Intensive Abdominal Drawing-In Maneuver After Unipedal Postural Stability in Nonathletes With Core Instability

A convenience sample of 19 nonathletes with core instability (4 women: age = 22.3 ± 1.3 years, height = 164.0 ± 1.7 cm, mass = 56.0 ± 4.6 kg; 15 men: age = 24.6 ± 2.8 years, height = 172.6 ± 4.7 cm, mass = 66.8 ± 7.6 kg) participated in this study (Table 1). The sample size was estimated based on a power of 80% at large differences (0.8) in effect size between the tests. Participants who were unable to successfully perform the formal test 3 consecutive times at baseline were included. However, participants were excluded if they had any neurologic or musculoskeletal impairment that could affect performance during the experimental tests. Initially, we recruited 20 participants, but we excluded 1 participant at the pretest because he did not meet the criterion of core instability, which was manifested in the formal test. All participants provided written informed consent, and the experimental protocol was approved by the Yonsei University College of Health Science Human Studies Committee.

A formal test was used to determine the participants' core stability in the prone position. The ADIM-contributed core-stabilization practice required the participants to lie in a prone position. A pressure biofeedback unit (PBU; Chattanooga Group Inc, Hixson, TN) was placed under the lower abdomen with the lower edge in line with the anterior-superior iliac spine and inflated to 70 mm Hg (Figure 1).^18,27,28  According to the ADIM technique used by Richardson et al,^15,18  each participant was instructed to draw in the lower abdomen below the navel gently and gradually, without moving the upper abdomen, while maintaining a neutral pelvic position. The formal test was considered successful when the participant met the following criteria: (1) performed the ADIM while reducing the pressure by approximately 4 to 10 mm Hg and maintaining the target pressure level monitored in the PBU (Figure 2)^18,27,28  and (2) performed the ADIM without substitutions, such as excessive contraction of the EO and ES muscles (<15% reference voluntary contraction [RVC]),¹⁵  as evidenced by electromyography or any evasive movements (ie, pelvic rotation, lumbar arching, shoulder elevation, or upper chest expansion).¹³  If a participant met only 1 of these criteria, the test was defined operationally as a failure or core instability. This definition was derived conceptually from previous studies.^{13,15,18,27,28} 

All participants underwent a pretest followed by intensive ADIM training and a posttest after training. The pretest and posttest measurements included changes in unipedal postural-stability test with the ADIM using RUSI, electromyography, and Matscan (Tekscan Inc, Boston, MA) COP measurements. The dominant foot was determined through the implementation of the single-legged–stance balance test, whereby the participant was instructed to maintain balance as stably as possible, and the more stable side was considered dominant.²⁹  To perform the unipedal postural-stability test, the participant stood on the dominant foot with the upper extremities across the chest and eyes closed for 10 seconds while maintaining equilibrium as stably as possible. The trial was repeated if the participant used the free lower extremity to touch either the ground or the stance extremity or if the upper extremities came off the chest. Participants initially performed 3 practice trials followed by 5 consecutive test trials, which were used for data analysis.

A real-time brightness-mode ultrasound apparatus (model SonoAce X8; Medison Inc, Seoul, South Korea) was used to assess changes in abdominal muscle (TrA, IO, and EO) thickness and to provide visual biofeedback about muscle contraction. We used a 4.5-cm linear transducer (Medison Inc) with a frequency of 10 MHz.¹⁸  Participants were instructed to lie prone. The examiner palpated the inferior borders of the rib cage and iliac crest to provide anatomic reference points for the abdominal muscles. Ultrasound gel (Aquasonic 100; Parker Inc, Orange, NJ) then was applied to the transducer head, which was placed in a transverse orientation on the anterolateral aspect of the abdomen halfway between the iliac crest and the inferior border of the rib cage (12th rib).^30,31  The angle was manipulated until the sharpest possible images of all abdominal muscles were obtained.³²  All images were captured and stored immediately at the end of the expiration phase, and muscle thickness was determined with an onscreen caliper by 1 investigator (N.G.L.). A vertical reference line positioned 1 cm from the myofascial junction of the TrA muscle was used to measure the TrA, IO, and EO muscle thicknesses (Figure 2).³¹  The myofascial boundaries were defined as the target layers of the hypoechoic area where the heterogeneous boundaries were manifested in contrasting pixels from dark to light.^32,33  We calculated the contraction ratios of each abdominal muscle and the TrA preferential contraction ratio to identify changes in the relative muscle thicknesses compared with the relaxed state using the following equations: (1) contraction ratio of each abdominal muscle (TrA, IO, and EO) = abdominal muscle contracted/relaxed, and (2) TrA preferential contraction ratio = (TrA contracted/[TrA + IO + EO contracted]) – (TrA relaxed/[TrA + IO + EO relaxed]).³²  The pretest scanning location was marked on a transparent sheet to ensure identical placement throughout the experiment and in the posttest. The anatomic reference locations for the iliac crest and 12th rib were palpated to identify and mark their locations with a permanent marker. Next, we superimposed the transparent sheet over these locations and made corresponding markings on it with the permanent marker.

