Multivariate Analysis in the Maximum Strength Performance

Abstract

This study performed an exploratory analysis of the anthropometrical and morphological muscle variables related to the one-repetition maximum (1RM) performance. In addition, the capacity of these variables to predict the force production was analyzed. 50 active males were submitted to the experimental procedures: vastus lateralis muscle biopsy, quadriceps magnetic resonance imaging, body mass assessment and 1RM test in the leg-press exercise. K-means cluster analysis was performed after obtaining the body mass, sum of the left and right quadriceps muscle cross-sectional area (∑CSA), percentage of the type II fibers and the 1RM performance. The number of clusters was defined a priori and then were labeled as high strength performance (HSP1RM) group and low strength performance (LSP1RM) group. Stepwise multiple regressions were performed by means of body mass, ∑CSA, percentage of the type II fibers and clusters as predictors' variables and 1RM performance as response variable. The clusters mean ± SD were: 292.8 ± 52.1 kg, 84.7 ± 17.9 kg, 19249.7 ± 1645.5 mm(2) and 50.8 ± 7.2% for the HSP1RM and 254.0 ± 51.1 kg, 69.2 ± 8.1 kg, 15483.1 ± 1104.8mm(2) and 51.7 ± 6.2%, for the LSP1RM in the 1RM, body mass, ∑CSA and muscle fiber type II percentage, respectively. The most important variable in the clusters division was the ∑CSA. In addition, the ∑CSA and muscle fiber type II percentage explained the variance in the 1RM performance (Adj R2=0.35, p=0.0001) for all participants and for the LSP1RM (Adj R2=0.25, p=0.002). For the HSP1RM, only the ∑CSA was entered in the model and showed the highest capacity to explain the variance in the 1RM performance (Adj R2=0.38, p=0.01). As a conclusion, the muscle CSA was the most relevant variable to predict force production in individuals with no strength training background.