Abstract
Background
We sought to investigate whether growth hormone (GH) treatment increases the risk of scoliosis in children with short stature and analyse a dose-response relationship.
Methods
A retrospective cross-sectional and cohort study design was used. The cross-sectional study included 1160 children with short stature (355 and 805 in the GH-exposed and non-exposed groups, respectively). The scoliosis prevalence was compared. A total of 313 short stature children without scoliosis at baseline were enrolled in the cohort study (158 and 155, respectively, in the GH and non-GH groups). The median follow-up time was approximately 4.5 years. The relationship between risk factors and GH dose and the occurrence of scoliosis was evaluated with the outcome of new-onset scoliosis.
Results
The incidence of scoliosis in the GH exposure group was significantly higher than that in the non-exposure group (5.40% vs. 2.40%, p=0.008). The risk of scoliosis in the GH group was 2.726 times higher than that in the non-GH group (HR=2.726, 95%CI: 1.326-5.605, p=0.006). The total GH dose was positively correlated with scoliosis risk, and there was a dose-response relationship between the total GH dose and the risk (HR=1.002, 95%CI: 1.001–1.003, p<0.001). A dose-response relationship was confirmed, with the predicted probability of scoliosis increasing continuously with cumulative GH dose.
Conclusions
GH therapy is an independent risk factor for scoliosis in children with short stature, with a significant dose-response relationship. The cumulative GH dose can be used as a preliminary reference indicator for risk stratification and personalized monitoring.
Keywords: Cumulative dose, dose-response relationship, growth hormone, risk factor, scoliosis, short stature
Introduction
Injection of growth hormone (GH) has become a common therapy to promote growth, increase bone and muscle mass, and improve quality of life in children with short stature [1,2]. However, adolescent idiopathic scoliosis is considered to be closely related to rapid bone growth, which can be achieved through GH injection therapy [3,4]. Therefore, whether GH treatment increases the risk of scoliosis in children with short stature has also attracted much attention.
Short stature is defined as a height less than 2 standard deviations of the average height of the general population for the same age, sex, and growth stage [5]. Scoliosis is a three-dimensional deformity defined as a coronal Cobb Angle ≥10 degrees [6]. Park et al. found that the incidence of new-onset scoliosis in children with idiopathic short stature treated with GH was 3.7%, similar to the incidence of idiopathic scoliosis in the general adolescent population (1% to 4%) [7]. However, Hong et al. found that the incidence of new-onset scoliosis in children with idiopathic short stature after GH treatment was 32.97%, much higher than the incidence of scoliosis in the general adolescent population [8]. These disparities may stem from differences in total GH dose, daily dose, and follow-up time in previous studies, factors that have not been discussed in depth previously.
Therefore, the primary objective of this study was to investigate whether GH treatment is an independent risk factor for scoliosis in children with short stature. The secondary objective was to explore, for the first time, the existence of a dose-response relationship between the cumulative GH dose and the risk of developing scoliosis.
Materials and methods
Study design and population
This study was a retrospective cross-sectional and cohort study. A total of 1160 children with short stature were included in this cross-sectional study. These children were consecutively admitted to our hospital from January 2015 to December 2020 and underwent conventional whole spine and left wrist joint anteroposterior and lateral radiographs. The whole spine X-ray and the use of GH in the same period were recorded. These patients were divided into GH exposure group (n = 355) and non-GH exposure group (n = 805). A total of 313 short stature children without scoliosis at baseline were enrolled in the cohort study, including 158 in the GH group and 155 in the non-GH group. These children were regularly followed up in the outpatient clinic and received ≥2 anteroposterior and lateral radiographs of the whole spine. With the occurrence of scoliosis (Cobb Angle ≥10°) as the outcome event, the survival curve was used to describe the scoliotic-free survival rate of the two groups of children with short stature in our cohort. Multivariate Cox regression analysis was used to explore the risk factors for scoliosis in children with short stature.
