5.5 The Athlete’s Brain: Functional Neuroanatomy for Performance

Executive summary. Muscle is the last link in a long chain. A voluntary movement begins as a decision in the prefrontal cortex, is planned by premotor areas, released by the primary motor cortex, gated by the basal ganglia, corrected by the cerebellum, and finally executed by spinal circuits and motor units. Knowing which structure does what explains a lot of ordinary training experience: Why strength appears before size, why a skill feels effortless one day and clumsy the next, why fatigue degrades decision-making before it degrades force, and why the same programme produces different results in a distracted athlete than in a focused one. This article is the anatomical foundation for the rest of Module 5.

‌

Key takeaways

1. Movement is produced by a hierarchy, not a single command centre. Intention, planning, release, gating, correction and execution are separable, and any one of them can be the limiting factor.
2. The corticospinal tract gives direct, fractionated control. It is why humans can produce fine, independent finger and ankle movements that most animals cannot.
3. The basal ganglia decide which action is released and with how much vigour. They are also where habits live, so they matter as much for adherence as for movement.
4. The cerebellum compares predicted with actual sensory feedback and corrects the difference. It is why an unfamiliar load feels clumsy before it feels heavy.
5. The prefrontal cortex is the most fatigue-sensitive link. Technique usually breaks down because supervision fails, not because muscle fails.
6. Almost every early training adaptation is neural. Force output rises before cross-sectional area does, because the nervous system gets better at recruiting what already exists.

‌

Beginner section: A simple map of the brain that trains

If you have ever wondered why a new exercise feels awkward for two weeks and then suddenly feels normal, or why you can lift more on a day when you feel switched on, the answer is not in the muscle. It is in the control system driving it.

‌

Think of the brain during training as a company with five departments. Each has a job, and the movement you see is the result of all five doing their job well enough at the same time.

‌

• The decision-maker. The front of the brain, the prefrontal cortex, decides the set is worth doing and keeps you on task while it hurts. This is the department that quits first when you are tired, stressed or underslept.
• The planner. Just behind it, the premotor areas work out the order of operations. Brace, then unrack, then descend. Planning happens before the first muscle contracts, which is why a rushed set-up produces a poor rep.
• The dispatcher. The primary motor cortex sends the actual signal down the spinal cord. How strongly and how quickly it sends that signal is trainable, and this is a large part of what heavy and explosive lifting improves.
• The gatekeeper. The basal ganglia sit deep in the brain and decide which action is released and how forcefully. They also automate anything you repeat often, whether that is a clean technique or a bad one.
• The proofreader. The cerebellum predicts what a movement should feel like and corrects the gap between prediction and reality, continuously and faster than you can consciously notice.

‌

‌

Two practical consequences fall out of this picture immediately. First, when a movement is new, most of your training effect is going into the planner and the proofreader rather than the muscle. That is why the first few weeks of any new lift produce fast progress with little visible change. Second, when you are tired in a way that feels mental rather than physical, the department failing is the decision-maker, and the fix is not more effort but better session design.

‌

‌

One more idea worth having before the detail. The nervous system does not send a single “lift” command. It sends a continuous stream of adjustments, informed by a continuous stream of feedback from muscle, tendon, joint, skin, eye and inner ear. Movement is a conversation, not an instruction, and most coaching that works does so by improving the quality of that conversation.

‌

Advanced section: Structure, circuitry, and what each level computes

The three descending routes

Cortical motor areas do not influence the spinal cord through one pathway. They use three, and the distinction matters because they fail differently and they are trained differently.

‌

‌

The corticospinal tract is the direct route. Roughly a million axons descend from motor and premotor cortex, most crossing at the medullary pyramids, and in primates a meaningful fraction synapse directly onto spinal motor neurons (Lemon, 2008). This monosynaptic component supports fractionated movement: The ability to move one digit, or one joint, largely independently of its neighbours. Sports demanding fine distal control, from goalkeeping to the turnover in a clean, lean heavily on it.

