Accepted Manuscript The Influence of Experimental Anterior Knee Pain During Running on Electromyography and Articular Cartilage Metabolism W. Matt Denning , MS Scott Woodland , MS Jason G. Winward , Michael G. Leavitt , MS Allen C. Parcell , PhD J. Ty Hopkins , PhD Devin Francom , MS Matthew K. Seeley , PhD PII:
S1063-4584(14)01085-1
DOI:
10.1016/j.joca.2014.05.006
Reference:
YJOCA 3149
To appear in:
Osteoarthritis and Cartilage
Received Date: 18 November 2013 Revised Date:
1 May 2014
Accepted Date: 7 May 2014
Please cite this article as: Denning WM, Woodland S, Winward JG, Leavitt MG, Parcell AC, Hopkins JT, Francom D, Seeley MK, The Influence of Experimental Anterior Knee Pain During Running on Electromyography and Articular Cartilage Metabolism, Osteoarthritis and Cartilage (2014), doi: 10.1016/ j.joca.2014.05.006. This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
ACCEPTED MANUSCRIPT 1 Title The Influence of Experimental Anterior Knee Pain During Running on Electromyography and Articular Cartilage Metabolism
RI PT
Authors
Jason G. Winward Brigham Young University
[email protected] Allen C. Parcell, PhD Brigham Young University
[email protected] EP
J. Ty Hopkins, PhD Brigham Young University
[email protected] TE D
Michael G. Leavitt, MS Brigham Young University
[email protected] M AN U
Scott Woodland, MS Brigham Young University
[email protected] SC
W. Matt Denning, MS 106 SFH Brigham Young University Provo, UT 84602, USA Phone: 1.801.422.9156 Email:
[email protected] Devin Francom, MS University of California Santa Cruz
[email protected] AC C
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46
Matthew K. Seeley, PhD Brigham Young Univeristy
[email protected] Running Title
Experimental knee pain, EMG, and COMP
ACCEPTED MANUSCRIPT 2 Abstract
48
Objective: To determine whether anterior knee pain (AKP), during running, acutely affects
49
lower-extremity electromyography (EMG) and articular cartilage metabolism.
50
Methods: A within-subjects design was used. Each of 12 able-bodied subjects ran on a treadmill
51
for 30 minutes for three different sessions: control (no infusion), sham (0.9% NaCl infusion into
52
the involved-leg infrapatellar fat pad), and pain (5.0% NaCl infusion into the involved-leg
53
infrapatellar fat pad). Bilateral surface EMG was monitored for the vastus medialis (VM), vastus
54
lateralis (VL), and gastrocnemius (GA). Serum cartilage oligomeric matrix protein (COMP)
55
concentration was determined before, after, and 60 minutes after the run. A functional analysis
56
approach was used to compare EMG amplitude, across the entire stance phase, between sessions
57
and legs. Mixed-model analysis of covariance was used to compare serum COMP concentration
58
between sessions, across time.
59
Results: Relative to the uninvolved leg, greater involved-leg VL and GA EMG amplitude existed
60
during midstance for the sham and control sessions (p < 0.01). During the painful session,
61
however, involved-leg VM, VL, and GA EMG amplitude was 5-10% less than for the
62
uninvolved leg. COMP concentration immediately post-run was 14% and 21% greater than pre-
63
run (P = 0.01) and 60 minutes post-run concentrations (P < 0.01), respectively. Session,
64
however, did not significantly influence COMP.
65
Conclusion: During a 30-minute run, AKP acutely alters midstance VM, VL, and GA EMG
66
amplitude. AKP during a 30-minute run does not, however, acutely influence articular cartilage
67
metabolism.
68
Key Words: COMP; Knee Pain; Electromyography; Running
AC C
EP
TE D
M AN U
SC
RI PT
47
ACCEPTED MANUSCRIPT 3 69 70
Introduction Knee pathology and the associated pain is common; 600,000 total knee arthroplasties were performed in the United States in 20101. Cram et al.1 estimated that arthroplasty, alone, cost
72
the Unites States 9 billion dollars in 2009, and that 3 million procedures will be performed
73
annually by 20302. Although signs and symptoms of knee pathologies vary, pain is a chief
74
symptom3. Knee pain has even been described as a musculoskeletal epidemic4. Knee pathology
75
and the associated pain are especially common for distance runners. Although distance running
76
has not been directly linked to a particular knee injury, as many as 79% of all distance runners
77
become injured and 50% of these injuries are knee injuries5. Combining all running injures, 42%
78
affect the knee, and anterior knee pain (AKP) is associated with 62% of these knee injuries6.
SC
M AN U
79
RI PT
71
AKP is associated with abnormal lower-extremity muscle activation characteristics7-9. Lower-extremity activation characteristics influence knee joint load, and the activation
81
characteristics for the vastii and gastrocnemius (GA) are especially influential on knee load10,11.
82
AKP is also associated with abnormal lower-extremity kinematics12-14. Small changes in
83
kinematics can influence knee load10. Because AKP causes neural alterations, and because
84
articular cartilage is sensitive to lower-extremity kinematics15-17 and the resulting mechanical
85
conditions18,19, AKP likely affects articular cartilage. It is unclear, however, how long AKP must
86
persist before articular cartilage condition is affected. In felines, knee joint degeneration can
87
begin to occur, as a direct result of decreased muscle force during cyclical loading, in as little as
88
30 minutes20. An experimental AKP model that allows researchers to acutely produce AKP in
89
otherwise able-bodied subjects could allow researchers to evaluate potential acute effects of AKP
90
on articular cartilage, independent of other potentially confounding knee pathology factors (e.g.,
91
knee joint effusion or degradation)14,21.
