It is well-documented that repeated sessions of heavy strength training (HST) induces lasting adaptations at many levels of the neuromuscular system (Sale, 1988; Aagaard et al., 2002; Carroll et al., 2011) resulting in overall strength gains (for review see Zatsiorsky, 2008). In particular, adaptations to the central nervous system such as an increase in corticospinal excitability and release of short-interval intra-cortical inhibition (SICI) following 2- to 8-week strength training programs have been commonly observed (Deschenes et al., 1994; Kidgell et al., 2010; Latella et al., 2012; Weier et al., 2012; Hendy and Kidgell, 2013). While there is strong evidence to suggest that significant neural adaptations occur following multiple resistance training sessions (Kidgell et al., 2010; Latella et al., 2012), and acute changes in corticospinal excitability with sustained submaximal isometric exercise (Nuzzo et al., 2016), few studies have systematically investigated the acute central and peripheral neural responses associated with a single HST session.

Previously, acute corticospinal responses following a single-session of exercise is thought to reflect central fatigue or acute neuroplastic responses to exercise (Smith et al., 2007; Teo et al., 2012). These studies have commonly showed a reduction in corticospinal excitability, as measured by a decrease in motor-evoked potential (MEP) amplitude, and an increase in SICI following maximal and submaximal exercise. Further, peripheral changes such as a reduction in motorneurone excitability and maximal strength production have also been reported (Todd et al., 2003). While these studies provide some insights into the initial corticospinal and peripheral responses to exercise, they only provide a short “window” of observation to the neural responses, mostly only up to 60 min post-exercise, which limits our understanding of the time-course and recovery of neuromuscular system following HST.

To better understand the time-course and body’s physiological responses to exercise, Bompa and Haff (2009) previously proposed the super-compensation model that consists of four distinct phases: (1) fatigue (0–2 h); (2) compensation back to baseline (24–48 h); (3) super-compensation beyond baseline (48–72 h); and (4) involution (>72 h). They suggested that while the recovery period (i.e., fatigue and compensation phase) may differ depending on the type of exercise and intensity, more neurally-demanding exercises, one such as HST, may require up to 24–48 h to recover back to baseline and for super-compensation to occur. However, to the best of our knowledge, no studies to date have investigated the time-course of corticospinal responses following HST and compared it to the current super-compensation model.

Therefore, the purpose of this study was to map the acute time-course of corticospinal excitability, intra-cortical inhibition and facilitation, peripheral nerve excitability and maximal force production of the biceps brachii (BB) up to 72 h following HST. Specifically, we aim to determine if the changes in corticospinal and peripheral responses would coincide with the four stages of the super-compensation model proposed by Bompa and Haff (2009). Based upon evidence that strength improvements may still be observed even if HST was performed with less than 48 h rest in between HST sessions (Raastad et al., 2012; Cook et al., 2014), we hypothesized that the fatigue, compensation and super-compensation phases of the cycle would be shorter than current suggestions of a 24–48 h recovery period back to basal levels.

The aim of the study was to map the time-course of corticospinal adaptations and maximal force output, up to 72 h post-training, following a single session of HST of the BB. The results indicated an immediate reduction in MVIC, followed by a super-compensatory increase at 6 h post-training. M_MAX showed a reduction immediately post-training that lasted for 30 min, before increasing at 6 h. A reduction in MEP amplitude was observed immediately post-training up to 20 min, which was followed by an increase at the 48 and 72 h mark. Collectively, our results suggest that changes in corticospinal excitability, intra-cortical inhibition and facilitation, peripheral nerve excitability and maximal force production following HST of the BB may follow a shorter time-course of fatigue, recover and super-compensation as previously suggested by Bompa and Haff (2009).

The main finding from this study showed that MVIC returned to baseline levels within 1 h, with an increase at 6 h post-training indicative of a super-compensation effect. This pattern of recovery within 1 h and super-compensation by 6 h post-training suggests a significantly shorter time-course of recovery and super-compensation compared to the current proposed model. A possible explanation for this effect may be that training in the morning may improve performance measures later in the day. A study by Cook et al. (2014) found that when a 3 RM back squat protocol was performed in the morning, strength was improved when tested in the afternoon. Similarly, Ekstrand et al. (2013) showed that resistance exercise in the morning improved explosive power 4–6 h later. However, afternoon and evening strength training can be positively influenced by the body’s natural circadian rhythm and optimal core temperature (Teo et al., 2011a,b) and thus may contribute at least in part to these findings. Previously, the super-compensation model has suggested a significant reduction in force output observed immediately post-training, remaining impaired up until 1 h (Ruotsalainen et al., 2014), followed by a proposed return to baseline between 24 h and 48 h (Ide et al., 2011). It is therefore recommended that a 72 h rest period between HST sessions to prevent overtraining. This concept is in line with work by Howatson et al. (2016) suggesting force is impaired until at least 24 h after strength exercise. However, studies prescribing strength training frequencies as high as five times per week have also shown greater increases, maximum bench press than those who trained four or less times per week (McKenzie, 1981; Serra et al., 2015). Other high frequency protocols (Raastad et al., 2012) have also supported the idea of increased training frequency. The current inconsistencies on performance recovery, improvement and training frequency is likely due to the differing protocols used in previous literature thus making comparison between studies difficult. The findings of the current study suggest a shortened time course of recovery with strength training in the arm and supports previous research (McKenzie, 1981; Raastad et al., 2012; Serra et al., 2015) applying high frequency training for optimal strength gains.

