ArticlePDF Available

Abstract and Figures

This study aimed to determine the minimum time required for assessing spatiotemporal variability during continuous running at different submaximal velocities and, thereby, the number of steps required. Nineteen trained endurance runners performed an incremental running protocol, with a 3-min recording period at 10, 12, 14 and 16 km/h. Spatiotemporal parameters (contact and flight times, step length and step frequency) were measured using the OptoGait system and step variability was considered for each parameter, in terms of within-participants standard deviation (SD) and coefficient of variation (CV%). Step variability was considered over six different durations at every velocity tested: 0-10s, 0-20s, 0-30s, 0-60s, 0-120s and 0-180s. The repeated measures ANOVA revealed no significant differences in the magnitude of the four spatiotemporal parameters between the recording intervals at each running velocity tested (p≥0.05, ICC>0.99). The post-hoc analysis confirmed no significant differences in step variability (SD and CV% of each spatiotemporal parameter at any velocity tested) between measurements. The Bland-Altman limits of agreement method showed that longer recording intervals yield smaller systematic bias, random errors, and narrower limits of agreement, regardless of running velocity. The results suggest that the duration of the recording period required to estimate spatiotemporal variability plays an important role in the accuracy of the measurement, regardless of running velocity (10-16 km/h).
Content may be subject to copyright.
Dear author,
Please note that changes made in the online proofing system will
be added to the article before publication but are not reflected in
this PDF.
We also ask that this file not be used for submitting corrections.
1
2Short communication
4Minimum time required for assessing step variability during running at
5submaximal velocities
6
7
8Felipe García-Pinillos
a,
, Pedro A. Latorre-Román
b
, Rodrigo Ramírez-Campillo
c
, Juan A. Párraga-Montilla
b
,
9Luis E. Roche-Seruendo
d
10
a
Department of Physical Education, Sports and Recreation, Universidad de La Frontera, Temuco, Chile
11
b
University of Jaen, Department of Corporal Expression, Campus de Las Lagunillas s/n, D2 Building, Dep. 142, 23071 Jaen, Spain
12
c
Department of Physical Activity Sciences, Research Nucleus in Health, Physical Activity and Sport, Univesidad de Los Lagos, Osorno, Chile
13
d
Universidad San Jorge, Campus Universitario, A23 km 299, 50830, Villanueva de Gállego, Zaragoza, Spain
14
16
article info
17 Article history:
18 Accepted 4 September 2018
19 Available online xxxx
20 Keywords:
21 Biomechanics
22 Gait variability
23 Movement variability
24 Running
25
26
abstract
27
This study aimed to determine the minimum time required for assessing spatiotemporal variability dur-
28
ing continuous running at different submaximal velocities and, thereby, the number of steps required.
29
Nineteen trained endurance runners performed an incremental running protocol, with a 3-min recording
30
period at 10, 12, 14 and 16 km/h. Spatiotemporal parameters (contact and flight times, step length and
31
step frequency) were measured using the OptoGait system and step variability was considered for each
32
parameter, in terms of within-participants standard deviation (SD) and coefficient of variation (CV%). Step
33
variability was considered over six different durations at every velocity tested: 0–10 s, 0–20 s, 0–30 s, 0–
34
60 s, 0–120 s and 0–180 s. The repeated measures ANOVA revealed no significant differences in the mag-
35
nitude of the four spatiotemporal parameters between the recording intervals at each running velocity
36
tested (p 0.05, ICC > 0.90). The post-hoc analysis confirmed no significant differences in step variability
37
(SD and CV% of each spatiotemporal parameter at any velocity tested) between measurements. The
38
Bland-Altman limits of agreement method showed that longer recording intervals yield smaller system-
39
atic bias, random errors, and narrower limits of agreement, regardless of running velocity. The results
40
suggest that the duration of the recording period required to estimate spatiotemporal variability plays
41
an important role in the accuracy of the measurement, regardless of running velocity (10–16 km/h).
42
Ó2018 Elsevier Ltd. All rights reserved.
43
44
45
46
1. Introduction
47
The influence of running velocity on spatiotemporal gait charac-
48
teristics has been previously examined in the literature (Ogueta-
49
Alday et al., 2014; Padulo et al., 2012; Roche-Seruendo et al.,
50
2018). The overall decrease in contact time (CT) and increase in
51
flight time (FT), step length (SL), and step frequency (SF) with
52
increasing running speed has been previously shown (Ogueta-
53
Alday et al., 2014; Padulo et al., 2012; Roche-Seruendo et al.,
54
2018). Nevertheless, the evidence about the influence of running
55
velocity on step variability is quite limited.
56
Step variability seems to be related to both injuries (Hamill
57
et al., 2012; Meardon et al., 2011) and endurance performance
58
(Nakayama et al., 2010), although its accurate assessment remains
59
problematic. In 1995, Belly and colleagues (Belli et al., 1995) indi-
60
cated that step variability during running was difficult to estimate
61
due to the lack of measurement devices. Today, many devices pro-
62
vide real-time feedback on spatiotemporal parameters while run-
63
ning (e.g., OptoGait
TM
system). Therefore, nowadays the limitation
64
is not how to collect the data, but how long should last the data
65
collection period to obtain accurate assessments of step variability.
66
Belli et al. (1995) suggested that 32–64 consecutive steps are
67
required to assess step variability, which represents approximately
68
15–20 s when running at submaximal velocities. To the best of the
69
authors’ knowledge, no more studies have reconsidered this topic
70
during running, adapted to the new devices. A recent study exam-
71
ined the minimum number of steps required to accurately estimate
72
spatial and temporal step kinematic variability of subjects walking
73
on a treadmill (Owings and Grabiner, 2003), concluding that at
74
least 400 steps are required. Likewise, König et al. (2014) proposed
https://doi.org/10.1016/j.jbiomech.2018.09.005
0021-9290/Ó2018 Elsevier Ltd. All rights reserved.
Corresponding author at: Department of Physical Education, Sport and Recre-
ation, Universidad de La Frontera (Temuco, Chile), Calle Uruguay, 1980 Temuco,
Chile.
E-mail addresses: fegarpi@gmail.com (F. García-Pinillos), platorre@ujaen.es
(P.A. Latorre-Román), r.ramirez@ulagos.cl (R. Ramírez-Campillo), jparraga@ujaen.
es (J.A. Párraga-Montilla), leroche@usj.es (L.E. Roche-Seruendo).
Journal of Biomechanics xxx (2018) xxx–xxx
Contents lists available at ScienceDirect
Journal of Biomechanics
journal homepage: www.elsevier.com/locate/jbiomech
www.JBiomech.com
BM 8852 No. of Pages 10, Model 5G
14 September 2018
Please cite this article in press as: García-Pinillos, F., et al. Minimum time required for assessing step variability during running at submaximal velocities. J.
Biomech. (2018), https://doi.org/10.1016/j.jbiomech.2018.09.005
75
the collection of at least 50 cycles to reliably assess gait variability
76
during walking at a self-selected speed; whereas Bruijn et al.
77
(2009) concluded that longer data series (>150 strides) led to more
78
precise estimates for measures related to step variability. To fur-
79
ther examine the step variability during running, the aim of this
80
study is to determine the minimum time required for assessing
81
spatiotemporal variability during continuous running on an instru-
82
mented treadmill and, thereby, the number of steps required, as
83
well as to analyse the influence of running velocity.
84
2. Materials and methods
85
2.1. Participants
86
Nineteen trained male endurance runners (age: 34 ± 7 years;
87
height: 1.76 ± 0.05 m; body mass: 70.5 ± 6.2 kg) participated in
88
this study. Participants met the inclusion criteria: (1) older than
89
18 years old, (2) able to run 10 km in < 40 min, (3) training on a
90
treadmill at least once per week, (4) free from injury (points 3
91
and 4 refer to the 6 months preceding the study). After receiving
92
information on the objectives and procedures of the study, partic-
93
ipants signed an informed consent form, which complied with the
94
ethical standards of the World Medical Association’s Declaration of
95
Helsinki (2013). The study was approved by the local ethics
96
committee.
97
2.2. Procedures
98
Participants were tested on one day on a motorized treadmill
99
(HP cosmos Pulsar 4P, HP cosmos Sports & Medical, Gmbh, Ger-
100
many). A standardized 10-min warm-up (running at 10 km/h)
101
was performed since previous studies on human locomotion have
102
shown that accommodation to a new condition occurs in 6–8 min
103
(Schieb, 1986). After warming-up, four intervals were run for 3-
104
min at 10, 12, 14 and 16 km/h with a complete recording period.
105
Note that those running velocities are normal paces for these ath-
106
letes in both training and competition contexts. All participants
107
verbally reported feeling comfortable running at each set speed.
108
In order to control the potential effect of fatigue, and consistently
109
with a previous study (Mann et al., 2015), the intensity was mea-
110
sured using the 6–20 Borg scale (Borg, 1982) at the end of each
111
3-min acquisition period. If a participant indicated a score of 17
112
at any point, the remaining trials were not performed. Of note,
113
no participants indicated a score >16 in the Borg scale during the
114
protocol.
115
2.3. Materials and testing
116
Spatiotemporal parameters were measured using the Opto-
117
Gait
TM
system (Optogait; Microgate, Bolzano, Italy), which was pre-
118
viously validated for the assessment of spatiotemporal parameters
119
of the gait of young adults (Lee et al., 2014). The OptoGait
TM
system
120
is able to measure both CT and FT at 1000 Hz. The two parallel bars
121
of the OptoGait
TM
system were placed on the side edges of the
122
treadmill at the same level of the contact surface. The CT, FT, SL
123
and SF were measured for every step (Roche-Seruendo et al., 2018).
124
Step variability was assessed for each spatiotemporal parame-
125
ter through the within-participant standard deviation (SD) and
126
the coefficient of variation (CV%). Since previous studies have used
127
indistinctly the SD (Paquette et al., 2017) or CV% (Nakayama et al.,
128
2010), we incorporated both measures to make comparisons
129
easier. Step variability was examined over six recording intervals
130
within the 3-min recording period at each running velocity tested:
131
0–10 s, 0–20 s, 0–30 s, 0–60 s, 0–120 s and 0–180 s.
