Decision letter: Altered regulation of Ia afferent input during voluntary contraction in humans with spinal cord injury

Abstract

Article Figures and data Abstract Editor's evaluation Introduction Results Discussion Materials and methods Data availability References Decision letter Author response Article and author information Metrics Abstract Sensory input converging on the spinal cord contributes to the control of movement. Although sensory pathways reorganize following spinal cord injury (SCI), the extent to which sensory input from Ia afferents is regulated during voluntary contraction after the injury remains largely unknown. To address this question, the soleus H-reflex and conditioning of the H-reflex by stimulating homonymous [depression of the soleus H-reflex evoked by common peroneal nerve (CPN) stimulation, D1 inhibition] and heteronymous (d), [monosynaptic Ia facilitation of the soleus H-reflex evoked by femoral nerve stimulation (FN facilitation)] nerves were tested at rest, and during tonic voluntary contraction in humans with and without chronic incomplete SCI. The soleus H-reflex size increased in both groups during voluntary contraction compared with rest, but to a lesser extent in SCI participants. Compared with rest, the D1 inhibition decreased during voluntary contraction in controls but it was still present in SCI participants. Further, the FN facilitation increased in controls but remained unchanged in SCI participants during voluntary contraction compared with rest. Changes in the D1 inhibition and FN facilitation were correlated with changes in the H-reflex during voluntary contraction, suggesting an association between outcomes. These findings provide the first demonstration that the regulation of Ia afferent input from homonymous and heteronymous nerves is altered during voluntary contraction in humans with SCI, resulting in lesser facilitatory effect on motor neurons. Editor's evaluation This paper will be of interest to basic and clinical neurophysiologists who are focused on understanding neural mechanisms that influence recovery following spinal cord injury (SCI). The work compares the afferent regulation of motor output to soleus muscle in controls and individuals with SCI. The results indicate differences between groups such that there is less facilitation in the SCI group during muscle contraction. https://doi.org/10.7554/eLife.80089.sa0 Decision letter Reviews on Sciety eLife's review process Introduction Anatomical and physiological studies have shown that sensory input to motor neurons is altered following spinal cord injury (SCI) (DAmico et al., 2014). For example, lesions of descending motor tracts in animals result in aberrant sprouting of primary afferents, leading to symptoms of hyperreflexia (Murray and Goldberger, 1974; Wong et al., 2000), and prolonged excitatory postsynaptic potentials (EPSPs) are observed in motor neurons in response to brief sensory stimulation (Baker and Chandler, 1987). In agreement, humans with SCI show prolonged depolarization of motor neurons in response to stimulation of sensory nerves (Norton et al., 2008), exaggerated stretch reflexes (Chen et al., 2020), and decreased transmission in spinal inhibitory pathways compared with uninjured controls (Jones and Yang, 1994; Crone et al., 2003; DeForest et al., 2020). The functional consequences of this altered sensory input conveying to motor neurons following SCI remain largely unknown. Primary afferent fibers (Ia) are rapidly conducting sensory fibers that originate from muscle spindle primary endings, which constantly monitors the rate at which a muscle stretch changes (Matthews, 1972). Ia afferent fibers bifurcate on entering the spinal cord and run several segments in both rostral and caudal directions in the dorsal columns and make contact with motor neurons (Pierrot-Deseilligny and Burke, 2012). Different mechanisms can contribute to regulate Ia afferent input conveying to motor neurons. For decades, it was thought that sensory regulation was accomplished in part through axoaxonic contacts at the terminal of Ia sensory axons from GABAergic neurons that receive innervation from the brain and spinal cord through presynaptic inhibition (Frank and Fuortes, 1957; Eccles