‌

The cortico-basal ganglia loop is re-entrant: Cortex projects to striatum, striatum to internal pallidum and substantia nigra pars reticulata, those to thalamus, and thalamus back to cortex (Purves et al., 2018). The default state of this loop is inhibitory. The pallidum tonically suppresses thalamic output, and movement occurs when the direct pathway transiently releases that suppression. This is why the basal ganglia are better described as a selection and gating system than a movement generator. Damage does not remove the capacity to move; it removes the capacity to select and scale movement appropriately, which is the clinical picture in Parkinson’s disease and, at the other extreme, in Huntington’s disease (Kandel et al., 2013).

‌

The cortico-cerebellar loop runs cortex to pontine nuclei, into cerebellar cortex, out through deep cerebellar nuclei to thalamus, and back to cortex. Its computation is comparison. Climbing fibre input from the inferior olive carries error signals, and parallel fibre to Purkinje cell synapses adjust their strength accordingly (Kandel et al., 2013). Functionally, the cerebellum builds and maintains an internal model of your body and the objects you interact with, updating it whenever prediction and outcome diverge (Ito, 2008; Wolpert et al., 2011).

‌

‌

‌

What the primary motor cortex actually encodes

The old picture of a simple somatotopic map with one patch of cortex per muscle is a useful simplification and a poor description. Individual corticospinal neurons diverge onto multiple motor neuron pools, and stimulating a cortical site typically produces a coordinated multi-joint posture rather than a single muscle twitch (Lemon, 2008). What appears to be encoded is closer to a movement synergy scaled by intended force and direction.

‌

This has a direct training implication. When you train a lift you are not training a “quadriceps region” of cortex. You are consolidating a synergy: A particular pattern of co-activation, timing and postural support. That is why strength transfer is specific to pattern and position, and why a strong squatter is not automatically a strong split-squatter.

‌

• Recruitment. Motor units are generally recruited from small and fatigue-resistant to large and fatigable, in line with Henneman’s size principle (Henneman et al., 1965). Training does not reverse the order, but it improves access to the top of the range.
• Rate coding. Above roughly two-thirds of maximum force in many muscles, additional force comes mainly from higher firing frequency rather than added units (Enoka & Duchateau, 2017). Explosive intent trains this component preferentially.
• Synchronisation and doublets. Brief high-frequency bursts at contraction onset raise the rate of force development substantially without changing peak force (Enoka & Duchateau, 2017). This is one mechanism by which ballistic training improves early force.
• Antagonist coordination. Reducing inappropriate co-contraction is a large and underappreciated part of early strength gain (Sale, 1988). Getting stronger partly means stopping the brakes being applied.

‌

These four mechanisms are covered in depth in 1.5 The Nervous System and Athletic Performance, and their expression in fast movements in 2.5 Power Development and Explosive Strength.

‌

The prefrontal cortex: Supervision, and why it fails first

The prefrontal cortex maintains goals in working memory, inhibits competing actions, and monitors conflict between what you intended and what you did (Miller & Cohen, 2001). It is metabolically expensive, it does not scale well with simultaneous demands, and its performance degrades measurably with sleep loss, psychological stress and prolonged cognitive work.

‌

‌

Marcora and colleagues showed that ninety minutes of demanding cognitive work before a cycling time-to-exhaustion test reduced endurance performance without changing cardiorespiratory or metabolic markers, while raising perceived exertion at the same power output (Marcora et al., 2009). The muscle had not changed. The supervision had. In practice this means the state an athlete arrives in is part of the training stimulus, and a technically demanding session placed after a mentally punishing day is not the session you planned.

‌

‌

A useful heuristic follows. If technique degrades gradually across a set in proportion to local burning, treat it as peripheral fatigue and manage volume. If it degrades suddenly, after a distraction, or on a day that felt mentally heavy from the first warm-up set, treat it as supervisory fatigue and manage session order, complexity and environment instead.

‌

Reaction time, and where it is actually won

Simple reaction time in trained athletes has a floor of roughly 150 to 200 milliseconds, and that floor is largely fixed by conduction and electromechanical delay. What is not fixed is everything between detection and response selection.

