Skip to main content
NCBI home page
Wiley Open Access Collection logoLink to Wiley Open Access Collection
. 2026 Jul 11;31(3):e70278. doi: 10.1002/pri.70278

Acute Modulation of Lumbar Motor Unit Behavior During Experimentally Induced Low Back Pain: A Motor Unit Decomposition Analysis

Franciele Parolini 1,2,3,✉, Klaus Magno Becker 2, João Paulo Vilas‐Boas 2, Ulysses Fernandes Ervilha 2,4, Rubim Santos 3, Márcio Fagundes Goethel 2
PMCID: PMC13355871  PMID: 42435379

ABSTRACT

Background and Purpose

Acute low back pain is a prevalent musculoskeletal disorder that can disrupt motor control and compromise functional stability. Evidence suggests that pain alters motor unit discharge characteristics and the temporal organization of muscle activation; however, the specific neuromuscular adaptations associated with acute low back pain remain insufficiently understood. This study aimed to investigate the effects of experimentally induced acute low back pain on force steadiness and motor unit discharge characteristics (including recruitment threshold, motor unit firing rate, and action potential amplitude) during a sustained spinal extension task

Methods

Thirty‐three healthy participants (aged 18–40 years) performed a sustained spinal extension task at 20% of their maximum voluntary contraction under two experimental conditions: pre‐ and during pain induced by hypertonic saline injection, and pre‐ and during‐isotonic saline injection into the right lumbar region. Electromyography signals were recorded bilaterally from the longissimus muscles. Signals were decomposed into individual motor unit action potential trains using advanced algorithms and clustered using the K‐means method.

Results

The hypertonic saline injection successfully induced moderate pain on the numerical pain rating scale (NPRS) (median = 4.71; interquartile range (IQR) = 1.61), which was significantly higher than the isotonic control condition (NPRS median = 1.50; IQR = 1.00; p < 0.001). During the pain condition, significant lateralized adaptations were observed: motor units within the contralateral (left) longissimus muscle exhibited an increased firing rate (p = 0.020, r = 0.33), with no significant changes in the recruitment threshold or motor unit action potential amplitude. On the ipsilateral side, subtle modulations in recruitment threshold were detected. In contrast, the isotonic condition elicited only localized and nonsystemic neuromuscular changes, without the coordinated bilateral organization observed under acute pain.

Discussion

Acute low back pain induces a dynamic and lateralized modulation of motor unit recruitment, suggesting a compensatory redistribution of neural drive that may help preserve functional stability despite altered motor coordination. These findings provide novel insights into the neuromuscular adaptations underlying acute low back pain and have potential implications for physiotherapy assessment and intervention.

Keywords: electromyography, force steadiness, isometric contraction, low back pain, motor units, neuromuscular adaptation

1. Introduction

Low back pain is a highly prevalent musculoskeletal condition that compromises motor control and functional performance (Van Dieën et al. 2019a, 2019b). Evidence suggests that pain increases variability in muscle force control during painful contractions, highlighting the importance of these adaptations for functional performance (Schabrun et al. 2016; Arvanitidis et al. 2025). Despite advancements, there remains a significant gap in understanding how changes induced by low back pain manifest in force steadiness and motor unit pool organization during submaximal tasks (Arvanitidis et al. 2025). Previous investigations have reported divergent mechanisms in motor control in individuals with low back pain (Van Dieën et al. 2019a; Meier et al. 2018), but the role of selective redistribution of neural drive has not been fully explored.

Recent advances in electromyographic signal decomposition have enabled precise analysis of individual motor unit behavior during dynamic contractions, providing detailed insights into central nervous system strategies for force production modulation (Nawab et al. 2010; De Luca et al. 2006; De Luca and Contessa 2015). However, the interpretation of these data can vary significantly depending on the methodological approach used. It is essential to apply techniques that ensure greater precision in characterizing the motor unit pools (Martinez‐Valdes et al. 2020; Farina et al. 2004). While initial studies using surface EMG decomposition have provided valuable normative data on lumbar motor unit behavior, analyzing these complex signals during pain requires further refinement to avoid oversimplifying the neuromuscular response (Silva et al. 2017).

Force steadiness, typically quantified by the coefficient of variation of force, is primarily determined by the variability in motor unit discharge rates and the presence of low‐frequency oscillations in the common neural drive (Enoka and Farina 2021). Since acute pain has been shown to perturb the synaptic input to the motoneuron pool and alter the discharge behavior of active motor units (Farina et al. 2014), it is plausible that these neural changes will manifest as a measurable decrease in force steadiness. The method described by (Becker et al. 2024, 2022) allows for more precise and detailed analysis of motor units by segmenting neuromuscular activity data, separating critical events, and providing a clearer view of the interactions between recruitment, firing rate, and synchronization of action potentials.

Thus, our hypothesis is that experimentally induced acute low back pain triggers adaptive changes in force steadiness and motor unit behavior, reflecting an acute modulation of neuromuscular control. This study aimed to investigate the effects of experimentally induced acute low back pain on force steadiness and motor unit discharge characteristics (including recruitment threshold, firing rate, and action potential amplitude) during a sustained trunk extension task. By applying advanced EMG decomposition and clustering techniques, we aimed to provide a more comprehensive understanding of the mechanisms that maintain motor control in the presence of acute low back pain.

2. Methods

2.1. Study Design

This is an experimental study that involved participants attending two visits, with a 7‐day interval between them. The research protocol received ethical approval from the ethics committee of the Faculty of Sport of the University of Porto (CEFADE 28‐2023). All objectives and methodologies were meticulously developed and executed following the guidelines established by the Declaration of Helsinki. Before participation, all volunteers provided informed consent by signing an Informed Consent Form, ensuring their understanding of the objectives and procedures of the study.

2.2. Volunteers

Thirty‐three individuals participated in this study. The inclusion criteria required participants to be healthy individuals aged between 18 and 40 years, with no history of musculoskeletal disorders or recurrent pain in the preceding six weeks. Moreover, they should not have taken anti‐inflammatory or analgesic medication within 24 h before the experiment. Exclusion criteria encompassed individuals with a history of lumbar injury or surgery, pre‐existing chronic conditions such as disc degeneration, or other disorders that could affect muscle function. Pregnant women were also excluded. These criteria were established to ensure a homogeneous sample, thereby reducing the influence of confounding factors on the study outcomes.

