Content uploaded by Linda Mclean
Author content
All content in this area was uploaded by Linda Mclean on Jan 08, 2020
Content may be subject to copyright.
The role of lumbopelvic posture in pelvic floor muscle activation in
continent women
Angela Christine Capson, Joseph Nashed, Linda Mclean
⇑
Queen’s University, School of Rehabilitation Therapy, 32 George Street, Kingston, ON, Canada K7L 3N6
article info
Article history:
Received 12 April 2010
Received in revised form 21 June 2010
Accepted 30 July 2010
Keywords:
Pelvic floor muscles
Electromyography
Lumbopelvic posture
Postural control
Stress urinary incontinence
Low back pain
Lumbopelvic pain
Spinal stability
abstract
This study was undertaken to determine the effect of changing standing lumbopelvic posture on pelvic
floor muscle (PFM) activation amplitude and timing and the resultant vaginal manometry values
recorded during static and dynamic tasks. Sixteen nulliparous, continent women between the ages of
22 and 41 years performed five tasks (quiet standing, maximal effort cough, Valsalva manoeuvre, maxi-
mum voluntary contraction (MVC) of the PFMs, and a load-catching task) in three different standing pos-
tures (normal lumbopelvic posture, hyperlordosis and hypolordosis). Electromyographic (EMG) data
were recorded from the PFMs bilaterally using a Periform™ vaginal probe coupled to Delsys™ Bagnoli-
8 EMG amplifiers. In separate trials, vaginal manometry was obtained using a Peritron™ perineometer.
Lumbopelvic angle was recorded simultaneously with EMG and vaginal manometry using an Optotrak™
3D motion analysis system to ensure that subjects maintained the required posture throughout the three
trials of each task. All data were filtered using a moving 100 ms RMS window and peak values were deter-
mined for each trial and task. Repeated-measures analyses of variance were performed on the peak PFM
EMG, intra-vaginal pressure amplitudes, and lumbopelvic angles as well as activation onset data for the
cough and load-catching tasks.
There was significantly higher resting PFM activity in all postures in standing as compared to supine,
and in the standing position, there was higher resting PFM activity in the hypo-lordotic posture as com-
pared to the normal and hyperlordotic postures. During the MVC, cough, Valsalva, and load-catching
tasks, subjects generated significantly more PFM EMG activity when in their habitual posture than when
in hyper- or hypo-lordotic postures. Conversely, higher peak vaginal manometry values were generated
in the hypo-lordotic posture for all tasks in all cases. These results clearly indicate that changes in lum-
bopelvic posture influence both the contractility of the PFMs and the amount of vaginal pressure gener-
ated during static postures and during dynamic tasks. Lumbopelvic posture does not, however, appear to
have a significant effect on the timing of PFM activation during coughing or load-catching tasks.
Crown Copyright Ó2010 Published by Elsevier Ltd. All rights reserved.
1. Introduction
The primary roles of the pelvic floor muscles (PFMs) include the
maintenance of continence (Raizada and Mittal, 2008), the support
of the abdominal contents (Raizada and Mittal, 2008), and sexual
functioning (Dean et al., 2008) and have been well established in
the literature. Recent studies have examined additional roles of
the PFMs including assisting in ventilation (Hodges et al., 2007),
and contributing to spinal stability (Smith et al., 2007a,b) and pos-
tural control (Hodges et al., 2007). In women, there is an epidemi-
ological link between lumbopelvic dysfunction and urinary
incontinence (Smith et al., 2006), and between chronic pelvic pain
and urinary incontinence (Zondervan and Kennedy, 2005; Chaitow,
2007). Lumbopelvic posture may influence the activation of the
PFMs and their co-ordination with the trunk muscles and may be
an important factor when treating women with lumbopelvic pain
or incontinence.
It has also been well established that the PFMs work synergisti-
cally with the abdominal muscles (Madill and McLean, 2008;
Madill et al., 2009; Neumann and Gill, 2002; Sapsford et al.,
2001; Sapsford and Hodges, 2001; Bo and Stien, 1994), and physi-
cal therapists often incorporate postural correction training into
programs aimed at improving both continence function and pos-
tural control; however, the effect of postural changes on pelvic
floor and abdominal muscle function has not been studied. It is
important to establish whether postural changes induce
differences in the ability of the pelvic floor or abdominal muscles
to contract in response to postural or other perturbations.
In women with stress urinary incontinence, the dual demands
(postural control and maintaining continence) placed on the PFMs
may increase the risk of urinary incontinence particularly during
1050-6411/$ - see front matter Crown Copyright Ó2010 Published by Elsevier Ltd. All rights reserved.
doi:10.1016/j.jelekin.2010.07.017
⇑
Corresponding author. Tel.: +1 613 533 6101; fax: +1 613 533 6776.
E-mail address: mcleanl@queensu.ca (L. Mclean).
Journal of Electromyography and Kinesiology 21 (2011) 166–177
Contents lists available at ScienceDirect
Journal of Electromyography and Kinesiology
journal homepage: www.elsevier.com/locate/jelekin
tasks in which postural stability is challenged. For example, de-
layed or insufficient PFM activation during postural perturbations
may result in ineffective urethral closure during tasks that chal-
lenge spinal stability (Smith et al., 2007a). Hodges et al. (2007)
have shown that in continent women, the PFMs are activated
slightly before the prime mover as part of anticipatory postural
control, but that the activation of the PFMs is delayed in women
with urinary incontinence.
Lumbopelvic posture might influence the ability of the pelvic
floor muscles to contract effectively. Several studies have exam-
ined the effects of changing body position (i.e., from supine to
standing) on the tonic activity of the PFMs (Santiesteban, 1988;
Bo and Finckenhagen, 2003; Kelly et al., 2007). Sapsford et al.
(2006) showed that in the sitting position, greater PFM activity
was recorded during voluntary PFM contractions performed in an
upright unsupported posture as compared to in a slumped sup-
ported posture. Chen et al. (2005) examined the effect of changing
standing lumbopelvic posture on PFM activation, however they
used changes in degree of ankle dorsiflexion and plantarflexion
to induce changes in lumbopelvic posture without reporting the
degree to which this task actually changed lumbopelvic posture.
Uncovering the influence of lumbopelvic posture on resting, volun-
tary and reflex activation of the PFMs may help explain any poten-
tial role for postural correction in patients with SUI with or without
lumbopelvic pain.
The main purpose of this investigation was to determine
whether changes in standing lumbopelvic posture affect the acti-
vation amplitude or timing of PFM contractions or the amplitude
of intra-vaginal pressure measured using a manometer when nul-
liparous continent women stood quietly, while they performed
maximum voluntary PFM contractions, and while they performed
tasks that challenged either postural control or continence. The
secondary purpose was to determine the effect of changes in lum-
bopelvic posture on trunk muscle activation amplitudes and tim-
ing during these same tasks.
2. Methods
2.1. Research design and subjects
This study was conducted with ethics approval from the
Queen’s University Health Sciences and Affiliated Teaching Hospi-
tals Research Ethics Board (REH-439-08). A single sample, cross-
sectional repeated-measures design was used.
A convenience sample of sixteen women between the ages of 22
and 41 were recruited by word of mouth from the Queen’s Univer-
sity and Kingston, Ontario communities. Volunteers were excluded
from participation if they were parous, reported an incident of
involuntarily leaking urine in the past month, or had experienced
low back or pelvic pain in the past year. After providing informed
consent, participants provided demographic information including
age, height, weight, and physical activity level. Women who agreed
to participate attended a single 2-h testing session. Subjects were
briefed on the equipment to be used and were asked to void their
bladder before beginning the experiment in order to standardize
bladder volume. The participant was then provided with standard-
ized verbal instructions on how to place a Periform™ vaginal probe
in their vagina and was left in a private room to do so. The correct
placement of the probe was verified by the experimenters through
visual inspection, as was the cranial movement of the probe during
voluntary contraction and coughing.
2.2. Posture analysis
Skin markers compatible with the Optotrak™ 3200 3D motion
analysis system were used to measure lumbopelvic, hip, and knee
angles to determine subject’s posture during the selected tasks.
The markers were placed based on the methods described by
Grimstone and Hodges (2003). Single markers were adhered to
the skin overlying the spinous processes of L
5
and L
2
, the greater
trochanter, the lateral femoral epicondyle and the lateral malleo-
lus. All lower extremity markers were fixed to the anatomic land-
marks on the right side of the body using standard medical tape. A
lightweight dowel 1 m long was affixed to the body 2 cm inferior to
the top of the right iliac crest, and a marker was placed at each end
to accentuate changes in lumbopelvic posture. Motion marker data
were sampled at a rate of 100 Hz using NDI ToolBench software
and were temporally aligned with electromyographic (EMG)
recordings. See Fig. 1 for details of the experimental set-up.
