Comparison of two abdominal training devices with an abdominal crunch using strength and EMG measurements

J Sports Med Phys Fitness. 1999 Sep;39(3):253-8.


Background: The purpose of this study was to compare the training effects of the Ab-Flex (F), Ab-Roller (R) and standard crunch (C) on EMG production, isometric maximum voluntary contraction (MVC), and isokinetic average peak torque at 30 degrees/sec (ISO) of the abdominal muscles. It was hypothesized that the training devices would have similar value in a strength training program.

Experimental design: this was a prospective study involving 18 training sessions of progressively increasing repetitions.

Setting: Neuromuscular Research Laboratory, University of Pittsburgh.

Subjects: thirty-two subjects volunteered for this study, but only 26 completed the training. Each subject participated in recreational activity, but had not performed any abdominal training prior to starting this study. Each subject was randomly assigned to either the control group or one of the treatment groups.

Interventions: there were three interventions: two training devices (Ab-Flex and Ab-Roller) and the standard crunch, considered a control group.

Measures: the pretest consisted of skin fold measurements (%), EMG activity (V) during the three interventions, and peak torque (Nm) plus EMG during the MVC and ISO tasks. The 18 training sessions over three weeks consisted of three sets of exercise with increasing repetitions from 10 to 20, by 2, every three sessions. The difference in pretest/posttest scores were compared using a One-way ANOVA on the mean differences (Mdiff) for each of: MVC, ISO (peak torque), and EMG for upper rectus (UR), lower rectus (LR), internal oblique (IO), and external oblique (EO). A T-Test was used to detect significance for the body fat measures.

Results: Mean differences (Mdiff) were normally distributed about zero for both MVC and ISO (MVC = -0.55, ISO = 4.57). The analysis by group showed no difference (p = 0.596) on the reported means (Nm) -3.16 (C), 5.84 (F) and -4.83 (R). The change associated to the treatment during MVC was only 4% (eta = 0.04). For the ISO the Mdiff (Nm) were 1.39 (C), 13.66 (F) and -2.06 (R) which were not significant (p = 0.127). The Ab-Flex was the only group to have a 95% confidence interval above zero, increasing by an average of 16.5%. There were no significant differences for the EMG activity for Mdiff or between group scores.

Conclusions: No significant differences were found with this study. These results would suggest that using these devices does not add significantly to overall abdominal strength development, or reduction of body fat. A suggestion could be made that certain devices influence muscles differently.

Publication types

  • Clinical Trial
  • Comparative Study
  • Randomized Controlled Trial

MeSH terms

  • Abdominal Muscles / physiology*
  • Adult
  • Electromyography
  • Exercise / physiology*
  • Exercise Therapy / instrumentation
  • Female
  • Humans
  • Male
  • Muscle Contraction*