Validity of the iPhone M7 motion co-processor as a pedometer for able-bodied ambulation

Physical activity benefits for disease prevention are well-established. Smartphones offer a convenient platform for community-based step count estimation to monitor and encourage physical activity. Accuracy is dependent on hardware–software platforms, creating a recurring challenge for validation, but the Apple iPhone® M7 motion co-processor provides a standardised method that helps address this issue. Validity of the M7 to record step count for level-ground, able-bodied walking at three self-selected speeds, and agreement with the StepWatch^TM was assessed. Steps were measured concurrently with the iPhone® (custom application to extract step count), StepWatch^TM and manual count. Agreement between iPhone® and manual/StepWatch^TM count was estimated through Pearson correlation and Bland-Altman analyses. Data from 20 participants suggested that iPhone® step count correlations with manual and StepWatch^TM were strong for customary (1.3 ± 0.1 m/s) and fast (1.8 ± 0.2 m/s) speeds, but weak for the slow (1.0 ± 0.1 m/s) speed. Mean absolute error (manual–iPhone®) was 21%, 8% and 4% for the slow, customary and fast speeds, respectively. The M7 accurately records step count during customary and fast walking speeds, but is prone to considerable inaccuracies at slow speeds which has important implications for certain patient groups. The iPhone® may be a suitable alternative to the StepWatch^TM for only faster walking speeds.     

Measurements of drag and lift on tennis balls in flight

Measurements are presented of drag and lift on new tennis balls in flight. Two video cameras were used to measure the velocity and height of the balls at two positions separated horizontally by 6.4 m. The balls were fired from a ball launcher at speeds between 15 and 30 m/s and with topspin or backspin at rates up to 2,500 rpm. Significant shot-to-shot variations were found in both the drag and lift coefficients. The average drag coefficient was 0.507 ± 0.024, independent of ball speed or spin, and lower than the value usually observed in wind tunnel experiments. The lift coefficient increased with ball spin, on average, but significant lift was observed even at very low spin. The latter effect can be attributed to a side force arising from asymmetries in the ball surface, analogous to the side force responsible for the erratic path of a knuckleball in baseball.     

A three-dimensional analysis of the contributions of upper limb joint movements to horizontal racket head velocity at ball impact during tennis serving

In this study, we examined the relationship between upper limb joint movements and horizontal racket head velocity to clarify joint movements for developing racket head speed during tennis serving. Sixty-six male tennis players were videotaped at 200 Hz using two high-speed video cameras while hitting high-speed serves. The contributions of each joint rotation to horizontal racket velocity were calculated using vector cross-products between the angular velocity vectors of each joint movement and relative position vectors from each joint to the racket head. Major contributors to horizontal racket head velocity at ball impact were shoulder internal rotation (41.1%) and wrist palmar flexion (31.7%). The contribution of internal rotation showed a significant positive correlation with horizontal racket head velocity at impact (r = 0.490, P < 0.001), while the contribution of palmar flexion showed a significant negative correlation (r = ? 0.431, P < 0.001). The joint movement producing the difference in horizontal racket head velocity between fast and slow servers was shoulder internal rotation, and angular velocity of shoulder internal rotation must be developed to produce a high racket speed.     

The effects of eccentric exercise-induced muscle damage on running kinematics at different speeds

Abstract

This study investigated the effects of knee localised muscle damage on running kinematics at varying speeds. Nineteen young women (23.2 ± 2.8 years; 164 ± 8 cm; 53.6 ± 5.4 kg), performed a maximal eccentric muscle damage protocol (5 × 15) of the knee extensors and flexors of both legs at 60 rad · s^-1. Lower body kinematics was assessed during level running on a treadmill at three speeds pre- and 48 h after. Evaluated muscle damage indices included isometric torque, muscle soreness and serum creatine kinase activity. The results revealed that all indices changed significantly after exercise, indicating muscle injury. Step length decreased and stride frequency significantly increased 48 h post-exercise only at the fastest running speed (3 m · s^-1). Support time and knee flexion at toe-off increased only at the preferred transition speed and 2.5 m · s^-1. Knee flexion at foot contact, pelvic tilt and obliquity significantly increased, whereas hip extension during stance-phase, knee flexion during swing-phase, as well as knee and ankle joints range of motion significantly decreased 48 h post-exercise at all speeds. In conclusion, the effects of eccentric exercise of both knee extensors and flexors on particular tempo-spatial parameters and knee kinematics of running are speed-dependent. However, several pelvic and lower joint kinematics present similar behaviour at the three running speeds examined. These findings provide new insights into how running kinematics at different speeds are adapted to compensate for the impaired function of the knee musculature following muscle damage.     

