Specificity of early motor unit adaptations with resistive exercise training

After exposure of the human body to resistive exercise, the force‐generation capacity of the trained muscles increases significantly. Despite decades of research, the neural and muscular stimuli that initiate these changes in muscle force are not yet fully understood. The study of these adaptations is further complicated by the fact that the changes may be partly specific to the training task. For example, short‐term strength training does not always influence the neural drive to muscles during the early phase (<100 ms) of force development in rapid isometric contractions. Here we discuss some of the studies that have investigated neuromuscular adaptations underlying changes in maximal force and rate of force development produced by different strength training interventions, with a focus on changes observed at the level of spinal motor neurons. We discuss the different motor unit adjustments needed to increase force or speed, and the specificity of some of the adaptations elicited by differences in the training tasks.


Introduction
Strength training is a form of physical exercise aimed at improving strength and speed (Aagaard et al., 2002a;Balshaw et al. 2016;Enoka, 1988;Folland & Williams, 2007;Hakkinen et al., 1998;Maffiuletti et al., 2016;Sale, 1988).It is often used for the restoration of motor function, prevention of injuries and improvement in sports performance.The increase in muscle force is usually achieved by loading the neuromuscular system against external resistive loads during dynamic and isometric exercises (Fig. 1).The human body responds rather quickly to strength training: it takes no more than 4 weeks of training at high forces to increase maximal muscle force significantly (Del Vecchio, Casolo, et al., 2019;Duchateau et al., 2005;Enoka, 1988;Glover & Baker, 2020;Hakkinen et al., 1998;Sale, 1988;Tillin & Folland, 2014).This relatively rapid response is mainly achieved by adaptations in the neural drive (cumulative motor unit activity) to the involved muscles.
Muscle force depends on both the neural and muscular properties of motor units, as well as on the passive properties of the connective tissues involved in force transmission from the muscle fibres to the skeleton.Although there may be an initial decline in 0 Alessandro Del Vecchio is an Assistant Professor at the Friedrich-Alexander University Erlangen-Nürnberg, Department of Artificial Intelligence in Biomedical Engineering.He completed his studies in neuroscience and biomechanics at the University of Rome 'Foro Italico' and Imperial College London.He is interested in motor unit physiology and neural interfacing in healthy and pathological conditions.
force-transmission capacity due to the damage caused by the unusual stressor, the neural adaptations appear to dominate the response to strength training within the first 4 weeks.An important feature of the neural adaptations is the specificity of the response, as observed in many studies on neuroplasticity (Cramer et al., 2011;Kleim & Jones, 2008).In contrast to the clear results in changes in force capacity, for example, scientific evidence is less consistent in relation to the influence of strength training on the maximal rate of force development, even when the training combines classic resistance training and rapid contractions (Del Vecchio et al., 2022;Hakkinen et al., 1998;Lanza et al., 2019; Figs 1 and 2).Indeed, although 4 weeks of strength training combined with rapid contractions can improve maximal muscle force, some studies have shown minimal changes in the rate of force development and motor unit discharge characteristics during fast contractions (Del Vecchio, Negro, et al., 2019;Del Vecchio et al., 2022).While the dominant role of neural adaptations to the increase in maximal force in the first few weeks of strength training is well supported, our review focuses on the less established potential specificity of these neural adaptations and their influence on other characteristics of force generation, in particular speed (Fig. 1).

Motor unit adaptations with strength training
In the absence of muscular adaptations, the generation of strong contractions requires the development of stable and high motor unit discharge rates.The rates normally produced during slow maximal voluntary contractions are less than those required to achieve maximal force (Macefield et al., 1996).Therefore, there is an opportunity to improve performance by increasing the discharge rates of the active motor units.
