See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/7448744

Delayed Trunk Muscle Reflex Responses Increase the Risk of Low Back
Injuries

Article  in  Spine · January 2006

DOI: 10.1097/01.brs.0000188273.27463.bc · Source: PubMed

CITATIONS READS

243 1,307

7 authors, including:

Jacek Cholewicki Sheri Silfies

Michigan State University University of South Carolina
130 PUBLICATIONS   9,733 CITATIONS    39 PUBLICATIONS   909 CITATIONS   

SEE PROFILE SEE PROFILE

Norman Peter Reeves

Michigan State University
56 PUBLICATIONS   2,110 CITATIONS   

SEE PROFILE

Some of the authors of this publication are also working on these related projects:

Trunk motor control performance before and after spinal manipulation treatment View project

Changes in lumbar spine kinematics due to instability and chronic pain View project

All content following this page was uploaded by Norman Peter Reeves on 11 October 2017.

The user has requested enhancement of the downloaded file.


Delayed Trunk Muscle Reflex Responses Increase the
Risk of Low Back Injuries
Jacek Cholewicki, PhD,* Sheri P. Silfies, PhD, PT,† Riaz A. Shah, MD,*
Hunter S. Greene, MD,* N. Peter Reeves, MSc,* Kashif Alvi, MD,* and Barry Goldberg, MD*
Study Design. Prospective observational study with a
2- to 3-year follow-up.
Objectives. To determine whether delayed muscle re-
flex response to sudden trunk loading is a result of or a
risk factor for sustaining a low back injury (LBI).
Summary of Background Data. Differences in motor
control have been identified in individuals with chronic
low back pain and in athletes with a history of LBI when
compared with controls. However, it is not known whether
these changes are a risk for or a result of LBI.
Methods. Muscle reflex latencies in response to a
quick force release in trunk flexion, extension, and lateral
bending were measured in 303 college athletes. Informa-
tion was also obtained regarding their personal data, ath-
letic experience, and history of LBI. The data were entered
into a binary logistic regression model to identify the
predictors of future LBI.
Results. A total of 292 athletes were used for the final
analysis (148 females and 144 males). During the fol-
low-up period, 31 (11%) athletes sustained an LBI. The
regression model, consisting of history of LBI, body
weight, and the latency of muscles shutting off during
flexion and lateral bending load releases, predicted cor-
rectly 74% of LBI outcomes. The odds of sustaining LBI
increased 2.8-fold when a history of LBI was present and
increased by 3% with each millisecond of abdominal
muscle shut-off latency. On average, this latency was 14
milliseconds longer for athletes who sustained LBI in
comparison to athletes who did not sustain LBI (77 [36]
vs. 63 [31]). There were no significant changes in any of
the muscle response latencies on retest following the
injury.
Conclusions. The delayed muscle reflex response sig-
nificantly increases the odds of sustaining an LBI. These
delayed latencies appear to be a preexisting risk factor
and not the effect of an LBI.
Key words: low back injury, motor control, muscle
response, risk factors. Spine 2005;30:2614 –2620
From the *Biomechanics Research Laboratory, Department of Ortho-
paedics and Rehabilitation, Yale University School of Medicine New
Haven, CT; and †Rehabilitation Sciences Research Laboratory, Drexel
University, Philadelphia, PA.
Acknowledgment date: October 18, 2004. First revision date: August
4, 2005. Acceptance date: August 19, 2005.
Supported by NIH Grant No. 5R01 AR46844 from the National In-
stitute of Arthritis and Musculoskeletal and Skin Diseases.
The manuscript submitted does not contain information about medical
device(s)/drug(s).
No funds were received in support of this work. No benefits in any
form have been or will be received from a commercial party related
directly or indirectly to the subject of this manuscript.
Address correspondence and reprint requests to Jacek Cholewicki,
PhD, Biomechanics Research Laboratory, Department of Orthopae-
dics and Rehabilitation, Yale University School of Medicine, P.O. Box
208071, New Haven, CT 06520-8071; E-mail: jacek.cholewicki@
yale.edu
2614
SPINE Volume 30, Number 23, pp 2614 –2620
©2005, Lippincott Williams & Wilkins, Inc.
Risk of developing a low back disorder is multifactorial
in nature.1,2 Many large prospective and case-control
studies have identified a multitude of risk factors that can
be parsed into three general categories: biomechanical,
psychosocial, and personal. For example, biomechanical
factors may include frequency of bending and twisting or
the magnitude of cumulative loads imposed on the spine
in the workplace.3–5 Job dissatisfaction and depression
are the most often cited psychosocial risk factors associ-
ated with low back pain (LBP).6 – 8 Finally, age, weight,
selected anthropometric measurements, smoking, radio-
graphic spine abnormalities, back endurance, or a his-
tory of LBP are all considered personal or individual