Inclusion and exclusion criteria
The inclusion criteria were children with short stature who were consecutively treated in our hospital from January 2015 to December 2020. Exclusion criteria were the absence of anteroposterior and lateral radiographs of the whole spine, a genetic disorder (e.g. Turner syndrome, Prader–Willi syndrome, Noonan syndrome), a disorder known to cause scoliosis (congenital spinal deformity), a neuromuscular disorder (epilepsy), an endocrine disorder (precocious puberty, hyperthyroidism or hypothyroidism), and a chronic dystrophic condition (Figure 1).
Figure 1.
Study flow diagram showing patient selection, exclusions, and group allocation for the cross-sectional (n = 1160) and cohort (n = 313) study populations.
Data collection
Data collected included: children’s age, bone age, sex, body mass index, whole spine X-ray, insulin-like growth factor 1 (IGF-1), insulin-like growth factor binding protein 3 (IGFBP-3), GH stimulation test (maximum GH stimulation level <5 ng/mL was defined as complete deficiency; 5–10 ng/mL was defined as partial deficiency; >10 ng/mL was defined as normal GH release), initial daily GH dose, and total GH dose.
Definitions and measurements
Bone age was assessed from radiographs of the left wrist joint of the child. Cobb Angle was defined as the angle formed by two perpendicular lines between the upper endplate of the upper involved vertebra and the lower endplate of the lower involved vertebra [6]. Body mass index (BMI) was calculated as weight in kilograms divided by height in meters squared.
Statistical analysis
Qualitative data are expressed as frequency. Quantitative data conforming to a normal distribution are expressed as mean ± standard deviation. Quantitative data not conforming to normal distribution are expressed as median (interquartile range). Chi-square test, independent-sample t-test, and non-parametric test were used for statistical inference. Binary logistic regression analysis was used to adjust for statistically different indicators to confirm the difference in the prevalence of scoliosis between children with short stature and those without GH treatment. Variables with a P value less than 0.05 in univariate analysis were included in the multivariate regression model. Cox regression analysis was used to determine scoliosis-related factors. Survival curves were used to describe the occurrence of scoliosis between different groups.
All statistical analyses except the post-hoc power analysis and the generation of the dose-response probability plot were performed using IBM SPSS version 27.0. The post-hoc power analysis was conducted using R statistical software (version 4.2.0; R Foundation for Statistical Computing), based on the observed incidence of new-onset scoliosis in the GH-treated group (17.1%) vs. the non-treated group (8.4%), with sample sizes of 158 and 155. The calculated power (1-β) for the primary outcome was 0.74. While acceptable, this suggests that the study may not be sufficiently powered for subgroup analyses, which is acknowledged as a limitation. The dose-response probability plot was also generated using R software, based on the multivariate Cox regression model. A P value of less than 0.05 was considered to indicate statistical significance.
Data collection and study period
This retrospective study utilized electronic medical records from the Children’s Hospital of Chongqing Medical University. The data collection period for clinical records spanned from January 2015 to December 2020, during which consecutive children with short stature who met the inclusion criteria were enrolled. The data extraction and analysis phase for this research was conducted after ethical approval, specifically from July 1, 2025, to August 31, 2025.
Ethical approval
This study was performed in line with the principles of the Declaration of Helsinki. Approval was granted by the Ethics Committee of Children’s Hospital of Chongqing Medical University [Ethical Approval No. 147, 2025]. The requirement to sign an informed consent form was waived by the Ethics Committee of Children’s Hospital of Chongqing Medical University.
Results
Cross-sectional study findings
The prevalence of scoliosis was significantly higher in GH-exposed children than in unexposed children (5.40% vs. 2.40%, p = 0.008). The location and type of scoliosis did not differ significantly between the two groups; thoracic and thoracolumbar curves predominated, and double curves were rare (Table 1). The severity of scoliosis was also similar between groups, with nearly all cases measuring 10–20° (94.7% vs. 100%). In univariate analysis, age, bone age, IGFBP-3, and GH treatment were significantly associated with scoliosis. Multivariate binary logistic regression identified decreased IGFBP-3 and GH treatment as independent factors (Table 2). The odds ratio for scoliosis in GH-treated children was 2.686 (95%CI: 1.4–5.3, p = 0.004).
Table 1.
