Demographic characteristics of positive students stratified by curve magnitude

As shown in Table 1, the PPV was 55.4% for the maximal Cobb angle of 10–19°, 37.8% for 20–39°, and 2.6% for ≥ 40° in the 426 students who were screened as positive. Both our [20] and other studies [2] have previously depicted a higher incidence of AIS in women;, in the present analysis we further found that the proportion of women with AIS with different curve magnitudes was higher than that of males (Cobb angle 10–19°: 43.2 vs. 12.2%; 20–39°: 34.7 vs. 3.1%; ≥ 40°: 2.6 vs. 0.0%; Z = -3.929, P < 0.001), suggesting that women present with more severe scoliosis. Patients with AIS were mainly aged 7–12y (52.6%) and 13–15y (39.0%) with Cobb angles of 10–19° (31.9 and 21.8%) and 20–39° (19.5 and 16.0%), which are greater than those observed in adolescents aged 16–19y (χ2 = 6.300, P = 0.043). These results indicate that the sex and age of patients with AIS are significantly associated with the severity of the curve magnitude.

Table 1 Demographics and incorrect postures of participants stratified by the distribution of the Cobb angleIncorrect postures and ATR Associated with the curve magnitude

As shown in Table 1, incorrect postures, including shoulder height, scapula tilt, and rib hump, were significantly different among the different curve magnitude groups. The AIS patients with Cobb angles of 10–19°, 20–39°, and ≥ 40° had greater incidences of left shoulder height (23.5%, 14.1%, and 0.9%), right shoulder height (21.1%, 19.0%, and 1.4%), left scapula tilt (23.5%, 15.5%, and 0.7%), and right scapula tilt (20.7%, 17.1%, and 1.4%) than that of normal shoulder (10.8%, 4.7%, and 0.2%; χ2 = 9.126, P = 0.010) and scapula height (11.3%, 5.2%, and 0.5%; χ2 = 8.048, P = 0.018). Interestingly, patients with AIS with Cobb angles of 20–39° and ≥ 40° had a greater proportion of right rib hump (22.5% and 2.1%) than those with normal rib morphology (11.0% and 0.2%) and left rib hump (4.2% and 0.2%; χ2 = 12.721, P = 0.002). Correspondingly, the AIS group with Cobb angles 20–39° and ≥ 40° had significantly higher frequencies of students with an angle of right thoracic rotation ≥ 5° (20.0 and 2.1%) compared to that of normal thoracic rotation 0–4° (13.8 and 0.2%) and left thoracic rotation ≥ 5° (4.0 and 0.2%; χ2 = 13.882, P = 0.001). No significant differences were reported in the percentages of other incorrect postures (pelvic tilt, lumbar concave, flat back, thoracolumbar kyphosis, and lumbar kyphosis) or ATR (angle of thoracolumbar or lumbar rotation) in the different curve magnitude groups. Taken together, the incorrect postures, including shoulder-height difference, scapula tilt, a rib hump, and angle of thoracic rotation, were significantly associated with the curve magnitude in students with AIS.

Multifactorial analysis of correlative factors of curve magnitude

Univariate LR analysis models were applied to explore the factors associated with the curve magnitude. The results showed that female gender (OR = 2.92, 95%CI 1.67–5.09, P < 0.001), age 16–19y (OR = 2.83, 95%CI 1.10–7.28, P = 0.031), right shoulder height (OR = 2.15, 95%CI 1.23–3.75, P = 0.007), right scapula tilt (OR = 2.03, 95%CI 1.18–3.50, P = 0.010), right rib hump (OR = 1.88, 95%CI 1.23–2.85, P = 0.003), right thoracic rotation ≥ 5° (OR = 2.14, 95%CI 1.43–3.20, P < 0.001), and left thoracolumbar kyphosis (OR = 3.79, 95%CI 1.06–13.56, P = 0.041) were all significantly associated with the severity of the curve magnitude (Table 2).

Table 2 Univariate logistic regression analysis of incorrect postures associated with spinal curve magnitude

Furthermore, factors with a P-value < 0.05 in the univariate analysis were used for multivariate analysis. As shown in Table 3, women were more likely to develop a severe curve magnitude (AOR = 3.23, 95%CI 1.81–5.73, P < 0.001) than men; students aged 16–19y had 5.08 times (AOR = 5.08, 95%CI 1.86–13.91, P = 0.002) higher likelihood of a severe curve magnitude compared to those aged 7–12y. Interestingly, the presence of a right rib hump, but not left rib hump, was associated with a 1.72 times (AOR = 1.72, 95%CI 1.11–2.64, P = 0.015) higher likelihood of having severe curve magnitude than normal students. These results indicate that sex, age, and incorrect postures, especially a right rib hump, were significantly associated with a higher risk of developing severe AIS.

Table 3 Multivariate logistic regression analysis of incorrect postures associated with spinal curve magnitudeCompare the discrimination ability of influential factors for curve magnitude using ROC analysis

We employed ROC curves and AUC scores to compare the predictive effects of different influential factors for curve magnitude (Fig. 1). Similar to the results of LR models, sex, age, shoulder-height difference, scapula tilt, flat back, rib hump, angle of thoracic rotation could significantly distinguish different grades of major curve magnitude. The corresponding AUC scores were listed in Table S1.

Fig. 1

Results of ROC curve analysis by different risk factors for curve magnitude. ROC curve results for lumbar concave, pelvic tilt, thoracolumbar kyphosis, angle of thoracolumbar rotation, lumbar kyphosis, and angle of lumbar rotation were not shown because there was no statistical difference in their AUC value. G1: Cobb angle < 10°, G2: 10-19°, G3: 20-39°, G4: ≥ 40°