Clinical summary based on the natural history study by Murray, Weinstein and Spratt (JBJS Am 1993, 32 years of follow-up), Garrido's cohort of untreated patients (Spine Deform 2021, 27 years of follow-up), the assessment of radiographic criteria in the Rotterdam study (Makurthou, Spine 2013) and the only randomised trial of specific exercise published in this condition (Bezalel, Asian Spine J 2019): 58 references verified one by one on PubMed and cross-checked on CrossRef.

What should you take away in one reading?

Nine points that govern what to do. Each is taken up, detailed and sourced in the corresponding chapter.

Why is a Scheuermann kyphosis confused with bad posture?

It is the central error in the file, it is documented, and it is corrected in one minute of examination. This chapter sets out the test, then draws the consequences for treatment.

The complaint is nearly always the same. An adolescent, more often a boy, stands stooped. The parents have picked him up on it a hundred times. A PE teacher has spoken of "posture". They arrive at the clinic with a request for postural rehabilitation, and sometimes a prescription that says so in black and white.

How often is that reading wrong? Nobody knows precisely, but the orthopaedic literature has been flagging it as a recurring problem for a quarter of a century. Lowe, in two successive reviews, one of them devoted to the level of evidence, writes that the kyphotic deformity is frequently put down to "bad posture", which leads to a delay in diagnosis and treatment.15 A paediatric review puts it from the families' side: the parents of affected children often mistake it for bad posture.4 Twenty-five years later, the most recent concepts review repeats the same warning.2

The delay is not trivial, for a simple reason: the only treatment capable of altering the shape of the vertebrae works during growth and only during it.1330 A diagnosis made at seventeen in an adolescent at the end of growth closes a door that was open at thirteen.

The test that separates them

The distinction is mechanical, not aesthetic. A postural kyphosis is an attitude: the vertebral bodies are normal on the radiograph, the curve is supple and painless, it corrects by itself as soon as the person straightens up.3 By contrast, a Scheuermann kyphosis is a deformity: it is rigid, it rests on visible vertebral lesions, and in the wording of the 2023 review, the kyphotic deformity is rigid and does not correct passively in extension.2

The examination sequence takes four manoeuvres, in this order:

1. Standing, in profile. You look at the whole, not only the back: a thoracic hyperkyphosis is classically accompanied by compensatory lumbar and cervical hyperlordosis, with the head and neck carried forward.2
2. "Straighten up." This is active correction. A kyphotic attitude largely disappears. A structural kyphosis moves little.
3. Hyperextension, or prone. This is passive correction, and it is the one that settles the question. You ask for forced extension, or you place the adolescent prone, possibly over a bolster at the apex of the curve. Flexibility is assessed by physical examination in forced flexion and extension: that is what makes it possible to tell a postural kyphosis from a rigid one.2 A Scheuermann curve stays angular, with a short, sharp apex, instead of unfolding into a smooth arc.
4. The short chains. You look for the tightness that accompanies the deformity: iliopsoas, hamstrings, pectorals and the anterior aspect of the shoulder.2 They do not make the diagnosis, but they make part of the programme.

A kyphosis that corrects in hyperextension is not a Scheuermann. A kyphosis that does not correct is not a matter of posture.

The practical consequence, and it is uncomfortable

If the curve is structural, then postural correction exercises and reminders to stand up straight do not straighten it. This is not an editorial opinion: the 2023 concepts review concludes that there is no conclusive research demonstrating a lasting, significant improvement in kyphosis from exercise alone.2 A rehabilitation review devoted to the subject had already written it: physiotherapy is barely mentioned in the literature as an effective treatment for Scheuermann kyphosis, and there is little evidence that it can on its own alter the natural history of the disease.33

That does not mean physiotherapy has nothing to do. It means it changes target. What is measured and worked on is pain, the flexibility of the shortened chains, the strength of the trunk extensors and the monitoring of progression during growth. Chapter 7 sets out what each modality is really worth, with its level of evidence. Deficient trunk extension, in range as in strength, is in fact one of the few objective differences found at thirty-two years of follow-up between patients and matched controls: that is a direct argument for this target.26

What does the radiograph show, and from what threshold?

The diagnosis is radiographic. But the most cited criterion, the 45° angle, is also the most fragile: normal adolescent kyphosis is already close to it.

Sørensen's criterion, and what it says exactly

The reference investigation is a radiograph of the whole spine lateral, standing. The historical criterion, set out by Sørensen in 1964, is anterior wedging of at least 5° on at least three contiguous vertebrae.72 It matters to see what that criterion measures: the shape of the vertebrae, not the overall curve. A 70° kyphosis without wedging is not a Scheuermann; three wedged vertebrae with a 48° kyphosis is one.

Around that core, four signs classically accompany the picture and are read on the same film: irregularity of the vertebral endplates, Schmorl's nodes, narrowing of the disc spaces and disc degeneration.6 A Schmorl's node is a herniation of the nucleus pulposus through the cartilaginous and bony endplate into the adjacent vertebral body; it is common on imaging and most often asymptomatic, which rules out making a diagnosis from it alone.12

The problem with the 45° threshold

Many definitions add to the wedging criterion a thoracic kyphosis angle greater than 45°, measured between T5 and T12.13 That threshold circulates everywhere. It deserves a close look.

In 341 normal subjects aged 3 to 18, mean thoracic kyphosis is 44.0 ± 10.9°.9 In other words, the threshold supposed to define the abnormality almost coincides with the mean of normal. In adults, a systematic review of 34 studies covering 7,633 healthy people concludes that kyphosis exceeds 40° in 65 % of cases, and that the widely accepted 40° threshold therefore wrongly classifies a majority of normal subjects.10 Kyphosis also increases with age (Spearman correlation 0.52 between T5 and T12), with no difference between the sexes.10

An angle on its own therefore does not make the diagnosis. What does is the combination of characteristic wedging and clinical rigidity.

One last reason makes the overall angle uninformative on its own: thoracic kyphosis is not a single arc. In 373 asymptomatic young adults, it breaks down into two unequal arcs, a stable upper arc of 25.8° on average and a variable lower arc of 19.8° (p < 0.001), with an apex sitting around T8 and T9 depending on spinal morphology.11 Two patients with the same total angle may therefore have very different spines, and it is where the apex sits, not the sum, that governs mechanical behaviour.