factors.9 –14 A history of previous LBP is the single best
predictor of future back pain for the general working
population1,6,15,16 as well as for collegiate athletes.17
However, when the risk factors from all three categories
are measured, they appear to interact in an additive man-
ner.1,18 Several other risk factors have been hypothe-
sized, but their causality is not yet verified scientifically.
Impaired motor control of the lumbar spine has been
proposed as one of the possible mechanisms, which
would predispose an individual to suffer a low back in-
jury (LBI).19,20 This hypothesis is supported with find-
ings that individuals with LBP exhibited longer trunk
muscle response latencies than healthy controls when
confronted with sudden loading.21–25 In addition, epide-
miologic studies have suggested that there is a link be-
tween LBI and sudden and unexpected movements, such
as slips and trips while handling loads.26 –29 On the other
hand, spine stability depends on appropriate muscle con-
trol, which determines the trunk kinematic response to
sudden loading30 and perhaps the subsequent likelihood
of injury.31 An individual may be more likely to sustain
an injury if his or her muscle response to sudden and
unexpected trunk loading is inadequate, which could
compromise spinal stability.
Radebold et al24 and Cholewicki et al25 measured re-
flex responses from 12 major trunk muscles during quick
force release experiments in subjects with chronic LBP
and in athletes with a history of an acute LBI. These were
short-latency reflexes most likely associated with muscle
spindle activity.32,33 Subjects with LBP had significantly
longer latencies in the offset of agonistic as well as in the
onset of antagonistic muscles in response to force release
in flexion, extension, and lateral bending directions. Fur-
thermore, when compared with healthy controls, LBP
patients demonstrated a significant reduction in the num-
ber of trunk muscles that responded to quick force re-
 Delayed Trunk Muscle Reflex Response • Cholewicki et al 2615
lease. Athletes with a history of a single LBI episode
exhibited similar motor control changes.
An alternative to the hypothesis of impaired motor
control as a risk factor is that the changes in motor con-
trol could be caused by the injury or pain itself. Damage
to the receptors imbedded within the soft tissues of the
lumbar spine could impair the feedback control and in
turn delay the reflex response. This notion is supported
by poor trunk repositioning sense34 –37 and poor pos-
tural control37– 42 found in patients with LBP in compar-
ison to healthy controls. Alternatively, patients with LBP
may adopt different motor control strategies to compen-
sate for an injured and unstable spine43 or to avoid pain.
Finally, the presence of pain stimuli itself may alter the
neuronal excitability and, in turn, trunk muscle activa-
tion patterns.44 – 46
The question of whether the altered motor control is a
risk factor or a result of LBP has profound clinical reper-
cussions. Diametrically different prevention, diagnosis,
and rehabilitation methods must be used if delayed trunk
muscle reflex response is a risk factor for LBI or alterna-
tively a result of LBI. Therefore, the purpose of the
present study was to answer the above question in a
prospective experimental manner. We hypothesized that
athletes who exhibited delayed muscle response latencies
at baseline would be predisposed to suffer LBI during the
follow-up period. Furthermore, if the above hypothesis
were true, there should be no change in latencies before
and after an LBI.
Methods
A sample of college athletes was used as an alternative to the
general population cohort. Similar changes in motor control
between injured athletes and a general population of patients
suffering from LBP24,25 suggest that the mechanisms underly-
ing the association between the impaired motor control and
LBP are similar for both populations. However, in contrast to
the general population, the athletes constitute a more homoge-
neous group in terms of their overall health, motivation, age,
and fitness level.
Athletes were tested at baseline and then followed for 2 to 3
years to track any LBIs sustained during that period. Injured
athletes were invited for a retest.
A total of 303 Yale varsity athletes from 22 different sports
volunteered for the study. A total of 299 of the athletes partic-
ipating in this study competed at the varsity level. The addi-
tional four athletes competed at the intercollegiate club level in
the following sports: rugby, badminton, martial arts, and
weight lifting. None of the athletes experienced acute pain dur-
ing the experimental testing, and all finished the outlined pro-
tocol to the best of their ability. Before experimental testing,
every subject filled out a detailed, 45-item questionnaire per-
taining to personal data (height, weight, and age), athletic ex-
perience, varsity level sport(s) affiliation, and for those with a
