The effects of different sit- and curl-up positions on
activation of abdominal and hip flexor musculature
Kevin C. Parfrey, David Docherty, R. Chad Workman, and David G. Behm
Abstract: The purpose of this study was to evaluate abdominal muscle activation with variations in trunk flexion (sit or
curl up) positions, including the protocol currently used by the Canadian Society of Exercise Physiology (CSEP) Health
and Fitness Program. Electromyographic (EMG) data were collected during isometric contractions from the upper rectus
abdominis (URA), lower rectus abdominis (LRA), external obliques (EO), lower abdominal stabilizers (LAS), rectus femo-
ris (RF), and the biceps femoris (BF) in 14 subjects. Sit-up positions were varied and randomized through 3 variables: the
distance the hand traveled along the floor (5, 10, or 15 cm), bent knee or extended knee, and fixed or non-fixed feet. In re-
gard to the distance the hand traveled along the floor, the 10 cm position produced the highest activation of the LRA (p=
0.02), the 5 cm distance produced the lowest RF activation (p= 0.001), and the 15 cm distance produced the lowest acti-
vation of the URA (p= 0.001). There was no significant difference between bent-knee and extended-leg sit-up positions;
however, there was a trend (p= 0.1) showing that the bent-knee sit-up position produced higher levels of LAS activation
and lower levels of RF activation. Foot fixation resulted in significantly lower activation levels of all abdominal sites and
higher levels for the RF (p< 0.0001). The technique used for the CSEP Health and Fitness program partial curl- or sit-up
test produced the highest or equal activation levels for all abdominal muscle sites.
Key words: electromyography, rectus abdominus, external obliques, rectus femoris, CSEP Health and Fitness Program.
´:Le but de cette e
´tude est d’e
´valuer l’activation des muscles abdominaux en fonction des positions du tronc, soit
en position assise ou replie
´e, y compris la position utilise
´e dans le protocole du programme courant de Sante
physique de la Socie
´canadienne de physiologie de l’exercice (SCPE). On enregistre chez quatorze sujets l’activite
´lectrique (EMG) de la partie supe
´rieure du grand droit (URA), de sa partie infe
´rieure (LRA), des obliques externes (EO),
des stabilisateurs infe
´rieurs de l’abdomen (LAS), du droit ante
´rieur (RF) et du biceps crural (BF). Les diverses postures
´es au cours des redressements assis sont re
´atoirement en fonction de trois variables : la distance parcourue
par la main au sol (5, 10 ou 15 cm), le genou fle
´chi ou e
´tendu et le pied ancre
´ou non. En ce qui concerne la distance par-
courue par la main au sol, c’est la distance de 10 cm qui suscite le plus d’activation chez la LRA (p= 0,02), la distance
de 5 cm qui en suscite le moins chez le RF (p= 0,001) et la distance de 15 cm qui en suscite le moins chez l’URA (p=
0,001). On n’observe aucune diffe
´rence d’activation entre les positions de genou fle
´chi et de genou e
´tendu, mais on note
une tendance (p= 0,1) a
`une plus grande activation du LAS et une moins grande activation du RF quand le genou est fle
chi durant le redressement assis. Quand le pied est fixe
´, on observe de fac¸on significative les plus faibles niveaux d’acti-
vation de toutes les re
´gions de l’abdomen et les plus hauts niveaux d’activation du RF (p< 0,0001). La technique utilise
dans le programme Sante
´et condition physique de la SCPE, soit le redressement assis partiel, suscite le plus ou autant
d’activation dans tous les muscles de l’abdomen.
´lectromyographie, grand droit, obliques externes, droit ante
´rieur, Programme Sante
´et condition physique de
[Traduit par la Re
Well-developed abdominal musculature is important in
maintaining trunk and spine stability to reduce low back
pain and enhance athletic performance (Axler and McGill
1996; Escamilla et al. 2006b). A variety of exercises are
used to develop and test the abdominal musculature and its
contribution to trunk stability. Many modifications are made
to sit-up exercises in an attempt to maximize the activation
of abdominal muscles considered to contribute to trunk
stability and minimize the compressive forces on the lumbar
spine. Other modifications to the sit- or curl-up exercise are
made in an attempt to decrease the activation of the hip
flexors, which increase the risk of lower back injury if they
are over developed relative to the abdominal muscles
(Andersson et al. 1998; Axler and McGill 1997; Escamilla
et al. 2006a; Norris 1993). The modifications to the various
sit-up techniques include fixed or non-fixed feet, the degree
of knee flexion (from full extension to bent at 908), or the
distance the torso travels from the ground (often reflected
by the distance the hands are slid along the floor).
Different professional groups and associations recommend
Received 20 February 2008. Accepted 14 May 2008. Published
on the NRC Research Press Web site at apnm.nrc.ca on 10 July
K.C. Parfrey,1R.C. Workman, and D.G. Behm. School of
Human Kinetics and Recreation, Memorial University of
Newfoundland, St. John’s, NL A1C 5S7, Canada.
D. Docherty. School of Physical Education, University of
Victoria, Victoria, BC V8W 3P1, Canada.
1Corresponding author (e-mail: firstname.lastname@example.org).