132
2.4. Statistical analysis
133
Descriptive statistics are represented as mean (SD). Tests of nor-
134
mal distribution and homogeneity (Kolmogorov-Smirnov and
135
Levene’s test, respectively) were conducted on all data before anal-
136
ysis. One-way repeated measures ANOVA with Bonferroni post-hoc
137
corrections were conducted on the magnitude of each spatiotem-
138
poral parameter as well as on variability outcomes (i.e., SD and
139
CV%) to examine possible differences between the recording inter-
140
vals (0–10 s, 0–20 s, 0–30 s, 0–60 s, 0–120 s, 0–180 s) at each run-
141
ning velocity (10, 12, 14 and 16 km/h). The association of the
142
magnitude and variability of the spatiotemporal parameters
143
between the recording intervals was quantified through the intra-
144
class correlation coefficient (ICC). Based on the characteristics of
145
this experimental design and following the guidelines reported
146
by Koo and Li (2016), the authors decided to conduct a ‘‘two-way
147
random-effects” model (ICC [2,k]), ‘‘mean of measurements” type,
148
and ‘‘absolute” definition for the ICC measurement. The interpreta-
149
tion of the ICC was based on the benchmarks reported by a previ-
150
ous study (Landis and Koch, 1977): ICC < 0 reflects ‘poor’ reliability,
151
0–0.20 ‘slight’, 0.21–0.40 ‘fair’, 0.41–0.60 ‘moderate’, 0.61–0.80
152
‘substantial’, and >0.81 ‘almost perfect’ reliability. The Bland-
153
Altman (Bland and Altman, 1995) limits of agreement method
154
(mean difference ± 1.96 SD) was used to examine differences in
155
step variability (i.e., CV%) between each recording interval (0–
156
10 s, 0–20 s, 0–30 s, 0–60 s and 0–120 s) and the longest interval
157
(0–180 s) for each spatiotemporal parameter (CT, FT, SL and SF)
158
at each running velocity. The level of significance used was
159
p < 0.05. Data analysis was performed using the SPSS (version 21,
160
SPSS Inc., Chicago, Ill).
161
3. Results
162
The ANOVAs showed no significant differences in the magni-
163
tude of the spatiotemporal parameters between the recording
164
intervals at 10 km/h (p = 0.058 for CT, p = 0.051 for FT, p = 0.232
165
for SL and p = 0.573 for SF), at 12 km/h (p = 0.454 for CT,
166
p = 0.396 for FT, p = 0.433 for SL and p = 0.671 for SF), at 14 km/h
167
(p = 0.072 for CT, p = 0.113 for FT, p = 0.409 for SL and p = 0.506
168
for SF) or at 16 km/h (p = 0.111 for CT, p = 0.110 for FT, p = 0.056
169
for SL and p = 0.284 for SF). An almost perfect association was
170
observed in the magnitude of the four spatiotemporal parameters
171
between the recording intervals compared to the 0–180 s interval,
172
at each running velocity tested (ICC > 0.90) (Table 1).
173
The ANOVAs conducted on the within-participants SD (Table 2)
174
revealed no significant differences (p 0.05) for any spatiotempo-
175
ral parameter at 10 and 16 km/h. At 12 km/h significant differences
176
for CT (p = 0.035), FT (p = 0.007) and SF (p = 0.035) were found,
177
while no significant differences were observed for SL (p = 0.167).
178
At 14 km/h significant differences for CT (p = 0.010) were found,
179
with no significant differences for FT (p = 0.133), SL (p = 0.166)
180
nor SF (p = 0.113). Bonferroni pairwise comparisons never reached
181
statistical significance. An almost perfect association between the
182
recording intervals compared to the 0–180 s interval was observed
183
for the within-participants SD at each running speed tested
184
(ICC > 0.90).
185
Regarding the ANOVAs conducted on CV% values (Table 3), no
186
significant differences (p 0.05) for any spatiotemporal parameter
187
at 10 km/h were found. At 12 km/h, significant differences were
188
observed for FT (p = 0.025) and SF (p = 0.036), but not for CT
189
(p = 0.155) and SL (p = 0.172). At 14 km/h, only CT reported signif-
190
icant differences (p = 0.013), whereas at 16 km/h, only FT reached
191
statistical significance (p = 0.014). At both 14 and 16 km/h, no
192
other parameters showed significant differences (p 0.05). Bon-
193
ferroni pairwise comparisons never reached statistical significance.
2F. García-Pinillos et al. / Journal of Biomechanics xxx (2018) xxx–xxx
BM 8852 No. of Pages 10, Model 5G
14 September 2018
Please cite this article in press as: García-Pinillos, F., et al. Minimum time required for assessing step variability during running at submaximal velocities. J.
Biomech. (2018), https://doi.org/10.1016/j.jbiomech.2018.09.005
Table 1
Descriptive values and association between the magnitude of the spatiotemporal parameters obtained from six time intervals at every running velocity tested.
Mean (standard deviation) ICC (0–10 s vs 0–180 s)
(95% CI)
ICC (0–20 s vs 0–180 s)
(95% CI)
ICC (0–30 s vs 0–180 s)
(95% CI)
ICC (0–60 s vs 0–180 s)
(95% CI)
ICC (0–120 s vs 0–180 s)
(95% CI)
0–10 s 0–20 s 0–30 s 0–60 s 0–120 s 0–180 s
10 km/h CT (s) 0.294
(0.023)
0.294
(0.023)
0.295
(0.022)
0.296
(0.022)
0.297
(0.021)
0.297
(0.020)
0.974 (0.909–0.991) 0.982 (0.924–0.994) 0.986 (0.951–0.995) 0.993 (0.983–0.997) 0.998 (0.996–0.999)
FT (s) 0.072
(0.026)
0.071
(0.025)
0.070
(0.025)
0.068
(0.024)
0.067
(0.024)
0.067
(0.023)
0.960 (0.867–0.986) 0.969 (0.898–0.989) 0.974 (0.927–0.990) 0.987 (0.967–0.995) 0.997 (0.992–0.999)
SL (cm) 101.42
(4.27)
101.35
(4.13)
101.26
(4.07)
101.10
(4.04)
101.00
(4.23)
101.01
(4.30)
0.985 (0.958–0.994) 0.987 (0.966–0.995) 0.988 (0.970–0.995) 0.993 (0.982–0.997) 0.999 (0.997–0.999)
SF (step/
min)
164.83
(7.22)
164.98
(7.05)
165.19
(6.94)
165.33
(6.78)
165.44
(6.97)
165.40
(7.04)
0.985 (0.961–0.994) 0.988 (0.970–0.995) 0.987 (0.967–0.995) 0.993 (0.982–0.996) 0.999 (0.996–0.999)
12 km/h CT (s) 0.263
(0.019)
0.264
(0.018)
0.264
(0.018)
0.265
(0.017)
0.265
(0.017)
0.265
(0.017)
0.981 (0.951–0.993) 0.987 (0.965–0.995) 0.991 (0.976–0.997) 0.997 (0.992–0.999) 0.998 (0.997–0.999)
FT (s) 0.090
(0.025)
0.090
(0.024)
0.090
(0.024)
0.089
(0.023)
0.089
(0.022)
0.089
(0.022)
0.961 (0.900–0.985) 0.973 (0.931–0.990) 0.980 (0.950–0.952) 0.993 (0.981–0.997) 0.999 (0.998–1.00)
SL (cm) 118.09
(5.34)
118.04
(5.23)
118.09
(5.20)
117.99
(5.09)
117.95
(5.11)
117.99
(5.16)
0.996 (0.913–0.988) 0.980 (0.947–0.992) 0.984 (0.960–0.994) 0.994 (0.984–0.998) 0.999 (0.998–0.999)
SF (step/
min)
170.1
(7.58)
170.1
(7.56)
170.0
(7.50)
170.1
(7.34)
170.1
(7.36)
170.0
(7.41)
0.968 (0.917–0.988) 0.979 (0.942–0.996) 0.985 (0.961–0.994) 0.994 (0.984–0.998) 0.999 (0.998–1.00)
14 km/h CT (s) 0.237
(0.015)
0.237
(0.015)
0.238
(0.015)
0.239
(0.014)
0.239
(0.013)
0.239
(0.013)
0.971 (0.921–0.989) 0.982 (0.952–0.993) 0.989 (0.973–0.996) 0.997 (0.992–0.999) 0.999 (0.998–1.00)
FT (s) 0.108
(0.021)
0.107
(0.020)
0.106
(0.020)
0.105
(0.020)
0.105
(0.019)
0.105
(0.018)
0.950 (0.869–0.981) 0.969 (0.921–0.989) 0.979 (0.948–0.992) 0.993 (0.982–0.997) 0.999 (0.997–1.00)
SL (cm) 134.32
(5.85)
134.05
(5.81)
133.92
(5.93)
133.65
(5.98)
133.66
(5.88)
133.60
(5.70
0.971 (0.924–0.989) 0.984 (0.959–0.994) 0.989 (0.979–0.996) 0.995 (0.987–0.998) 0.999 (0.997–1.00)
SF (step/
min)
174.38
(7.56)
174.61
(7.48)
174.71
(7.56)
174.97
(7.60)
174.90
(7.42)
174.94
(7.18)
0.972 (0.929–0.989) 0.983 (0.957–0.993) 0.987 (0.968–0.995) 0.995 (0.986–0.998) 0.999 (0.997–1.00)
16 km/h CT (s) 0.216
(0.014)
0.216
(0.014)
0.216
(0.014)
0.216
(0.013)
0.216
(0.013)
0.217
(0.013)
0.991 (0.976–0.997) 0.994 (0.982–0.998) 0.989 (0.969–0.996) 0.996 (0.990–0.999) 0.998 (0.993–0.999)
FT (s) 0.119
(0.021)
0.118
(0.021)
0.118
(0.021)
0.117
(0.020)
0.117
(0.020)
0.116
(0.019)
0.972 (0.913–0.991) 0.984 (0.948–0.995) 0.987 (0.956–0.995) 0.996 (0.987–0.998) 0.998 (0.996–0.999)
SL (cm) 149.01
(6.87)
148.59
(6.98)
148.42
(6.96)
148.08
(6.95)
147.90
(6.86)
147.83
(6.65)
0.977 (0.904–0.992) 0.985 (0.955–0.995) 0.990 (0.971–0.996) 0.996 (0.989–0.998) 0.999 (0.998–0.999)
SF (step/
min)
179.60
(8.41)
180.03
(8.47)
180.25
(8.48)
180.63
(8.52)
180.84
(8.46)
180.91
(8.26)
0.977 (0.920–0.992) 0.986 (0.959–0.995) 0.991 (0.974–0.997) 0.997 (0.991–0.999) 0.999 (0.998–0.999)
CT, contact time; FT, flight time; SL, step length; SF, step frequency; ICC, intraclass correlation coefficient; 95% CI, 95% confidence interval.