AC C
EP
TE D
80
ACCEPTED MANUSCRIPT 4 92
Cartilage oligomeric matrix protein (COMP) is a noncollagenous extracellular matrix protein that is synthesized by chondrocytes and helps maintain the collagen matrix integrity by
94
binding to type II collagen fibers; COMP also contributes to the mechanical properties of
95
articular cartilage and is sensitive to load22-25. COMP is an established biomarker that reflects
96
cartilage metabolism in osteoarthritis (OA)26,27 and otherwise injured knee joints28. COMP has
97
been used to monitor articular cartilage degradation progress23,29-31. Additionally, COMP
98
concentration increases as a result of exercise in able-bodied, asymptomatic individuals, which
99
demonstrates that COMP concentration reflects normal articular cartilage turnover15,32-34. Serum
100
COMP concentration is influenced by knee load characteristics15,35, and has also been indirectly
101
related to muscle activation patterns for lower-extremity muscles during running32.
M AN U
SC
RI PT
93
The purpose of this study was to determine whether experimental AKP during distance
103
running acutely affects lower-extremity muscle activation characteristics and articular cartilage
104
metabolism that results from running. We hypothesized that experimental AKP during running
105
would decrease muscle activation for the involved leg, relative to the uninvolved leg. Also, we
106
hypothesized that experimental AKP during running would enlarge the increase in serum COMP
107
that is typically associated with running.
108
Methods
109
Subjects
EP
AC C
110
TE D
102
Twelve relatively young, able-bodied volunteers (seven males and five females; mean ± 1
111
standard deviation; age = 22 ± 4 years; mass = 72 ± 28 kg; height =1.79 ± 0.27 m; body mass
112
index = 22.2 ± 1.9 kg/m2) were recruited from a university setting. It was intended that these
113
subjects represent a young able-bodied, asymptomatic population; free from lower-extremity
ACCEPTED MANUSCRIPT 5 pathology and injury. Each subject participated in three different data collection sessions: pain,
115
control, and sham. No subject reported a history of lower-extremity injury within six months
116
prior to participation or any knee surgery in their lifetime. Because physical activity habits and
117
body mass index influence serum COMP concentration before and after physical activity35,
118
subjects with similar physical activity levels were recruited. Each subject was also required to (1)
119
be participating in moderate physical activity, as defined by the World Health Organization, at
120
least three times a week, and (2) have a normal body mass index (between 18.5 and 25). Subjects
121
were required to refrain from exercise between 24 hours prior to the first data collection session
122
and the completion of the final data collection session. Prior to participation, subjects completed
123
an informed consent form that was approved by the appropriate institutional review board. All
124
study procedures complied with the Helsinki Declaration. We considered sample size, a priori,
125
using pilot EMG data (EMG being the most variable) and a multivariate analysis approach.
126
Using this approach, however, the estimated sample size that was required to achieve 80% power
127
at a significance level of 0.05 was far beyond the scope of the current study. Consequently, a
128
feasible sample size of 12 subjects was chosen.
129
Experimental Protocol
SC
M AN U
TE D
EP
Subjects completed the three different data collection sessions in a counterbalanced
AC C
130
RI PT
114
131
fashion, 48 hours apart. For each session, subjects ran for 30 minutes on the same treadmill
132
(Quinton Cardiology, Bothwell, WA, USA) at a grade of zero. For the control session, subjects
133
ran on the treadmill in a typical fashion. For the sham or pain sessions, subjects ran while
134
receiving a continuous infusion of isotonic (0.9% NaCl) or hypertonic saline solution,
135
respectively, into the right (involved) infrapatellar fat pad, at a rate of 0.216 ml · min-1 (total
136
infused volume = 8 ml). This pain model (an identical solution concentration and volume) was
ACCEPTED MANUSCRIPT 6 used previously to induce consistent and satisfactory experimental AKP levels during running14.
138
For all data collection sessions, subjects ran in the same pair of their own running shoes, and
139
running shorts and shirts that were provided by the investigators.
140
RI PT
137
At the beginning of each data collection session, subjects rested in a seated position for 30 minutes to limit potential influence of preceding activity on serum COMP concentration.
142
Subjects then stood for 10 minutes to allow for body fluid distribution to adjust to the vertical
143
posture; electromyography (EMG) surface electrodes were applied to the subject during this
144
standing period. Next, the baseline blood sample was drawn. Subjects then completed one of
145
three running exercises: control (no infusion), sham (isotonic saline infusion), or pain
146
(hypertonic saline infusion). A second blood sample was drawn immediately after the 30-minute
147
run. After this blood draw, subjects rested in a seated position for 60 minutes, and a final blood
148
sample was then drawn. The general timeline for each session is shown in Figure 1.
M AN U
TE D
149
SC
141
For each data collection session, after the baseline blood sample was collected and EMG sensors were applied, subjects warmed up by running on the treadmill at one of three speeds (3.0,
151
3.5, or 4.0 m/s) for 5 minutes. The subjects were instructed to select the greatest of these three
152
speeds that they could run at continuously for 30 minutes. Subjects ran at the same speed for all
153
data collection sessions. After this 5-minute warm-up, subjects lay supine on a treatment table
154
while the isotonic or hypertonic saline infusions were initiated, for the sham or pain sessions,
155
respectively. Prior to these infusions, the skin was shaved, if needed, and prepared with iodine
156
and an alcohol wipe. The catheter was inserted into the lateral superior infrapatellar fat pad of the
157
right leg using a previously described procedure14. After placement, the catheter was secured
158
with tape, and used a plastic 0.762-m connection tube to connect the catheter (B. Braun Medical
159
Inc., Melsungen, Germany) to a portable syringe pump (Graseby Medical Ltd., Watford,
AC C
EP
150
ACCEPTED MANUSCRIPT 7 England, UK). Throughout the sham and pain running sessions, this pump was held in a pocket
161
that was sewn into a spandex running shirt (Figure 2). Subjects and investigators were blinded
162
regarding the saline treatment that was being administered. After the syringe pump was initiated,
163
subjects lay supine for three minutes, sat upright for two minutes, and then stood for two
164
minutes, in an attempt to familiarize subjects to the experimental AKP and decrease the
165
likelihood of syncope. The subjects then ran for 30 minutes at the selected speed while the saline
166
was infused. The control session was identical to the sham and pain sessions, except that a
167
catheter was not inserted and no infusion was administered. For each session, a 100-mm visual
168
analog scale was used to measure subject-perceived AKP immediately prior to and after the
169
needle stick, and then every three minutes until 60 minutes after the completion of the run.
