Clinically Optimized Adult Height Prediction from Key Bone and Pubertal Stages : Prospective Validation to Adult Height

Context Accurate adult height prediction remains a challenge in pediatric endocrinology. Traditional bone age (BA) based methods are time-consuming, software-dependent, and unreliable, while ignoring the critical impact of pubertal progression on growth potential. Methods A cross-sectional study was conducted in Beijing (2022–2023). We performed Tanner-Whitehouse 3/radius-ulna-short bone (TW3-RUS) bone grading and pubertal staging including Pre-puberty, On-puberty, and Completing-puberty. Spearman analysis identified key bone combinations most associated with BA and height. An integrated model combining bone grades and pubertal stage was developed and validated in an independent cohort followed to adult height. Results Key Spearman correlation revealed strong correlations of the Radius, Ulna and Metacarpal I grading with BA (ρ = 0.94-0.96), with bone combinations (ρ = 0.98-0.99) outperforming any single bones. Three types of bone combinations (Radius + Ulna, Radius + Metacarpal I, and Radius + Ulna + Metacarpal I) integrating with pubertal stages demonstrated approximately equivalent predictive performance for adult height prediction. Considering bone representativeness and feasibility, we prefer to propose the Radius + Metacarpal I combination with puberty stages as the clinically optimized model for adult height prediction. Independent validation cohort confirmed superior accuracy of the proposed model versus traditional BA based methods: mean prediction error was reduced from 0.71 cm to 0.02 cm, while the proportion of predictions error ≤3 cm increased from 66.9% to 73.5%. Conclusion The integrated bone-puberty model significantly improves prediction accuracy by incorporating skeletal maturity and pubertal dynamics. Its streamlined two-bone protocol offers a practical tool for growth monitoring and clinical decision-making.