Appl. Physiol. Nutr. Metab. 33: 888–895 (2008) doi:10.1139/H08-061 #2008 NRC Canada
different sit-up protocols or techniques for developing or
testing muscular fitness of the abdominal muscles. The
American College of Sport Medicine (ACSM) use a partial
sit-up test in which the knees are bent at 908without foot
fixation and the arms are fully extended at the side on the
mat with finger tips at the edge of a piece of tape. When
the age of the group being tested is over 45 years, the finger
tips travel 8 cm, whereas clients less than 45 years of age
slide their fingers tips 12 cm (Heyward 2002). Hoffman
(2006) recommends a bent-knee, fixed-foot sit-up with the
hands held behind the head and the endpoint identified as
the elbows touching the knees. The Canadian Society for
Exercise Physiology (CSEP) Health and Fitness Program
(Gledhill and Jamnik 2003) uses a protocol similar to that
of the ACSM, in which a bent-knee partial curl-up is per-
formed with arms extended at the side, finger tips at the
edge of a piece of tape, and without foot fixation. The
fingertips travel 10 cm along the floor. The test is ended if
the feet lose contact with the floor.
Foot fixation has been shown to increase the activation of
the hip flexors and hyperextend the lumbar spine during a
sit-up (Norris 1993). However, the literature that compares
fixed to non-fixed foot position and abdominal muscle acti-
vation is quite limited. Andersson et al. (1997) found no dif-
ference in abdominal muscle activation between the 2
conditions. However, they compared sit ups involving a dy-
namic movement, which may have compromised the validity
of the surface electromyography (EMG) readings (Deluca
The relationship between degree of knee flexion and acti-
vation of the hip flexors has been equivocal. The degree of
knee flexion has been found to have no effect (Juker et al.
1998), as well as decrease (Norris 1993) or increase
(Andersson et al. 1997; McGill 1995) activation of the hip
The ‘‘crunch’’ is an exercise in which the trunk is lifted
until the scapulae raise off the ground (Willett et al. 2001)
and has been shown to produce less activation of the hip
flexors than a full sit-up in which the entire back is lifted
off the floor (Andersson et al. 1997, 1998; Escamilla et al.
2006a, 2006b; Juker et al. 1998; Konrad et al. 2001; Warden
et al. 1999). It has also been found that performing a
‘‘crunch’’ produced high levels of activation of the rectus
abdominis (Escamilla et al. 2006a, 2006b), as well as the
external obliques (Escamilla et al. 2006a, 2006b; Konrad et
Previous research examining activation of the abdominal
muscles and hip flexors has measured the distance the trunk
travels through either a change in hip angle (Andersson et
al. 1997, 1998) or by a restraining rod (Piering et al. 1993;
Warden et al. 1999). The effect of the distance traveled by
the hands sliding along the floor on muscle activation has
not been examined despite this being a reliable and easy
method to administer and used in the CSEP Health and Fit-
ness Program abdominal endurance test (Gledhill and
The CSEP Health and Fitness program uses this test to as-
sess abdominal endurance, which is compositely used with
other tests to indicate the risk of low back pain. Validity for
the inclusion of the sit-up test in the CSEP Health and Fit-
ness program test battery has been established by the rela-
tionship between the number of partial curl ups performed
in a minute (to a maximum of 25) and the occurrence of
low back pain (Albert et al. 2001; Payne et al. 2000). How-
ever, to our knowledge, there was no inclusion of any stud-
ies that used EMG to quantify and identify the activation of
specific abdominal musculature and hip flexors.
The purpose of this study was, therefore, to investigate
the effect of several modifications of an exercise commonly
used to develop and test trunk stability on the activation of
the different abdominal musculature and hip flexors. Of par-
ticular interest was the activation of the various abdominal
musculature and hip flexors while performing the CSEP
Health and Fitness Program curl- or sit-up protocol. To
avoid the potential artifact on the EMG signal as a result of
movement (Deluca 1997), this study elected to monitor the
EMG signal during isometric held positions at different
trunk inclinations with the feet fixed or non-fixed and the
knees bent at 908or straight.
Materials and methods
A convenience sample of 14 male participants volun-
teered to participate for the study with a respective mean
age, height, and body mass of 24.8 (±7.4) y, 176.9
(±9.0) cm, and 78.9 (±13.0) kg. All participants were either
competitive rugby players or recreational athletes with no
known or apparent musculoskeletal injuries and able to per-
form the exercises correctly. All subjects were required to
read and sign a consent form before participation. Memorial
University’s Human Investigation committee approved the
Subjects participated in a familiarization session on a sep-
arate day prior to the experimental session where they at-
tempted and completed all variations of the sit-up protocol.
On the subsequent visit, subjects were instructed to lie flat
on a horizontal bench and fitted unilaterally with surface
electrodes on the upper rectus abdominis (URA), the lower
rectus abdominis (LRA), the external obliques (EO), the
lower abdominal stabilizers (LAS), the rectus femoris (RF),
and the biceps femoris (BF) muscles.
