The purpose of this investigation was to provide further insight into the functional interrelationships between the muscles of the anterior abdominal wall. Specifically, this involved the evaluation of a mathematical model which represented the interactive capabilities of the anterior abdominal wall muscles.

The subjects were 12 male volunteers between the ages of 20 and 30 years. The task involved performing isometric trunk flexion tests from a supine and a 30 degree trunk flexed position. The five levels of effort (100,75,50,25,0 percent maximum) performed in each test position were based on pretest measurements. Data derived from the two test positions were analyzed separately.

Photographic and goniometric data were used to adapt an interactive model to the individual geometry of each subject's pelvic, abdominal, and thoracic regions. The normalized mean electromyographic activities of the upper section of rectus abdominis (URA), lower section of rectus abdominis (LRA), external oblique (EO), and internal oblique (10) were utilized as model input variables.

The noninteractive model was used to generate a root mean squared (RMS) error score for each test position. Each RMS error score was calculated from the differences between the NEMG levels of URA and LRA. The interactive model was used to calculate RMS error scores based on URA, LRA, and optimized EO and IO NEMG activity. A one tailed paired t-test (p ≤ .05) was employed to test for differences between the paired noninteractive and interactive RMS values.

Results of the photographic and geometrical data indicated that URA and EO were capable of shortening independently from LRA and IO. EO was better geometrically positioned than 10 to produce forces parallel to URA and LRA.

In general, the NEMG data were highly variable. The magnitude of the noninteractive RMS values averaged one-quarter the magnitude of URA and LRA NEMG levels. The pooled interactive model analysis indicated that IO was capable of generating greater forces relative to EO. Combining the optimized relative strengths of each oblique muscle with its geometrical alignment factor (relative effect) indicated that the effect the oblique muscles had on the tendinous intersection were symmetrical in nature. The means of the relative effect factors were similar across test positions as well. This suggested that the influence an oblique muscle could have on the tendinous intersection had a fixed upper limit which was lower than the force these muscles were capable of producing. In both test positions, the interactive model produced significantly (p <​ .05) lower RMS error values than the noninteractive model. The pooled RMS error terms were reduced 31 percent by the interactive model. The general conclusion based on this investigation was that a mathematical model based on the interactive capabilities of the anterior abdominal wall muscles may be used to partially explain functional differences found between the segments of rectus abdominis.