What the choice of criteria changes: the measurement exists

The Rotterdam study did this work in 2,753 participants aged 45 to 89, applying Sørensen's criteria and those of Sachs to lateral radiographs.8 The results are instructive:

• 677 people (24.6 %) had endplate irregularities;
• 140 (5.1 %) vertebral wedging;
• 127 had both, of whom 111 with a kyphosis greater than 45°, giving a prevalence of 4.0 % (95 % CI 3.3 to 4.7) ;
• inter-observer agreement was 78.8 % for wedging and 79.4 % for endplate irregularity;
• and above all: lowering the threshold from 45° to 40° and then 35° changed prevalence only from 4.0 to 4.2 and then 4.4 %, with no significant difference.8

This last point is the most useful in the clinic. It is not the angular threshold that decides, it is the joint presence of the vertebral lesions. The authors indeed conclude that there is no absolute reference for the radiographic definition, but that standardised scoring of independent signs gives satisfactory inter-observer agreement.8

Thoracic or thoracolumbar form: why do they not behave the same?

Two locations, two presenting complaints, two decision thresholds. Treating a low form with the benchmarks of the high one leads to underestimating its severity.

The classic form sits in the thoracic spine and makes the hump. It is visible, it worries those around the patient, and it brings them to the clinic for cosmetic as much as painful reasons. The thoracolumbar or lumbar form differs on almost every point.

Blumenthal, Roach and Herring proposed the first classification of it in 1987, from cases at the Texas Scottish Rite Hospital.14 They describe two distinct radiographic pictures: a "classic" form and an "atypical" form made up of endplate changes, disc space narrowing and anterior Schmorl's nodes, without meeting Sørensen's criteria. Their observation matters in practice: this form tended to occur in more athletic adolescents, or in those with a history of increased axial loading of the spine.14

Greene, Hensinger and Hunter had described nineteen adolescents with this picture two years earlier. Symptoms sat mainly at the thoracolumbar junction, intense exertion or a traumatic event was clearly associated with the onset of trouble in 16 of the 19 patients, and a grade I or II spondylolysis or spondylolisthesis at L5-S1 was found in 32 % of them.17 None developed a progressive kyphotic deformity during follow-up, and most responded to rest, exercise and temporary avoidance of the activity involved.17

A comparative measurement supports this link with sporting load. In 120 elite skiers under 17 compared with 39 controls of the same age with no high-level practice, alpine skiers and ski jumpers had a significantly higher rate of anterior vertebral endplate lesions in the thoracolumbar spine. The authors put it down to excessive loads and repeated trauma to the immature spine at high speed, particularly in a forward-leaning position.18

The French update by Palazzo, Sailhan and Revel makes the observation that ought to change the clinical eye: lumbar involvement is probably as common as the thoracic form and may be more painful, and the course of thoracolumbar and lumbar forms is unknown.15 A recent CT study of 1,287 examinations in subjects aged 18 to 40 finds a prevalence of 2.8 % for the atypical form defined on the T8-L2 segment.16

The thresholds are not the same, and the gap is considerable

This is where the distinction stops being academic. Lowe writes that a kyphosis greater than 80° in the thoracic spine or 65° in the thoracolumbar spine is almost never treated successfully without surgery in a symptomatic patient.1 The JAAOS review puts the surgical indications at 70 to 75° for thoracic curves and to 25 to 30° for thoracolumbar curves.7

A figure of 28° at the junction is therefore anything but reassuring, whereas it would be in the thoracic spine. The reason is mechanical: the thoracolumbar junction is normally straight, and any kyphosis there is a reversal of the curve, not an exaggeration. A finite element model comparing twenty patients with a Scheuermann thoracolumbar kyphosis to twenty healthy subjects finds, in forward flexion, von Mises stresses up to 2.8 times higher at S1 and up to 2.33 times higher in the intervertebral disc.19

A methodological remark is needed here. The two large natural history series overwhelmingly concern thoracic forms: in Garrido, 57 of the 66 patients assessed had a thoracic kyphosis and only 9 a thoracolumbar deformity.27 What we know about the long-term course therefore describes mainly the high form. That is exactly what Palazzo flags in writing that the course of the low forms is unknown.15

Who develops Scheuermann's disease, and what makes it progress?

Strong heredity, a frequency that varies widely with the definition used, and a mechanism that remains a mechanical hypothesis on weakened ground.

A largely genetic disease

This is the best established point in the file, and it comes from a twin registry. Damborg and his team surveyed 46,418 Danish twins born between 1931 and 1982, from the Odense registry that covers every twin birth in the country over a hundred and thirty years.20 The results:

• self-reported prevalence of 2.8 %, that is 2.1 % in women and 3.6 % in men (p < 0.0001), in other words a boy-to-girl ratio close to 2 to 1 ;20
• pairwise concordance of 0.19 in monozygotic twins against 0.07 in dizygotic twins;20
• odds ratios of 32.9 in monozygotic twins against 6.25 in dizygotic twins (p < 0.01) ;20
• heritability of 74 %, confirmed five years later in 11,436 complete pairs with a 95 % confidence interval of 0.65 to 0.81, the environmental share being estimated at 0.26 (0.19 to 0.35).21

The same study finds no variation in prevalence over fifty years of ages examined.21 That is a strong argument against the idea of a disease caused by recent ways of living, and a direct counter-argument to the line that blames adolescent kyphosis on the phone or the school bag.

Prevalence depends mostly on the definition

Published figures range from 0.4 to 10 %.7 The low end corresponds to young populations examined with Sørensen's strict criteria: in 454 patients aged 15 to 40 screened by thoracic radiograph, prevalence is 2.2 % (95 % CI 0.9 to 3.5), with no significant difference between women (1.4 %) and men (2.8 %, p = 0.36), with excellent inter- and intra-observer agreement.22 That range is not a disagreement between authors: it is the direct consequence of the criteria used, the age of the subjects and the imaging method. The figure below sets them side by side with their populations.

Two practical lessons come out of it. First, the European EVOS study, conducted in 27 centres in more than ten thousand men and women over 50, finds a mean prevalence of 8 % identical in both sexes, but highly variable from one centre to another, high in Germany, Sweden, the United Kingdom and France, low in Hungary, Poland and Slovakia.23 Second, and this is useful in community practice, that same study establishes that Scheuermann's disease is not associated with bone mineral density, either at the spine or the hip: that is what makes it possible to tell it apart from an osteoporotic vertebral deformity on an older person's radiograph.23

The mechanism remains a hypothesis

The aetiology has not been elucidated. The dominant hypothesis describes an abnormality of growth in the vertebral endplate growth cartilage, which gives way under excessive mechanical loading during spinal growth, on genetically weakened ground.615 The multifactorial nature is accepted by all recent authors, and no unified theory has prevailed.26

Two observations put purely postural explanations into perspective. An analysis of 2,025 skeletons finds in affected subjects a significantly more horizontal sacrum than in controls (sacral orientation 44.4 ± 9.7° against 50.0 ± 9.9°, p < 0.001), which suggests a constitutional morphological contribution rather than a behavioural one.25 And the fact that prevalence has not moved in fifty years points the same way.21

What becomes of a Scheuermann kyphosis in adulthood?

