Changes in normalized mutual information in response to strength training: An ancillary analysis of a quasi-randomized controlled trial

The aim of the present investigation was twofold. (1) to assess test–retest reliability of normalized mutual information (NMI) values extracted from the surface electromyography (sEMG) signal of muscles pairs of the upper body during dynamic bench press at a high load, and (2) to assess changes in NMI values from before to after a five-week quasi-randomized controlled bench press training intervention. For test–retest reliability, 20 strength trained males (age 25 ± 2 years, height 1.81 ± 0.07 m) performed two three-repetition maximum (3RM) tests in bench press, while sEMG was recorded from six upper body muscles. Tests were separated by 8.2 ± 2.9 days. For the training intervention, 17 male participants (age 26 ± 5 years, height 1.80 ± 0.07 m) trained bench press specific strength training for 5 weeks (TRA), while 13 male participants (age 23 ± 3 years, height 1.80 ± 0.08 m) constituted a control group (CON). 3RM bench press test and sEMG recordings were carried out before and after the intervention period. The NMI values ranged from poor to almost perfect reliability, with the majority displaying substantial reliability. TRA displayed a significant decrease in NMI values during the concentric phase for two agonist–agonist muscle pairs, while one agonist–agonist and two agonist–antagonist muscle pairs increased the NMI values during the eccentric phase. The observed changes did not exceed the minimal detectable threshold, and we therefore cannot surely ascertain that the changes observed in NMI values reflect genuine neural adaptations.