2.3. Electromyography
A Delsys Bagnoli-8 EMG amplification system (gain 1000, band-
pass filter 20–450 Hz, CMRR > 100 db at 60 Hz, input imped-
ance > 100 M
X
) was used to acquire EMG and pressure data.
DE2.1 bipolar stainless steel electrodes were placed on the rectus
abdominus (RA; 4 cm lateral to the umbilicus), external obliques
(EO; just inferior to the 8th rib, mid-clavicular on a 45°angle),
internal obliques (IO; 2 cm medial and superior to the anterior
superior iliac crest), and erector spinae (ES; 5 cm from midline at
the level of L
3
) on the right side in accordance with Ng et al.
(1998), and used in a recent study by Smith et al. (2007b). A com-
mon reference electrode was placed on the right anterior superior
iliac spine (ASIS). Prior to electrode application, the skin was pre-
pared using a 70% ethanol solution. To record EMG signals from
the PFMs, a modified Periform™ vaginal probe was used. This
pear-shaped probe has stainless steel electrodes embedded on
Fig. 1. Experimenatal set-up illustrated during the load-catching task.
A.C. Capson et al. / Journal of Electromyography and Kinesiology 21 (2011) 166–177 167
the right and left sides and has been reported to record PFM activ-
ity with minimal cross-talk during tasks such as those used in the
current study (Smith et al., 2007a; Hodges et al., 2007). Unlike in
previous studies, however, the probe was wired such that monopo-
lar EMG activity was recorded from each side of the PFMs sepa-
rately. As such, separate reference electrodes were placed on
each ASIS, one for each side of the PFMs. EMG activity was sampled
at a rate of 1000 Hz and synchronized with the Optorak™ data
using NDI ToolBench™ software.
2.4. Testing procedure
First each participant was asked to lie supine on a plinth and to
completely relax her abdominal and PFMs while EMG data were
recorded from all channels. The participant was then asked to
move to a standing position and to perform five different tasks in
three different postures. The tasks included quiet standing, three
trials of a maximum voluntary contraction (MVC) of the PFMs,
three single-barrel maximum expiratory effort coughs, three Val-
salva manoeuvres, and 10 trials of a load-catching task. The last
task required the participant to catch a 1 kg load dropped from a
height of 30 cm into a milk crate held 5 cm in front of the body
with the elbows flexed at 90°. Five trials of this task were com-
pleted with the participant’s eyes open, to observe responses when
the perturbation was anticipated, and five trials were completed
with the participant’s eyes closed to observe the responses when
the timing of the perturbation was unknown. Each of these tasks
was performed while the participant stood in her habitual standing
posture, when she held an exaggerated lumbar lordosis (hyperlor-
dotic lumbopelvic posture) and when she held a flattened lumbar
lordosis (hypo-lordotic lumbopelvic posture). The Optotrak™ 3D
motion analysis system was used to monitor posture throughout
all tasks while EMG data were recorded using the Delsys™ EMG
amplifiers. Individuals were provided with standardized verbal
instruction, demonstration, and physical cuing to help them as-
sume and maintain the hypo-lordotic and hyperlordotic postures.
Body positioning was standardized for each participant by keeping
bare feet shoulder-width apart and the knees just slightly out of
hyperextension. Following the completion of the tasks, the entire
procedure was repeated using the Peritron™ perineometer in place
of the Periform™ vaginal EMG probe. To do so, the participant was
asked to return to the private room to remove the Periform™ probe
and then to insert the Peritron™ perineometer into her vagina.
Again the participant received standardized instructions on how
to place the probe. The Peritron™ perineometer was used to mea-
sure intra-vaginal pressure during each of the tasks and therefore
the experiment was repeated with the Peritron™ probe in situ,
with vaginal pressure data acquired at 1000 Hz and synchronized
with the Optotrak™ data.
2.5. Data analysis
All raw EMG and pressure data were processed using Matlab™
v.7.7.0. Data were first plotted, and visually inspected on a com-
puter monitor to ensure that they had a signal to noise ratio
>10 dB and no motion artefact. Files with excessive noise or motion
artefact were not included in the analysis, which resulted in data
from four subjects being removed from the Valsalva data set (see
below). A root-mean-square (RMS) filter using a moving window
of 100 ms in length with 99 ms overlap was applied to all raw data.
The average baseline activity from each muscle and in each posture
was computed from the resting trial for that muscle and posture
and was subtracted from the peak amplitudes recorded from that
muscle and posture during each task in order to remove subject
and instrumentation noise. This resulted in the removal of tonic
activity from the overall amplitude data reported and as such the
amplitude values reported in this paper refer to the increase in
activity induced by the different dynamic tasks and not the overall
activity.
The onset timing and the timing of peak muscle activation were
determined from the smoothed EMG data for the cough task and
for the load-catching tasks. The onset times were automatically
determined using a computer algorithm that marked the time at
which the EMG signal rose more than two standard deviations
above the mean level of baseline activity recorded before the task
began, and were verified using visual inspection and corrected if
necessary. For the cough task, the onset and peak activation times
for the right and left PFMs were averaged and the onset and peak
activation times of the other muscles were determined relative to
the onset and peak activation time of the PFMs. For the load-catch-
ing task, when the load landed in the box it activated a switch that
was used as the reference point for muscle activation timing. As
such, the onset times and the timing of peak activation for all mus-
cles were determined relative to the time the load first hit the bot-
tom of the box.
Optotrak™ motion marker data were also processed using Mat-
lab™ v.7.7.0. The lumbopelvic angle was calculated by creating
vectors using the L
5
and L
2
markers and the front and rear dowel
markers. The hip angle was calculated in a similar fashion using
the greater trochanter and lateral epicondyle markers, relative to
the positioning of the dowel. Knee angle was calculated as the an-
gle between the vectors created by the greater trochanter and lat-
eral epicondyle markers, and the lateral maleolus and lateral
epicondyle markers, respectively. Each of these three angles was
calculated at the beginning and end of the trial and in the load-
catching tasks, at the time the load contacted the box. Trials in
which the angles at the start, end and during the load catch lay
outside the 95% confidence interval for the angle computed across
the trial would have indicated that the participant did not hold the
posture throughout the task and that trial therefore would be ex-
cluded from the analysis; however this was not found in any of
the data sets. Position data that were missing due to a blocked
marker were not included in the calculation of the average angles
throughout the task. Based on this methodology, the kinematic
data were standardized within subjects however, based on the
inherent variability of the angle at which the dowel sat relative
to the pelvis, between subject comparisons of lumbopelvic angle
were not meaningful and are therefore not reported.
2.6. Statistical analysis
Two-way repeated-measures analyses of variance (RM-ANO-
VAs) using peak EMG amplitude as the outcome variable and
including task and posture as a main effects were first performed
separately on the PFM EMG data, since this was the main muscle
of interest in the study. Since the two-way interaction between
task and posture was not significant, Tukey’s one-way post hoc
analyses were used to determine the effect of posture on PFM acti-
vation amplitude for each task. For the trunk muscles, a three-way
RM-ANOVA was performed to determine the effect of muscle, pos-
ture and task on the peak EMG activity. There were no three-way
interactions. Where there were significant two-way interactions
Tukey’s two-way post hoc analyses were used to (i) fix task and
test the effect of posture among the muscles then (ii) fix muscle
and test for the effect of posture among the tasks. For the muscle
activation timing data, separate two-way RM-ANOVAs including
muscle and posture as main effects were performed for each of
the tasks analyzed (cough, load catch with eyes open and load
catch with eyes closed). A two-way RM-ANOVA was conducted
for the manometry data to compare peak intra-vaginal pressure
among the postures and tasks. Finally average lumbopelvic angles
were evaluated using one-way RM-ANOVAs to ensure that the
168 A.C. Capson et al. / Journal of Electromyography and Kinesiology 21 (2011) 166–177
postures were significantly different from one another. An alpha le-
vel of 0.05 was used for all statistical tests.
Since the experiment was designed using repeated measures,
each participant served as her own control. In this way, the data
did not need to be normalized in order to remove between-subject
variability. As such, peak rectified and smoothed EMG data are pre-
sented – as with all EMG data, the actual values are not relevant,
however the differences in amplitude among postures and tasks
are.
3. Results
3.1. Subject demographics
Sixteen continent, nulliparous women between the ages of 22
and 41 (mean 27.1, SD = 5.48), with an average BMI of 22.8
(SD = 1.57), mean height of 164.5 cm (SD = 6.72), and mean weight
of 61.8 kg (SD = 7.58) participated. The participants reported par-
ticipating in regular physical activity of moderate to high intensity:
two reported exercising 6–7 days per week, eight reported exercis-
ing 4–5 days per week and four exercising 2–3 days per week. One
participant reported exercising for more than 15 h per week, four
reported exercising from 10 to 15 h per week, eight reported exer-
cising for 5–10 h per week, and three reported exercising from 0 to
5 h per week.