It is the question every family asks, and there are two solid answers: one at 32 years of follow-up, the other at 27. They agree on the essentials and complement each other on the rest.

The landmark study: 67 patients, 32 years of follow-up

Murray, Weinstein and Spratt reviewed 67 patients whose mean kyphosis angle was 71°, on average thirty-two years after diagnosis (10 to 48 years), and compared them with 34 controls matched for age and sex.26 Fifty-four had a clinical examination and radiographs, fifty-two pulmonary function testing, forty-five a measurement of trunk muscle strength.

What was different about the patients:

• back pain that was more intense, and differently located;
• jobs that tended to require less physical activity ;
• one reduced trunk extension range ;
• one reduced trunk extension strength.26

What, by contrast, did not differ significantly from the controls: educational level, days off work for low back pain, interference of pain with daily activities, numbness in the lower limbs, concern about appearance, self-esteem, social limitations, use of pain medication and level of leisure activity.26 The authors further note that the patients reported little concern about their physical appearance.

On the respiratory side, function was normal or above normal in patients whose kyphosis stayed below 100°. A restrictive pattern appeared only in those whose kyphosis exceeded 100° and whose apex sat between T1 and T8.26 Finally, a mild scoliosis was common, and no spondylolisthesis was observed.26

At thirty-two years of follow-up, the measurable difference between a Scheuermann back and a control back comes down to four points: pain, its location, trunk extension and extensor strength. Three of those four are exactly what physiotherapy knows how to work on.

The modern confirmation: 27 years, and the rate quantified

Garrido and his team traced 113 untreated patients identified in a national database before the year 2000, at a time when the centre did not offer surgery. Sixty-six agreed to clinical assessment, forty-seven to radiological assessment. Mean age 45.1 years (31 to 65), mean follow-up 27 years (16–36).27

• The kyphosis went from 66° at skeletal maturity to 78° (p < 0.001).27
• Long-term progression is 0.45° a year (n = 47).27
• SRS-22 and SF-36 scores are lower and the Oswestry Disability Index higher than population normative values.27
• And above all: it is the sagittal vertical axis that correlates with disability (r = 0.59; p = 0.001 with the Oswestry), not the kyphosis angle on its own.27 Level of evidence III.

A natural history review published in 2025 converges: after maturity, the curve grows by only a third to half a degree a year, very few patients ever exceed 100°, pain and daily function generally remain acceptable, and radiographic size alone does not predict symptoms: it is the combination of a deformity and a loss of global alignment that flags the situations to watch.28

The qualification that changes what you say to the patient

A Swedish cohort provides a counterpoint it would be dishonest to leave out. In the MrOS Sweden study, 1,417 men aged 69 to 81 had readable radiographs: 92 of them (6.5 %, 95 % CI 5.3 to 7.9) had Scheuermann's disease.24 Yet 51 % of those affected reported back pain against 55 % of those unaffected (p = 0.4), and 31 % neck pain in both groups (p = 0.90). Among the affected men with back pain, none reported severe pain: 57 % moderate, 43 % mild, against 7 % severe in the unaffected.24

These results do not contradict Murray or Garrido: they concern much older subjects, not recruited for a deformity, and therefore on average less severe forms. But they license a message families rarely hear: at that age, in that population, the disease was not associated with pain. Tsirikos and Jain sum up the consensus position: a benign prognosis, with settling of symptoms and stabilisation of the deformity at skeletal maturity, is expected in most patients.29

The brace: for whom, when, how many hours and for what gain?

It is the only non-surgical treatment shown to alter the shape of the vertebrae. It has three conditions, and they are not negotiable.

The principle is bone remodelling: during growth, sustained pressure can redirect the growth of wedged vertebral bodies.13 The consequence fits in one sentence: outside growth, the brace remodels nothing. In an adult it may relieve, it does not correct.

The three conditions for effectiveness

1. Growth remaining. Treatment must be started before skeletal maturity; every series reporting remodelling did so in growing subjects.23013 One review even concludes that the brace is more effective if the diagnosis is made before the curve exceeds 50° in a patient who is still growing.33
2. A curve in the right window. The 2023 concepts review speaks of a result that is reproducible and broadly favourable for kyphoses between 50 and 80° if treatment is started before skeletal maturity.2 Lowe gave a window of 55 to 80°.1
3. Real wearing time. The same review is explicit: a minimum of 16 hours a day is required until correction is obtained.2 The Italian series with ten years of follow-up prescribed 16 to 20 hours a day.13

What the series actually measure

The historical series remains that of the Twin Cities Scoliosis Center. Of 274 patients treated with a Milwaukee brace between 1960 and 1978, 120 were analysed with at least five years of follow-up after the end of treatment.30 Among those who wore the brace diligently, 76 improved, 24 worsened and 10 stayed stable; seven of the twenty-four who worsened had been operated on before the review. Among the ten patients with irregular wear, two improved and eight worsened, three of them fused.30 And one figure that bounds expectations: four of the fourteen patients whose initial kyphosis exceeded 74° ended up with a fusion.30

Three recent series refine that result:

• Aulisa 2023, 158 patients with a thoracic kyphosis, antigravity brace 16 to 20 hours a day, mean age 14, mean treatment duration 28 months. The Cobb angle goes from 57.6 ± 6.3° to 43.3 ± 7.8° at the end of treatment, and holds at 44.5 ± 7.4° ten years later. The wedging angle of the apical vertebra goes from 14.4 ± 2.5° to 8.6 ± 3.6° and holds there.13 It is the most direct demonstration of vertebral remodelling.
• Piazzolla 2021, 192 adolescents, mean age 13.1 years, mean thoracic kyphosis 61.9 ± 11.3°, mean follow-up 57 months. Self-reported adherence good in 84.9 %. Correction in the brace of 37.4 %, final correction of 31.6 %. At last follow-up, 60.4 % curve reduction and 29.7 % stabilisation. Mental health and self-image scores, lower at baseline in the most deformed group, improve significantly.31
• Etemadifar 2017, Milwaukee brace combined with physiotherapy: mean angle of 63.2 ± 10.0° at presentation, 36.5 ± 13.4° at the end of the study (p < 0.001), and progression halted in 97.5 % of patients, including for curves of up to 90°.32

Three caveats that have to be stated

First caveat: part of the correction is lost. Tsirikos and Jain report a loss of correction after the brace is stopped, with recurrence of anterior wedging, in about a third of patients.29 Aulisa's series, on the other hand, shows it holding at ten years: the difference probably lies in the protocol and the weaning, carried out in Aulisa's case until complete recovery of vertebral geometry or the end of growth.13

Second caveat: we do not know who would have progressed without a brace. That is the most serious methodological objection, and it is put by the authors themselves: since we cannot predict which curves will progress, we cannot determine the effectiveness of brace treatment.33

Third caveat: the overall level of evidence remains low. A review devoted to spinal deformities concludes that for hyperkyphosis, there is no high-quality evidence in favour of physiotherapy, bracing or surgery in adolescents and adults, while acknowledging that the brace reduces thoracic hyperkyphoses between 55 and 80° in adolescents.34 A systematic review of 45 studies covering 1,829 patients further confirms that both surgical approaches correct more than the brace, which is expected, without that settling the question of clinical benefit.35

What is exercise really worth in Scheuermann's disease?

