Resistance training (RT) involves resisted movements with the overall goal to increase an individual’s strength. Numerous muscular and neurophysiological effects can be seen in the target muscles, such as, an increase in muscle volume and physiological cross-sectional area (Maden-Wilkinson et al., 2020) and neural adaptations (Gabriel et al., 2001, 2006; Škarabot et al., 2019; Aagaard et al., 2020; Hortobágyi et al., 2021; Pearcey et al., 2021). The increase in strength in response to RT occurs before morphological changes such as hypertrophy (i.e., muscle size increase) are measurable (Akima et al., 1999). A recent systematic review suggests that plastic changes at different levels of central nervous system such as, decreased activity of inhibitory networks in the primary motor cortex (Kidgell et al., 2017), increased corticospinal axon excitability at the spinal level (Mason et al., 2019) and changes in motor unit activation (Siddique et al., 2020) are responsible for early changes in strength output (Siddique et al., 2020). Additionally, skeletal muscle protein adaptations have been shown to occur within the first 2–4 weeks of resistance training (Staron et al., 1994) as recently discussed by Pearcey et al. (2021).

There is a growing body of research investigating changes in motor unit firing properties following RT. Amongst others, the most commonly investigated measures are: motor unit recruitment threshold (MURT), discharge rate (MUDR), and discharge rate variability (MUDRV) (Vila-Chã et al., 2010; Vila-Chã and Falla, 2016). For example, Del Vecchio et al. (2019a) reported a significant increase in MUDR and a significant decrease in MURT following 4-weeks of isometric RT of the tibialis anterior in young healthy adults. However, others have shown only early increases in MUDR with no changes after weeks of training or no change across the whole RT intervention. For instance, Patten et al. (2001), despite an initial significant increase in MUDR after 2 days of isometric RT of the digiti minimi, found no significant increase in MUDR after 42 days. Further, Pucci et al. (2006) found no significant change in MUDR after 3-weeks of isometric RT. These diverse findings make changes in motor unit firing properties following a regimen of RT unclear, and no systematic review has evaluated the overall evidence of changes in motor unit firing properties in response to RT in order to synthesize the available evidence and draw conclusions from the available studies. Thus, this systematic review aims to synthesize current evidence on the effect of RT on motor unit firing properties in order to determine the direction and strength of evidence.

The reporting of this review follows the Preferred Reporting Items for Systematic Reviews and Meta-Analysis Protocols (PRISMA-P) (Page et al., 2021). The review was prospectively registered with PROSPERO (CRD42021236376) on the 11th of February 2021.

This is the first systematic review to evaluate the current evidence surrounding changes in motor unit firing properties following RT vs. a control group. This review found a moderate level of evidence supporting a decrease of MUDRV following RT and a very low certainty of evidence for no variation in MUDR (SMD = 0.7, P = 0.43), with a high likelihood of heterogeneity (I² = 91%). Additionally, a low to moderate level of evidence was found for a change in the linear slope coefficient relationships between MUDR and MURT. None of the seven studies included in this review was rated as low risk of bias, therefore the outcomes should be interpreted with caution. More high-quality studies are needed before conclusions can be drawn. Despite this, important recommendations for future research can be drawn from this systematic review.

The studies included in this review had small sample sizes and lacked statistical power, not only reducing the chance of detecting a true effect but also making statistically significant results less likely to reflect a true effect (Button et al., 2013). Future studies should recruit larger sample sizes and report sample size calculations within their methods. Additionally, future research should strive to reduce imprecision. For example, the inclusion of women within future trials may account for observed sex differences in motor unit behavior during submaximal contractions (Peng et al., 2018; Inglis and Gabriel, 2020).

A large source of clinical heterogeneity within the current evidence comes from different RT regimens, as presented in Supplementary Appendix 7. Some trials used a constant RT regime, recording MVC to determine the intensity of contractions performed per session without altering these throughout the trial (Rich and Cafarelli, 2000; Pucci et al., 2006; Del Vecchio et al., 2019a) showing a variable strength increase. Other trials used a linear periodization system (Vila-Chã et al., 2010; Vila-Chã and Falla, 2016; Sterczala et al., 2020). A higher total volume of training occurs in the periodization model, an important factor for increasing strength (Ratamess et al., 2009). Increased intensity has also been shown to affect neural properties such as resting membrane potential and voltage threshold of α-motor neurons (Gardiner, 2006). Differences in single-joint vs. multi-joint exercises present an additional source of heterogeneity (Boccia et al., 2019). The current evidence on the physiological response to different training methods is thus inconclusive, which precludes a recommendation for future research (Schoenfeld et al., 2019).

