Rectus abdominis activity, but not femoris, is similar in different core training exercises: A statistical parametric mapping analysis

Rectus abdominis (RA) and rectus femoris (RF) differently activate in hip and trunk flexion movements. Previous studies using surface electromyography (sEMG) showed heterogenous results in mean and maximum activity of Upper (URA), Middle (MRA), and Lower (LRA) RA. Statistical Parametric Mapping (SPM) and Time-Varying Multi-muscle Co-activation function (TMCf) are emerging methods for whole-time-course continuous comparisons of muscle activity and co-activation. The aim was to analyse RA and RF activity in three core training exercises with continuous approach. Eight subjects performed Curl-up, Sit-up 45° and Sit-up 90° (where 45° is half and 90° is complete trunk flexion). sEMG from URA, MRA, LRA and RF were normalized to the maximum voluntary contraction root-mean-square (RMSMVC), and compared among muscles/exercises with SPM and TMCf. Angular displacement was assessed using stereophotogrammetry. Curl-up, Sit-up 45° and Sit-up 90° did not differently activate URA, MRA or LRA, and in their common range of motion (ROM) RA activity was similar among the three exercises. Conversely, Sit-up exercises elicited higher RF activity than Curl-up (on average 45% vs. 25% RMSMVC) mainly due to the wider ROM. RA-RF co-activation pattern was different in concentric and eccentric phases between Sit-up exercises and Curl-up, leading to significantly higher co-activation in Sit-up exercises.