A winged scapula, where the shoulder blade lifts away from the ribcage and becomes visibly prominent, is most often caused by weakness or paralysis of the serratus anterior muscle. The serratus anterior holds the scapula flat against the thorax and rotates it upward when you raise your arm, so when it fails, the inner border of the scapula juts out like a small wing. The usual culprit is damage to the long thoracic nerve, the single nerve responsible for powering this muscle, though the way that damage happens and how well it recovers varies considerably from person to person.
What the Serratus Anterior Actually Does
The serratus anterior is a broad, fan-shaped muscle that wraps around the side of your ribcage, attaching from the upper eight or nine ribs to the underside of the scapula. Its middle and lower portions are the only muscles around the shoulder blade that can both rotate the scapula upward and tilt it backward at the same time. That combination is critical every time you lift your arm overhead: the scapula needs to rotate so the socket stays under the ball of the upper-arm bone, and it needs to tilt back so the bony roof of the shoulder does not pinch the rotator cuff tendons passing beneath it. The serratus anterior also pulls the scapula forward around the ribcage (protraction) and, perhaps most fundamentally, presses it snugly against the thorax to create a stable base for the entire shoulder complex.
When the serratus anterior stops working, all of these functions are lost at once. The scapula drifts inward and tilts forward, and the medial (inner) border peels away from the ribs. Overhead reach drops dramatically because the scapula can no longer rotate to accommodate the movement. Over time, the altered mechanics can lead to subacromial impingement, where the rotator cuff tendons get compressed under the shoulder’s bony arch during arm elevation, a condition linked to the muscle imbalances and abnormal scapular motion that serratus anterior paralysis creates.
How the Long Thoracic Nerve Gets Injured
The long thoracic nerve is unusually vulnerable. It originates from the C5, C6, and C7 nerve roots in the neck, threads through or between the scalene muscles, passes beneath the clavicle, and then runs down the entire lateral chest wall to reach the serratus anterior. With an average length of about 22 centimeters measured from the upper border of the muscle alone, it is one of the longest motor nerves in the body, and its superficial course leaves it exposed to stretch and compression at multiple points.
When the arm is raised, the bundle of nerves and vessels in the armpit shifts upward, but the long thoracic nerve is tethered to the chest wall. At the point where it exits the axillary sheath, the nerve bends at roughly a 30-degree angle and gets stretched. That mechanical stress is thought to be one reason why repetitive overhead activities can injure the nerve over time.
In a surgical case series looking at isolated long thoracic nerve palsy, the causes broke down into traumatic injury in about 58 percent of patients, Parsonage-Turner syndrome (a sudden inflammatory nerve condition) in about 32 percent, and complications from shoulder surgery in the remaining 10 percent. Traumatic causes range from direct blows to the lateral chest wall, to carrying heavy loads on the shoulder, to surgical procedures near the axilla or chest. Sports that involve repetitive overhead motion or sudden forceful stretching of the shoulder are frequently implicated.
Parsonage-Turner syndrome, also called neuralgic amyotrophy, deserves special attention because it is probably underdiagnosed. It typically starts with severe, acute pain in the shoulder or upper arm that comes on seemingly out of nowhere, followed within days to weeks by weakness and muscle wasting. The long thoracic nerve is one of the nerves most commonly affected. Triggers can include viral illnesses, vaccinations, surgery under general anesthesia, and strenuous exercise, though many cases have no identifiable trigger at all. MR neurography studies have shown that the nerve can develop hourglass-like constrictions in Parsonage-Turner syndrome, which may explain why recovery is sometimes incomplete.
Anatomical Variations That Raise Risk
Not everyone’s long thoracic nerve follows the same path, and those differences help explain why some people develop winging after seemingly minor events while others tolerate the same activities without trouble. Cadaver studies have documented at least seven distinct patterns of how the nerve roots emerge and converge. The spot where the upper roots (from C5 and C6) join the lower root (from C7) can range from high in the neck above the first rib all the way down to the third intercostal space. A nerve that converges lower has a longer unsupported segment and may be more susceptible to stretch.
The nerve’s relationship to the scalene muscles in the neck is another source of variability. Some nerve roots pass between the middle and posterior scalene muscles, some pierce directly through the middle scalene, and some course over it. In a clinical study of patients with thoracic outlet syndrome, two main variants were identified: an extramuscular course, seen in about 60 percent, and an intramuscular course where the nerve passed through the scalene muscle belly, seen in about 40 percent. Scapular winging and dyskinesia occurred in 66 percent of patients with the intramuscular variant compared to just 11 percent of those with the extramuscular variant, a striking difference that suggests the intramuscular path creates a built-in vulnerability to compression.