There is much debate surrounding what causes an initial increase in strength before morphological changes occur in the muscular tissue. Although a combination of multiple neural factors are likely involved (Siddique et al., 2020; Pearcey et al., 2021; Škarabot et al., 2021; Tallent et al., 2021), it is not clear which factors are most relevant. A recent review concluded that the reticular formation and inhibitory cortical interneurons are involved in the neural response to RT (Škarabot et al., 2021). Furthermore, Siddique et al. (2020) found that the mediation of short-interval intracortical inhibition via GABA-ergic inhibitory circuits might also play a role in the increased efferent drive to agonists after RT, contributing to the early increases in strength. However, these findings illustrate subtle changes and are unlikely to be the only factors involved.

The present review has highlighted the potential of MUDR as a property that should be subject to further research. Despite low-quality trials and a high risk of bias, the significant increases in MUDR found by Vila-Chã et al. (2010) and Del Vecchio et al. (2019a) demonstrate there is potential for significant changes. Specifically, larger sample sizes, trials which incorporate a standardized RT program using a linear periodization model may produce significant results.

Another MU property highlighted by this review as likely to be changed by RT is MURT. Significant decreases in MURT were found in Del Vecchio et al. (2019a) after a 4-week RT program. These results seem to concur with a previous study which reported that MURT can initially decrease by as much as 40% in the first 24-h following eccentric exercises (Dartnall et al., 2011). The maintenance in MURT found by Sterczala et al. (2020), from a longer, 8-week program contradicts Del Vecchio’s results. This inconsistency is ultimately undermining the quality of the evidence, further reduced by the heterogeneity between studies. We hypothesize that the initial decrease in MURT observed by Dartnall et al. (2011), may continue for some weeks following the commencement of an RT program. However, a significant decrease cannot be observed at 8-weeks because of a gradual normalization of MURT back to pre-training levels, as muscle fiber changes begin to occur.

The different methods employed to identify motor units and assess their firing properties across the included studies are another important factor to consider. These different techniques may present some challenges since the examination of motor unit firing properties with methods with low-spatial resolution such as intramuscular EMG, may increase the likelihood of sampling different populations of motor units (i.e., low-vs. high-threshold motor units) across the training intervention. This is problematic since properties such as MUDR vary according to the MURT [i.e., low-threshold motor units usually show higher MUDR compared to high-threshold motor units (De Luca and Erim, 1994)]. Additionally, there is a need for a large number of motor units to be recorded to obtain reliable data for a pool of motor units with similar properties. Conversely, recordings with higher spatial resolution such as high-density surface EMG enable the assessment of similar populations of motor units across testing sessions (Martinez-Valdes et al., 2016) and in some cases (when changes in muscle volume are not too substantial) allow the tracking of the same motor units across the training intervention (Martinez-Valdes et al., 2017b) as recently shown by Martinez-Valdes et al. (2017a) and Del Vecchio et al. (2019b). Therefore, high-density EMG techniques might have the potential to further clarify the contribution of motor unit firing properties to the increase in strength observed after RT. However, their use in longer training interventions, inducing significant changes in muscle mass needs to be confirmed in future studies.

Other challenges when comparing studies assessing motor unit firing properties following a period of RT should be considered when interpreting the results of this systematic review. Important considerations are that different muscles have been examined, using different detection methods, and different training regimes have been employed. Some studies included in the review involved training with constant loads (requiring coordination between muscles), other used ballistic training, and some involved isometric training. These different training protocols most likely result in different changes in motor unit activity which impacts on drawing conclusions across multiple studies. Additionally, and as mentioned above, some studies use conventional intramuscular recordings, which is restricted to recordings during low force isometric contractions. In contrast, other studies used high-density EMG and a decomposition technique to extract motor unit activity from a large motor unit population during higher force contractions, but this is usually limited to more superficial motor units that are likely to be different to the motor units targeted with the conventional technique. Further, studies have obtained motor unit recordings from different muscles, which differ in the mode of action and the magnitude of coordination with other muscles. This may also contribute to different changes in motor unit activity with RT. Additionally, due to lack of evidence of the effects of RT in the same motor unit population (i.e., motor units tracked longitudinally across the period of training), it remains unclear whether the adaptations are uniform within the motor pool. All of the above impacts on the ability to draw conclusions from the current body of evidence and should be considered when interpreting the results of this review.

There is a lack of high-quality evidence to demonstrate the effect of RT on motor unit firing properties. The meta-analysis showed that there is no change in MUDR following RT although this should be interpreted with caution given the heterogeneity across studies particularly in relation to the muscle examined, the detection method used to assess MU activity and the type of RT used. A narrative synthesis revealed a mixed direction of evidence for most outcomes, with very low-to-moderate certainty of evidence. Clinical heterogeneity lies throughout this body of research, methodological differences between studies undermine the certainty of evidence, and differences between types of RT may have an impact on motor unit behavior.

The original contributions presented in the study are included in the article/Supplementary Material, further inquiries can be directed to the corresponding author.

DF, EE-C, and EE conceived and designed research. EE first reviewed, EE-C second reviewed, and DF third reviewed the manuscript. EE and EC drafted the manuscript. EE, EE-C, EM-V, and DF revised the manuscript and approved final version of the manuscript. All authors contributed to the article and approved the submitted version.

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.