Prevalence and associated factors of screen-positive scoliosis among primary and secondary school students in Taizhou, China: a cross-sectional analysis based on ATR thresholds with identification of the apparent 5° measurement clustering

Background Adolescent idiopathic scoliosis (AIS) affects 0.47%–5.2% of school-aged children globally. The angle of trunk rotation (ATR) measurement using a scoliometer is the primary screening tool, though threshold controversy exists between ≥5° and ≥7°. This study aimed to determine the prevalence of screen-positive scoliosis and its associated factors among students in Taizhou, China, with particular focus on exploring potential measurement clustering patterns at the 5° ATR threshold. Methods A cross-sectional study was conducted among 813 students (ages 10–17 years) from four schools in Taizhou. Screening included general examination, Adam's forward bend test, ATR measurement with scoliometer, spinal movement test, and classification. The primary outcome was post-exercise ATR≥7°. Secondary outcomes included initial screening ATR≥5°. Univariate and multivariable logistic regression analyses were performed. Model performance was evaluated using AUC, Hosmer-Lemeshow test, and decision curve analysis. Results Among 813 participants, 215 (26.4%) were classified as screen-positive: Grade I (5°≤ATR<7°): 165 (76.7%), Grade II (7°≤ATR<10°): 40 (18.6%), Grade III (ATR≥10°): 10 (4.7%). Significant clustering was observed at ATR = 5° (144/813 = 17.7%, ratio 5.5:1 compared to ATR 1–4°; χ 2 = 794.88, P < 0.001 for deviation from interpolated expectation). The primary model showed female sex (OR = 3.509, 95%CI: 1.917–6.425, P < 0.001) and screen time (OR = 1.391, 95%CI: 1.070–1.809, P = 0.013) as significantly associated factors. BMI showed marginal significance (OR = 0.930, P = 0.052). A significant sex × screen time interaction was identified (OR = 1.68, P = 0.046). The main model demonstrated acceptable discrimination (AUC = 0.714) and calibration (HL P = 0.680). Conclusion This study identifies statistically significant measurement clustering at the 5° ATR threshold ( χ 2 = 794.88, P < 0.001), which should be considered when interpreting screening results. Female sex and screen time were independently associated with screen-positive status, with screen time showing a stronger effect in females. Using ATR≥7° rather than ≥5° yields a substantially lower screen-positive rate (6.1% vs. 26.8%) that is more consistent with prevalence ranges reported in the literature, and may reduce unnecessary referrals and parental anxiety. However, these findings require validation against radiographic confirmation in future studies.