Baseline clinical characteristics and scoliosis prevalence in the cross-sectional study cohort (n = 1160), comparing GH-exposed and GH-unexposed children with short stature.
| Variables | GH group | GH-free group | P-value |
|---|---|---|---|
| Total, n | 355 | 805 | |
| Sex, n (%) | 0.249α | ||
| male female |
208 (58.6%) 147 (41.4%) |
442 (54.9%) 363 (45.1%) |
|
| Age, M(IQR) (yrs) | 3.9 (3.2,4.9) | 4.0 (3.0,5.0) | 0.726β |
| BMI, M(IQR) (kg/m2) Bone age, M(IQR) (yrs) Scoliosis, n (%) Types of scoliosis, n (%) T(T2-T10) TL(T11-T12) L(L1-L5) S Cobb’s degree |
15.11 (14.32,15.98) 3.3 (2.5,4.3) 19 (5.4%) 8 (42.1%) 7 (36.8%) 3 (15.8%) 1 (5.3%) |
15.09 (14.15,16.04) 3.5 (2.6,4.9) 19 (2.4%) 7 (36.8%) 3 (15.8%) 5 (26.3%) 4 (21.1%) |
0.508β 0.011β 0.008α 0.265α 1.000α |
| 10°–20°, n (%) Sex, n (%) male female |
18 (94.7%) 9 (50.0%) 9 (50.0%) |
19 (100%) 7 (36.8%) 12 (63.2%) |
0.219α |
| ≥20°, n (%) Sex, n (%) male female |
1 (5.3%) / 1 (100) |
0 (0) / / |
/ |
n: number of patients. M(IQR)=Median (Interquartile Range). BMI: body mass index. αPearson Chi-square test. βNon-parametric test. T: Thoracic curvature, TL: Thoracolumbar curvature, L: Lumbar spine curvature, S: double curve.
Table 2.
Univariate and multivariate binary logistic regression analyses of factors associated with scoliosis in children with short stature (cross-sectional study).
| Variables | Scoliosis (n = 38) | No scoliosis (n = 1122) | Univariate analysis | Binary logistic regression analysis |
||
|---|---|---|---|---|---|---|
| P-value | P-value | Odds Ratio | 95%CI | |||
| Sex, n (%) | 0.096# | |||||
| Male | 16 (42.1%) | 634 (56.5%) | ||||
| Female | 22 (57.9%) | 488 (43.5%) | ||||
| Age, M ± SD (yrs) | 7.6 ± 3.0 | 4.6 ± 2.7 | <0.001∗ | |||
| BMI, M ± SD (kg/m2) | 15.36 ± 1.68 | 15.23 ± 2.02 | 0.688∗ | |||
| Bone age, M ± SD (yrs) | 4.8 ± 2.9 | 4.0 ± 2.4 | 0.041∗ | 0.408 | 1.059 | 0.9-1.2 |
| IGF-1, n (%) | 0.223# | |||||
| Normal | 25 (65.8%) | 869 (77.5%) | ||||
| Decrease | 13 (34.2%) | 249 (22.2%) | ||||
| Increase | 0 (0) | 4 (0.4%) | ||||
| IGFBP-3, n (%) | 0.004# | |||||
| Normal | 31 (81.6%) | 1024 (91.3%) | Reference | |||
| Decrease | 7 (18.4%) | 54 (4.8%) | 0.010 | 3.992 | 1.4-11.5 | |
| Increase | 0 (0) | 44 (3.9%) | 0.998 | 0 | 0 | |
| GH stimulation test, n (%) | 0.276# | |||||
| Normal | 23 (60.5%) | 631 (56.2%) | ||||
| Partial deficiency | 8 (21.1%) | 353 (31.5%) | ||||
| Complete deficiency | 7 (18.4%) | 138 (12.3%) | ||||
| Used GH, n (%) | 0.012# | |||||
| No | 19 (50.0%) | 786 (70.1%) | Reference | |||
| Yes | 19 (50.0%) | 336 (29.9%) | 0.004 | 2.686 | 1.4-5.3 | |
Pearson Chi-square test.
Independent sample T-test. M ± SD=Mean ± Standard Deviation.
