Finger forces in fastball baseball pitching

Forces imparted by the fingers onto a baseball are the final, critical aspects for pitching, however these forces have not been quantified previously as no biomechanical technology was available. In this study, an instrumented baseball was developed for direct measurement of ball reaction force by individual fingers and used to provide fundamental information on the forces during a fastball pitch. A tri-axial force transducer with a cable having an easily-detachable connector were installed in an official baseball. Data were collected from 11 pitchers who placed the fingertip of their index, middle, ring, or thumb on the transducer, and threw four-seam fastballs to a target cage from a flat mound. For the index and middle fingers, resultant ball reaction force exhibited a bimodal pattern with initial and second peaks at 38–39 ms and 6–7 ms before ball release, and their amplitudes were around 97 N each. The ring finger and thumb produced single-peak forces of approximately 50 and 83 N, respectively. Shear forces for the index and middle fingers formed distinct peak at 4–5 ms before release, and the peaks summed to 102 N; a kinetic source for backspin on the ball. An additional experiment with submaximal pitching effort showed a linear relationship of peak forces with ball velocity. The peak ball reaction force for fastballs exceeded 80% of maximum finger strength measured, suggesting that strengthening of the distal muscles is important both for enhancing performance and for avoiding injuries.     

Multi-body dynamic coupling mechanism for generating throwing arm velocity during baseball pitching

The purpose of this study was to identify the detailed mechanism how the maximum throwing arm endpoint velocity is determined by the muscular torques and non-muscular interactive torques from the perspective of the dynamic coupling among the trunk, thorax and throwing and non-throwing arm segments. The pitching movements of ten male collegiate baseball pitchers were measured by a three-dimensional motion capture system. Using the induced-segmental velocity analysis (IVA) developed in this study, the maximum fingertip velocity of the throwing arm (MFV) was decomposed into each contribution of the muscular torques, passive motion-dependent torques due to gyroscopic moment, Coriolis force and centrifugal force, and other interactive torque components. The results showed that MFV (31.6 ± 1.7 m/s) was mainly attributed to two different mechanisms. The first is the passive motion-dependent effect on increasing the angular velocities of three joints (thorax rotation, elbow extension and wrist flexion). The second is the muscular torque effect of the shoulder internal rotation (IR) torque on generating IR angular velocity. In particular, the centrifugal force-induced elbow extension motion, which was the greatest contributor among individual joint contributions, was caused primarily by the angular velocity-dependent forces associated with the humerus, thorax, and trunk rotations. Our study also found that a compensatory mechanism was achieved by the negative and positive contributions of the muscular torque components. The current IVA is helpful to understand how the rapid throwing arm movement is determined by the dynamic coupling mechanism.     

Proprioceptive impairments associated with knee osteoarthritis are not generalized to the ankle and elbow joints

The mechanisms for proprioceptive changes associated with knee osteoarthritis (OA) remain elusive. Observations of proprioceptive changes in both affected knees and other joints imply more generalized mechanisms for proprioceptive impairment. However, evidence for a generalized effect remains controversial. This study examined whether joint repositioning proprioceptive deficits are localized to the diseased joint (knee) or generalized across other joints (elbow and ankle) in people with knee OA. Thirty individuals with right knee OA (17 female, 66 ± 7 [mean ± SD] years) of moderate/severe radiographic disease severity and 30 healthy asymptomatic controls of comparable age (17 female, 65 ± 8 years) performed active joint repositioning tests of the knee, ankle and elbow in randomised order in supine. Participants with knee OA had a larger relative error for joint repositioning of the knee than the controls (OA: 2.7 ± 2.1°, control: 1.6 ± 1.7°, p = .03). Relative error did not differ between groups for the ankle (OA: 2.2 ± 2.5°, control: 1.9 ± 1.3°, p = .50) or elbow (OA: 2.5 ± 3.3°, control: 2.9 ± 2.8°, p = .58). These results are consistent with a mechanism for proprioceptive change that is localized to the knee joint. This could be mediated by problems with mechanoreceptors, processing/relay of somatosensory input to higher centers, or joint-specific interference with cognitive processes by pain.     

