Growth rates and the prevalence and progression of scoliosis in short-statured children on Australian growth hormone treatment programmes
© Day et al; licensee BioMed Central Ltd. 2007
Received: 03 August 2006
Accepted: 22 February 2007
Published: 1 December 2007
Study design and aim
This was a longitudinal chart review of a diverse group (cohort) of patients undergoing HGH (Human Growth Hormone) treatment. Clinical and radiological examinations were performed with the aim to identify the presence and progression of scoliosis.
Methods and cohort
185 patients were recruited and a database incorporating the age at commencement, dose and frequency of growth hormone treatment and growth charts was compiled from their Medical Records. The presence of any known syndrome and the clinical presence of scoliosis were included for analysis. Subsequently, skeletally immature patients identified with scoliosis were followed up over a period of a minimum four years and the radiologic type, progression and severity (Cobb angle) of scoliosis were recorded.
Four (3.6%) of the 109 with idiopathic short stature or hormone deficiency had idiopathic scoliosis (within normal limits for a control population) and scoliosis progression was not prospectively observed. 13 (28.8%) of 45 with Turner syndrome had scoliosis radiologically similar to idiopathic scoliosis. 11 (48%) of 23 with varying syndromes, had scoliosis. In the entire cohort, the growth rates of those with and without scoliosis were not statistically different and HGH treatment was not ceased because of progression of scoliosis.
In this study, there was no evidence of HGH treatment being responsible for progression of scoliosis in a small number of non-syndromic patients (four). An incidental finding was that scoliosis, similar to the idiopathic type, appears to be more prevalent in Turner syndrome than previously believed.
Research indicating that juveniles who develop idiopathic scoliosis have a markedly elevated growth rate compared to controls fostered a popular belief that growth hormone influences the progression of scoliosis[1, 2]. Serum levels of growth hormone in children with idiopathic scoliosis were found to be elevated between the ages of seven and twelve years compared to controls [3–7]. Progression of structural scoliosis was reported in a short-statured patient during Human Growth Hormone (HGH) treatment  and stimulated a systematic clinical audit of children with scoliosis on an HGH treatment programme in one institution . The administration of HGH appeared to accelerate the progression of scoliosis in six of ten children on that programme (cohort size 250) . Of the ten, five had idiopathic scoliosis, one had acute lymphocytic leukaemia and two with Turner syndrome were diagnosed with scoliosis prior to treatment. An accelerated growth rate was also recorded in eight of ten with scoliosis.
Armed with this information, it was hypothesized that children with scoliosis, who are administered Human Growth Hormone, are more likely to have progression of their scoliosis and an accelerated growth rate compared to those without scoliosis. The aims of this study are to determine the prevalence of scoliosis in four Australian Institutions' Growth Hormone Treatment Programmes and to confirm the results of previous research that HGH treatment can be associated with progression of pre-existing scoliosis.
Methods and subjects
Syndromes/conditions known to be associated with scoliosis
Number of patients
Number with scoliosis
14 (1 congenital)
Russell Silver dwarf
Klippel-Feil with Sprengel shoulder
Fanconi Syndrome and Rickets
Pituitary Deep Xray therapy
Deep Xray therapy to the brain
Conditions other than idiopathic short stature, not usually associated with scoliosis
Number of patients
Exclusion criteria for clinical examination included five children less than six years old and nineteen who commenced the OZGROW (The Australasian Paediatric Growth Hormone database) programme over the age of twelve years, because of their proximity to menarche. Two who signed Consent Forms and subsequently changed their minds were excluded. The Medical Records of four deceased were available for perusal, but were excluded from analysis.
Turner syndrome scoliosis (non-congenital)
12° Right thoracic, hypokyphosis
30° Right Thoraco-lumbar
18° Right thoracic, hypokyphosis
30° Right Thoraco-lumbar
No Xrays Mild clinical deformity Right thoracic
17° Left Thoraco-lumbar
18° Right thoracic, hypokyphosis
55° Right thoracic Scoliosis surgery
45° Right thoracic, hypokyphosis
10° Right lumbar
Diagnosed South Africa. Karyotype not available.
20° Left Thoraco-lumbar
12° left Thoraco-lumbar
45XO/46XY mosaic Undiagnosed syndrome
45° Right Thoracic, hypokyphosis
Growth chart information
Actual and percentile heights were recorded at three monthly intervals during the HGH treatment programme. Percentile, rather than actual, heights were included for analysis because the growth chart with actual heights for Turner syndrome has a different scale from the National Centre for Health Statistics scale.
