Which Shoulder Is Higher in Scoliosis?

In the most common form of adolescent idiopathic scoliosis, where the main thoracic curve bends to the right, the right shoulder sits higher than the left. But scoliosis is not one condition with one shape, and the answer changes depending on which part of the spine curves, in which direction, and how multiple curves interact with each other. Roughly seven in ten adolescents with idiopathic scoliosis have a right-convex thoracic curve, so “right shoulder high” is what most people encounter, yet a sizable minority present with the opposite pattern or with shoulders that look level despite a significant spinal curve underneath.

Why the Right Shoulder Tends to Be Higher

The thoracic spine runs through the upper and middle back, and a curve in this region tilts the ribcage and the structures attached to it, including the shoulder girdle. When the main thoracic curve bows to the right (meaning the apex of the curve points rightward), the vertebrae at the top of that curve push the right side upward. The collarbone, scapula, and soft tissues on the convex side of the curve all get carried along. The result is a visibly elevated right shoulder, often accompanied by a more prominent right rib hump when the person bends forward.

The shoulder’s height depends not only on the main thoracic curve but also on what happens above it. Research has shown that shoulder level is driven by the main thoracic curve when the body’s overall tilt leans toward the same side as that curve, and by the upper (proximal) thoracic curve when the tilt goes the other way.¹ In practical terms, the spine doesn’t operate as a single arc. There are compensatory curves above and below the main one, and each curve tugs the shoulders in its own direction. Which one “wins” depends on the relative size and stiffness of each curve.

When the Left Shoulder Is Higher Instead

A left-convex thoracic curve, though less common in typical adolescent scoliosis, elevates the left shoulder by the same mechanical logic: the convex side pushes upward. Left thoracic curves are more often seen in thoracolumbar or lumbar scoliosis patterns, and clinicians pay extra attention to them because a leftward thoracic curve sometimes signals an underlying condition rather than the “idiopathic” variety.

Even in the dominant right-thoracic pattern, the left shoulder can end up higher if the proximal thoracic curve (the smaller curve in the upper back that bends leftward as a compensatory response) is large or stiff enough to overpower the main curve’s effect. This is sometimes called a “double thoracic” pattern, and failing to recognize it before surgery is a well-documented source of postoperative shoulder asymmetry. When surgeons correct the main rightward curve without accounting for the upper leftward curve, the left shoulder can become noticeably elevated after the operation.²

It Is Not Just Up and Down

Shoulder asymmetry in scoliosis involves more than one shoulder sitting higher than the other. The shoulders also rotate in the horizontal plane. A study using CT scans found that in nearly all adolescent idiopathic scoliosis patients examined, the right shoulder was rotated forward (anteriorly) regardless of the specific curve type. Thirteen of fourteen cases showed this anticlockwise axial torsion, and the single exception was a patient with a left-dominant (Lenke type V) curve pattern who had level shoulders. The degree of forward rotation did not depend on curve severity: large curves produced similar torsion to smaller ones.³

This rotational component means that looking at someone from behind may not tell the full story. The shoulder that appears higher from the back may also be the one that juts forward when viewed from above. And the scapula (shoulder blade) on the convex side of the curve tends to sit in a different resting position: more laterally displaced, more anteriorly tilted, or more “winged.” Patients with scoliosis show measurably different scapular positions and shoulder function compared with people without the condition.⁴ So the question “which shoulder is higher” captures only one dimension of a three-dimensional deformity.

What X-Rays Show vs What People See

Here is something that catches both patients and clinicians off guard: the shoulder that looks higher on an X-ray is not always the one that looks higher in the mirror. Radiographic measures of shoulder balance (things like T1 tilt, clavicle angle, and coracoid height difference) only partially reflect what the shoulder actually looks like when you stand in front of someone. One study of double-thoracic curves found that none of the commonly used radiographic parameters correlated strongly enough with cosmetic appearance to serve as a reliable proxy. The correlation coefficients between X-ray measurements and visual shoulder balance were all below 0.8, meaning the X-ray captured some of the picture but missed a lot.⁵