Acceptable timing error at ball-bat impact for different pitches and its implications for baseball skills

In baseball hitting, batters need high precision timing control to hit the ball with bat’s sweet spot. Knowing the acceptable range of timing error for hitting the ball in the aimed direction for various pitch types is helpful to understand whether the cause of the batter's mis-hit is a spatial or temporal error and highlight the motor skills required by the batter. The purpose of this study was to determine the acceptable timing error in different baseball pitches and the impact characteristics of mis-hits. Twenty-six high school baseball players hit a ball launched from a pitching machine with three types of pitches: fastballs, curveballs, and slowballs. We recorded the three-dimensional behavior of the ball, bat, and human body (pelvis) using an optical motion capture system. We then defined the optimal impact location based on timing accuracy, and determined the acceptable range of timing error by the interactive relationship between the horizontal orientation of the bat’s long axis at the time of ball impact and the horizontal direction of the batted ball. The ±30° width in the horizontal direction of the batted ball was set as the precondition for the tolerance of timing. The acceptable timing error was ±7.9 ms for fastballs, ±10.7 ms for curveballs, and ±10.7 ms for slowballs, and the optimal timing for outside pitches was approximately 10 ms later than that for inside pitches. The timing error was also explained 38.1% by variation in the impact location along the long axis of the bat (R² = 0.381, P < 0.001) and was minimized at a position close to the bat’s sweet spot. These results suggest that the optimal impact location and acceptable range of timing error depend on the pitching course and speed and that timing accuracy is essential to achieve the spatial accuracy required to hit the ball at the bat’s sweet spot.     

An inferential investigation into how stride length influences temporal parameters within the baseball pitching delivery

Motion analyses of lower body mechanics offer new schemas to address injury prevention strategies among baseball pitchers, where the influence of stride length remains unknown. This study examined the temporal effect of stride length at constituent pitching events and phases. Nineteen competitive pitchers (15 collegiate, 4 high school) were randomly assigned to pitch two simulated, 80-pitch games at ±25% of their desired stride length. An integrated, three-dimensional motion capture system recorded each pitch. Paired t-tests were used to determine whether differences between stride conditions at respective events and within phases were significantly different. The results demonstrate the shorter strides mediated earlier onset of stride foot contact, reduced time in single support whereas double support intervals increased (p < .001). The opposite was observed with the longer strides. However, the acceleration phase, which comprises the highest throwing arm kinematics and kinetics, remained unchanged. The interaction between stride length, stride foot contact onsets, and time in single support is inferentially evidenced. The equivalent acceleration phases suggest stride length alone influenced time in single and double support by altering the onset of stride foot contact, which perhaps affects the mechanics in preparing the throwing arm for maximal external shoulder rotation.     

Variability of a “force signature” during windmill softball pitching and relationship between discrete force variables and pitch velocity

This study assessed reliability of discrete ground reaction force (GRF) variables over multiple pitching trials, investigated the relationships between discrete GRF variables and pitch velocity (PV) and assessed the variability of the “force signature” or continuous force-time curve during the pitching motion of windmill softball pitchers. Intraclass correlation coefficient (ICC) for all discrete variables was high (0.86–0.99) while the coefficient of variance (CV) was low (1.4–5.2%). Two discrete variables were significantly correlated to PV; second vertical peak force (r(5) = 0.81, p = 0.03) and time between peak forces (r(5) = −0.79; p = 0.03). High ICCs and low CVs support the reliability of discrete GRF and PV variables over multiple trials and significant correlations indicate there is a relationship between the ability to produce force and the timing of this force production with PV. The mean of all pitchers’ curve-average standard deviation of their continuous force-time curves demonstrated low variability (CV = 4.4%) indicating a repeatable and identifiable “force signature” pattern during this motion. As such, the continuous force-time curve in addition to discrete GRF variables should be examined in future research as a potential method to monitor or explain changes in pitching performance.     

Contralateral pelvic drop during gait increases knee adduction moments of asymptomatic individuals

PurposeThe current study purpose was to investigate the effects of contralateral pelvic drop gait on the magnitude of the knee adduction moment (KAM) within asymptomatic individuals.Methods15 participants walked on a dual belt instrumented treadmill while segment motions and ground reaction forces were recorded. Participants completed typical gait trials and pelvic drop gait trials. The net external KAM was calculated using inverse dynamics. Peak and impulse were identified. Frontal plane hip abduction/adduction and pelvic drop were determined. Correlations and paired t-tests were used for statistical hypothesis testing (alpha = 0.05).ResultsPeak hip adduction angle reached 4° (±6°) during pelvic drop trials compared to 0° (±6°) in the typical gait trials (p < 0.05) equating to 4° of pelvic drop. KAM impulse was higher in the pelvic drop trial (0.16 Nm s/kg ± 0.04) compared to the typical gait trial (0.13 Nm s/kg ± 0.05) (p < 0.001). Peak KAM was higher in the pelvic drop trial (0.55 Nm/kg ± 0.15) compared to the typical gait trial (0.40 Nm/kg ± 0.109) (p < 0.001). Correlations between change in KAM and change in hip adduction moment and pelvic drop were r > 0.80 (p < 0.001).ConclusionPelvic drop gait increased KAM peak and impulse. Results have implications for understanding relationships between frontal plane hip movement and the knee adduction moment during gait.     