3.2. Electromyography
3.2.1. PFM EMG amplitude
The EMG data recorded during the supine and quiet standing
activities are presented in Fig. 2. The baseline activity of the PFMs
was significantly lower in supine compared to all of the standing
postures. Tonic PFM EMG activity in the standing hypo-lordotic
posture was significantly higher than that recorded in the habitual
posture. The PFM EMG results from the MVC, coughing, Valsalva,
and load-catching tasks are presented in Fig. 3. The two-way ANO-
VA revealed that there was no significant interaction between pos-
ture and task for either side of the PFMs. During the PFM MVC, both
the right and left PFMs showed significantly lower EMG ampli-
tudes in the hyperlordotic and hypo-lordotic postures as compared
to the activity recorded in the habitual posture. Similar results
were observed during coughing, where there was significantly
lower PFM activation in both the hyper- and hypo-lordotic pos-
tures as compared to the habitual posture. For the Valsalva data,
as noted above, 4 of the 16 subjects had motion artefact evident
in their PFM EMG data and thus the data from these four subjects
were omitted from the analysis. The resultant ANOVA (n= 12), was
consistent with the results seen in the other tasks, both the left and
right PFMs showed significantly lower activation in both the
hyperlordotic and hypo-lordotic position, as compared to the
habitual posture. Finally, during the load-catching task, similar
PFM EMG results were found in both the eyes open and the eyes
closed conditions. In the hyperlordotic and hypo-lordotic postures,
both the right and left PFM were activated to a significantly lower
level than in the habitual posture.
3.2.2. Trunk muscle EMG amplitude
As with the PFMs, the tonic EMG activity recorded from the
trunk muscles was significantly lower in supine than in standing.
EMG activation while quietly standing in the different postures
was not significantly different for any of the muscles studied.
In the three-way ANOVA including muscle, posture and task as
main effects, there was no significant three-way interaction, but
there were significant two-way interactions between each pair of
factors (i.e., muscle by posture, muscle by task and posture by
task). The post hoc pairwise comparisons with posture fixed indi-
cated that for all muscles except the ES, the cough produced higher
EMG activity than all other tasks (Fig. 3). In ES there was no signif-
icant difference in EMG activity across all of the tasks.
The EMG activation amplitudes of all muscles across all pos-
tures is presented in Fig. 4. Higher EMG activation was generally
seen when tasks were performed in the habitual posture. The
EMG activation of all muscles across all tasks is presented in
Fig. 5. All muscles except ES were activated at a higher level during
coughing than during the other tasks. Across tasks, for the EO IO
and PFMs, the PFM MVC and Valsalva produced similar levels of
EMG activation, which were significantly higher than the amount
of activation produced during the load-catching tasks. For RA there
was no difference in the amount of EMG activation produced dur-
ing the PFM MVC, the Valsalva and the load-catching tasks.
Fig. 2. Pelvic floor muscle EMG tonic activation amplitude changes related to posture. Ldenotes the activity recorded from the left side of the PFMs. Rindicates activity
recorded from the right side of the PFMs. RMS = root mean square. Bars represent the mean value and whiskers represent the 95% confidence interval for the mean. Note that,
as denoted by ‘‘*”, tonic EMG activity was higher in all standing postures than in the supine position, and tonic EMG was higher in the hypo-lordotic posture than the other
postures.
A.C. Capson et al. / Journal of Electromyography and Kinesiology 21 (2011) 166–177 169
3.2.3. EMG activation timing
Timing data were only analyzed for the perturbation tasks (i.e.,
the cough and the load-catching tasks) since the PVM MVC and
Valsalva tasks were graded and thus had no abrupt onset making
timing data highly variable and less relevant. For the cough task,
there was no significant muscle by posture interaction and no
muscle or posture main effect, suggesting that the muscles were
not significantly different in terms of their activation timing and
that the timing was not significantly affected by posture. The tim-
ing of the RA activation was, however, significantly different from
the onset of the PFM activity, where the RA was activated 106
(SD = 33) ms before the PFMs. The timing of the muscle activation
peak, was different among the muscles (p= 0.003) and showed a
trend towards being different among the postures (p= 0.06), but
this trend did not reach statistical significance. With all postures
combined, the EO and IO reached peak activation simultaneously
with the PFMs, and the ES and RA reached their peak activation
450 (SD = 91) ms and 292 (SD = 93) ms after the PFMs reached
their peak, respectively.
For the load-catching task performed with eyes open, there was
a significant difference in activation timing among muscles
(p= 0.032) and again a marginal effect of posture on activation
timing (p= 0.085). The Tukey’s pairwise comparisons revealed that
the EO was activated significantly earlier (26 ± 50 ms before the
load hit the box) than the ES (120 ± 19 ms after the load hit the
box) but that none of the other muscles demonstrated significant
Fig. 3. PFM EMG amplitude presented by posture and by task. EMG amplitude is presented as the peak root mean square activation across each data set. MVC denotes the
maximum voluntary contraction task, Box EO denotes the load-catching task performed with eyes open, and Box EC denotes the load-catching task performed with the eyes
closed. Hyper denotes the hyperlordotic posture and Hypo denotes the hypo-lordotic posture. Ldenotes activity recorded from the left PFMs and Rdenotes activity recorded
from the right PFMs. The between-task results are indicated by the # symbols. The cough task, labeled with # produced higher EMG activation than the tasks labeled with ##
and the tasks labeled with ## produced higher EMG activation than the tasks labeled with ###. Within each task, the posture marked with * produced higher EMG activation
than the postures marked with **.
Fig. 4. EMG activity of the different muscles across the three postures. EMG amplitude is presented as the peak root mean square activation across each data set. EO denotes
the external oblique muscle, ES denotes the erector spinae muscles, IO denotes the internal oblique muscle, PFML denotes the left pelvic floor muscle, PFMR denotes the right
pelvic floor muscle and RA denotes the rectus abdominus muscle. Hyper denotes the hyperlordotic posture and Hypo denotes the hypo-lordotic posture. Postures marked with
* resulted in significantly higher muscle activity than that recorded in the other postures.
170 A.C. Capson et al. / Journal of Electromyography and Kinesiology 21 (2011) 166–177
differences in activation onset timing. For the timing of the peak
EMG activation, the activity of all muscle peaked significantly later
(583 ± 41 ms) than the time the load landed in the box and there
was no significant interaction between posture and muscle.
For the load-catching task performed with eyes closed, there
was again a significant difference in activation timing among mus-
cles (p= 0.035) and no significant difference between postures
(p= 0.123). The IO were activated before the load hit the box
(91 ± 39 ms), whereas there were no significant differences in acti-
vation timing among the other muscles, neither relative to the time
the load hit the box nor relative to each other. There were signifi-
cant differences in the timing of peak activation between postures
(p= 0.003) and muscles (p= 0.05). As with the same task per-
formed with the eyes open, all muscle activation peaked after the
load hit the bottom of the box. The activity of RA peaked
412 ± 70 ms before the activity of ES. Interestingly, the average
time for the muscles to reach peak activation was 335 ± 49 ms later
when the task was performed in the hyperlordotic posture as com-
pared to habitual posture.
3.3. Vaginal manometry
As with the EMG results the baseline intra-vaginal pressure was
significantly lower in supine as compared to habitual, hyperlordot-
ic, or hypo-lordotic standing postures. In standing, the baseline
pressure was significantly higher in hyperlordosis as compared
to normal, and was significantly higher in the hypo-lordotic pos-
ture as compared to the hyperlordotic posture. In general, as pre-
sented in Fig. 6, across the tasks the highest changes in intra-
vaginal pressure were generated in the hypo-lordotic posture,
while the lowest pressure changes were generated in the habitual
standing posture. These results were consistent across four of the
five tasks (MVC, coughing, and load catching with both eyes open
and eyes closed). During the Valsalva manoeuvre there was as
much intra-vaginal pressure generated in the habitual posture as
in the hypo-lordotic posture. During the load-catching task, the re-
sults were consistent between the eyes open and eyes closed con-
ditions, and both the hyper- and hypo-lordotic postures resulted in
higher intra-vaginal pressure than the same task performed in the
habitual posture.