Even the branching pattern on the chest wall varies. In roughly four out of five people, the nerve travels as a single trunk along the serratus anterior, but in about one in five, it splits into two equal major branches. That bifurcation matters to surgeons performing thoracic procedures, since an unrecognized second branch could be inadvertently damaged.
Medial Winging Versus Lateral Winging
Not all winged scapulae look the same, and the pattern of winging tells you which muscle has failed. Serratus anterior paralysis produces medial winging, where the inner (medial) border of the scapula lifts off the chest wall. The winging typically becomes more prominent when you push against a wall or try to raise your arm forward. By contrast, paralysis of the trapezius (from spinal accessory nerve injury) or the rhomboids (from dorsal scapular nerve injury) produces lateral winging, where the scapula shifts outward and its lateral border becomes prominent. The distinction is clinically important because the underlying nerve injuries, prognoses, and treatment strategies differ for each pattern.
Testing for Serratus Anterior Weakness
The classic bedside test for a winged scapula is the wall push-up: you place your hands flat against a wall and push, and a clinician watches from behind to see if the scapula lifts off. It is widely taught and instantly recognizable, but research suggests it is far less reliable than its popularity would imply. One comparative study of various methods for bringing out winging in neuromuscular disorders found that the wall push-up ranked fourth in sensitivity, detecting only about 60 percent of cases. The most sensitive method was simply having the patient slowly lower their arms from a raised position, which picked up winging in 100 percent of cases.
Another study comparing the wall push-up test to a shoulder flexion resistance test found a different but related problem: in a group of 50 patients who had the scapular border visibly lifting off the chest wall, the wall push-up was positive in every single one, but 45 of those were false positives for actual serratus anterior deficiency. In other words, the wall push-up caught everything but could not distinguish serratus anterior weakness from other causes of scapular prominence. The shoulder flexion resistance test, by contrast, was both 100 percent sensitive and 100 percent specific in that study. A systematic review of scapular physical examination tests for shoulder disorders reached a sobering conclusion overall: no single physical examination test of the scapula was found to be reliably useful for differentially diagnosing specific shoulder pathologies.
The practical takeaway is that visual inspection and wall push-ups are reasonable starting points, but they are not enough to confirm that the serratus anterior is the problem. If winging is present, forward arm lowering and resisted shoulder flexion give the examiner more useful information. And when the clinical picture remains uncertain, electrodiagnostic testing and imaging become essential.
The Role of Nerve Studies and Imaging
Electrodiagnostic studies, including nerve conduction tests and needle electromyography (EMG), are the standard way to confirm which nerve is involved and how severely it is affected. A thorough evaluation should include bilateral nerve conduction studies of both the long thoracic nerve and the spinal accessory nerve, along with needle EMG of the muscles those nerves supply, since the clinical picture can be misleading.
That caution is well-founded. In a small series of patients who appeared to have isolated long thoracic nerve palsy, 57 percent had additional electrodiagnostic abnormalities beyond the territory of the long thoracic nerve, and 71 percent had MRI evidence of nerve enlargement or muscle wasting outside the nerve’s distribution. These findings suggest that what looks like a straightforward serratus anterior problem on the surface may be part of a broader nerve injury, possibly an incomplete form of Parsonage-Turner syndrome or a brachial plexus lesion. Missing that broader picture can change treatment decisions significantly.
MRI is increasingly used not just to image the shoulder but to visualize the nerve itself. MR neurography can detect the hourglass-like constrictions that characterize Parsonage-Turner syndrome, while quantitative muscle MRI can measure the degree of fatty infiltration and atrophy in the serratus anterior, giving clinicians an objective measure of how much muscle damage has already occurred.
Conservative Treatment and Rehabilitation
Because many cases of long thoracic nerve palsy do recover on their own, the standard first-line approach is conservative management. Most experts recommend waiting at least 12 to 18 months before considering surgery, using that window for physical therapy focused on restoring shoulder range of motion and strengthening the muscles that can partially compensate for a weak serratus anterior, particularly the trapezius and rhomboids. A physical therapy program specifically designed for scapular winging showed highly significant improvements across all measured domains of a shoulder-specific quality-of-life index, suggesting that structured rehab makes a meaningful difference even when full nerve recovery does not occur.