F. García-Pinillos et al. / Journal of Biomechanics xxx (2018) xxx–xxx 3
BM 8852 No. of Pages 10, Model 5G
14 September 2018
Please cite this article in press as: García-Pinillos, F., et al. Minimum time required for assessing step variability during running at submaximal velocities. J.
Biomech. (2018), https://doi.org/10.1016/j.jbiomech.2018.09.005
Table 2
Descriptive values and association observed for the within-participant standard deviation of the spatiotemporal parameters obtained from six time intervals at every running velocity tested.
Mean (standard deviation) ICC (0–10 s vs 0–180 s)
(95% CI)
ICC (0–20 s vs 0–180 s)
(95% CI)
ICC (0–30 s vs 0–180 s)
(95% CI)
ICC (0–60 s vs 0–180 s)
(95% CI)
ICC (0–120 s vs 0–180 s)
(95% CI)
0–10 s 0–20 s 0–30 s 0–60 s 0–120 s 0–180 s
10 km/h CT (s) 0.008
(0.006)
0.008
(0.005)
0.008
(0.006)
0.008
(0.004)
0.008
(0.004)
0.009
(0.003)
0.908 (0.799–0.953) 0.928 (0.814–0.972) 0.911 (0.804–0.958) 0.946 (0.891–0.979) 0.978 (0.943–0.991)
FT (s) 0.010
(0.005)
0.010
(0.004)
0.010
(0.003)
0.011
(0.003)
0.011
(0.003)
0.011
(0.002)
0.901 (0.800–0.952) 0.913 (0.835–0.924) 0.902 (0.804–0.950) 0.910 (0.769–0.965) 0.942 (0.854–0.978)
SL (cm) 3.60
(1.12)
3.50
(1.12)
3.46
(1.07)
3.43
(1.08)
3.43
(1.02)
3.42
(0.91)
0.902 (0.824–0.927) 0.905 (0.791–0.958) 0.932 (0.823–0.974) 0.978 (0.943–0.992) 0.990 (0.974–0.996)
SF (step/
min)
5.30
(3.12)
5.28
(2.54)
5.74
(4.36)
5.45
(3.47)
5.27
(2.56)
5.22
(2.09)
0.943 (0.852–0.978) 0.977 (0.941–0.991) 0.904 (0.798–0.937) 0.931 (0.822–0.973) 0.988 (0.968–0.995)
12 km/h CT (s) 0.006
(0.002)
0.006
(0.002)
0.007
(0.002)
0.007
(0.002)
0.007
(0.002)
0.007
(0.002)
0.905 (0.845–0.937) 0.907 (0.853–0.962) 0.962 (0.903–0.985) 0.985 (0.962–0.994) 0.996 (0.990–0.998)
FT (s) 0.008
(0.002)
0.008
(0.002)
0.009
(0.002)
0.009
(0.002)
0.009
(0.001)
0.009
(0.002)
0.904 (0.808–0.907) 0.901 (0.766–0.922) 0.909 (0.873–0.950) 0.936 (0.833–0.975) 0.991 (0.976–0.996)
SL (cm) 3.52
(1.18)
3.49
(0.98)
3.66
(0.96)
3.63
(0.95)
3.56
(0.96)
3.56
(0.91)
0.909 (0.850–0.929) 0.940 (0.847–0.977) 0.957 (0.890–0.983) 0.963 (0.907–0.986) 0.988 (0.969–0.995)
SF (step/
min)
4.65
(1.54)
4.59
(1.44)
4.88
(1.38)
4.77
(1.29)
4.76
(1.30)
4.72
(1.27)
0.909 (0.861–0.941) 0.933 (0.829–0.974) 0.967 (0.915–0.987) 0.980 (0.948–0.992) 0.995 (0.987–0.998)
14 km/h CT (s) 0.005
(0.001)
0.005
(0.002)
0.005
(0.002)
0.006
(0.002)
0.006
(0.002)
0.006
(0.002)
0.901 (0.828–0.963) 0.906 (0.827–0.941) 0.954 (0.851–0.986) 0.979 (0.947–0.992) 0.994 (0.986–0.998)
FT (s) 0.007
(0.002)
0.007
(0.002)
0.008
(0.002)
0.008
(0.002)
0.008
(0.002)
0.008
(0.002)
0.903 (0.801–0.925) 0.904 (0.803–0.911) 0.908 (0.878–0.956) 0.921 (0.800–0.921) 0.985 (0.963–0.994)
SL (cm) 2.99
(1.02)
3.14
(1.10)
3.27
(1.40)
3.41
(1.59)
3.41
(1.54)
3.49
(1.44)
0.901 (0.795–0.917) 0.904 (0.808–0.907) 0.960 (0.890–0.985) 0.978 (0.943–0.991) 0.994 (0.993–0.998)
SF (step/
min)
3.80
(1.21)
4.03
(1.41)
4.08
(1.41)
4.26
(1.56)
4.33
(1.47)
4.43
(1.31)
0.901 (0.786–0.955) 0.920 (0.741–0.972) 0.940 (0.795–0.979) 0.967 (0.915–0.987) 0.989 (0.972–0.996)
16 km/h CT (s) 0.005
(0.002)
0.005
(0.002)
0.005
(0.002)
0.005
(0.003)
0.005
(0.003)
0.006
(0.003)
0.908 (0.836–0.942) 0.911 (0.860–0.951) 0.917 (0.853–0.971) 0.930 (0.816–0.954) 0.977 (0.936–0.992)
FT (s) 0.007
(0.002)
0.007
(0.002)
0.007
(0.002)
0.007
(0.003)
0.007
(0.002)
0.007
(0.002)
0.907 (0.855–0.955) 0.901 (0.717–0.966) 0.968 (0.890–0.989) 0.968 (0.915–0.988) 0.993 (0.982–0.998)
SL (cm) 3.10
(1.26)
3.11
(1.30)
3.01
(1.04)
2.89
(0.80)
2.89
(0.80)
2.93
(0.74)
0.908 (0.858–0.941) 0.912 (0.808–0.944) 0.927 (0.806–0.972) 0.967 (0.912–0.987) 0.983 (0.956–0.994)
SF (step/
min)
3.75
(1.08)
3.80
(0.99)
3.80
(0.89)
3.85
(0.86)
4.05
(1.23)
4.27
(1.18)
0.908 (0.804–0.959) 0.912 (0.810–0.960) 0.909 (0.788–0.960) 0.911 (0.750–0.976) 0.936 (0.827–0.976)
CT, contact time; FT, flight time; SL, step length; SF, step frequency; ICC, intraclass correlation coefficient; 95% CI, 95% confidence interval.
4F. García-Pinillos et al. / Journal of Biomechanics xxx (2018) xxx–xxx
BM 8852 No. of Pages 10, Model 5G
14 September 2018
Please cite this article in press as: García-Pinillos, F., et al. Minimum time required for assessing step variability during running at submaximal velocities. J.
Biomech. (2018), https://doi.org/10.1016/j.jbiomech.2018.09.005
Table 3
Descriptive values and association observed for the coefficient of variation (%) of the spatiotemporal parameters obtained from six time intervals at every running velocity tested.
