Motor Units & Force Grading

A motor unit consists of a single alpha motor neuron in the spinal cord (or brainstem) and every skeletal muscle fiber it innervates via its branching axon. Because every fiber in a motor unit fires together whenever the motor neuron fires — each neuromuscular junction normally produces a suprathreshold end-plate potential — the motor unit is functionally indivisible. It is, in effect, the smallest quantum of controllable force a muscle can produce. A single muscle contains many motor units, typically tens to a few hundred, and each contributes a small, fixed increment of force whenever it is activated. Muscles used for fine, precise movements (extraocular muscles, hand intrinsics) have small innervation ratios — as few as 3–10 fibers per motor neuron — while muscles built for gross, powerful movements (gastrocnemius, gluteus maximus) have innervation ratios in the thousands. Because each motor unit's fibers are of one histochemical type, motor units themselves are classified into types that parallel fiber type: S (slow, fatigue-resistant, small force), FR (fast, fatigue-resistant, intermediate force), and FF (fast, fatigable, large force). This typing was pioneered by Burke and colleagues using glycogen-depletion histochemistry combined with intracellular physiology in cat hindlimb muscle. The type of a motor unit is not an independent property — it is imposed on the muscle fibers by the identity and firing pattern of the innervating motor neuron, a phenomenon proven directly by cross-reinnervation experiments (below). The nervous system has exactly two levers to grade whole-muscle force output: recruitment, the sequential activation of additional motor units, and rate coding, the modulation of firing frequency in units already active. Both levers are graded by the size principle, which links a motor neuron's intrinsic electrical size to its recruitment threshold. Understanding force grading, therefore, is understanding how the size principle organizes recruitment order, how rate coding builds tetanic force once a unit is on, and how fiber-type heterogeneity shapes the force and fatigue landscape a muscle can produce.

Step 1 — Anatomical substrate. A muscle's motor neuron pool resides in the ventral horn of the spinal cord (for limb and trunk muscles) or brainstem motor nuclei (for cranial muscles), organized somatotopically into longitudinal columns dedicated to individual muscles. Each alpha motor neuron sends a single myelinated axon out the ventral root; that axon arborizes and innervates a scattered set of muscle fibers distributed throughout the muscle's cross-section (not clustered together), which is why a single motor unit's contraction produces smooth, spatially averaged force rather than a lopsided twitch.

Step 2 — Motor neuron size varies systematically. Motor neurons innervating a given muscle form a continuum in cell body diameter, dendritic tree size, and axon diameter. Small motor neurons have small cell bodies, thin axons, few synaptic inputs, and high input resistance (Ohm's law: for the same synaptic current, a higher-resistance cell reaches threshold at lower input). Large motor neurons have large somata, thick fast-conducting axons, and low input resistance, requiring much larger synaptic drive to reach firing threshold.

Step 3 — Common drive recruits neurons in size order. Descending commands from the cortex and brainstem, plus segmental reflex inputs (e.g., Ia afferents from muscle spindles), are broadcast relatively uniformly across the motor neuron pool for a given muscle — a principle called common drive. Because recruitment threshold scales inversely with neuron size (Henneman's size principle), as synaptic drive to the pool ramps up, the smallest, highest-resistance motor neurons depolarize to threshold first, followed progressively by larger ones. This produces an obligatory, repeatable recruitment order: S units first, then FR, then FF, regardless of the specific behavioral task (voluntary contraction, stretch reflex, or postural sway) — though descending pathways can bias overall drive to the pool, task-dependent reordering of individual units is rare.

Step 4 — Recruitment adds force in discrete steps. Each newly recruited motor unit adds its own twitch/tetanic force to the whole-muscle output. Because small (S) units are recruited first and are weak, initial force increments are small and finely graded — ideal for delicate tasks. Because large (FF) units are recruited last and are strong, later increments are large — appropriate since fine control is no longer the priority at high force, where speed and power matter more. This is the same logic as a graduated set of weights: small denominations for making change precisely, large denominations for bulk value.

Step 5 — Rate coding adds force within a unit. Once a motor unit is recruited, it does not simply switch on at fixed force — its force output climbs with firing rate through tetanic summation, exactly as in single-fiber excitation-contraction coupling: successive action potentials arrive before the sarcoplasmic reticulum has finished re-sequestering Ca²⁺ and before cross-bridges have fully detached, so contractions summate mechanically (unfused, then fused tetanus). Firing rates for human motor units in voluntary contraction typically range:

Step 6 — The two levers combine across the force range. In most human limb muscles, recruitment accomplishes the bulk of force grading from 0 up to roughly 50–90% of maximal voluntary contraction (the exact crossover varies by muscle — small hand muscles like the first dorsal interosseous complete recruitment by ~50% MVC, while large muscles like the biceps or deltoid continue recruiting new units up to 80–90% MVC). Beyond that point, essentially all motor units in the pool are active, and further force increases come almost entirely from rate coding driving those units toward fused tetanus.