The honest answer comes in two parts: the volume of evidence is very low, and what is proven does not target what patients expect.

First, measure the volume of evidence

Before discussing a modality, you have to know what you are discussing. MEDLINE indexes 708 references under the MeSH descriptor "Scheuermann Disease". Among them: 4 randomised trials, 7 systematic reviews, 47 papers crossing the subject with the descriptor "Braces" and 23 with "Exercise Therapy". On the same database and with the same query method, scoliosis has 332 randomised trials.

That imbalance is nothing new. A systematic review devoted to the subject already put it in 2009: most published opinions and recommendations are weakly supported by levels of evidence, and prospective studies with validated questionnaires and long follow-up comparing treated, operated and untreated patients, which would document the natural history, remain unavailable.57 Seventeen years later, the observation still holds.

These counts are reproducible: they come from PubMed's E-utilities interface, queried on 15 August 2026 with the queries "Scheuermann Disease"[MeSH] AND randomized controlled trial[pt] and "Scoliosis"[MeSH] AND randomized controlled trial[pt]. They measure an indexing, not the truth of the literature. The evidence lies in the four trials themselves, which are worth naming:

• one trial of rehabilitation, suspension therapy against ball exercises in 40 adolescent girls;38
• one comparison of conservative treatment against surgery in 63 adolescents with a curve of 50 to 65°;46
• one surgical trial on the choice of distal fusion level in 54 patients;45
• and a comparative study of elite skiers, indexed as a randomised trial, which is not about a treatment.18

In other words, only one of those four trials evaluates a rehabilitation programme. And the count further misses the most directly relevant trial, Bezalel's on Schroth therapy: it is not indexed under that pairing of descriptor and publication type.36 That is the limitation inherent in this kind of measurement, and it is better stated than hidden: the void counted is a void of indexing. The order of magnitude, though, is not open to argument.

The randomised trial that does concern an exercise programme

There is only one, and it deserves close reading rather than blanket citation. Bezalel and his team allocated 50 young adults with Scheuermann's disease between Schroth therapy (n = 25) and conventional antigravity exercises (n = 25), with one individual session a week for a few weeks, daily home exercises logged in a diary, and follow-up of twelve months. Single-blind randomised trial.36

• Significant effect of time on thoracic kyphosis (F[1] = 5.72; p = 0.02) and on the kyphotic apex line (F[1] = 5.76; p = 0.02).36
• Significant group by time interaction on thoracic kyphosis (F[1] = 4.91; p = 0.03) and on the deformity measured with an inclinometer (F[1] = 4.05; p = 0.02): the Schroth group improves more.36
• The authors conclude that back exercises in general, and Schroth therapy in particular, constitute an effective treatment for the thoracic Cobb angle and for symptoms.36

Two caveats are needed, and they do not come from us. First, a methodological review of Schroth trials published in 2023 examined the seven existing trials, of which only one concerned Scheuermann kyphosis: the Schroth method was genuinely used in only four of the seven, only three used certified therapists, none had prospectively registered its protocol, and overall methodological rigour is judged sub-optimal, to the point of compromising evidence synthesis.37 Second, the trial concerns young adults, not growing adolescents: transposing its result to the moment when bracing is under discussion would be an abuse.

A second randomised trial deserves citing for what it is. Forty adolescent girls aged 16.3 ± 0.5 years with juvenile vertebral osteochondrosis (ICD-10 code M42.0) were allocated between suspension therapy and ball exercises, at fifteen thirty-minute sessions over three weeks. Both groups improve; suspension does better on pain, trunk muscle endurance and standing posture.38 Three weeks of follow-up says nothing about the trajectory of the curve, and the M42.0 code covers a broader set than Scheuermann's disease alone.

What exercise can legitimately target

The target is not the angle. It is what the measured deficit points to: at thirty-two years of follow-up, patients had reduced trunk extension range and strength, and pain that was more intense and differently located.26 Those three targets can be worked on, measured in the clinic, and require no promise of straightening. To them can be added the consistently described tightness of the psoas, hamstrings, pectorals and anterior aspect of the shoulder.2

When does surgery become an option, and at what price?

Rare, but not exceptional. And the most useful piece of data for the physiotherapist is the one showing that it is not the angle that decides.

The indications as they are written

The JAAOS review lists the situations that put fusion up for discussion: a thoracic curve greater than 70–75°, a thoracolumbar curve greater than 25–30°, intractable pain, neurological deficit, cardiopulmonary impact, or major cosmetic harm.7 Lowe adds documented progression and loss of sagittal balance.1 The German review takes a Stagnara angle greater than 70 or 75° combined with progression despite conservative treatment, increasing pain or a neurological deficit.47

What really decides: the most instructive figure in the file

The spinal deformity study group followed 150 consecutive patients with a Scheuermann kyphosis, 77 of them treated without surgery and 73 operated on, the choice resting with the surgeon and the patient.40 The comparison of the two groups is revealing:

The line that counts is the fourth: the maximum sagittal Cobb angle did not differ significantly between operated and non-operated patients (73 against 70°, p = 0.11).40 What did differ was pain, self-image, age and body mass index. The authors conclude that factors other than the radiographic measurement probably contribute to the surgical decision.40 Level of evidence II.

Two adolescents with the same angle will not follow the same path. What separates them is what they experience, not what the radiograph measures. That is also what gives the physiotherapy consultation its weight.

