Responsiveness and minimal clinically important changes of surface topography parameters in adolescents with idiopathic scoliosis: results from the schroth exercise trial.

Abstract

Adolescent idiopathic scoliosis (AIS) affects 2-3% of adolescents, causing spinal curvature and functional limitations. Traditional assessment using the Cobb angle may only partially capture patient concerns about aesthetics. The Schroth scoliosis-specific exercises (SSE), focusing on posture correction, have shown promise for reducing the Cobb angle and preventing progression, though research on its cosmetic impact is limited. Surface topography (ST) offers a radiation-free alternative to assess AIS, but determining what is a meaningful change for patients remains underexplored. The objective of this study is to determine the Minimal Clinically Important Changes (MCIC) in ST parameters in AIS after undergoing 6 months of Schroth SSE exercises. This is a secondary analysis from a randomized controlled trial. Participants (n = 124) were recruited from a Scoliosis Clinic and randomized into standard care (observation or bracing) and Schroth intervention added to standard care groups (one-hour weekly supervised session, 30-40 min daily home exercises). A global Rating of Change (GRC) was self-reported after six months, and asymmetry parameters of root mean square (RMS) and maximum deviation (MaxDev) over the asymmetry patch corresponding to the spinal curve were obtained through surface torso scans captured at baseline and six months. Pearson correlation and receiver-operating characteristic (ROC) curve analysis were performed to determine the MCICs. Subgroup analyses were also conducted to ascertain MCICs for thoracic and lumbar curve types. GRC ratings correlated with changes in RMS (r=-0.510, p < 0.001) and MaxDev (r=-0.409, p < 0.001). Participants who reported improved (GRC ≥ 2) posture saw a 1.76 ± 2.9 mm and 3.29 ± 6.5 mm decrease in RMS and MaxDev, respectively. In contrast, RMS and MaxDev increased by 1.03 ± 3.0 mm and 1.26 ± 5.6 mm, respectively, among individuals who stated that their posture had deteriorated or not changed (GRC < 2). Using ROC analysis, MCICs for RMS and MaxDev were determined to be -0.27 mm (area under the curve (AUC) 0.746, sensitivity 67%, specificity 74%) and - 0.49 mm (AUC 0.717, sensitivity 64%, specificity 68%), respectively, for overall improvement perception. Having met both thresholds reduced sensitivity to 62% and achieved 74% specificity. MCICs for thoracic curve types were - 0.58 mm (AUC 0.618, sensitivity 60%, specificity 53%) for RMS and - 1.32 mm (AUC 0.632, sensitivity 73%, specificity 92%) for MaxDev. MCICs for lumbar curve types were - 0.26 mm for RMS (AUC 0.881, sensitivity 73%, specificity 92%) and - 0.61 mm for MaxDev (AUC 0.811, sensitivity 68%, specificity 83%). Changes in RMS and MaxDev were aligned to GRC score reflecting perceived improvements in back condition. Stronger associations were observed between ST parameters and perceived improvement in lumbar than thoracic or combining all curves.