170
Electromyography
SC
M AN U
Surface EMG (1000 Hz; Delsys, Boston, MA, USA) was used to monitor bilateral muscle
TE D
171
RI PT
160
activation patterns for the gastrocnemius (GA), vastus medialis (VM), and vastus lateralis (VL).
173
Previously described recommendations for skin preparation procedures and electrode application
174
locations were used36. Thirty seconds of EMG data at two different times of the run (10 and 20
175
minutes after beginning the run) were collected, and then processed the EMG data using custom
176
algorithms that were written in MatLab (MathWorks, Natick, MA, USA). The DC offset was
177
first removed and then a root mean square algorithm (moving window = 50ms) was applied to
178
rectify and smooth the EMG signal32. In an effort to minimize the influence of electrode
179
placement, the amplitude of the smoothed signal was then normalized to a reference value that
180
was the average of ten amplitude peaks from ten different stance phases during the warm-up run.
181
Consequently, all EMG amplitude data were reported as a percentage of this mean peak
182
amplitude (i.e., from the warm-up run). Also, the EMG signal was time normalized to 100% of
AC C
EP
172
ACCEPTED MANUSCRIPT 8 the stance phase, for five stance phases from 10 minutes after the beginning of the run and five
184
stance phases from 20 minutes after the beginning of the run. Stance phase was manually
185
determined using ten high-speed video cameras (200 Hz; VICON, Santa Rosa, CA, USA) that
186
completely surrounded each subject, and reflective markers that were placed bilaterally on the
187
heel and toe. The video and EMG data were synchronized.
188
Serum COMP
SC
RI PT
183
All blood samples (6 ml) were collected from an antecubital vein using a 20 gage
190
shielded I.V. catheter (Becton Dickinson & Company, Franklin Lakes, NJ, USA). Immediately
191
after catheter placement and every 15 minutes thereafter, the catheter was flushed with 1ml of
192
isotonic saline (0.9% NaCl) to prevent clotting. A 1ml waste sample was drawn before each
193
blood sample was collected. Blood samples were placed in EDTA vacutainers (Becton
194
Dickinson & Company, Franklin Lakes, NJ, USA), centrifuged with an Eppendorf 5403
195
refrigerated centrifuge (Hamburg, Germany) for 15 min at 3,000 × gravity, and stored in a
196
freezer at -20°C until further analysis. Serum COMP concentration was determined using a
197
commercially available enzyme-linked immunosortbent assay kit (ELISA; R&D Systems, Inc.,
198
Minneapolis, MN, USA). Serum samples were analyzed in triplicate across each ELISA plate
199
according to the manufacturer’s instructions. The present ELISA assay kits had inter- and intra-
200
plate coefficients of variation of 13.66% and 2.61%, respectively, for a 273.3 ± 33.8 ng/ml
201
sample. Differences due to inter-plate variation were minimized by comparing serum COMP
202
concentration within subjects and by testing subject replicates on the same plate.
203
Statistical Analysis
AC C
EP
TE D
M AN U
189
ACCEPTED MANUSCRIPT 9 204
The independent variables for this study were session, time, and leg. The dependent variables were subject-perceived AKP, EMG amplitude, and serum COMP concentration. We
206
assumed that the current data would be normally distributed. Prior to statistical analyses, this
207
assumption was confirmed via visual inspection. Another assumption that our statistical
208
procedure employed, and that we feel is logical, was that the true EMG functions are continuous
209
(by true EMG functions we mean that we observed representations of some underlying EMG
210
over time, which we would call the true EMG functions). A mixed model repeated measures
211
ANOVA was used to compare subject-perceived AKP between sessions every three minutes
212
throughout most of each session (Figure 3; α = 0.05). A mixed model repeated measures
213
ANOVA was also used to compare serum COMP concentration between sessions across three
214
times: (1) baseline (pre-run), (2) immediately post-run, and (3) one hour post-run (Figure 1; α =
215
0.05). If a session × time interaction was detected for subject-perceived AKP or serum COMP
216
concentration, Tukey’s post hoc comparisons were used to evaluate potential between-time
217
differences for each session. Functional linear models were used to evaluate influence of session
218
and leg (involved and uninvolved) on EMG amplitude, modeled as polynomial functions, across
219
the stance phase of running. This functional analysis evaluated the effect of session and leg
220
across all of stance (heel strike to toe-off) in a single statistical test, rather than using multiple
221
tests at separate discrete time points within stance, increasing statistical power. For this particular
222
analysis, differences were significant if the mean difference and corresponding 99% confidence
223
interval crossed zero. All ANOVA analyses were performed using JMP Pro 10 software (SAS
224
Institute Inc., Cary, NC, USA). The functional linear models were produced using R statistical
225
software.