Running speed increases plantar load more than per cent body weight on an AlterG® treadmill

AlterG® treadmills allow for running at different speeds as well as at reduced bodyweight (BW), and are used during rehabilitation to reduce the impact load. The aim of this study was to quantify plantar loads borne by the athlete during rehabilitation. Twenty trained male participants ran on the AlterG® treadmill in 36 conditions: all combinations of indicated BW (50–100%) paired with different walking and running speeds (range 6–16 km · hr^–1) in a random order. In-shoe maximum plantar force (Fmax) was recorded using the Pedar-X system. Fmax was lowest at the 6 km · hr^–1 at 50% indicated BW condition at 1.02 ± 0.21BW and peaked at 2.31 ± 0.22BW for the 16 km · hr^–1 at 100% BW condition. Greater increases in Fmax were seen when increasing running speed while holding per cent BW constant than the reverse (0.74BW–0.91BW increase compared to 0.19–0.31BW). A table is presented with each of the 36 combinations of BW and running speed to allow a more objective progression of plantar loading during rehabilitation. Increasing running speed rather than increasing indicated per cent BW was shown to have the strongest effect on the magnitude of Fmax across the ranges of speeds and indicated per cent BWs examined.     

Hammer acceleration due to thrower and hammer movement patterns

Ground reaction force and wire tensile force were measured during test throws by three hammer throwers: the Asian record holder, who had a personal best of 83.47 m at the time of the investigation, and two university athletes, with personal bests of 59.95 m and 46.30 m respectively. They were filmed using three high-speed video cameras (250 Hz). The displacements of the hammer head and the athletes' centres of mass were calculated using three-dimensional analysis procedures. The Asian record holder's centre of mass and the hammer head on the final two turns exhibited approximate conjunctions of the hammer high point and the thrower's low point and vice versa about the hammer's azimuth angle. It is conjectured that the reason why the thrower's movement is asynchronous with the hammer's movement by approximately half a turn is to accelerate the hammer head in a manner similar to the way that the amplitude of a pendulum increases when it is pulled upward by a string against the downward movement of the swinging weight.     

Sprint mechanical differences at maximal running speed: Effects of performance level

ABSTRACT

As the effect of performance level on sprinting mechanics has not been fully studied, we examined mechanical differences at maximal running speed (MRS) over a straight-line 35 m sprint amongst sprinters of different performance levels. Fifty male track and field sprinters, divided in Slow, Medium and Fast groups (MRS: 7.67 ± 0.27 m?s^?1, 8.44 ± 0.22 m?s^?1, and 9.37 ± 0.41 m?s^?1, respectively) were tested. A high-speed camera (250 Hz) recorded a full stride in the sagittal plane at 30–35 m. MRS was higher (p < 0.05) in Fast vs. Medium (+11.0%) and Slow (+22.1%) as well as in Medium vs. Slow (+10.0%). Twelve, eight and seven out of 21 variables significantly distinguished Fast from Slow, Fast from Medium and Medium from Slow sprinters, respectively. Propulsive phase was signi?cantly shorter in Fast vs. Medium (?17.5%) and Slow (?29.4%) as well as in Medium vs. Slow (?14.4%). Fast sprinters had significantly higher vertical and leg stiffness values than Medium (+44.1% and +18.1%, respectively) and Slow (+25.4% and +22.0%, respectively). MRS at 30–35 m increased with performance level during a 35-m sprint and was achieved through shorter contact time, longer step length, faster step rate, and higher vertical and leg stiffness.     