Step 7 — Fatigue and fiber-type feedback into grading. Because S units are recruited constantly (postural, low-force, oxidative-friendly tasks) and FF units are recruited rarely and briefly (fatigable, powerful, glycolytic tasks), the size principle also functions as a fatigue-management strategy: the units most often "on" are precisely the ones biochemically built to sustain activity without fatiguing, while short, forceful bursts recruit units that can afford to fatigue because they are used only briefly.

Henneman's size principle (Elwood Henneman, 1957–1965). Working with anesthetized cats, Henneman and colleagues impaled individual motor axons or recorded intracellularly from motor neuron somata in the spinal cord while electrically stimulating dorsal roots (activating Ia afferents from muscle spindles, the monosynaptic stretch reflex pathway) with graded stimulus intensity. As stimulus strength increased, they recorded the order in which different motor neurons began firing. The consistent finding: motor neurons with smaller cell bodies (identified by lower antidromic action potential thresholds and higher input resistance, measured by intracellular current injection) always recruited before larger ones, across repeated trials and across different behavioral contexts (reflex activation, direct current injection into the motor neuron, and later, voluntary contraction in humans). Crucially, the order was fixed and predictable from a purely passive electrical property (input resistance) — not from descending "intention" or task type. Henneman's 1965 paper in the Journal of Neurophysiology ("Excitability and inhibitability of motoneurons of different sizes") established that recruitment threshold correlates inversely with motor neuron soma size and, by extension, axon diameter — the "size principle."

Milner-Brown, Stein, and Yemm (1973) — human voluntary contraction confirms the size principle. Using fine-wire and surface EMG in humans performing voluntary isometric contractions of the hand, they showed motor units recruited at progressively higher force thresholds also produced progressively larger twitch tensions, and that recruitment order was highly reproducible across repeated contractions in the same subject — direct behavioral confirmation that Henneman's reflex-based finding generalizes to voluntary human movement.

Burke, Levine, Zajac, Tsairis (1971–1973) — motor unit typing. Burke's group in cat medial gastrocnemius combined intracellular recording of individual motor axons (measuring twitch contraction time, tetanic force, and fatigue index by repeated tetanic stimulation) with glycogen depletion histochemistry: they stimulated one identified motor unit repeatedly until its glycogen stores were selectively depleted, then froze the muscle and stained serial sections for glycogen (periodic acid–Schiff, PAS). Only the fibers belonging to the stimulated motor unit showed pale, depleted staining, allowing the experimenters to identify precisely which histochemical fiber type belonged to a physiologically characterized motor unit. This proved that motor units are histochemically and physiologically homogeneous (all fibers in one unit share one type) and established the S/FR/FF classification with matched contraction speed, force, and fatigue index for each type.

Cross-reinnervation experiments (Buller, Eccles & Eccles, 1960). To test whether fiber type is intrinsic to the muscle fiber or imposed by its nerve, these researchers surgically cross-connected the nerves of a fast muscle (flexor digitorum longus) and a slow muscle (soleus) in kittens, so each muscle was reinnervated by the "wrong" nerve. After regeneration, the fast muscle reinnervated by the slow nerve became slower-contracting, and the slow muscle reinnervated by the fast nerve became faster-contracting — contractile properties switched to match the new nerve, not the original muscle. This was decisive proof that the motor neuron (via its characteristic firing pattern and trophic signaling) actively determines the fiber-type properties of the muscle fibers it innervates, rather than fiber type being a fixed, autonomous property of the muscle.

Adrian and Bronk (1929) — the original demonstration of rate coding. Using concentric needle electrodes to record extracellular action potentials from human and animal muscle during voluntary contraction (among the earliest EMG recordings), Edgar Adrian and Detlev Bronk showed that motor unit firing frequency increased systematically with the strength of voluntary effort, and that stronger contractions recruited additional, previously silent units — the first direct electrophysiological evidence for both rate coding and recruitment as parallel mechanisms of force grading, decades before the size principle formalized why recruitment proceeds in a fixed order.

Example 1 — Estimating motor unit force distribution in a hand muscle. The first dorsal interosseous (FDI) muscle in humans has been studied extensively with motor unit decomposition techniques. Suppose an FDI motor neuron pool contains 120 motor units, and twitch tensions across the pool range from 0.1 g (smallest S unit) to 100 g (largest FF unit) — a roughly 1000-fold range, consistent with published human FDI data (Milner-Brown et al.). If twitch forces are distributed exponentially across the recruitment order (a common empirical finding), then the cumulative force contributed by the first 50% of units recruited (the small, weak ones) is only a small fraction of total muscle force — often well under 25% — while the last 10–20% of units recruited (the largest FF units) can contribute nearly half of maximal force. This is exactly the functional logic of the size principle: fine force steps are available at low force (good for pinch-grip precision when threading a needle), while the muscle still has large force reserves available for a forceful, less-precise grip (crushing a walnut) via the last few, largest units. A clinician measuring grip force loss after ulnar nerve injury (which disproportionately affects the deep intrinsic hand muscles including much of FDI's innervation) should expect precision tasks to degrade early, since exactly the deficits appear first in the fine, low-force end of the recruitment curve where remaining discrete steps become coarser.