The price: junctional kyphosis

The characteristic complication is junctional kyphosis, proximal or distal, that is the appearance of an angulation of more than 10° above or below the construct.1 The figures:

• In a multicentre series of 78 patients of mean age 16.7 years, maximum kyphosis goes from 78.8° to 51.4°; proximal junctional kyphosis of at least 10° occurs in 25 patients (32.1 %) and distal junctional kyphosis in 4 (5.1 %). Loss of correction is 3.2° after a combined approach and 6.4° after a posterior-only approach.41
• In adults, a series of 22 patients of median age 23 years with two years of follow-up finds at least one complication in 77 % of them, a major complication in 32 %, proximal junctional kyphosis in 55 %, and a revision rate of 18 % at two years.42
• The choice of distal level counts: in 94 patients, distal junctional kyphosis occurs in 59 % when the lowest instrumented vertebra sits one vertebra below the first lordotic disc, against 22 % when it sits two vertebrae lower (odds ratio 3.2; 95 % CI 1.28 to 8.18).44
• This point is disputed. The only published randomised trial on the question, in 54 patients allocated by the distal level chosen, finds no significant difference in distal junctional kyphosis between the groups (3, 5 and 3 cases), and concludes that there is no need to extend the fusion to the sagittal stable vertebra.45 The two pieces of work do not contradict each other term for term, but they do not point the same way: that is the real state of the file.
• Lowe traces these complications to two technical causes: not including every level of the kyphosis, or overcorrecting by more than 50 %.1

On the surgical approach, a meta-analysis of 13 studies and 586 patients (300 with a combined anterior and posterior approach, 286 with a posterior-only approach) finds no significant difference, neither in the correction obtained (33.3°; 95 % CI 27.5 to 39.2 against 31.2°; 27.0 to 35.4), nor in the incidence of proximal junctional kyphosis. The posterior-only approach comes with less blood loss and a shorter operating time, which explains the current trend.4335

A Jordanian study deserves a mention and a caveat. Sixty-three adolescents with a kyphosis of 50 to 65° were followed, conservative treatment against surgery, and the operated group does better on every score.46 But the authors themselves put the deterioration of the conservative group down to poorer cooperation from the patients, which made the curve less supple for exercises and the brace.46 So it is not a comparison of treatments, it is a comparison of treatments as delivered, one of which was poorly followed. Conversely, this study indirectly confirms what Sachs says: adherence decides.30

What do concrete clinical cases teach us?

Five published case reports, chosen because each illustrates a point the series do not show.

Two brothers in presentation, two opposite trajectories

Lemire and colleagues report two boys seen in the same year, and the interest of that publication lies in their contrast.50

The first is 13 and has had low back pain for two years. Radiographs show irregularity of the upper lumbar endplates, with Schmorl's nodes. No wedging, no hump: this is the lumbar form. Conservative treatment, improvement maintained at six months of follow-up.50 That is exactly the picture chapter 3 describes: low pain with no visible deformity, in an adolescent nobody would have referred for "bad posture".

The second is 14 and attends orthopaedics. Radiographs show anterior wedging of T6, T7 and T8 with a thoracic kyphosis of 72°. A brace is prescribed; he does not wear it regularly. The deformity progresses by 72°–92°. He is operated on. Six years later he is well, with no pain and no complications, with a curve at 65°.50

That second case is the individual translation of Sachs's figure: eight of the ten patients with irregular wear worsened, three were fused.30 Twenty degrees lost is the cost of adherence that went unsupported. The authors indeed conclude that significant progression remains rare in both types, typical and atypical, but that it can happen.50

The case that reminds you why you examine the neurology

Kapetanos and his team describe a boy aged 14 years with Scheuermann's disease, presenting with a significant neurological deficit due to a thoracic disc herniation at the apex of the kyphosis. He was treated with anterior decompression and combined anterior and posterior fusion in one sitting, with a plate, a cage and segmental instrumentation. Complete neurological recovery.49

That is rare. But 9 % of a surgical series of 69 cases had an abnormal neurological examination preoperatively, and the abnormality sometimes went no further than simple trunk paraesthesia, easy not to look for.48

The case that questions the limit of growth

Berdishevsky reports a woman aged 76 years, diagnosed in adolescence, with a T1-T12 kyphosis of 85°, a lumbar hyperlordosis of 70°, a lumbar scoliosis of 21° and a sagittal trunk translation of 4.5 cm. Intermittent low back pain rated 6 out of 10 at worst, SRS-22 quality of life 3.8. Protocol: one-hour exercise sessions three times a week for six months following the Barcelona school, a home programme, and a SpinoMed brace two hours a day. At six months, the kyphosis is 70° and the lordosis 57°; a year later, 64° and 55°. Pain at 2, SRS-22 at 4.5, translation improved by 2.2 cm.39

This case has to be read for what it is: a single patient, with no control group, and with part of the angular variation attributable to radiographic positioning. The author claims nothing more. It is cited here because it is the only published case with imaging in an older adult, and because it forbids concluding too quickly that there is nothing left to do after growth. It grounds no recommendation.

The sportsperson's case, and the family's

Wilson and Lindseth describe three patients whose back pain was made worse by the butterfly stroke, and in whom the diagnosis of Scheuermann kyphosis was made subsequently. They were treated conventionally, but allowed to swim outside the brace and encouraged to do so, while avoiding butterfly. Mean curve correction of 27 % at a mean follow-up of 1.6 years, symptoms resolved in every case.53 The authors' conclusion concerns the athlete's psychological and emotional wellbeing: continued participation did not compromise the result.

Finally, Dai and his team describe a Chinese family of 17 members across three generations, three of whom had a lumbar form with wedging from L1 to L3 and Cobb angles of 37, 70 and 73°.52 It is the clinical illustration of the 74 % heritability measured in the Danish twins,20 and a reminder that the family history is worth taking. A case of atypical lumbar form in a man of 18 with chronic low back pain without radiation or deficit further recalls that the diagnosis remains one of exclusion when the classic criteria are not met.51

How do you apply this in the clinic?

What the physiotherapist can do from the first session, what they must refer on, and what they must not promise.

The first session

1. Do the test before anything else. Active correction, then hyperextension or prone. Record it in the notes: it is the piece of data that directs everything, and it is reproducible from one session to the next.2
2. Measure what you are going to follow. Trunk extension range, extensor endurance, hamstring and pectoral flexibility, pain on a numerical scale. Those four measures are exactly the ones that differed between patients and controls at thirty-two years of follow-up.26
3. Examine the neurology. Sensation in the trunk and lower limbs, reflexes, strength. Nine per cent of abnormalities in a surgical series justifies devoting three minutes to it.48
4. Place the growth. Spurt under way, bone age if available, pubertal stage. That is what determines whether the brace window is open or closed.1330
5. Refer if necessary. No recent radiograph, a rigid curve, or an angle close to the surgical thresholds: orthopaedic opinion.7

What is said, and what is not said

The shared context of growth

Scheuermann's disease belongs to the family of growth osteochondroses, in the same way as the apophyseal conditions of the lower limb that the sporting adolescent presents with at the knee or the heel. They share a background: growth cartilage subjected to repeated loading during a growth spurt. An older series of 185 osteochondroses in young athletes already placed Scheuermann's disease among the most common, alongside Osgood-Schlatter disease and Sever's disease.58 The spine, however, obeys its own thresholds and its own imaging: what precedes does not transpose to the lower limb, and the reverse is true too.

