Date Published: October 9, 2018
Publisher: Public Library of Science
Author(s): Pierre Lemaire, Gwénaëlle Duhil de Bénazé, Dick Mul, Sabine Heger, Wilma Oostdijk, Raja Brauner, Lisa Swartz Topor.
A previous single-center study established a mathematical model for predicting the adult height (AH) in girls with idiopathic central precocious puberty (CPP).
To perform internal and external validations by comparing the actual AH to the calculated AH established by this model and to update it.
The original formula, calculated AH (cm) = 2.21 (height at initial evaluation, SD) + 2.32 (target height, SD) – 1.83 (luteinizing hormone/follicle-stimulating hormone peaks ratio) + 159.68, was established in a sample of 134 girls (group 4) and was applied to additional girls with CPP seen in the same center (group 1, n = 35), in Germany (group 2, n = 43) and in the Netherlands (group 3, n = 72). This formula has been updated based on these extended data, and both versions are available at the following location: http://www.kamick.org/lemaire/med/girls-cpp15.html.
Despite the differences among the 4 groups in terms of their characteristics at the initial evaluation and the percentages of patients treated with the gonadotropin-releasing hormone analogue, they have similar calculated and actual AHs. The actual AHs are 162.2±7.0, 163.0±7.6, 162.4±7.7 and 162.1±5.6 cm in groups 1 to 4, respectively. They are highly correlated with the AHs calculated by the formula established in the original group (group 4), with R at 0.84, 0.67 and 0.69 in groups 1 to 3, respectively. When the actual AHs and the AHs predicted by the Bayley and Pinneau method are compared, the R is 0.76, 0.51 and 0.64 in groups 1 to 3, respectively.
This study validates and updates the previously established formula for predicting AH in girls with CPP. This updated formula can help clinicians to make treatment decisions.
Central precocious puberty (CPP) in girls is defined as the development of sexual characteristics before the age of 8 years due to the premature activation of the hypothalamic-pituitary-ovarian axis. In girls, CPP is idiopathic in more than 80% of cases [1–3]. Secondary to this activation, the secretion of estradiol increases the growth rate and accelerates bone maturation. These events can shorten the growing period, resulting in short adult height (AH). Treatment with a gonadotropin-releasing hormone (GnRH) analogue has been used to block the pituitary-ovarian axis and thus the estradiol secretion in girls with CPP for more than 30 years . However, the question of whether to treat girls with idiopathic CPP remains. The reported height gain (AH-predicted AH at the onset of treatment, according to the Bayley and Pinneau method ) varies from 0.3 to 9.8 cm .
Despite the differences in the initial characteristics established around one year after the onset of breast development and in the percentages of patients treated with the GnRH analogue between the original group (used to establish the formula) and the three native groups, the actual and calculated AHs are similar in the four groups. The model established in the original population gave a coefficient of correlation between the actual and the calculated AHs that was better in the internal validation but that was slightly lower and had more girls with absolute differences greater than 1 SD in the Dutch populations. However, its results are more accurate than those given by the Bayley and Pinneau method. The original model has been updated by including all populations. This model can be used to predict the AH in girls with idiopathic CPP. The prediction of the AH, together with the prediction of the age at first menstruation, may help clinicians to make treatment decisions for patients. We recently optimized our previous formula for predicting age at first menstruation in untreated girls based on the LH/FSH peak ratio  by adding the serum inhibin B concentration.
The occurrence of CPP in a given girl exposes her to decreases in growth potential and age at first menstruation. The current study allows us to validate the previously established formula for predicting AH in girls with idiopathic CPP. We use this formula in our daily practice to predict at the initial evaluation the adult height to help us to take decisions on the treatment.