3.4. Lumbopelvic angle
The lumbopelvic angle was calculated to ensure that the pos-
tures differed significantly from one another during each of the
tasks. There were consistent differences between the subject’s nor-
mal lumbopelvic angle and the hypo-lordotic and hyperlordotic
angles across the tasks and during both the EMG and the pressure
trials. The mean (SD) angles in the hyperlordotic and hypo-lordotic
postures averaged across the four tasks relative to angle recorded
in the habitual posture were 8.1 (0.92) and 2.6 (1.4) degrees,
respectively. There were similarly significant differences in the pel-
vic angle recorded in the three postures while pressure data were
recorded. The mean (SD) hyperlordotic and hypo-lordotic angles
averaged across the tasks relative to the habitual posture angle
were 4.4 (2.3) and 3.3 (0.82), respectively.
4. Discussion
The main aim of this study was to determine the effect of
changing lumbopelvic posture on the activity of the PFMs and on
the amount of vaginal squeeze pressure generated at rest and dur-
ing tasks that challenged both the lumbar stability and urinary
continence systems. The results showed that there was signifi-
cantly higher resting PFM activity in all postures in standing as
compared to supine, and that in the standing position, there was
higher resting PFM activity in the hypo-lordotic posture as com-
pared to the normal and hyperlordotic postures. During the MVC,
cough, Valsalva, and load-catching tasks, subjects generated signif-
icantly more PFM EMG activity when in their normal posture than
when in hyper- or hypo-lordotic postures (p< 0.05 in all cases).
Conversely, higher peak intra-vaginal pressures were generated
in the hypo-lordotic posture for all tasks (p< 0.05 in all cases).
These results clearly indicate that changes in lumbopelvic posture
influence both the contractility of the PFMs and the amount of in-
tra-vaginal pressure generated during static postures and during
dynamic tasks. They also highlight that manometry measures
within that vaginal canal do not solely reflect PFM activation
amplitude. Lumbopelvic posture does not appear to have a signif-
icant effect on the timing of PFM activation during coughing or
load-catching tasks.
Fig. 5. EMG activity of the different muscles across the five tasks. EMG amplitude is presented as the peak root mean square activation across each data set. MVC denotes the
maximum voluntary contraction task, Box EO denotes the load-catching task performed with eyes open, and Box EC denotes the load-catching task performed with the eyes
closed. EO denotes the external oblique muscle, ES denotes the erector spinae muscles, IO denotes the internal oblique muscle, PFML denotes the left pelvic floor muscle, PFMR
denotes the right pelvic floor muscle and RA denotes the rectus abdominus muscle. Within each muscle, the task marked with * produced higher EMG activation than the
tasks marked with ** and the tasks marked with ** produced higher EMG actiation than the unmarked tasks.
A.C. Capson et al. / Journal of Electromyography and Kinesiology 21 (2011) 166–177 171
Changes in lumbopelvic posture to have a similar effect on the
amplitude of abdominal muscle activation during the tasks stud-
ied, and no effect on muscle activation timing.
4.1. Electromyography data
4.1.1. PFM activation amplitude
In this study, tonic PFM EMG activity was higher in all standing
positions as compared to supine. This corroborates results reported
by Shafik et al. (2003), who found higher EMG activity of the leva-
tor ani in a standing position as compared to a recumbent position.
Considering the role of the PFMs in supporting the pelvic contents,
it is intuitive that the tonic activity of the PFM would be higher in
gravity-dependent (standing) positions where the weight of the
viscera is more directly applied to the PFM, rather than in a grav-
ity-independent (supine) position where the viscera rest against
the posterior abdominal wall.
The tonic activity of the PFMs was significantly higher in the
hypo-lordotic posture as compared to habitual posture
(p= 0.0054). The angle of the sacrum is related to the degree of
lumbar lordosis (Nakipoglu et al., 2008; Stagnara et al., 1982;
Vaz et al., 2002), and the degree of lumbar lordosis is related to
the degree of pelvic tilt (Harrison et al., 2002; Levine and Whittle,
1996). In a hypo-lordotic position, the pelvis is tilted posteriorly,
which may result in an antero-inferior movement of the sacrum
and a postero-superior movement of the pubis. As such, creating
a hypo-lordotic deviation from an individual’s habitual standing
posture may shorten the PFMs by changing the orientation of their
attachments at the sacrum, coccyx and pubis. Although this has not
been substantiated empirically, the hypo-lordotic standing posture
may bring the attachments of the pubococcygeus muscle at the
pubis and the coccyx closer together, thereby shortening the key
PFM, the pubococcygeus muscle, from which EMG activity was re-
corded (Enck and Vodusek, 2006; Deindl et al., 1994; Brown, 2008).
There is some evidence that a muscle may receive more excitatory
input from the central nervous system when it is held in a short-
ened position (Pasquet et al., 2005; Del Valle and Thomas, 2004),
which may explain the increase in resting PFM activity when sub-
jects were standing with a reduced lumbar lordosis. The degree of
muscle shortening, and therefore the degree of neuromodulation,
is likely dependent on the compliance of the ligaments in the lum-
bopelvic region, and therefore may be different in parous or older
women.
The results seen in the current study, whereby PFM EMG activa-
tion amplitude was affected by changes in lumbopelvic posture,
augment the results of previous studies that investigated differ-
ences in PFM activation amplitude among different seated postures
(Sapsford et al., 2006, 2008). Sapsford et al. (2006) examined PFM
activation in slumped supported, tall unsupported, and very tall
unsupported sitting postures and found the highest EMG activation
of the PFM when subjects assumed a very tall unsupported posi-
tion. These results may appear to conflict, with the findings of
the current study, where the hyperlordotic posture resulted in low-
er EMG activation than the other two postures. However the seated
postures studied by Sapsford et al. (2006) are not equivalent to the
standing postures used in the current study. In particular, the very
tall, unsupported sitting posture does not necessarily result in a
larger lumber lordosis than tall supported sitting and as such, the
differences in PFM EMG activation reported by Sapsford et al.
(2006) may be related to the increase in abdominal muscle activa-
tion generated by the very tall unsupported posture (O’Sullivan
et al., 2002), whereby the resultant increase in intra-abdominal
pressure creates activation of the PFMs to support the abdominal
and pelvic organs.
During all of the dynamic tasks tested in this study, the PFM
EMG activation was higher in the neutral posture than in the exag-
gerated or diminished lordosis postures. Several different contrac-
tion levels were tested in this study, and yet these results were
consistent across all tasks (p< 0.05 in all cases). As discussed above
regarding tonic muscle activation, it can be postulated that the
contractility of the PFMs is affected by postural changes as a result
of alterations induced in muscle length, although, again it is not
clear that the different postures that were held by the participants
in this study had any appreciable effect on changing the length of
the PFMs. Changes in muscle length may be confirmed by studying
changes in the dimensions of the levator hiatus induced by the dif-
ferent postures. If the different postures assumed by participants in
this study did induce changes in muscle length, then this could
have resulted in alterations in the length-tension relationships of
the PFMs (Odegard et al., 2008; Beck et al., 1998). Studies on other
limb muscles have produced conflicting results, with some muscles
being activated to a higher degree when held in a lengthened
Fig. 6. Intra-vaginal pressure measured across tasks and postures. Peak pressure is presented as the rise in pressure over that recorded at rest. MVC denotes the maximum
voluntary contraction task, Box EO denotes the load-catching task performed with eyes open, and Box EC denotes the load-catching task performed with the eyes closed. Hyper
denotes the hyperlordotic posture and Hypo denotes the hypo-lordotic posture. The tasks marked with * produced significantly more intra-vaginal pressure than the
unmarked tasks.
172 A.C. Capson et al. / Journal of Electromyography and Kinesiology 21 (2011) 166–177
position and others being activated to a higher degree when held in
a shortened position (Suter and Herzog, 1997; Forti and Guirro,
2008; Cresswell et al., 1995; Brondino et al., 2002). Changes in
muscle length may also change the moment arm length of the
PFMs, which would suggest that more or less muscle activation
would be required to accomplish a given task (Nourbakhsh and
Kukulka, 2004). Changes in muscle length may also affect the cen-
tral modulation of muscle activation (Pasquet et al., 2005; Del Valle
and Thomas, 2004). In the current study, it must be noted that the
changes in EMG activation seen with changes in lumbopelvic pos-
ture may simply be related to changes in the number of active
muscle fibres within the recording volume of the EMG electrode,
or due to changes in the amount of cross-talk recorded from near-
by muscles (Auchincloss and McLean, 2009) and not to changes in
muscle length.
The results of two previous studies that examined the effect of
ankle position on dynamic PFM activation in standing (Chen et al.,
2005; Cerruto et al., 2008) conflict with the results of the current
study. In both studies, the study design was based on the assump-
tion that inducing passive ankle dorsiflexion (DF) with subjects in
standing would facilitate an anterior pelvic tilt (hyperlordosis), and
that inducing passive ankle plantarflexion (PF) would create a pos-
terior pelvic tilt (hypolordosis). In the first study by Chen et al.