The gains in muscle force achieved with a few weeks of strength training are usually accompanied by significant increases in the amplitude of surface EMG recordings during submaximal contractions (Del Balso & Cafarelli, 2007;Penzer et al., 2015;Tracy et al., 2004;Vila-Chã et al., 2010).This finding indicates that such interventions increase the number of muscle fibre action potentials during a submaximal contraction (Dideriksen et al., 2011), which is consistent with the finding of significantly greater motor unit discharge rates during submaximal contractions of trained individuals (Del Vecchio, Casolo, et al., 2019;Vila-Chã et al., 2010).
Although strength training elicits significant and consistent adaptations in motor unit activity during sub-maximal contractions, less is known about discharge rate during maximal force contractions (MVCs) and the findings are mixed (Duchateau et al., 2005).For example, Kamen and colleagues found conflicting results in two studies on changes in maximal motor unit discharge rate after a few weeks of strength training.In one study, Kamen and Knight (2004) reported that peak discharge rates of motor units in vastus lateralis during isometric MVCs increased (19%) from the first to the second familiarization session but then, despite progressive increases in MVC force, did not change during the subsequent 6 weeks of strength training for young adults whereas it increased for older adults.There was a strong correlation between the changes in MVC force and discharge rate during the familiarization period for both groups of participants, but not during the training intervention.In another study, Christie and Kamen (2010) measured the change in peak discharge rate of motor units in tibialis anterior of young and older adults during isometric MVCs after 2 weeks of strength training the dorsiflexor muscles.Both groups exhibited an increase in peak discharge rate, but the average increase was greater for the older adults (24%) than it was for the young The training stimulus can comprise rapid movements with relatively low loads and brief durations with an emphasis on fast movements (top panel in green) or sustained movements at high contraction forces (classic strength training, red).These two training conditions can also be combined.Some evidence suggests that rapid-contraction training predominantly increases the rate of force development, with moderate effects on maximal force.In contrast, strength training evokes changes in maximal muscle force, but results are conflicting on the rate of force development, with some, but not all, studies showing minimal changes in maximal contractile capacity.Although it seems reasonable to assume that the combination of strength and rapid-contraction training would increase strength and rate of force development, current results suggest that the combination may have a greater influence on maximal force than rate of force development, although the evidence is still very limited.In the figure, max force corresponds to the increase in peak of force during a maximal force contraction (MVC), whereas speed corresponds to the increase in maximal rate of force development, and motor unit output to an increase in the neural drive to muscles.Note that the question marks indicate conclusions that are mainly based on assumptions by the authors, and that more experimental evidence is probably needed.adults (7%).Moreover, there was a statistically significant correlation between the increases in MVC force and maximal discharge rate.
When we consider the task of performing fast contractions, the neural requirements appear to differ from those needed to produce strong contractions slowly.Fast changes in muscle force require a rapid increase in the net synaptic input to motor neurons to hasten recruitment and reach high discharge rates in the shortest time possible; recruitment speed is positively correlated with the rate of increase in synaptic input.Moreover, fast contractions are characterized by greater discharge rates than those achieved during slow contractions up to maximal force (Del Vecchio, 2022).For example, Del Vecchio et al. ( 2019) observed peak discharge rates in the tibialis anterior muscle greater than 100 pulses per second (pps), which are more than double the rates observed during steady maximal force contractions (Christie & Kamen, 2010).Since the first observations of discharge rates during rapid contractions (Desmedt & Godaux, 1977), however, it has been consistently observed that the much greater discharge rates at the beginning of ballistic contractions (>100 pps for two to three discharges) are followed by a fast decline in rate.A rapid increase in rate above values normally observed during slow contractions followed by a relatively fast decrease are also observed in fast ramp-and-hold contractions (Bawa & Calancie, 1983;Del Vecchio et al., 2022).In these contractions, the high rate that allows a fast rate of force development is followed by a decrease in rate that soon matches the rates normally observed during slow contractions at the same target force.