history of LBI, the number of episodes of LBI, date of last
episode, and the length of time out of competition/practice. The
level of pain while injured was recorded with a visual analog
pain scale,47 and the level of disability was assessed with a
Roland-Morris Disability Questionnaire.48 All subjects under-
Figure 1. A subject positioned in a multidirectional, quick force
release apparatus. Flexion (A), extension (B), and lateral bending
(C) loads were applied via a system of pulleys.
stood the experimental protocol and signed the consent form,
both approved by the Human Investigation Committee.
Muscle Response to a Quick Force Release. A quick force
release in three separate directions of isometric trunk exertions
was used for assessing the trunk muscle response to sudden
unloading. Subjects were placed in a specially built apparatus
in the semi-seated position, which was designed to permit iso-
metric contraction in trunk flexion, extension, and lateral
bending (Figure 1). The apparatus restrained pelvic motion,
leaving the upper body free to move in any direction. It ex-
cluded any postural adjustments through joints other than the
spine (i.e., hip, knee and ankle). The semi-seated position al-
lowed the subjects to assume their most comfortable lumbar
spine geometry before their pelvis was restrained. A cable at-
tached to a chest harness at approximately T9 was held with
an electromagnet and served as a resisting force for isometric
exertions.
Each subject performed five trials at a constant force level
corresponding to 30% of the maximal isometric trunk exertion
for an average healthy male (108 N) or female (72 N) estab-
lished empirically in our preliminary study. The force level was
displayed on an oscilloscope to provide visual feedback to the
athlete helping them reach the target force. The resisted force
was suddenly released with an electromagnet at random time
intervals after the target was reached.
EMG signals were recorded from 12 major trunk muscles
both pre and post release using bipolar, Ag-AgCl, surface, dis-
posable electrodes. The electrodes were placed with a center-
to-center spacing of 3 cm over the following muscles on each
side of the body: rectus abdominis (3 cm lateral to the umbili-
cus), external oblique (approximately 15 cm lateral to the um-
bilicus), internal oblique (approximately midway between the
anterior superior iliac spine and symphysis pubis, above the
inguinal ligament), latissimus dorsi (lateral to T9 over the mus-
cle belly), thoracic erector spinae (3 cm lateral to T9 spinous
process), and lumbar erector spinae (3 cm lateral to L4 spinous
process). In our previous studies, this electrode place-
ment proved to maximize signal-to-noise ratio with insignifi-
cant levels of cross talk.20,49 All EMG signals were band-pass
filtered between 20 and 420 Hz, differentially amplified (input
impedance ⫽ 100 GW, CMRR ⬎ 140 dB) and A/D converted
at a sample rate of 1,600 Hz.
Agonistic muscles were defined as muscles that were active
before the force release and were expected to shut off after the
release. Antagonistic muscles were inactive before the force
release and were expected to respond with increased electrical
activity after the release (Figure 2). In flexion, flexors acted as
agonists (FlexOFF) and extensors as antagonists (FlexON). In
extension, extensors acted as agonists (ExtOFF) and flexors as
2616 Spine • Volume 30 • Number 23 • 2005
Figure 2. Muscle response to sudden trunk unloading. An example
of EMG traces for a muscle shut off and switch on. The trunk force
was released at time zero.
antagonists (ExtON). In lateral bending to the left, the ipsilat-
eral muscles (left side) acted as agonists (LatOFF) and con-
tralateral muscles (right side) acted as antagonists (LatON).
The detection of onsets and offsets of muscle activity was
fully automated using a model-based algorithm developed by
Staude and Wolf50 and implemented in the Matlab software
environment (Math Works, Inc., Natick, MA). This algorithm
showed superior performance when compared with the tradi-
tional, amplitude threshold based methods.50 The onsets and
offsets were detected when significant changes occurred in the
statistical properties of the EMG signal modeled as a gaussian
random process. All detected events, marked with alarm times,
were ranked based on their maximum likelihood statistics. An
appropriate event (expected onset or offset) that reached a pre-
determined probability level was selected at the earliest alarm
time in the interval of 15 to 150 milliseconds following the
force release. The assumption was made that reflex responses
could not occur earlier than 15 milliseconds after the stimu-
lus45 and that responses longer than 150 milliseconds could be
voluntary and no longer reflexive.
To quantify the overall trunk muscles’ response to quick
force release, the average latency and the number of muscles
that responded to quick release were computed as in our pre-
