Beyond Muscle Quantity: Rethinking Sarcopenia as a Disorder of Brain-Muscle Communication?

The Three-Level Neuromuscular Organization

Framework 
Neural control of movement is a multilevel process that unfolds across interconnected stages rather than a process reducible to any single measure. Thus, we organize the current evidence around three complementary levels, each accessible through distinct neurophysiological approaches, and we argue that each one is necessary, yet insufficient on its own, to capture the neural mechanisms contributing to sarcopenia.

Brain-muscle interaction can be characterized through a range of functional connectivity metrics that quantify statistical interdependence between simultaneously recorded neural and peripheral electrophysiological signals, each with distinct assumptions and limitations (2, 15). Among these, corticomuscular coherence (CMC) quantifies the frequency-domain oscillatory synchronization between cortical activity measured by electroencephalography (EEG) and muscular activity measured by surface bipolar electromyography (EMG). CMC findings in aging are genuinely heterogeneous: studies report both increases and decreases in CMC in older relative to younger adults, with outcomes depending largely on the muscle studied, the task constraints, and the frequency band examined (14, 18). CMC is nonetheless the most widely used metric in this context, generally treated as a marker of aging (19) as well as of effective corticospinal interaction and neural drive to skeletal muscle (15). 

Classically computed CMC reflects the pooled and rectified discharge of an entire MU population and cannot disentangle individual MU activity. This is a genuine methodological ceiling, because the MU is the final common pathway for force generation, and aging is known to alter MU behavior directly: reduced excitatory synaptic input (24), diminished intrinsic motoneuron excitability (26), and structural remodeling (30) collectively reduce discharge rates and impair the muscle’s capacity for volitional force modulation (27). Recent decomposition algorithms applied to high-density surface EMG (HDsEMG) now allow non-invasive identification of large MU populations, opening a new observation window onto motoneuron behavior in aging and training (4).

The methodological leap that matters most for the argument of this concept study is the recent coupling of HDsEMG decomposition with simultaneous high-density EEG, enabling coherence to be computed not against the pooled EMG signal but against the firing-event train of a single, identified MU, termed cortico-motoneuronal coherence (CMnC) (38). Such an approach differs substantially from classical CMC, which typically assesses brain and muscle synchrony at the population level. With CMnC, the focus is instead on identifying which specific MUs within a motor-unit pool, presumably heterogeneous in fiber-type composition and recruitment threshold, receive cortically synchronized drive and which do not.

Applied to sarcopenia, this methodology could help distinguish between alternative interpretations of the increased CMC reported in sarcopenic older adults compared to non-sarcopenic individuals (16), of which compensatory upregulation is one plausible but currently unconfirmed hypothesis, alongside alternatives such as reduced afferent modulation discussed below. For instance, CMnC could clarify whether the increased CMC observed in sarcopenic older adults reflects a genuine increase in corticospinal drive, which should manifest as a greater proportion or magnitude of coherent MUs, or instead reflects reduced afferent modulation, which would leave MU coherence largely unchanged. Current MU-extraction methods, however, remain largely limited to static contractions (10) constraining this distinction in practice. To our knowledge, this methodology has not yet been applied to a sarcopenic population and therefore represents a clear and immediate next step toward disentangling the neural mechanisms underlying the increased CMC observed in sarcopenia.

Independent MU evidence provides further support for the ecological-validity argument developed in this manuscript. In a recent preprint investigation, older adults with sarcopenia displayed greater MU discharge variability and altered common synaptic input specifically under dual-task conditions, relative to non-sarcopenic controls and master athletes (28). This is important for two reasons. First, it shows that MU alterations in sarcopenia are at least partly task-dependent, potentially latent under single-task isometric conditions and apparent only under added cognitive load. Second, it converges with (and helps interpret) the CMC literature’s own well-documented sensitivity to task constraints (15), suggesting that both the population-level (CMC) and single MU signatures of sarcopenia may share a common dependency on ecological task demand.

That is, whereas corticospinal drive and MU behavior characterize the neural control of individual muscles, functional movement requires the coordination of activity across multiple muscles. In this context, muscle synergies address this higher-order level of motor organization by capturing low-dimensional, spatiotemporal patterns of coordinated multi-muscle activation, typically derived through non-negative matrix factorization of several bipolar surface EMG sensors (1).

Notably, aging is associated with measurable reorganization of muscle synergies during walking: changes in number, structure, and temporal activation generally interpreted as reduced motor flexibility, particularly evident under increased task difficulty (8, 32). However, it remains unknown whether sarcopenia is associated with a reorganization of muscle synergies that extends beyond, or differs qualitatively from, the effects of chronological aging, and whether such reorganization is exacerbated under dual-task conditions. To date, and to our knowledge, no study has examined muscle synergy structure in a sarcopenia-defined population or linked synergy-level organization with concurrently measured corticospinal drive or MU behavior within the same individuals. A critical next step for the field will be to integrate muscle synergy analysis with measures of brain-muscular communication and MU behavior within the same participants and during locomotor tasks.