The initial rapid increase in discharge rate is a specific response of the motor neuron to fast synaptic input that cannot be maintained over time.Such rate coding is consistent with observations in classic animal studies in response to rapid current injections (Baldissera et al., 1987;Kernell, 1965) and may therefore reflect a difference in the response of motor neurons to fast than to slower net excitatory synaptic input.The decrease in motor unit discharge rate as a function of rapid force contraction has been shown in simulations and in vitro and is related to spike-frequency adaptation (Revill & Fuglevand, 2011;Sawczuk et al., 1995).In addition, motor neurons presumably function in open-loop mode during ballistic contractions with the synaptic input being dominated by descending signals from supraspinal centres.
Given the requirements for fast contractions, it is reasonable to assume that speed training is most suited to increase the rate of force development.Speed training can be performed by repetition of ballistic contractions, which represent an impulse of force, or by rapid increases of force to a peak value and then either a brief sustained contraction or a more gradual decline back to baseline.The rate of force development can be increased by both types of speed training.For example, in a seminal study, Van Cutsem et al. (1998) found that 12 weeks of training the dorsiflexors with rapid dynamic (not isometric) contractions against a resistive load (30-40% of maximum) increased MVC torque (30%), rate of torque development (82%) and the discharge rates of motor units in tibialis anterior during ballistic contractions.They found that average discharge rate calculated from the first three interspike intervals increased from 69 pps before training to 96 pps after training.As discussed above, these discharge rates are substantially greater than observed during MVC of the tibialis anterior muscle before or after strength training (Christie & Kamen, 2010).Significantly, there was an increase in the number of motor units (before training, 5%; after training, 33%) that discharged doublets (≤5 ms between consecutive action potentials) at the onset of the ballistic contractions.Accordingly, Tillin and Folland (2014) found that 4 weeks of force-development (fast and hard) training increased the initial rate of force development by 53% in the knee extensors of young adults, whereas maximal force increased only slightly by 11%.However, although it is known that the rate of force development during rapid contractions is significantly correlated with both initial discharge rate and the speed at which motor units are recruited (Del Vecchio, Casolo, et al., 2019), there are no data on the change in maximal motor unit discharge rate after force-development training.
Although the influence of speed training on rate of force development is well supported, there is less consensus on whether classic strength training (slow contractions) modulates the rate of force development or is specific to maximal force.The results on this issue are mixed.In the study by Tillin and Folland (2014) cited above, they had a second group of individuals perform strength training.They observed that strength training increased MVC force (21%) but did not change the initial rate of force development.Other studies, however, did find an increase in the rate of force development after strength training alone (Aagaard et al., 2002a;Blazevich et al., 2008;Bruhn et al., 2006;Del Balso & Cafarelli, 2007;Holtermann et al., 2007;Jenkins et al., 2016;Vila-Chã et al., 2010).Therefore, the current evidence on the degree of specificity of the effects of strength training is largely mixed and the number of studies that compared training regimes systematically is too small to reach firm conclusions.
Training interventions that combine both training protocols produce a mixture of adaptations.For example, Del Vecchio et al. (2022) examined the adaptations elicited by a 4 week intervention during which participants performed both hard-and-fast (speed) and sustained (strength) isometric contractions with the dorsiflexors.The target force was 75% MVC torque for both sets of contractions.The training protocol produced a significant increase in MVC torque (12%) but no change in the rate of torque development.The intervention did not alter the initial discharge rate as determined from the first three interspike intervals (∼55 pps) or recruitment speed during fast contractions, but it did increase (before: 25.4 ± 3.4 pps; after: 27.9 ± 3.5 pps) discharge rate during a steady isometric contraction at 75% MVC torque.The absence of changes in the rate of torque development, initial discharge rate and recruitment speed was attributed to an inadequate training stimulus during the intervention.Similar results on maximal force and rate of force development were also obtained by Lanza et al. (2019) for mixed strength and speed training with isometric contractions.