vious studies.24,25 Only those muscles that responded to force
release were included in the average latency. In addition, the
censored average latency was computed, in which the muscles
that did not respond to force release were assigned the maxi-
mum latency of 150 milliseconds. This censored average com-
bined the latencies and the number of responding muscles into
one quantity. The following parameters were considered in this
procedure: 1) number of agonistic muscles that shut off; 2)
number of antagonistic muscles that switched on; 3) offset la-
tencies of agonistic muscle activities; 4) onset latencies of an-
tagonistic muscle activities; 5) censored offset latencies of ago-
nistic muscle activities; and 6) censored onset latencies
of antagonistic muscle activities. There were six muscles ex-
pected to switch on and six to shut off in the trunk flexion,
extension, and lateral bending trials.
Retesting Protocol. Athletes were instructed to notify a mem-
ber of the experimental team if they sustained an LBI during the
follow-up period. The injured athlete was then scheduled for a
retest at their convenience. The retest consisted of a retest ques-
tionnaire and motor control testing as described above. For the
purpose of this study, an injury was defined as any LBP that
caused an athlete to miss or not fully participate in at least 3
days of practice or competitions, and resulted in a visit to a
sports physician or athletic trainer. To help ensure a high cap-
ture rate for the injured athletes, we performed regularly sched-
uled electronic mailings to study participants inquiring if they
had had an LBI. In addition, we reviewed the training room and
team physician records for low back injuries among the partic-
ipating athletes. Several injured athletes underwent a second
and third retest to gain longitudinal data.
To verify the reproducibility of our measures, 15 athletes
were selected as controls for a retest. They did not have a
history of LBI and had not sustained an injury before their
retesting. The reproducibility was quantified with an intraclass
correlation coefficient using a (2, k) model according to Portney
and Watkins.51 The retests were conducted at an average of
6.85 months (SD, 1.75 months) after original testing.
Statistical Analysis. The quick force release test generated 18
parameters: 3 directions, each with muscle response latencies
for onsets and offsets, censored muscle response latencies for
onsets and offsets, along with number of muscles switching on
and shutting off (3 ⫻ (2 ⫹ 2 ⫹ 2) parameters). These parame-
ters were averaged across all trials for each athlete and served
as input into a backwards stepwise binary logistic regression
model (Minitab 12.23, State College, PA) for predicting LBI. A
parameter contributing the least to the model was removed at
each step until only the parameters with P ⬍ 0.1 remained. To
determine the initial most appropriate set of parameters for
regression analysis, a two-factor ANOVA was used to identify
the parameters that differed significantly between the injured
and uninjured athletes (P ⬍ 0.05). The two factors were as
follows: LBI vs. no LBI during the follow-up, and history of LBI
vs. no history of LBI at baseline as a possible confounding
variable. Some of the parameters were not normally distributed
(Anderson-Darling normality test, Minitab 12.23, State Col-
lege, PA). Therefore, these parameters underwent logarithmic
transformation to correct them to a normal distribution before
a two-factor ANOVA. The duration of the follow-up period
was added to the final regression model to test if it had any
effect on model predictions.
Results
Our results reflected 292 college athletes who could be reliably
classified as injured or not injured based on the availability of
records and our definition of injury. This group was made up of
148 females and 144 males. The athletes were on average 19.4
years of age, 1.77 m tall, and were 72.7 kg. There were 60
(20.5%) athletes that had a history of LBI lasting longer than 3
days and 232 (79.5%) athletes with no history of LBI. During
the follow-up period (2–3 years), there were a total of 31 (11%)
athletes with LBI, which fit our established criteria (Table 1).
The reproducibility of muscle response parameters used in this
study ranged from fair to very good (intraclass correlation co-
efficient ⫽ 0.37– 0.88).
The initial ANOVA returned significant main effects of LBI
for muscles shutting off in response to sudden load release in
trunk flexion and in lateral bending (Table 2). The mean (stan-
dard deviation) latencies were longer in athletes who sustained
an LBI during the follow-up period as compared with those
who did not (77 [36] vs. 63 [31] ms for FlexOFF and 63 [25] vs.
55 [20] ms for LatOFF). Furthermore, the LBI athletes shut off
a significantly smaller number of muscles in trunk flexion (0.8
 Delayed Trunk Muscle Reflex Response • Cholewicki et al 2617