et al., 1961; Eccles et al., 1962; Rudomin and Schmidt, 1999). However, recent evidence in animal and humans suggested that facilitation of Ia-mediated EPSPs in motor neurons likely occurs when axon nodes of Ranvier are depolarized from activation of nodal GABAA receptors, which contributes to reduce branch point failure in Ia afferent fibers (Hari et al., 2021; Metz et al., 2021). GABAergic neurons known to innervate terminals of Ia afferent fibers in the spinal cord also innervate nodes of Ranvier contributing to prevent failure of sodium spike transmission at branch points, which can facilitate sensory transmission and reflexes (Hari et al., 2021). Furthermore, in humans, cutaneous, proprioceptive, and corticospinal pathways can facilitate rather than inhibit Ia afferents facilitating the propagation of action potentials to motor neurons (Metz et al., 2021). Thus, GABAergic networks can have both facilitatory and inhibitory actions on afferent transmission within the spinal cord at different sites within Ia afferent fibers (Metz et al., 2022). A critical question is how Ia afferent transmission is regulated during voluntary contraction after SCI. In uninjured humans, evidence showed that Ia afferent transmission decreases at the onset of a voluntary contraction (Hultborn et al., 1987b; Nielsen and Kagamihara, 1993) and during tonic contractions that last for 1–2 min (Meunier and Pierrot-Deseilligny, 1989; Nielsen and Kagamihara, 1993) compared to rest and it changes according to the task requirements (Capaday and Stein, 1986; Crone and Nielsen, 1989). Following SCI, descending motor pathways converging onto GABAergic interneurons thought to contribute to regulate Ia afferent transmission (Jankowska and Edgley, 2006) are likely altered. Indeed, humans with SCI show lesser corticospinal (Davey et al., 1998; Bunday et al., 2014) and H-reflex (Yang et al., 1991; Phadke et al., 2010) modulation during voluntary behaviors compared with control participants. We hypothesized that during voluntary contraction, Ia afferent input exerts a lesser facilitatory effect on motor neurons in SCI compared with control participants. To test this hypothesis, the soleus H-reflex and conditioning of the H-reflex by stimulating homonymous and heteronymous nerves by measuring the depression of the soleus H-reflex evoked by common peroneal nerve (CPN) stimulation [D1 inhibition] and the monosynaptic Ia facilitation of the soleus H-reflex evoked by femoral nerve (FN) stimulation [FN facilitation] were tested at rest and during 30% of tonic maximal voluntary contraction (MVC). The position of the ankle joint was maintained constant across conditions to standardize the possible effect of other inputs into the testing procedures (Figure 1). Figure 1 Download asset Open asset Experimental setup. (A) Schematic representation of afferent fibers and motor neurons stimulated during our procedures. The soleus reflex was evoked by electrical stimulation of Ia afferents on the posterior tibial nerve (PTN). We assessed Ia afferent input to motor neurons by measuring the depression of the soleus H-reflex evoked by stimulating Ia afferents on the common peroneal nerve (CPN; referred as ‘D1 inhibition’) and the monosynaptic Ia facilitation of the soleus H-reflex evoked by stimulating Ia afferents on femoral nerve (referred as ‘FN facilitation’) at rest, and during tonic voluntary contraction. (B) Representative traces showing the soleus H-reflex evoked by PNT stimulation, the D1 inhibition evoked by stimulation of the CPN preceding the PTN at a conditioning-test interval of 15 ms, and the FN facilitation evoked by stimulation of the FN after the PTN at a conditioning-test interval of –8 ms (negative value of the interval indicates that the stimuli to the PTN precedes the FN stimuli). Results EMG Figure 2A shows raw rectified EMG data in the soleus muscle in representative control and SCI participants at rest and during 30% of MVC. One-way ANOVA showed an effect of GROUP on MVCs (controls = 0.3 ± 0.1 mV and SCI = 0.1 ± 0.1 mV, p<0.001; F1,38=13.7, p=0.001, η2p=0.3; Figure 2B) and the maximal motor response (M-max) (controls = 13.2 ± 3.3 mV and SCI = 10.3 ± 4.8 mV; F1,38=5.1, p=0.03, η2p=0.2) but not the maximal H-reflex (H-max) (controls = 6.2 ± 2.7 mV and SCI = 6.5 ± 4.0 mV, F1,38=0.06, p=0.8) in the soleus muscle. We also found a group effect on H-max/M-max ratio (controls = 48.5% ± 19.7%, SCI = 62.0 ± 19.1%, p=0.03, η2p=0.2, Figure 2C). Note that there was no difference in the activation of the tibialis anterior muscle between controls and SCI participants during 30% of plantarflexion MVC (controls = 7.3% ± 3.6% of tibialis anterior MVC and SCI = 9.2% ± 10.6% of tibialis anterior MVC, p=0.2). Figure 2 Download asset Open asset Voluntary contraction and maximal H-reflex and maximal motor response (M-max) ratio (H-max/M-max ratio). (A) Electromyographic (EMG) traces tested at rest and during 30% of maximal voluntary contraction (MVC) with the soleus muscle in a control and in a spinal cord injury (SCI) participant. (B) Bar graph shows the MVC group data. The abscissa shows the groups tested (controls = blue bar, SCI = orange bar) and the ordinate shows the MVC (in millivolts). (C) Bar graph shows the H-max/M-max ratio group data. The abscissa shows the groups tested (controls = blue bar, SCI = orange bar) and the ordinate shows the H-max/M-max ratio. *p<0.05, one-way ANOVA with Holm-Sidak post-hoc analysis. n = 20 per group, error bars show standard diviation (SD). Soleus H-reflex Figure 3A illustrates raw traces showing the soleus H-reflex in a control and a SCI participant. Note that the H-reflex increased in both participants during voluntary contraction compared with rest but to a lesser extent in the individual with SCI. Repeated measures ANOVA showed an effect of GROUP (F1,38=15.6, p<0.001, η2p=0.3), CONTRACTION (F1,28=100.3, p<0.001, η2p=0.7) and in their interaction (F1,38=15.6, p<0.001, η2p=0.3) on the H-reflex size. Post hoc analysis showed that the H-reflex was larger during 30% of MVC (controls = 245.6% ± 88.7%, p<0.001; SCI = 163.2 ± 23.2%, p<0.001) compared to rest in both groups. Additionally, the H-reflex size increased to a lesser extent in SCI compared to control subjects at 30% of MVC (p<0.001, Figure 3B, C and D). Note that the soleus H-reflex was tested during 30% of MVC (controls = 29.8% ± 7.1% of MVC and SCI = 33.2% ± 4.0% of MVC, p=0.2) into plantarflexion in both groups. Figure 3 Download asset Open asset Soleus H-reflex. (A) Representative EMG traces showing the soleus H-reflex tested at rest and during 30% of MVC in a control and in a SCI participant. (B) Graph shows the group H-reflex data. The abscissa shows the groups tested (controls = blue bar, SCI = orange bar) and the ordinate shows the H-reflex size during 30% of MVC expressed as a % of the H-reflex size at rest. Graphs show individual H-reflex data in controls (C) and SCI (D) participants. The abscissa shows the conditions tested (rest, 30% of MVC) and the ordinate shows the H-reflex size during 30% of MVC expressed as a % of the H-reflex size at rest. *p<0.05, repeated measures ANOVA with Holm-Sidak post-hoc analysis. n = 20 per group, error bars show SD. We conducted an additional control experiment where we matched absolute EMG level across groups during 30% of MVC by asking control participants to perform the similar EMG activity as SCI participants. Repeated measures ANOVA showed an effect of GROUP (F1,18=6.0, p=0.02, η2p=0.3), CONTRACTION (F1,18=52.4, p<0.001, η2p=0.7), and in their interaction (F1,18=6.1, p=0.02, η2p=0.3) on the H-reflex size. Post hoc analysis showed that the H-reflex was larger during 30% of MVC (controls = 218.9% ± 73.7%, p<0.001; SCI = 158.12 ± 23.1%, p<0.001) compared to rest in both groups. Note that the increases in H-reflex size were lesser in SCI compared with control at 30% of MVC (p=0.02). D1 inhibition Figure 4A and B illustrates raw traces showing the D1 inhibition measured in a representative control and SCI participant. Compared with rest, it shows that the D1 inhibition was abolished during 30% of MVC in the control participant while the D1 inhibition remained unchanged in the SCI participant. Repeated measures ANOVA showed an effect of GROUP (F1,32=6.4, p=0.01, η2p=0.1), CONTRACTION (F1,32=36.6, p<0.001, η2p=0.6), and in their interaction (F1,32=20.6, p<0.001, η2p=0.7) on the D1 inhibition. Post hoc analysis showed that the D1 inhibition decreased during 30% of MVC in controls (rest = 