, which are estimated to be approximately 86 billion dollars in the year 2020 [3,4]. Slip and fall events represent a complex interaction of environmental and human factors [5,6]. Slips occur when the friction demand of an individual exceeds the friction available from the shoe/floor interface [7,8]. The utilized coefficient of friction (COFu) of a person can be measured from a force plate and is defined as the ratio of the shear and vertical ground reaction forces [9,10]. The available friction of shoe/floor interface can be estimated using a footwear slip resistance tester. Several footwear characteristics are thought to influence the available friction at the shoe/floor interface. These include shoe design, sole material, tread pattern, and heel geometry [11–13]. One major factor contributing to the slip resistance of the shoe/

* Corresponding author. Tel.: +1 323 442 1928; fax: +1 323 442 1515. E-mail address: [email protected] (C.M. Powers). 0966-6362/$ – see front matter ß 2009 Elsevier B.V. All rights reserved. doi:10.1016/j.gaitpost.2009.05.019

floor interface is the hardness of the sole material. Based on mechanical testing, harder soled shoes have been shown to provide less available friction than soft soled shoes [14,15]. This would imply that persons wearing harder soled shoes could be at greater risk for slip initiation. However, human ambulation studies have reported that the vertical ground reaction force pattern differs when wearing shoes of varying sole hardness. The loading rate of vertical force has been shown to increase when wearing harder soled shoes, [16,17] although the peak vertical force does not change [16,18]. An increase in the loading rate could have an influence on COFu as a rapid rise of the vertical force relative to the resultant shear force would result in a decrease in the peak COFu during weight acceptance. Evidence linking shoe sole hardness and peak COFu has been provided by Fendley and Medoff who demonstrated the peak COFu decreased when harder soled shoes were worn [19]. Their data however, were obtained from a single subject, and cannot be generalized to the population at large. Furthermore, whether or not the observed changes in peak COFu were the result of behavioral gait adaptations or simply were reflective of the harder sole material was not evaluated. The purpose of the current study was to quantify the influence of footwear sole hardness on peak COFu during walking. A secondary purpose of this study was to compare differences in gait

304

Y.-J. Tsai, C.M. Powers / Gait & Posture 30 (2009) 303–306

characteristics (i.e., altered lower extremity and/or total body center of mass kinematics) between shoe conditions that may explain the observed difference in peak COFu. Information obtained from this study is an important step in establishing whether or not humans make meaningful adjustments in their COFu and gait kinematics to accommodate the decrease in available friction while wearing hard soled shoes. 2. Methods 2.1. Subjects Fifty-six healthy adults (28 males and 28 females) between the ages of 23 and 38 years participated in this study (mean age: 27.7  3.4 years; height: 168.0  8.5 cm; weight: 64.4  11.4 kg). Subjects reported no orthopaedic or any medical condition that would adversely affect their gait. Each signed an informed consent approved by the Institutional Review Board of the University of Southern California. 2.2. Instrumentation Ground reaction forces were recorded using three force platforms (Model OR6-6-1000; Advanced Mechanical Technology Inc., Watertown, MA). Each was covered with smooth vinyl composition tile (similar to the rest of laboratory floor) and aligned in the middle of a 10 m walkway. Analog Force plate data were collected at 1560 Hz. Full body kinematics was obtained using an eight-camera motion analysis system (Vicon 612, Vicon Motion Systems, Lake Forest, CA). Kinematic data were sampled at 120 Hz. In order to quantify lower extremity and total body center of mass (COM) kinematics, 37 light reflective markers (25 mm spheres) were attached to bilateral upper and lower extremities and the trunk. Two pairs of commercially available Oxford style dress shoes (Bates Footwear Inc., Rockford, MI) that differed only in outsole hardness were provided for each subject. The material and appearance of the uppers of the two pairs were identical, and the outsole of each shoe was made from Styrene Butadiene Rubber. The sole hardness for the soft shoes was Shore 75A while the sole hardness for the hard shoes was Shore 54D. The outsole hardness was determined by American Society of Testing and Materials standard test method D2240-04 for Rubber Property-Durometer Hardness. The outsoles of both sets of shoes were smooth with no tread pattern. The available friction between the soft and hard shoe outsoles and laboratory floor surface (smooth vinyl composition tile) was 0.91 and 0.62, respectively as determined using the Shoes and Allied Trade Research Association Physical Test Method (SATRA PM144 slip resistance tester). This method determines the dynamic coefficient of friction between the footwear outsole and a given floor surface under conditions that simulate heel strike. 2.3. Procedures Subject height, weight, and anthropometric measures including leg length, hand thickness, elbow, wrist, ankle, and knee width were recorded for use in estimating joint centers necessary for kinematic calculations. Reflective markers were then attached bilaterally to the anterior temple, posterior lateral head, acromion process, upper arm, lateral epicondyle, medial and lateral wrist, third metacarpal phalangeal joint, anterior superior iliac spine, posterior superior iliac spine, lateral thigh, lateral femoral epicondyle, lateral malleolus, heel, dorsal second toe. Additional markers were attached to the 7th cervical vertebrae, 10th thoracic vertebrae, the right upper posterior trunk, the sterno-clavicular notch, and the inferior sternum.