AC C
EP
TE D
M AN U
SC
RI PT
205
ACCEPTED MANUSCRIPT 10 226 227
Results Average running speed for the present sample was 3.46 ± 0.43 m/s. A significant session × time interaction was observed for subject-perceived AKP (P < 0.01). Throughout the 30-
229
minute run, subject-perceived AKP was significantly greater during the pain session (30 mm,
230
95% confidence interval (CI): 27, 33), relative to the sham (5 mm, 95% CI: 4, 6) and control
231
sessions (0 mm, 95% CI: 0, 0; Figure 3).The subject-perceived AKP significantly influenced
232
EMG amplitude at various times within stance for all of the observed muscles. No bilateral EMG
233
amplitude differences larger than 10-14%, however, were observed. For the control and sham
234
sessions, involved-leg GA EMG amplitude was 5-10% greater than uninvolved-leg GA EMG
235
amplitude throughout mid-stance (P < 0.01; Figures 4B, 4C, 4E and 4F). Conversely, for the
236
pain session, involved-leg GA EMG amplitude was 5% less than uninvolved-leg GA EMG
237
amplitude for a short time around 30% of the stance phase (P < 0.01; Figure 4A and 4D). The
238
VM and VL exhibited similar alterations due to AKP. For VM EMG amplitude, bilateral
239
symmetry existed for most of the stance phase during the control and sham sessions (Figures 5B,
240
5C, 5E, and 5F). For the pain session, however, involved-leg VM EMG amplitude was 5-10%
241
less than uninvolved-leg VM EMG amplitude for much of the middle part of stance (P < 0.01;
242
Figures 5A and 5D). For VL EMG amplitude, bilateral asymmetries existed for each session, in
243
the same direction: VL EMG amplitude for the uninvolved leg was ~14%, ~10%, and ~6%
244
greater than for the involved leg during parts of midstance for the pain, sham, and control
245
sessions, respectively (P < 0.01; Figures 6A-6F). Time significantly influenced serum COMP
246
concentration as a main effect (P < 0.001; Figure 7). Average serum COMP concentration,
247
immediately post-run (132 ng·ml-1, 95% CI: 94, 170), was 14% greater than immediately pre-run
248
(116 ng·ml-1, 95% CI: 78, 154, P = 0.0092) and 21% greater than 60 minutes post-run (109
AC C
EP
TE D
M AN U
SC
RI PT
228
ACCEPTED MANUSCRIPT 11 ng·ml-1, 95% CI: 70, 147; P= 0.0002). Serum COMP concentration was not significantly
250
different between pre-run and 60 minutes post-run (P = 0.29). No session × time interaction for
251
serum COMP concentration (P = 0.57; Figure 7) was observed. Furthermore, as a main effect,
252
session did not significantly affect serum COMP concentration (P = 0.48). When certain subject
253
characteristics (running speed, gender, age, and body mass index) were included as covariates,
254
all of the reported results remained the same (Appendix A).
255
Discussion
SC
This study was performed to understand whether AKP during distance running acutely
M AN U
256
RI PT
249
affects lower-extremity muscle activation and resulting articular cartilage metabolism. It was
258
hypothesized that, relative to pain free running, running for 30 minutes with AKP would (1)
259
decrease EMG amplitude for the involved leg, relative to the uninvolved leg, and (2) enlarge the
260
serum COMP concentration increase that typically results from running. The present results
261
supported the first hypothesis, but did not support the second hypothesis: EMG amplitudes for
262
the involved leg (for muscles that significantly influence knee joint load) were decreased,
263
relative to the uninvolved leg, during AKP running; however, pain free and painful running had
264
similar effects on serum COMP concentration.
EP
AC C
265
TE D
257
The observed EMG alterations are important findings from this study. Although knee
266
load was not directly measured, the present decreases for GA, VM, and VL EMG amplitude that
267
were due to AKP likely indicate altered knee load: muscle forces significantly and directly
268
impact tibial-femoral contact load10,20,37. Specifically, vastii and GA activity influence knee load
269
during the stance phase of human locomotion10,11. Although it is admittedly difficult to determine
270
whether EMG differences of 10-14% are clinically meaningful, the observed reduction for VM
ACCEPTED MANUSCRIPT 12 EMG amplitude likely contributes to abnormal patellar tracking within the trochlear groove38,39.
272
This idea is congruent to the clinical treatment of AKP in young adults that is strengthening the
273
oblique VM40. Further, the present statistical analyses of EMG involved confidence intervals,
274
which are one way to evaluate clinical significance41. Whenever present confidence intervals did
275
not cross the zero line (Figures 4-6), the described differences can be considered as clinically
276
significant41.
SC
277
RI PT
271
Normal knee joint load that (1) results from lower-extremity muscle force and (2) is required to perform everyday function is important to maintain articular cartilage. For example,
279
reduced quadriceps strength has been linked to symptomatic and morphological progression of
280
knee osteoarthritis42. Additionally, decreased lower-extremity muscle force during cyclical
281
loading has been associated with joint degeneration in as little as 30 minutes20. To date, it
282
remains unclear if experimental AKP can influence the onset and/or progression of articular
283
cartilage degradation. Further research is required to investigate the potential effect of persistent
284
experimental AKP (> 30 minutes) on EMG and serum COMP concentration.
TE D
M AN U
278
The present findings suggest that AKP alters neuromuscular activation patterns.
286
Previously, researchers have theorized that AKP is a result of altered neuromuscular activation
287
patterns43,44. The present data indicate that the reverse is also true: altered EMG patterns result
288
from AKP. Additionally, bilateral EMG data from the present control session demonstrated
289
subtle asymmetries for neural activation for the GA, VM, and VL muscles during able-bodied
290
running (Figures 4-6). This fits with previous research documenting lower-extremity EMG
291
asymmetries for the GA, VM, and VL muscles during able-bodied walking45,46. While
292
attempting to interpret the present EMG data, we were surprised to learn that no one has
293
previously evaluated bilateral symmetry for EMG during able-bodied running, especially using
AC C
EP
285
ACCEPTED MANUSCRIPT 13 the relatively novel statistical approach (functional linear models) that was presently used. Like
295
other previously published EMG data, the reliability of the present EMG data is limited due to
296
numerous issues that are inherent to EMG (e.g., varying electrode placement, muscle fiber
297
length, and contraction velocity). However, efforts were made to minimize such limitations
298
where possible (e.g., normalization of EMG amplitude).