(2005), their results indicated that there was higher tonic PFM
EMG activity in DF (consistent with the higher tonic PFM EMG acti-
vation in hypolordosis in current study), as compared to neutral or
PF, and that there was a trend towards inducing higher PFM EMG
activation during an MVC when subjects moved from a position
of ankle PF (hyperlordosis) to neutral (habitual lordosis) to DF
(hypolordosis). These changes in PFM activity could be related to
the intra-abdominal pressure generated through abdominal mus-
cle contraction required to maintain postural control on a tilted
surface as easily as it could be related to changes in PFM length
or contractility. While the results of the current study support that
there is decreased maximum voluntary activation of the PFM in the
hypo-lordotic posture, there was also reduced PFM activation
when MVCs were performed in the hyperlordotic posture. In the
second study by Cerruto et al. (2008) which was of similar design
to that of Chen et al. (2005), the authors reported that there were
no differences in maximal PFM EMG activation when continent
women stood with different ankle angles (n= 35). The results of
Chen et al. (2005) and Cerruto et al. (2008) must be interpreted
with caution, however, as neither group provided evidence that
the changes in ankle position used in their studies actually induced
significant changes in lumbopelvic posture. The effect of changing
ankle position from DF to PF may well have an effect on lumbo-
pelvic angle, however the magnitude of this effect is likely quite
small.
4.1.2. Trunk muscle activation amplitudes
The activation amplitudes of the abdominal muscles followed a
pattern similar to that seen in the PFMs, and did not vary substan-
tially among the postures. The abdominal muscles contracted most
strongly during the cough. These findings are not surprising since
the abdominal muscles have been shown to work synergistically
with the PFMs, and since they contribute to the increase in intra-
abdominal pressure generation during coughing (Madill et al.,
2009) and functional tasks (Smith et al., 2007a,b). The only muscle
that was affected by the change in posture is IO, which contracted
to a higher level when subjects performed the functional tasks in
their habitual posture as compared to a hypo-lordotic posture.
Since the IO muscle can act as a trunk flexor (Moore and Dalley,
2006), the increase in activation seen may reflect the fact the mus-
cle was statically active to hold the hypo-lordotic posture, and due
to this dual task demand, it could not increase its activation to the
same extent during tasks performed in that posture. The reader is
reminded that the RMS of the tonic muscle activity was removed
from each data file during processing.
The only trunk extensor muscle included in the data set was the
lumbar erector spinae, which showed no significant difference in
activation amplitude among the postures or the tasks. The ES are
not directly involved in the generation of IAP and were not ex-
pected to have activation amplitudes modulated by the PFMs.
4.1.3. PFM and trunk muscle activation timing
Postural changes did not affect the activation onset timing of
any of the muscles studied. To the author’s knowledge, there have
been no previous studies investigating posture effects on muscle
activation timing. Since there is some evidence to suggest that
the timing of PFM and trunk muscle activation is an important con-
tributing factor in stress urinary incontinence (Madill et al., 2009;
van der Kooi et al., 1984; Heidler et al., 1979; Barbic et al., 2003)it
does not appear that postural correction would be an effective
treatment strategy for women with this condition. The pattern of
muscle activation found in the load-catching task in this study
was not consistent with the results found by Smith et al.
(2007a,b) during a rapid arm motion task. In particular, Smith
et al. (2007a,b) used a rapid arm motion task to act as a postural
perturbation and reported that the PFMs contracted before the del-
toid muscle (the prime mover) in continent women and after the
deltoid muscle in women with stress urinary incontinence. The
continent women in the current study did not contract their PFMs
until the time the load hit the bottom of the box, which does not
suggest that the women in the current study used their PFMs in
feed-forward control. This is further substantiated by the fact that
the onset timing results did not differ in the current study between
the eyes open and eyes closed conditions. The task used in the cur-
rent study was, however, very different from the task used by
Smith et al. (2007a). In a later study Smith et al. (2007b) used
the load-catching task in a study comparing PFM and abdominal
muscle activation amplitude, but did not report muscle activation
timing information.
4.2. Vaginal manometry findings
In contrast with the findings that maximal PFM activation was
recorded during dynamic activities when subjects were in their
habitual posture, this position did not result in the highest intra-
vaginal pressures. Instead, the highest vaginal manometry values
generated during static and dynamic activities were produced in
the hypo-lordotic position. At first glance, this contradiction seems
puzzling. It should be noted, however, that the intra-vaginal
manometry measurements made using the Peritron™ perineome-
ter are not simply representative of the force output from the PFMs
alone, but rather result from the action both of the PFMs and the
abdominal muscles (Madill and McLean, 2008) as well as pressures
caused by the gravitational forces acting on the pelvic organs (Bo
and Fickenhagen, 2003) and other increases in intra-abdominal
pressure, including diaphragmatic contraction (Hodges et al.,
2007). Although it has been shown that intra-vaginal pressure is
moderately correlated with EMG activity of the PFMs (r= 0.75)
(Workman et al., 1993), such correlations do not suggest that the
PFMs alone are responsible for the generation of intra-vaginal pres-
sure, particularly during dynamic tasks. Therefore, it is feasible that
intra-vaginal pressures could be highest in a hypo-lordotic position
despite sub-maximal PFM EMG activation in that position. In fact,
in the current study the manometry values recorded during PFM
MVC and during coughing are very similar, suggesting that there
is not a large effect of increased intra-abdominal pressure on this
measure. However, the difference in intra-vaginal pressure values
recorded during the load catching tasks are much higher than
those recorded during the MVC or the cough. Since abdominal
A.C. Capson et al. / Journal of Electromyography and Kinesiology 21 (2011) 166–177 173
muscle activity was much higher during the cough and the MVC
tasks than during the load catching tasks, the large difference in
pressure may be driven by diaphragmatic contraction and breath
holding during this task.
This phenomenon can also be explained by investigating the ac-
tions induced by contraction of the PFMs. PMF contraction results
in both (1) cephalad movement and (2) anterior translation of the
pelvic floor. The vaginal (and therefore presumably the urethral)
closure forces generated by each action and the effect of changing
lumbopelvic posture on these forces occur in concert.
The anterior translation of the PFMs generates a closure force at
the vaginal, and subsequently the urethral, lumen (Schaer et al.,
1995). If the vagina is oriented perpendicular to the PFMs
(h=90°), then all of the anterior translational force will be trans-
mitted across the vaginal wall (i.e., F
closure
=Fsin h)inFig. 7a,
resulting in vaginal closure. The true angle between the vaginal lu-
men and the pelvic floor is likely somewhat variable among wo-
men, but, based on ultrasound illustrations presented in Peng
et al. (2007) and Kruger et al. (2008), is approximately 45–60°
when women stand in an upright posture. As the angle (h) between
the vaginal lumen and the PFMs decreases, then the closure force
induced on the vaginal wall by anterior translation of the PFMs
pressing into the vagina decreases, and as the angle (h) between
the vaginal lumen and the PFM increases (up to 90°) the closure
force induced on the vaginal wall by PFM contraction alone in-
creases. Assuming the contributions of intra-abdominal pressure
are similar among the postures for each task (since there was no
difference noted in abdominal muscle activation among the pos-
tures), this may explain the effect of changing lumbopelvic posture
on the vaginal closure pressure demonstrated in the current work.
Specifically, in the hypo-lordotic posture, hmay be larger relative
to the angle seen in the normal and hyperlordotic postures and
as such the closure pressure transmitted to the vagina through
the anterior translation component of the PFM contraction may
be higher.
The cranial movement of the PFMs must also be considered. The
PFMs provide a firm surface that is able to resist downward forces
acting on the pelvic organs (Delancey, 1990; Ashton-Miller and
Delancey, 2007). If the PFMs and fascial support structures are in-
tact, then downward forces transmitted to the vaginal wall (i.e., in-
creases in intra-abdominal pressure due to abdominal muscle
activation, diaphragm descent, and due to gravity acting on the
pelvic organs), are equal and opposite to the upward forces trans-
mitted to the vagina from the PFMs and their connective tissues
(Ashton-Miller and Delancey, 2007). As seen in Fig. 7b, if a normal
lumbopelvic posture reflects a situation where the angle (h) be-
tween the pelvic floor and the vagina (h) is approximately 45°
(through inspection of Peng et al., 2007; Kruger et al., 2008) then
closure force acting on the vagina in this position is the vertical
component of the force directed cephalad by the PFMs, i.e.,
F
closure
=F
cephalad
cos hin Fig. 7b. This occurs regardless of whether
the PFMs actively contract or whether the passive tension in the
tissues resist the downward forces in the absence of muscle
contraction. When the angle between the vagina and the PFMs in-
creases or decreases above or below 45°of horizontal, the compo-
nent of the vaginal closure force attributed to the cranial
movement of the PFMs increases.