The participants ranged in age from 18 to 40 years (mean: 29.06 ± 5.96 years). Among them, eight were women, with an average height of 164.66 ± 6.3 cm, an age of 29.77 ± 6.2 years, and a body mass of 65.14 ± 6.20 kg. The remaining 25 participants were men, with a mean height of 175.20 ± 5.5 cm, an age of 28.79 ± 5.9 years, and a body mass of 80.00 ± 9.3 kg. Height and body mass were assessed using a bioimpedance system (InBody 230, InBody Co. Ltd., Korea).

2.3. Electromyography (EMG) Signals

Surface EMG signals were collected bilaterally from the longissimus muscles using a dEMG Galileo sensor (Delsys, Natick, MA, USA). The sensor measures 23 × 30 mm, weighs 19 g, incorporates four electrodes with a 5‐mm inter‐electrode distance, and was applied to dry skin. All the gears were plugged into an uninterruptible power supply with an external battery to isolate all the possible noise from the net wire. The active site of the sensor placement was cleaned and shaved, and the electrode was placed in the middle of the muscle belly, according to the SENIAM recommendations (Hermens et al. 2000). Further, the quality of the signal was checked based on two principles: the minimal offset amplitude accepted was around 20 and no more than 40 μV and the other caution was to check the signal‐to‐noise ratio, to ensure the quality of the signal collected. Galileo sensors were placed over the right and left longissimus muscles. Data collection was performed using the Delsys EMGWorks 4.8.0 software (Delsys, USA) with a sampling frequency of 2000 Hz.

In accordance with the requirements of the Delsys Precision Decomposition algorithm (De Luca et al. 2006), no digital filtering was applied prior to decomposition, as the algorithm requires unaltered signals to maintain the accuracy of the template matching process. Furthermore, rigorous signal quality protocols were implemented before each data collection session: skin impedance was minimized, and the signal‐to‐noise ratio was verified to ensure high‐fidelity recordings. To eliminate potential interference from the electrical grid, the EMG system was powered by an uninterruptible power supply battery during all collection procedures. The decomposed EMG simulator was used to evaluate the accuracy of the decomposed motor unit, using the algorithm developed by De Luca et al. (2006) and refined by Nawab et al. (2010), utilizing the NeuroMap System software (Delsys, USA). Only motor units with a decomposition accuracy higher than 80% were included in further analysis. To verify the high fidelity of our data, the actual mean decomposition accuracy achieved across all analyzed contractions in this study was 91.2% ± 3.4% (ranging from 84.5% to 96.1%), ensuring an excellent signal‐to‐noise ratio and highly reliable identification of discharge timings.

2.4. Experimental Setup

2.4.1. Force Assessment

The experimental setup is illustrated in Figure 1. Force was measured using a dynamometer equipped with a single‐axis load cell, mounted on a stable metal structure. Participants were positioned in a prone position on a treatment table, with the dynamometer placed at the midline between the superior angles of the scapulae (Parolini et al. 2025). Prior to data collection, volunteers underwent a familiarization protocol consisting of five repetitions of the maximal voluntary contraction (MVC) and three repetitions of the submaximal (20% MVC) trapezoidal task to assess force Steadiness (Arvanitidis et al. 2025; Valentin and Maribo 2014). During these trials, volunteers were asked to perform spinal extension and maintain the isometric contraction for 30 s (Parolini et al. 2025). Rest intervals of 1 min were provided between repetitions.

FIGURE 1.

FIGURE 1

Open in a new tab

Experimental setup: (a) Initial position of the volunteer, (b) Volunteer performing trunk extension, and (c) Screen display of the trapezoidal model, curve (red line) and the superimposed force output (blue line), demonstrating how the visual feedback was provided to ensure compliance with the target 20% MVC force, using Delsys EMG Works 4.8.0 software (Delsys, Natick, MA, USA).

Following a 2‐min rest period, the experimental protocol was conducted: one MVC was performed to establish the reference value, followed by one repetition of the submaximal isometric contraction at 20% of the MVC. This task followed a trapezoidal profile (5s ramp‐up, 20s plateau, and 5s ramp‐down), as shown in Figure 1c. Real‐time visual feedback of the force curve was provided throughout the trials to help participants maintain force steadiness. The same procedure was strictly repeated on the second visit, after a 7‐day interval (Parolini et al. 2025).

2.4.2. Experimentally Induced Acute Low Back Pain

Experimental acute low back pain and a control condition were obtained, respectively, by intramuscular injections of hypertonic and isotonic saline solutions, administered 7 days apart (Kellgren 1938). During the injections, the participants were positioned in the prone position on a trunk wedge with a 20° incline, as illustrated in Figure 1. The inclination was measured in the cephalocaudal direction, while the spine was extended, which helped to promote muscle relaxation. The study employed a randomized crossover design, where each participant received a single 2.5 mL dose (Canestri et al. 2021) of either a hypertonic saline solution (6.0%) to induce pain, or an isotonic saline solution (0.9%) (Graven‐Nielsen 2006), as a placebo, ensuring the test was conducted during the pain phase.

The injections were administered in the intervertebral space between L3 and L4, at the level of the spinous process, located approximately 2–3 cm from the midline (Hermens et al. 2000). To ensure experimental consistency, the injection was consistently performed on the right side of the body for all participants. The bolus was delivered into the multifidus muscle due to its primary stabilizing role and its known tendency for reflex inhibition under nociceptive conditions (Devecchi et al. 2022; MacDonald et al. 2006). This standardization was adopted because trunk extension is a bilateral, synergistic task in which paraspinal muscles operate as a functional unit. Previous studies have demonstrated that lumbar muscle activation is highly symmetrical during isometric contractions, with minimal influence of limb dominance (McGill 1991; Hodges and Richardson 1996; Ng et al. 2002).