Mean (standard deviation) ICC (0–10 s vs 0–180 s)
(95% CI)
ICC (0–20 s vs 0–180 s)
(95% CI)
ICC (0–30 s vs 0–180 s)
(95% CI)
ICC (0–60 s vs 0–180 s)
(95% CI)
ICC (0–120 s vs 0–180 s)
(95% CI)
0–10 s 0–20 s 0–30 s 0–60 s 0–120 s 0–180 s
10 km/h CT
(%)
0.008
(0.006)
0.008
(0.006)
0.008
(0.006)
0.008
(0.005)
0.008
(0.004)
0.009
(0.003)
0.908 (0.870–0.963) 0.918 (0.793–0.968) 0.901 (0.681–0.963) 0.940 (0.844–0.977) 0.977 (0.941–0.9991)
FT
(%)
0.010
(0.005)
0.010
(0.004)
0.010
(0.004)
0.011
(0.003)
0.011
(0.003)
0.011
(0.002)
0.920 (0.786–0.969) 0.934 (0.808–0.976) 0.948 (0.864–0.980) 0.974 (0.933–0.990) 0.988 (0.968–0.995)
SL
(%)
3.57 (1.17) 3.47 (1.15) 3.42 (1.09) 3.40 (1.08) 3.40 (1.02) 3.39 (0.91) 0.902 (0.746–0.934) 0.901 (0.749–0.960) 0.933 (0.826–0.974) 0.978 (0.942–0.991) 0.990 (0.973–0.996)
SF
(%)
3.19 (1.80) 3.18 (1.45) 3.45 (2.51) 3.28 (2.01) 3.18 (1.47) 3.14(1.21) 0.941 (0.846–0.977) 0.975 (0.936–0.990) 0.905 (0.696–0.948) 0.931 (0.822–0.973) 0.988 (0.968–0.995)
12 km/h CT
(%)
2.36 (0.81) 2.32 (0.82) 2.52 (0.83) 2.59 (0.76) 2.58 (0.77) 2.70 (0.77) 0.901 (0.681–0.946) 0.904 (0.697–0.949) 0.964 (0.907–0.986) 0.987 (0.966–0.995) 0.996 (0.990–0.998)
FT
(%)
10.29
(4.94)
10.14
(4.15)
10.92
(3.80)
10.22
(3.60)
11.42
(4.06)
11.44
(4.02)
0.905 (0.689–0.944) 0.901 (0.685–0.964) 0.959 (0.894–0.984) 0.982 (0.953–0.993) 0.998 (0.995–0.999)
SL
(%)
3.01 (1.08) 2.98 (0.88) 3.12 (0.86) 3.10 (0.87) 3.03 (0.88) 3.04 (0.84) 0.901 (0.657–0.943) 0.948 (0.866–0.980) 0.961 (0.901–0.985) 0.968 (0.919–0.988) 0.990 (0.973–0.996)
SF
(%)
2.71 (0.83) 2.69 (0.79) 2.86 (0.76) 2.79 (0.71) 2.78 (0.71) 2.77 (0.69) 0.901 (0.732–0.951) 0.927 (0.815–0.972) 0.964 (0.906–0.986) 0.977 (0.940–0.991) 0.994 (0.986–0.998)
14 km/h CT
(%)
1.97 (0.50) 2.14 (0.78) 2.26 (0.73) 2.42 (0.90) 2.45 (0.90) 2.48 (0.85) 0.905 (0.615–0.954) 0.907 (0.707–0.919) 0.952 (0.772–0.985) 0.977 (0.941–0.991) 0.995 (0.986–0.998)
FT
(%)
6.96 (2.62) 7.26 (2.69) 7.62 (2.77) 8.07 (3.06) 8.18 (2.94) 8.26(2.83) 0.906 (0.612–0.939) 0.901 (0.635–0.927) 0.957 (0.838–0.985) 0.976 (0.938–0.991) 0.995 (0.987–0.995)
SL
(%)
2.24 (0.70) 2.35 (0.86) 2.45 (1.11) 2.57 (1.26) 2.57 (1.23) 2.63 (1.14) 0.907 (0.700–0.948) 0.905 (0.699–0.943) 0.962 (0.895–0.985) 0.980 (0.950–0.992) 0.995 (0.985–0.998)
SF
(%)
2.17 (0.67) 2.30 (0.78) 2.33 (0.77) 2.42 (0.84) 2.46 (0.79) 2.52 (0.70) 0.901 (0.613–0.932) 0.916 (0.734–0.970) 0.936 (0.784–0.977) 0.963 (0.904–0.986) 0.987 (0.967–0.995)
16 km/h CT
(%)
2.13 (0.87) 2.18 (0.96) 2.27 (1.24) 2.44 (1.75) 2.46 (1.32) 2.59 (1.30) 0.913 (0.802–0.945) 0.915 (0.648–0.958) 0.935 (0.795–0.977) 0.936 (0.831–0.976) 0.982 (0.950–0.993)
FT
(%)
5.57 (1.71) 5.86 (1.59) 5.97 (1.69) 6.38 (2.46) 6.37 (2.05) 6.50 (2.11) 0.904 (0.604–0.965) 0.902 (0.588–0.963) 0.951 (0.763–0.985) 0.974 (0.931–0.990) 0.991 (0.975–0.997)
SL
(%)
2.08 (0.86) 2.10 (0.90) 2.04 (0.72) 1.95 (0.56) 1.96 (0.57) 1.99 (0.51) 0.903 (0.750–0.954) 0.905 (0.760–0.951) 0.928 (0.809–0.973) 0.968 (0.915–0.988) 0.984 (0.957–0.994)
SF
(%)
2.09 (0.62) 2.11 (0.54) 2.11 (0.48) 2.13 (0.45) 2.23 (0.65) 2.36 (0.63) 0.901 (0.632–0.938) 0.905 (0.700–0.940) 0.906 (0.711–0.941) 0.909 (0.776–0.966) 0.930 (0.810–0.974)
CT, contact time; FT, flight time; SL, step length; SF, step frequency; ICC, intraclass correlation coefficient; 95% CI, 95% confidence interval.
F. García-Pinillos et al. / Journal of Biomechanics xxx (2018) xxx–xxx 5
BM 8852 No. of Pages 10, Model 5G
14 September 2018
Please cite this article in press as: García-Pinillos, F., et al. Minimum time required for assessing step variability during running at submaximal velocities. J.
Biomech. (2018), https://doi.org/10.1016/j.jbiomech.2018.09.005
199199199199199199
An almost perfect association between the recording intervals
200
compared to the 0–180 s interval was observed for the CV% at each
201
running speed tested (ICC > 0.90).
202
The Bland-Altman plots (Figs. 1–4) depict the mean difference
203
and 95% limits of agreement referred to the CV% of the spatiotem-
204
poral parameters (CT in Fig. 1,FTinFig. 2,SLinFig. 3, and SF in
205
Fig. 4) obtained from different time intervals (0–10 s, 0–20 s,
206
0–30 s, 0–60 s and 0–120 s) compared to 0–180 s. Longer recording
207
intervals yield smaller systematic bias, random errors, and
208
narrower limits of agreement for each spatiotemporal parameter,
209
regardless of running velocity (10–16 km/h).
210
4. Discussion
211
This study aimed to determine the minimum time required for
212
assessing spatiotemporal variability during continuous running on
213
an instrumented treadmill at different submaximal velocities and,
214
thereby, the number of steps required. The results demonstrated
Fig. 1. Bland-Altman plot with the coefficients of variation (CV%) of the contact time (CT) obtained from different time intervals (0–10 s, 0–20 s, 0–30 s, 0–60 s and 0–120s,
compared to 0–180 s). The plot includes the mean difference (dotted line) and 95% limits of agreement (dashed lined), along with the regression line (solid line). Systematic
bias and Pearson’s multivariate coefficient of determination (r
2
) are also presented. From left to right: 10 km/h; 12 km/h; 14 km/h and 16 km/h. From top to bottom: 0–10 s
vs. 0–180 s; 0–20 s vs. 0–180 s; 0–30 s vs. 0–180 s; 0–60 s vs. 0–180 s; 0–120 s vs. 0–180 s.
6F. García-Pinillos et al. / Journal of Biomechanics xxx (2018) xxx–xxx
BM 8852 No. of Pages 10, Model 5G
14 September 2018
Please cite this article in press as: García-Pinillos, F., et al. Minimum time required for assessing step variability during running at submaximal velocities. J.
Biomech. (2018), https://doi.org/10.1016/j.jbiomech.2018.09.005
215
that mean spatiotemporal gait parameters during running, and
216
variability in those parameters, can be accurately assessed (with
217
an assumable error) in a short period of time (10 s) and with rela-
218
tively few steps (25–30 steps), regardless of running velocity (10–
219
16 km/h). However, based on the information reported by the
220
Bland-Altman plots, the systematic bias and random error
221
assumed in step variability (CV%) during shorter time intervals
222
(i.e., 10 s, 20 s, 30 s, 60 s or 120 s) is higher than during longer
223
intervals (i.e., 180 s), and this should be taken into consideration.
224
Despite the importance of step variability, very few studies have
225
focused on determining how many steps are required to accurately
226
estimate spatial and temporal step variability during running.
227
Indeed, the authors found just one study from 1995 (Belli et al.,
228
1995), which examined step variability during running in terms
229
of step duration and vertical body displacement. Belli et al.
230
(1995) concluded that 32–64 consecutive steps (20 s) are
231
required to accurately estimate step variability during running at
232
submaximal velocities. Despite methodological differences, the
233
current results are consistent with those, highlighting that mean
234
spatiotemporal gait parameters during running, and variability in
235
those parameters, can be accurately assessed through the data col-
236
lected over a time period of only 10 s. Nevertheless, the current
Fig. 2. Bland-Altman plot with the coefficients of variation (CV%) of the flight time (FT) obtained from different time intervals (0–10 s, 0–20 s, 0–30 s, 0–60 s and 0–120 s,
compared to 0–180 s). The plot includes the mean difference (dotted line) and 95% limits of agreement (dashed lined), along with the regression line (solid line). Systematic
bias and Pearson’s multivariate coefficient of determination (r
2
) are also presented. From left to right: 10 km/h; 12 km/h; 14 km/h and 16 km/h. From top to bottom: 0–10 s
vs. 0–180 s; 0–20 s vs. 0–180 s; 0–30 s vs. 0–180s; 0–60 s vs. 0–180 s; 0–120 s vs. 0–180 s.
F. García-Pinillos et al. / Journal of Biomechanics xxx (2018) xxx–xxx 7
BM 8852 No. of Pages 10, Model 5G
14 September 2018
Please cite this article in press as: García-Pinillos, F., et al. Minimum time required for assessing step variability during running at submaximal velocities. J.
Biomech. (2018), https://doi.org/10.1016/j.jbiomech.2018.09.005
237
study builds up those findings by incorporating the Bland-Altman
238
limits of agreement method to examine differences in step vari-
239
ability between the shorter and the longest recording intervals
240
for each spatiotemporal parameter at each running velocity. This
241
analysis shows that longer recording intervals yield smaller sys-
242
tematic bias, random errors, and narrower limits of agreement,
243
regardless of running velocity or spatiotemporal parameter ana-
244
lyzed. These results seem to be consistent with previous studies
245
indicating that a greater number of steps are required to accurately
246
estimate step variability. As mentioned earlier, Owings and
247
Grabiner (2003) indicated that at least 400 steps are required,
248
König et al. (2014) proposed the collection of at least 50 cycles,
249
whereas Bruijn et al. (2009) concluded that longer data series
250
(>300 steps) led to more precise estimates for measures related
251
to step variability.
252
The lack of methodological consensus (i.e. the aforementioned
253
studies were conducted during walking at different velocities and
254
with different populations) might partially explain the differences
255
in the results. In this context, the authors suggest that caution
256
should be exercised to correctly interpret the data. In the current
257
study, from a practical standpoint, the magnitude of the differences
258
between longer and shorter recording intervals was rather small.