Once the subject was in the appropriate position (knee an-
gle, extent of foot fixation, and prescribed distance of the
finger tips) for the specific sit-up technique, he was then di-
rected to perform the action to the desired position. The sub-
ject was permitted 3 s to establish the correct position and
hold for at least 2 s, during which time the EMG recording
was taken and used for analysis. If the investigator noted
that the subject took longer than 3 s to get to a stable iso-
metric position, data collection was delayed until the subject
was stable. There were 12 conditions tested: 3 hand posi-
tions (5, 10, and 15 cm) 2 knee positions (bent and
extended) 2 stabilization conditions (feet fixed or not
As there are various interpretations of sit ups, curl ups,
and crunches in the literature, this study offers the following
definitions. The ‘‘crunch’’ is an exercise in which the trunk
is lifted until the scapulae raise off the ground (Willett et al.
2001), which in this study corresponds to the 5 cm move-
Parfrey et al. 889
#2008 NRC Canada
ment of the hands. Sit ups and curl ups are used synonym-
ously in this study to indicate a movement whereby the
scapulae and trunk are elevated off the surface with the feet
remaining in constant contact with the surface. The sit up
involving a 10 cm movement of the hands in the present
study corresponds to the CSEP Health and Fitness Program
partial curl up (Gledhill and Jamnik 2003), whereas the
15 cm hand movement sit up corresponds to the bent knee
sit ups illustrated by Juker et al. (1998).
All surface electrodes (Meditrace 130 ECG Conductive
Adhesive Ag–AgCl Electrodes, Tyco Healthcare Group LP,
Mansfield, Mass.) were placed on the right side of the body
on 6 different muscle sites. To reduce resistance of the sig-
nal, all sites for electrode placement were shaved, scrubbed
with sand paper, and rubbed with an alcohol-soaked paper
towel. This process removed body hair, dead skin cells, and
oils. All electrodes were placed parallel to the muscle fibres,
with an inter-electrode difference of 2 cm. The URA was
identified as the midpoint between the xiphoid process and
the umbilicus and 3 cm to the right of the linea alba. The
LRA area was identified as 3 cm to the right of the linea
alba and perpendicular to the iliac crest. The electrodes for
the EO were placed 5 cm superior to the iliac crest and at
an oblique angle (~458). Electrodes for LAS (an area that
encompassed the stabilizing muscles of the transversus ab-
domonis and internal obliques) were placed immediately
medial to the anterior superior iliac spine, a site used in
previous published studies from this laboratory (Anderson
and Behm 2005; Behm et al. 2005, 2006; Hamlyn et al.
2007) The BF electrode location was the mid point between
the gluteal fold and the popliteal space and the electrode
placement for RF was immediately distal to the anterior
superior iliac spine and inferior to the inguinal ligament.
The RF site has been shown to provide reliable and valid
estimates of the EMG activity of the iliopsoas muscle
group (McGill et al. 1996). All ground electrodes were
placed on the nearest bony prominence for each pair of ac-
All EMG signals were collected over a 5 s period,
sampled at 2000 Hz, with a Blackman –61 band-pass filter
between 10–500 Hz, and amplified (500) (Biopac Systems
MEC bi-polar differential 100 amplifier, Santa Barbara,
Calif.; input impedance = 2MU, common mode rejection
ratio > 110 db min (50/60 Hz), noise > 5 UV). EMG activity
was then directed through a 12 bit analog-to-digital con-
verter (Biopac MP 100) and stored on a computer (Sona,
St. John’s, Nfld.). The data were later transferred to a per-
sonal computer for further analysis. The EMG signal was
rectified and integrated over the final 2 s of the 5 s iso-
metric contraction. An average of the 2 rectified and inte-
grated trials was used for statistical analysis. Because the
focus of this study was on changes in activation of indi-
vidual muscles caused by different exercises, and not be-
tween muscles or subjects, normalization of the EMG
signal to a maximal voluntary contraction was considered
unnecessary. The study was a repeated-measures design
that was completed in a single experimental session (no
change in electrode position), therefore absolute data were
The study had 3 independent variables: the distance the
hand traveled during the trunk position or inclination based
on hand positions (5, 10, or 15 cm), holding the sit-up posi-
tion with the knees bent at 908or legs fully extended (1808),
and holding the sit-up with feet fixed or non-fixed. Each
subject performed the 12 different exercise conditions (3
hand conditions 2 knee positions 2 stabilization condi-
tions) at least twice. For each sit-up position the participant
was instructed to keep both arms extended by their side. The
tips of the subject’s fingers on their left hand were placed on
a ruler. As the subject performed a sit-up the fingers slid
along the ruler indicating the distance traveled (extent of
trunk flexion). A randomly predetermined distance of 5, 10,
or 15 cm was chosen prior to each sit-up and set with a
stopping block. On the investigator’s command, the subject
performed the sit up, stopped, stabilized once he hit the
stopping block, and held the position for 5 s.