The present COMP data and data from previous researchers15,16,33,34, demonstrate that the
SC
299
RI PT
294
magnitude and duration of serum COMP concentration increase, due to physical activity, is
301
dependent upon exercise duration and intensity. For example, serum COMP concentration
302
increases 90% and 60% as a result from a 200-km ultramarathon and marathon33, respectively,
303
while only a 9.7% serum COMP concentration increase resulted from a less intense 30-minute
304
walk35. The present observation of a 14% increase for serum COMP concentration (Figure 7),
305
resulting from a 30-minute run, fits with these previously reported data. Additionally, walking
306
for 30 minutes resulted in an increased COMP concentration duration of 30 minutes35, while
307
running durations of 30 minutes16, 3-4 hours34, and 3-33 hours33 resulted in elevated COMP
308
concentration for 1hour, 24 hours, and 6 days, respectively. Fitting with previous results15,16, the
309
present COMP concentration returned to baseline levels within 1 hour (Figure 7). It appears that
310
exercise sessions involving greater durations and intensities, and greater cumulative knee load,
311
elicit greater amounts of COMP released from articular cartilage.
TE D
EP
AC C
312
M AN U
300
Interpretation of increased COMP, as a result of exercise, is unclear. Although some
313
authors have hypothesized that elevated COMP concentration may indicate detrimental articular
314
cartilage degradation15,34, other authors have speculated that increased COMP concentration
315
represents typical, healthy articular cartilage turnover16,33. Previously, researchers15,32,35 have
316
demonstrated that as serum COMP concentration increases, articular cartilage volume decreases,
ACCEPTED MANUSCRIPT 14 even in a healthy population. Kim et al.33 speculated that running distance and the corresponding
318
magnitude and duration of COMP concentration increase may each be positively correlated with
319
the degree of articular cartilage damage. It is assumed that the presently observed COMP
320
concentration increases reflected healthy articular cartilage turnover. The present subjects ran, on
321
average, only 6.2 Km, and COMP levels increased only 14% and returned to pre exercise levels
322
in less than 60 minutes (Figure 7). The present data are limited to a single running session, and
323
more data are needed to determine the chronic effects of AKP on articular cartilage health.
SC
The use of this experimental AKP model to understand effects of pain on articular
M AN U
324
RI PT
317
cartilage is another novel feature of this study. This pain model was presently used to induce
326
consistent experimental AKP during a 30-minute run (Figure 3). This 30-minute duration is
327
nearly 70% greater than the previous longest duration14. For this reason, the present results
328
further demonstrate that this pain model can be used to study independent effects of AKP during
329
various exercises over relatively extended durations. As with any novel method, however, certain
330
aspects related to the present AKP model are unclear. Although the magnitude of subject-
331
perceived AKP (Figure 3) appears to have been sufficient to represent clinical AKP47, it is
332
unclear whether other characteristics of this AKP pain are clinically representative. However,
333
previous researchers propose that experimental pain, induced by hypertonic saline, is
334
representative of joint pain related to knee pathology21,48-50. The nature, quality, and distribution
335
of pain elicited by hypertonic saline has been shown to be similar to clinical AKP21. Also, results
336
from animal experiments suggest that group III and IV nociceptive afferents are active due to
337
hypertonic saline49,50, providing more evidence that the pathways involved during this
338
experimental pain are similar to musculoskeletal pain. Furthermore, hypertonic saline induces
AC C
EP
TE D
325
ACCEPTED MANUSCRIPT 15 339
the release of substance P from C-fiber neurons, which is also consistent with musculoskeletal
340
pain48. In conclusion, there are three important findings from this study. First, experimental AKP
RI PT
341
generally caused involved-leg EMG amplitude to decrease, during running, relative to
343
uninvolved-leg EMG amplitude (Figures 4-6). Second, experimental AKP did not significantly
344
affect systemic serum COMP concentration increase due to running (Figure 4). Third, the
345
infusion of hypertonic saline effectively produced consistent subject-perceived AKP throughout
346
a 30-minute run (Figure 3).
347
Author contributions
M AN U
SC
342
Conception and design: Woodland, Parcell, Hopkins, Seeley
349
Acquisition of data: Denning, Woodland, Winward, Leavitt, Parcell, Seeley
350
Analysis and interpretation of data: Denning, Woodland, Leavitt, Parcell, Hopkins,
353 354
Drafting and revising of article: Denning, Woodland, Winward, Leavitt, Parcell, Hopkins, Francom, Seeley
EP
352
Francom, Seeley
Final approval of the version of the article to be published: Denning, Woodland,
AC C
351
TE D
348
355
Winward, Leavitt, Parcell, Hopkins, Francom, Seeley
356
Competing Interests Statement
357
There are no completing interests for any of the authors.
358
Appendix A. Summary of the serum COMP results when running speed, gender, age, and body
359
mass index were used as covariates. None of the covariates included in this model were
ACCEPTED MANUSCRIPT 16 statistically significant. Neither running speed (P = 0.55), gender (P = 0.63), age (P = 0.71), nor
361
body mass index (P = 0.50) influenced serum COMP concentration. Time was the only
362
significant main effect that influenced serum COMP (P < 0.001).
363
References
364
1.
RI PT
360
Cram P, Lu X, Kates SL, Singh JA, Li Y, Wolf BR. Total knee arthroplasty volume, utilization, and outcomes among Medicare beneficiaries, 1991-2010. JAMA
366
2012;308(12):1227-36. 2.
Kurtz SM, Lau E, Ong K, Zhao K, Kelly M, Bozic KJ. Future young patient demand for
M AN U
367
SC
365
368
primary and revision joint replacement: national projections from 2010 to 2030.
369
Clin Orthop Relat Res 2009;467(10):2606-12.