On the other hand, the longissimus was monitored with EMG because it is a superficial muscle, highly involved in the extension of the spine and frequently recruited in a compensatory manner when there is dysfunction in the deep muscles (Raabe and Chaudhari 2018). After cleaning the area with 70% alcohol, the solution was injected into the right side of the body (ipsilateral side) using a 25 × 28 mm needle, 25 gauge, at a depth of 30 mm from the skin surface. The participants were blinded to the solution type. The order of injections (hypertonic‐pain vs. Isotonic‐control) was randomized, with a 7‐day washout period between sessions, while performing a spinal extension task in the prone position.

2.5. Data Analysis—Analyses Force Steadiness and Clustering, EMG

For the analysis of force steadiness, a 10‐s stable window was systematically selected from the center of the trapezoidal force plateau (from the 10th to the 20th second of the task; Figure 2). This specific duration was selected to ensure a steady‐state signal, avoiding the initial and final periods of the plateau where force fluctuations are typically more pronounced. This segment provides a sufficient period for a reliable estimation of mean firing rates and recruitment thresholds while maintaining high decomposition accuracy. Additionally, force steadiness was quantified as the coefficient of variation (CV) of the force signal (CV = standard deviation/mean force × 100) during this same 10‐s window. All force data were normalized to the participant's MVC before analysis to ensure experimental consistency.

FIGURE 2.

FIGURE 2

Open in a new tab

Schematic representation of the data analysis workflow adopted in the current work.

The clustering method (Becker et al. 2024, 2022) utilized firing rate (FR), recruitment threshold (RT), and motor unit action potential amplitude (MUAPAMP) as input variables for the K‐means algorithm. A signal processing routine was developed in MATLAB (R2022a, Natick, MA, USA). This process partitioned the motor unit pool of each muscle into two distinct clusters, labeled G1 and G2, reflecting the physiological diversity of the motoneuron pool. Following the framework described by Becker et al. (2022, 2024), Cluster G1 (red) represents ‘low threshold’ motor units. These are characterized by earlier recruitment, lower discharge rates, and smaller amplitudes, which are typical of Type I (slow‐twitch) fibers responsible for sustained postural control. In contrast, Cluster G2 (blue) represents ‘high‐threshold’ motor units, characterized by later recruitment and larger amplitudes, consistent with Type II (fast‐twitch) fibers that are typically mobilized for force adjustments or as compensatory strategies during painful contraction (Becker et al. 2024, 2022). A two‐cluster model was selected a priori based on this physiological distribution of the lumbar motoneuron pool. To mathematically validate this specific two‐cluster configuration and ensure appropriate cluster separation across our sample, a Silhouette analysis was performed. The average Silhouette coefficient achieved was 0.68, confirming a robust and well‐defined cluster structure without significant overlapping, thereby justifying the binary classification framework.

2.6. Statistical Analysis

The Shapiro‐Wilk test was used to assess the normality of the data, and because the assumption of normality was not met, nonparametric tests were used for statistical analysis. The Friedman test was applied to compare the four experimental conditions (pre‐hypertonic, during pain, pre‐isotonic, and during‐isotonic) regarding changes in motor unit behavior (FR, RT, and MUAPAMP). Additionally, the Wilcoxon signed‐rank test was used to assess intra‐group differences for cluster G1 and cluster G2 under each experimental condition. To control for Type I error inflation due to multiple comparisons across clusters, sides, and conditions, a formal Bonferroni‐Dunn post hoc correction was applied to all pairwise comparisons, and adjusted p‐values are reported. The effect size for all nonparametric pairwise comparisons was calculated using the Wilcoxon signed‐rank test and expressed as r, and computed as r = Z/√N, where Z is the standardized test statistic and N is the total number of observations. Effect size magnitude was interpreted according to non‐parametric criteria. A post hoc statistical power analysis was integrated into our design rationale; based on our final sample size (N = 33) and the observed nonparametric effect sizes for our primary outcomes (firing rate shifts), the achieved statistical power estimated via G*Power 3.1.7 (University of Kiel, Germany) exceeded 0.80, validating that the study was sufficiently powered to detect robust neuromuscular adaptations despite the conservative multi‐comparison correction. All statistical analyses were conducted using SPSS Statistics (IBM Corporation, Version 27), with a significance level set at α = 0.05. Data are presented as medians and interquartile ranges (parentheses), as they did not follow a normal distribution.

3. Results

3.1. Pain Intensity and Profile

Pain intensity, assessed via the Numerical Pain Rating Scale (NPRS), significantly increased during the hypertonic (pain) condition compared to the no‐pain baseline (median = 4.71; IQR = 1.61 vs. 0.00; p < 0.001). Similarly, the isotonic (placebo) condition elicited higher scores than the no‐pain condition (median = 1.50; IQR = 1.00; p < 0.001), though scores were significantly lower than in the hypertonic condition. To ensure the integrity of the force signal and the motor unit decomposition, participants were not interrupted for verbal pain scoring during the 20‐s trapezoidal contraction. However, monitoring immediately before and every minute after the task confirmed that pain remained stable and at its peak throughout the effort. As illustrated in Figure 3, the nociceptive stimulus peaked within the first 2 min post‐injection and showed a progressive decay thereafter.

FIGURE 3.

FIGURE 3

Open in a new tab

Temporal profile of pain intensity. The graph displays the group median and interquartile range (IQR) of pain intensity, measured via the Numerical Pain Rating Scale (NPRS, 0–10). Following the hypertonic saline injection (red line), pain peaked within the first 2 minutes and gradually declined until complete resolution at the seventh minute. The shaded area indicates the trapezoidal contraction task, confirming that the 10‐s neuromuscular analysis window was synchronized with the peak and stable phase of the experimental pain stimulus. The isotonic condition (blue line) served as a control, reaching lower peak values and resolving significantly faster.

All participants reported complete pain resolution (NPRS = 0) within 7 min, confirming that the 10‐s analysis window (seconds 10–20 of the plateau) was perfectly synchronized with the period of maximum sustained pain.