‌

‌

Hick’s law describes the relationship: Choice reaction time rises approximately with the logarithm of the number of alternatives (Hick, 1952). This is why reducing an athlete’s effective number of options through pattern recognition produces faster apparent reactions than any amount of light-board training. An experienced boxer is not conducting nerve impulses faster than a novice; they are reading a smaller, better-organised set of possibilities. The applied version is covered in 3.10 Agility and Change of Direction.

‌

Practical section: Training each level deliberately

You cannot train “the brain” as a single entity, but you can bias a session towards a particular level of the hierarchy. The list below is not a programme; it is a diagnostic tool for deciding what to change when progress stalls.

‌

• To train descending drive: Heavy loads with maximal intent, and light loads moved with maximal intent. Both raise the demand on cortical output. Low repetitions, long rests, and stop the exercise when speed drops meaningfully.
• To train selection and gating: High repetition volume in a consistent context. Same order, same cues, same equipment. Consistency is what lets the basal ganglia chunk a sequence into one retrievable unit (Graybiel, 2008).
• To train internal models: Vary the task, not the context. Different loads, tempos, implements and surfaces, with feedback available. Variability forces the cerebellum to update rather than repeat (Wolpert et al., 2011).
• To train supervision: Add constraints deliberately. Time pressure, fatigue, noise, an audience, or a competing cognitive task, introduced in small doses during general preparation rather than the week before competition.

‌

Sequencing within a session should respect the fatigue sensitivity of each level. Skill and decision work goes early, while supervision is intact. Heavy strength work goes next. Volume and conditioning go last, because they tolerate degraded supervision better than skill work does.

‌

‌

A final point about time course. Neural adaptations to a new movement are fast, often measurable within a handful of sessions, but comparatively fragile early on. That argues for higher frequency and lower per-session volume when learning something new, and the reverse once the pattern is stable.

‌

Sport applications

• Combat sports. The limiting factor is almost never conduction speed. It is response selection under a large option set, trained by constrained sparring rather than isolated reaction drills.
• Team sports. Decision quality falls with accumulated physical and cognitive load. Placing tactical work after a heavy conditioning block trains athletes to make poor decisions efficiently.
• Weightlifting and powerlifting. Corticospinal and gating contributions dominate. Consistency of context is a performance variable rather than a superstition, which is why elite lifters guard their warm-up routine so carefully.
• Endurance sport. Supervisory fatigue is a genuine performance limiter. Mental durability is trainable, but it should be periodised like any other quality rather than assumed.
• Skill-dominant sports. Gymnastics, diving and throwing rely on cerebellar internal models, which is why they are unusually sensitive to changes in equipment, footwear or body mass.

‌

Common mistakes

• Treating early strength gains as muscle growth. For the first several weeks of a new stimulus most of the change is neural (Sale, 1988). Judging a programme by size in week three tells you almost nothing.
• Adding variety to a movement that is not yet stable. Variation updates internal models, but only once there is a model to update. Vary after competence, not before.
• Training reaction with lights instead of with opponents. Generic reaction drills load the stage of processing that is least trainable and least sport-specific.
• Scheduling skill work late in the session. Supervision is the first resource to deplete and skill work is the most supervision-dependent thing you do.
• Interpreting mental fatigue as low motivation. Prefrontal load has measurable performance effects independent of how much the athlete wants to train.
• Changing several context variables at once. New shoes, new bar, new order and a new cue in the same week makes it impossible to attribute any change to anything.

‌

Coaching cues

• Set up the same way every time. Consistency of context is what lets a sequence become automatic.
• Move with intent even when the load is light. Intent is a cortical variable, not a barbell variable.
• When a movement feels clumsy rather than heavy, the problem is prediction, not strength. Slow it down and get feedback.
• Do the thinking early in the session and the grinding late.
• If technique collapses suddenly rather than gradually, end the set. That is supervision failing, and pushing through teaches the wrong pattern.
• Reduce the number of options before trying to react faster.

‌

FAQs

Does the brain get stronger, or just the muscle?

Both change, but on different timescales and by different mechanisms. Measurable increases in voluntary force can occur within a few sessions, well before any detectable change in muscle cross-sectional area, and these early gains are attributed to improved recruitment, higher firing rates, better antagonist coordination and reduced inhibition. Muscle growth becomes the dominant contributor later. Neither is more real than the other; they arrive in a predictable order.