As said, the conflicting results of the relatively few studies in the literature make it difficult to draw firm conclusions on the degree of specificity of the training effects.In addition to variability in experimental settings, another factor that influences the variability of results across training studies is the way in which changes in force characteristics are quantified.For example, there are several approaches that can be used to quantify the rate of force development (see Del Vecchio, 2022, for more detailed discussion on this topic).The results shown in Fig. 2 indicate that the peak rate of force development, when computed from the first derivative of force, occurs quite fast and often before 100 ms, with great variability among participants.However, most studies that have examined the rate of force development after strength training with isometric contractions have made an analysis of rate of force development in small arbitrary intervals of time and in this way have failed to obtain significant changes in absolute rate of force development during the early phase of contraction.None of these studies have assessed the full force-time curve or the maximum peak of the derivative, so it is challenging to compare data before and after training across studies (see Maffiuletti et al., 2016, for all the papers that use time intervals).Therefore, future studies should not only assess force in different time windows but also compute the maximal rate of force development as displayed in Fig. 2B  and C.This approach will reduce some of the variability in the values extracted from arbitrary time points along the force-time curve.
Finally, we note that due to the specificity of some adaptations, maximal force may differ when producing slow or fast contractions.For example, individuals who are trained to perform contractions with high loads (e.g.strength-trained athletes) may exhibit forces during an isometric MVC that are substantially greater than those during maximal rapid movements, even after familiarization with the isometric rate of force development contractions.In rapid actions, however, the maximal rate of force development is often achieved J Physiol 602.12 within 100 ms from the onset of the contraction, which leaves little time to engage other muscles to increase the rate of force development.Therefore, caution is needed when comparing maximal relative (%MVC) force and rate of force development in strength-trained individuals; high MVC force may decrease the maximal rate of force development when expressed relative to the maximal force.Therefore, we suggest reporting both absolute and normalized rates of force development in parallel with estimates of motor unit discharge rates.

Potential mechanisms underlying motor unit adaptations
The motor unit adaptations that can be elicited by physical training include alterations in the intrinsic excitability and properties of neurons (Orssatto et al., 2023;Zhang & Linden, 2003), increases in the strength, number and density of synaptic contacts (Kleim & Jones, 2008), and increases in net excitatory synaptic inputs (increase in excitation or decrease in inhibition) from the motor cortex or brain stem down to the spinal cord (Aagaard et al., 2002b(Aagaard et al., , 2020;;Glover & Baker, 2020;Penzer et al., 2015).Moreover, although evidence is still conflicting, as we discussed, it seems safe to assume that at least some of the adaptations are specific to the type of training and therefore underlie different physiological responses.
Christie and Kamen (2010) observed a weak inverse association between the duration of the after-hyperpolarization phase (AHP) of motor neurons and the maximal discharge rates, both before and after strength training (Christie & Kamen, 2010).In addition to the AHP, the sensitivity of motor neurons to excitatory input may change due to changes in intrinsic ion channels that underlie neuromodulation of motor neuron function (Zhang & Linden, 2003).As a result of such an adaptation, motor neurons that are close to saturation (Fuglevand et al., 2015) may be driven well above this limit by changes in their intrinsic properties.As summarized in MacDonell and Gardiner (2018), several groups have measured the adaptations in the properties of motor neurons that innervate hindlimb muscles of rodents after exposure to different types of physical-activity interventions (Gardiner et al., 2006;MacDonell & Gardiner, 2018).One of these interventions involved resistance-type training for 1 h per day, 5 days per week, for 5 weeks (Krutki et al., 2017).The 5 week intervention increased the tetanic force elicited by electrical stimulation of motor axons, but more so for slow-than fast-contracting motor units (49% vs. 21%, respectively).There were no statistically significant changes in several passive and threshold properties of motor neurons that innervate medial gastrocnemius or lateral gastrocnemius-soleus, including resting membrane potential, AHP amplitude or time course (contrary to the human study by Christie & Kamen, 2010), input resistance, or steady-state discharge rate.In contrast, there were substantial changes in some properties related to the repetitive discharge characteristics of motor neurons.These included decreases in the spike rise time (depolarization rate), the threshold current for the repetitive discharge of action potentials and membrane excitability (rheobase), and increases in the slope of the current-frequency relationship and initial discharge rate.Notably, there was no change in the incidence of doublets after the weight-lifting intervention, although the initial triplets (two subsequent interspike intervals <5 ms) increased and were observed for 22 motoneurons (only 10 for the control group).These types of adaptations are mediated by changes in ion conductances due to alterations in the number and location of motor neuron ion channels and receptors and their modulation.Simulations based on these findings suggest that an increase in sodium conductance plays a major role in contributing to the adaptations elicited by physical training (MacDonell and Gardiner, 2018).
