Background: Central precocious puberty (CPP) is a common condition in girls and has been associated with deterioration of final height (FH). Height prognosis (HP) is critical for the decision of treatment in CPP. There are several methods for predicting FH in CPP but none is completely reliable. Most methods consider bone age (BA), which is very imprecise, but there is no consensus on which method is the best to estimate FH.
Objective: To compare the accuracy of three methods in the prediction of FH in girls with untreated CPP.
Patients and methods: 143 girls consulted for CPP between 1993 2000, 93 had reached their FH without being treated with LHRH analogues and they came to measure their current height. FH was considered with BA > 15 years. Parents heights were registered. BA was reported by the same four observers. We compared the initial HP, by three methods: Bayley-Pinneau (BP), Roche-Wainer-Thissen (RWT) and midparental height; with their FH. Statistical correlation between HP and FH was evaluated with Bland and Altman method.
Results: The age at diagnosis of the 93 girls was 8.3±1.2 years (6.410.5); their BA was 9.9±1.7 years and their current age was 19.1±3.3 years. The BP method vs. FH showed x: −1.01 cm difference, but with great interindividual variation (+8.87 cm: and −10.89 cm, ±2DS respectively). Similarly RWT method showed x: +0.96 cm difference (+9.65 and −7.72±2DS respectively). In contrast, midparental height showed less dispersion and variation: (x: +0.05 cm difference with +6.19 and −6.10 cm, ±2DS respectively).
Conclusions: All methods for predicting FH showed good results in average, but they have a considerable individual variation which may be explained by the subjectivity of the radiological interpretation of BA. Midparental height, which does not consider BA, demonstrated superior ability to predict FH.
01 - 03 Oct 2015
European Society for Paediatric Endocrinology