That said, some radiographic measurements do a better job than others. In one study, the clavicle-chest cage angle difference showed the strongest correlation with both inner and outer shoulder height as seen clinically.⁶ The practical takeaway is that no single X-ray number tells you exactly how someone’s shoulders look in a T-shirt. Surgeons increasingly use multiple radiographic parameters together, and newer deep-learning models are being developed to automate these measurements from X-rays to reduce human measurement error.⁷

The disagreement extends to the surgeons themselves. When multiple spine surgeons were asked to assess the same scoliosis patients for shoulder balance, their agreement with each other was only fair to moderate, whether they were looking at X-rays, front-view photographs, or back-view photographs. Even individual surgeons showed only moderate consistency when re-evaluating the same images at different times.⁸ This is a genuine weakness in the field: there is no universally agreed-upon way to define or measure shoulder balance, which complicates everything from diagnosis to surgical planning.

What Patients Actually Notice About Their Shoulders

You might assume that someone whose shoulders are measurably uneven would notice the asymmetry. Often they don’t. A study examining the relationship between clinical shoulder measurements and patients’ own perceptions found no significant correlation between the two. Roughly half of the patients who were clinically classified as having imbalanced shoulders (a difference of 3 degrees or more) perceived themselves as “totally balanced.” Meanwhile, one in ten patients whose shoulders were clinically balanced reported feeling that their shoulders were uneven.⁹

This disconnect matters because patient satisfaction after treatment hinges on perceived appearance, not on what the X-ray says. Instruments like the SRS-22 questionnaire (a standard scoliosis quality-of-life survey) and the Trunk Appearance Perception Scale do correlate modestly with measured shoulder height differences, but the correlation is not strong.¹⁰ If you’re a parent or patient trying to understand what “shoulder balance” means for everyday life, the answer is that a clinically detectable difference of a centimeter or two may be invisible to you and everyone around you, while a subtler difference paired with rotational asymmetry might bother you more than the numbers would suggest.

How the Pelvis Affects Shoulder Height

Shoulder asymmetry doesn’t always start at the spine. The pelvis is the foundation the spine sits on, and if one side of the pelvis is higher than the other, the spine tilts to compensate, which cascades upward to the shoulders. In adolescents with scoliosis, pelvic obliquity (one hip sitting higher) is associated with changes in radiographic shoulder height. Patients with a rightward pelvic tilt showed greater absolute shoulder height differences, while those with a leftward pelvic tilt showed smaller differences.¹¹

Leg length differences are one common cause of pelvic obliquity. Even a modest discrepancy in leg length can tilt the pelvis, introduce a compensatory spinal curve, and shift the shoulders.¹² This is worth knowing because not every uneven shoulder is scoliosis. A shoe lift or other intervention at the foot can sometimes level the pelvis and reduce the shoulder asymmetry. Clinicians evaluating shoulder imbalance typically check for pelvic obliquity and leg length discrepancy before attributing the problem entirely to the spinal curve itself.

What Happens to Shoulder Balance After Surgery

One of the trickiest aspects of scoliosis surgery is getting the shoulders to look balanced afterward. Correcting the main thoracic curve straightens the middle of the spine but can unmask or worsen asymmetry at the top if the upper curve isn’t addressed proportionally. A systematic review identified key risk factors for postoperative shoulder imbalance: a main thoracic curve greater than 80 degrees, preoperative level shoulders (paradoxically), a preoperatively high left shoulder, and greater skeletal maturity. Shoulder imbalance greater than 2 centimeters was the most commonly used threshold for a clinically meaningful problem.¹³

The balance between how much you correct the upper curve versus the main curve turns out to be critical. When the proximal thoracic curve was corrected by more than about 55%, the vast majority of patients achieved balanced shoulders. But when the proximal curve was corrected less aggressively and the main thoracic curve was corrected by more than 56%, about two-thirds of patients developed postoperative shoulder imbalance.¹⁴ In other words, overcorrecting the big curve while undercorrecting the smaller upper curve pulls the shoulders out of alignment.