3.2. Clustering of Motor Units

RT was the most significant variable for the clustering of motor units, as shown in Figure 4. Since the clustering method used three variables, we employed a three‐dimensional plot to visualize their interaction in both conditions, hypertonic and isotonic (Figure 4a,b, respectively). This spatial distribution effectively highlights the mathematical separation between the low‐threshold (G1) and high‐threshold (G2) motor unit cohorts. To contextualize how this clustered neural drive impacts mechanical output, the coefficient of variation of the force signal was concurrently evaluated across the same experimental conditions, as detailed sequentially in Figure 5.

FIGURE 4.

FIGURE 4

Open in a new tab

Three‐dimensional plots of the three variables considered in the clustering analyses. (a) Clusters of MUs in hypertonic condition. (b) Clusters of MUs in isotonic condition. Each point represents a motor unit, identified as belonging to Group 1 (red) or Group 2 (blue), according to the recruitment threshold value. Data represent states for Pre hypertonic or isotonic, During hypertonic, and During isotonic. MUs, motor units; MVC, maximum voluntary contraction; pps, pulses per second.

FIGURE 5.

FIGURE 5

Open in a new tab

The graph presents the effect of acute pain on force steadiness. Violin plots represent the distribution of force CV (%), with the central horizontal dashed line representing the median and outer dotted lines representing the interquartile range (IQR). A significant shift in median force CV was observed during the acute low back pain (hypertonic saline‐induced) condition (p = 0.011), while no significant effect was found in the isotonic (no‐pain control) protocol. *indicates p < 0.05.

3.3. The Effect of Pain in Force Steadiness

Force steadiness, quantified by the coefficient of variation of the force signal, exhibited a median value of 0.55% (IQR: 0.51) during the pre‐hypertonic (baseline) condition. During the acute induced low back pain (hypertonic) condition, the median CV shifted to 0.65% (IQR: 0.72). Statistical analysis revealed a significant modulation in force steadiness during the pain condition (p = 0.011, Figure 5). In contrast, the isotonic (no‐pain) protocol did not elicit a significant effect on force steadiness, maintaining stable coefficient of variation values across the baseline and post‐injection states (Figure 5).

3.4. Pain Effect on Motor Units' Pool

In G1, when comparing the conditions before and during the induced pain, statistically significant differences were observed in the left longissimus muscle (contralateral to the injection), as shown in Figure 6, for the firing rate (FR) (p = 0.020, r = 0.33). No significant differences were found for recruitment threshold (RT) (p = 0.151, r = 0.25) or motor unit action potential amplitude (MUAPAMP) (p = 0.345, r = 0.16). In the right longissimus muscle (ipsilateral to the injection site), no significant differences were observed for RT (p = 0.783, r = 0.03), FR (p = 0.709, r = 0.04), or MUAPAMP (p = 0.437, r = 0.13). In the isotonic condition, on the contralateral side, only MUAPAMP showed a significant difference (p = 0.003, r = 0.51), while RT (p = 0.156, r = 0.19) and FR (p = 0.519, r = 0.10) were not significantly different. On the ipsilateral side, RT showed a significant difference (p = 0.023, r = 0.33), whereas FR (p = 0.013, r = 0.36) and MUAPAMP (p = 0.333, r = 0.06).

FIGURE 6.

FIGURE 6

Open in a new tab

Motor unit (MU) properties during hypertonic (pain) and isotonic (control) protocols for both the Ipsilateral (Right) and Contralateral (Left) muscles. (a) Firing rate, expressed in pulses per second (pps). (b) Recruitment threshold, expressed as a percentage of maximal voluntary contraction (% MVC). (c) Motor unit action potential amplitude (MUAPAMP), expressed in millivolts (mV). Red violins represent low‐threshold units (G1) and blue violins represent high‐threshold units (G2). Dashed lines within violins indicate the median and dotted lines indicate the interquartile range (IQR). *p < 0.05; **p < 0.01.

In G2, in the pre and during hypertonic pain condition, significant differences were observed for FR in the left longissimus (p = 0.003, r = 0.35) and MUAPAMP (p = 0.050, r = 0.25), while RT showed no significant differences (p = 0.155, r = 0.19). On the ipsilateral side, RT was significantly altered (p = 0.025, r = 0.28), whereas FR (p = 0.142, r = 0.18) and MUAPAMP (p < 0.01, r = 0.42) showed modulation. In the isotonic condition, on the contralateral side, were observed in RT (p < 0.001, r = 0.12), FR (p = 0.013, r = 0.36), or MUAPAMP (p = 0.333, r = 0.06). On the ipsilateral side, no significant differences were found for RT (p = 0.825, r = 0.05), FR (p = 0.519, r = 0.10), or MUAPAMP (p = 0.652, r = 0.08).

4. Discussion

The results of this study provide substantial support for our hypothesis that experimentally induced acute low back pain triggers selective, asymmetric modulations in motor unit discharge patterns, which directly reflects the observed reduction in force steadiness. The analysis of motor unit firing patterns was grounded in the pain adaptation model, which suggests that the nervous system modulates muscle activation to protect painful structures while maintaining overall task performance (Hodges and Tucker 2011). This lateralized effect of pain was confirmed by effect size analysis. In Cluster G1, the transition from an isotonic stimulus to a painful one shifted the moderate effect of firing rate from the ipsilateral side (r = 0.461) to the contralateral side (r = 0.516). This finding is highly consistent with a redistribution of neural drive away from the nociceptive site to preserve spinal stability (Hodges and Tucker 2011). Interestingly, this contralateral activation pattern in FR suggests an adaptive modulation of motor control, potentially involving changes in motor unit synchronicity (Meier et al. 2018; Tsao et al. 2008) to prioritize functional stability.

It is critical to note that significant, albeit transient, alterations in recruitment thresholds and firing rates were also observed during the isotonic condition, particularly across clusters and muscle sides. These responses were generally smaller in magnitude and less consistent than those observed during the hypertonic condition, suggesting a combination of procedural and non‐specific neuromuscular influences rather than a true nociceptive‐driven adaptation. The mechanical stimulation associated with needle insertion and tissue perturbation likely contributed to a localized protective response, reflected in subtle modulations of motor unit behavior (Graven‐Nielsen 2006). However, unlike the hypertonic condition, the isotonic protocol did not elicit a coherent or systematically organized pattern of bilateral modulation, reinforcing the interpretation that these changes are predominantly non‐nociceptive and related to experimental manipulation rather than pain‐induced neural reorganization.