Kinetics of elite unilateral below-elbow amputee running

The role of the upper limbs in human locomotion and their influence on ground reaction force (GRF) have been extensively examined for walking and slow running. However, research has focused on unimpaired populations and has not evaluated high-speed running. In this study, the GRFs of an unilateral upper limb amputee athlete [missing right forearm, personal bests (PB): 400 m: 0:48.45 min, 800 m: 1:50.92 min] running at speeds of 5.4 and 8 m/s were collected using four floor-mounted force plates in a 100 m tartan track recording at 1000 Hz. The amputee athlete also performed trials with a running speed of 8 m/s wearing a weighted cuff on his impaired arm (0.5 and 1 kg, respectively). GRF data (without additional weight) were compared to those of an unimpaired athlete with similar PBs and anthropometry (age, height, weight). All data were evaluated for anterior–posterior, medial–lateral and vertical GRF as well as for stance phase (SP) duration and free moment (FM) values and a paired Student-t test (\(\alpha =5\) %) was performed on maximum and minimum values of the respective data sets comparing the left and right side of both athletes with each other, but no intersubjective comparisons were performed. The results revealed that vertical GRF showed significant differences for the impaired athlete at both running speeds comparing left and right foot, whereas the unimpaired athlete showed no significant differences in this matter at all. Medial–lateral GRF showed highly significant differences between the left and right foot of both athletes at both running speeds, whereas the results for anterior–posterior GRF and FM, however, showed ambiguous results. The trials with additional weight on the impaired limb led to significant differences for SP duration, but not for all conditions, and left–right differences did not change with additional weight. GRFs were found to differ in minimum medial–lateral GRF with no and heavy weight, maximum medial–lateral GRF within all weight conditions and maximum vertical GRF with no and heavy weight. For the FM, the overall pattern changed drastically and maximum FM showed a highly significant difference between the left and right foot, but none between the different conditions for both feet. Even though more significant asymmetries could be revealed for the impaired athlete, no general conclusion can be drawn at this point, given the limitations of the here presented study (low number of subjects available, anthropometric data of the impaired athlete could not be assessed).     

Using Principal Components Analysis to Identify Individual Differences in Vertical Jump Performance

Purpose: The purpose of this study was to determine the reliability of cardiorespiratory and pelvic kinematic responses to simulated horseback riding (SHBR) and to characterize responses to SHBR relative to walking in apparently healthy children. Method: Fifteen healthy children (M_age = 9.5 ± 2.6 years) completed SHBR on a commercially available simulator at low intensity (0.27 Hz) and high intensity (0.65 Hz) during 3 sessions on different occasions. Heart rate (HR), blood pressure, and respiratory gases were measured at rest and during steady-state exercise at both intensities. Pelvic displacement was measured during steady-state exercise. Rate of energy expenditure, mean arterial pressure, and rate pressure product (RPP) were calculated. Participants also walked on a treadmill for 26.8 m/min to 80.5 m/min in 13.4-m/min increments at 0% grade during 1 session to compare cardiorespiratory responses with those of SHBR. Results: Physiological variables across all 3 SHBR sessions were similar at both intensities (p>.05 for all). Intraclass correlation coefficients (ICCs) and coefficients of variation indicate good to modest reliability of cardiorespiratory measures during SHBR (ICCs = .542–.996 for oxygen consumption, energy expenditure, and RPP). Cardiorespiratory variables, except for HR, were 2% to 19% greater, and pelvic displacement was up to 37% greater with high-intensity riding. Treadmill walking at all speeds elicited greater physiological responses compared with SHBR (p < .05). Conclusion: Cardiorespiratory responses and pelvic kinematics are reproducible with SHBR in young children, and these responses were lower than those elicited by slow treadmill walking.     

Development of a measurement and feedback training tool for the arm strokes of high-performance luge athletes

Abstract

Previous studies have shown that the start plays a critical role in sliding events and explains more than 55% of the variance of the final time in luge. Experts evaluate the contribution of the arm strokes to be 23% of the total starting performance. The aim of the present study was to develop a measurement and feedback training tool (Speedpaddler) for the arm strokes of high-performance luge athletes. The construction is an aluminium alloy framework with a customary belt conveyor system, which is driven by two synchronized servo motors. Training is possible with constant speeds up to 12 m · s^?1 or several speed curves, which simulate the acceleration of different luge tracks. The construction facilitates variations in the inclination and speed of the conveyor belts and thereby the resistance and movement speed. If the athlete accelerates the conveyor belts during arm-paddling, the torque of the motors decreases. Torque measurements and high-speed video offer valuable insights into the several technique criteria. Comparisons of arm-paddle cycle durations on ice and on the Speedpaddler with 18 luge athletes (national team and juniors) showed no statistical differences. The Speedpaddler might be a useful tool to improve starting performance all year round.     