69.6% ± 15.4%, 30% of MVC = 100.5% ± 7.2%, p<0.001) but not in SCI (rest = 81.2% ± 7.8%, 30% of MVC = 80.2% ± 6.6%, p=0.5; Figure 4C) participants. Because D1 inhibition at rest is decreased in SCI (81.2% ± 7.8%) compared with control (69.6% ± 15.4%, p=0.02) participants, we tested the D1 inhibition in a subgroup of SCI participants (n=10) by adjusting to magnitude of the D1 inhibition to match the values obtained in control subjects (D1 inhibitionadj, see Materials and methods). Here, we found that the D1 inhibitionadj decreased during 30% of MVC compared with rest in SCI participants (rest = 72.9% ± 6.1%, 30% of MVC = 81.1% ± 8.2%, p<0.001; Figure 4C–D) but to a lesser extent than controls. Figure 4 Download asset Open asset D1 inhibition. Representative traces showing the H-reflex (control reflex, in black) and the H-reflex conditioned by common peroneal nerve (CPN) stimulation (conditioned H-reflex, in gray) tested at rest and during 30% of MVC in a control (A) and in a SCI (B) participant. Note that during 30% of MVC we show the test H-reflex adjusted. The bar graph shows the conditioned H-reflex normalized to the control H-reflex in both groups (C). The abscissa shows the groups tested at rest and during 30% of MVC (controls = blue bars, SCI = orange bars, SCIadj = brown bars). Note that here the SCIadj condition refers to testing of the D1 inhibitionadj. The ordinate shows the size of conditioned H-reflex expressed as a % of the control H-reflex (use to assess the D1 inhibition). Data from individual subjects (D) showing the conditioned H-reflex normalized to the control H-reflex in all groups tested (controls = blue circles, SCI = orange circles, SCIadj = brown circles). *p<0.05, repeated measures ANOVA with Holm-Sidak post-hoc analysis. Controls n = 14, SCI n = 20, SCIadj n = 10, error bars show SD. FN facilitation Figure 5A and B illustrates raw traces showing the FN facilitation measured in a control and in an SCI participant. We found that the FN facilitation increased in the control but not in the SCI participant during 30% of MVC compared with rest. Repeated measures ANOVA showed an effect of GROUP (F1,32=4.8, p=0.04, η2p=0.1), CONTRACTION (F1,32=30.8, p<0.001, η2p=0.5), and in their interaction (F1,30=51.5, p<0.001, η2p=0.6) on the FN facilitation. Post hoc analysis showed that the FN facilitation increased during 30% of MVC in controls (rest = 111.0% ± 7.1%, 30% of MVC = 126.3% ± 8.2%, p<0.001) but not in SCI (rest = 119.2% ± 9.3%, 30% of MVC = 117.2% ± 7.6%, p=0.16; Figure 5C) participants. The FN facilitation at rest was increased in SCI (119.2% ± 9.3%) compared with control (111.0% ± 7.1%, p=0.007) participants. Thus, we tested the FN facilitation in a subgroup of SCI participants (n=10) by adjusting to magnitude of the intensity of the conditioning pulse to match the level of FN facilitation obtained in control participants (FN facilitationadj, see Materials and methods). We found that the FN facilitationadj increased during 30% of MVC and rest in SCI participants (rest = 113.2% ± 6.2%, 30% of MVC = 120.2% ± 11.1%, p=0.01; Figure 5C–D) but to a lesser extent than controls. Figure 5 Download asset Open asset FN facilitation. Representative traces showing the H-reflex (control reflex, in black) and the H-reflex conditioned by FN stimulation (conditioned H-reflex, in gray) tested at rest and during 30% of MVC in a control (A) and in a SCI (B) participant. Note that during 30% of MVC we show the test H-reflex adjusted. The bar graph shows the conditioned H-reflex normalized to control H-reflex in both groups (C). The abscissa shows the groups tested at rest and during 30% of MVC (controls = blue bars, SCI = orange bars, SCIadj = brown bars). Note that here the SCIadj condition refers to the testing of the FN facilitationadj. The ordinate shows the size of the conditioned H-reflex expressed as a % of the control H-reflex (used to assess the FN facilitation). Data from individual subjects (D) showing the conditioned H-reflex normalized to the control H-reflex in all groups tested (controls = blue circles, SCI = orange circles, SCIadj = brown circles). *p<0.05, repeated measures ANOVA with Holm-Sidak post-hoc analysis. Controls n = 14, SCI n = 20, SCIadj n = 10, error bars