Clinical confirmation of spine deformity
Patients presented dressed in a white Hospital gown for a forward bend test (Adams), having removed their shoes . One examiner (GD) used a scoliometer to measure for thoracic and lumbar spine rotation (Bunnell test) . Patients with a straight spine and no truncal rotation were not asked to present for spine radiographs.
Radiologic confirmation of the spine deformity
Spine radiographs of all patients with clinical signs of scoliosis have been performed with the exception of one with Turner syndrome and another who had undergone deep Xray therapy. Both were skeletally mature at the time of clinical examination. Standard postero-anterior and lateral complete spine radiographs in the erect posture were reviewed. Images at the time of clinical examination were in a hard-copy form and images after 2002 were in digital format. Cobb angles were drawn by the treating scoliosis surgeons and recorded.
Measurement of progression of scoliosis
Two patients detected with previously unrecognised spine lateral deviation/trunk rotation were referred to the Scoliosis Clinic in their hospital, and subsequently underwent spine imaging every six-months to skeletal maturity. A third patient with previously unrecognised spine deformity was skeletally mature and underwent a single spine radiograph. Twenty-four with known scoliosis had undergone similar regular follow-up in Scoliosis Clinics, prior to clinical examination in this study (two continued follow-up after the 2002 clinical examination). Follow-up spine imaging ceased when patients became skeletally mature or Human Growth Hormone treatment ceased. An increase in Cobb angle of 10° was the minimum measurement used to define progression of scoliosis.
Institutional Ethics approval was granted from four hospitals in three Australian States to clinically examine all their patients treated through the OZGROW program.
Statistical analysis for the cohort of 185. Predictors of Scoliosis: Logistic Regression – Univariate Models Results
Mg per Kg1
Mg per Kg2
Mg per Kg3
Mg per Kg4
At review in 2001/2002, the mean age of the patients was 13.6 years (range 6–27 years). 91 were male and 94 were female. The mean age at commencement on the OZGROW programme was 8.5 years (range from birth to 12 years, and followed a normal distribution). The mean duration of HGH therapy up to the time of review was 4.7 years (range 2 months to 12 years). The findings from routine clinical and radiological follow-up for those in Scoliosis Clinics ceased in 2006, spanning 4–5 years duration (mean 4.6 years).
Presence of scoliosis
29 of 185 patients had scoliosis. The duration of HGH therapy and total dose of human growth hormone were not related to the presence of scoliosis, although the average of the first two doses of growth hormone per patient, expressed in milligrams per kilogram body weight, was related to the presence of scoliosis. (p = 0.021). Having Turner syndrome was identified as the only highly significant factor for the presence of scoliosis (p = 0.001).
Progression of scoliosis
The mean duration of HGH treatment for the entire cohort was 4.7 years (SD 2.8). A number of different HGH regimens were administered to the cohort including Saizen™, Norditropin™, Somaton™, Genotropin™ and Humatropin™. The mean dose of HGH treatment was 0.60 mg/Kg body weight (Range 0.25 – 1.14 mg/Kg body weight and SD 0.227). HGH therapy was ceased and restarted at a later stage in 13 children for reasons other than the presence of scoliosis. One male with Cushing's syndrome and mild thoracic scoliosis had a temporary three-month cessation of HGH treatment. HGH therapy was re-started and continued for another two years, with no radiologic progression of scoliosis. Progression of scoliosis was not recorded in two patients with idiopathic scoliosis and two with other conditions, followed up in Scoliosis Clinics between 2001/2 and 2006.
Progression of scoliosis had already occurred in five Turner and two 'other condition' patients at the time of clinical review in 2001/2002 (Cobb angles between 30° and 55°).
Idiopathic short stature and hormone deficiency subgroup
Four (3.4%) of the 117 in this subgroup had scoliosis. All four had idiopathic right thoracic scoliosis less than 20 degrees Cobb angle. Three were male and all were unaware of their scoliosis until they were examined in this study. Two of the males had idiopathic short stature and the other had multiple pituitary hormone deficiency. Two of the three were skeletally mature and the other was age 13 at the time of clinical examination and continued review in the Scoliosis Clinic of his hospital for three years. The only female in this group had growth hormone deficiency and commenced HGH supplementation at age three years and remained on the OZGROW programme at age 13 years (in 2001). Both who continued clinical review demonstrated no radiologic progression of their scoliosis, from 2001/2 to 2006.