Subjects were tested under two conditions: (1) hard soled shoes, and (2) soft soled shoes. The order of footwear conditions was randomized and subjects were blinded to the shoe condition being evaluated. All subjects were instructed to ambulate at a selfselected fast walking speed. To control for the potential influence of walking speed between shoe conditions, the speed of all walking trials was constrained to fall within 5% of the first acceptable trial. A walking trial was considered acceptable if the right foot fully contacted one of the three force plates without intentional targeting. For all subjects, three trials were recorded for each footwear condition. Kinematic and force plate data were recorded simultaneously and synchronized. Following walking trials, subjects were asked to provide a subjective rating of perceived footwear slipperiness. Each subject was asked to slide the entire right foot back and forth on the laboratory floor surface to evaluate the slipperiness of the hard and soft shoes. Subjects provided a subjective rating of slipperiness (Visual Analog Scale) by placing a vertical mark along a 10 cm horizontal line. One end of the line was anchored with a ‘‘0’’ (very slippery) while the other end of the line was anchored with ‘‘10’’ (not slippery). 2.4. Data management and analysis Ground reaction force data were filtered at 350 Hz using a low pass Butterworth fifth-order filter with zero lag compensation (DATAPAC 2K2 software; RUN Technologies, Mission Viejo, CA). Using Eq. (1), COFu was calculated as the ratio of the shear (algebraic resultant of the anterior-posterior and medial-lateral forces) to vertical ground reaction forces throughout the stance phase. COFu ¼ ¼

resultant shear GRF vertical GRF qffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi ðF anteriorposterior Þ2 þ ðF mediallateral Þ2 F vertical

(1)

The peak COFu during weight acceptance that would contribute to a forward foot slip was identified, as were the vertical and resultant shear ground reaction forces at the time of the peak COFu (Fig. 1). To avoid spuriously high values occurring when shear forces were divided by small vertical forces, only COFu data in which the vertical ground reaction force exceeded 50 N were analyzed [10]. Initial contact was defined when the vertical ground reaction force exceeded 5 N. COM position data were extrapolated and synchronized with ground reaction force data using DATAPAC 2K2 software (RUN Technologies, Mission Viejo, CA). Kinematic data were analyzed using the VICON 612 Workstation software (Vicon Motion Systems, Lake Forest, CA). A 15-component link-segment model was used to estimate the location of the total body COM based on the estimated body segment parameters [20]. The position of total body COM was defined from a weighted sum of the COM of all 15 segments. Kinematic data was filtered using a Woltering quintic spline filter with a predicted mean square error of 20 mm. The horizontal acceleration of the total body COM prior to initial contact was calculated using the following procedures. First, the instantaneous acceleration of the total body COM at the frame just prior to initial contact was obtained from the instantaneous velocity of the total body COM in the horizontal (forward progression) direction. COM instantaneous accelerationki ¼ ½V kiþ1  V ki1 =2Dt

(2)

where, k = the initial contact frame. This procedure was repeated for the six frames immediately proceeding initial contact. The horizontal acceleration of the total body COM prior to initial contact was then calculated by taking the

Y.-J. Tsai, C.M. Powers / Gait & Posture 30 (2009) 303–306

305

Table 2 Horizontal acceleration of the total body center of mass (COM) and heel mean (standard deviation). Soft soled shoes COM acceleration before contact (m/s2) COM acceleration after contact (m/s2) Heel acceleration before contact (m/s2) Heel acceleration after contact (m/s2) *

Fig. 1. Representative tracings of COFu and ground reaction forces during the stance phase of walking. Note that the initial spuriously high spikes in the COFu at the beginning and end of the stance are due to a relatively low vertical ground reaction forces. The vertical line indicates point of peak COFu during weight acceptance and corresponding vertical and resultant shear forces.

