Analyzing gait parameters in trail runners using wireless trunk accelerometry during real-world and treadmill incline running

Bergh, Oloff Charles Wessel (2020-12)

Thesis (M Sport Sc)--Stellenbosch University, 2020.

Thesis

ENGLISH ABSTRACT: Few studies have explored dynamic stability when running over complex or challenging surfaces, and no studies have investigated how trail terrain could affect components of dynamic stability. The aim of this study was to measure the acute changes in dynamic stability when running at incline, between treadmill and trail surfaces. Twelve recreational trail runners (age 25.2 ±2.6 years; mass 78.8 ±5.9 kg; height 183.6 ±7.1 cm) participated and completed all aspects of testing. They ran at 10 km.h-1with an eight-degree incline, over both treadmill and trail surfaces. Each participant had a single Noraxon®myoMotion Research PRO inertial measurement unit (IMU) attached to their third lumbar vertebrae region, capable of collecting wireless acceleration data. Linear acceleration data was captured up to 200 Hz and ± 16 g at the trunk region in three-dimensions, namely the vertical (VT), anterior-posterior (AP) and mediolateral (ML). Data was streamed to the Noraxon®myo RESEARCH software. Thereafter, the data was filtered using a zero-lag 4thorder low-pass Butterworth filter with a cut-off frequency of 50 Hz. Filtered acceleration data was imported into MATLAB R2020a (Version 9.6), with a custom written code performing an autocorrelation procedure of each participant over both treadmill and trail surfaces. The autocorrelations provided information regarding the step and stride regularity, as well as the symmetry of the individual over the two terrains, based on the three-dimensional accelerations at the trunk. Furthermore, mean step and stride times, as well as their coefficients of variations (CV) were calculated from the filtered data. Results were reported in the article (Chapter Four) and indicated that step and stride regularity was decreased (p< 0.01) in all three-dimensions when running over the more complex trail surface, compared to the steady treadmill surface. The AP and ML directions indicated a greater degree of diminution compared to the VT and is evident in the symmetry values. Symmetry decreased over the trail surface for both the AP (z= -3.06, p< 0.01) and ML (p< 0.01) directions, but not in the VT (z= -1.65, p= 0.10) direction. Additionally, there was no change in mean step (p= 0.45) and stride (p= 0.33) times, but a significant increase was observed for both step CV (p< 0.01) and stride CV (p< 0.01) when running on the trail surface. The first null hypothesis was rejected, as the coefficients of variation for both step and stride times indicated a significant difference when comparing the treadmill and trail surfaces. The second null hypothesis was rejected, as the trail surface did indicate a general decrease in dynamic stability components compared to the treadmill. In conclusion the trail demonstrated a higher degree of step and stride variability, and low symmetry, primarily due to the inconsistent nature of the trail surface. Future studies could investigate the role of cognition during trail running, by examining the decision-making process while traversing complex terrain such as the trail environment. Furthermore, future studies in the field of sports biomechanics could aim to incorporate a greater degree of software technology, such as adopting a more algorithmic approach to analysing data.

AFRIKAANSE OPSOMMING: Min studies het voorheen dinamiese stabiliteit ondersoek wanneer individue oor ingewikkelde of uitdagende oppervlaktes hardloop en geen studies het ondersoek ingestel na hoe die terrein van die voetslaanpad, komponente van dinamiese stabiliteit kan beïnvloed. Die doel van hierdie studie was om die akute veranderinge in dinamiese stabiliteit te meet wanneer individue teen ‘n helling, op trapmeul en voetslaanpaaie hardloop.Twaalf amateur voetslaanpad hardlopers (ouderdom 25.2 ± 2.6 jaar; massa 78.8 ± 5.9 kg; hoogte 183.6 ± 7.1 cm) het deelgeneem en alle aspekte van die toetsing voltooi. Hulle het met 'n helling van agt grade teen 10 km.h-1gehardloop, oor beide trapmeul en voetslaanpad oppervlaktes. Elke deelnemer het 'n enkele Noraxon® myoMotionResearch PRO –traagheidsmetingeenheid (IMU) gekoppel aan hul derde lumbale werwel gedra, wat draadlose versnellingsdata kon versamel. Die IMU kon in drie dimensies lineêre versnellingsdata tot 200 Hz en ± 16 g opneem, naamlik die vertikale (VT), anteroposterior (AP) en mediolaterale (ML). Data is na die Noraxon® myoRESEARCH gestroom. Daarna is die data gefiltreer met behulp van 'n nul-lag 4de-orde laagdeurlaat Butterworth-filter met 'n afsny frekwensie van 50 Hz. Die gefiltreerde versnellingsdata is in MATLAB R2020a (Weergawe 9.6) ingevoer, met 'n self-geskrewe kode wat 'n outokorrelasie prosedure van elke deelnemer oor beide die trapmeul en voetslaanpad oppervlaktes uitvoer. Die outokorrelasies het inligting oor die tree en stap reëlmatigheid, sowel as die simmetrie van die individu oor die twee terreine gegee, gebasseer op die drie-dimensionele versnellings by die romp. Verdere berekeninge van die gemiddelde tree en stap tye, sowel as hul koëffisiënte van variasies (CV), was bereken uit die gefiltreerde data. Resultate is in die artikel (Hoofstuk Vier) gerapporteer en het aangedui dat die reëlmatigheid van tree en stap in alle drie-afmetings (p< 0.01) verminder is wanneer die individue oor die meer ingewikkelde voetslaanpad hardloop, in vergelyking met die stabiele trapmeul oppervlakte. Die AP en ML rigtings van meting, het aangedui op 'n groter mate van afname in vergelyking met die VT en kan waargeneem word in die simmetriewaardes. Simmetrie het oor die voetslaanpad afgeneem vir beide die AP (z= -3.06, p< 0.01) en ML (p< 0.01) rigtings, maar nie in die VT (z= -11.65, p= 0.10) rigting nie. Boonop was daar geen verandering in die gemiddelde tree (p= 0.45) en die stap (p= 0.33) tye nie, maar 'n beduidende toename is waargeneem vir beide tree CV (p< 0.01) en stap CV (p< 0.01) tydens hardloop op die voetslaanpad oppervlakte. Die eerste null hipotese is nie aanvaar nie, aangesien die koëffisiënte van variasies vir die stap tye en tree tye 'n beduidende verskil in die vergelykings tussen die trapmeul oppervlak en die voetslaanpad oppervlakte. Die eerste nul hipotese is nie aanvaar nie, aangesien die koëffisiënte van variasies vir die stap tye en tree tye 'n beduidende verskil in die vergelykings tussen die trapmeul oppervlak en die voetslaanpad aandui. Die tweede nul hipotese is nie aanvaar nie, aangesien die voetslaanpad se oppervlak wel 'n algemene afname in dinamiese stabiliteit toon, in vergelyking met die trapmeul. Ter afsluiting het die voetslaanpad 'n hoër mate van tree en stap veranderlikheid en 'n lae simmetrie getoon, hoofsaaklik as gevolg van die inkonsekwente aard van die spooroppervlak. Toekomstige studies kan die rol van kognisie tydens voetslaanpad hardloop ondersoek, deur die besluitnemingsproses te ondersoek terwyl komplekse terrein, soos die spooromgewing, gekruis word. Verder kan toekomstige studies in die gebied van sportbiomeganika ten doel hê om 'n groter mate van sagteware tegnologie te inkorporeer, soos om 'n meer algoritmiese benadering tot die ontleding van data te gebruik.

Please refer to this item in SUNScholar by using the following persistent URL: http://hdl.handle.net/10019.1/109299
This item appears in the following collections: