Evidence for the Use of Dynamic Maximum Normalization Method of Muscle Activation During Weight Back Squats

The purpose of this study was to evaluate intra and inter-participant variability and reliability of muscle activations when analyzed using maximum voluntary isometric contractions (MVIC) and dynamic maximum (DMVC) normalization methods. Muscle activations were collected on twenty-seven participants (13 females, 14 males) performing one-repetition maximum (1RM) and submaximum (80%) back squats. Data from submaximum squats were normalized to MVICs and DMVCs. Data were compared using intra-class correlations over two testing days, variance ratio, and coefficients of variation. Mixed-model ANOVAs were used to elucidate the influence on intra- (method) and inter- (sex) subject variability. Reliability levels were good for rectus femoris (RF) (Interclass correlation coefficient (ICC): 0.776), vastus medialis (VM) (ICC: 0.820), and biceps femoris (BF) (ICC: 0.822) for DMVC across testing sessions. Reliability levels were also good for RF (ICC: 0.735) and BF (ICC: 0.795) during MVIC across testing sessions. Reliability of VM during MVICs was moderate, reaching an ICC of 0.660. Variance Ratio, peak coefficient of variation (CV), and mean CV for all muscles were reduced by 24%, 48%, and 56% in DMVC compared to MVIC, respectively. A significant normalization method by sex interaction was found for both peak and mean BF activation levels (p=0.005 and p=0.007, respectively). Post hoc tests revealed that differences between normalization methods were more pronounced in females than males for both peak (T=3.043, p=0.005, d=1.171) and mean (T=2.821, p=0.013, d=1.103) activations. Post hoc tests also found greater sex differences when normalizing to MVIC than DMVC for both peak (T=2.541, p=0.026, d=0.757) and mean (T=2.629, p=0.022, d=0.920) BF activations.