Cohort study findings
Among children without scoliosis at baseline, the incidence of new-onset scoliosis was significantly higher in the GH-treated group than in the untreated group (17.1% vs. 8.4%, p = 0.027), after a median follow-up of approximately 4.5 years. However, there was no increased need for bracing or surgery (Table 3). Multivariate Cox regression analysis showed that bone age and GH treatment were independently associated with new-onset scoliosis (Table 4). The risk of scoliosis in GH-treated children was 2.726 times that of untreated children (HR = 2.726, 95%CI: 1.326–5.605, p = 0.006) (Figure 2).
Table 3.
Baseline characteristics and new-onset scoliosis outcomes in children without scoliosis at baseline (cohort study), comparing GH-treated and untreated groups.
| Variables | GH group (n = 158) | GH-free group (n = 155) | P-value |
|---|---|---|---|
| Age, M (IQR) (yrs) | 3.8 (3.1,4.9) | 3.3 (3.0,4.2) | 0.011# |
| Sex, n (%) | 0.428※ | ||
| Male | 78 (49.4%) | 69 (44.5%) | |
| Female | 80 (50.6%) | 86 (55.5%) | |
| BMI, M (IQR) (kg/m2) Bone age, M (IQR) (yrs) Follow-up time, M (IQR) (yrs) De novo scoliosis, n (%) Sex, n(%) Male Female Need for braces, n (%) Need for surgery, n (%) Cobb Angle of novo scoliosis, M (IQR) (°) Age at novo scoliosis, M (IQR) (yrs) |
15.35 (14.34,15.87) 3.9 (2.8,4.5) 4.3 (2.8,5.8) 27 (17.1%) 10 (37.0%) 17 (63.0%) 1 (3.7%) 0 11.36 (10.42,12.65) 8.80 (6.30,10.30) |
15.04 (14.24,17.70) 3.3 (2.3,4.3) 4.6 (3.3,5.7) 13 (8.4%) 4 (30.8%) 9 (69.2%) 0 0 11.86 (10.36,12.59) 8.10 (7.55,9.25) |
0.350# 0.020# 0.281# 0.027※ 0.740※ / / 0.848# 0.949# |
M (IQR)=Median (Interquartile Range). ※Pearson Chi-square test. #Non-parametric test.
Table 4.
Univariate and multivariate cox regression analyses of variables associated with new-onset scoliosis in the cohort study (n = 313).
| Variables | novo scoliosis (n = 40) |
No scoliosis (n = 273) |
Univariate analysis | Multivariate COX regression analysis |
||
|---|---|---|---|---|---|---|
| P-value | P-value | HR | 95%CI | |||
| Sex, n (%) | 0.127# | |||||
| Male | 14(35.0%) | 133(48.7%) | ||||
| Female | 26(65.0%) | 140(51.3%) | ||||
| Age, M ± SD (yrs) | 4.1 ± 1.9 | 4.2 ± 2.1 | 0.963∗ | |||
| BMI, M ± SD (kg/m2) | 15.37 ± 1.62 | 15.15 ± 1.77 | 0.457∗ | |||
| Bone age, M ± SD (yrs) | 4.0 ± 1.9 | 3.7 ± 2.2 | 0.462∗ | 0.015 | 1.212 | 1.038-1.415 |
| IGF-1, n (%) | 0.356# | |||||
| Normal | 30(75.0%) | 213(78.0%) | ||||
| Decrease | 9(22.5%) | 59(21.6%) | ||||
| Increase | 1(2.5%) | 1(0.4%) | ||||
| IGFBP-3, n (%) | 0.767# | |||||
| Normal | 36(90.0%) | 254(93.0%) | ||||
| Decrease | 2(5.0%) | 7(2.6%) | ||||
| Increase | 2(5.0%) | 12(4.4%) | ||||
| GH stimulation test, n (%) | 0.685# | |||||
| Normal | 22(55.0%) | 171(62.6%) | ||||
| Partial deficiency | 14(35.0%) | 78(28.6%) | ||||
| Complete deficiency | 4(10.0%) | 24(8.8%) | ||||
| GH treatment, n (%) | 0.027# | |||||
| No | 13(32.5%) | 142(52.0%) | Reference | |||
| Yes | 27(67.5%) | 131(48.0%) | 0.006 | 2.726 | 1.326-5.605 | |
M ± SD= Mean ± Standard Deviation. HR = Hazard Ratio. ∗Independent sample T-test. #Pearson Chi-square test.