Prior to each new condition the subject was instructed to
either bend his knees to 908or to fully extend his knees
(1808). The angle of the knee was measured with a gonio-
meter. The subject was instructed to keep his heels in con-
tact with the bench during the entire isometric contraction
for both conditions. If the subject’s heels lifted off the bench
the data were discarded and the subject performed another
When the condition required foot fixation the feet of the
subject were held by one of the investigators. The investiga-
tor held the lower legs with the fingers just superior to the
lateral maleoli. When the sit up was performed with feet in
the fixed condition the subject was allowed to use the resist-
ance provided by the investigators to aid in trunk flexion
and obtain the required position. When holding the non-
fixed sit-up position the subject was not allowed to lift his
heels off the bench.
All subjects were given a verbal and visual demonstration
of each sit-up condition and allowed to practice the condi-
tion. Incorrect procedures such as excessive shoulder pro-
traction and retraction were monitored by the investigators.
Each subject performed 12 different exercise conditions at
least twice. A third or fourth sit up was performed if one of
the investigators decided the participant performed the sit up
incorrectly. The order of sit-ups that were performed was
randomized. Thirty seconds rest was allocated between each
type of sit up. Subjects unanimously agreed that fatigue was
not a factor throughout the testing period.
A series of separate 2-way analysis of variance (ANOVA)
repeated measures for each muscle site was used to test for
differences in activation levels for each condition. When
statistical significance was found, the Bonferroni–Dunns
post hoc test was used to reveal the differences between
conditions. Effect sizes (ES is calculated by dividing the
mean change by the standard deviation of the sample scores)
were also calculated and reported (Cohen 1988). Qualitative
descriptors were allocated for the effect sizes with ratios of
0.35–0.80, 0.80–1.5, and >1.5 indicating small, moderate,
and large effects, respectively (Rhea 2004). Descriptive sta-
tistics include means ± standard deviation (SD).
890 Appl. Physiol. Nutr. Metab. Vol. 33, 2008
#2008 NRC Canada
Main effect for feet fixed vs non-fixed
Sit-up positions in which the feet were non-fixed com-
pared with fixed resulted in 27.9%, 19.2%, 27.8%, 22.7%,
and 55.1% significantly greater EMG activation of the EO
(p= 0.01; ES = 0.27), LAS (p< 0.0001; ES = 0.41), LRA
(p< 0.001; ES = 0.35), URA (p= 0.000; ES = 0.36), and
BF (p= 0.03; ES = 0.40), respectively (Fig. 1). Only the
RF had significantly (p< 0.0001; ES = 0.91) less activity
(42.1% less) when the feet were non-fixed.
Main effect for sit-up distance
Sliding the hands 10 cm and holding this position resulted
in greater activation of the EO and LRA muscle groups. A
10 cm position provided 9.9% (ES = 0.13) and 16.5%
(ES = 0.21) significantly (p= 0.02) greater LRA activation
than the positions in which the hands were slid 5 and 15 cm,
respectively. There was a trend (p= 0.1) towards greater EO
activation with 29.1% (ES = 0.22) higher activation holding
the 10 cm position compared with the 5 cm position. The
position at 15 cm provided 24.1% (ES = 0.47) and 27%
(ES = 053) less EMG activity for the URA compared with
the 10 and 5 cm positions, respectively (p= 0.001). The RF
was 37.8% (ES = 0.47) and 42.7% (ES = 0.59) significantly
(p= 0.001) less activated at the 5 cm position than at 10 and
15 cm, respectively (Fig. 2).
Main effect for knee position
There were no significant differences in muscle activation
in regard to knee position. However, the LAS tended to
have a 19.4% (ES = 0.35) higher level of activation when
the knees were flexed (p= 0.1), whereas the RF showed
21.4% (ES = 0.54) more activity when the knees were ex-
tended (p= 0.1) (Fig. 3).
Feet fixation distance interaction
The lowest activation of the URA occurred when the po-
sition involved fixed feet and the hands slid 15 cm. A 15 cm
hand position with fixed feet resulted in 22.2% (ES = 0.43)
and 39.2% (ES = 0.97) less activity than when the hands
were held at 10 and 5 cm, respectively (p< 0.0001). Com-
pared with the non-fixed sit-up position, a fixed 15 cm posi-
tion resulted in 30.3% (ES = 0.66), 48.8% (ES = 1.44), and
40.9% (ES = 1.04) less URA activity than the non-fixed 15,
10, and 5 cm positions, respectively (p< 0.0001). A non-
fixed 10 cm position provided the greatest URA activity,
26.6% (ES = 0.45), 15.9% (ES = 0.27), 34.3% (ES = 0.59),
Fig. 1. The effect of foot fixation during a sit up on electromyo-
graphic (EMG) activity of the upper rectus abdominis (URA),
lower rectus abdominis (LRA), lower abdominal stabilizer (LAS),
external obliques (EO), rectus femoris (RF), and biceps femoris
(BF). Asterisks (*) indicate significant difference (p£0.05) be-
tween the two conditions at the same muscle site.
Fig. 2. The effect of travelling different distances during a sit up on
electromyographic (EMG) activity of the upper rectus abdominis
(URA), lower rectus abdominis (LRA), lower abdominal stabilizer
(LAS), external obliques (EO), rectus femoris (RF), and biceps fe-
moris (BF). Asterisks (*) indicate significant difference (p£0.05)
between the 3 conditions at the same muscle site.