The increase in doublets after training (Van Cutsem et al. 1998) may involve the rapidly inactivating sodium channels (Eijkelkamp et al., 2012;Ulbricht, 2005).Voltage-gated sodium channels are critical for the generation of action potentials in that they allow an inward current that depolarizes the membrane potential, but then they are inactivated and remain refractory for a few milliseconds.Rapidly inactivating sodium channels appear to recover from the refractory state more quickly and be available to respond sooner to an excitatory state.An up-regulation of these channels after speed training may account for the increased incidence of doublets, although there is not yet any experimental evidence to support this possibility.One potential explanation for why speed training may distinctly target sodium channels is that the synaptic inputs are greater during rapid actions than during submaximal contractions.This may generate a specific adaptation given that these channels regulate the spike-frequency adaptation response of the motor neurons (Miles et al., 2005).This could be also one of the reasons why Krutki et al. found an increase in the presence of triplets after training in anesthetized rats.Whatever the underlying adaptation responsible for the increase in doublets and triplets, their presence greatly increases the temporal summation of twitch responses and the rate of force development (Burke et al., 1976;Thomas et al., 1999).
The reticulospinal pathway can increase the ionotropic and metabotropic modulation of spinal motor neurons (Gardiner et al., 2006;Heckman et al., 2003Heckman et al., , 2009)), presumably via sodium channels, which would ultimately increase motor unit discharge rate.In addition to adaptations in the intrinsic properties of motor neurons, changes in motor unit discharge rates with strength training may be mediated by the synaptic input they receive.Indeed, Glover and Baker (2020) concluded that changes in motor neuron properties were not the dominant factor in contributing to increases in the amplitude of motor-evoked potentials after 8-9 weeks of resistance training.This conclusion was based on invasive measurements performed on two macaque monkeys in which responses were evoked by stimulating the primary motor cortex, medial longitudinal fasciculus (reticulospinal pathway) and pyramidal tract (corticospinal tract).Based on differences in the amplitudes of the motor-evoked potentials, they found that strength training generated neural adaptations primarily in intracortical and reticulospinal networks, including an increase in the synaptic efficacy of reticulospinal inputs onto interneurons and motor neurons in the spinal cord.
The role of the reticulospinal pathway may also underlie the specificity of the adaptations observed after speed training.When a loud acoustic stimulus is delivered while a person prepares to perform a reaction-time task, the subsequent action is produced more quickly with a greater rate of force development (Valls-Solé et al., 1995, 1999).Due to the latency of the response, the acoustic stimulus is presumed to engage the startle reflex pathways, including the reticulospinal pathway (Smith et al., 2019;Valls-Solé et al., 1999).Given that speed training requires the performance of rapid contractions, it seems likely that the details of the descending reticulospinal volley may be modulated according to the intended speed of each action.For example, the initial discharge rate and rate of force development for single motor units depend critically on the rate at which a current was injected into the motor neuron (Baldissera et al., 1987).Presumably, the dynamic response of motor neurons can be modulated by J Physiol 602.12 the intensity of reticulospinal input they receive (Škarabot et al., 2022).