Although Singh et al. (2001) demonstrated that the normal
levator-vaginal angle (defined as the angle between a perpendicu-
lar line from the levator plate and the upper axis of the vagina) on
radiographic films was approximately 90°(±5°) in normal subjects
in supine, this value is likely an over-estimate of the angle be-
tween the pelvic floor and the vagina seen when women are in
the standing position, where the weight of the abdominal and pel-
vic organs push against the pelvic floor. Their results do suggest,
however, that in a hypo-lordotic position, the perpendicular com-
ponent of the orientation of the PFMs relative to the vaginal canal
responsible for vaginal closure through anterior translation is
maximized (h=90°in Fig. 7a), and the parallel component respon-
sible for vaginal closure through abdominal force transmission is
minimized.
4.3. Differences in EMG and manometry measured among the tasks
In the habitual posture, the maximal cough task generated sig-
nificantly higher EMG amplitudes compared to all other tasks and
the load-catching task (both the eyes open and eyes closed condi-
tions) resulted in significantly lower amplitudes compared to all
other tasks. No significant differences were found in the EMG
amplitude between the MVC and Valsalva tasks, but the Valsalva
task generated significantly higher intra-vaginal pressures as com-
pared to MVC. This higher pressure recording during a Valsalva
manoeuvre and the load-catching tasks, in the absence of higher
PFM EMG activation illustrates the ability of women to generate
intra-vaginal manometry values through the transmission of
intra-abdominal pressure to the fascial support structures at the
Fig. 7. Closure force exerted on the bladder wall through anterior translation of the pelvic floor. (a) Schematic of the bladder and urethra where the urethra is at a right angle
relative to the levator ani. The cross-sectional closure force acting on the urethra is the sine of the angle between the urethra and the pelvic floor (sin(h)), which in this case
h=90°and therefore sin(90°) = 1. In this case, therefore, all force exerted through anterior translation of the pelvic floor is directed such that it closes the cross-section of the
urethra. (b) Same schematic as above however now the angle between the urethra and the pelvic floor is 45°. Since sin(45°) = 0.5, only half the force exerted through anterior
translation of the pelvic floor will act to close the cross-section of the urethra.
174 A.C. Capson et al. / Journal of Electromyography and Kinesiology 21 (2011) 166–177
pelvic floor. This result might be different in women in whom there
is damage to the fascial supports at the pelvic floor, such as parous
women and particularly women with pelvic organ prolapse, where
active contraction of the PFMs may be necessary to take up the
slack in the connective tissues in order to transmit the appropriate
forces.
Of the four tasks performed by the participants in this study, the
Valsalva manoeuvre was expected to display the most variability in
results. Previous studies have shown there to be high variability in
subject technique and intra-abdominal pressures generated during
a Valsalva (Tunn et al., 2005; Greenland et al., 2005), and an indi-
vidual’s performance (i.e., degree of PFM contraction) may be influ-
enced by fatigue, degree of abdominal contraction, bladder
fullness, and fear of embarrassment due to uncontrolled loss of ur-
ine or gas (Greenland et al., 2007). In the current study there was
slightly higher variability in EMG amplitudes with the Valsalva
manoeuvre as compared to the other tasks when all sixteen sub-
jects were included (CV = 99.8% and 89.0% for the left and right
PFM, respectively) and when the four subjects whose data were
contaminated with motion artefact were removed (CV = 90.1%
and 87.5%). In the current study, the task that had the lowest var-
iability was the maximal coughing task, where the coefficients of
variation of the data from the left and right PFM EMG
(CVs = 60.9% and 55.6%) were lower than those observed during
the other tasks. There was no significant difference in the variabil-
ity of the data recorded during the load-catching task when the
task was performed with the eyes open (CV left PFMs = 93.4%; CV
right PFMs = 85.2%) or eyes closed (CV left PFMs = 73.3%; CV right
PFMs = 90.6%). Since the results of the current study demonstrated
similar EMG activation amplitude and timing and similar variabil-
ity in the data between the load-catching task performed with the
eyes open as compared to with the eyes closed, our data suggest
that expectation of the load did not alter the PFM contraction char-
acteristics. Work by Smith et al. (2007a,b) demonstrated that in an
unexpected condition (eyes closed), EMG activity of the PFM was
significantly greater than in an expected condition (eyes open).
Other studies have also shown that unexpected loading increases
the magnitude of the postural response on EMG (Moseley et al.,
2003; Lavender and Marras, 1995). Therefore, it is recommended
that further studies continue to utilize both the unexpected and
expected conditions.
With respect to selecting a model with which to measure auto-
matic PFM activation and resultant intra-vaginal pressures, the
tasks appear to produce similar levels of variability in the data,
however, the magnitude of the detectable response (either by
EMG or perineometry) is significantly lower during the load-catch-
ing task as compared to coughing, MVCs, or Valsalva manoeuvres
and thus these latter activities may be better positioned to illus-
trate group differences. Despite the large variation in the levels
of activity demanded by each task, the same pattern emerged
across tasks for the EMG trials. This suggests that the static and dy-
namic activity of the PFM is affected by lumbopelvic posture,
regardless of the task performed and the intensity of contraction.
Several researchers have examined PFM function using one or
two of the tasks used in this study (Bo and Stien, 1994; Smith
et al., 2007b; Thompson et al., 2006), but to date the current study
is the only one to examine trends across several dynamic tasks. Of
significance are the strikingly consistent results observed across all
of the dynamic tasks. EMG activation amplitude was significantly
higher when subjects stood in their habitual standing postures as
compared to either a hypo- or hyperlordotic posture, and maximal
intra-vaginal pressures were significantly higher in the hypo-lor-
dotic posture relative to normal or hyperlordotic postures. These
trends were observed in the EMG data when baseline activation
amplitudes were removed and when they were not, suggesting
that both the tonic (PFM activity required to sustain a given pos-
ture) and phasic (PFM activity during tasks) activity of the PFM
was affected by posture.
4.4. Lumbopelvic angle measurement
In the current study, lumbopelvic angle between the three
standing postures varied by approximately 5°. Pelvic angles were
measured relative to each individual and were not referenced to
an external coordinate system. This has two major implications
for the interpretation of the results in this study. Firstly, it is pos-
sible that a given subject’s habitual standing posture had an exag-
gerated or reduced lordosis to begin with and therefore, we cannot
assume that her habitual posture was ‘‘normal”. For example, a
subject with an initially reduced lordosis moving into a position
of relative hyperlordosis would, in fact, be in a more neutral (or
‘‘normal”) lumbopelvic position. However, all subjects were free
from low back and pelvic pain, and were drawn from a healthy, ac-
tive population.
Secondly, while the power of the statistical analysis was im-
proved through the use of a repeated-measures design, the conclu-
sions drawn from this study relate only to intra-individual changes
in PFM activity and intra-vaginal pressure relative to changes lum-
bopelvic posture. In order to form conclusions relating PFM activa-
tion and intra-vaginal pressure to the effect of postural correction
on PFM contractility, future studies should be performed in which
lumbopelvic angle is referenced to a standardized external coordi-
nate system such that comparisons between subjects with differ-
ent habitual postures can be made.
The results obtained from postural analysis displayed some
within-subject variability in the habitual angles recorded during
the EMG as compared to the pressure trials despite the fact that
participants were given the same instructions when each set of
data were collected. The mean (SD) habitual angle calculated
across all tasks when the EMG data were recorded was 82.7
(1.5)°whereas it was 87.1 (0.93)°when the pressure data were re-
corded. The instrumentation was adjusted during the transition
between the Periform™ EMG probe and the Peritron™ perineome-
ter. Therefore, the initial position of the pelvic rod may have chan-
ged and may explain the difference in lumbopelvic angle means
between EMG and pressure trials. However, relative changes in
the hyper- and hypo-lordotic positions are comparable within par-
ticipants and between postures.
5. Conclusions and clinical implications
There was significantly higher resting PFM activity in all pos-
tures in standing as compared to supine, and in the standing posi-
tion, there was higher resting PFM activity in the hypo-lordotic
posture as compared to the normal and hyperlordotic postures.