Fig. 3. The effect of alterations in knee position on electromyo-
graphic (EMG) of the lower abdominal stabilizers (LAS). URA,
upper rectus abdominis; LRA, lower rectus abdominis; EO, external
obliques, RF, rectus femoris; BF, biceps femoris.
Parfrey et al. 891
#2008 NRC Canada
and 48.8% (ES = 0.84) higher than with non-fixed feet at
15 cm and fixed feet at 5, 10, and 15 cm positions, respec-
tively (p< 0.0001). There were no significant URA activity
differences between non-fixed sit-up position with the hands
at 5 or 10 cm (Fig. 4).
Similarly, the least activation of the LRA occurred with
the fixed 15 cm sit-up position. The fixed-feet 15 cm sit-up
position produced 29.7% (ES = 0.34), 42.2% (ES = 0.48),
70.7% (ES = 0.81), 94.7% (ES = 1.09), and 49.9% (ES =
0.57) less EMG activity than the fixed 10 and 5 cm posi-
tions, as well as the non-fixed sit-up positions held at 15,
10, and 5 cm, respectively (p< 0.0001). The greatest activa-
tion of the LRA also occurred in the non-fixed 10 cm sit-up
position. The non-fixed 5 cm and fixed 5, 10, and 15 cm sit-
up positions had 22.9% (ES = 0.30), 26.9% (ES = 0.36),
33.3% (ES = 0.44), and 48.6% (ES = 0.64) less activity
than the non-fixed 10 cm sit-up position (p< 0.0001). There
were no significant LRA activity differences between non-
fixed sit-up positions at 10 or 15 cm (Fig. 5).
Feet fixation knee position interaction
Significantly (p< 0.0001) higher activation of the LAS
occurred in the non-fixed, flexed- knee, sit-up position. The
non-fixed extended knee, fixed flexed knee, and fixed ex-
tended knee positions produced 23.1% (ES = 0.37), 22.3%
(ES = 0.43), and 38.2% (ES = 0.76) less EMG activity, re-
spectively, than the non-fixed flexed knee sit-up position
There were no significant findings for knee position
distance interaction or knee position foot fixation dis-
This study compared 12 variations of the sit up performed
by most healthy people. The most important finding of this
study was that the 10 cm, non-fixed, bent-knee sit-up posi-
tion produced the highest activation levels in 3 of the ab-
dominal muscle sites monitored. Activation of the hip
flexors (RF) was lowest in the 5 cm non-fixed position, but
this position did not produce high levels of activation of the
abdominal muscles and may, therefore, have less effect on
stabilizing the trunk.
In agreement with Norris (1993), a sit-up with fixed feet
produced significantly higher RF activation levels. McGill
et al. (1996) showed that activity of the RF is a valid and
reliable indicator of the hip flexor activity, especially the
iliopsoas. However, Norris (1993) did not mention how
fixed feet affected activation levels of the abdominal
muscles compared with non-fixed feet. This study found
that a sit-up position with non-fixed feet increased activation
levels of the URA, LRA, LAS, and EO. These data conflict
with Andersson et al. (1997), who found that foot fixation
did not affect abdominal muscle activity; however,
Andersson et al. (1997) compared dynamic movements,
Fig. 4. The effect of foot fixation and distance traveled on electro-
myographic (EMG) activity of the upper rectus abdominis during a
sit up. A single asterisk (*) indicates a significant difference (p£
0.05) compared with all other conditions. Double asterisks (**) in-
dicate a significant difference compared with all other conditions
except the non-fixed feet, 5 cm sit up.
Fig. 5. The effect of foot fixation and distance traveled on electro-
myographic (EMG) activity of the lower rectus abdominis during a
sit up. A single asterisk (*) indicates a significant difference (p£
0.05) compared to all other conditions. Double asterisks (**) indi-
cate a significant difference compared to all other conditions except
the non-fixed feet, 15 cm sit up.
Fig. 6. The effect of alterations in foot fixation and knee position
on electromyographic (EMG) activity of the lower abdominal sta-
bilizers. Asterisk (*) indicates a significant difference (p£0.05)
compared with all other conditions. FFBK, feet fixed with bent
knee; FFEK, feet fixed with extended knee; NFBK, non-fixed feet
with bent knee; NFEK, non-fixed feet with extended knee.
892 Appl. Physiol. Nutr. Metab. Vol. 33, 2008
#2008 NRC Canada
whereas the current study used isometric contractions. In the
present study, the contractions were held at prescribed dis-
tances, whereas the increased variability associated with
monitoring the EMG of the muscles in which the fibre
length is changing, as was the case with Andersson et al.