Choosing where to start the instrumentation (the “upper instrumented vertebra”) also matters, though the evidence here is surprisingly unsatisfying. One study found that following the standard recommendations for where to place the top of the rods actually resulted in more shoulder imbalance than going against those recommendations, and the researchers concluded they could not identify any single set of selection criteria that reliably predicted balanced shoulders after surgery.¹⁵ Proximal balance parameters do tend to improve over time regardless of the initial instrumentation choice, but residual differences between groups persist.¹⁶ Patients with lower preoperative right shoulder elevation were at higher risk of postoperative imbalance when the main thoracic curve was corrected aggressively.¹⁷ Vertebral rotation in the upper thoracic spine also plays a role: greater proximal vertebral rotation before or after surgery is linked to worse shoulder balance outcomes.¹⁸

Can Exercise or Bracing Improve Shoulder Position

For mild to moderate scoliosis, or as a complement to other treatments, targeted exercise programs show some promise in improving shoulder-related parameters. The Schroth method, a physiotherapy approach designed specifically for scoliosis, has been studied in randomized trials. One trial found that Schroth exercises combined with scapular stabilization exercises were more effective than Schroth alone at reducing neck tilt and the angle of scapular asymmetry after eight weeks.¹⁹ Another trial comparing Schroth-based exercises to conventional core training found greater improvements in curve angles, trunk rotation, and patient self-image scores in the Schroth group.²⁰

Bracing, the standard non-surgical treatment for moderate curves in growing adolescents, affects shoulder mechanics in ways that aren’t always positive. A study examining the immediate effects of rigid thoracolumbosacral braces found that they altered scapular movement significantly, increasing anterior tilt and reducing the range of overhead arm motion on both sides of the body.²¹ Soft bracing, a newer approach using engineered textiles, has shown some improvements in overall postural balance for mild scoliosis, though responses vary from person to person.²² The honest summary is that exercise and bracing can improve some aspects of shoulder alignment, but the evidence for dramatic, lasting shoulder rebalancing through conservative treatment alone is still thin.

Surface Scanning as an Alternative to X-Rays

Because shoulder balance is partly a cosmetic concern, measuring it with radiation-based imaging has always felt like overkill for monitoring purposes. Three-dimensional surface topography, where a scanner captures the shape of the body’s surface without any radiation, is increasingly used as a supplement or alternative. These systems can measure shoulder height differences, shoulder slope differences, and trunk asymmetry in a few seconds. One pilot study found that shoulder parameters made up 40% of a composite “anterior aesthetic deformity score” developed from surface scans, reflecting how central shoulder asymmetry is to the visible deformity of scoliosis.²³

Surface topographic scanning has been validated as a reliable method for evaluating spinal range of motion and surface asymmetry in both scoliotic and non-scoliotic adolescents, and it avoids the cumulative radiation exposure that comes with repeated X-rays during years of scoliosis monitoring.²⁴ For tracking how shoulder balance changes over time in a growing teenager, these scans offer a practical, radiation-free option. They also capture the three-dimensional reality of shoulder deformity better than a flat X-ray can, which matters given how poorly X-ray parameters predict what shoulders actually look like to the naked eye.

Handedness and Other Non-Scoliosis Causes of Uneven Shoulders

Not every case of one shoulder sitting higher than the other signals scoliosis. Habitual posture, heavy one-sided bag carrying, muscular imbalances from sports, and even handedness can create visible shoulder asymmetry. In right-handed women, the degree of right-handedness has been correlated with asymmetric activation of shoulder muscles during overhead tasks, meaning that dominant-side muscle development and recruitment patterns can subtly shift shoulder position over time.²⁵ This kind of functional asymmetry is usually small and doesn’t involve a structural spinal curve, but it can add to or mask a mild scoliotic shoulder imbalance.

Conditions like Sprengel’s deformity (a congenitally elevated scapula), Klippel-Feil syndrome, muscular torticollis, or a simple postural habit can all produce a visually higher shoulder without any lateral spinal curve. The classic screening test for scoliosis, the Adam’s forward bend test, helps distinguish structural scoliosis from these other causes. If one shoulder is higher but the spine appears straight and there’s no rib hump when bending forward, the asymmetry is more likely postural or muscular than scoliotic. When in doubt, a clinician can use a scoliometer or a single low-dose X-ray to check for a curve.