In addition, training-specific changes in strength and rate of force development are encoded in different brain centres.It is likely that the changes in rapid force development after a few days of training are mediated by specific neurons encoding force and speed.Interestingly, preliminary evidence indicates that when the training stimulus is combined (strength and speed), the changes may favour strength gains (Fig. 3), presumably due to the absolute duration of exposure (i.e. the total 'work' performed by the populations of neurons across the nervous system in the two conditions, Fig. 3) to the stressor.The longer training stimulus may augment the intrinsic excitability of neurons (Zhang & Linden, 2003).It is tempting to speculate that these training-induced alterations in motor neuron throughput after strength and speed training constitute the engram that underlies the specificity of training (Moyer et al., 1996;Zhang & Linden, 2003).It has been indeed shown that training produces an increase in excitability that is manifested as specific alterations in voltage threshold and neuronal discharge characteristics, which may underlie the specificity of the adaptations.

Summary
We have discussed adaptations that augment the neuromuscular responses elicited by short-term strength training.Although the scientific evidence is still scarce, especially at the level of the motor unit, and more studies are needed with standardized protocols, strength training seems to elicit functional gains in muscle force with some degrees of specificity.For example, although there is strong evidence for an early influence of strength training on maximal force, strength training, either alone or combined with speed training, does not always impact the early phase of fast contractions (<100 ms).There is a need for studies with standardized training regimes and control conditions to shed light on the divergent neuromuscular adaptations produced by combined and specific training stimuli.

Figure 1 .
Figure 1.Neural adaptations with strength trainingThe training stimulus can comprise rapid movements with relatively low loads and brief durations with an emphasis on fast movements (top panel in green) or sustained movements at high contraction forces (classic strength training, red).These two training conditions can also be combined.Some evidence suggests that rapid-contraction training predominantly increases the rate of force development, with moderate effects on maximal force.In contrast, strength training evokes changes in maximal muscle force, but results are conflicting on the rate of force development, with some, but not all, studies showing minimal changes in maximal contractile capacity.Although it seems reasonable to assume that the combination of strength and rapid-contraction training would increase strength and rate of force development, current results suggest that the combination may have a greater influence on maximal force than rate of force development, although the evidence is still very limited.In the figure, max force corresponds to the increase in peak of force during a maximal force contraction (MVC), whereas speed corresponds to the increase in maximal rate of force development, and motor unit output to an increase in the neural drive to muscles.Note that the question marks indicate conclusions that are mainly based on assumptions by the authors, and that more experimental evidence is probably needed.

Figure 2 .
Figure 2. Force-time derivatives during rapid contractionsA, force-time curve (normalized to the maximal value) for an individual.B, the change over time of the force signal (x) with respect to the initial value (i) and the time window (T).The values were calculated as the difference between the force value at each instant t and the initial value, divided by the interval between the time instant of the initial value and the time instant t.The curve is normalized by the maximum value.C, the first derivative of the force signal normalized by the maximum value.D-F, the same curves as in A-C but for 20 participants (the best three contractions for each individual), without normalization.In E, the maximum values of the curves for three contractions are indicated by black dots.G, the time it takes to reach the maximum RFD for the two different derivative methods.Note the difference in the maximal derivative when computed with the two methods.

Figure 3 .
Figure 3. Influence of controlled isometric training on the applied force and motor unit discharge rates (Del Vecchio, Casolo, et al., 2019; Del Vecchio et al., 2022) A, ankle-dorsiflexion dynamometer.B and C, training protocol comprising sets of brief rapid (B) and longer duration isometric contractions (C).D-F, the results demonstrated that the combination of these two types of contractions increased the maximal discharge of the motor units during slow movements (D), but not fast movements (E and F).The strength-training component of the intervention depressed the early changes in rate of force development that are commonly observed with speed training.