(1997), may explain the differences in the findings. The
combination of increased activation levels for the hip flexors
and decreased levels for the abdominal muscles may predis-
pose an individual to lower back injuries (Norris 1993). This
type of exercise might over-develop the hip flexor muscle
group, pulling the pelvis into an anterior tilt increasing lord-
osis, and placing increased compressional forces on the lum-
bar spine, particularly at the L4–L5 vertebral levels
(Escamilla et al. 2006a). Axler and McGill (1997) reported
that a curl-up with feet not fixated had the lowest compres-
sional loads on the spine of any abdominal exercise tested in
their study, making it a good and practical choice for health
and fitness assessment. With a decrease in abdominal activa-
tion caused by foot fixation the abdominal musculature
might not develop at the same rate as the hip flexors, pro-
ducing a muscle imbalance and subsequent reduced spine-
stabilizing capabilities (Konrad et al. 2001). The increased
lordosis coupled with decreased training of the abdominal
musculature may increase the possibility of lower back in-
Bending the knees while performing a sit up is often pre-
scribed to reduce the contribution of the hip flexors during
the exercise (Norris 1993). Norris suggests that having the
knees flexed to 908would reduce the tension by 40%–50%
of its maximum. One of the few studies examining this as-
sumption found that performing a sit up with the legs straight
produced lower levels of hip flexor activity compared with a
bent-knee sit up (Andersson et al. 1997). They suggested that
the higher activation levels were due to the decreased force-
producing capabilities from the shortened length of the hip
flexors in the bent-knee position. However, Juker et al.
(1998) found no significant difference in the activation levels
of the hip flexors between bent-knee and straight-leg condi-
tions. The present study supports the findings of Juker et al.
(1998), but did find a trend for the knee-flexed position to
produce 21.4% less RF activity than the knee-extended posi-
tion. Norris (1993) contends that bending the knees during a
sit up will shorten the hip flexors, which will decrease the
tension in the muscle and its subsequent activation. It appears
more research is needed to clarify the activation of the hip
flexors as it relates to the position of the legs.
This study showed a trend that the bent-knee sit-up posi-
tion produced higher activation (19.4%; p= 0.1) of the LAS
than the straight-leg sit-up position. This trend disagrees
with previous research that found no difference in any of ab-
dominal muscle sites when bent-knee and straight-leg sit ups
were compared (Andersson et al. 1997; McGill 1995).
McGill (1995) and Andersson et al. (1997) provided 2
reasons why there would be no change in abdominal muscle
activation when comparing a bent- and straight-leg sit up.
First, bending the knees produces a decrease in hip flexor
length and should, therefore, affect only the hip flexors and
not the abdominal muscle group (Andersson et al. 1997).
McGill (1995) took a biomechanical view and stated that a
knee bend had no effect on torso mechanics and, therefore,
should not affect levels of activation. However, the present
study agreed with Juker et al. (1998) who found a trend
(p= 0.1) for higher levels of activation for the LAS during
the bent-knee sit up compared with the straight-leg sit up.
With the knees in a bent position and the trunk raised off
the ground, the base of support for the body would be much
smaller than if the legs were extended. This decreased base
of support would decrease stability and consequently in-
crease the work required by the LAS group.
The final variable examined in this study was the distance
covered by the hands as the trunk was elevated during a sit-
up procedure. In previous research, the height of the sit up
has been determined by hip angle (Andersson et al. 1997,
1998), a restraining rod (Piering et al. 1993; Warden et al.
1999), or the subjects performing the exercise with verbal
instructions (Clark et al. 2003; Escamilla et al. 2006b; Juker
et al. 1998; Konrad et al. 2001; Lehman and McGill 2001;
Sarti et al. 1996). No studies have used the distance covered
along the floor as a means of controlling the height the trunk
is raised off of the floor. This technique was chosen for this
study because the testing protocol for the CSEP Health and
Fitness program abdominal endurance test uses distance and
not angle or a restraining rod to establish how high the trunk
is raised. This study used the 5 cm sit-up distance to reflect
a ‘‘crunch’’, the 10 cm distance because it is the CSEP
Health and Fitness program curl- or sit-up protocol, and the
15 cm position as a ‘‘full sit up’’ for comparative purposes.
The ‘‘crunch’’ has been shown to minimize hip flexor ac-
tivity when compared with the full sit up (Andersson et al.
1997, 1998; Escamilla et al. 2006a, 2006b; Juker et al.
1998; Konrad et al. 2001; Warden et al. 1999). However, it
has also been found to induce greater levels of URA and
LRA (Escamilla et al. 2006a, 2006b) activity, whereas the
full sit up produces more EO activity (Escamilla et al.
2006a, 2006b; Konrad et al. 2001).
The present study showed a higher activation level for the
RF (37.8%) and a trend of higher levels of EO activity (p=
0.1) for the 10 cm sit-up position than for the 5 cm
‘‘crunch’’, which is in agreement with previous research
(Escamilla et al. 2006b; Konrad et al. 2001). However, in
contrast to some previous research (Escamilla et al. 2006a,
2006b), the present study showed no difference for the URA
when the 5 cm crunch and 10 cm sit-up positions were com-
pared, but it did show that the LRA produced 9.9% and
16.5% more activity in the 10 cm sit-up position than in the
5 cm crunch and 15 cm sit-up positions, respectively.
The 10 cm sit-up position elicited a significant increase in
the LRA, but not in the URA; this suggests that certain ex-
ercises may stimulate separate sections of the RF. Such a
finding is in agreement with previous research that com-
pared the URA and LRA to examine if it is possible to se-
lectively activate different portions of the rectus abdominis
(Sarti et al. 1996; Warden et al. 1999; Willett et al. 2001).