Even under a low‐intensity challenge, the presence of pain induces specific alterations in neuromuscular activity, with distinct responses in the ipsilateral and contralateral muscles relative to the pain induction site. Analysis of clustered groups G1 and G2 revealed significant changes in electrophysiological parameters, particularly in the FR and the MUAPAMP. In the left longissimus muscle (contralateral to pain induction), an increase in FR was observed in the G1 group, suggesting that the central nervous system may redistribute neural drive toward the non‐painful side as a compensatory strategy (Van Dieën et al. 2019a; Martinez‐Valdes et al. 2020; Zhang et al. 2024). This response may be related to the predominance of type 1 fibers in the lumbar musculature, which are known for their stabilizing function and gradual recruitment (Agten et al. 2020; Sĭrca and Kostevc 1985).

However, the interpretation of a definitive ‘lateralized neuromuscular adaptation’ must be made with caution. While contralateral changes in motor unit firing rates were evident, these adaptations were not accompanied by uniform, parallel shifts in recruitment thresholds or motor unit action potential amplitudes across both clusters (Martinez‐Valdes et al. 2020; Farina et al. 2004). Furthermore, the mixed outcomes yielded by the cluster analyses suggest that the observed changes reflect a highly selective, non‐uniform redistribution of neural drive rather than evidence of a structural reorganization of spinal motor pathways (Becker et al. 2024, 2022). Therefore, these findings should be interpreted as acute, localized compensatory adjustments designed to redistribute mechanical stress under nociceptive threat (Hodges and Tucker 2011), rather than a permanent or comprehensive reorganization of spinal motor control pathways.

A notable finding in this study was the distinct lateralized behavior observed in Cluster G1. Specifically, while the pain was induced on the right side, G1 participants exhibited a significant modulation of motor unit firing rates in the contralateral (left) longissimus. This lateralization suggests a sophisticated neural strategy aimed at shifting the mechanical demand away from the nociceptive source to preserve spinal stability. Such findings are in line with the ‘redistribution of activity' hypothesis, where the central nervous system optimizes muscle recruitment patterns to minimize pain while fulfilling the task requirements (Hodges and Tucker 2011). The fact that this was more evident in G1 than in G2 indicates that neuromuscular signatures of pain are not uniform across individuals, pointing toward different ‘phenotypes’ of motor adaptation (Hodges and Tucker 2011). Contralateral activation pattern FR suggests an adaptive modulation of motor control: a possible change in synchronicity (Meier et al. 2018; Tsao et al. 2008). The central nervous system appears to prioritize functional stability by selectively increasing the activation of both groups of Mus on the contralateral side, elucidating again a compensatory strategy to preserve neuromuscular function (Zhang et al. 2024; Wang et al. 2023).

The observed changes in FR and MUAPAMP together indicate a dynamic process of acute neural modulation in which the adjustment of neural activation aims to preserve motor execution (Brumagne et al. 2019). The differential modulation of muscle activity between the ipsilateral and contralateral sides suggests a refined adaptive mechanism within the central nervous system (Heming et al. 2019). The redistribution of motor unit activation may reflect adjustments in motor control in response to pain, thereby maintaining functional stability (Brumagne et al. 2019; Schmid et al. 2021). This differential response may result from both peripheral influences, such as changes in motor unit excitability, and central mechanisms involved in motor control adjustments (Meier et al. 2018; Cheung et al. 2005; Corti et al. 2022). On the ipsilateral side to pain induction (right longissimus muscle), the changes were more subtle, with no major variations in FR or MUAPAMP amplitude, but with adjustments in recruitment latency. This finding suggests that although FR modulation was less pronounced than on the contralateral side, the central nervous system adjusts the muscular response to local nociceptive input (Martinez‐Valdes et al. 2020; Farina et al. 2004; Sterling et al. 2001). The absence of substantial FR modifications suggests that the central nervous system maintains a baseline level of ipsilateral motor regulation, possibly to prevent interference that could compromise overall movement coordination (Tazoe and Perez 2014; Porcaro et al. 2021).

The observed lateralized adjustments have important implications for understanding motor control mechanisms in the presence of pain. Previous studies have shown that pain is often associated with functional deficits and reduced force‐generation capacity (Van Dieën et al. 2019a; Becker et al. 2022; Hodges and Richardson 1999). However, our results indicate that pain not only affects force steadiness, but also induces an asymmetric redistribution of muscle activation, reflecting a neurophysiological adaptation aimed at preserving functional stability despite the limitations imposed by pain (Arvanitidis et al. 2025). Thus, even during submaximal contractions, acute low back pain significantly impacts force steadiness and variability. These changes are dynamic and affect not only the ability to generate force, but also its accuracy and consistency over time. Our findings suggest that experimentally induced acute low back pain promotes a redistribution of muscle activation, resulting in a lateralized modulation of force steadiness control. This directly affects the neural mechanisms involved in motor control and the maintenance of postural stability.

4.1. Limitations

The limitations of this study include the cross‐sectional nature of the experimental design, which does not allow the evolution of neural responses over time to be assessed, and the fact that the results were obtained in a controlled environment, which may limit generalization to natural conditions of acute low back pain. Furthermore, the evaluation was restricted to a single experimental pain model, which may not reflect the full complexity of acute and chronic low back pain in clinical populations.

Notably, this study did not account for the potential influence of psychosocial factors, sex differences, or genetic predispositions. As explored in our recent work on the interplay between mental health and pain definitions (Parolini et al. 2023) and the quantification of pain‐related facial expressions, pain is a multidimensional experience (Parolini et al. 2026). The absence of these variables limits the broader interpretation of the observed phenotypes. Future research should integrate these multidimensional factors to better characterize individual signatures of redistribution of neural drive. Although the observed changes suggest acute adaptations in motor unit behavior, the cross‐sectional design and short duration of the experimental pain model do not allow conclusions regarding long‐term neuroplastic changes or structural reorganization of the motor system.