show SD. Correlation Figure 6 shows the correlation between the size of H-reflex, D1 inhibition, and FN facilitation during 30% of MVC compared with rest. We found that the H-reflex size positively correlated with the D1 inhibition (r=0.6, p<0.001; Figure 6A) and the FN facilitation (r=0.4, p=0.024; Figure 6B) during 30% of MVC compared with rest. The D1 inhibition positively correlated with the FN facilitation (r=0.7, p<0.001; Figure 6C) during 30% of MVC compared with rest. No correlation was found between the size of H-reflex and the D1 inhibition (r=0.2, p=0.2) and the size of H-reflex and the FN facilitation (r=0.1, p=0.6) during 30% of MVC compared with rest when adjusted data was used for the analysis. Figure 6 Download asset Open asset Correlation. Graphs show individual data from controls (blue circles) and SCI (orange circles) participants. The abscissa shows the size of the H-reflex during 30% of MVC expressed as a % of the H-reflex tested at rest (A and B) and the ordinate shows the D1 inhibition during 30% of MVC expressed as a % of the D1 inhibition tested at rest (A), the FN facilitation during 30% of MVC expressed as a % of the FN facilitation tested at rest (B). Data from individual subjects (C) showing the correlation between the FN facilitation during 30% of MVC expressed as a % of the FN facilitation tested at rest (C, abscissa) and the D1 inhibition during 30% of MVC expressed as a % of the D1 inhibition tested at rest (C, ordinate). Dashed lines represent the regression line of all the data points included in the plot. *p<0.05, Controls n = 14, SCI n = 20. Discussion Our electrophysiological data supports the hypothesis that during voluntary contraction the regulation of Ia afferent input to motor neurons is altered following SCI. The size of the H-reflex in the soleus muscle increases in controls and SCI participants during voluntary contraction but to a lesser extent in people with SCI. Two observations suggest that altered regulation of Ia afferent input from homonymous and heteronymous nerves during voluntary contraction might contribute to these results. First, the D1 inhibition was decreased in controls but it was still present in SCI participants during voluntary contraction compared with rest. Second, the FN facilitation was increased in controls but not in SCI participants during contraction compared with rest. Changes in the D1 inhibition and the FN facilitation correlated with changes in the H-reflex size during voluntary contraction. Together, these findings indicate that during voluntary contraction Ia afferent input from homonymous and heteronymous nerves exert a lesser facilitatory effect on motor neurons in humans with SCI compared with control subjects. Regulation of Ia afferent input after SCI Although there have been multiple studies in animals and humans showing that sensory input conveying to motor neurons is altered following SCI (DAmico et al., 2014), the functional consequences of these changes remain largely unknown. Here, for the first time, we examined the regulation of Ia afferent input during voluntary contraction in humans with chronic incomplete SCI. Studies have used H-reflex conditioning paradigms in control humans to make inferences about the regulation of Ia afferent input (Hultborn et al., 1987a, Hultborn et al., 1987b; Meunier and Pierrot-Deseilligny, 1989; Nielsen and Kagamihara, 1993). For example, the depression of the soleus H-reflex evoked by CPN stimulation (referred here as D1 inhibition) is thought to be caused by presynaptic inhibition at the terminal of Ia afferents on soleus motor neurons (Mizuno et al., 1971) and the FN facilitation is thought to reflect the size of the monosynaptic EPSP in the soleus motor neurons evoked by activation of Ia afferents from the quadriceps muscle (Hultborn et al., 1987a). Both outcomes likely provide independent information about Ia afferent transmission that help to rule out changes in the recruitment gain of soleus motor neurons (Nielsen and Kagamihara, 1993). In the current study, the D1 inhibition was decreased during voluntary contraction in controls subjects but still present in SCI participants compared with rest. In addition, the FN facilitation was increased in control subjects