Figure 2.
Kaplan–Meier curves for scoliosis-free survival in children without scoliosis at baseline, comparing GH-treated (n = 158) and untreated (n = 155) groups. Log-rank test: χ2=5.398, p = 0.020.
Dose–response relationship
Initial daily GH dose was not significantly correlated with scoliosis. However, multivariate Cox regression demonstrated a significant positive correlation between total cumulative GH dose and scoliosis risk (HR = 1.002 per mg, 95%CI: 1.001–1.003, p < 0.001) (Table 5). To visualize this relationship, we plotted the predicted probability of developing scoliosis as a function of total GH dose, based on the multivariate Cox regression model. As shown in Figure 3, the risk of scoliosis rises progressively with increasing cumulative GH dose, illustrating a continuous dose-response gradient rather than a simple threshold effect.
Table 5.
Association of GH dose parameters with new-onset scoliosis in GH-treated children: univariate and multivariate cox regression analyses.
| Variables | Novo scoliosis (n = 27) |
No scoliosis (n = 105) |
Univariate analysis | Multivariate COX regression analysis |
||
|---|---|---|---|---|---|---|
| P-value | P-value | HR | 95%CI | |||
| Sex, n (%) | 0.133# | |||||
| Male | 10 (30.0%) | 57 (57.3%) | ||||
| Female | 17 (63.0%) | 48 (45.7%) | ||||
| Age, M ± SD (yrs) | 4.1 ± 1.3 | 3.6 ± 1.1 | 0.030∗ | |||
| BMI, M ± SD (kg/m2) | 15.11 ± 1.58 | 14.98 ± 1.94 | 0.747∗ | |||
| Bone age, M ± SD (yrs) | 4.0 ± 1.3 | 3.1 ± 1.3 | 0.002∗ | 0.005 | 1.552 | 1.145-2.103 |
| IGF-1, n (%) | 0.312# | |||||
| Normal | 23 (85.2%) | 78 (74.3%) | ||||
| Decrease | 4 (14.8%) | 27 (25.7%) | ||||
| Increase | 0 | 0 | ||||
| IGFBP-3, n (%) | 0.106# | |||||
| Normal | 25 (92.6%) | 104 (99.0%) | ||||
| Decrease | 1 (3.7%) | 0 | ||||
| Increase | 1 (3.7%) | 1 (1.0%) | ||||
| GH stimulation test, n (%) | 0.322# | |||||
| Normal | 13 (48.1%) | 62 (59.0%) | ||||
| Partial deficiency | 12 (44.4%) | 31 (29.5%) | ||||
| Complete deficiency | 2 (7.4%) | 12 (11.4%) | ||||
| Initial GH does, M ± SD (mg/kg·d) | 0.049 ± 0.005 | 0.049 ± 0.006 | 0.902∗ | |||
| Total dose of GH, M (IQR) (mg) | 590.0 (320.0,800.0) | 116.0 (40.0,428.5) | <0.001※ | <0.001 | 1.002 | 1.001-1.003 |
M ± SD= Mean ± Standard Deviation. M(IQR)=Median (Interquartile Range). ∗Independent sample T-test. #Pearson Chi-square test. ※Non-parametric test.
Figure 3.
Dose-response relationship between cumulative GH dose and predicted probability of new-onset scoliosis. The curve was derived from the multivariate Cox regression model (adjusting for bone age at the mean value). Shaded area = 95% confidence interval.
Discussion
Herein, we demonstrate that the incidence of scoliosis in children with short stature who had used GH was significantly higher than that in children with short stature who had not used GH, and the history of GH use was an independent factor for scoliosis in children with short stature. In the cohort study, the risk of scoliosis in children with short stature without baseline scoliosis after GH treatment was 2.726 times higher than that in children without GH treatment. GH treatment was an independent risk factor for scoliosis in children with short stature. There was a significant dose–response relationship between the total GH dose and the occurrence of scoliosis in children with short stature.