Example 2 — Diagnosing motor neuron loss from an EMG interference pattern (clinical reasoning). A 58-year-old presents with progressive weakness and muscle wasting in the hand. Needle EMG of the FDI during moderate voluntary contraction shows: (1) individual motor unit action potentials that are markedly larger in amplitude (5–10 mV vs. a normal 0.5–2 mV) and longer in duration than normal, and (2) a reduced interference pattern — fewer distinct motor units are recruited even at high effort, and the patient reports needing much higher perceived effort to generate ordinary forces, with firing rates of the surviving units abnormally high (up to 40–50 Hz) even at submaximal force. Reasoning: normal motor unit action potentials reflect the number and synchrony of fibers within one unit. Enlarged, long-duration potentials indicate that surviving axons have sprouted collateral branches and reinnervated denervated fibers whose original motor neurons died — a compensatory process ("collateral sprouting") that increases the innervation ratio of surviving units, consistent with a motor neuron disease such as amyotrophic lateral sclerosis (ALS) rather than a primary muscle disease. In myopathies (e.g., muscular dystrophy), by contrast, individual muscle fibers die while motor neurons and innervation ratios remain normal, producing small, short-duration, polyphasic motor unit potentials and an early, "full" interference pattern at low force because more units (each now weaker) must be recruited to achieve a given force. The abnormally high firing rates in this patient reflect the nervous system's attempt to compensate for a shrunken recruitable pool by leaning harder on rate coding — exactly what the two-lever model of force grading predicts when the recruitment lever has lost range.

Elwood Henneman (1915–1996) trained in neurophysiology and spent most of his career at Harvard Medical School, where he built a research program around one deceptively simple question: when a muscle contracts more forcefully, which motor neurons turn on, and in what order? Rather than relying on indirect force measurements, Henneman went to the source — impaling individual spinal motor neurons with microelectrodes in anesthetized cats and correlating each neuron's intrinsic electrical properties (input resistance, axonal conduction velocity, antidromic spike threshold) with the stimulus intensity at which it began to fire. His insistence on measuring passive electrical properties directly, rather than inferring recruitment order from muscle twitch force alone, is what allowed him to identify soma size — not task, not intention, not even fiber type per se — as the causal variable governing recruitment order. This work was controversial when first proposed, because it implied a rigid, hard-wired constraint on voluntary motor control that seemed at odds with the flexibility of real movement; decades of subsequent human EMG studies (Milner-Brown, Stein, De Luca, and others) have repeatedly confirmed the size principle's robustness across muscles, species, and behavioral contexts, with only limited, muscle-specific exceptions (e.g., some evidence for task-dependent reordering in a few multifunctional muscles).

Clinical electrodiagnosis. Needle EMG and nerve conduction studies exploit motor unit physiology directly: motor unit action potential size, duration, and recruitment pattern distinguish neurogenic disease (ALS, radiculopathy, peripheral neuropathy — large, reinnervated units, reduced recruitment) from myopathic disease (muscular dystrophy, polymyositis — small, short-duration units, early full recruitment at low force). This is one of the most common diagnostic uses of the physiology described here.

Rehabilitation and strength training. Resistance training programs are designed around the size principle: near-maximal loads are needed to recruit FF (Type IIx) motor units at all, which is why hypertrophy and power training protocols use heavy loads or explosive intent, while endurance training targeting S units can be achieved with lighter, sustained loads. Electrical stimulation for rehabilitation (e.g., after ACL surgery) recruits motor units in reverse order from voluntary effort — largest axons first, because artificial electrical fields preferentially depolarize the largest-diameter axons — a key asymmetry exploited and compensated for in neuromuscular electrical stimulation (NMES) protocol design.

Neuroprosthetics and brain-machine interfaces. Decoding intended force or movement from motor cortex or peripheral nerve signals requires models of recruitment and rate coding to translate neural firing into realistic, graded muscle commands for prosthetic limbs and functional electrical stimulation systems restoring movement after spinal cord injury.

Aging and sarcopenia. Aging preferentially causes loss of large, fast motor neurons and denervation of FF fibers, with collateral reinnervation by surviving (often slower) axons — a physiological signature detectable by EMG and a major contributor to age-related loss of peak force and power, even when endurance capacity is relatively preserved.

Sports physiology and doping science. Explosive athletic performance (sprinting, jumping, Olympic lifting) depends on the capacity to recruit FF units rapidly and synchronously; training and, in banned contexts, some performance-enhancing interventions target this recruitment/rate-coding machinery, motivating EMG-based research in elite sport science.

Ask the tutor: How does motor unit recruitment order change (if at all) during a fast, ballistic movement versus a slow, controlled contraction?