Fig. 3. Bland-Altman plot with the coefficients of variation (CV%) of the step length (SL) obtained from different time intervals (0–10 s, 0–20 s, 0–30 s, 0–60 s and 0–120 s,
compared to 0–180 s). The plot includes the mean difference (dotted line) and 95% limits of agreement (dashed lined), along with the regression line (solid line). Systematic
bias and Pearson’s multivariate coefficient of determination (r
2
) are also presented. From left to right: 10 km/h; 12 km/h; 14 km/h and 16 km/h. From top to bottom: 0–10 s
vs. 0–180 s; 0–20 s vs. 0–180 s; 0–30 s vs. 0–180 s; 0–60 s vs. 0–180 s; 0–120 s vs. 0–180 s.
8F. García-Pinillos et al. / Journal of Biomechanics xxx (2018) xxx–xxx
BM 8852 No. of Pages 10, Model 5G
14 September 2018
Please cite this article in press as: García-Pinillos, F., et al. Minimum time required for assessing step variability during running at submaximal velocities. J.
Biomech. (2018), https://doi.org/10.1016/j.jbiomech.2018.09.005
259
For example, the systematic bias ± random error for variability in
260
SL at 14 km/h (see Fig. 3) in the 0–10 s interval compared to the
261
0–180 s interval was 0.13 ± 0.46%, whereas in the 0–120 s vs.
262
0–180 s comparison was 0.06 ± 0.17%. Such small-magnitude dif-
263
ference between recording intervals must be gauged in practical
264
contexts, such as when assessing big groups of athletes, or where
265
logistical issues difficult long-lasting assessment protocols.
266
Some previous studies have examined the influence of running
267
velocity on spatiotemporal gait characteristics (Ogueta-Alday et al.,
268
2014; Padulo et al., 2012; Roche-Seruendo et al., 2018). Neverthe-
269
less, to the best of the authors
´knowledge, no previous studies have
270
examined the influence of running velocity on step variability and
271
whether the time required to accurately estimate variability, varies
272
with changes in running velocity. The data obtained in this study
273
indicate that some changes occur in step variability when running
274
velocity increases, but the time required to accurately estimate
275
variability seems to be unaffected by running velocity. In this con-
276
text, and considering that step variability seems to be related to
277
both injuries (Hamill et al., 2012; Meardon et al., 2011) and endur-
278
ance performance (Nakayama et al., 2010), further research is
279
clearly needed to highlight the role of step variability in endurance
280
running. The current work provides some insights into the devel-
281
opment of methodological basis to evaluate step variability in
282
endurance runners.
Fig. 4. Bland-Altman plot with the coefficients of variation (CV%) of the step frequency (SF) obtained from different time intervals (0–10 s, 0–20 s, 0–30 s, 0–60 s and 0–120 s,
compared to 0–180 s). The plot includes the mean difference (dotted line) and 95% limits of agreement (dashed lined), along with the regression line (solid line). Systematic
bias and Pearson’s multivariate coefficient of determination (r
2
) are also presented. From left to right: 10 km/h; 12 km/h; 14 km/h and 16 km/h. From top to bottom: 0–10 s
vs. 0–180 s; 0–20 s vs. 0–180 s; 0–30 s vs. 0–180s; 0–60 s vs. 0–180 s; 0–120 s vs. 0–180 s.
F. García-Pinillos et al. / Journal of Biomechanics xxx (2018) xxx–xxx 9
BM 8852 No. of Pages 10, Model 5G
14 September 2018
Please cite this article in press as: García-Pinillos, F., et al. Minimum time required for assessing step variability during running at submaximal velocities. J.
Biomech. (2018), https://doi.org/10.1016/j.jbiomech.2018.09.005
283
Finally, it is important to note that these results could be
284
restricted to trained endurance runners (Nakayama et al., 2010)
285
and to a running protocol performed on a treadmill at fixed sub-
286
maximal velocities (Riley et al., 2008; Roche-Seruendo et al.,
287
2018). One more limitation of the present study could be that
288
the footwear was not standardized, but all runners wore their
289
own footwear to increase the ecological validity of the study.
290
In conclusion, the results suggest that the duration of the
291
recording period required to estimate spatiotemporal variability
292
plays an important role in the accuracy of the measurement,
293
regardless of running velocity (10–16 km/h). The analysis con-
294
ducted shows that longer recording intervals yield smaller system-
295
atic bias, random errors, and narrower limits of agreement.
296
Therefore, if maximum accuracy is required (e.g., scientific
297
approach), longer recording periods (i.e., 3 min) must be used.
298
However, from a practical standpoint, shorter recording intervals
299
might be a time-efficient option for clinicians or coaches working
300
with big groups of athletes, or where logistical issues difficult
301
long-lasting assessment protocols (e.g., athletes with pain during
302
running).
303
5. Conflict of interest statement
304
The authors declare that they have no conflict of interests.
305
6. Competing interests
306
The authors declare that they have no competing interests.
307
7. Funding sources
308
This research did not receive any specific grant from funding
309
agencies in the public, commercial, or not-for-profit sectors.
310
References
311 Belli, A., Lacour, J.R., Komi, P.V., Candau, R., Denis, C., 1995. Mechanical step
312 variability during treadmill running. Eur. J. Appl. Physiol. Occup. Physiol. 70,
313 510–517.
314 Bland, J.M., Altman, D.G., 1995. Comparing methods of measurement: why plotting
315 difference against standard method is misleading. Lancet 346, 1085–1087.
316 https://doi.org/10.1016/S0140-6736(95)91748-9.
317
Borg, G., 1982. Psychophysical bases of perceived exertion. Med. Sci. Sports Exerc. 318
14, 377–381. 319
Bruijn, S.M., van Dieën, J.H., Meijer, O.G., Beek, P.J., 2009. Statistical precision and 320
sensitivity of measures of dynamic gait stability. J. Neurosci. Meth. 178, 327– 321
333. https://doi.org/10.1016/j.jneumeth.2008.12.015.322
Hamill, J., Palmer, C., Van Emmerik, R.E.A., 2012. Coordinative variability and 323
overuse injury. Sport. Med. Arthrosc. Rehabil. Ther. Technol. 4, 45. 324
König, N., Singh, N.B., von Beckerath, J., Janke, L., Taylor, W.R., 2014. Is gait 325
variability reliable? An assessment of spatio-temporal parameters of gait 326
variability during continuous overground walking. Gait Post. 39, 615–617. 327
https://doi.org/10.1016/j.gaitpost.2013.06.014.328
Koo, T.K., Li, M.Y., 2016. A guideline of selecting and reporting intraclass correlation 329
coefficients for reliability research. J. Chiropr. Med. 15, 155–163. https://doi.org/ 330
10.1016/j.jcm.2016.02.012.331
Landis, J.R., Koch, G.G., 1977. The measurement of observer agreement for 332
categorical data. Biometrics 33, 159–174. 333
Lee, M.M., Song, C.H., Lee, K.J., Jung, S.W., Shin, D.C., Shin, S.H., 2014. Concurrent 334
validity and test-retest reliability of the OPTOGait photoelectric cell system for 335
the assessment of spatio-temporal parameters of the gait of young adults. J. 336
Phys. Ther. Sci. 26, 81–85. https://doi.org/10.1589/jpts.26.81.337
Mann, R., Malisoux, L., Nührenbörger, C., Urhausen, A., Meijer, K., Theisen, D., 2015. 338
Association of previous injury and speed with running style and stride-to-stride 339
fluctuations. Scand. J. Med. Sci. Sports 25, e638–e645. https://doi.org/10.1111/ 340
sms.12397.341
Meardon, S.A., Hamill, J., Derrick, T.R.T.R., 2011. Running injury and stride time 342
variability over a prolonged run. Gait Post. 33, 36–40. https://doi.org/10.1016/ 343
j.gaitpost.2010.09.020.344
Nakayama, Y., Kudo, K., Ohtsuki, T., 2010. Variability and fluctuation in running gait 345
cycle of trained runners and non-runners. Gait Post. 31, 331–335. https://doi. 346
org/10.1016/j.gaitpost.2009.12.003.347
Ogueta-Alday, A., Rodríguez-Marroyo, J.A., García-López, J., 2014. Rearfoot striking 348
runners are more economical than midfoot strikers. Med. Sci. Sports Exerc. 46, 349
580–585. https://doi.org/10.1249/MSS.0000000000000139.350
Owings, T.M., Grabiner, M.D., 2003. Measuring step kinematic variability on an 351
instrumented treadmill: how many steps are enough? J. Biomech. 36, 1215– 352
1218. https://doi.org/10.1016/S0021-9290(03)00108-8.353
Padulo, J., Annino, G., Migliaccio, G.M., D’Ottavio, S., Tihanyi, J., 2012. Kinematics of 354
running at different slopes and speeds. J. Strength Cond. Res. https://doi.org/ 355
10.1519/JSC.0b013e318231aafa.356
Paquette, M.R., Milner, C.E., Melcher, D.A., 2017. Foot contact angle variability 357
during a prolonged run with relation to injury history and habitual foot strike 358
pattern. Scand. J. Med. Sci. Sports 27, 217–222. https://doi.org/10.1111/ 359
sms.12647.360
Riley, P.O., Dicharry, J., Franz, J., Della Croce, U., Wilder, R.P., Kerrigan, D.C., 2008. A 361
kinematics and kinetic comparison of overground and treadmill running. Med. 362
Sci. Sports Exerc. 40, 1093–1100. https://doi.org/10.1249/ 363
MSS.0b013e3181677530.364
Roche-Seruendo, L.E., García-Pinillos, F., Haicaguerre, J., Bataller-Cervero, A.V., Soto- 365
Hermoso, V.M., Latorre-Román, P.Á., 2018. Lack of influence of muscular 366
performance parameters on spatio-temporal adaptations with increased 367
running velocity. J. Strength Cond. Res. 32, 409–415. https://doi.org/10.1519/ 368
JSC.0000000000001845.369
Schieb, D.A., 1986. Kinematic accommodation of novice treadmill runners. Res. Q. 370
Exerc. Sport 57, 1–7.