Hard soled shoes

P-value*

0.4 (0.7)

0.2 (0.7)

0.002

0.7 (1.2)

0.4 (1.0)

0.004

48.0 (7.6)

48.5 (8.3)

0.497

29.1 (11.0)

29.8 (10.4)

0.362

Paired t-tests.

the soft soled shoes (1.90  0.16 m/s vs.1.92  0.16 m/s; P = 0.019; Table 1). Similarly, the average stride length while wearing the harder soled shoes was significantly less when compared to the soft soled shoe condition (1.74  0.18 m vs. 1.78  0.16 m; P = 0.002). No differences in average cadence where observed between the two shoe conditions (138.7  12.5 steps/min vs. 137.5  11.8 steps/min; P = 0.054). 3.2. Acceleration of total body COM and heel

average of the six data points. The same procedures were used to calculate the horizontal acceleration of the total body COM immediately following initial contact. Similar methods were repeated to calculate to the horizontal acceleration of the heel prior to initial contact and horizontal acceleration of the heel immediately following initial contact. Heel accelerations were derived from the heel marker position data. Lower extremity kinematics at initial contact was quantified from the segment motions about the flexion/extension axes in the hip, knee, and ankle joints. Stride characteristics including walking velocity, cadence and stride length were obtained for each subject. In addition, subjective perception of footwear slipperiness was rated as a continuous variable, ranging from 0 (very slippery) to 100 (not slippery). 2.5. Statistical analysis Paired t-tests were performed to determine the influence of footwear sole hardness on the peak COFu, as well as vertical and resultant shear ground reaction forces at the time of peak COFu. Paired t-tests also were used for the secondary variables of interest, including COM and heel accelerations (prior to and immediately following initial contact), lower extremity kinematics at initial contact, stride characteristics, and subjective perception of footwear slipperiness. Analyses were performed using SPSS 11.5 statistical software (SPSS, Chicago, IL). A significance level of 0.05 was used for all statistical comparisons.

The horizontal acceleration of the total body COM prior to initial contact was significantly less when wearing the harder soled shoes when compared to the soft soled shoes (0.2  0.7 m/s2 vs. 0.4  0.7 m/s2; P = 0.002; Table 2). Similarly, the horizontal acceleration of the COM after initial contact was significantly less with the harder soled shoes when compared to the soft soled shoes (0.4  1.0 m/s2 vs. 0.7  1.2 m/s2; P = 0.004). No differences in the horizontal acceleration of the heel prior to and immediately following initial contact were observed (Table 2). 3.3. Peak COFu and ground reaction forces On average, the peak COFu while wearing the harder soled shoes was significantly lower when compared to the soft soled shoes (0.23  0.03 vs. 0.26  0.04, P < 0.001; Table 3). The peak COFu for soft soled shoes was 13% greater than the harder soled shoes. The vertical ground reaction forces at the time of peak COFu was not significantly different between two shoes conditions (Table 3). However, the resultant shear forces at the time of peak COFu was significantly lower while wearing the harder soled shoes when compared to the soft soled shoes (152.2  48.4 N vs. 168.5  49.1 N; P = 0.016; Table 3).

Table 3 Peak utilized coefficient of friction (COFu) and corresponding ground reaction forces mean (standard deviation).

3. Results Peak COFu Vertical forces at peak COFu (N) Shear forces at peak COFu (N)

3.1. Stride characteristics On average, the self-selected fast walking speed when wearing the harder soled shoes was significantly slower than when wearing Table 1 Stride characteristics mean (standard deviation).

Velocity (m/s) Cadence (steps/min) Stride length (m) *

Paired t-tests.

*

Soft soled shoes

Hard soled shoes

P-value*

0.26 (0.04) 651.0 (189.5) 168.5 (49.1)

0.23 (0.03) 662.0 (199.4) 152.2 (48.4)

<0.001 0.705 0.016

Paired t-tests.

Table 4 Lower extremity joint angles at initial contact mean (standard deviation).