Published studies on the role of GH therapy in the development or progression of scoliosis have reached different conclusions [7–10]. Similar to the results of the present study, Hong et al. found that GH treatment significantly increased the risk of scoliosis and the need for bracing treatment in children with idiopathic short stature who did not have scoliosis at baseline [8]. However, children with short stature treated with GH in this study did not show a significant increase in the need for bracing treatment, which is closely related to the insufficient sample size in this study and may mask a true association. Using propensity score matching, Yun et al. found that the spinal Cobb Angle increased 1.0 degree per year in children with idiopathic short stature in the GH treatment group [10], potentially suggesting that GH treatment promoted the development of scoliosis in children with short stature.
However, Park et al. analysed 1128 children with idiopathic short stature found that the development or progression of scoliosis in children with idiopathic short stature was not affected by GH treatment and duration [7]. Park et al. analysed clinical data from 113 children with short stature before and after GH treatment for 12 months and found that growth hormone treatment was not associated with the development or exacerbation of scoliosis [9]. Allen et al. looked at the spinal curves of 34,550 children with idiopathic short stature treated with various doses of GH, and the prevalence of scoliosis was 0.2% [11]. Similar studies did not find an increased risk of scoliosis with GH treatment in the study population. Possible reasons include unmentioned variations in the duration and dose of GH and length of follow-up. Previous diagnostic and screening methods for scoliosis were also different. For example, Quigley et al. used a forward flexion test to examine scoliosis in children with short stature after GH treatment [12], which may have underestimated the incidence of scoliosis. Additionally, Çetiner et al. recently examined GH therapy outcomes in neurosecretory dysfunction and found variations in treatment response [13], further highlighting the complexity of GH effects across different short stature etiologies.
We found a significant dose–response relationship between the total GH dose and the development of scoliosis in children with short stature. Yun et al. found that the spinal Cobb Angle increased by 1° per year in children with short stature treated with GH [10], confirming a GH dose-dependent increase. It has been suggested that abnormalities in other hormones, such as melatonin, oestrogen, leptin, and GH-releasing peptide, may also be associated with scoliosis [14]. However, these factors were not included in this study. Future studies should focus on exploring the dose–response relationship between the dose of GH and the development of scoliosis in children with short stature and build prediction models to provide a scientific basis for clinicians to accurately identify high-risk groups and formulate follow-up plans.
This study has several limitations. First, its retrospective, single-center design may introduce selection bias. Second, the sample size, while adequate for the primary analysis, was relatively small, limiting the statistical power for more granular subgroup analyses and the inclusion of all potential confounders. Third, a key diagnostic limitation is the sole reliance on the Cobb angle measured from standing X-rays without routine use of the forward bending test or Adam’s test for all patients at every visit, which may have led to an underestimation of the true prevalence of very mild curves. Fourth, we were unable to adjust for all potential confounders, such as genetic predisposition (e.g. melatonin receptor genes) or other hormonal factors (e.g. leptin, estrogen), which have been associated with scoliosis. Furthermore, gait characteristics, which have been associated with secondary scoliosis in other pediatric conditions [15], were not assessed in our study.
Conclusion
In conclusion, this study strengthens the evidence that GH therapy increases the risk of scoliosis in children with short stature and establishes a significant dose-response relationship. The risk increases progressively with cumulative GH dose, supporting a shift from general risk awareness to personalized monitoring based on cumulative exposure. Future prospective studies are needed to validate these findings and integrate them into clinical risk prediction models.
Acknowledgments
Pan Ren contributed to conceptualization, study design, data collection, drafting of the original manuscript, and critical revision of the manuscript; Yuxi Su contributed to study design, data collection, assistance in drafting the manuscript, and approval of the final version; Wenquan Cai contributed to statistical analysis, data interpretation, response to reviewer comments, major revision of the manuscript, and overall supervision. All authors read and approved the final manuscript.
Funding Statement
The author(s) reported there is no funding associated with the work featured in this article.
Disclosure statement
No potential conflict of interest was reported by the author(s).
Data availability statement
The data that support the findings of this study are available from the corresponding author upon reasonable request.