‌

Can I train my nervous system directly?

You can bias training towards neural qualities, but not isolate them. Heavy loads, maximal intent at any load, high rates of force development, unfamiliar tasks and constrained decision-making all shift demand towards the control system rather than the tissue. There is no exercise that trains the nervous system in isolation, because the nervous system is only ever engaged through movement.

‌

Why do I lift better on some days for no obvious reason?

The most common contributors are sleep quality, accumulated cognitive load, time of day relative to your circadian rhythm, and psychological stress. All four act on levels of the hierarchy above the muscle. This is normal variation rather than a failing programme, and it is one reason autoregulated loading tends to outperform rigid percentage prescriptions.

‌

Is mental fatigue actually different from being unmotivated?

Yes. Experimental work has shown that prolonged cognitive tasks reduce subsequent endurance performance and raise perceived exertion at a given workload without changing heart rate, oxygen uptake or blood lactate. That dissociation between subjective effort and physiological cost is what distinguishes it from simple lack of motivation.

‌

How long does it take for a movement to become automatic?

There is no single number, and published estimates vary widely because automaticity depends on movement complexity, consistency of context and number of exposures rather than on calendar time. A simple, highly repeatable action in a stable context automates far faster than a complex multi-joint skill performed under varying conditions.

‌

Does damage to one of these regions end an athletic career?

That depends entirely on the region, the extent and the sport, and it is a clinical question rather than a coaching one. Anyone with a suspected neurological injury, including concussion, should be assessed and cleared by a qualified medical professional before returning to training.

‌

Recommended videos

Each video below was chosen because it covers a specific part of this article in more depth than text alone allows.

‌

‌

‌

‌

‌

‌

Related reading on FitXplor

• 1.5 The Nervous System and Athletic Performance
• 1.2 Human Anatomy for Athletes
• 2.5 Power Development and Explosive Strength
• 3.10 Agility and Change of Direction
• 5.1 The Psychology Behind Peak Performance

‌

References

Purves, D., Augustine, G. J., Fitzpatrick, D., Hall, W. C., LaMantia, A.-S., Mooney, R. D., Platt, M. L., & White, L. E. (Eds.). (2018). Neuroscience (6th ed.). Oxford University Press.

Kandel, E. R., Schwartz, J. H., Jessell, T. M., Siegelbaum, S. A., & Hudspeth, A. J. (Eds.). (2013). Principles of Neural Science (5th ed.). McGraw-Hill.

Lemon, R. N. (2008). Descending pathways in motor control. Annual Review of Neuroscience, 31, 195–218.

Graybiel, A. M. (2008). Habits, rituals, and the evaluative brain. Annual Review of Neuroscience, 31, 359–387.

Wolpert, D. M., Diedrichsen, J., & Flanagan, J. R. (2011). Principles of sensorimotor learning. Nature Reviews Neuroscience, 12(12), 739–751.

Henneman, E., Somjen, G., & Carpenter, D. O. (1965). Functional significance of cell size in spinal motoneurons. Journal of Neurophysiology, 28, 560–580.

Enoka, R. M., & Duchateau, J. (2017). Rate coding and the control of muscle force. Cold Spring Harbor Perspectives in Medicine, 7(10), a029702.

Sale, D. G. (1988). Neural adaptation to resistance training. Medicine and Science in Sports and Exercise, 20(5 Suppl), S135–S145.

Marcora, S. M., Staiano, W., & Manning, V. (2009). Mental fatigue impairs physical performance in humans. Journal of Applied Physiology, 106(3), 857–864.

Hick, W. E. (1952). On the rate of gain of information. Quarterly Journal of Experimental Psychology, 4(1), 11–26.

Miller, E. K., & Cohen, J. D. (2001). An integrative theory of prefrontal cortex function. Annual Review of Neuroscience, 24, 167–202.

Ito, M. (2008). Control of mental activities by internal models in the cerebellum. Nature Reviews Neuroscience, 9(4), 304–313.