During the MVC, cough, Valsalva, and load-catching tasks, subjects
generated significantly more PFM EMG activity when in their nor-
mal posture than when in hyper- or hypo-lordotic postures. Con-
versely, higher peak intra-vaginal pressures were generated in
the hypo-lordotic posture for all tasks (p< 0.05 in all cases). These
results clearly indicate that changes in lumbopelvic posture
influence both the contractility of the PFMs and the amount of in-
tra-vaginal pressure generated during static postures and during
dynamic tasks. Lumbopelvic posture does not, however, appear
to have a significant effect on the timing of PFM activation during
coughing or load-catching tasks. Changes in lumbopelvic posture
do not appear to have any significant effect on the amplitude and
timing of trunk muscle activation during the tasks studied. The
only exception was that IO was activated to a higher level when
the tasks were performed in habitual posture as compared to
hyper- or hypo-lordotic postures.
A.C. Capson et al. / Journal of Electromyography and Kinesiology 21 (2011) 166–177 175
It should first be noted that the current study was conducted
using a convenience sample of healthy, continent women and
therefore the results cannot be generalized to individuals with
SUI or with other conditions such as lumbopelvic pain. Further re-
search should investigate whether the observed trends are consis-
tent in a symptomatic or pathological population of women or in
men.
The results of this study may have important implications for
facilitating PFM contraction and for training the PFMs. Since the
hypo-lordotic posture resulted in higher tonic activation of the
PFMs, this posture may be a good posture in which to facilitate
PFM contraction in women who have difficulty activating the PFMs
voluntarily. Since the habitual posture was that which resulted in
the highest activation during voluntary PFM contractions and dur-
ing functional tasks, therapists should ensure that women are not
deviating from their habitual posture when attempting to contract
their PFMs during strengthening exercises. Further, in young,
healthy, nulliparous women, coughing induced a larger PFM con-
traction than the other tasks and as such may be a useful approach
to initiate strengthening of the PFMs in women who have difficulty
activating their PFMs voluntarily. Women with SUI who present
with postural defects (exaggerated lumber lordosis or flattened
lordosis) and decreased PFM activity (Thompson et al., 2006),
may benefit from postural interventions to improve tonic and pha-
sic PFM activation. Again, a future study should investigate the
effectiveness of assuming a hypo-lordotic posture on the incidence
and severity of urine leakage in women with SUI.
When considering women with lumbopelvic pain, the results of
this study indicate that neutral pelvic positioning in an individual’s
relaxed standing posture facilitates greater PFM activation during
MVC and during functional tasks. In this position, the PFMs have
the greatest potential to co-contract with the abdominal muscles
and with the diaphragm to maximize the generation of intra-
abdominal pressure to promote spinal stability. Consequently,
deviations from a participant’s normal standing posture may ad-
versely affect both spinal stability and continence control. Assum-
ing that the posture in which maximal PFM EMG is generated
equates with the posture in which the PFMs can generate the most
force to support the pelvic organs, the maximal PFM contractility
noted in the subjects’ habitual posture may be important for pa-
tients with pelvic pain syndromes. These results support the idea
of performing functional tasks with a ‘‘neutral spine”, where func-
tion is maximized and the risk of injury is minimized (Oxland and
Panjabi, 1992; Kiefer et al., 1997).
The manometry results from this study serve to highlight con-
cerns with using manometry as and outcome measure to quantify
the force generated during a PFM contraction. Since the tasks and
postures that resulted in the highest PFM EMG activation do not
match the postures and tasks that generate the highest intra-vag-
inal pressures, clinicians should be wary of using manometry dur-
ing functional tasks if their aim is to study PFM function.
References
Ashton-Miller JA, Delancey JOL. Functional anatomy of the female pelvic floor. Ann.
N. Y. Acad. Sci. 2007;1101:266–96.
Auchincloss C, McLean L. The reliability of surface EMG recorded from the pelvic
floor muscles. J. Neurosci. Methods 2009;182(1):85–96.
Barbic M, Kralj B, Cor A. Compliance of the bladder neck supporting structures:
importance of activity pattern of levator ani muscle and content of elastic fibers
of endopelvic fascia. Neurourol. Urodyn. 2003;22:269–76.
Beck J, Sinderby C, Lindstrom L, Grassino A. Effects of lung volume on diaphragm
EMG signal strength during voluntary contractions. J. Appl. Physiol.
1998;85(3):1123–34.
Bo K, Finckenhagen HB. Is there any difference in measurement of pelvic floor
muscle strength in supine and standing position? Acta Obstet. Gynecol. Scand.
2003;82:1120–4.
Bo K, Stien R. Needle EMG registration of striated urethral wall and pelvic floor
muscle activity patterns during cough, Valsalva, abdominal, hip adductor and
gluteal muscle contractions in nulliparous healthy females. Neurourol.
Urodynam. 1994;13:35–41.
Brondino L, Suter E, Lee H-D, Herzog W. Elbow flexor inhibition as a function of
muscle length. J. Appl. Biomech. 2002;18(1):46–56.
Brown, C., 2008. Reliability of Electromyography Detection Systems for the Pelvic
Floor Muscles. Master’s thesis, Queen’s University, Kingston, ON, Canada.
Cerruto M, Vedovi E, Rossi S, Mantovani W, Pozzo A, Cangemi A, et al. The effect of
ankle inclincation in supine and standing position on the electromyographic
activity of abdominal and pelvic floor muscles in women with and without
stress urinary incontinence: preliminary results from a pilot study. Eur. Urol.
Suppl. 2008;7(3):145.
Chaitow L. Chronic pelvic pain: pelvic floor problems, sacroiliac dysfunction and the
trigger point connection. J. Bodywork Mov. Ther. 2007;11:327–39.
Chen C-H, Huang M-H, Chen T-W, Weng M-C, Lee C-L, Wang G-J. Relationship
between ankle position and pelvic floor muscle activity in female stress urinary
incontinence. Urology 2005;66:288–92.
Cresswell G, Loscher WN, Thorstensson A. Influence of gastrocnemius muscle length
on triceps surae torque development and electromyographic activity in man.
Exp. Brain Res. 1995;105(2):283–90.
Dean N, Wilson D, Herbison P, Glazener C, Aung T, MacArthur C. Sexual function,
delivery mode history, pelvic floor muscle exercises and incontinence: a cross-
sectional study six years post-partum. Aust. N. Z. J. Obs. Gynecol.
2008;48:302–11.
Deindl FM, Vodusek DB, Hesse U, Schussler B. Pelvic floor activity patterns:
comparison of nulliparous continent and parous urinary stress incontinent
women. A kinesiological EMG study. Br. J. Urol. 1994;73:413–7.
Del Valle A, Thomas CK. Motor unit firing rates during isometric voluntary
contractions performed at different muscle lengths. Can. J. Physiol. Pharmacol.
2004;82:769–76.
Delancey JOL. Anatomy and physiology of urinary continence. Clin. Obstet. Gynecol.
1990;33:298–307.
Enck P, Vodusek DB. Electromyography of pelvic floor muscles. J. Electromyogr.
Kinesiol. 2006;16(6):568–77.
Forti F, Guirro RJ. Comparison of the electromyographic variables at different
muscle lengths and contraction intensities. Electromyogr. Clin. Neurophysiol.
2008;48(1):3–8.
Greenland H, Bryant A, Hosker G, Smith A. Can the Valsalva manoeuvre be
standardized? Neurourol. Urodyn. 2005;22(5):357–548.
Greenland H, Hosker G, Smith A. A valsalvometer can be effective in standardizing
the Valsalva manoeuvre. Int. Urogynecol. J. 2007;18:499–502.
Grimstone SK, Hodges PW. Impaired postural compensation for respiration in
people with recurrent low back pain. Exp. Brain Res. 2003;151:218–24.
Harrison DE, Cailliet R, Harrison DD, Janik TJ. How do anterior/posterior translations
of the thoracic cage affect the sagittal lumbar spine, pelvic tilt, and thoracic
kyphosis? Eur. Spine J. 2002;11:287–93.
Heidler H, Wölk H, Jonas U. Urethral closure mechanism under stress conditions.
Eur. Urol. 1979;5:110–2.
Hodges PW, Sapsford R, Pengel LH. Postural and respiratory functions of the pelvic
floor muscles. Neurourol. Urodyn. 2007;26:362–71.
Kelly M, Tan B-K, Thompson J, Carroll S, Follington M, Arndt A, et al. Healthy adults
can more easily elevate the pelvic floor in standing than in crook-lying: an
experimental study. Aust. J. Physiother. 2007;53:187–91.
Kiefer A, Shirazi-Adl A, Parnianpour M. Stability of the human spine in neutral
postures. Eur. Spine J. 1997;6:45–53.
Kruger JA, Heap SW, Murphy BA, Dietz HP. Pelvic floor function in nulliparous
women using three-dimensional ultrasound and magnetic resonance imaging.
Obset. Gynecol. 2008;111:631–8.
Lavender SA, Marras WS. The effects of a temporal warning signal on the
biomechanical preparation for sudden loading. J. Electromyogr. Kinesiol.