during voluntary contraction but not in SCI participants when compared with rest. This is consistent with studies showing that, during a ramp-up and hold tonic contraction, in control subjects the FN facilitation decreased at the onset of a voluntary contraction (Hultborn et al., 1987a; Meunier and Pierrot-Deseilligny, 1989) and during a tonic contraction lasting for 1–2 min (Nielsen and Kagamihara, 1993) as performed in the present study. This is also consistent with findings in control subjects showing that vibratory inhibition of the soleus H-reflex decreases during tonic voluntary contraction (Iles and Roberts, 1987). These results were previously interpreted as a decrease in presynaptic inhibition during a voluntary contraction. For decades, it was thought that sensory regulation was accomplished in part through axoaxonic contacts at the terminal of Ia sensory axons through presynaptic inhibition (Frank and Fuortes, 1957; Eccles et al., 1961; Eccles et al., 1962; Rudomín et al., 1998). However, recent evidence in rats and control humans suggested that facilitation of Ia-mediated EPSPs in motor neurons likely occurs when axon nodes are depolarized from the activation of nodal GABAA receptors (Hari et al., 2021; Metz et al., 2021). Thus, it is possible that the regulation of Ia afferent input tested in our study occurs not at the terminal of the Ia sensory fiber but at different sites within the Ia afferent fiber. It is also possible that the suppression of the H-reflex evoked by conditioning of a homonymous nerve is related, at least in part, to post-activation depression or any direct effect on the soleus motor neurons (Metz et al., 2022). Regardless of the site of Ia afferent regulation, together, these findings suggest that during voluntary contraction proprioceptive input from homonymous and heteronymous nerves exert a lesser facilitatory effect on motor neurons after SCI. Both peripheral (Eccles et al., 1963; Jankowska et al., 2021) and central (Carpenter et al., 1963; Andersen et al., 1964; Fetz, 1968) mechanisms have been shown to contribute to regulate Ia sensory transmission. GABAergic neurons contributing to regulate Ia sensory transmission receive innervation from the brain and spinal cord (Rudomin and Schmidt, 1999). At rest, decreases in the D1 inhibition and increases in the FN facilitation in people with SCI compared with control participants could be related to decreased input from descending motor pathways. This is supported by multiple studies showing that activation of descending motor pathways, including the corticospinal pathway, is altered after the injury having a higher threshold resulting in the use of higher stimulus intensities in SCI compared with control participants (Bunday and Perez, 2012; Bunday et al., 2014; Cirillo et al., 2016). Then, how do we explain the facilitatory effect of Ia afferent input on motor neurons in control participants during voluntary contraction? A possibility is that this is, in part, related to descending inhibition of GABAergic interneurons, which overrides the suppression originated from peripheral sources during a voluntary contraction. Hence, in SCI participants, a lesser facilitatory effect of Ia afferent input on motor neurons might be present during voluntary contraction because of the abnormal and/or decreased contribution from descending pathways, which might not be strong enough to override peripheral sources. Participants in the present study had incomplete injuries and were able to perform voluntary contraction. Evidence showed that corticospinal excitability increases in controls and SCI participants during voluntary contraction but to a lesser extent in SCI participants (Davey et al., 1998; Bunday et al., 2014; Tazoe and Perez, 2021). Thus, a possibility is that the lesser facilitatory effect of Ia afferent input on motor neurons after SCI during voluntary contraction reflects altered contribution from descending motor pathways. Note that when the D1 inhibition and FN facilitation were tested at matching levels between groups, we observed a small but significant decrease in the D1 inhibition and increase in the FN facilitation during voluntary contraction, suggesting that to some extent similar mechanisms might