We used a TeleMyo 2000 EMG system (Noraxon USA, Inc, Scottsdale, AZ) to record the mean values for the EO, ES, gluteus medius, vastus medialis oblique, tibialis anterior, and medial gastrocnemius muscles.^34,35  These electromyographic data were used to ensure proper muscle activation and provide visual feedback during the ADIM training. The skin was prepared carefully to reduce skin impedance to less than 5 kΩ by dry shaving hair with a disposable razor and cleansing the skin with a 2% alcohol swab. When the skin was dry, we placed pregelled bipolar Ag/AgCl disposable electrodes (Bio-Protech Inc, Wonju, South Korea) with a contact length and width of 36 × 23 mm over the EO, ES, gluteus medius, vastus medialis oblique, tibialis anterior, and medial gastrocnemius muscle bellies³⁶  on the side of the body with the weight-bearing limb during the unipedal postural stability test at an interelectrode distance of 2.0 cm. A reference electrode was positioned over the anterior-superior iliac spine. Two representative peak electromyographic amplitude values out of the 5 trials were obtained at the maximal expiratory volume during the unipedal stance position by monitoring with a spirometer (TECKSCIENCE Inc, Seoul, South Korea),¹⁵  and the middle 0.35-second electromyographic signals were averaged to provide a stable reference value. We used this reference anchor to calculate the percentage of the RVC for normalization of the aforementioned muscles. The raw electromyographic signal data were recorded at a sampling frequency of 1500 Hz and processed with a 60-Hz notch filter using MyoResearch software (version 1.06; Noraxon USA, Inc, Scottsdale, AZ) to reduce the noise associated with electrical interference, including 60-Hz power lines or radio frequencies and electric or magnetic devices. The root mean square electromyographic amplitude for each EO, ES, gluteus medius, vastus medialis oblique, tibialis anterior, and medial gastrocnemius muscle was computed for 10 seconds during the ADIM. The raw electromyographic signal was full-wave rectified and bandpass filtered at 15 to 500 Hz.³⁷ 

The center-of-pressure (COP) measurement was used to determine the COP sway measures: anteroposterior COP (COP_AP) and mediolateral COP (COP_ML) sway velocities (cm/s), COP_AP and COP_ML standard deviations (SDs; cm), and COP sway areas (cm²). The COP_AP and COP_ML sway velocities were the mean values of the instantaneous velocity of the COP in a given direction during a given time. The COP_AP and COP_ML SDs were the total SDs of the sway during a given time for a given number of trials. The COP sway area represented the maximal anterior, posterior, medial, and lateral sways during a given time.³⁸  Before data acquisition, automatic calibration was implemented based on each participant's body weight. The COP data were acquired at a sampling rate of 40 Hz for 10 seconds while the participant maintained equilibrium as stably as possible during the unipedal postural-stability test.³⁹  The raw COP data that we collected were converted to ASCII using F-Scan (version 6.0; Tekscan Inc, Boston, MA) software. We used MATLAB (version R2008A; The MathWorks, Inc, Natick, MA) to compute the COP data using the formulae provided in Table 2.³⁸ 

Before the intervention, we established the intrarater reliability of the RUSI and Matscan COP measurements by measuring the thickness of the abdominal muscles and COP sway velocity during unipedal stance on 2 occasions approximately 24 hours apart. An ultrasound measurement for test-retest reliability was obtained in the relaxed state (without the ADIM) during the unipedal stance for all participants.

We determined the baseline core-stabilization performance of each participant. Next, we introduced the ADIM exercise using the PBU, RUSI, and electromyography in the supine hook-lying, prone, quadruped, and bilateral-stance positions followed by a unipedal stance in progression (Figure 3). According to the ADIM procedures described by Richardson et al,^15,18  each participant was instructed to relax the upper chest and abdomen and to gently and gradually draw in the lower abdomen below the navel without moving the upper abdomen or spine while maintaining a neutral pelvic position and exhaling with lateral chest expansion for 10 seconds, which was monitored using a digital watch (Casio Inc, Tokyo, Japan). After each ADIM practice, a 20-second rest interval was provided. This intervention lasted for 20 minutes each day, 7 days each week, over a 2-week period.