Anatomically, there is segmental innervation of the antero-
lateral abdominal musculature from the ventral rami of the
lower 6 or 7 thoracic nerves. In addition, the rectus abdomi-
nus is separated into different sections by its tendonous in-
scriptions, which should increase its ability to segmentally
contract (Clark et al. 2003; Sarti et al. 1996). However,
from a biomechanical standpoint, this may not be possible.
The rectus abdominis is a strap-like muscle. Superiorly, the
rectus abdominis is attached to the 5th, 6th, and 7th ribs, as
Parfrey et al. 893
#2008 NRC Canada
well as to the xiphoid process. Inferiorly, it is attached to the
crest of the pubis, the pubic tubercle, and the front of the
pubic symphysis. When the rectus abdominis contracts it
should, with an equal and balanced force, pull on all inser-
tion points (Clark et al. 2003), thereby making segmental
contraction improbable. Clark et al. (2003), Lehman and
McGill (2001), and Piering et al. (1993) have all shown that
the rectus abdominis lacks the ability to selectively activate
its different segments. Although the participants who volun-
teered for this study were not elite athletes they were a gen-
erally well-trained group and may have had an increased
ability to selectively activate individual sections of their rec-
tus abdominis (Lehman and McGill 2001).
The lowest levels of activation for URA in this study oc-
curred when the feet were fixated in the 15 cm sit-up posi-
tion. This sit-up position also provided the highest levels of
activation for the RF. This sit up was considered the ‘‘full
sit-up’’ and probably had the lowest level of activation be-
cause the exercise required holding the torso in a position
where the resistive torque of the trunk was at its lowest of
the 3 held positions. This finding concurs with Andersson
et al. (1997) who found that abdominal activation decreased
after the subject reached a hip-to-ground angle of 308.
In the present study, the highest levels of activation came
from the 10 cm sit-up position with non-fixed feet and bent
knees. This is the type of sit-up used in the trunk endurance
test for the CSEP Health and Fitness program protocol
(Gledhill and Jamnik 2003). This study showed that the
CSEP Health and Fitness program abdominal strength and
endurance test provides high activation of the abdominal
musculature with minimal activation of the hip flexors.
However, it is acknowledged that the 5 cm position–exercise
is possibly a good option for individuals with low back pain,
in that it would seem to minimize the compression on the
lower back and still provide good activation of the abdomi-
The findings of this study have practical applications
when selecting a test of abdominal performance or prescrib-
ing an abdominal exercise program. The condition of the in-
dividual and health of the individual’s back will influence
which exercise should be performed. If there is history of
low back pain or recovery from lower back injury, the indi-
vidual should be informed to proceed no farther than 5 cm
for this will limit hip flexor activity and increase abdominal
stability. If the subject is healthy and seeking a more intense
exercise, the individual can implement greater trunk flexion
(i.e., equal to a 10 cm movement of the hands along the
floor). It should be recommended that the subject not per-
form a full sit up (i.e., equal to a 15 cm movement of the
hands along the floor), because the final stage of this type
of sit-up exercise seems to target the hip flexors. Going be-
yond a 10 cm position of the hands or raising the trunk to an
angle greater than 308may provide a rest period (reduced
trunk muscle activation) for the abdominal muscles, but the
hip flexors may remain activated. Foot fixation during ab-
dominal exercise should be avoided and other strategies
used to increase the strength and endurance of the trunk
flexors. Although this study did not find significant effects
for knee position on muscle activation, it did find a trend to-
wards greater activation of the abdominal musculature and
lower RF activation when the knees were bent. It must be
kept in mind that the present study analyzed EMG activity
during an isometric contraction, which may not fully repre-
sent the responses during a dynamic contraction.
Albert, W.J., Bonneau, J., Stevenson, J.M., and Gledhill, N. 2001.
Back fitness and back health assessment considerations for the
Canadian physical activity, fitness and lifestyle appraisal. Can.
J. Appl. Physiol. 26: 291–317. PMID:11441232.
Anderson, K., and Behm, D.G. 2005. Trunk muscle activity in-
creases with unstable squat movements. Can. J. Appl. Physiol.
30: 33–45. PMID:15855681.
Andersson, E.A., Nilsson, J., Ma, Z., and Thorstensson, A. 1997.
Abdominal and hip flexor muscle activation during various
training exercises. Eur. J. Appl. Physiol. 75: 115–123. doi:10.
Andersson, E.A., Ma, Z., and Thorstensson, A. 1998. Relative
EMG levels in training exercises for abdominal and hip flexor
muscles. Scand. J. Rehabil. Med. 30: 175–183. doi:10.1080/
Axler, C.T., and McGill, S.M. 1997. Low back loads over a variety
of abdominal exercises: searching for the safest abdominal chal-
lenge. Med. Sci. Sports Exerc. 29: 804–811. PMID:9219209.
Behm, D.G., Leonard, A.M., Young, W.B., Bonesy, W.A., and
Mackinnon, S.N. 2005. Trunk muscle electromyographic activity
with unstable and unilateral exercises. J. Strength Cond. Res. 19:
Behm, D.G., Burry, S.M., Greeley, G.E.D., Poole, A.C., and
MacKinnon, S.N. 2006. An unstable base alters limb and ab-
dominal activation strategies during the flexion-relaxation re-
sponse. J. Sports Sci. Med. 5: 323–332.