Difference of motor overflow depending on the impaired or unimpaired hand in stroke patients

The aim of this study was to investigate the patterns of contralateral motor overflow (i.e. mirror movement) between the homologous body parts on the right and left side, in stroke patients during single-finger and multi-finger maximum force production tasks. Forty subjects, including stroke (n = 20) and normal subjects (n = 20), participated in this study. The stroke subjects maximally pressed force sensors with their fingers in a flexed position using a single (index, middle, ring, or little) or all fingers (all 4 fingers) using the impaired (IH) or unimpaired (UIH) hand, while the non-patient subjects used their right hands for the same tasks. The maximal voluntary forces in the ipsilateral and unintended pressing forces of each contralateral finger were recorded during the tasks. The magnitude of motor overflow to the contralateral side was calculated using the index of contralateral independence (CI). During the single finger tasks, the finger CI was significantly decreased in the UIH (91%) compared with that in the IH (99%) or normal hands (99%). Likewise, the multiple finger tasks showed that the CI was significantly lower in the UIH (84%) compared with that in the IH (96%) or normal hands (99%). However, the maximal forces were significantly lower in the IH relative to those in the UIH and normal hands. These data demonstrate that stroke patients have greater motor overflow from the UIH to the IH.     

Biomechanical analyses of prolonged handwriting in subjects with and without perceived discomfort

Since wrist-joint position affects finger muscle length and grip strength, we studied its biomechanical relevance in prolonged handwriting. We recruited participants from young adults, aged 18–24, and separated them into control (n = 22) and in-pain (n = 18) groups, based whether or not they experience pain while handwriting. The participants then performed a writing task for 30 min on a computerized system which measured their wrist-joint angle and documented their handwriting kinematics. The in-pain group perceived more soreness and had a less-extended wrist joint, longer on-paper time, and slower stroke velocity compared to control group. There was no significant difference in handwriting speed and quality between the two groups. The wrist extension angle significantly correlated with perceived soreness. Ergonomic and biomechanical analyses provide important information about the handwriting process. Knowledge of pen tip movement kinematics and wrist-joint position can help occupational therapists plan treatment for individuals with handwriting induced pain.     

Three-dimensional kinematics of upper limb anatomical movements in asymptomatic adults: Dominant vs. non-dominant

The effect of dominance on upper limb (UL) kinematics has only been studied on scapular movements. Moreover, when an anatomical UL movement is performed in a specific plane, secondary movements in the remaining planes involuntarily occur. These secondary movements have not been previously evaluated. The aim of this study was to compare the kinematics of primary and secondary angles of dominant and non-dominant UL during anatomical movements in asymptomatic adults.25 asymptomatic adults performed 6 anatomical movements bilaterally: shoulder flexion-extension, abduction-adduction, horizontal abduction-adduction, internal-external rotation, elbow flexion-extension and wrist pronation-supination. Kinematics of the dominant and non-dominant UL were compared by their ranges of motion (ROM) and their angular waveforms (Coefficient of Multiple Correlations, CMC).The comparison between dominant and non-dominant UL kinematics showed different strategies of movement, most notably during elbow flexion-extension (CMC = 0.29): the dominant UL exhibited more pronation at maximal elbow flexion. Significant secondary angles were found on most of the UL anatomical movements; e.g. a secondary ROM of shoulder (humero-thoracic) external-internal rotation (69° ± 16°) was found when the subject intended to perform maximal shoulder abduction-adduction (119° ± 21°).Bias of dominance should be considered when comparing pathological limb to the controlateral one. Normative values of primary and secondary angles during anatomical movements could be used as a reference for future studies on UL of subjects with neurological or orthopedic pathologies.     