371
10 F. García-Pinillos et al. / Journal of Biomechanics xxx (2018) xxx–xxx
BM 8852 No. of Pages 10, Model 5G
14 September 2018
Please cite this article in press as: García-Pinillos, F., et al. Minimum time required for assessing step variability during running at submaximal velocities. J.
Biomech. (2018), https://doi.org/10.1016/j.jbiomech.2018.09.005
... The overall decrease in the contact time (CT) and the increase in flight time (FT), step length (SL), and step frequency (SF) with increasing running speed have been previously shown (Brughelli et al., 2010;Ogueta-Alday et al., 2014;Padulo et al., 2012;Roche-Seruendo et al., 2018). However, new devices for examining running biomechanics have arisen (e.g., OptoGait™ system) facilitating the assessment of new parameters, e.g., step angle [SA] (Santos-Concejero et al., 2014) and step variability (García-Pinillos et al., 2018a). Moreover, there is limited evidence about the dynamics of these parameters in amateur endurance runners as running velocity increases, especially regarding the interpretation of these parameters. ...
... Some studies have suggested a relationship between step variability and both injuries (Hamill et al., 2012;Meardon et al., 2011) and endurance performance (Nakayama et al., 2010). Despite the importance of step variability during running, the evidence about potential influencing factors such as running velocity is quite limited (García-Pinillos et al., 2018a). A previous study (García-Pinillos et al., 2018a) described the step variability during running at different submaximal velocities (i.e., 10, 12, 14 and 16 km/h), although did not assess the impact of running velocity on the magnitude of step variability. ...
... Despite the importance of step variability during running, the evidence about potential influencing factors such as running velocity is quite limited (García-Pinillos et al., 2018a). A previous study (García-Pinillos et al., 2018a) described the step variability during running at different submaximal velocities (i.e., 10, 12, 14 and 16 km/h), although did not assess the impact of running velocity on the magnitude of step variability. ...
Article
Full-text available
This study aimed to analyse the effects of running velocity on spatiotemporal parameters and step variability in amateur endurance runners, according to sex. A group of 51 males and 46 females performed an incremental running test on a treadmill (10-16 km/h). Spatiotemporal parameters (contact and flight time, step length, step frequency and step angle [CT, FT, SL, SF, SA]) and step variability, in terms of within-participant standard deviation (SD), were recorded through the OptoGait System. The ANOVA showed significant differences in the magnitude of the spatiotemporal parameters as running velocity increased (p < 0.001). It also revealed significant differences in step variability (p < 0.005) over the entire running protocol. Between-sex differences in CT, SL, SL-normalized and SF (p < 0.05, ES = 0.4-0.8) were found. Differences were also found in step variability at high velocities (15-16 km/h), with males showing a greater SD than females. In conclusion, increasing running velocity makes CT shorter, FT and SL longer, and SF and SA greater in amateur endurance runners, changing step variability, regardless of sex. Additionally, some between-sex differences were found in spatiotemporal parameters and step variability.
... Just after warming up, an accommodation program over 8 minutes [30] was developed by increasing speed by 1 km/h every minute from 8 to 12 km/h. After that, subjects ran under the first footwear condition (shod or barefoot) at a speed of 12 km/h for 3 minutes [31], −6 and −8 strides were analyzed to obtain representative data in healthy adults (95% confidence intervals within 5% of error) [32]. Thereafter, subjects ran under the next footwear condition at 12 km/h for another 3 minutes. ...
... Kvert (kN/m) 31 Table 3 shows a comparison of connective tissue morphology parameters between those sub-groups (higher stiffness group [HSG] vs. lower stiffness group [LSG]) in both barefoot and shod running conditions. In the barefoot condition, no significant differences in the morphology of connective tissue were found between the HSG and the LSG (p ≥ 0.05). ...
Article
Full-text available
Background: The lower limb behaves like a spring compressing and decompressing during running, where lower-limb stiffness is one of the most influential factors. This prospective observational study is aimed at examining the relationship between the connective tissue morphology and lower-limb stiffness and investigating whether the barefoot/shod condition influences on such relationship. Methods: 14 male amateur runners (10-km time trial <50') were included. Data were recorded over one session, where participants ran 2 trials (i.e., barefoot and shod conditions) of 3 minutes at 12 km/h, where running spatiotemporal parameters and vertical (Kvert) and leg stiffness (Kleg) were obtained. Prior to testing trials, thickness and cross-sectional area (CSA) were recorded for Achilles (AT) and patellar tendons (PT) and plantar fascia (PF) with ultrasound. Results: Under barefoot condition, a positive correlation was found between Kleg and AT-thickness and CSA and PF-thickness; and between Kvert and AT-thickness and PF thickness. Under shod condition, a positive correlation was found between Kleg and PT-CSA and PT-thickness, and between Kvert and PT-CSA and PT-thickness. Conclusions: The results reveal a specificity of the relationship between the lower-limb stiffness and the morphology of the connective tissue. Greater tendon shows higher lower-limb stiffness when that tendon is specially demanded by the function.
... For each trial, participants completed two successive 3 min running bouts (i.e., shod for the first and barefoot for the latter), separated by a 2 min period to change from shod to barefoot condition. Since power output [19] and spatiotemporal parameters [22] reach a steady state in less than 2 min, data were recorded during both running trials and 6-8 strides were analysed [23]. ...
Article
Full-text available
Several studies have already analysed power output in running or the relation between VO2max and power production as factors related to running economy; however, there are no studies assessing the difference in power output between shod and barefoot running. This study aims to identify the effect of footwear on the power output endurance runner. Forty-one endurance runners (16 female) were evaluated at shod and barefoot running over a one-session running protocol at their preferred comfortable velocity (11.71 ± 1.07 km·h−1). The mean power output (MPO) and normalized MPO (MPOnorm), form power, vertical oscillation, leg stiffness, running effectiveness and spatiotemporal parameters were obtained using the Stryd™ foot pod system. Additionally, footstrike patterns were measured using high-speed video at 240 Hz. No differences were noted in MPO (p = 0.582) and MPOnorm (p = 0.568), whereas significant differences were found in form power, in both absolute (p = 0.001) and relative values (p < 0.001), running effectiveness (p = 0.006), stiffness (p = 0.002) and vertical oscillation (p < 0.001). By running barefoot, lower values for contact time (p < 0.001) and step length (p = 0.003) were obtained with greater step frequency (p < 0.001), compared to shod running. The prevalence of footstrike pattern significantly differs between conditions, with 19.5% of runners showing a rearfoot strike, whereas no runners showed a rearfoot strike during barefoot running. Running barefoot showed greater running effectiveness in comparison with shod running, and was consistent with lower values in form power and lower vertical oscillation. From a practical perspective, the long-term effect of barefoot running drills might lead to increased running efficiency and leg stiffness in endurance runners, affecting running economy.
... Immediately after, an 8-min treadmill running accommodation program [21] was implemented by increasing speed by 1 km·h −1 every minute from 8 to 12 km·h −1 . As spatiotemporal parameters reach a steady state soon [22], a 3-min running bout at a speed of 12 km·h −1 was proposed being 6-8 strides analysed [23]. Data were recorded for subsequent analysis. ...
Article
Full-text available
Background: Musculotendinous reactive strength is a key factor for the utilization of elastic energy in sporting activities such as running. Aim: To evaluate the relationship between musculotendinous reactive strength and lower-limb stiffness during running as well as to identify age-related differences in both variables. Methods: Fifty-nine amateur endurance runners performed three 20-cm drop jumps and a constant 3-min easy run on a motorized treadmill. Reactive strength index and dynamic lower-limb stiffness were calculated with a photoelectric cell system by jumping and running, respectively. Additionally, sit to stand difference in plantar arch height was assessed as a static lower-limb stiffness measure. The cluster analysis allows the comparison between younger and older runners. Results: No significant correlations were found between jumping reactive strength and running lower-limb stiffness. The younger group performed better at drop jumps (p = 0.023, ES = 0.82), whereas higher-but-no-significant results were found for reactive strength index and stiffness-related metrics. Conclusions: Musculotendinous vertical reactiveness may not be transferred to combined vertical and horizontal movements such as running.
... Then, each individual completed a warm-up period consisting of 5 min of brisk walking and 5 min of running at the test speed. Next, we recorded the 3D marker trajectories for 30 s yielding 35-40 full gait cycles per participant, which has been shown to be sufficient for accurately estimating kinematic running patterns (Dingenen et al., 2018;García-Pinillos et al., 2018). Throughout the entire treadmill protocol, participants were wearing a safety harness to avoid the risk of injury in the case of a fall or slip. ...
Article
Full-text available
There is a sex bias for common overuse running injuries that are associated with sex-specific hip kinematics. Gait retraining programs aimed at altering hip kinematics may be more efficient if they incorporated an understanding of how hip kinematics are correlated with the movement of the remaining body segments. We applied a principal component analysis to structure the whole-body running kinematics of 23 runners (12 ♀) into k = 12 principal movements (PMk), describing correlated patterns of upper and lower body movements. We compared the time-dependent movement amplitudes with respect to each PMk between males and females using a waveform analysis and interpreted our findings according to stick figure animations. The movement amplitudes of two PMs (PM6 and PM8) showed statistically significant effects of ‘sex’, which were independent of running speed. According to PM8, females showed more hip adduction, which correlated with increased transverse rotation of the pelvis and upper body compared to men. We propose that increased hip adduction and upper body rotation in female runners may be a strategy to compensate for a less efficient arm and upper body swing compared to men. Gait interventions aimed at reducing hip adduction and running-related injuries in female runners should consider instructions for both upper and lower body to maximize training efficacy.
... Since previous studies on spatiotemporal parameters during running have shown that longer recording intervals resulted in smaller systematic bias, random errors, and narrower limits of agreement, 30 the differences in power assessment are also expected to decrease for longer recording periods. However, in the aforementioned work, 30 it was stated that the variability in the spatiotemporal parameters of running can be accurately calculated within 25-30 steps (i.e., approximately 10 s), so the 1-min intervals selected should fit the purpose of this study. In addition, some methodological differences between systems could also play a role in this discrepancy. ...