Frequently asked questions

Can a Scheuermann kyphosis be corrected with exercises?

No, not the curve itself. The most recent concepts review concludes that there is no conclusive research demonstrating a lasting, significant improvement in kyphosis from exercise alone.2 The only randomised trial available does show a gain in angle with a specific programme in young adults, but in 50 participants and with the methodological caveats the whole field carries.3637 What exercise defensibly improves is pain, flexibility and extensor strength.26

From what angle do we speak of Scheuermann's disease?

The question is badly put, and that is instructive. The criterion is not an angle but a vertebral deformity: at least 5° of anterior wedging on at least three contiguous vertebrae.7 The 45° threshold often added discriminates poorly, since normal thoracic kyphosis in children and adolescents is already 44 ± 11°,9 and varying it from 45 to 35° changed the measured prevalence only from 4.0 to 4.4 %.8

Should my teenager give up sport?

Nothing in the literature justifies that as a general rule. Three swimmers whose pain was brought on by butterfly were treated conventionally while continuing to swim, avoiding that one stroke: 27 % mean correction and resolution of symptoms.53 Conversely, the thoracolumbar form is associated with repeated axial loading,14 and Greene's series finds a sporting or traumatic trigger in 16 of the 19 patients, who improved with rest and temporary avoidance of the activity involved.17 The sensible approach is therefore to adapt the painful movement, not to stop the activity.

Does the brace work in adults?

It no longer remodels. The mechanism assumes growth is under way: it is growth that allows wedged vertebral bodies to be redirected.13 In an adult, a brace may relieve, it does not correct. A single published case with imaging describes an angular improvement in a 76-year-old woman on intensive exercise and a brace worn two hours a day: it is an isolated case, with no control, and it grounds no practice.39

Is an MRI needed?

Not routinely. A standing lateral radiograph of the whole spine is enough for the diagnosis in the great majority of cases.6 MRI is justified in the face of a neurological sign, however subtle, and in the preoperative work-up: in one surgical series, 9 % of patients had an abnormal neurological examination before the operation.48

Is Scheuermann's disease painful in adulthood?

The data diverge according to the population studied, and that is worth knowing. At thirty-two years of follow-up, in patients whose mean kyphosis was 71°, pain was more intense than in controls, with no difference in sickness absence, interference with daily activities or self-esteem.26 In 1,417 Swedish men aged 69 to 81 not recruited for a deformity, by contrast, the disease was associated with neither neck pain nor back pain, and no affected man reported severe pain.24

Can you have Scheuermann's disease without a hump?

Yes, and that is the trap of chapter 3. The thoracolumbar or lumbar form shows itself as pain at the junction, often in a sporting adolescent, with irregular endplates, anterior Schmorl's nodes and disc space narrowing, without meeting the classic wedging criteria.14 Lumbar involvement is reportedly as common as the thoracic form, and potentially more painful.15

Is this disease linked to osteoporosis?

No, and the demonstration is explicit. The European EVOS study, conducted in 27 centres in more than ten thousand people over 50, finds a prevalence of 8 % with no association whatsoever with bone mineral density at the spine or the hip.23 The authors present that result as an aid to differential diagnosis: a wedged vertebra in an older person is not necessarily an osteoporotic fracture.

What is the risk of ending up operated on?

It depends mostly on the starting angle and on adherence. In the historical Milwaukee brace series, four of the fourteen patients whose initial kyphosis exceeded 74° ended up fused, as did three of the ten patients with irregular wear.30 The decision itself is not dictated by the angle: in a cohort of 150 patients, the maximum sagittal Cobb angle did not differ between operated and non-operated patients (73 against 70°, p = 0.11); what did differ was pain and self-image.40

Is there a link with scoliosis?

The two often coexist. A mild scoliosis was common in the patients followed for thirty-two years,26 and a family with a lumbar form also had idiopathic scoliosis.52 They are nonetheless two distinct deformities, one in the frontal plane with rotation, the other in the sagittal plane: their screening, their thresholds and their treatments do not overlap. The subject is covered separately in our article on scoliosis.