Clark, K.M., Laurence, E.H., and Sinyard, J. 2003. Electromyographic
comparison of the upper and lower rectus abdominis during abdom-
inal exercises. J. Strength Resistance Cond. 17: 475–483. doi:10.
Cohen, J. 1988. Statistical power analysis for the behavioral
sciences. 2nd ed. L. Erlbaum Associates, Hillsdale, N.J.
Deluca, C.J. 1997. The use of surface electromyography in bio-
mechanics. J. Appl. Biomech. 13: 135–163.
Escamilla, R.F., Babb, E., Dewitt, R., Jew, P., Kelleher, P., and
Burnham, T. 2006a. Electromyographic analysis of traditional
and nontraditional abdonminal exercises: implications for
rehabilitation and training. Phys. Ther. 86: 656–671.
Escamilla, R.F., McTaggart, M.S.C., Fricklas, E.J., Dewitt, R.,
Kelleher, P., and Taylor, M.K. 2006b. An electromyographic
analysis of commercial and common abdominal exercises: im-
plications for rehabilitation and training. J. Orthop. Sports Phys.
Ther. 36: 45–57. PMID:16494072.
Gledhill, N., and Jamnik, V. (editors). 2003. The Canadian physical
activity, fitness and lifestyle approach. 3rd ed. Canadian Society
for Exercise Physiology. Ottawa, Ont. pp. 741–742.
Hamlyn, N., Behm, D.G., and Young, W.B. 2007. Trunk muscle ac-
tivation during dynamic weight-training exercise and isometric
instability activities. J. Strength Cond. Res. 21: 1108–1112.
Heyward, V.H. 2002. Advanced fitness assessment and exercise
prescription. 4th ed. Edited by M.S. Bahrke. Human Kinetics,
894 Appl. Physiol. Nutr. Metab. Vol. 33, 2008
#2008 NRC Canada
Hoffman, J. 2006. Norms for fitness, performance, and health. Edi-
ted by M.S. Bahrke. Human Kinetics, Windsor, Ont.
Juker, D., McGill, S., Kpopf, P., and Steffen, T. 1998. Quantitative
intramuscular myoelectric activity of lumbar portions of psoas
and the abdominal wall during a wide variety of tasks. Med.
Sci. Sports Exerc. 30: 301–310. PMID:9502361.
Konrad, P., Schmitz, K., and Denner, A. 2001. Neuromuscular eva-
luation of trunk-training exercises. J. Athl. Train. 36: 109–118.
Lehman, G.J., and McGill, S.M. 2001. Quantification of the differ-
ences in electromyographic activity magnitude between upper
and lower portions of the rectus abdominis muscle during selected
trunk exercises. Phys. Ther. 81: 1096–1101. PMID:11319934.
McGill, S.M. 1995. The mechanics of torso flexion: sit-ups and
standing dynamic flexion maneuvers. Clin. Biomech. (Bristol,
Avon), 10: 184–192. doi:10.1016/0268-0033(95)91396-V. PMID:
McGill, S.M., Juker, D., and Kropf, P. 1996. Appropriately placed
surface EMG electrodes reflect deep muscle activity (psoas,
quadratus lumborum, abdominal wall) in the lumbar spine. J.
Biomech. 29: 1503–1507. doi:10.1016/0021-9290(96)84547-7.
Norris, C.M. 1993. Abdominal muscle training in sport. Br. J.
Sports Med. 27: 19–27. PMID:8457806.
Payne, N., Gledhill, N., Katzmarzyk, P.T., and Jamnik, V. 2000.
Health-related fitness, physical activity, and history of back
pain. Can. J. Appl. Physiol. 25: 236–249. PMID:10953062.
Piering, A.W., Janowski, A.P., Moore, M.T., Snyder, A.C., and
Wehrenberg, W.B. 1993. Electromyographic analysis of four
popular abdominal exercises. J. Athl. Train. 28: 120–126.
Rhea, M.R. 2004. Determining the magnitude of treatment effects
in strength training research therough the use of the effect size.
J. Strength Cond. Res. 18: 918–920. doi:10.1519/14403.1.
Sarti, M.A., Monfort, M., Fuster, M.A., and Villaplana, L.A. 1996.
Muscle activity in upper and lower rectus abdominus during ab-
dominal exercises. Arch. Phys. Med. Rehabil. 77: 1293–1297.
Warden, S.J., Wajswelner, H., and Bennell, K.L. 1999. Comparison
of abshaper and conventionally performed abdominal exercises
using surface electromyography. Med. Sci. Sports Exerc. 31:
1656–1663. doi:10.1097/00005768-199911000-00024. PMID:
Willett, G.M., Hyde, J.E., Uhrlaub, M.B., Wendel, C.L., and Karst,
G.M. 2001. Relative activity of abdominal muscles during
commonly prescribed strengthening exercises. J. Strength Cond.
Res. 15: 480–485. doi:10.1519/1533-4287(2001)015<0480:
Parfrey et al. 895
#2008 NRC Canada