Table 1. Demographic Data for Athletes Who Did (LBI) Table 3. Predictive Factors for LBI Included in the Final
and Did Not (No LBI) Sustain Low Back Injury During the Binary Regression Model
Follow-up Period
Factor Coefficient P Odds Ratio
No LBI LBI
Constant ⫺9.330000 0.000 —
Males: 106 Females: 107 Males: 11 Females: 8 History of LBI 1.045100 0.013 2.84
Body weight 0.033510 0.018 1.03
No history of LBI Age 19.3 (1.2) Age 19.4 (1.1) Censored latency (Flexion OFF) 0.026280 0.042 1.03
Height 1.76 (0.10) Height 1.80 (0.10) Latency (Lateral Bending OFF) 0.015229 0.067 1.02
Weight 71.7 (12.2) Weight 79.2 (14.5)
Males: 23 Females: 25 Males: 4 Females: 8
associated with each parameter, an athlete with a history of LBI
History of LBI Age 19.7 (2.4) Age 19.6 (0.9)
Height 1.76 (0.10) Height 1.80 (0.07)
was 2.8 times more likely to suffer an LBI than an athlete
Weight 73.3 (13.2) Weight 78.4 (17.1) without a history of LBI (Table 3). Moreover, an athlete’s odds
of LBI increased by 3% for every millisecond delay in censored
Values are mean (SD).
muscle response latency in FlexOFF and by 2% for every mil-
lisecond delay in muscle response latency in LatOFF. Odds of
sustaining an LBI also increased by 3% for each kilogram in-
[1.0] vs. 1.3 [1.1]) making the censored latency also signifi- crease in body weight (Table 3). The duration of the follow-up
cantly longer (138 [16] vs. 129 [20] ms) (Table 2). period did not have any effect on model predictions (P ⫽ 0.34).
The only effect of a history of LBI was in the number of The regression analysis was repeated with stricter defini-
muscles switching on in trunk extension trials (ExtON) (P ⫽ tions of injury. When the definition of an injured athlete was
0.03). Athletes with no history of LBI responded with a greater limited to only those who were referred to a sports medicine
number of muscles than athletes with a history of LBI (5.6 [0.5] physician, or those who were further directed for MRI or other
vs. 5.5 [0.6]). There was no interaction between a history of imaging tests, the percentage of predicted probabilities concor-
LBI at baseline and LBI during the follow-up in any of the dant with the observed outcomes increased to 81% and 87%,
measured parameters. respectively.
The final binary logistic regression model consisted of ath- The regression analysis was also repeated separately for ath-
letes’ history of LBI at baseline, their body weight, and letes with and without a history of LBI to examine whether the
FlexOFF and LatOFF response latencies (Table 3). The model model would perform equally well for both groups of subjects.
predicted 74% of LBI probabilities concordant with the ob- The number of concordant observations was 71% and 72%,
served outcomes during the follow-up. Based on the odds ratios respectively.
Analysis of the retest data showed that there were no signif-
icant differences in the pre and post injury test values for muscle
Table 2. Main Effects of Low Back Injury (LBI) response times in athletes who had no history of LBI and who
Sustained During the Follow-up Period on Trunk Muscle sustained an LBI (Table 4). There were also no significant
Responses to Sudden Load Release Measured During the changes in motor control over time following an LBI (Figure 3).
Initial Test
Discussion
Confidence
Parameter No LBI LBI P Intervals There is no doubt that risk factors predisposing an indi-
vidual to develop LBP are multifactorial and their inter-
Flexion, agonists OFF
Mean latency (ms)* 63 (31) 77 (36) 0.03 2/32 action is extremely complex. At the same time, the
No. of muscles responding* 1.3 (1.1) 0.8 (1.0) 0.03 ⫺0.9/⫺0.1 knowledge of these risk factors is crucial for designing
Censored latency (ms)* 129 (20) 138 (16) 0.01 2/17 more effective prevention, diagnosis, and rehabilitation
Extension, agonists OFF
Mean latency (ms) 48 (12) 51 (17) 0.13 ⫺1/9 strategies for LBP. The present study identified yet an-
No. of muscles responding 2.7 (0.9) 2.6 (0.9) 0.64 ⫺0.5/0.3 other risk factor that was not previously confirmed with
Censored latency (ms) 103 (18) 105 (17) 0.47 ⫺0/10 a prospective experimental design. The results from a
Lateral bending, agonists OFF
Mean latency (ms)* 55 (20) 63 (25) 0.04 0/16 binary logistic regression analysis supported the hypoth-
No. of muscles responding 1.5 (0.8) 1.4 (0.8) 0.50 ⫺0.4/0.2 esis that the delayed muscle response to sudden trunk
Censored latency (ms) 125 (15) 127 (16) 0.35 ⫺3/9 loading is a significant predictor (risk factor) of a future
Flexion, antagonists ON
Mean latency (ms) 59 (9) 58 (10) 0.26 ⫺5/1 LBI. This conclusion was further supported by the lack
No. of muscles responding 5.4 (0.5) 5.4 (0.5) 0.31 ⫺0.3/1.0 of any significant change in muscle reflex latencies fol-
Censored latency (ms) 68 (12) 67 (12) 0.92 ⫺5/4 lowing an LBI among athletes who reported no history of
Extension antagonists ON
Mean latency (ms) 58 (8) 58 (8) 0.99 ⫺3/3 injury (Table 4).
No. of muscles responding 5.6 (0.5) 5.6 (0.5) 0.88 ⫺0.2/0.2 In addition to the muscle reflex response latencies,
Censored latency (ms) 64 (11) 64 (11) 0.79 ⫺4/5 history of LBI and body weight were also identified as
Lateral bending, antagonists ON
Mean latency (ms) 65 (8) 66 (8) 0.66 ⫺2/4 significant predictors of future LBI. In this context, im-