Bibliography

The 58 references in this article

1. Lowe TG. Scheuermann's kyphosis. Neurosurg Clin N Am. 2007;18(2):305-15. PMID 17556132 · doi:10.1016/j.nec.2007.02.011
2. O'Donnell JM, Wu W, Youn A, et al. Scheuermann Kyphosis: Current Concepts and Management. Curr Rev Musculoskelet Med. 2023;16(11):521-530. PMID 37615931 · doi:10.1007/s12178-023-09861-z
3. Garoflid N, Fragnière B, Dutoit M. ["Round back" in children and adolescents]. Rev Med Suisse Romande. 2000;120(10):815-20. PMID 11109912
4. Ali RM, Green DW, Patel TC. Scheuermann's kyphosis. Curr Opin Pediatr. 1999;11(1):70-5. PMID 10084088 · doi:10.1097/00008480-199902000-00014
5. Lowe TG, Line BG. Evidence based medicine: analysis of Scheuermann kyphosis. Spine (Phila Pa 1976). 2007;32(19 Suppl):S115-9. PMID 17728677 · doi:10.1097/BRS.0b013e3181354501
6. Kaur S, Lalam R. Scheuermann's Disease. Semin Musculoskelet Radiol. 2023;27(5):522-528. PMID 37816360 · doi:10.1055/s-0043-1771334
7. Sardar ZM, Ames RJ, Lenke L. Scheuermann's Kyphosis: Diagnosis, Management, and Selecting Fusion Levels. J Am Acad Orthop Surg. 2019;27(10):e462-e472. PMID 30407981 · doi:10.5435/JAAOS-D-17-00748
8. Makurthou AA, Oei L, El Saddy S, et al. Scheuermann disease: evaluation of radiological criteria and population prevalence. Spine (Phila Pa 1976). 2013;38(19):1690-4. PMID 24509552 · doi:10.1097/BRS.0b013e31829ee8b7
9. Mac-Thiong JM, Labelle H, Berthonnaud E, et al. Sagittal spinopelvic balance in normal children and adolescents. Eur Spine J. 2007;16(2):227-34. PMID 16311754 · doi:10.1007/s00586-005-0013-8
10. Zappalá M, Lightbourne S, Heneghan NR. The relationship between thoracic kyphosis and age, and normative values across age groups: a systematic review of healthy adults. J Orthop Surg Res. 2021;16(1):447. PMID 34243795 · doi:10.1186/s13018-021-02592-2
11. Sebaaly A, Silvestre C, Rizkallah M, et al. Revisiting thoracic kyphosis: a normative description of the thoracic sagittal curve in an asymptomatic population. Eur Spine J. 2021;30(5):1184-1189. PMID 33222002 · doi:10.1007/s00586-020-06670-7
12. Kyere KA, Than KD, Wang AC, et al. Schmorl's nodes. Eur Spine J. 2012;21(11):2115-21. PMID 22544358 · doi:10.1007/s00586-012-2325-9
13. Aulisa AG, Marsiolo M, Calogero V, et al. Long-term outcome after brace treatment of Scheuermann's kyphosis: an observational controlled cohort study. Eur J Phys Rehabil Med. 2023;59(4):529-534. PMID 37746785 · doi:10.23736/S1973-9087.23.08070-X
14. Blumenthal SL, Roach J, Herring JA. Lumbar Scheuermann's. A clinical series and classification. Spine (Phila Pa 1976). 1987;12(9):929-32. PMID 3441839
15. Palazzo C, Sailhan F, Revel M. Scheuermann's disease: an update. Joint Bone Spine. 2014;81(3):209-14. PMID 24468666 · doi:10.1016/j.jbspin.2013.11.012
16. Souza GV, Oki LY, Portelinha AM, et al. Study of the Prevalence of Atypical Scheuermann's Kyphosis Using Computed Tomography Scans. Rev Bras Ortop (Sao Paulo). 2024;59(6):e854-e860. PMID 39711633 · doi:10.1055/s-0044-1792119
17. Greene TL, Hensinger RN, Hunter LY. Back pain and vertebral changes simulating Scheuermann's disease. J Pediatr Orthop. 1985;5(1):1-7. PMID 3156876 · doi:10.1097/01241398-198501000-00001
18. Rachbauer F, Sterzinger W, Eibl G. Radiographic abnormalities in the thoracolumbar spine of young elite skiers. Am J Sports Med. 2001;29(4):446-9. PMID 11476384 · doi:10.1177/03635465010290041101
19. Wan C, Shen X, Wu X, et al. Assessing the biomechanics of scheuermann's kyphosis affected thoracolumbar spine in forward flexion at the tissue-level using a finite element model. Sci Rep. 2025;15(1):27408. PMID 40721849 · doi:10.1038/s41598-025-12968-7
20. Damborg F, Engell V, Andersen M, et al. Prevalence, concordance, and heritability of Scheuermann kyphosis based on a study of twins. J Bone Joint Surg Am. 2006;88(10):2133-6. PMID 17015588 · doi:10.2106/JBJS.E.01302
21. Damborg F, Engell V, Nielsen J, et al. Genetic epidemiology of Scheuermann's disease. Acta Orthop. 2011;82(5):602-5. PMID 21895506 · doi:10.3109/17453674.2011.618919
22. Urrutia J, Narvaez F, Besa P, et al. Scheuermann's disease in patients 15-40 years old: A study to determine its prevalence and its relationship with age and sex using chest radiographs as screening tool. J Orthop Sci. 2019;24(5):776-779. PMID 30685093 · doi:10.1016/j.jos.2018.12.024
23. Armbrecht G, Felsenberg D, Ganswindt M, et al. Vertebral Scheuermann's disease in Europe: prevalence, geographic variation and radiological correlates in men and women aged 50 and over. Osteoporos Int. 2015;26(10):2509-19. PMID 26021761 · doi:10.1007/s00198-015-3170-6
24. Jönsson A, Damm H, Hofvander M, et al. In old men Scheuermann's disease is not associated with neck or back pain: a Swedish cohort study. Acta Orthop. 2023;94:236-242. PMID 37170780 · doi:10.2340/17453674.2023.12358
25. Peleg S, Dar G, Steinberg N, et al. Sacral orientation and Scheuermann's kyphosis. Springerplus. 2016;5:141. PMID 26933639 · doi:10.1186/s40064-016-1772-x
26. Murray PM, Weinstein SL, Spratt KF. The natural history and long-term follow-up of Scheuermann kyphosis. J Bone Joint Surg Am. 1993;75(2):236-48. PMID 8423184 · doi:10.2106/00004623-199302000-00011
27. Garrido E, Roberts SB, Duckworth A, et al. Long-term follow-up of untreated Scheuermann's kyphosis. Spine Deform. 2021;9(6):1633-1639. PMID 34212306 · doi:10.1007/s43390-021-00354-y
28. Bennett DM. The Natural History of Scheuermann's Kyphosis. J Pediatr Soc North Am. 2025;13:100263. PMID 41126901 · doi:10.1016/j.jposna.2025.100263
29. Tsirikos AI, Jain AK. Scheuermann's kyphosis; current controversies. J Bone Joint Surg Br. 2011;93(7):857-64. PMID 21705553 · doi:10.1302/0301-620X.93B7.26129
30. Sachs B, Bradford D, Winter R, et al. Scheuermann kyphosis. Follow-up of Milwaukee-brace treatment. J Bone Joint Surg Am. 1987;69(1):50-7. PMID 3100538
31. Piazzolla A, Bizzoca D, Solarino G, et al. Maria Adelaide brace in the management of Scheuermann's Kyphosis. Spine Deform. 2021;9(2):549-557. PMID 33206353 · doi:10.1007/s43390-020-00225-y
32. Etemadifar MR, Jamalaldini MH, Layeghi R. Successful brace treatment of Scheuermann's kyphosis with different angles. J Craniovertebr Junction Spine. 2017;8(2):136-143. PMID 28694598 · doi:10.4103/jcvjs.JCVJS_38_16