1995;5:45–56.
Levine D, Whittle MW. The effects of pelvic movement on lumbar lordosis in the
standing position. JOSPT 1996;24(3):130–6.
Madill S, Harvey MA, McLean L. Women with SUI demonstrate motor control
differences during voluntary pelvic floor muscle contractions. Int. Urogynaecol.
J. 2009;20(4):447–59.
Madill S, McLean L. Quantification of abdominal and pelvic floor muscle synergies in
response to voluntary pelvic floor muscle contractions. J. Electromyogr.
Kinesiol. 2008;18:955–64.
Moseley GL, Hodges PW, Gandevia SC. External perturbation of the trunk in
standing humans differentially activates components of the medial back
muscles. J. Physiol. (Camb). 2003;547:581–7.
Moore Keith L, Dalley Arthur R. Clinically Oriented Anatomy. fifth ed. Lippincott
Williams and Wilkins; 2006.
Nakipoglu GF, Karagoz A, Ozgirgin N. The biomechanics of the lumbosacral region in
acute and chronic low back pain patients. Pain Physician. 2008;11:505–11.
Neumann P, Gill V. Pelvic floor and abdominal muscle interaction: EMG activity and
intra-abdominal pressure. Int. Urogynecol. J. 2002;13:125–32.
Ng JK, Kippers V, Richardson CA. Muscle fibre orientation of abdominal muscles and
suggested surface EMG electrode positions. Electromyogr. Clin. Neurophysiol.
1998;38:51–8.
Nourbakhsh MR, Kukulka CG. Relationship between muscle length and moment
arm on EMG activity of human triceps surae muscle. J. Electromyogr. Kinesiol.
2004;14(2):263–73.
O’Sullivan PB, Grahamslaw KM, Kendell M, Lapenskie SC, Moller NE, Richards KV.
The effect of different sitting and standing postures on trunk muscle activity in
a pain free population. Spine 2002;27:1238–44.
176 A.C. Capson et al. / Journal of Electromyography and Kinesiology 21 (2011) 166–177
Odegard GM, Haut Donahue TL, Morrow DA, Kaufman KR. Constitutive modeling of
skeletal muscle tissue with an explicit strain–energy function. J. Biomech. Eng.
2008;130:061017.
Oxland TR, Panjabi MM. The onset and progression of spinal injury: a
demonstration of neutral zone sensitivity. J. Biomech. 1992;25(10):1165–72.
Pasquet B, Carpentier A, Duchateau J. Change in muscle fascicle length influences
the recruitment and discharge rate of motor units during isometric
contractions. J. Neurophysiol. 2005;94:3126–33.
Peng Q, Jones R, Shishido K, Christos E, Constantinou CE. Ultrasound evaluation of
dynamic responses of female pelvic floor muscles. Ultrasound Med. Biol.
2007;33(3):342–52.
Raizada V, Mittal RK. Pelvic floor anatomy and applied physiology. Gastroenterol.
Clin. North Am. 2008;37:493–509.
Santiesteban A. Electromyographic and dynamometric characteristics of female
pelvic floor musculature. Phys. Ther. 1988;68:344–50.
Sapsford RR, Hodges PW. Contraction of the pelvic floor muscles during abdominal
maneuvers. Arch. Phys. Med. Rehabil. 2001;82:1081–8.
Sapsford RR, Hodges PW, Richardson CA, Cooper DH, Markwell SJ, Jull GA. Co-
activation of the abdominal and pelvic floor muscles during voluntary exercises.
Neurol. Urodyn. 2001;20:31–42.
Sapsford RR, Richardson CA, Stanton WR. Sitting posture affects pelvic floor muscle
activity in parous women: an observational study. Aust. J. Physiother.
2006:219–22.
Sapsford RR, Richardson CA, Maher CF, Hodges PW. Pelvic floor muscle activity in
different sitting postures in continent and incontinent women. Arch. Phys. Med.
Rehabil. 2008;89:1741–7.
Schaer GN, Koechli OR, Schuessler B, Haller U. Perineal ultrasound for evaluating the
bladder neck in urinary stress incontinence. Obstet. Gynecol. 1995;85:220–4.
Shafik A, Doss S, Asaad S. Etiology of the resting myoelectric activity of the levator
ani muscle: physioanatomic study with a new theory. World J. Surg.
2003;27:309–14.
Singh K, Reid WMN, Berger LA. Assessment and grading of pelvic organ prolapse by
use of dynamic magnetic resonance imaging. Am. J. Obstet. Gynecol.
2001;185:71–7.
Smith MD, Russell A, Hodges PW. Disorders of breathing and continence have a
stronger association with back pain than obesity and physical activity. Aust. J.
Physiother. 2006;52:11–6.
Smith MD, Coppieters MW, Hodges PW. Postural activity of the pelvic floor muscles
is delayed during rapid arm movements in women with stress urinary
incontinence. Int. Urogynecol. J. 2007a;18:901–11.
Smith MD, Coppieters MW, Hodges PW. Postural response of the pelvic floor and
abdominal muscles in women with and without incontinence. Neurourol.
Urodyn. 2007b;26:377–85.
Stagnara P, De Mauroy JC, Dran G, Gonon GP, Costanzo G, Dimnet J, et al. Reciprocal
angulation of vertegral bodies in sagittal plane: approach to references for the
evaluation of kyphosis and lordosis. Spine 1982;7(4):335–43.
Suter E, Herzog W. Extent of muscle inhibition as a function of knee angle. J.
Electromyogr. Kinesiol. 1997;7(2):123–30.
Thompson JA, O’Sullivan PB, Brija NK, Neumann P. Altered muscle activation
patterns in symptomatic women during pelvic floor muscle contraction and
Valsalva manoeuvre. Neurourol. Urodynam. 2006;25:268–76.
Tunn R, Schaer G, Peschers U, Bader W, Gauruder A, Hanzal A, et al. Updated
recommendations on ultrasonography in urogynecology. Int. Urogynecol. J.
Pelvic Floor Dysfunct. 2005;16(3):236–41.
van der Kooi JB, van Wanroy PJA, de Jonge MC, Kornelis JA. Time separation between
cough pulses in bladder, rectum and urethra in women. J. Urol.
1984;132:1275–8.
Vaz G, Roussouly P, Berthonnaud E, Dimnet J. Sagittal morphology and equilibrium
of pelvis and spine. Eur. Spine J. 2002;11:80–7.
Workman DE, Cassisi JE, Dougherty MC. Validation of surface EMG as a measure of
intra-vaginal and intra-abdominal activity: implications for biofeedback-
assisted Kegel exercises. Psychophysiology 1993;30:120–5.
Zondervan KT, Kennedy SH. Epidemiology of chronic pelvic pain. Int. Congr. Ser.
2005;1279:77–84.
Angela Capson graduated from the concurrent
Bachelor of Physical and Health Education and
Bachelor of Science program at Queen’s University in
2007 and went on to complete her Master of Science
in Physical Therapy at Queen’s in 2009. She is cur-
rently working as a Physiotherapist at Lifemark
Health Esquimalt in Victoria, BC. Her clinical interests
include whiplash-associated disorders, lumbopelvic
dysfunction, and chronic pain. She is currently pur-
suing the post-graduate Diploma of Advanced
Orthopaedic Manual and Manipulative Physiother-
apy.
Joe Nashed received a B.Sc. degree in Electrical
Engineering and M.Sc. degree in Rehabilitation Sci-
ence from Queen’s University in 2006 and 2008,
respectively. He is currently a doctoral student in the
Centre for Neuroscience Studies at Queen’s Univer-
sity. His research interests include understanding the
interaction between motor behaviour, limb
mechanics and neural processing as well as biological
signal processing and electrophysiology.
Linda McLean received her Bachelor of Science in
Physiotherapy at McGill University in 1990, an M.Sc.
in Electrical Engineering and a Ph.D. in Biomedical
Engineering from the University of New Brunswick.
She is currently Associate Professor in the School of
Rehabilitation Therapy at Queen’s University in
Kingston, Ontario, Canada and Chair of the Graduate
Program in Rehabilitation Science. She is an Execu-
tive Committee member of the International Society
for Electromyography and Kinesiology and is on the
Editorial Board of the Journals of Electromyogaphy
and Kinesiology, NeuroEngineering and Rehabilita-
tion and Workplace Health Management. Her
research interests include biological signal processing, electrophysiology, chronic
muscle pain and dysfunction due to overuse, and physical therapy interventions
with specific foci including urinary incontinence in women, pelvic pain and
repetitive strain injuries.
A.C. Capson et al. / Journal of Electromyography and Kinesiology 21 (2011) 166–177 177