The choice of exercise matters. Electromyographic studies have identified which movements activate the serratus anterior most strongly in healthy individuals, and those same exercises form the foundation of rehab programs aimed at retraining the muscle as nerve function returns. Exercises that combine upward scapular rotation with protraction consistently produce the highest activation levels. The push-up plus (a standard push-up with an extra push at the top to protract the shoulder blades), the dynamic hug (a bear-hug motion with resistance), and the serratus anterior punch (a straight-arm punching motion while lying on your back) all reliably generate meaningful muscle activity. Shoulder abduction in the scapular plane above 120 degrees and diagonal patterns combining flexion with horizontal flexion and external rotation also produce strong activation.
For patients who cannot yet handle those loads, modified versions on unstable surfaces or at reduced body-weight angles can still produce substantial serratus anterior activation. A review of push-up variations found moderate evidence that standard push-ups on an unstable surface, elbow push-ups on both stable and unstable surfaces, and wall push-ups on an unstable surface all produced mean serratus anterior activation of at least 50 percent of maximum voluntary contraction. The principle is to start where the patient can manage and progress as tolerated.
When Surgery Becomes the Answer
The traditional view is that most patients will recover spontaneously within two years. However, there is growing evidence that the natural history is not as optimistic as textbooks suggest. A significant percentage of patients are left with residual winging, persistent muscle weakness, and fatigability even after the standard observation period. When conservative treatment fails to restore adequate function, two broad surgical strategies exist: muscle transfer and nerve reconstruction.
The most established muscle-transfer procedure involves moving the pectoralis major, the large chest muscle, to the inferior border of the scapula. The transferred muscle acts as a dynamic stabilizer, pulling the scapula against the chest wall in place of the paralyzed serratus anterior. Several technical variations have been described, including split transfers with allograft tissue augmentation, which aim to provide secure fixation with acceptable cosmetic results and better preservation of shoulder range of motion compared to older techniques.
Nerve-based procedures are a newer and less widely available option but can produce excellent results when performed early enough. Transferring a healthy donor nerve, typically a branch of the thoracodorsal nerve, to the long thoracic nerve has been reported to eliminate winging and restore full shoulder function in individual cases, with one patient showing no winging and no functional restriction at six and a half years of follow-up. A two-level nerve transfer approach, using branches from both the thoracodorsal nerve and a pectoral fascicle, has also achieved near-total improvement. These procedures work best when the serratus anterior has not yet undergone irreversible fatty degeneration, which is why the timing question, how long to wait before abandoning conservative management, remains one of the most debated issues in this field. Proponents of earlier surgical intervention have argued that decompression of the long thoracic nerve, even beyond 12 months after symptom onset, can yield satisfactory results, challenging the conventional wait-and-see timeline.
The Overlooked Burden of Living With It
Winged scapula is sometimes dismissed as a cosmetic issue, but qualitative research with patients paints a very different picture. People living with the condition describe physical and functional limitations that significantly affect daily activities, independence, and social engagement. Simple tasks like reaching into a cabinet, carrying groceries, or getting dressed become frustrating ordeals. Many report a feeling of lost identity, particularly if the condition ended an athletic career or made it impossible to perform their job.
The mental health burden is substantial. Patients describe feelings of isolation compounded by the invisibility of the condition, since people around them cannot see the nerve damage and often do not understand why a seemingly fit-looking person cannot lift their arm properly. Interactions with the healthcare system can add to the frustration: delayed diagnoses, insufficient knowledge among clinicians who may not have encountered the condition before, and wide variation in treatment recommendations are recurring themes in patient accounts.
Pediatric cases add another layer of complexity. Winged scapula in children can follow trauma, as in one reported case of a four-year-old boy who developed persistent winging after a fall onto his back. At two years of follow-up, conservative management with physiotherapy had not improved his winging, and while he was pain-free at rest, swimming caused scapular and shoulder pain, limiting his activity. He was eventually referred for surgical evaluation. Children’s ongoing growth and high activity demands make prolonged waiting periods particularly challenging for families.
An Evolutionary Footnote
The serratus anterior is not just a clinical curiosity; it carries an evolutionary signature. Comparative electromyographic studies of the muscle in primates have shown that the human serratus anterior is clearly adapted for arm-raising functions, a pattern consistent with descent from a small ape with a thoracic shape similar to certain New World monkeys that swing through trees. The same anatomical design that allows us to reach, throw, and lift overhead is exactly what makes the shoulder blade’s stability so dependent on this one muscle and its one nerve. The engineering that lets us raise our arms freely is, in a sense, the same engineering that makes a winged scapula possible when something goes wrong.