33. Bezalel T, Carmeli E, Been E, et al. Scheuermann's disease: current diagnosis and treatment approach. J Back Musculoskelet Rehabil. 2014;27(4):383-90. PMID 24898440 · doi:10.3233/BMR-140483
34. Bettany-Saltikov J, Turnbull D, Ng SY, et al. Management of Spinal Deformities and Evidence of Treatment Effectiveness. Open Orthop J. 2017;11:1521-1547. PMID 29399227 · doi:10.2174/1874325001711011521
35. Huq S, Ehresman J, Cottrill E, et al. Treatment approaches for Scheuermann kyphosis: a systematic review of historic and current management. J Neurosurg Spine. 2020;32(2):235-247. PMID 31675699 · doi:10.3171/2019.8.SPINE19500
36. Bezalel T, Carmeli E, Levi D, et al. The Effect of Schroth Therapy on Thoracic Kyphotic Curve and Quality of Life in Scheuermann's Patients: A Randomized Controlled Trial. Asian Spine J. 2019;13(3):490-499. PMID 30669825 · doi:10.31616/asj.2018.0097
37. Schreiber S, Whibley D, Somers EC. Schroth Physiotherapeutic Scoliosis-Specific Exercise (PSSE) Trials-Systematic Review of Methods and Recommendations for Future Research. Children (Basel). 2023;10(6). PMID 37371186 · doi:10.3390/children10060954
38. Dudoniene V, Kuisma R, Juodzbaliene V. Sling suspension therapy is an effective treatment method of juvenile spinal osteochondrosis in adolescent girls. J Back Musculoskelet Rehabil. 2022;35(4):873-879. PMID 34744069 · doi:10.3233/BMR-210099
39. Berdishevsky H. Outcome of intensive outpatient rehabilitation and bracing in an adult patient with Scheuermann's disease evaluated by radiologic imaging-a case report. Scoliosis Spinal Disord. 2016;11(Suppl 2):40. PMID 27785478 · doi:10.1186/s13013-016-0094-7
40. Polly DW, Ledonio CGT, Diamond B, et al. What Are the Indications for Spinal Fusion Surgery in Scheuermann Kyphosis?. J Pediatr Orthop. 2019;39(5):217-221. PMID 30969249 · doi:10.1097/BPO.0000000000000931
41. Lonner BS, Newton P, Betz R, et al. Operative management of Scheuermann's kyphosis in 78 patients: radiographic outcomes, complications, and technique. Spine (Phila Pa 1976). 2007;32(24):2644-52. PMID 18007239 · doi:10.1097/BRS.0b013e31815a5238
42. Heegaard M, Bari TJ, Ohrt-Nissen S, et al. Complications Following Surgery for Adult Scheuermann's Kyphosis: A 2-Year Follow-Up in 22 Patients. Turk Neurosurg. 2022;32(3):471-480. PMID 35416264 · doi:10.5137/1019-5149.JTN.35974-21.2
43. Li Q. Surgical Procedures Used for Correction of Scheuermann's Kyphosis: A Meta-Analysis. Pain Res Manag. 2021;2021:2142964. PMID 34725561 · doi:10.1155/2021/2142964
44. Luzzi A, Sardar Z, Cerpa M, et al. Risk of distal junctional kyphosis in scheuermann's kyphosis is decreased by selecting the LIV as two vertebrae distal to the first lordotic disc. Spine Deform. 2022;10(6):1437-1442. PMID 35840788 · doi:10.1007/s43390-022-00542-4
45. Yanik HS, Ketenci IE, Coskun T, et al. Selection of distal fusion level in posterior instrumentation and fusion of Scheuermann kyphosis: is fusion to sagittal stable vertebra necessary?. Eur Spine J. 2016;25(2):583-9. PMID 26195078 · doi:10.1007/s00586-015-4123-7
46. Audat ZA, Kheirallah KA, Ababneh BF, et al. Assessment of Quality of Life for Scheuermann's Kyphosis Patients with Cobb's Angle 50°-65° Treated Conservatively or Surgically in North Jordan: A Prospective Comparative Study. Clin Orthop Surg. 2022;14(2):244-252. PMID 35685985 · doi:10.4055/cios20219
47. Diaremes P, Braun S, Meurer A. [Scheuermann's disease]. Orthopade. 2022;51(4):339-348. PMID 35290495 · doi:10.1007/s00132-022-04239-4
48. Cho W, Lenke LG, Bridwell KH, et al. The prevalence of abnormal preoperative neurological examination in Scheuermann kyphosis: correlation with X-ray, magnetic resonance imaging, and surgical outcome. Spine (Phila Pa 1976). 2014;39(21):1771-6. PMID 25029218 · doi:10.1097/BRS.0000000000000519
49. Kapetanos GA, Hantzidis PT, Anagnostidis KS, et al. Thoracic cord compression caused by disk herniation in Scheuermann's disease: a case report and review of the literature. Eur Spine J. 2006;15 Suppl 5(Suppl 5):553-8. PMID 16421745 · doi:10.1007/s00586-005-0053-0
50. Lemire JJ, Mierau DR, Crawford CM, et al. Scheuermann's juvenile kyphosis. J Manipulative Physiol Ther. 1996;19(3):195-201. PMID 8728463
51. Singh M, Bansal M, Singh A, et al. Atypical Lumbar Scheuermann's Disease: A Rare Entity and Literature Review. J Orthop Case Rep. 2023;13(2):48-50. PMID 37144077 · doi:10.13107/jocr.2023.v13.i02.3552
52. Dai Y, Li Y, Li P, et al. Familial lumbar Scheuermann disease with idiopathic scoliosis in China: First case report. Medicine (Baltimore). 2017;96(25):e7100. PMID 28640085 · doi:10.1097/MD.0000000000007100
53. Wilson FD, Lindseth RE. The adolescent "swimmer's back". Am J Sports Med. 1982;10(3):174-6. PMID 6214197 · doi:10.1177/036354658201000309
54. Sugrue PA, OʼShaughnessy BA, Blanke KM, et al. Rapidly progressive Scheuermann's disease in an adolescent after pectus bar placement treated with posterior vertebral-column resection: case report and review of the literature. Spine (Phila Pa 1976). 2013;38(4):E259-62. PMID 23202355 · doi:10.1097/BRS.0b013e31827fc5cb
55. Shah SA, Saller J. Evaluation and Diagnosis of Back Pain in Children and Adolescents. J Am Acad Orthop Surg. 2016;24(1):37-45. PMID 26589458 · doi:10.5435/JAAOS-D-14-00130
56. Achar S, Yamanaka J. Back Pain in Children and Adolescents. Am Fam Physician. 2020;102(1):19-28. PMID 32603067
57. Tsirikos AI. Scheuermann's Kyphosis: an update. J Surg Orthop Adv. 2009;18(3):122-8. PMID 19843436
58. Orava S, Virtanen K. Osteochondroses in athletes. Br J Sports Med. 1982;16(3):161-8. PMID 7139227 · doi:10.1136/bjsm.16.3.161

Every reference in this article was verified one by one on 15 August 2026 against two independent databases: PubMed's E-utilities interface for the identifier, the journal, the year, the pagination, the full author list and the content of the abstract, then the CrossRef API for the DOI and a second opinion on the title, the journal and the year. Discrepancies found were arbitrated source by source, and the pagination of a 1999 Lippincott record was taken in the version on which PubMed and Europe PMC agree, against that of CrossRef. The figures cited reproduce the values published by the authors, with their confidence intervals, their sample sizes and the caveats they themselves state about risk of bias and level of evidence. The bibliometric counts in chapter 7 are reproducible queries, whose exact wording is given in the text.