Technique determinants of knee joint loads during cutting in female soccer players

The aim of this study was to investigate the relationships between technique characteristics and knee abduction moments during 90° cuts. A cross sectional design involving 26 elite and sub-elite female soccer players (mean ± SD; age: 21 ± 3.2 years, height: 1.68 ± 0.07 m, and mass: 59.1 ± 6.8 kg) was used to explore relationships between pre-determined technical factors on knee abduction moments during cutting. Three dimensional motion analyses of 90° cuts on the right leg were performed using ‘Qualisys Pro Reflex’ infrared cameras (240 Hz). Ground reaction forces were collected from two AMTI force platforms (1200 Hz) embedded into the running track to examine 2nd last and last footfalls. Pearson’s correlation coefficients, co-efficients of determination and hierarchical multiple regression were used to explore relationships between a range of technique parameters and peak knee abduction moments. Significance was set at p < .05. Hierarchical multiple regression revealed that initial knee abduction angle, lateral leg plant distance and initial lateral trunk lean could explain 67% (62% adjusted) of the variation in peak knee abduction moments (F_(1,22) = 8.869, p = .007). These findings reveal potential modifiable technical factors to lower peak knee abduction moments during cutting.     

Changes in muscular activity and lumbosacral kinematics in response to handling objects of unknown mass magnitude

The aim of this study was to evaluate the main and interaction effects of mass knowledge and mass magnitude on trunk muscular activity and lumbosacral kinematics. Eighteen participants performed symmetric box lifts of three different mass magnitudes (1.1 kg, 5 kg, 15 kg) under known and unknown mass knowledge conditions. Outcome measures were normalized peak electromyography of four trunk muscles in addition to three dimensional lumbosacral angles and acceleration. The results indicated that three out of four muscles exhibited significantly greater activity when handling unknown masses (p < .05). Meanwhile, only sagittal angular acceleration was significantly higher when handling unknown masses (115.6 ± 42.7°/s²) compared to known masses (109.3 ± 31.5°/s²). Similarly, the mass magnitude and mass knowledge interaction significantly impacted the same muscles along with the sagittal lumbosacral angle and angular acceleration (p < .05) with the greatest difference between knowledge conditions being consistently occurring under the 1.1 kg mass magnitude condition. Thus, under these conditions, it was concluded that mass magnitude has more impact than mass knowledge. However, handling objects of unknown mass magnitude could be hazardous, particularly when lifting light masses, in that they can increase mechanical burden on the lumbosacral spine due to increased muscular exertion and acceleration.     

Evaluating feedback time delay during perturbed and unperturbed balance in handstand

Feedback delays in balance are often assessed using muscle activity onset latencies in response to discrete perturbations. The purpose of the study was to calculate EMG latencies in perturbed handstand, and determine if delays are different to unperturbed handstand. Twelve national level gymnasts completed 12 perturbed and 10 unperturbed (five eyes open and five closed) handstands. Forearm EMG latencies during perturbed handstands were assessed against delay estimates calculated via: cross correlations of wrist torque and COM displacement, a proportional and derivative model of wrist torque and COM displacement and velocity (PD model), and a PD model incorporating a passive stiffness component (PS-PD model). Delays from the PD model (161 ± 14 ms) and PS-PD model (188 ± 14 ms) were in agreement with EMG latencies (165 ± 14 ms). Cross correlations of COM displacement and wrist torque provided unrealistically low estimates (5 ± 9 ms). Delays were significantly lower during perturbed (188 ± 14 ms) compared to unperturbed handstand (eyes open: 207 ± 12 ms; eyes closed: 220 ± 19 ms). Significant differences in delays and model parameters between perturbed and unperturbed handstand support the view that balance measures in perturbed testing should not be generalised to unperturbed balance.     

Contextual effects on activity profiles of domestic field hockey during competition and training

Game context is widely accepted to influence performance but most data available is ‘categorical’ and addresses performance rather than activity. This study assessed direct effects of opposition and team ranking in field hockey to establish influences on activity. One hundred and eight (n = 108) female field hockey players (age 16–39 years) participated, giving 186 competition and 48 training analyses. Team average distance (mean ± SD) observed in a mid-ranked team during competition ranged from 5949 ± 611 to 7719 ± 257 m demonstrating an opposition rank effect (Pearson’s r = .71; adjusted R² = .42). However, multiple linear regression analysis demonstrated a stronger relationship in lower ranked teams (Pearson’s r = .99; adjusted R² = .97 through 3–9). In contrast, no team rank effect was observed during randomly monitored competition where team average distance ranged from 5177 ± 444 to 7316 ± 241 m (Pearson’s r = .15 and adjusted R² = .12). In training, however, a team rank effect was observed in distance-related performance indicators where team average distance during small-sided games ranged from 5877 ± 188 to 3551 ± 193 m drill (per 70 min) with Pearson’s r = .95 and adjusted R² = .87. The presence of contextual effects has significant practical implications for team sports where the training load assumed from competition may be overestimated.     