4.2. Future Research

Future studies could investigate how the chronicity of pain modifies the adaptations observed in acute pain, as well as examine the temporal modulation and intensity of pain. It would also be relevant to explore compensatory motor recruitment strategies and postural adjustments to mitigate pain. The longitudinal analysis of therapeutic interventions and the impact of visual and proprioceptive feedback on neuromuscular recovery are also promising areas.

5. Conclusion

In conclusion, acute low back pain acts as a potent modulator of the motor system, intensifying asymmetric neuromuscular adjustments and affecting force steadiness. While minor adaptations were observed in the control condition likely reflecting a nonspecific protective response to the mechanical stimulus of the needle, the presence of chemical nociception triggered a distinct, more robust, and strategically organized electrophysiological signature. These findings suggest that acute pain induces a selective modulation of motor unit activity, particularly evident in distinct motor phenotypes (clusters), aimed at maintaining functional stability while potentially offloading the painful area.

5.1. Implications of Physiotherapy Practice

The selective modulation of motor unit firing rates observed in this study provides key translational insights for physical therapy and low back pain rehabilitation. Rather than inducing a generalized inhibition of the lumbar spine musculature, acute nociception triggered an asymmetric, cluster‐specific redistribution of neural drive. Clinically, this suggests that acute low back pain patients naturally adopt lateralized, offloading motor strategies even during symmetrical functional tasks (such as bilateral spinal extension). For clinicians and physiotherapists, these findings underscore that rehabilitation in the early stages of pain should move away from generic, global core‐strengthening programs. Instead, therapeutic interventions should target the resolution of asymmetric motor unit recruitment and lateralized motor control imbalances. Addressing these acute, protective neural adaptations early through targeted movement retraining and unilateral motor control exercises may be crucial to prevent the consolidation of long‐term compensatory movement patterns, which are highly associated with chronic mechanical overloading and recurrent low back pain.

Funding

This research was funded by FCT Portuguese national funding agency for science, research and technology, Grant No. UI/BD/151415/2021.

Ethics Statement

This study was approved by the Ethics Committee of the Faculty of Sport at the University of Porto (protocol number 28‐2023).

Consent

Informed consent was obtained from all participants involved in the study.

Conflicts of Interest

The authors declare no conflicts of interest.

Acknowledgments

During the preparation of this manuscript, the authors utilized artificial intelligence tools solely for data visualization purposes, specifically to generate and refine the code used for plotting the figures (including Figures 2 and 3). No AI tools were used for data collection, statistical analysis, literature interpretation, or the drafting and editing of the manuscript text. The authors reviewed and verified all final outputs, maintaining full human oversight, and take sole responsibility for the originality, scientific integrity, accuracy, and references of the work.

Data Availability Statement

The data presented in this study are available on request from the corresponding author.