Beyond these three levels, remodeling at the neuromuscular junction itself, including axonal sprouting and cyclical denervation-reinnervation (30), represents a complementary, peripheral substrate of impaired brain-muscle communication that lies outside the connectivity‑based framework developed here (23, 30, 31). At the cellular level, alterations remain conceptual at this stage but may independently contribute to the disproportionate loss of muscle strength and function in sarcopenia and interact with the neural mechanisms outlined above. For instance, altered intracellular Ca2+ handling and excitation-contraction coupling, together with impaired sarcolemmal excitability, may directly constrain force generation and transmission independently of neural drive (22). Muscle excitability, for example, depends in part on Na+/K+-ATPase activity, which is stimulated by β2-adrenergic receptor signaling and by calcitonin gene-related peptide released from intact sensory nerve endings, but can be inhibited by reactive oxygen species (ROS) (36). This provides a potential point of convergence through which sensory innervation and redox state may jointly influence a key electrophysiological determinant of force development. Mitochondrial dysfunction may further amplify these disturbances through impaired energy availability and increased ROS production (22), with redox imbalance potentially further compromising Na+/K+-ATPase activity and muscle excitability (36). These mechanisms remain conceptual, as do the central mechanisms proposed here. Nevertheless, they emphasize that the neural framework developed in this manuscript is not intended to replace peripheral or cellular explanations of sarcopenia-related weakness, but rather to represent one complementary layer within a broader, multifactorial model.
Interestingly, a recurring theme across all three levels of the neuromuscular organization is representesd by task-dependence. CMC values are strongly shaped by task constraints: static, isometric, or otherwise constrained paradigms producing connectivity patterns that differ systematically from those observed during dynamic and ecologically valid tasks such as overground or dual-task walking (8, 14, 16). For instance, increased MU discharge variability in sarcopenia appears specifically under dual-task, not single-task, conditions (28), and muscle synergy reorganization with age is most apparent when task difficulty is deliberately increased (32).

This convergence points to a critical methodological limitation: neurophysiological studies of aging and sarcopenia have predominantly relied on isolated motor conditions, including single-joint isometric contractions or seated tasks performed under minimal cognitive load. While these paradigms maximize experimental control and signal-to-noise ratio, they may mask neural alterations that emerge only when motor control is challenged under ecologically relevant conditions, such as walking while processing information or maintaining balance while monitoring one’s surroundings and responding to environmental demands. Many activities of daily living inherently involve the simultaneous performance of motor and cognitive tasks (29), and under such dual-task conditions older adults show greater movement variability, as well as reduced force control and locomotor stability, relative to single-task performance (37).
Mobile brain/body imaging (MoBI) enables the investigation of brain-body dynamics via wearable wireless sensors during unconstrained, ecologically relevant locomotor tasks such as overground walking (15). Previous studies employing MoBI and related neurophysiological approaches have provided insight into efferent neural drive during gait in older adults (3, 8, 14, 16), highlighting its dependence on task complexity. Thus, building on this task-dependence, future studies should consider locomotor single- and dual-task paradigms not simply as ecologically valid extensions of conventional isometric protocols, but as potentially necessary conditions for revealing specific components of the neural alterations associated with sarcopenia. For instance, a single-task, isometric assessment may yield a seemingly unremarkable neurophysiological profile in individuals whose corticomuscular or MU alterations emerge primarily under the combined motor-cognitive dual-task demands encountered repeatedly during real-world everyday activities.

The Conceptual Model
Skeletal muscle is not a homogeneous tissue, but comprises fiber types with distinct contractile, metabolic, and innervation properties. Type II (fast-twitch) fibers are typically associated with larger and higher-threshold MUs, which discharge at lower mean rates and with greater discharge rate variability than lower-threshold MUs (9). Moreover, a substantial body of evidence indicates that aging preferentially affects type II fiber size relative to type I (20). If this selective vulnerability extends to the MUs and fiber populations recruited during everyday tasks that demand a substantial fraction of available force-generating capacity in older adults, such as walking, stair climbing, and sit-to-stand transitions, then the resulting deficit could be disproportionately expressed during these functionally critical movements. This may occur even when the function of low-threshold MUs and the integrity of type I fibers are relatively preserved.

A hypothesis in this context is that the functional consequences of sarcopenia may depend not only on the extent of muscle loss, but also on which MU populations are required to compensate for the resulting reduction in force-generating capacity. Lower-limb muscles are repeatedly challenged during gait, stair climbing, postural transitions, and other weight-bearing activities, often under conditions in which the force demands represent a substantial fraction of an older individual’s available capacity. As maximal force-generating capacity declines, these tasks may increasingly require recruitment of higher-threshold MUs, potentially amplifying the functional consequences of selective vulnerability within fast-twitch fiber populations. This framework predicts that the relationship between muscle mass, MU recruitment, and functional performance may differ across muscle groups and levels of task demand, with the effects becoming most apparent during demanding or ecologically relevant motor behaviors. Importantly, this remains a hypothesis rather than an established mechanism and will require future studies combining measures of muscle morphology, MU behavior, and functional performance to test it directly (figure 1). 