References
- 1.Danowitz M, Grimberg A.. Clinical indications for growth hormone therapy. Adv Pediatr. 2022;69(1):203–217. doi: 10.1016/j.yapd.2022.03.005. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2.Maghnie M, Ranke MB, Geffner ME, et al. Safety and efficacy of pediatric growth hormone therapy: results from the Full KIGS Cohort. J Clin Endocrinol Metab. 2022;107(12):3287–3301. doi: 10.1210/clinem/dgac517. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3.Latalski M, Danielewicz-Bromberek A, Fatyga M, et al. Current insights into the aetiology of adolescent idiopathic scoliosis. Arch Orthop Trauma Surg. 2017;137(10):1327–1333. doi: 10.1007/s00402-017-2756-1. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4.Shi B, Mao S, Liu Z, et al. Spinal growth velocity versus height velocity in predicting curve progression in peri-pubertal girls with idiopathic scoliosis. BMC Musculoskelet Disord. 2016;17(1):368. doi: 10.1186/s12891-016-1221-6. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.Backeljauw P, Cappa M, Kiess W, et al. Impact of short stature on quality of life: a systematic literature review. Growth Horm IGF Res. 2021;57-58:101392. doi: 10.1016/j.ghir.2021.101392. [DOI] [PubMed] [Google Scholar]
- 6.Horng M-H, Kuok C-P, Fu M-J, et al. Cobb angle measurement of spine from X-ray images using convolutional neural network. Comput Math Methods Med. 2019;2019:6357171–6357118.). doi: 10.1155/2019/6357171. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7.Park S-J, Lee K-H, Lee C-S, et al. Impact of growth hormone treatment on scoliosis development and progression: analysis of 1128 patients with idiopathic short stature. J Pediatr Endocrinol Metab. 2021;34(2):243–250. doi: 10.1515/jpem-2020-0393. [DOI] [PubMed] [Google Scholar]
- 8.Hong H, Pan X, Song J, et al. Idiopathic short stature and scoliosis in children treated with growth hormone. Bone Jt J. 2023;105-B(4):439–448. doi: 10.1302/0301-620X.105B4.BJJ-2022-0898.R1. [DOI] [PubMed] [Google Scholar]
- 9.Park M, Kim YJ, Oh KE, et al. The association between idiopathic scoliosis and growth hormone treatment in short children. Ann Pediatr Endocrinol Metab. 2022;27(3):207–213. doi: 10.6065/apem.2142186.093. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10.Yun Y-H, Kwon S-S, Koh Y, et al. Influence of growth hormone treatment on radiographic indices of the spine: propensity-matched analysis. J Orthop Surg Res. 2017;12(1):130. doi: 10.1186/s13018-017-0630-z. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11.Allen DB. Safety of growth hormone treatment of children with idiopathic short stature: the US experience. Horm Res Paediatr. 2011;76 Suppl 3:45–47. doi: 10.1159/000330159. [DOI] [PubMed] [Google Scholar]
- 12.Bell J, Parker KL, Swinford RD, et al. Long-term safety of recombinant human growth hormone in children. J Clin Endocrinol Metab. 2010;95(1):167–177. doi: 10.1210/jc.2009-0178. [DOI] [PubMed] [Google Scholar]
- 13.Çetiner EB, Donbaloğlu Z, Singin B, et al. Growth hormone therapy in neurosecretory dysfunction. Turk Arch Pediatr. 2025;60(3):294–300. doi: 10.5152/TurkArchPediatr.2025.24319. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14.Liang Z-T, Guo C-F, Li J, et al. The role of endocrine hormones in the pathogenesis of adolescent idiopathic scoliosis. Faseb J. 2021;35(9):e21839. doi: 10.1096/fj.202100759R. [DOI] [PubMed] [Google Scholar]
- 15.Leblebici G, Kısa EP, Tarakcı E, et al. Gait characteristics in children with juvenile idiopathic arthritis and secondary scoliosis. Children (Basel). 2026;13(1):83. doi: 10.3390/children13010083. [DOI] [PMC free article] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Data Availability Statement
The data that support the findings of this study are available from the corresponding author upon reasonable request.