The validity and reliability of a novel indoor player tracking system for use within wheelchair court sports

Abstract

The aim of the current study was to investigate the validity and reliability of a radio frequency-based system for accurately tracking athlete movement within wheelchair court sports. Four wheelchair-specific tests were devised to assess the system during (i) static measurements; (ii) incremental fixed speeds; (iii) peak speeds; and (iv) multidirectional movements. During each test, three sampling frequencies (4, 8 and 16 Hz) were compared to a criterion method for distance, mean and peak speeds. Absolute static error remained between 0.19 and 0.32 m across the session. Distance values (test (ii)) showed greatest relative error in 4 Hz tags (1.3%), with significantly lower errors seen in higher frequency tags (<1.0%). Relative peak speed errors of <2.0% (test (iii)) were revealed across all sampling frequencies in relation to the criterion (4.00 ± 0.09 m · sˉ¹). Results showed 8 and 16 Hz sampling frequencies displayed the closest-to-criterion values, whilst intra-tag reliability never exceeded 2.0% coefficient of variation (% CV) during peak speed detection. Minimal relative distance errors (<0.2%) were also seen across sampling frequencies (test (iv)). To conclude, the indoor tracking system is deemed an acceptable tool for tracking wheelchair court match play using a tag frequency of 8 or 16 Hz.     

Stance limb kinetics of older male athletes endurance running performance

The aim of this study was to examine the age-based, lower limb kinetics of running performances of endurance athletes. Six running trials were performed by 24 male athletes, who were distinguished by three age groupings (S35: 26–32 years, M50: 50–54 years, M60+: 60–68 years). Lower limb coordinate and ground reaction force data were collected using a nine camera infra-red system synchronised with a force plate. A slower anteroposterior (M ± SD S35 = 4.13 ± 0.54 m/s: M60+ = 3.34 ± 0.40 m/s, p < 0.05) running velocity was associated with significant (p < 0.05) decreases in step length and discrete vertical ground contact force between M60+and S35 athletes. The M60+athletes simultaneously generated a 32% and 42% reduction (p < 0.05) in ankle joint moment when compared to the M50 and S35 athletes and 72% (p < 0.05) reduction in knee joint stiffness when compared to S35 athletes. Age-based declines in running performance were associated with reduced stance phase force tolerance and generation that may be accounted for due to an inhibited force–velocity muscular function of the lower limb. Joint-specific coaching strategies customised to athlete age are warranted to maintain/enhance athletes' dynamic performance.     

Gender difference in cycling speed and age of winning performers in ultra-cycling – the 508-mile “Furnace Creek” from 1983 to 2012

We analysed (i) the gender difference in cycling speed and (ii) the age of winning performers in the 508-mile “Furnace Creek 508”. Changes in cycling speeds and gender differences from 1983 to 2012 were analysed using linear, non-linear and hierarchical multi-level regression analyses for the annual three fastest women and men. Cycling speed increased non-linearly in men from 14.6 (s = 0.3) km · h^?1 (1983) to 27.1 (s = 0.7) km · h^?1 (2012) and non-linearly in women from 11.0 (s = 0.3) km · h^?1 (1984) to 24.2 (s = 0.2) km · h^?1 (2012). The gender difference in cycling speed decreased linearly from 26.2 (s = 0.5)% (1984) to 10.7 (s = 1.9)% (2012). The age of winning performers increased from 26 (s = 2) years (1984) to 43 (s = 11) years (2012) in women and from 33 (s = 6) years (1983) to 50 (s = 5) years (2012) in men. To summarise, these results suggest that (i) women will be able to narrow the gender gap in cycling speed in the near future in an ultra-endurance cycling race such as the “Furnace Creek 508” due to the linear decrease in gender difference and (ii) the maturity of these athletes has changed during the last three decades where winning performers become older and faster across years.     

Facial feature tracking: a psychophysiological measure to assess exercise intensity?

The primary aim of this study was to determine whether facial feature tracking reliably measures changes in facial movement across varying exercise intensities. Fifteen cyclists completed three, incremental intensity, cycling trials to exhaustion while their faces were recorded with video cameras. Facial feature tracking was found to be a moderately reliable measure of facial movement during incremental intensity cycling (intra-class correlation coefficient = 0.65–0.68). Facial movement (whole face (WF), upper face (UF), lower face (LF) and head movement (HM)) increased with exercise intensity, from lactate threshold one (LT1) until attainment of maximal aerobic power (MAP) (WF 3464 ± 3364mm, P < 0.005; UF 1961 ± 1779mm, P = 0.002; LF 1608 ± 1404mm, P = 0.002; HM 849 ± 642mm, P < 0.001). UF movement was greater than LF movement at all exercise intensities (UF minus LF at: LT1, 1048 ± 383mm; LT2, 1208 ± 611mm; MAP, 1401 ± 712mm; P < 0.001). Significant medium to large non-linear relationships were found between facial movement and power output (r² = 0.24–0.31), HR (r² = 0.26–0.33), [La^?] (r² = 0.33–0.44) and RPE (r² = 0.38–0.45). The findings demonstrate the potential utility of facial feature tracking as a non-invasive, psychophysiological measure to potentially assess exercise intensity.     

