Background: For children with retarded bone ages such as in constitutional delay of growth and puberty (CDGP) there are no specific methods to predict adult height based on bone age. Widely used methods such as Bayley-Pinneau (BP) tend to overestimate adult height in CDGP. Therefore, we aimed to develop a specific adult height prediction model for boys in pubertal age with retarded bone age >1 year.
Methods: Based on the adult heights of 68 males (median age 22.5 years) a new height prediction model was calculated based on 105 height measurements and bone age determinations at a median age of 14.0 years. The new model was adapted for the degree of bone age retardation and validated in an independent cohort of 32 boys with CDGP. The new model for predicting adult height was based on the following procedure: Achieved adult height was divided by the measured height at the recent bone age determination. For every bone age, the median of these calculated factors at the respective bone age was chosen. Afterwards, regression models (linear, exponential, potential, logarithmic and hyperbolic) were calculated based on the median factor to predict adult height of each bone age. The model with the highest r² was chosen.
Results: The BP method led to an overestimation of adult height (median +1.2cm; P=0.282), which was more pronounced in boys with a bone age retardation >2 years (median +1.6cm; P=0.027). In the validation study, there was no significant difference between reached and predicted adult height based on the new model (P=0.196), while the BP model led to a significant overestimation of predicted adult height (median +4.1cm; P=0.009).
|Bone age [years]||BP||New model for bone age retardation >1 and <2 years||New model for bone age retardation > 2 years|
Conclusions: The new model to predict adult height in boys with CDGP provides novel indices for height predictions in bone ages >13 years and is adapted to different degrees of bone age retardation. The new prediction model has a good predictive capability and overcomes some of the shortcomings of the BP model.
19 - 21 Sep 2019
European Society for Paediatric Endocrinology