Similarly, the observed changes in M_MAX post-training do not adhere to the time-course as shown in previous super-compensation models (Bompa and Haff, 2009). When analyzed over time, there was a significant suppression of M_MAX post-training lasting up until 30 min. M_MAX has previously been shown to decline with muscle fatigue in the elbow flexors with maximal contractions (Todd et al., 2003). A likely reason for the suppression of M_MAX may be explained by a reduction in sodium-potassium pump efficiency of the sarcolemma following HST (Kirkendall, 1990; Nielsen and Clausen, 2000; Tucker et al., 2005; Mileva et al., 2012). Other mechanisms have also been implicated, such as fatigue-related changes in neurotransmitter release at the neuromuscular junction that may affect electrical nerve conduction (Kirkendall, 1990; Deschenes et al., 1994). Following this, increases in M_MAX showed an excitatory increase above baseline as early as 6 h whereby increased peripheral drive may reflect alterations in motor neuron recruitment and firing rate (Aagaard, 2003). Similar to force production, the time-course of M_MAX fatigue, recovery and super-compensation was much shorter than proposed by previous literature (Bompa and Haff, 2009) and coincides with an increase in force generating capacity. The super-compensation window appears to begin earlier than the proposed time line in this model. The efficacy of peripheral nerve excitability recovery may contribute, at least in part, to the shortened time-course for super compensatory effects after a single strength training session.

In this study, normalized MEP responses from TMS showed an immediate decrease in corticospinal excitability that lasted up to 30 min post-training. Previous studies have showed mixed evidence on the changes in MEP amplitude following HST (Kidgell et al., 2010; Weier et al., 2012; Hendy and Kidgell, 2013; Ruotsalainen et al., 2014). Ruotsalainen et al. (2014) reported an increase in MEP amplitude at the start of a task, while McNeil et al. (2011a) showed a decrease in MEP amplitude after sustained muscular effort at 25% of MVC over 10 min. Our study showed a significant reduction in MEP amplitude immediately post-training, followed by an increase at 72 h. Our findings are in agreement with previous studies (Todd et al., 2003; McNeil et al., 2011a) where a reduction in MEP amplitude following sustained submaximal muscular contractions in the elbow flexors was observed (Todd et al., 2003), and has similarly been found in other muscle groups after isometric voluntary contractions (McNeil et al., 2011a).

An interesting finding from our study was the lack of intra-cortical changes after a single session HST. This is dissimilar to studies reporting an increase in corticospinal excitability and a reduction in SICI after 2 and 4 weeks of strength training in the upper and lower limb (Weier et al., 2012; Hendy and Kidgell, 2013). Our findings suggest that, with no evident change in ICF or inhibition and a concurrent decrease in peripheral neural excitability the reductions in MEP amplitude appear to be primarily driven by mechanisms downstream of the M1. It is possible that spinal inhibitory mechanisms contribute to this finding (McNeil et al., 2011a). Further, changes in excitability have commonly been found to occur at sub-cortical spinal levels with acute and early strength training (Aagaard, 2003; Nuzzo et al., 2016). Therefore the acute responses may reflect perturbations at subcortical levels.

In light of our findings that differ to previous studies, we acknowledge that several limitations in our study that may have contributed to disparate findings. First, one strength session may not be a sufficient stimulus to induce long-lasting changes, but rather, changes downstream of the M1 may primarily drive super-compensatory responses after training. Second, maximal force production has been shown to increase with training, without any apparent increases in sEMG (Cannon and Cafarelli, 1987; McNeil et al., 2011a) and that maximal voluntary activation increases may not be reflected in sEMG signals (Latella et al., 2012). This would imply that any increase could not be completely attributed to changes in central nervous system. Third, HST recruits small and large motor units and testing the corticospinal tract at rest with TMS may target different neurons in the motor neuron pool and not be a direct representation of activated motor units (McNeil et al., 2011a). Likewise, the difference between gross movement employing large muscle groups, synergists and stabilizers in comparison to simple or isolated tasks is not known, and it is possible that exercise complexity may influence the output from the central nervous system. Although the primary aim of this study was to investigate acute neurophysiological responses, other factors outside the scope of this study such as mechanical, metabolite and hormonal responses may also contribute to the super-compensation cycle. Fourth, it is believed that alteration of the LICI inter stimulus interval from 100 ms to 150 ms may influence the activation of pre and post synaptic GABA_b receptors and thus should be considered to clarify the locus of GABA_b mediated inhibition (Vallence et al., 2014). Lastly, we acknowledge that the findings support a shortened super-compensation cycle in only recreationally trained populations and may not translate to other populations such as older adults or even in athletes of different training status. Future studies should compare different population groups; novice, or elite, and within training factors such as the effects of increased volume, which may present a different fatigue and recovery response profile due to physiological factors outside the scope of this study.

In conclusion, our findings reveal that after a single strength training stimulus, the time-course of fatigue and recovery and possible super-compensation from an acute HST session in the BB appears to be shorter than that proposed in the current super-compensation model for recreationally trained populations. We acknowledge that other factors may also contribute to the super-compensation cycle, however, the observed neurophysiological changes appear to be primarily driven by peripheral neural mechanisms downstream of the M1. Our results from this study may have significant implications for coaches and strength and power athletes who may program their training based on the current super-compensation model. Based on our current findings, it may be that optimal frequency of strength training can be scheduled sooner than 72 h to enhance strength and neuromuscular adaptations associated with HST. We believe that investigating the basic of post-training neurophysiological changes and comparing it to the super-compensation model may provide evidence for better exercise prescription in future.

CL, AJP and DVW conceptualized the study design. CL, AMH and W-PT collected and analyzed the data. All authors contributed equally to the write-up of this manuscript.

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.