Turner syndrome subgroup
Other syndromes known to be associated with the presence of scoliosis
Scoliosis was present in 11 of the other 23 (48%) with varying other syndromes (Refer Table 1). One of this group had kypho-scoliosis.
There was a significant statistical correlation between the first and last percentiles of height in the idiopathic short stature and hormone deficiency subgroup (r = 0.514, p < 0.001) – ie. an acceleration of growth rate compared to peers was not statistically evident on the growth charts, between the commencement of the human growth hormone treatment programme and its cessation. The growth velocity of the four with scoliosis (mean 8.3 mm/year, range 7.0 to 10.6) in the idiopathic short stature and hormone deficiency subgroup was no different from the other 113 in the subgroup (mean 9.0 mm/year, standard deviation 7.53).
Prevalence of scoliosis on the Human Growth Hormone treatment programme
In this study, the 97 patients with idiopathic short stature had a similar prevalence of idiopathic scoliosis to a control Caucasian population  and served as a 'control' group for comparisons with the other groups. The prevalence of scoliosis in this study was similar to previously published research, involving 31,462 patients on Human Growth Hormone treatment programmes [14–18]. Scoliosis is not a specific feature of idiopathic short stature and its prevalence should mirror that of the control population . Although scoliosis has been extensively reported in specific syndromes associated with short stature, the commonest spinal deformity in short-stature syndromes and skeletal dysplasias is kyphosis .
Prevalence of scoliosis in Turner syndrome on the HGH treatment programme
The 28.8% prevalence of scoliosis in Turner syndrome from this study was higher than previously published research, which includes those not treated with HGH (11–12%) [20–22]. In a previous HGH treatment study, a 'subgroup with Turner syndrome and those with a history of brain tumour or vertebral abnormalities had a higher prevalence of new and/or severe scoliosis than other groups'. In a larger HGH treatment study, 72 of 24,000 children were diagnosed with scoliosis, of whom 10 had Turner syndrome .
Progression of scoliosis on the Growth Hormone treatment programme
In this study, prospective follow-up from the four with scoliosis and idiopathic short stature/hormone deficiency indicated that there was no progression of the scoliosis during HGH treatment. The deformities in the current study were mainly single curve thoracic scoliosis and the results are similar to those of a previous study involving four single curves . In the previous study, progression of scoliosis was associated with six double curves during HGH treatment, however it was not stated whether there was a temporary/permanent cessation of HGH treatment at the time of scoliosis progression .
In this study, retrospective data from the Turner syndrome patients' medical records indicated that the scoliosis deformity progressed in five of thirteen during HGH therapy and that the HGH treatment was not stopped because of this scoliosis progression.
Growth velocity and progression of scoliosis
In this study, the growth velocity was similar in patients with and without idiopathic scoliosis. This result contrasts with that of a previous study, where accelerated growth was recorded in eight of ten with scoliosis compared to those without scoliosis .
This study supports the contention that the administration of Human Growth Hormone to a diverse group of short-statured children does not lead to an increased prevalence of scoliosis. An accelerated growth rate during HGH treatment was not recorded in the small numbers with idiopathic scoliosis, contrasting with the results of the only other similar previous study. In this study, patients on HGH treatment with progressive scoliosis had syndromes in which the presence of scoliosis is commoner than an age-matched population.
A curious incidental finding was the 28.8% prevalence of scoliosis in Turner syndrome, which could mean that scoliosis is commoner in Turner syndrome than previously believed or that treating Turner syndrome girls with human growth hormone leads to an increased prevalence of scoliosis.
The authors wish to thank Dr. R. Kattaranga and Mr P Cundy (South Australia), Mr A J Geddes and Dr B Brankoff (Western Australia), Dr M Thomsett, Dr A Cotterill and H Mercer (Queensland) and especially Mrs Maree Grant (Western Australia), without whom this study could not have been completed. Dr C Swanson (University of Queensland) provided Statistical Analysis.
This study was partly funded by a $7000.00 grant from the Australian Orthopaedic Association Research Foundation. (#008RF).