Soft soled shoes

Hard soled shoes

P-value*

1.92 (0.16) 137.5 (11.8) 1.8 (0.2)

1.90 (0.16) 138.7 (12.5) 1.7 (0.2)

0.019 0.054 0.002

Ankle dorsiflexion (deg) Knee flexion (deg) Hip flexion (deg) *

Paired t-tests.

Soft soled shoes

Hard soled shoes

P-value*

9.3 (4.1) 4.7 (5.4) 36.7 (5.7)

8.6 (4.2) 5.6 (5.1) 36.7 (5.6)

0.026 0.079 0.916

306

Y.-J. Tsai, C.M. Powers / Gait & Posture 30 (2009) 303–306

3.4. Lower extremity joint angles at initial contact On average, the ankle dorsiflexion angle at initial contact was significantly smaller with the harder soled shoes when compared to the soft soled shoe condition (8.6  4.2 degrees vs. 9.3  4.1 degrees; P = 0.026; Table 4). No differences in the knee and hip flexion angles at initial contact were observed between the two shoe conditions (Table 4). 3.5. Subjective perception of shoe slipperiness On average, subjects perceived the harder soled shoes to be more slippery than the soft soled shoes (47.1  21.1 mm vs. 70.0  19.9 mm; P < 0.001). 4. Discussion The results of the current study support the hypothesis that the friction demand during weight acceptance is reduced when wearing harder soled shoes compared to when soft soled shoes are worn. This finding is in agreement with Fendley and Medoff who observed a reduction in peak COFu while wearing shoes with a harder sole [19]. In the current study, the decrease in peak COFu when wearing the hard soled shoes was the result of a reduction in the resultant shear forces at the time of peak COFu as no differences in the vertical ground reaction forces were observed. On average, the resultant shear forces were 10% lower during the hard soled walking trials. The decrease in resultant shear forces at the time of peak COFu can be attributed to the decrease in the total body COM acceleration prior to and immediately following initial contact. According to Newton’s second law of motion, the ground reaction forces generated by an individual are proportional to the acceleration of the center of mass in a given direction (F = ma). For example, a reduction in the horizontal acceleration of the total body COM would be expected to decrease the anterior-posterior ground reaction force, which in turn would decrease the resultant shear force. The reduction in the horizontal COM acceleration is consistent with the observed decrease in walking velocity during the hard sole walking trials. Although the walking speed for each subject was constrained to be within 5% of the first acceptable trial, small but significant shoe condition differences were still observed. Apart from a decrease in horizontal COM acceleration and walking velocity when wearing the hard soled shoes, additional differences in gait characteristics also were identified. For example, subjects shortened their stride length and decreased their ankle dorsiflexion angle at initial contact. Given that subjects perceived the harder soled shoes to be more slippery, we believe that the changes in kinematics and stride characteristics represent subtle gait modifications aimed at decreasing the friction demand during walking (i.e., protective strategy). Our findings are in agreement with previous studies in that individuals who perceive a floor surface to be slippery adjusted their gait patterns by decreasing walking velocity, shortening stride length and decreasing ankle impact angle [21,22]. An alternative explanation for the observed changes in stride characteristics with the hard soled shoes may be related to the fact that harder soled shoes were less flexible at the midfoot and forefoot regions. Although, the degree of stiffness was not quantified as part of the current study, it is conceivable that the harder soled shoes could have restricted forward propulsion by limiting forefoot flexion. In turn, this could have resulted in a shortened stride length and decrease in walking velocity. The probability of slip increases when either the friction demand of an individual increases or the available friction from the shoe/floor interface decreases [7,8]. Although the results of the current study suggest that individuals employ subtle gait adjustments aimed at