No. of muscles responding 5.2 (0.6) 5.3 (0.5) 0.43 ⫺0.1/0.3 paired motor control can be interpreted as follows. Be-
Censored latency (ms) 76 (12) 76 (10) 0.81 ⫺5/4 cause of the multifactorial nature of LBP, the single best
Values are mean (SD). predictor of the future LBP episodes is a history of
*Significant effects (P ⬍ 0.05).
LBP.1,6,15,17 Simply, individuals who had LBP are likely
2618 Spine • Volume 30 • Number 23 • 2005
Table 4. Pre- and Post-Injury Reflex Latencies for Athletes Without History of LBI (N ⴝ 16) and Test/Retest Latencies
for Control Athletes Who Had No History of LBI and Who Did Not Sustain LBI (N ⴝ 15)
Extension Extension
LBI Test OFF ON
No history of LBI Initial test (ms) 52 (18) 58 (5)
Retest (ms) 49 (16) 62 (8)
Controls Initial test (ms) 49 (17) 57 (7)
Retest (ms) 53 (14) 59 (8)
Values are mean (SD). No significant differences were found in any of the quick force release directions.
to possess or to be exposed to a combination of many
risk factors. These risk factors will predispose them to
future LBP. Therefore, it is also likely that studies com-
paring patients with history of LBP to healthy controls
will find differences in the measures that quantify some of
these risk factors. For example, in previous studies, de-
layed muscle reflex response to sudden trunk loading
was found in patients with chronic LBP21–24 and in ath-
letes with a history of LBI.25 Probably, most of these
individuals already possessed impaired motor control,
which predisposed them to LBP in the first place. How-
ever, not all people with LBP have impaired motor con-
trol and not all individuals with impaired motor control
will suffer LBP. The number of all risk factors and their
combination determine the likelihood of LBP, and this is
reflected in the single variable of “history of LBP.” In our
analyses, the impaired motor control parameters and
body weight percolated to the forefront as independent
predictors of LBP from all other risk factors encom-
passed by history of LBP. The delayed muscle reflex la-
tencies were found to be related to a future LBI and not
to a history of LBP (Table 2).
The present results are consistent with several previ-
ous findings. Similar to Cholewicki et al,25 we found that
the largest difference between athletes with a history of
LBI and controls was in the offset latency of the abdom-
inal muscles (FlexOFF). Our FlexOFF response latencies
(88 [38] ms for injured and 64 [31] ms for noninjured
Figure 3. The latency of trunk muscles shutting off (mean ⫾ SD) in
response to quick force release in trunk flexion (FlexOFF) as a
function of test interval after the injury. There was no significant
effect of time following the injury (P ⬎ 0.05).
Flexion Flexion Lateral Bending Lateral Bending
OFF ON OFF ON
69 (32) 56 (9) 64 (22) 66 (9)
60 (21) 59 (9) 51 (13) 67 (9)
66 (26) 58 (6) 58 (16) 66 (7)
61 (23) 58 (5) 62 (23) 61 (7)
athletes) matched closely with those found in the
Cholewicki et al study25 (90 [37] ms for injured and
59 [29] ms for noninjured athletes). These results are
very encouraging for several reasons. First, the algorithm
for determining onsets and offsets was different yet the
flexion offset delays were the largest in the injured ath-
letes in both studies. Second, the largest differences were
seen in flexion offset latencies, which match previous
studies with chronic LBP patients.24,42 Finally, the cur-
rent study supported the findings from an earlier pro-
spective study of 679 Yale athletes, in which a history of
LBI was the greatest predictor of future LBI.17
The biggest weakness in our study was the identifica-
tion of athletes who had LBI during the follow-up pe-
riod. Although our classification of injury was well de-
fined, it relied on either the athletes to report injury or on
the accuracy and completeness of the training room and
team physician records to flag all LBI. We, moreover, did
not physically assess any of the athletes in the postinjury
phase or pursue an objective diagnosis. However, we
repeated the analysis with stricter criteria in the defini-
tion of an injury and obtained the same outcome. There-
fore, the results of our study are robust in the context of
the definition of injury and are likely to hold if a physical
examination was performed to classify LBI. This study
was limited to assessment of motor control parameters
as potential risk factors and did not evaluate any psycho-
social factors related to LBI in athletes. However, our
previous study showed that there was no significant cor-
relation between these factors and risk of suffering an
LBI.17 The collegiate athletes constitute a very homoge-
neous group, and psychosocial factors may not be as
important in this cohort as in general population with
LBP.1 Although specific results obtained from testing
athletes may not be directly extrapolated to the working
population, the findings of the delayed muscle reflex re-
sponse being a risk factor for LBI could be.
Our results indicated that delayed FlexOFF and Lat-
OFF latencies are a significant predictor of a future LBI in
athletes, irrespective of their history of LBI. These results
are important for two reasons. First, our understanding
of the mechanisms of LBI is enhanced in favor of the
hypothesis that delayed muscle reflex response increases
vulnerability of the spine to injury under sudden loading