References

  1. Agten, A. , Stevens S., Verbrugghe J., Eijnde B. O., Timmermans A., and Vandenabeele F.. 2020. “The Lumbar Multifidus Is Characterised by Larger Type I Muscle Fibres Compared to the Erector Spinae.” Anatomy & Cell Biology 53, no. 2: 143–150. 10.5115/acb.20.009. [DOI] [PMC free article] [PubMed] [Google Scholar]
  2. Arvanitidis, M. , Falla D., Sanderson A., and Martinez‐Valdes E.. 2025. “Does Pain Influence Control of Muscle Force? A Systematic Review and Meta‐Analysis.” European Journal of Pain 29, no. 2: e4716. 10.1002/ejp.4716. [DOI] [PMC free article] [PubMed] [Google Scholar]
  3. Becker, K. , Goethel M., Fonseca P., Vilas‐Boas J. P., and Ervilha U.. 2022. “The Strategy of the Brain to Maintain the Force Production in Painful Contractions‐A Motor Units Pool Reorganization.” Cells 11, no. 20: 3299. 10.3390/cells11203299. [DOI] [PMC free article] [PubMed] [Google Scholar]
  4. Becker, K. , Goethel M., Sousa M., et al. 2024. “The Importance of Analysing a Single Common Window of a Concentric and Eccentric Dynamic Muscle Contraction: Unveiling the Components of EMG Data and Motor Unit Pool.” In Computer Methods in Biomechanics and Biomedical Engineering Imaging & Visualization. [DOI] [PubMed] [Google Scholar]
  5. Brumagne, S. , Diers M., Danneels L., Moseley G. L., and Hodges P. W.. 2019. “Neuroplasticity of Sensorimotor Control in Low Back Pain.” Journal of Orthopaedic & Sports Physical Therapy 49, no. 6: 402–414. 10.2519/jospt.2019.8489. [DOI] [PubMed] [Google Scholar]
  6. Canestri, R. , Franco‐Alvarenga P. E., Brietzke C., et al. 2021. “Effects of Experimentally Induced Muscle Pain on Endurance Performance: A Proof‐of‐Concept Study Assessing Neurophysiological and Perceptual Responses.” Psychophysiology 58, no. 6: e13810. 10.1111/psyp.13810. [DOI] [PubMed] [Google Scholar]
  7. Cheung, V. C. K. , d’Avella A., Tresch M. C., and Bizzi E.. 2005. “Central and Sensory Contributions to the Activation and Organization of Muscle Synergies During Natural Motor Behaviors.” Journal of Neuroscience 25, no. 27: 6419–6434. 10.1523/jneurosci.4904-04.2005. [DOI] [PMC free article] [PubMed] [Google Scholar]
  8. Corti, E. J. , Marinovic W., Nguyen A. T., Gasson N., and Loftus A. M.. 2022. “Motor Cortex Excitability in Chronic Low Back Pain.” Experimental Brain Research 240, no. 12: 3249–3257. 10.1007/s00221-022-06492-7. [DOI] [PMC free article] [PubMed] [Google Scholar]
  9. De Luca, C. J. , Adam A., Wotiz R., Gilmore L. D., and Nawab S. H.. 2006. “Decomposition of Surface EMG Signals.” Journal of neurophysiology 96, no. 3: 1646–1657. 10.1152/jn.00009.2006. [DOI] [PubMed] [Google Scholar]
  10. De Luca, C. J. , and Contessa P.. 2015. “Biomechanical Benefits of the Onion‐Skin Motor Unit Control Scheme.” Journal of Biomechanics 48, no. 2: 195–203. 10.1016/j.jbiomech.2014.12.003. [DOI] [PMC free article] [PubMed] [Google Scholar]
  11. Devecchi, V. , Falla D., Cabral H., and Gallina A.. 2022. “Neuromuscular Adaptations to Experimentally Induced Pain in the Lumbar Region: Systematic Review and Meta‐Analysis.” Pain 164, no. 6: 1159–1180. 10.1097/j.pain.0000000000002819. [DOI] [PubMed] [Google Scholar]
  12. Enoka, R. M. , and Farina D.. 2021. “Force Steadiness: From Motor Units to Voluntary Actions.” Physiology 36, no. 2: 114–130. 10.1152/physiol.00027.2020. [DOI] [PubMed] [Google Scholar]
  13. Farina, D. , Arendt‐Nielsen L., Merletti R., and Graven‐Nielsen T.. 2004. “Effect of Experimental Muscle Pain on Motor Unit Firing Rate and Conduction Velocity.” Journal of Neurophysiology 91, no. 3: 1250–1259. 10.1152/jn.00620.2003. [DOI] [PubMed] [Google Scholar]
  14. Farina, D. , Merletti R., and Enoka R. M.. (1985)2014. “The Extraction of Neural Strategies From the Surface EMG: An Update.” Journal of Applied Physiology 117, no. 11: 1215–1230. 10.1152/japplphysiol.00162.2014. [DOI] [PMC free article] [PubMed] [Google Scholar]
  15. Graven‐Nielsen, T. 2006. “Fundamentals of Muscle Pain, Referred Pain, and Deep Tissue Hyperalgesia.” Supplement, Scandinavian Journal of Rheumatology 35, no. sup122: 1–43. 10.1080/03009740600865980. [DOI] [PubMed] [Google Scholar]
  16. Heming, E. A. , Cross K. P., Takei T., Cook D. J., and Scott S. H.. 2019. “Independent Representations of Ipsilateral and Contralateral Limbs in Primary Motor Cortex.” eLife 8: e48190. 10.7554/elife.48190. [DOI] [PMC free article] [PubMed] [Google Scholar]
  17. Hermens, H. , Freriks B., Disselhorst‐Klug C., and Rau G.. 2000. “Development of Recommendations for SEMG Sensors and Sensor Placement Procedures.” Journal of Electromyography and Kinesiology: official journal of the International Society of Electrophysiological Kinesiology 10, no. 5: 361–374. 10.1016/s1050-6411(00)00027-4. [DOI] [PubMed] [Google Scholar]
  18. Hodges, P. W. , and Richardson C. A.. 1996. “Inefficient Muscular Stabilization of the Lumbar Spine Associated With Low Back Pain. A Motor Control Evaluation of Transversus Abdominis.” Spine 21, no. 22: 2640–2650. 10.1097/00007632-199611150-00014. [DOI] [PubMed] [Google Scholar]
  19. Hodges, P. W. , and Richardson C. A.. 1999. “Altered Trunk Muscle Recruitment in People With Low Back Pain With Upper Limb Movement at Different Speeds.” Archives of Physical Medicine and Rehabilitation 80, no. 9: 1005–1012. 10.1016/s0003-9993(99)90052-7. [DOI] [PubMed] [Google Scholar]
  20. Hodges, P. W. , and Tucker K.. 2011. “Moving Differently in Pain: A New Theory to Explain the Adaptation to Pain.” Supplement, Pain 152, no. 3: S90–S98. 10.1016/j.pain.2010.10.020. [DOI] [PubMed] [Google Scholar]
  21. Van Dieën, J. H , Reeves N. P., Kawchuk G. N., Kawchuk G. N., Van Dillen L. R., and Hodges P. W.. 2019a. “Analysis of Motor Control in Patients With Low Back Pain: A Key to Personalized Care?” Journal of Orthopaedic & Sports Physical Therapy 49, no. 6: 380–388. 10.2519/jospt.2019.7916. [DOI] [PMC free article] [PubMed] [Google Scholar]
  22. Van Dieën, J. H , Reeves N. P., Kawchuk G. N., Kawchuk G. N., Van Dillen L. R., and Hodges P. W.. 2019b. “Motor Control Changes in Low Back Pain: Divergence in Presentations and Mechanisms.” Journal of Orthopaedic & Sports Physical Therapy 49, no. 6: 370–379. PMCID: PMC7393576. 10.2519/jospt.2019.7917. [DOI] [PMC free article] [PubMed] [Google Scholar]