reducing friction demand, the decrease in peak COFu was only 0.03. In contrast, the difference in the available COF between the hard and soft soled shoes was quite large (0.29). This would suggest that once the available friction from a floor surface is decreased (i.e., in the presence of a contaminant), the observed reduction in friction demand resulting from gait modifications when wearing hard soles shoes may not be sufficient to prevent a slip from occurring. This hypothesis is consistent with a recent publication by Tsai and Powers [23] who demonstrated that persons wearing hard soled shoes are more likely to experience a slip event when compared to persons wearing soft soled shoes. 5. Conclusions The findings of the current study suggest that the friction demand during weight acceptance of an individual is reduced when wearing hard soled shoes compared to when soft soled shoes are worn. We believe the reduction in peak COFu is the result of gait modifications aimed at reducing the friction demand during walking when wearing shoes that is perceived to be more slippery. References [1] Layne LA, Pollack KM. Nonfatal occupational injuries from slips, trips, and falls among older workers treated in hospital emergency departments, United States 1998. Am J Ind Med 2004;46(1):32–41. [2] BLS. National Census of Fatal Occupational Injuries in 2004. Bureau of Labor Statistics 2005; http://www.bls.gov/iif/oshcfil.htm. [3] Englander F, Hodson TJ, Terregrossa RA. Economic dimensions of slip and fall injuries. J Forensic Sci 1996;41(5):733–46. [4] Leamon TB, Murphy PL. Occupational slips and falls: more than a trivial problem. Ergonomics 1995;38(3):487–98. [5] Bentley TA, Hide S, Tappin D, Moore D, Legg S, Ashby L, et al. Investigating risk factors for slips, trips and falls in New Zealand residential construction using incident-centered and incident-independent methods. Ergonomics 2006; 49(1):62–77. [6] Steinberg M, Cartwright C, Peel N, Williams G. A sustainable programme to prevent falls and near falls in community dwelling older people: Results of a randomised trial. J Epidemio Commun H 2000;54:227–32. [7] Hanson JP, Redfern MS, Mazumdar M. Predicting slips and falls considering required and available friction. Ergonomics 1999;42(12):1619–33. [8] Burnfield JM, Powers CM. Prediction of slips: An Evaluation of utilized coefficient of friction and available slip resistance. Ergonomics 2006;49(10): 982–95. [9] Burnfield JM, Powers CM. Influence of age and gender on utilized coefficient of friction during walking at different speed. In: Marpet MI, Sapienza MA, editors. Metrology of Pedestrian Locomotion and Slip Resistance. West Conshohocken, PA: ASTM International; 2002. p. 3–16. [10] Buczek FL, Banks SA. High-resolution force plate analysis of utilized slip resistance in human walking. J Test Eval 1996;24(6):353–8. [11] Menz HB, Lord ST, McIntosh AS. Slip resistance of casual footwear: implications for falls in older adults. Gerontology 2001;47(3):145–9. [12] Manning DP, Jones C. The effect of roughness, floor polish, water, oil and ice on underfoot friction: current safety footwear solings are less slip resistant than microcellular polyurethane. Appl Ergon 2001;32(2):185–96. [13] Gao C, Abeysekera J, Hirvonen M, Gronquist R. Slip resistant properties of footwear on ice. Ergonomics 2004;47(6):710–6. [14] Redfern MS, Bidanda B. Slip resistance of the shoe-floor interface under biomechanically relevant conditions. Ergonomics 1994;37(3):511–24. [15] Chaffin DB, Woldstad JC, Trujillo A. Floor/shoe slip resistance measurement. Am Ind Hyg Assoc J 1992;53(5):283–9. [16] McCaw ST, Heil ME, Hamill J. The effect of comments about shoe construction on impact forces during walking. Med Sci Sports Exer 2000;32(7):1258–64. [17] Milani TL, Hennig EM, Lafortune MA. Perceptual and biomechanical variables for running in identical shoe constructions with varying midsole hardness. Clin Biomech 1997;12(5):294–300. [18] De Wit B, De Clercq D, Aerts P. Biomechanical analysis of the stance phase during barefoot and shod running. J Biomech 2000;33(3):269–78. [19] Fendley AE, Medoff HP. Required coefficient of friction versus top-piece/outsole hardness and walking speed: significance of correlations. J Forensic Sci 1996;41(5):763–9. [20] Winter DA. Anthropometry. In: Biomechanics and motor control of human movement. Hoboken: Wiley-Interscience; 2005. 59–85 pp.. [21] Fong DT-P, Hong Y, Li JX. Lower-extremity gait kinematics on slippery surfaces in construction worksites. Med Sci Sports Exer 2005;37(3):447–54. [22] Cham R, Redfern MS. Changes in gait when anticipating slippery floors. Gait Posture 2002;15:159–71. [23] Tsai Y-J, Powers CM. The influence of footwear sole hardness on slip initiation in young adults. J Forensic Sci 2008;53(4):884–8.