The utility of single-item readiness screeners in middle school

This study examined the benefit of utilizing one-item academic and one-item behavior readiness teacher-rated screeners at the beginning of the school year to predict end-of-school year outcomes for middle school students. The Middle School Academic and Behavior Readiness (M-ABR) screeners were developed to provide an efficient and effective way to assess readiness in students. Participants included 889 students in 62 middle school classrooms in an urban Missouri school district. Concurrent validity with the M-ABR items and other indicators of readiness in the fall were evaluated using Pearson product-moment correlation coefficients, with the academic readiness item having medium to strong correlations with other baseline academic indicators (r = ± 0.56 to 0.91) and the behavior readiness item having low to strong correlations with baseline behavior items (r = ± 0.20 to 0.79). Next, the predictive validity of the M-ABR items was analyzed with hierarchical linear regressions using end-of-year outcomes as the dependent variable. The academic and behavior readiness items demonstrated adequate validity for all outcomes with moderate effects (β = ± 0.31 to 0.73 for academic outcomes and β = ± 0.24 to 0.59 for behavioral outcomes) after controlling for baseline demographics. Even after controlling for baseline scores, the M-ABR items predicted unique variance in almost all outcome variables. Four conditional probability indices were calculated to obtain an optimal cut score, to determine ready vs. not ready, for both single-item M-ABR scales. The cut point of “fair” yielded the most acceptable values for the indices. The odd ratios (OR) of experiencing negative outcomes given a “fair” or lower readiness rating (2 or below on the M-ABR screeners) at the beginning of the year were significant and strong for all outcomes (OR = 2.29 to OR = 14.46), except for internalizing problems. These findings suggest promise for using single readiness items to screen for varying negative end-of-year student outcomes.     

H-reflex excitability is inhibited in soleus,but not gastrocnemius,at the short-latency response of a horizontal jump-landing task

Impaired spinal-level neuromuscular control is suggested to contribute to instability and injury during dynamic landing tasks. Despite this suggestion, spinal-level neuromuscular control is yet to be examined during a horizontal jump-landing task. The aim of the current study was to assess changes in H-reflexes and its reliability at the short-latency response of landings from short and long distances. Eight healthy individuals (five male, three female; age, 22 ± 1.2 yrs; height, 178 ± 8.1 cm; weight, 72 ± 15.7 kg) participated in the study. H-reflexes were evoked at the SLR in the soleus and medial gastrocnemius muscles, during two landing conditions: 25% and 50% of maximal broad jump distance. H-reflexes were expressed relative to the background electromyography (EMG) and maximal M-wave responses (M-max). Soleus H-reflexes were inhibited when landing from shorter distance (25%, 13.9 ± 7.6%; 50%, 8.3 ± 6.5%; p < 0.01). No change in H-reflex excitability was observed in medial gastrocnemius. Background EMG was unaltered across landing conditions. Inhibition of soleus H-reflex excitability from 25% to 50% landing condition indicates a reduced contribution of Ia-afferent feedback to the alpha-motor neuron during landings from greater distances, which may contribute to stiffness regulation at the ankle joint. Unaltered H-reflex excitability of medial gastrocnemius is most likely attributed to its functional role during the landing task.     

Fatigue-induced dissociation between rate of force development and maximal force across repeated rapid contractions

We examined whether the presence of fatigue induced by prolonged running influenced the time courses of force generating capacities throughout a series of intermittent rapid contractions. Thirteen male amateur runners performed a set of 15 intermittent isometric rapid contractions of the knee extensor muscles, (3 s/5 s on/off) the day before (PRE) and immediately after (POST) a half marathon. The maximal voluntary contraction force, rate of force development (RFD_peak), and their ratio (relative RFD_peak) were calculated. At POST, considering the first (out of 15) repetition, the maximal force and RFD_peak decreased (p < 0.0001) at the same extent (by 22 ± 6% and 24 ± 22%, respectively), resulting in unchanged relative RFD_peak (p = 0.6). Conversely, the decline of RFD_peak throughout the repetitions was more pronounced at POST (p = 0.02), thus the decline of relative RFD_peak was more pronounced (p = 0.007) at POST (−25 ± 13%) than at PRE (−3 ± 13%). The main finding of this study was that the fatigue induced by a half-marathon caused a more pronounced impairment of rapid compared to maximal force in the subsequent intermittent protocol. Thus, the fatigue-induced impairment in rapid muscle contractions may have a greater effect on repeated, rather than on single, attempts of maximal force production.     