370
3.
Merle-Vincent F, Couris CM, Schott AM, Perier M, Conrozier S, Conrozier T, et al. Cross-sectional study of pain and disability at knee replacement surgery for
372
osteoarthritis in 299 patients. Joint Bone Spine 2007;74(6):612-6. 4.
5.
van Gent RN, Siem D, van Middelkoop M, van Os AG, Bierma-Zeinstra SM, Koes BW. Incidence and determinants of lower extremity running injuries in long distance
376
runners: a systematic review. Br J Sports Med 2007;41(8):469-80; discussion 480.
377 378
EP
Rheum Dis 2001;60(2):89-90.
374 375
Symmons DP. Knee pain in older adults: the latest musculoskeletal "epidemic". Ann
AC C
373
TE D
371
6.
Taunton JE, Ryan MB, Clement DB, McKenzie DC, Lloyd-Smith DR, Zumbo BD. A
379
retrospective case-control analysis of 2002 running injuries. Br J Sports Med
380
2002;36(2):95-101.
ACCEPTED MANUSCRIPT 17 381
7.
Toumi H, Best TM, Pinti A, Lavet C, Benhamou CL, Lespessailles E. The role of muscle
382
strength & activation patterns in patellofemoral pain. Clin Biomech2013;28(5):544-
383
8. 8.
Van Tiggelen D, Cowan S, Coorevits P, Duvigneaud N, Witvrouw E. Delayed Vastus
RI PT
384
Medialis Obliquus to Vastus Lateralis Onset Timing Contributes to the Development
386
of Patellofemoral Pain in Previously Healthy Men A Prospective Study. Am J Sports
387
Med 2009;37(6):1099-105.
388
9.
SC
385
Willson JD, Kernozek TW, Arndt RL, Reznichek DA, Scott Straker J. Gluteal muscle activation during running in females with and without patellofemoral pain
390
syndrome. Clin Biomech 2011;26(7):735-40. 10.
force during normal walking. J Biomech 2010;43(14):2780-4.
392 393
Sasaki K, Neptune RR. Individual muscle contributions to the axial knee joint contact
11.
Shelburne KB, Torry MR, Pandy MG. Contributions of muscles, ligaments, and the
TE D
391
M AN U
389
394
ground-reaction force to tibiofemoral joint loading during normal gait. J Orthop Res
395
2006;24(10):1983-90.
Boehm JE. Effect of patellofemoral pain on strength and mechanics after an
397
exhaustive run. Med Sci Sports Exerc 2013;45(7):1331-9.
398 399
13.
range of motion boundary. Gait Posture 2013.
401
403
Rodrigues P, Tenbroek T, Van Emmerik R, Hamill J. Evaluating runners with and without anterior knee pain using the time to contact the ankle joint complexes'
400
402
Bazett-Jones DM, Cobb SC, Huddleston WE, O'Connor KM, Armstrong BS, Earl-
EP
12.
AC C
396
14.
Seeley MK, Park J, King D, Hopkins JT. A novel experimental knee-pain model affects perceived pain and movement biomechanics. J Athl Train 2013;48(3):337-45.
ACCEPTED MANUSCRIPT 18 404
15.
Niehoff A, Kersting UG, Helling S, Dargel J, Maurer J, Thevis M, et al. Different
405
mechanical loading protocols influence serum cartilage oligomeric matrix protein
406
levels in young healthy humans. Eur J Appl Physiol 2010;110(3):651-7. 16.
Niehoff A, Muller M, Bruggemann L, Savage T, Zaucke F, Eckstein F, et al.
RI PT
407
Deformational behaviour of knee cartilage and changes in serum cartilage
409
oligomeric matrix protein (COMP) after running and drop landing. Osteoarthritis
410
Cartilage 2011;19(8):1003-10.
411
17.
SC
408
Teichtahl AJ, Davies-Tuck ML, Wluka AE, Jones G, Cicuttini FM. Change in knee angle influences the rate of medial tibial cartilage volume loss in knee osteoarthritis.
413
Osteoarthritis Cartilage 2009;17(1):8-11. 18.
early cartilage degeneration. Comput Math Methods Med 2013;2013:862903. 19.
mechanobiology: a research perspective. Bone 2003;33(1):1-13.
417 20.
degeneration. Langenbecks Arch Surg 2003;388(5):305-15.
419 420
21.
Res 2004;22(1):116-21.
422
424
Bennell K, Hodges P, Mellor R, Bexander C, Souvlis T. The nature of anterior knee pain following injection of hypertonic saline into the infrapatellar fat pad. J Orthop
421
423
Herzog W, Longino D, Clark A. The role of muscles in joint adaptation and
EP
418
Wong M, Carter DR. Articular cartilage functional histomorphology and
TE D
415 416
Dabiri Y, Li LP. Altered knee joint mechanics in simple compression associated with
AC C
414
M AN U
412
22.
Giannoni P, Siegrist M, Hunziker EB, Wong M. The mechanosensitivity of cartilage oligomeric matrix protein (COMP). Biorheology 2003;40(1-3):101-9.
ACCEPTED MANUSCRIPT 19 425
23.
Hedbom E, Antonsson P, Hjerpe A, Aeschlimann D, Paulsson M, Rosapimentel E, et
426
al. Cartilage Matrix Proteins - an Acidic Oligomeric Protein (Comp) Detected Only in
427
Cartilage. J Biol Chem 1992;267(9):6132-6. 24.
Rosenberg K, Olsson H, Morgelin M, Heinegard D. Cartilage oligomeric matrix
RI PT
428 429
protein shows high affinity zinc-dependent interaction with triple helical collagen. J
430
Biol Chem 1998;273(32):20397-403.
rheumatoid arthritis and knee osteoarthritis. Clin Rheumatol 2005;24(3):278-84.
