Milner-Brown, Stein, and Lee (1975) were the first to demonstrate experimentally that strength training results in directly measurable changes in the way motor units are recruited during effort. In their experiment, they determined the degree of synchronization of motor unit recruitment during effort by comparing the firing pattern of single motor units, measured with intramuscular electrodes, with the pattern of the surface EMG. Generally, their experimental paradigm asked the question, When the single motor unit fires, does the surface EMG show that a lot of other units are firing at the same time? They compared weightlifters with untrained control subjects, and they also studied a group of subjects before and after 6 weeks of weight training. Their study revealed a higher degree of synchronization of motor units in the first dorsal interosseus muscle in trained subjects than in untrained subjects. They also found that weight training increased the amplitudes of V2 and V3 responses (reflexes evoked during voluntary activation) measured at the muscle in response to stimulation of the peripheral nerve. The V2 and V3 responses—with latencies of about 56 and 83 milliseconds, respectively, following nerve stimulation—represent transcortical reflexes. Thus, increased synchronization of motor units during effort in strength-trained individuals is accompanied by enhanced transcortical reflexes, suggesting that firing synchronization may be linked to a supraspinal mechanism.
A word on synchronization of motor units is warranted at this point. The time course of synchronization that I refer to here is of little consequence to the rapidity of force development: The relatively small difference in the degree of motor unit synchronization ratios between trained and untrained individuals does not translate into differences in power, as we might be inclined to conclude. For instance, a certain degree of asynchrony is expected in peak force generation among motor units, even if the appearance of their action potential in the EMG is perfectly synchronized, because of differences in contractile properties. Rather, synchronization is seen as an index of the strength of presynaptic influences from common sources on motoneurons that determine the degree of coincident generation of action potentials in motoneurons within the pool (Semmler and Nordstrom 1998).
Thus, the results of Milner-Brown, Stein, and Lee (1975) suggest an effect of strength training on supraspinal mechanisms that are responsible for the recruitment of motor units, possibly by an enhancement of the efficacy of synapses on motoneurons from supraspinal sources. These original findings have been supported by more recent studies.
Sale, MacDougall, and colleagues (1983) substantiated the findings of enhanced reflex responses after strength training. They found increased V1 and V2 responses in all muscles trained during the study (which included hypothenar muscles, extensor digitorum brevis, brachioradialis, and soleus) after 9 to 22 weeks of training. Since in this study the reflexes were evoked while the subjects were performing an MVC (unlike in the 1975 study of Milner-Brown, Stein, and Lee, in which the effort was submaximal), they suggested that the elevated reflex responses indicated increased excitation of motoneurons during MVC. This again suggests that activation is not maximal in untrained subjects. However we interpret the source of this training-induced effect, it does support the notion that the nervous system responds to strength training.
Motor unit synchronization appears to adapt to chronic activity. Motor unit synchronization is measured as the near-coincident discharge of pairs of motor units during voluntary movement and reflects the distribution of shared inputs to motor units from corticospinal pathways. Semmler and Nordstrom (1998) have confirmed an increased synchrony of motor unit activation in the first dorsal interosseus of strength-trained subjects and interpreted their findings as an increased corticospinal activity accompanying this task as a result of training. Increased motor unit synchronization would benefit the rate of force development during a voluntary contraction and the coordination of synergists during a complex movement (Semmler 2002). An interesting observation was that skilled individuals (experienced musicians who use their fingers extensively to play their instruments) demonstrated levels of motor unit synchronization, common drive to all motor units, and tremor amplitude during maintained submaximal contractions that were all less than those of controls and strength-trained individuals. This may signify that, in some cases, reduced motor unit synchronization might be a beneficial adaptation to allow for independent synergies among synergistic muscles that might be seen in, say, the hand muscles of a piano or flute player. For the skilled subjects, the task involved the trained muscles but not the task used in the daily training. For the strength-trained subjects, on the other hand, the task included neither trained muscles nor training task. While it is clear from this literature that the function of inputs to motor units adapts to chronic activity, the precise mechanisms remain to be elucidated. More recently, investigators have found adaptations in motor unit coherence, which is an estimate, once again using discharge properties of pairs of motor units, of the oscillatory input to motor units that originates in cortical and subcortical areas (Semmler et al. 2004). The finding that this measure also differs between dominant and nondominant hands and among individuals with differing training status suggests adaptations that involve higher nervous system levels.
Evidence from Van Cutsem, Duchateau, and Hainaut (1998) provides substantial evidence for a neural effect of training. In their study, they asked subjects to train their ankle dorsiflexors for 12 weeks (5 times per week) by moving a load representing 30% to 40% of 1RM as quickly as possible. At the end of the study, recruitment of motor units during ballistic contractions was examined using intramuscular electrodes. The researchers found that ballistic contractions after the training program were faster, with a more rapid onset of EMG. They also found that maximal instantaneous firing rates of motor units during ballistic contractions were higher and showed less decrease
in frequency after training. In addition, the percentage of motor units showing incidents of doublets (two spikes of the same motor unit separated by 5 milliseconds or less) increased from 5.2% of the control units to 32.7% of the trained units. The authors suggested that ballistic training causes increased motoneuron excitability that leads to the previously described changes during voluntary excitation.
More recent evidence from Carroll and colleagues (2002) suggests that resistance training alters the functional properties of the corticospinal pathway. In their experiment, they resistance trained the first dorsal interosseus muscles of 16 individuals for 16 weeks and measured the EMG responses to TMS and TES, at rest and at various levels of voluntary contraction, before and after the training. Since TMS-evoked responses are more influenced by the excitability state of the motor cortex than TES-evoked responses are, the investigators were able to estimate cortical and spinal mechanisms. Their conclusion was that resistance training changes the organization of synaptic circuitry in the spinal cord, which can lead to altered recruitment of motor units during the task, with no major adaptations occurring at the level of the cortex.