The influence of the human TMJ eminence inclination on predicted masticatory muscle forces

Aim of this paper was to investigate the change in masticatory muscle forces and temporomandibular joint (TMJ) reaction forces simulated by inverse dynamics when the steepness of the anterior fossa slope was varied. We used the model by de Zee et al. (2007) created in AnyBody™. The model was equipped with 24 musculotendon actuators. Mandibular movement was governed by the trajectory of the incisal point. The TMJ was modelled as a planar constraint canted 5° medially and the caudal inclination relative to the occlusal plane was varied from 10° to 70°. Our models showed that for the two simulated movements (empty chewing and unilateral clenching) the joint reaction forces were smallest for the eminence inclination of 30° and 40° and highest for 70°. The muscle forces were relatively insensitive to change of the eminence inclination for the angles between 20° and 50°. This did not hold for the pterygoid muscle, for which the muscle forces increased continually with increasing fossa inclination. For empty chewing the muscle force reached smaller values than for clenching. During clenching, the muscle forces changed by up to 200 N.     

Effects of short-term fatigue on biomechanical and physiological aspects of double poling in high-level cross-country skiers

The study aim was to evaluate biomechanical and physiological alterations in double poling technique (DP) after a short-term fatiguing exercise. Eight high-level skiers performed a sub-maximal DP trial (20 km h⁻¹, 1°) before (PRE) and after (POST) a DP test to exhaustion while roller skiing on a treadmill. An integrated analysis of DP technique during PRE and POST included measurement of pole, joint, and centre of mass (COM) kinematics, poling forces, cycle timing, and metabolic parameters. Muscle fatigue in three upper-body muscles was assessed by calculating the Dimitrov’ fatigue index (FI_nms5) of specific electromyographic segments. FI_nms5 tended to increase in the latissimus dorsi and teres major muscles (P = 0.023 and P = 0.030, respectively) across consecutive DP cycles, as did blood lactate concentration (P = 0.001) and rating of perceived exertion (P = 0.005). The changes indicated a state of fatigue during POST and coincided with the reduction in poling force exertion capacity (P = 0.020). Pole, joint and COM kinematics did not differ between PRE and POST (P > 0.050), whereas recovery phase and cycle times were shorter at POST (P < 0.001 and P = 0.001, respectively). Short-term fatigue led to a reduction in poling force exertion capacity and cycle time in high-level skiers, without altering body and pole kinematics.     

Profiling movement quality and gait characteristics according to body-mass index in children (9–11 y)

Obese children move less and with greater difficulty than their normal-weight counterparts. Whilst the effect of high BMI on cardiovascular fitness is well known, the effect on movement quality characteristics during a standardised fitness test has not been investigated. The aims of this study were, to characterise the movement quality of children performing the multi-stage fitness test (MSFT), and, report how movement quality characteristics cluster according to weight status. One hundred and three children (10.3 ± 0.6 y, 1.42 ± 0.08 m, 37.8 ± 9.3 kg, BMI; 18.5 ± 3.3 kg m²) performed the MSFT whilst wearing an ankle mounted accelerometer. BMI groups were used to classify children as underweight (UW), normal weight (NW), overweight (OW) and obese (OB). Characteristics of movement were profiled using a clustering algorithm. Spearman’s rho was used to assess relationship with BMI group, and a Mann-Whitney U test was used to assess differences between BMI groups. Obese children had significantly lower spectral purity than every other group and significantly lower time to exhaustion (TTE) than UW and NW children (P < 0.05). BMI was clustered with stride profile and TTE with spectral purity. Significant negative correlations (P < 0.05) were found between BMI and TTE (r = −0.25), spectral purity (r = −0.24), integrated acceleration (r = −0.22), stride angle (r = −0.23) and stride variability (r = −0.22). This was the first study to report the spectral purity of children’s gait. Further analysis unveiled key performance characteristics that differed between BMI groups. These were (i) representative of children’s performance during the MSFT and, (ii) significantly negatively correlated with BMI.