A related but distinct question is directionality: whether selective fiber-type or MU loss underlies impaired neural drive, or conversely, whether altered neural drive contributes to selective peripheral loss, remains unresolved. In the first scenario, primary loss of fast-twitch fibers or their MUs could alter peripheral feedback, such that increased CMC in sarcopenic participants would reflect a compensatory cortical response rather than a primary abnormality.In the second scenario, reduced descending drive to higher‑threshold motor units could chronically under‑recruit fast‑twitch fibers, promoting disuse‑related atrophy-consistent with findings that cortical drive and maximal voluntary activation decline with age (21), and emerging evidence that motor unit and neuromuscular junction impairments are already detectable at the pre-sarcopenic stage, alongside a progressive decline in muscle strength and mass across sarcopenia stages (33). However, conventional CMC metrics assess non-directional, instantaneous brain-muscle synchronization and thus cannot distinguish between these scenarios. Future studies should therefore leverage directional connectivity metrics (i.e., Granger causality, Directed Transfer Function, or Partial Directed Coherence) to disentangle the causal interactions between cortical and muscular activity.

One possibility is that neuromuscular decline in sarcopenia reflects none of the above scenarios, but rather a self-reinforcing interaction between central and peripheral mechanisms. That is, an initial perturbation at either level of the neuromuscular system, whether a modest reduction in descending drive or a modest loss or dysfunction of specific MUs, could alter neural and sensory signaling between these levels, thereby increasing vulnerability at the other level. If such interactions contribute to the progression of neuromuscular decline, this framework might have important implications for future intervention design. Because the underlying process may operate as a self-reinforcing loop rather than a unidirectional cascade, interventions targeting either central or peripheral components could, in principle, influence the other through adaptive changes across the system. This possibility provides a rationale for combining resistance-training approaches with motor-cognitive, neurophysiologically-informed exercise paradigms, as discussed in the next section. However, these proposed mechanisms remain untested in integrated longitudinal designs.

Training and Translational Implications
Resistance training remains the most consistently supported intervention for sarcopenia, with established benefits for muscle strength and more modest effects on muscle mass (17). However, if sarcopenia-related weakness is partly driven by neural dysfunction, resistance training alone may not fully address the underlying neuromuscular impairment. This is particularly important given that neural drive changes have been shown to emerge primarily under dual-task or ecologically relevant demands (3, 28), making combined resistance and motor-cognitive training a plausible complementary strategy. However, it remains unclear whether currently available motor-cognitive training systems adequately integrate structured resistance-training principles. Furthermore, whether such systems have been validated for sarcopenic and older populations is similarly unresolved, and this represents an important translational gap given these populations’ narrow tolerance for non-individualized training loads. Relatedly, neural adaptations to resistance training may themselves contribute substantially to functional improvements in older adults (35) , as they appear to retain the capacity for MU adaptations following strength training, with increases in discharge rate linked to improved force production (4). Thus, neural mechanisms should not be considered secondary to muscle adaptations, and different training modalities may engage distinct aspects of neuromuscular control, particularly when combined with motor-cognitive dual-task demands.

A key next step is therefore to determine whether neurophysiological profiling can help identify which components of neuromuscular function may be most responsive to targeted training. The framework outlined above, integrating CMnC during isometric tasks, MU discharge characteristics during single- and dual-task conditions, and CMC alongside muscle synergy analyses during dynamic, ecologically valid tasks (i.e., overground walking), could provide complementary insights into the mechanisms underlying individual differences in functional impairment. Moreover, such profiling may help identify individuals for whom altered central drive, MU function, or multi-muscle coordination, as well as greater susceptibility to cognitive-motor interference, differentially contribute to functional limitations. This information could then be used to investigate whether training demands (i.e., motor, cognitive, or motor-cognitive components) can be adapted to an individual’s neurophysiological profile, rather than applying a uniform training stimulus across individuals with potentially different neuromuscular constraints.

However, establishing whether such neurophysiologically informed training improves functional outcomes will require prospective longitudinal studies integrating mechanistic profiling with intervention and repeated functional assessments. An important remaining question is whether tailoring training demands to an individual’s neurophysiological profile can enhance strength, mobility, and dual-task adaptations beyond conventional resistance training or standardized motor-cognitive intervention (e.g., exergaming). Ultimately, the goal is not to replace traditional exercise strategies, but to determine whether combining their established peripheral and central effects with task-specific motor-cognitive demands can provide a more targeted approach to the heterogeneous neuromuscular impairments observed in sarcopenia.