The descriptive statistics included means and SDs. We used a paired t test to assess the mean differences in the TrA, IO, and EO muscle thicknesses; the electromyographic amplitudes; and the COP sway measures between the pretest and posttest. The intraclass correlation coefficient (ICC [1,2]) was calculated to determine intrarater reliability. Further analysis regarding the power and effect size of the TrA preferential contraction ratio also was conducted. Findings were considered different at P < .05. We used the SPSS for Windows software (version 12.0; SPSS Inc, Chicago, IL) for the statistical analysis.

Initially, no participants were unable to complete the formal test, but after 2 weeks of ADIM training, all could successfully reduce the pressure on the biofeedback unit by a range of 4 to 10 mm Hg from 70 mm Hg and maintain it at 60 to 66 mm Hg with minimal EO and ES muscle contraction (<15% RVC) 3 consecutive times. The mean electromyographic amplitudes of the superficial EO (t₁₈ = 3.691, P = .002) and ES (t₁₈ = 2.823, P = .01) muscles were decreased after the intervention, confirming that superficial abdominal-muscle activation was inhibited effectively after ADIM training (Table 3).

The TrA contraction ratio in the unipedal-stance position differed between the pretest and posttest (t₁₈ = −2.327, P = .03; Table 4).This finding indicates that the thickness of the TrA muscle was improved after ADIM training with RUSI and electromyographic feedback. The preferential contraction ratio for the TrA muscle tended to increase during the unipedal-stance position in the posttest (t₁₈ = −2.074, P = .053). The power analysis and effect size of the TrA preferential contraction ratio showed relatively moderate power (0.596) and a small effect size (0.231), respectively.

A comparison of the normalized electromyographic amplitudes (% RVC) between the pretest and posttest with and without ADIM in the unipedal-stance position is presented in Table 5. Activation of the EO muscle differed between the pretest and posttest (t₁₈ = 3.172, P = .005).

We found differences in the COP_AP and COP_ML velocities (t₁₈ = 4.720 and 5.775, respectively) and SDs (t₁₈ = 5.558 and 3.953, respectively) and in the COP area (t₁₈ = 4.628) between the pretest and posttest (P < .001; Figures 4 through 6).

Intrarater reliability of the abdominal muscle thickness and COP measures in the unipedal-stance position was determined. An ICC analysis showed high intrarater reliability for the muscle-thickness measure (ICC [1,2] = 0.76–0.98) and COP_AP and COP_ML sway velocities (ICC [1,2] = 0.96–0.97), suggesting that both the muscle thickness (Table 6) and COP_AP and COP_ML sway-velocity measures (Table 7) were consistently reliable.

Our study highlights the effects of ADIM training on static core stability and unipedal postural stability in nonathletes with core instability. As anticipated, static core stability and unipedal postural stability were enhanced after ADIM training augmented with RUSI and electromyographic feedback. Given that we present the first evidence in the literature of the important role of core-stability training in unipedal postural stability, comparing our data with those of other studies is difficult.

Most importantly, all of the nonathletes with initial core instability successfully performed a formal core-stability test with substantially decreased excessive EO (from 41.5% to 13.3% RVC) and ES (from 19.2% to 7.1% RVC) muscle activation after the intervention. This core-stabilization–training effect was evident even during the higher hierarchy of the unipedal postural-stability test. Our data are in line with those of previous studies^15,22,40,41  in which the authors examined the effect of the ADIM on lumbopelvic core stability in uninjured participants²²  and participants with LBP. Richardson et al¹⁵  compared the differential effects of abdominal bracing and the ADIM on sacroiliac joint laxity in a healthy population using the Doppler color image of the vibrating (200 Hz) sacrum and ilium and showed a superior effect of the ADIM on sacroiliac laxity or stiffness compared with a control population. The probable reason for such an improvement is that the ADIM activates the deep TrA and IO muscles independent of the superficial EO and ES muscles, stiffening the sacroiliac joint. Drysdale et al⁴⁰  examined the normalized peak electromyographic amplitude of the rectus abdominis and EO muscles in healthy young adults and found greater reduction of the global abdominal rectus abdominis and EO muscles during the ADIM than during pelvic tilting, regardless of the testing position. In a recent randomized clinical trial, Chon et al⁴¹  evaluated the augmented effects of combined ADIM and sequential voluntary cocontraction training via resisted ankle dorsiflexion on TrA and IO electromyographic activation patterns, abdominal muscle thickness, and associated pain reduction in patients with LBP. This ADIM, combined with cocontraction, enhanced muscle activity with associated morphologic changes in the TrA and IO muscles, thereby reducing LBP. In this study, an increase in TrA muscle thickness was accompanied by an increase in IO muscle thickness. Certainly, this finding suggests that the ADIM does not activate the TrA muscle independently but rather coactivates it with the IO muscle to optimize static core stability, which is a prerequisite for static postural stability.