The impact phase of drop punt kicking for maximal distance and accuracy

Impact is an important aspect of the kicking skill. This study examined foot and ball motion during impact and compared distance and accuracy punt kicks. Two-dimensional high-speed video (4000 Hz) captured data of the shank, foot and ball through impact of 11 elite performers kicking for maximal distance and towards a target 20 m in distance. Four phases were identified during impact, with an overall reduction in foot velocity of 5.0 m · s^?1 (± 1.1 m · s^?1) and increase in ball velocity of 22.7 m · s^?1 (± 2.3 m · s^?1) from the start to end of contact. Higher foot velocity was found in distance compared to accuracy kicks (22.1 ± 1.6 m · s^?1 vs. 17.7 ± 0.9 m · s^?1, P < 0.05), and was considered to produce the significant differences in all impact characteristics excluding foot-to-ball speed ratio. Ankle motion differed between the kicking tasks; distance kicks were characterised by greater rigidity compared to accuracy kicks evident by larger force (834 ± 107 N vs. 588 ± 64 N) and smaller change in ankle angle (2.2 ± 3.3° vs. 7.2 ± 6.4°). Greater rigidity was obtained by altering the position of the ankle at impact start; distance kicks were characterised by greater plantarflexion (130.1 ± 5.8° vs. 123.0 ± 7.9°, P < 0.05), indicating rigidity maybe actively controlled for specific tasks.     

An analysis of two styles of arm action in the vertical countermovement jump

The aim of this study was to determine the effect of two arm swing techniques, the simultaneous arm swing and the early arm swing, on vertical countermovement jump performance and on the contribution of the arms to vertical movement at the centre of mass (CM) during the propulsion phase. Participants were 28 athletes practicing sports in which the vertical jump constitutes a basic ability. Ground reaction forces were recorded by a force platform and the sagittal plane motion was recorded by a video camera. Although at take-off the vertical velocity (2.7 ± 0.2 m/s for simultaneous technique vs. 2.8 ± 0.2 m/s for early technique; p = 0.040) and position (1.18 ± 0.06 m for simultaneous vs. 1.17 ± 0.05 m for early; p = 0.033) of the CM were significantly different, no difference was observed in jump height (1.56 ± 0.01 m in both techniques). The arm action differed during the initial and final propulsion phases in both styles but the accumulated vertical contribution was similar. The practical implication in sports is that the use of the arm-swing technique to reach the maximum jump height should be determined by tactical demands instead of the technical execution of the arms.     

The influence of shaft stiffness on potential energy and puck speed during wrist and slap shots in ice hockey

The purpose of this study was to explore the relationship between hockey stick shaft stiffness and puck speed with mechanical energy considerations during stationary wrist and slap shots. Thirty left-handed pro-model composite hockey sticks, submitted by eleven hockey stick manufacturers, were subjected to a mechanical cantilever bend test to determine the shaft stiffness of each stick. Eight sticks representing the entire spectrum of stiffnesses were then used by five elite male hockey players to perform stationary wrist and slap shots in a laboratory setting. Eight infra-red high-speed digital video cameras were used to capture shaft deformation and puck speed. A second mechanical test then replicated the loading patterns applied to each stick during shooting. Force-deformation data from this test were used to determine the shaft stiffness and potential energy storage and return associated with each stick during shooting. The results of this study suggest that shaft stiffness has an influence on puck speed in wrist but not slap shots. During a wrist shot, a given player should realise higher puck speeds with a stick in which they store increased elastic potential energy in the shaft. In general, flexible sticks were found to store the most energy. However, how the athlete loads the stick has as much influence on puck speed as stick construction. Energy considerations were unable to explain changes in puck speed for the slap shot. For this type of shot it is the athlete and not the equipment influencing puck speed, but the governing mechanisms have yet to be elucidated.     