conditions. Therefore, clinicians should attempt to mod-
ify such muscle recruitment patterns in patients with
 Delayed Trunk Muscle Reflex Response • Cholewicki et al 2619
LBP. Second, our results could be used in future research
to develop a tool for screening athletes at risk of LBI.
With this information, athletes could be properly coun-
seled as to their risk of sustaining an LBI.
Future research should focus on the ability to improve
muscle response to sudden loading. This has been shown
to be possible in back muscles22 and in hamstring mus-
cles in anterior cruciate ligament deficient knees.52 Also,
further investigations should look for the reasons behind
delayed muscle responses. This could perhaps involve the
quantity and quality of the neuromuscular receptors
around the spine and their ability to modify muscle re-
sponse to stimuli. Finally, an investigation into whether
improvements in muscle response latencies actually re-
sult in a decreased risk of injury is warranted.
Key Points
● A prospective study with a 2- to 3-year follow-up
was conducted to determine whether delayed mus-
cle reflex response to sudden trunk loading is a
result of or a risk factor for sustaining a low back
injury (LBI).
● Muscle reflex latencies in response to a quick
force release in trunk flexion, extension, and lateral
bending were measured in 303 college athletes.
● The regression model consisting of history of
LBI, body weight, and latency of muscles shutting
off during flexion and lateral bending load releases
predicted correctly 74% of LBI outcomes during
the follow-up.
● The odds of sustaining an LBI increased 2.8-fold
when a history of LBI was present and increased by
3% with each millisecond of abdominal muscle
shut-off latency.
● The delayed muscle reflex response latencies ap-
pear to be a significant preexisting risk factor and
not the effect of an LBI.
References
1. Kerr MS, Frank JW, Shannon HS, et al. Biomechanical and psychosocial risk
factors for low back pain at work. Am J Public Health 2001;91:1069 –75.
2. Stevenson JM, Weber CL, Smith JT, et al. A longitudinal study of the devel-
opment of low back pain in an industrial population. Spine 2001;26:
1370 –7.
3. Marras WS, Allread WG, Burr DL, et al. Prospective validation of a low-back
disorder risk model and assessment of ergonomic interventions associated
with manual materials handling tasks. Ergonomics 2000;43:1866 – 86.
4. Hartvigsen J, Bakketeig LS, Leboeuf-Yde C, et al. The association between
physical workload and low back pain clouded by the ‘healthy worker’ effect:
population-based cross-sectional and 5-year prospective questionnaire
study. Spine 2001;26:1788 –92; discussion 92–3.
5. Elders LA, Heinrich J, Burdorf A. Risk factors for sickness absence because
of low back pain among scaffolders: a 3-year follow-up study. Spine 2003;
28:1340 – 6.
6. Bigos SJ, Battie MC, Spengler DM, et al. A longitudinal, prospective study of
industrial back injury reporting. Clin Orthop 1992;279:21–34.
7. Hoogendoorn WE, Bongers PM, de Vet HC, et al. Psychosocial work char-
acteristics and psychological strain in relation to low-back pain. Scand J
Work Environ Health 2001;27:258 – 67.
8. Carroll LJ, Cassidy JD, Cote P. Depression as a risk factor for onset of an
episode of troublesome neck and low back pain. Pain 2004;107:134 –9.
9. Biering-Sorensen F. Physical measurements as risk indicators for low-back
trouble over a one-year period. Spine 1984;9:106 –19.
10. Battie MC, Bigos SJ, Fisher LD, et al. A prospective study of the role of
cardiovascular risk factors and fitness in industrial back pain complaints.
Spine 1989;14:141–7.
11. Battie MC, Bigos SJ, Fisher LD, et al. Anthropometric and clinical measures
as predictors of back pain complaints in industry: a prospective study. J Spi-
nal Disord 1990;3:195–204.
12. Feldman DE, Rossignol M, Shrier I, et al. Smoking: a risk factor for devel-
opment of low back pain in adolescents. Spine 1999;24:2492– 6.
13. Takala EP, Viikari-Juntura E. Do functional tests predict low back pain?
Spine 2000;25:2126 –32.
14. Iwamoto J, Abe H, Tsukimura Y, et al. Relationship between radiographic
abnormalities of lumbar spine and incidence of low back pain in high school
and college football players: a prospective study. Am J Sports Med 2004;32:
781– 6.
15. Bigos SJ, Battie MC, Spengler DM, et al. A prospective study of work per-
ceptions and psychosocial factors affecting the report of back injury. Spine
1991;16:1– 6 [published erratum appears in Spine 1991;16:688].
16. Mannion AF, Dolan P, Adams MA. Psychological questionnaires: do ‘ab-
normal’ scores precede or follow first-time low back pain? Spine 1996;21:
2603–11.
17. Greene HS, Cholewicki J, Galloway MT, et al. A history of low back injury
is a risk factor for recurrent back injuries in varsity athletes. Am J Sports Med
2001;29:795– 800.
18. Thorbjornsson CO, Alfredsson L, Fredriksson K, et al. Psychosocial and
physical risk factors associated with low back pain: a 24 year follow up
among women and men in a broad range of occupations. Occup Environ
Med 1998;55:84 –90.
19. Panjabi MM. The stabilizing system of the spine: I. Function, dysfunction,
adaptation, and enhancement. J Spinal Disord 1992;5:383–9.
20. Cholewicki J, McGill SM. Mechanical stability of the in vivo lumbar spine:
implications for injury and chronic low back pain. Clin Biomech 1996;11:
1–15.
21. Hodges PW, Richardson CA. Inefficient muscular stabilization of the lumbar
spine associated with low back pain: a motor control evaluation of transver-
sus abdominis. Spine 1996;21:2640 –50.
22. Magnusson ML, Aleksiev A, Wilder DG, et al. Unexpected load and asym-
metric posture as etiologic factors in low back pain. Eur Spine J 1996;5:
23–35.
23. Hodges PW, Richardson CA. Delayed postural contraction of transversus
abdominis in low back pain associated with movement of the lower limb.
J Spinal Disord 1998;11:46 –56.
24. Radebold A, Cholewicki J, Panjabi MM, et al. Muscle response pattern to
sudden trunk loading in healthy individuals and in patients with chronic low
back pain. Spine 2000;25:947–54.
25. Cholewicki J, Polzhofer GK, Galloway MT, et al. Neuromuscular function in
athletes following recovery from an acute low back injury. J Orthop Sports
Phys Ther 2002;32:569 –76.
26. Magora A. Investigation of the relation between low back pain and occupa-
tion: IV. Physical requirements: bending, rotation, reaching and sudden
maximal effort. Scand J Rehabil Med 1973;5:186 –90.
27. Manning DP, Shannon HS. Slipping accidents causing low-back pain in a
gearbox factory. Spine 1981;6:70 –2.
28. Manning DP, Mitchell RG, Blanchfield LP. Body movements and events
contributing to accidental and nonaccidental back injuries. Spine 1984;9:
734 –9.
29. Omino K, Hayashi Y. Preparation of dynamic posture and occurrence of low
back pain. Ergonomics 1992;35:693–707.
30. Cholewicki J, Simons APD, Radebold A. Effects of external trunk loads on
lumbar spine stability. J Biomech 2000;33:1377– 85.
31. McGill SM, Grenier S, Kavcic N, et al. Coordination of muscle activity to
assure stability of the lumbar spine. J Electromyogr Kinesiol 2003;13:353–9.
32. Matthews PB. Evolving views on the internal operation and functional role
of the muscle spindle. J Physiol 1981;320:1–30.
33. Angel RW, Weinrich M. Stretch and unloading reflexes in a human hand
muscle. Exp Neurol 1986;94:348 –58.
34. Parkhurst TM, Burnett CN. Injury and proprioception in the lower back.
J Orthop Sports Phys Ther 1994;19:282–95.
35. Gill KP, Callaghan MJ. The measurement of lumbar proprioception in indi-
viduals with and without low back pain. Spine 1998;23:371–7.
36. Taimela S, Kankaanpaa M, Luoto S. The effect of lumbar fatigue on the
ability to sense a change in lumbar position: a controlled study. Spine 1999;
24:1322–7.
37. Leinonen V, Kankaanpaa M, Luukkonen M, et al. Lumbar paraspinal mus-
cle function, perception of lumbar position, and postural control in disc
herniation-related back pain. Spine 2003;28:842– 8.
2620 Spine • Volume 30 • Number 23 • 2005
38. Byl NN, Sinnott PL. Variations in balance and body sway in middle-aged
adults: subjects with healthy backs compared with subjects with low-back
dysfunction. Spine 1991;16:325–30.
39. Takala E-P, Korhonen I, Viikari-Juntura E. Postural sway and stepping response
among working population: reproducibility, long-term stability, and associa-
tions with symptoms of the low back. Clin Biomech 1997;12:429 –37.
40. Luoto S, Aalto H, Taimela S, et al. One-footed and externally disturbed
two-footed postural control in patients with chronic low back pain and
healthy control subjects: a controlled study with follow-up. Spine 1998;23:
2081–9.
41. Mientjes MI, Frank JS. Balance in chronic low back pain patients compared
with healthy people under various conditions in upright standing. Clin Bio-
mech 1999;14:710 – 6.
42. Radebold A, Cholewicki J, Polzhofer GK, et al. Impaired postural control of
the lumbar spine is associated with delayed muscle response times in patients
with chronic idiopathic low back pain. Spine 2001;26:724 –30.
43. van Dieën JH, Cholewicki J, Radebold A. Trunk muscle recruitment patterns
in patients with low back pain enhance the stability of the lumbar spine.
Spine 2003;28:834 – 41.
44. Moe-Nilssen R, Ljunggren AE, Torebjork E. Dynamic adjustments of walk-
ing behavior dependent on noxious input in experimental low back pain.
Pain 1999;83:477– 85.
View publication stats
45. Zedka M, Prochazka A, Knight B, et al. Voluntary and reflex control of
human back muscles during induced pain. J Physiol 1999;520:591–
604.
46. Hodges PW, Moseley GL, Gabrielsson A, et al. Experimental muscle pain
changes feed forward postural responses of the trunk muscles. Exp Brain Res
2003;151:262–71.
47. Bijur PE, Silver W, Gallagher EJ. Reliability of the visual analog scale for
measurement of acute pain. Acad Emerg Med 2001;8:1153–7.
48. Roland M, Morris R. A study of the natural history of back pain: I. Devel-
opment of a reliable and sensitive measure of disability in low-back pain.
Spine 1983;8:141– 4.
49. Cholewicki J, Panjabi MM, Khachatryan A. Stabilizing function of trunk
flexor-extensor muscles around a neutral spine posture. Spine 1997;22:
2207–12.
50. Staude G, Wolf W. Objective motor response onset detection in surface
myoelectric signals. Med Eng Phys 1999;21:449 – 67.
51. Portney LG, Watkins MP. Foundations of Clinical Research: Applications to
Practice. Norwalk, CT: Appleton & Lange, 1993.
52. Wojtys EM, Huston LJ, Taylor PD, et al. Neuromuscular adaptations in
isokinetic, isotonic, and agility training programs. Am J Sports Med 1996;
24:187–92.