Article
The advent of portable power meters has revolutionized training in cycling, allowing an accurate field-based assessment of athletes. In a similar way, researchers have recently developed low-cost gait analysis equipment to assess running power in a more natural environment. The purpose of this study was to evaluate the absolute reliability of two different power meters and the agreement between these two wearable devices (i.e., Stryd and RunScribe) for measuring power during treadmill running. About 49 endurance runners performed a running protocol on a treadmill at self-selected comfortable velocities. Power output (W) was measured using the Stryd and RunScribe systems, which were attached to the same shoe. The absolute reliability, based on coefficient of variation, was 0.32 6 0.29% for Strydä and 1.68 6 1.49% for RunScribe, while the standard error of the mean were 0.3 6 0.2 W and 2.6 6 2.5 W for Stryd and RunScribe, respectively. Data from both devices showed significant correlations (r = 0.783, p \ 0.001) and the ICC (r = 0.855) reported an almost perfect reliability. Bland-Altman plots revealed no heteroscedasticity of error (r 2 = 0.030), although a moderate systematic bias (212.3 6 26.6 W), and wide limits of agreement (39.8-64.3 W) were found. Considering the increased popularity of using power meter devices in running, scientists, coaches, athletes, and general users should be aware that data from these devices are reliable, but not interchangeable, due to the variation shown for running power output data.
Article
Full-text available
Movement variability is defined as the normal variations in motor performance across multiple repetitions of a task. However, the term “movement variability” can mean different things depending on context, and when used by itself does not capture the specifics of what has been investigated. Within sport, complex movements are performed repeatedly under a variety of different constraints (e.g. different situations, presence of defenders, time pressure). Movement variability has implications for sport performance and injury risk management. Given the importance of movement variability, it is important to understand the terms used to measure and describe it. This broad term of “movement variability” does not specify the different types of movement variability that are currently being assessed in the sporting literature. We conducted a scoping review (1) to assess the current terms and definitions used to describe movement variability within sporting tasks and (2) to utilise the results of the review for a proposed framework that distinguishes and defines the different types of movement variability within sporting tasks. To be considered eligible, sources must have assessed a sporting movement or skill and had at least one quantifiable measure of movement variability. A total of 43 peer-reviewed journal article sources were included in the scoping review. A total of 280 terms relating to movement variability terminology were extracted using a data-charting form jointly developed by two reviewers. One source out of 43 (2%) supplied definitions for all types of movement variability discussed. Moreover, 169 of 280 terms (60%) were undefined in the source material. Our proposed theoretical framework explains three types of movement variability: strategic, execution, and outcome. Strategic variability describes the different approaches or methods of movement used to complete a task. Execution variability describes the intentional and unintentional adjustments of the body between repetitions within the same strategy. Outcome variability describes the differences in the result or product of a movement. These types emerged from broader frameworks in motor control and were adapted to fit the movement variability needs in sports literature. By providing specific terms with explicit definitions, our proposed framework can ensure like-to-like comparisons of previous terms used in the literature. The practical goal of this framework is to aid athletes, coaches, and support staff to gain a better understanding of how the different types of movement variability within sporting tasks contribute to performance. The framework may allow training methods to be tailored to optimise the specific aspects of movement variability that contribute to success. This review was retrospectively registered using the Open Science Framework (OSF) Registries (https://osf.io/q73fd).
Article
This study aimed to determine the influence of footwear condition, foot-strike pattern and step frequency on running spatiotemporal parameters and lower-body stiffness during treadmill running. Thirty-one amateur endurance runners performed a two-session protocol (shod and barefoot). Each session consisted of two trials at 12 km · h −1 over 5 minutes altering step frequency every minute (150, 160, 170, 180 and 190 spm). First, participants were instructed to land with the heel first; after completion, the same protocol was repeated landing with the forefoot first. Repeated measures ANOVAs showed significant differences for footwear condition, foot-strike pattern and step frequency for each variable: percent contact time, percent flight time, vertical stiffness and leg stiffness (all p < 0.001). The results demonstrate greater estimated vertical and leg stiffness when running barefoot for both foot-strike patterns showing the largest values for barefoot+forefoot condition. Likewise, both vertical and leg stiffness became greater as step frequency increased. The proper manipulation of these variables facilitates our understanding of running performance and assist in training programmes design and injury management.
Thesis
The study of the spring-mass model variables in running resulted in a great contribution to the understanding of the behaviour of such model not only in humans, but in animals as well. Although the study of the running spatiotemporal parameters has contributed to obtain a deeper knowledge about the spring-mass model and its capacity to estimate and predict kinematic variables, the contribution of lower-limb stiffness to this model needed further research. The main aim of the present PhD Thesis was to determine the effect of various influential factors on lower-limb stiffness while treadmill running in healthy adults. Three different studies were executed to accomplish the main aim of this PhD Thesis: a unilateral cross-over study aiming at examining the test-retest reliability of the OptoGait photoelectric system for spatiotemporal parameters and lower-body stiffness analysis while treadmill running in healthy adults (Study 1). This first study is key as the entire development of this PhD Thesis has been based on the material and methods implemented and the findings reported; a unilateral cross-over study to clarify the likely relationship between reactive strength index while jumping and lower-limb stiffness while treadmill running in amateur endurance runners as well as sex differences (Study 2); and, ultimately, a unilateral cross-over study to identify the effects of footwear, foot-strike pattern, and step frequency on spatiotemporal parameters and lower-body stiffness (Study 3). The main findings derived from this PhD Thesis suggest that: the OptoGait system can be used confidently for running spatiotemporal parameters analysis and lower body stiffness at a constant velocity for healthy adults. The spring-mass model reacts differently to tasks based on their specificity principle. Additionally, sex-related differences must be considered when assessing the stretch-shortening cycle. Lower-limb stiffness responds differently to changes in footwear condition, foot-strike pattern, and step frequency. The findings reported here update the knowledge of lower-body stiffness while running and offer new scopes of action. A reliable and user-friendly system for running spatiotemporal parameters and lower-body stiffness analysis has been provided. Moreover, although both the SSC and lower-limb stiffness are key within the neuromuscular behaviour when elastic energy is used in sport, the specificity principle of each individual sporting task may make them behave differently; additionally, the menstrual cycle should be considered when working with female athletes since musculotendinous properties change over it. Ultimately, it is highly recommended to avoid measuring the effect of different variables on lower-limb stiffness individually as it has been shown that they influence each another, therefore, the behaviour of the spring-mass model when altering variables such as footwear, foot-strike pattern (FSP), and step frequency (SF) needs to be examined should be analysed attentively.
Article
Despite the widespread use of the OptoGait photoelectric cell system for the analysis of running spatiotemporal parameters, its reliability has not been proved. Consequently, this study intends to determine the test–retest reliability of the system when applied to treadmill running spatiotemporal parameters and lower body stiffness at a constant velocity. Amateur endurance runners (n = 31; age: 34.42 ± 9.26 years; height: 171.54 ± 9.15 cm; body mass: 66.63 ± 11.3 kg) voluntarily consented to participate in this study. Data for each participant were recorded twice per session across two testing sessions. The intra-session and inter-session reliabilities of spatiotemporal parameters and lower body stiffness were determined through test–retest analysis. Although mean comparisons revealed significant differences between measurements in spatiotemporal running gait characteristics and lower body stiffness for intra-session (p < 0.05 in all parameters), the effect size was always small (<0.4). Moreover, the relationship between measurements was very large for spatiotemporal parameters and lower body stiffness (r > 0.7). The intraclass correlation coefficients revealed an almost perfect correlation between measurements (intraclass correlation coefficients >0.81), except Kleg with substantial reliability (intraclass correlation coefficient = 0.788). The inter-session reliability revealed some significant differences in contact time (p = 0.009) and Kleg (p = 0.013), although Cohen’s d indicated small effect size (<0.31). The relationship between sessions was very large for spatiotemporal parameters and lower body stiffness (r > 0.8), and the intraclass correlation coefficients revealed an almost perfect inter-session association (intraclass correlation coefficients >0.881). The results found here show that the OptoGait system can be used confidently for running spatiotemporal parameters analysis and lower body stiffness at a constant velocity for healthy adults.
Article
Full-text available
This study aimed to analyse the influence of muscular performance parameters on spatio-temporal gait characteristics during running when gradually increasing speed. 51 recreationally trained male endurance runners (age: 28 ± 8 years) voluntarily participated in this study. Subjects performed a battery of jumping tests (squat jump, countermovement jump, and 20 cm drop jump), and after that, the subjects performed an incremental running test (10 to 20 km/h) on a motorized treadmill. Spatio-temporal parameters were measured using the OptoGait system. Cluster k-means analysis grouped subjects according to the jumping test performance, by obtaining a group of good jumpers (GJ, n = 19) and a group of bad jumpers (BJ, n = 32). With increased running velocity, contact time was shorter, flight time and step length longer, whereas cadence and stride angle were greater (p < 0.001). No significant differences between groups (p ≥ 0.05) were found at any running speed. The results obtained indicate that increased running velocity produced no differences in spatio-temporal adaptations between those runners with good jumping ability and those with poor jumping ability. Based on that, it seems that muscular performance parameters do not play a key role in spatio-temporal adaptations experienced by recreational endurance runners with increased velocity. However, taken into consideration the well-known relationship between running performance and neuromuscular performance, the authors suggest that muscular performance parameters would be much more determinant in the presence of fatigue (exhausted condition), or in the case of considering other variables such as running economy or kinetic.