We determined unipedal postural stability and associated neuromuscular performance by the COP sway with parallel measurements of RUSI muscle thickness and electromyographic muscle activity during a unipedal stance. The COP sway analysis showed that the COP_AP and COP_ML sway velocities, SDs, and areas were improved markedly after the ADIM training. This finding corroborates the observations of other recent authors^7,8  who investigated the effect of core-stabilization exercises on bipedal postural stability. Muthukrishnan et al⁸  compared the long-term effects of core-stabilization exercise (progression from static to dynamic) and conventional physical therapy in patients with chronic LBP and found improvements in the dynamic postural-stability measures (force, moment, and COP) that were comparable with those of healthy controls. Similarly, Kaji et al⁷  evaluated the immediate effects of core stabilization (elbow-toe and hand-heel) exercises on static postural stability in 17 healthy young adults and showed changes in the maximal ranges and SDs of the COP_ML sway and COP_AP and COP_ML sway velocities.

Our concurrent RUSI and electromyographic data analyses of core and postural muscles during unipedal postural-stability testing revealed that TrA muscle thickness increased, EO muscle activation decreased, and tibialis anterior and ES muscle activity tended to decrease after ADIM biofeedback training. These results suggest that ADIM training helped to stimulate TrA muscle activation while inhibiting superficial EO and ES muscle activation. This efficient neuromuscular control increased lumbopelvic stability, subsequently diminishing or neutralizing the demand of excessive ankle muscle activity. One important mechanism by which the ADIM improved the neuromuscular function of the TrA muscle and associated lumbar spinal stability was the neuromechanical stiffening of the thoracolumbar fascia. The synergistic contraction of the TrA muscle and posterior fibers of the IO muscle increases posterolateral lumbar tension on the thoracolumbar fascia, which connects to the contralateral gluteal and hamstrings muscles via the sacrotuberous ligament. Coactivation of the TrA and IO muscles, together with the thoracolumbar fascia, increases intra-abdominal pressure and creates a more rigid, or stiffer, cylinder of the abdomen. This core stiffness or stability of the lumbar spine results in static unipedal postural stability.^9,16,19–23  Static core stability and unipedal postural stability are important elements for clinical diagnosis, the early phase of sport rehabilitation, and the prevention of recurrent LBP and associated postural instability.

Notwithstanding the clinical implications of our novel findings, our study had several shortcomings. First, the neuromuscular-control mechanism of the deep core muscles (ie, TrA, IO, mulitifidus, and quadratus lumborum) during the stability tests was not determined by surface electromyography. However, muscle thickness is an important indicator for neuromuscular control.⁴²  Second, our primary goal was to examine the ADIM intervention-related effects on unipedal postural stability in nonathletes with core instability rather than just compare the relative changes or effects of the ADIM between the relaxed and contracted (ADIM) conditions during the static unipedal stance. We did not measure static unipedal postural stability (COP measure) in the relaxed state, yet the relative effects of the ADIM on postural stability warrant further evaluation. Third, the preferential contraction ratio of the TrA was different at P = .053, indicating that the ratio tended to improve as a function of ADIM training, which was manifested consistently in the TrA muscle-thickness data. Further analysis regarding the power (0.596) and effect size (0.231) of the TrA preferential contraction ratio showed that they were of relatively moderate power and small effect size, which may indicate the need for a larger sample size. Fourth, in our preliminary investigation, we explored the static core stability and unipedal postural-stability mechanism of ADIM training in pain-free individuals with core instability. Our results cannot be generalized to other populations with differential pathologic conditions or dynamic core- and postural-stability conditions.⁴³  Our findings invite researchers to examine the relationship between dynamic core and unipedal postural stability. Nevertheless, our results make an important contribution to the existing body of knowledge on core-stabilization exercise of the abdominal muscles in populations with core instability.

We demonstrated that ADIM training using RUSI and electromyography effectively improved morphologic changes in TrA muscle thickness and the neuromuscular pattern of an overactive superficial EO muscle, contributing to static core stability and unipedal postural stability. Our findings may provide clinical insights for sport rehabilitation to improve static core and postural stability in injured athletes with core instability.

We thank research assistants Kyoung-seo Kim and Hae-man Ou for their assistance in data collection.