contribute to the modulatinon of Ia afferent input in controls and SCI participants. This is also consistent with our results showing that H-reflex size increased during voluntary contraction in controls and SCI participants but to a lesser extent in people with SCI. Similarly, evidence showed that during voluntary contraction, motor neuron excitability (as measured by F waves) increases in people with SCI but to a lesser extent than in control participants (Vastano and Perez, 2020). Stretch reflexes (Woolacott and Burne, 2006) and H-reflexes Faist et al., 1994; Faist et al., 1996 have been reported to increase to a lesser extent or not at all during voluntary contraction in people with SCI compared with control subjects. Another important question is if changes in H-reflex size were related to changes in the D1 inhibition and FN facilitation. We did not find a correlation between these variables in each group. However, when we looked at the groups together, we found a strong positive correlation showing that increases in H-reflex size were associated with lesser D1 inhibition and larger FN facilitation, suggesting a relation between these variable in the overall population. Functional consequences During a voluntary contraction, the ‘excitability’ of motor neurons increases. Thus, regulation of Ia afferent input to motor neurons can have implications for the generation of motor output. Indeed, evidence from animal studies and modelling analysis suggested that presynaptic regulation of spinal sensory feedback contributes to ensure smooth execution of movement (Fink et al., 2014) and motor stability (Stein and Oğuztöreli, 1976). In intact humans, lesser facilitatory effect of Ia afferent input on motor neurons have been related to the optimization needed to improve motor performance during motor skill learning (Perez et al., 2005). Thus, it is possibly that the lesser facilitatory effect of Ia afferent input on motor neurons during voluntary contraction in SCI compared with controls contributes to regulate the ongoing voluntary contraction. Because after SCI prolonged EPSPs are observed in motor neurons in response to even brief sensory stimulation (Norton et al., 2008), a lesser facilitatory effect of Ia afferent input on motor neurons at the spinal level in SCI participants might be beneficial to control small levels of ongoing voluntary contraction as the one performed in our study. It is unclear if this adaptation contributes to the lack of task-dependent modulation of the H-reflex observed in humans with SCI during the gait cycle (Yang et al., 1991; Phadke et al., 2010). H-reflex modulation differs during sitting, standing, and walking in humans with and without SCI (Hayashi et al., 1992; Phadke et al., 2010) requiring that future studies assess the impact of our results on gait-based and other conditions. In controls, the greater facilitatory effect of Ia afferent input onto motor neurons might be functionally appropriate during the tonic voluntary contraction that we tested, but this might also change to a different extent during performance of more skilled motor behaviors. Materials and methods Subjects Twenty individuals with SCI (50.9±17.2 years, 2 women) and 20 control subjects (41.5±13.6 years, 6 women; F1,38=2.8, p=0.1) participated in the study. All subjects were provided written consent to experimental procedures, which were approved by the local ethics committee at Northwestern University (IRB protocol #STU00209996). Participants with SCI had a chronic injury (≥1 year) and were classified using the International Standards for Neurological Classification of Spinal Cord Injury (ISNCSCI) as having a C2-T10 SCI. Five out of the 20 subjects were categorized by the American Spinal Cord Injury Impairment Scale (AIS) as AIS C and the remaining 15 subjects were classified as AIS D. Nine individuals with SCI were currently taking anti-spastic medication (baclofen and/or tizanidine and/or gabapentin; Table 1) at the time of enrollment. These participants were asked to stop anti-spastic medication on the day of testing (at least 12 hr since last dosage). Spasticity was assessed using the Modified Ashworth Scale (MAS). All the SCI and controls participated in the H-reflex experiment, 6 of 20