432 433
Wislowska M, Jablonska B. Serum cartilage oligomeric matrix protein (COMP) in
SC
25.
26.
Petersson IF, Boegard T, Svensson B, Heinegard D, Saxne T. Changes in cartilage and
M AN U
431
434
bone metabolism identified by serum markers in early osteoarthritis of the knee
435
joint. Brit J Rheumatol 1998;37(1):46-50.
436
27.
Saxne T, Heinegard D. Cartilage Oligomeric Matrix Protein - a Novel Marker of Cartilage Turnover Detectable in Synovial-Fluid and Blood. Brit J Rheumatol
438
1992;31(9):583-91.
439
28.
TE D
437
Dahlberg L, Roos H, Saxne T, Heinegard D, Lark MW, Hoerrner LA, et al. Cartilage Metabolism in the Injured and Uninjured Knee of the Same Patient. Ann Rheum Dis
441
1994;53(12):823-7.
443 444 445
29.
AC C
442
EP
440
Garnero P, Piperno M, Gineyts E, Christgau S, Delmas PD, Vignon E. Cross sectional evaluation of biochemical markers of bone, cartilage, and synovial tissue metabolism in patients with knee osteoarthritis: relations with disease activity and joint damage. Ann Rheum Dis 2001;60(6):619-26.
ACCEPTED MANUSCRIPT 20 446
30.
Kuhne SAM, Neidhart M, Everson MP, Hantzschel H, Fine PR, Gay S, et al. Persistent
447
high serum levels of cartilage oligomeric matrix protein in a subgroup of patients
448
with traumatic knee injury. Rheumatol Int 1998;18(1):21-5. 31.
Neidhart M, Hauser N, Paulsson M, DiCesare PE, Michel BA, Häuselmann HJ. Small
RI PT
449 450
fragments of cartilage oligomeric matrix protein in synovial fluid and serum as
451
markers for cartilage degradation. Br J Rheumatol 1997;36(11):1151-60. 32.
Kersting UG, Stubendorff JJ, Schmidt MC, Bruggemann GP. Changes in knee cartilage
SC
452
volume and serum COMP concentration after running exercise. Osteoarthritis
454
Cartilage 2005;13(10):925-34.
455
33.
M AN U
453
Kim HJ, Lee YH, Kim CK. Changes in serum cartilage oligomeric matrix protein (COMP), plasma CPK and plasma hs-CRP in relation to running distance in a
457
marathon (42.195 km) and an ultra-marathon (200 km) race. Eur J Appl Physiol
458
2009;105(5):765-70.
459
34.
TE D
456
Neidhart M, Muller-Ladner U, Frey W, Bosserhoff AK, Colombani PC, Frey-Rindova P, et al. Increased serum levels of non-collagenous matrix proteins (cartilage
461
oligomeric matrix protein and melanoma inhibitory activity) in marathon runners.
462
Osteoarthritis Cartilage 2000;8(3):222-9. 35.
Mundermann A, Dyrby CO, Andriacchi TP, King KB. Serum concentration of cartilage oligomeric matrix protein (COMP) is sensitive to physiological cyclic loading in
464
healthy adults. Osteoarthritis Cartilage 2005;13(1):34-8.
465 466
AC C
463
EP
460
36.
Hermens HJ, Freriks B, Disselhorst-Klug C, Rau G. Development of recommendations
467
for SEMG sensors and sensor placement procedures. J Electromyogr Kinesiol
468
2000;10(5):361-74.
ACCEPTED MANUSCRIPT 21 469
37.
Manal K, Buchanan TS. An electromyogram-driven musculoskeletal model of the
470
knee to predict in vivo joint contact forces during normal and novel gait patterns. J
471
Biomech Eng 2013;135(2):021014. 38.
Lin F, Wilson NA, Makhsous M, Press JM, Koh JL, Nuber GW, et al. In vivo patellar
RI PT
472 473
tracking induced by individual quadriceps components in individuals with
474
patellofemoral pain. J Biomech 2010;43(2):235-41. 39.
MacIntyre NJ, Hill NA, Fellows RA, Ellis RE, Wilson DR. Patellofemoral joint
SC
475
kinematics in individuals with and without patellofemoral pain syndrome. J Bone
477
Joint Surg Am 2006;88(12):2596-605.
478
40.
M AN U
476
Cowan SM, Bennell KL, Crossley KM, Hodges PW, McConnell J. Physical therapy
479
alters recruitment of the vasti in patellofemoral pain syndrome. Med Sci Sports Exerc
480
2002;34(12):1879-85. 41.
Armijo-Olivo S, Warren S, Fuentes J, Magee DJ. Clinical relevance vs. statistical
TE D
481 482
significance: Using neck outcomes in patients with temporomandibular disorders as
483
an example. Man Ther 2011;16(6):563-72.
Reduced quadriceps strength relative to body weight: a risk factor for knee
485
osteoarthritis in women? Arthritis Rheum 1998;41(11):1951-9.
486 487
43.
Noehren B, Hamill J, Davis I. Prospective evidence for a hip etiology in patellofemoral pain. Med Sci Sports Exerc 2013;45(6):1120-4.
488 489
Slemenda C, Heilman DK, Brandt KD, Katz BP, Mazzuca SA, Braunstein EM, et al.
EP
42.
AC C
484
44.
Pal S, Besier TF, Draper CE, Fredericson M, Gold GE, Beaupre GS, et al. Patellar tilt
490
correlates with vastus lateralis: vastus medialis activation ratio in maltracking
491
patellofemoral pain patients. J Orthop Res 2012;30(6):927-33.
ACCEPTED MANUSCRIPT 22 45.
during Walking in Normal Adults. Electroen Clin Neuro 1989;72(5):429-38.
493 494
46.
Pierotti SE, Brand RA, Gabel RH, Pedersen DR, Clarke WR. Are leg electromyogram profiles symmetrical? J Orthop Res 1991;9(5):720-9.
