ESPE Abstracts

Background: Adult height prediction (AHP) based on bone age appears as an incomplete procedure – it does not reveal the path from the present to the end-point. Growth charts offer little help in this respect because they average over children with different age of growth spurt (AGS).

Objective and hypotheses: To extend AHP by also forecasting AGS and the entire height and velocity trajectories until adulthood and displaying this in a growth chart made from the same data as the AHP model.

Method: The First Zurich Longitudinal study of 231 normal children born in 1955 was previously used to derive the AHP model (JCEM 2009). We used the same data to derive a model to predict AGS, and we derive height velocity curves corresponding to different AGS values. We also estimate a standard growth chart, which like the AHP model can be scaled to any mean population height. For a new child, the method predicts AGS, which is used to select the most likely height velocity trajectory. This is normalised and integrated to form the most likely height trajectory that ends up in the most likely value predicted by the previously developed AHP model.

Results: The method is implemented as a freely available, interactive tool available on www.BoneXpert.com/adult-height-predictor. One enters gender, age, bone age and current height (and optionally parental height or height at menarche), and the tool displays a growth chart with the most likely height trajectory, and the most likely velocity curve. The AHP and AGS are also shown with their 1 SD uncertainties.

Conclusion: The tool can provide a useful illustration in clinical practice and in the dialogue with the patients. It conveys the important message that the shape of an individual’s growth curve is quite different from the shape of the growth chart. At the same time the illustration is anchored in the well-validated AHP model.

Conflict of interest: HHT is the owner of Visiana which develops and markets the BoneXpert product for automated determination of bone age.

Funding: D Martin is supported by Tübingen University.