- Willner S: A Study of Growth in Girls with Adolescent Idiopathic Structural Scoliosis. Clin Orthop Rel Res. 1974, 101: 129-135.Google Scholar
- Nissinen M, Heliovaara M, Seitsamo J, Poussa M: Trunk asymmetry, posture, growth and risk of scoliosis. A three-year follow-up of Finnish prepubertal school children. Spine. 1993, 18: 8-13. 10.1097/00007632-199301000-00002.View ArticleGoogle Scholar
- Skogland LB, Miller JAA: Growth Related Hormones in Idiopathic Scoliosis. Acta Orthop Scand. 1980, 51: 779-789.View ArticleGoogle Scholar
- Willner S, Nilsson KO, Kastrup K, Bergstrand CG: Growth Hormone and Somatomedin A in Girls with Adolescent Idiopathic Scoliosis. Acta Paediatr Scand. 1976, 65: 547-552.View ArticleGoogle Scholar
- Ahl T, Albertsson-Wikland K, Kalen R: Twenty-four-hour Growth Hormone Profiles in Pubertal Girls with Idiopathic Scoliosis. Spine. 1988, 13 (2): 139-142. 10.1097/00007632-198802000-00001.View ArticleGoogle Scholar
- Misol S, Ponseti IV, Samaan N, Bradbury JT: Growth Hormone Blood Levels in Patients with Idiopathic Scoliosis. Clin Orthop and Related Research. 1971, 81: 122-125. 10.1097/00003086-197111000-00019.View ArticleGoogle Scholar
- Yamada K, Yamamoto H, Nakagawa Y, Tezuka A, Tamura T, Kawata S: Etiology of Idiopathic Scoliosis. Clin Orth Related Research. 1984, 184: 50-57.Google Scholar
- Dymling J, Willner S: Progression of a Structural Scoliosis During Treatment with Growth Hormone. Acta Orthop Scand. 1978, 49 (3): 264-268.View ArticleGoogle Scholar
- Wang ED, Drummond DS, Dormans JP, Moshang T, Davidson RS, Gruccio D: Scoliosis in Patients Treated with Growth Hormone. J Ped Orthop. 1997, 17: 708-711. 10.1097/00004694-199711000-00003.Google Scholar
- Lyon A, Preece M, Grant D: Growth charts for girls with Turner syndrome. Arch Dis in Childhood. 1985, 60: 932-935.View ArticleGoogle Scholar
- Adams W: Lectures on curvatures of the spine. Lecture IV. Lateral curvature of the spine, External characters and morbid anatomy. 1865, London, Churchill, 89-121.Google Scholar
- Bunnell W: An objective criterion for scoliosis screening. J Bone Joint Surg. 1984, 66A: 1381-1387.View ArticleGoogle Scholar
- Rogala E, Drummond D, Gurr J: Scoliosis: Incidence and natural history. J Bone Joint Surg. 1978, 60A: 173-176.View ArticleGoogle Scholar
- Cowell CT, Dietsch S: Adverse Events During Growth Hormone Therapy. J Pediat Endocrinology Metabolism. 1995, 8: 243-252.View ArticleGoogle Scholar
- Blethen SL: Monitoring Growth Hormone Treatment: Safety Considerations. The Endocrinologist CME Review Article #27. 1996, 369-374.Google Scholar
- Allen DB: Safety of Human Growth Hormone Therapy: Current Topics. J Pediatrics. 1996, 128 (5): S8-S13. 10.1016/S0022-3476(96)70003-3.View ArticleGoogle Scholar
- Wilton P: K I G S: Adverse Events Report No 11. Pharmacia International growth Database. To August 2000.Google Scholar
- Quigley C, Gill A, Crowe B, Robling K, Chipman J, Rose S, Ross J, Cassoria F, Wolka A, Wit J, Rekers-Mambarg L, Cutler G: Safety of growth hormone treatment in pediatric patients with idiopathic stature. J Clin Endocrinol Metab. 2005, 90: 5188-5196. 10.1210/jc.2004-2543.View ArticleGoogle Scholar
- Tolo V: Spinal deformity in short-stature syndromes. Instr Course Lect. 1990, 39: 399-405.PubMedGoogle Scholar
- Lippe B: Turner Syndrome. Endocrinology and Metabolism Clinics of North America. 1991, 20: 121-152.PubMedPubMed CentralGoogle Scholar
- Kim J, Rosenfeld S, Keyak J: Increased prevalence of scoliosis in Turner Syndrome. J Ped Orthop. 2001, 765-766. 10.1097/00004694-200111000-00012.Google Scholar
- Smith DW, Jones KL: Recognizable Patterns of Human Malformation. 1988, W B Saunders Company, 72-ThirdGoogle Scholar
This article is published under license to BioMed Central Ltd. This is an Open Access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/2.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.