495 496
Ounpuu S, Winter DA. Bilateral Electromyographical Analysis of the Lower-Limbs
47.
RI PT
492
Boling MC, Padua DA, Alexander Creighton R. Concentric and eccentric torque of the hip musculature in individuals with and without patellofemoral pain. J Athl Train
498
2009;44(1):7-13.
499
48.
SC
497
Garland A, Jordan JE, Necheles J, Alger LE, Scully MM, Miller RJ, et al. Hypertonicity, but not hypothermia, elicits substance P release from rat C-fiber neurons in primary
501
culture. J Clin Invest 1995;95(5):2359-66.
502
49.
M AN U
500
Kumazawa T, Mizumura K. Thin-fibre receptors responding to mechanical, chemical, and thermal stimulation in the skeletal muscle of the dog. J Physiol
504
1977;273(1):179-94.
Paintal AS. Functional analysis of group III afferent fibres of mammalian muscles. J Physiol 1960;152:250-70.
EP
506
50.
AC C
505
TE D
503
ACCEPTED MANUSCRIPT
Table 1. Average (95% CI: lower limit – upper limit) serum COMP concentration (ng·ml-1) for each session and time when running speed, gender, age, and body mass index were used as covariates.
Average
108.7
125.6
97.1
(65.3 – 152.2)
(82.1 – 169.1)
(53.7 – 140.6)
112.9
132.2
109.5
(69.5 – 156.4)
(88.7 – 175.6)
(66.0 – 1 52.9)
120.2
132.5
112.5
(76.7 – 163.6)
(89.1 – 176.0)
113.9
130.1*
(71.1 – 156.7)
(87.3 – 172.9)
RI PT
Pain
60 Minutes Post Run
SC
Sham
Post Run
(69.0 – 155.9) 106.4
M AN U
Control
Pre Run
(63.6 – 149.2)
AC C
EP
TE D
* Significantly different (P < 0.05) from pre run and 60 minute post run concentration.
ACCEPTED MANUSCRIPT 23 501
Figure 1. A description of the general timeline for each data collection session. EMG #1 and
502
EMG #2 indicate the times, within the run, that electromyography data were collected.
503 Figure 2. The knee saline infusion set-up. A pocket (A) that was sewn into the front of a spandex
505
shirt held the portable syringe pump and syringe. A connection tube (B) was used to connect the
506
syringe and the catheter (C) that was inserted into the infrapatellar fat pad.
RI PT
504
SC
507
Figure 3. Means and 95% confidence intervals for subject-perceived pain across time for each
509
data collection session (95% confidence intervals are not included for the control session because
510
they equaled zero at each time point). Subject-perceived pain significantly increased across time
511
during the pain session only (p < 0.01). Asterisks show statistically significant differences when
512
the pain session was compared to the sham and control sessions.
M AN U
508
TE D
513
Figure 4. Grand ensembles (99% confidence intervals are only shown for involved (right) leg to
515
increase clarity) for gastrocnemius (GA) electromyography (EMG) amplitude (Amp), for the
516
pain (A), control (B), and sham (C) sessions. Figures D-F show results of bilateral functional
517
comparisons, where mean bilateral differences and corresponding 99% confidence intervals are
518
plotted as a function of time. When the shaded area does not overlap with the zero line, a
519
bilateral difference (p < 0.01) is indicated. Positive differences indicate that involved-leg (right)
520
EMG is greater, while negative differences indicate the opposite.
AC C
521
EP
514
522
Figure 5. Grand ensembles (99% confidence intervals are only shown for involved (right) leg to
523
increase clarity) for vastus medialis (VM) electromyography (EMG) amplitude (Amp), for the
ACCEPTED MANUSCRIPT 24 pain (A), control (B), and sham (C) sessions. Figures D-F show results of bilateral functional
525
comparisons, where mean bilateral differences and corresponding 99% confidence intervals are
526
plotted as a function of time. When the shaded area does not overlap with the zero line, a
527
bilateral difference (p < 0.01) is indicated. Positive differences indicate that involved-leg (right)
528
EMG is greater, while negative differences indicate the opposite.
529
RI PT
524
Figure 6. Grand ensembles (99% confidence intervals are only shown for involved (right) leg to
531
increase clarity) for vastus lateralis (VL) electromyography (EMG) amplitude (Amp), for the
532
pain (A), control (B), and sham (C) sessions. Figures D-F show results of bilateral functional
533
comparisons, where mean bilateral differences and corresponding 99% confidence intervals are
534
plotted as a function of time. When the shaded area does not overlap with the zero line, a
535
bilateral difference (p < 0.01) is indicated. Positive differences indicate that involved-leg (right)
536
EMG is greater, while negative differences indicate the opposite.
M AN U
TE D
537
SC
530
Figure 7. Means and 95% confidence intervals for serum cartilage oligomeric matrix protein
539
(COMP) concentration at three times for each session. If pooled from each session, serum
540
COMP concentration was significantly greater immediately post run, relative to pre run and 60
541
minutes post run; this trend was consistent between sessions.
AC C
EP
538
AC C
EP
TE D
M AN U
SC
RI PT
ACCEPTED MANUSCRIPT
AC C
EP
TE D
M AN U
SC
RI PT
ACCEPTED MANUSCRIPT
AC C
EP
TE D
M AN U
SC
RI PT
ACCEPTED MANUSCRIPT
AC C
EP
TE D
M AN U
SC
RI PT
ACCEPTED MANUSCRIPT
AC C
EP
TE D
M AN U
SC
RI PT
ACCEPTED MANUSCRIPT
AC C
EP
TE D
M AN U
SC
RI PT
ACCEPTED MANUSCRIPT
AC C
EP
TE D
M AN U
SC
RI PT
ACCEPTED MANUSCRIPT