Article
Full-text available
Running-related injuries remain problematic among recreational runners. We evaluated the association between having sustained a recent running-related injury and speed, and the strike index (a measure of footstrike pattern, SI) and spatiotemporal parameters of running. Forty-four previously injured and 46 previously uninjured runners underwent treadmill running at 80%, 90%, 100%, 110%, and 120% of their preferred running speed. Participants wore a pressure insole device to measure SI, temporal parameters, and stride length (Slength) and stride frequency (Sfrequency) over 2-min intervals. Coefficient of variation and detrended fluctuation analysis provided information on stride-to-stride variability and correlative patterns. Linear mixed models were used to compare differences between groups and changes with speed. Previously injured runners displayed significantly higher stride-to-stride correlations of SI than controls (P = 0.046). As speed increased, SI, contact time (Tcontact), stride time (Tstride), and duty factor (DF) decreased (P < 0.001), whereas flight time (Tflight), Slength, and Sfrequency increased (P < 0.001). Stride-to-stride variability decreased significantly for SI, Tcontact, Tflight, and DF (P ≤ 0.005), as did correlative patterns for Tcontact, Tstride, DF, Slength, and Sfrequency (P ≤ 0.044). Previous running-related injury was associated with less stride-to-stride randomness of footstrike pattern. Overall, runners became more pronounced rearfoot strikers as running speed increased.
Article
Full-text available
[Purpose] The purpose of this study was to investigate the concurrent validity and test-retest reliability of the recently introduced OPTOGait Photoelectric Cell System for the assessment of spatio-temporal parameters of gait. [Subjects] Twenty healthy young adults (mean age = 27.35, SD = 7.4) were asked to walk 3 times on walkway at a comfortable speed. [Methods] Concurrent validity was assessed by comparing data obtained using the OPTOGait and GAITRite systems, and reliability was assessed by comparing data from the first and third OPTOGait sessions. [Results] Concurrent validity, as identified by intra-class correlation coefficients (ICC (2, 1) = 0.929-0.998), coefficients of variation (CVME = 0.32-11.30%), and 95% limits of agreement, showed high levels of correlation. In addition, the test-retest reliability of the OPTOGait Photoelectric Cell System was demonstrated as showing a high level of correlation with all spatio-temporal parameters by intra-class correlation coefficients (ICC (3, 1) = 0.785-0.952), coefficients of variation (CVME = 1.66-4.06%), 95% limits of agreement, standard error of measurement (SEM = 2.17-5.96%), and minimum detectable change (MDC95% = 6.01-16.52%). [Conclusion] The OPTOGait Photoelectric Cell System has strong concurrent validity along with relative and absolute test-retest reliabilities. This portable system with easy-to-use features can be used for clinical assessments or research purposes as an objective means of assessing gait.
Article
Full-text available
PURPOSE: To analyze the influence of foot strike pattern on running economy and biomechanical characteristics in sub-elite runners with a similar performance level. METHODS: Twenty sub-elite long-distance runners participated and were divided into two groups according to their foot strike pattern: rearfoot (RF, n= 10) and midfoot strikers (MF, n= 10). Anthropometric characteristics were measured (height, body mass, BMI, skinfolds, circumferences and lengths); physiological (V˙O2max, anaerobic threshold and running economy) and biomechanical characteristics (contact and flight times, step rate and step length) were registered during both incremental and submaximal tests on a treadmill. RESULTS: There were no significant intergroup differences in anthropometrics, V˙O2max or anaerobic threshold measures. RF strikers were 5.4, 9.3 and 5.0% more economical than MF at submaximal speeds (11, 13 and 15 km·h respectively, though the difference was not significant at 15 km·h, p=0.07). Step rate and step length were not different between groups, but RF showed longer contact time (p<0.01) and shorter flight time (p<0.01) than MF at all running speeds. CONCLUSIONS: The present study showed that habitually rearfoot striking runners are more economical than midfoot strikers. Foot strike pattern affected both contact and flight times, which may explain the differences in running economy.
Article
Full-text available
Overuse injuries are generally defined as a repetitive micro-trauma to tissue. Many researchers have associated particular biomechanical parameters as an indicator of such injuries. However, while these parameters have been reported in single studies, in many instances, it has been difficult to verify these parameters as causative to the injury. We have investigated overuse injuries, such as patella-femoral pain syndrome, using a dynamical systems approach. Using such methods, the importance of the structure of coordinative variability (i.e. the variability of the interaction between segments or joints) becomes apparent. We view coordinative variability as functionally important to the movement and different from end-point or goal variability. Using concepts derived from the work of Bernstein, we conducted studies using a continuous relative phase and/or modified vector coding approaches to investigate the coordinative variability of overuse injuries. Consistently, we have found that the higher variability state of a coordinative structure is the healthy state while the lower variability state is the unhealthy or pathological state. It is clear that very high coordinative variability could also result in injury and that there must be a window of 'higher variability' in which non-injured athletes function. While this finding that coordinative variability is functional has been shown in several studies, it is still not clear if reduced variability contributes to or results from the injury. Studies are currently underway to determine the potential reasons for the reduced variability in injured athletes. Nevertheless, our laboratory believes that this understanding of how joints interact can be important in understanding overuse injuries.
Article
Objective: Intraclass correlation coefficient (ICC) is a widely used reliability index in test-retest, intrarater, and interrater reliability analyses. This article introduces the basic concept of ICC in the content of reliability analysis. Discussion for researchers: There are 10 forms of ICCs. Because each form involves distinct assumptions in their calculation and will lead to different interpretations, researchers should explicitly specify the ICC form they used in their calculation. A thorough review of the research design is needed in selecting the appropriate form of ICC to evaluate reliability. The best practice of reporting ICC should include software information, "model," "type," and "definition" selections. Discussion for readers: When coming across an article that includes ICC, readers should first check whether information about the ICC form has been reported and if an appropriate ICC form was used. Based on the 95% confident interval of the ICC estimate, values less than 0.5, between 0.5 and 0.75, between 0.75 and 0.9, and greater than 0.90 are indicative of poor, moderate, good, and excellent reliability, respectively. Conclusion: This article provides a practical guideline for clinical researchers to choose the correct form of ICC and suggests the best practice of reporting ICC parameters in scientific publications. This article also gives readers an appreciation for what to look for when coming across ICC while reading an article.
Article
Foot strike pattern and movement variability have each been associated with running injuries. Foot contact angle (FCA) is a common measure of strike pattern. Thus, variability in FCA could be an important running injury risk factor. The purposes of this study were to compare 1) foot contact angle (FCA) and its variability between runners with and without injury history and, 2) FCA variability between habitual rearfoot strike (RFS) and non-RFS runners during a prolonged run. 23 runners with and 21 without injury history participated. Motion capture was used to collect kinematic data during a 40 minute treadmill run. Average FCA and its variability were compared between injury groups and among four time points. FCA and its variability were not different between runners with and without injury history or among time points during the run. FCA variability was lower in non-RFS compared to RFS runners (p < 0.001). Lower FCA variability in non-RFS runners may have implications for higher injury risks due to repeated localized tissue loading. Prospective analyses on the effects of lower FCA variability on injury risk are needed.
Article
The purpose of this study was to determine the effects of treadmill training on the kinematic accommodation and habituation process of novice treadmill runners. Six experienced male college distance runners, but novice treadmill runners, trained on a treadmill operating at 4.0 m[mdot]s−1, 15 min daily for 10 days. Subjects were filmed three times each day in the frontal and sagittal planes, at Minutes 1, 8, and 14 of the run. Stride length, temporal data, and vertical and lateral horizontal displacements of the center of gravity were determined with a computer digitizer system. Analysis of variance revealed that significant alterations occurred in treadmill running kinematics between Days 1 and 2 of the 10-day treadmill training period. Further, for Days 1 through 3, significant within-day stride changes occurred between Minutes 1 and 8, but not between Minutes 8 and 14. These results suggest that minimal amounts of treadmill training are necessary for a subject to fully accommodate to the treadmill.
Article
The assessment of gait variability has become an important indicator for quantifying motor performance. However, the use of treadmills is known to influence the temporal rhythm of gait, while non-continuous (i.e. stop-start) overground walking alters gait variability, leading to erroneous results. Through establishing the "8-walk", an overground walking protocol that allows the collection of a high number of consecutive gait cycles, the aim of this study was to determine the conditions under which gait variability can be assessed reliably. Twelve healthy subjects performed continuous barefoot walking at their preferred speed in a path shaped as an "8". Kinematic data of the dominant foot was collected while subjects walked along the straight 10m sections of the 8-walk during sessions on two different days. Mean spatio-temporal parameters of gait and gait variability were computed for 10, 20, 30, 40, 50 and 60 consecutive cycles. All mean parameters of gait showed excellent reliability (ICC: 0.88-0.98) with only 10 cycles included in the analysis. However, the reliability of spatial and temporal parameters of gait variability improved with increasing number of cycles (ICC: 0.60-0.90) but levelled-off after 50 consecutive cycles, revealing an inter-day test-retest variability of ∼13%. To reliably assess gait variability and evaluate human motor performance, we propose the collection of at least 50 cycles and the use of an 8-walk protocol, which avoids the limitations of treadmill and non-consecutive walking protocols.
Article
The aim of this study was to verify the influence of the combination of different running speeds and slopes based on main kinematic parameters in both groups of elite (RE) and amateur (RA) marathon runners. All subjects performed various tests on a treadmill at 0, 2, and 7% slopes at different speeds: 3.89, 4.17, 4.44, 4.72, and 5.00 m·s. A high speed digital camera, 210 Hz, has been used to record; Dartfish 5.5Pro has been used to perform a 2D video analysis. Step length (SL), step frequency (SF), flight time (FT), and contact time (CT) were determined and used for comparison. SL, SF, and FT parameters increased, and CT parameter decreased as speed increased. As slopes increased, SL and FT decreased and SF increased in both groups and only CT decreased in RE, whereas in RA, it increased. Data were fitted to the linear regression line (R > 0.95). The 2 groups were significantly different (p < 0.05) in FT, SL, and SF at all speeds in level running. A significant difference between the 2 groups was found in FT at 2 and 7% slopes at all speeds (p < 0.05). Percentage alterations in all variables were greater in the RA group. In conclusion, the choice of optimum SL and SF, through efficient running can be maintained, is influenced not only by speed but also by slopes. Elite runners perform more efficiently than amateur runners who have less experience.