Effect of team rank and player classification on activity profiles of elite wheelchair rugby players

The aim of the current study was to establish which indicators of mobility are associated with successful wheelchair rugby performance and determine whether these indicators differed across classification. Data were collected from 11 international teams during 30 matches (353 match observations) using a radio-frequency-based, indoor tracking system across two tournaments. Players (n = 111) were first grouped by team rank as determined by their International Wheelchair Rugby Federation (IWRF) world ranking (LOW, MID, HIGH) and then into one of four groups based on their IWRF classification: Group I (0.5), Group II (1.0–1.5), Group III (2.0–2.5) and Group IV (3.0–3.5). The volume of activity (relative distance and mean speed), peak speed and time spent within classification-specific arbitrary speed zones were calculated for each individual. Although no differences were identified in the volume of activity, playing time was significantly reduced in LOW (34:51 ± 8:35) compared to MID (48:54 ± 0:51) and HIGH (45:38 ± 9:53), which was further supported by the greater number of substitutions performed by LOW. HIGH achieved greater peak speeds (3.55 ± 0.40 m · sˉ¹) than LOW (3.27 ± 0.42 m · sˉ¹) and MID (3.45 ± 0.41 m · sˉ¹). Peak speed was further shown to be classification-dependent (P ≤ 0.005), whereby HIGH Groups III and IV players achieved greater peak speeds than LOW and MID. The time spent performing high-intensity activities was also greater in HIGH compared to LOW and MID, whilst further influenced by classification (P ≤ 0.0005). To conclude, peak speed and the ability to perform a greater number of high-intensity activities were associated with successful performance in wheelchair rugby.     

Predicting centre of mass horizontal speed in low to severe swimming intensities with linear and non-linear models

We aimed to compare multilayer perceptron (MLP) neural networks, radial basis function neural networks (RBF) and linear models (LM) accuracy to predict the centre of mass (CM) horizontal speed at low-moderate, heavy and severe swimming intensities using physiological and biomechanical dataset. Ten trained male swimmers completed a 7 × 200 m front crawl protocol (0.05 m.s^?1 increments and 30 s intervals) to assess expiratory gases and blood lactate concentrations. Two surface and four underwater cameras recorded independent images subsequently processed focusing a three-dimensional reconstruction of two upper limb cycles at 25 and 175 m laps. Eight physiological and 13 biomechanical variables were inputted to predict CM horizontal speed. MLP, RBF and LM were implemented with the Levenberg-Marquardt algorithm (feed forward with a six-neuron hidden layer), orthogonal least squares algorithm and decomposition of matrices. MLP revealed higher prediction error than LM at low-moderate intensity (2.43 ± 1.44 and 1.67 ± 0.60%), MLP and RBF depicted lower mean absolute percentage errors than LM at heavy intensity (2.45 ± 1.61, 1.82 ± 0.92 and 3.72 ± 1.67%) and RBF neural networks registered lower errors than MLP and LM at severe intensity (2.78 ± 0.96, 3.89 ± 1.78 and 4.47 ± 2.36%). Artificial neural networks are suitable for speed model-fit at heavy and severe swimming intensities when considering physiological and biomechanical background.     

Acute and chronic effects of aquatic treadmill training on land treadmill running kinematics: A cross-over and single-subject design approach

The aim of this study was to determine if selected kinematic measures (foot strike index [SI], knee contact angle and overstride angle) were different between aquatic treadmill (ATM) and land treadmill (LTM) running, and to determine if these measures were altered during LTM running as a result of 6 weeks of ATM training. Acute effects were tested using 15 competitive distance runners who completed 1 session of running on each treadmill type at 5 different running speeds. Subsequently, three recreational runners completed 6 weeks of ATM training following a single-subject baseline, intervention and withdrawal experiment. Kinematic measures were quantified from digitisation of video. Regardless of speed, SI values during ATM running (61.3 ± 17%) were significantly greater (P = 0.002) than LTM running (42.7 ± 23%). Training on the ATM did not change (pre/post) the SI (26 ± 3.2/27 ± 3.1), knee contact angle (165 ± 0.3/164 ± 0.8) or overstride angle (89 ± 0.4/89 ± 0.1) during LTM running. Although SI values were different between acute ATM and LTM running, 6 weeks of ATM training did not appear to alter LTM running kinematics as evidenced by no change in kinematic values from baseline to post intervention assessments.