  23. Kellgren, J. H. 1938. “Referred Pains From Muscle.” British Medical Journal 1, no. 4023: 325–327. 10.1136/bmj.1.4023.325. [DOI] [PMC free article] [PubMed] [Google Scholar]
  24. MacDonald, D. A. , Lorimer Moseley G., and Hodges P. W.. 2006. “The Lumbar Multifidus: Does the Evidence Support Clinical Beliefs?” Manual Therapy 11, no. 4: 254–263. 10.1016/j.math.2006.02.004. [DOI] [PubMed] [Google Scholar]
  25. Martinez‐Valdes, E. , Negro F., Farina D., and Falla D.. 2020. “Divergent Response of Low‐ Versus High‐Threshold Motor Units to Experimental Muscle Pain.” Journal of Physiology 598, no. 11: 2093–2108. 10.1113/jp279225. [DOI] [PubMed] [Google Scholar]
  26. McGill, S. M. 1991. “Electromyographic Activity of the Abdominal and Low Back Musculature During the Generation of Isometric and Dynamic Axial Trunk Torque: Implications for Lumbar Mechanics.” Journal of Orthopaedic Research 9, no. 1: 91–103. 10.1002/jor.1100090112. [DOI] [PubMed] [Google Scholar]
  27. Meier, M. L. , Vrana A., and Schweinhardt P.. 2018. “Low Back Pain: The Potential Contribution of Supraspinal Motor Control and Proprioception.” Neuroscientist 25, no. 6: 583–596. 10.1177/1073858418809074. [DOI] [PMC free article] [PubMed] [Google Scholar]
  28. Nawab, S. H. , Chang S.‐S., and De Luca C. J.. 2010. “High‐Yield Decomposition of Surface EMG Signals.” Clinical Neurophysiology 121, no. 10: 1602–1615. 10.1016/j.clinph.2009.11.092. [DOI] [PMC free article] [PubMed] [Google Scholar]
  29. Ng, J. K. , Richardson C. A., Parnianpour M., and Kippers V.. 2002. “EMG Activity of Trunk Muscles and Torque Output During Isometric Axial Rotation Exertion: A Comparison Between Back Pain Patients and Matched Controls.” Journal of Orthopaedic Research 20, no. 1: 112–121. 10.1016/s0736-0266(01)00067-5. [DOI] [PubMed] [Google Scholar]
  30. Parolini, F. , Goethel M., Becker K., et al. 2023. “Breaking Barriers: Artificial Intelligence Interpreting the Interplay Between Mental Illness and Pain as Defined by the International Association for the Study of Pain.” Biomedicines 11, no. 7: 2042. 10.3390/biomedicines11072042. [DOI] [PMC free article] [PubMed] [Google Scholar]
  31. Parolini, F. , Goethel M., Robalino J., et al. 2025. “Precision and Reliability of a Dynamometer for Trunk Extension Strength and Steadiness Assessment.” Applied Sciences 15, no. 8: 4081. 10.3390/app15084081. [DOI] [Google Scholar]
  32. Parolini, F. , Pires R., dos Santos S. D., et al. 2026. “The Face of Low Back Pain: A Preliminary Method for Quantifying Pain‐Related Facial Expressions.” Applied Sciences 16, no. 6: 2830. 10.3390/app16062830. [DOI] [Google Scholar]
  33. Porcaro, C. , Mayhew S. D., and Bagshaw A. P.. 2021. “Role of the Ipsilateral Primary Motor Cortex in the Visuo‐Motor Network During Fine Contractions and Accurate Performance.” International Journal of Neural Systems 31, no. 6: 2150011. 10.1142/s0129065721500118. [DOI] [PubMed] [Google Scholar]
  34. Raabe, M. , and Chaudhari A.. 2018. “Biomechanical Consequences of Running With Deep Core Muscle Weakness.” Journal of Biomechanics 67: 98–105. 10.1016/j.jbiomech.2017.11.037. [DOI] [PubMed] [Google Scholar]
  35. Schabrun, S. M. , Christensen S. W. M., Mrachacz‐Kersting N., and Graven‐Nielsen T.. 2016. “Motor Cortex Reorganization and Impaired Function in the Transition to Sustained Muscle Pain.” Cerebral Cortex 26, no. 5: 1878–1890. 10.1093/cercor/bhu319. [DOI] [PubMed] [Google Scholar]
  36. Schmid, S. , Bangerter C., Schweinhardt P., and Meier M. L.. 2021. “Identifying Motor Control Strategies and Their Role in Low Back Pain: A Cross‐Disciplinary Approach Bridging Neurosciences With Movement Biomechanics.” Frontiers in Pain Research 2: 715219. 10.3389/fpain.2021.715219. [DOI] [PMC free article] [PubMed] [Google Scholar]
  37. Silva, M. F. , Dias J. M., Pereira L. M., et al. 2017. “Determination of the Motor Unit Behavior of Lumbar Erector Spinae Muscles Through Surface EMG Decomposition Technology in Healthy Female Subjects.” Muscle & Nerve 55, no. 1: 28–34. 10.1002/mus.25184. [DOI] [PubMed] [Google Scholar]
  38. Sĭrca, A. , and Kostevc V.. 1985. “The Fibre Type Composition of Thoracic and Lumbar Paravertebral Muscles in Man.” Journal of Anatomy 141: 131–137. PMCID: PMC1166395; PMID: 2934358. [PMC free article] [PubMed] [Google Scholar]
  39. Sterling, M. , Jull G., and Wright A.. 2001. “The Effect of Musculoskeletal Pain on Motor Activity and Control.” Journal of Pain 2, no. 3: 135–145. 10.1054/jpai.2001.19951. [DOI] [PubMed] [Google Scholar]
  40. Tazoe, T. , and Perez M. A.. 2014. “Selective Activation of Ipsilateral Motor Pathways in Intact Humans.” Journal of Neuroscience 34, no. 42: 13924–13934. 10.1523/jneurosci.1648-14.2014. [DOI] [PMC free article] [PubMed] [Google Scholar]
  41. Tsao, H. , Galea M. P., and Hodges P. W.. 2008. “Reorganization of the Motor Cortex Is Associated With Postural Control Deficits in Recurrent Low Back Pain.” Brain: Journal of Neurology 131, no. Pt 8: 2161–2171. 10.1093/brain/awn154. [DOI] [PubMed] [Google Scholar]
  42. Valentin, G. , and Maribo T.. 2014. “Hand‐Held Dynamometry Fixated With a Tripod Is Reliable for Assessment of Back Extensor Strength in Women With Osteoporosis.” Osteoporosis International 25, no. 8: 2143–2149. 10.1007/s00198-014-2743-0. [DOI] [PubMed] [Google Scholar]
  43. Wang, X. , Luo Z., Zhang M., et al. 2023. “The Interaction Between Changes of Muscle Activation and Cortical Network Dynamics During Isometric Elbow Contraction: A sEMG and fNIRS Study.” Frontiers in Bioengineering and Biotechnology 11: 1176054. 10.3389/fbioe.2023.1176054. [DOI] [PMC free article] [PubMed] [Google Scholar]
  44. Zhang, C. , Li Y., Zhang Z., et al. 2024. “Motor Control Exercise Modulates the Neural Plasticity of the Default Mode Network in Patients With Chronic Low Back Pain.” Pain Physician 27, no. 1: E55–E64. PMID: 38285031. [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Data Availability Statement

The data presented in this study are available on request from the corresponding author.

Articles from Physiotherapy Research International are provided here courtesy of Wiley

RESOURCES