RM Global Health

{mosimage}Mathematical models, particularly those incorporating both growth and endocrine data, can predict which short children will benefit from treatment with growth hormone (GH), according to a new study.

The study, published online in BMC Medical Informatics and Decision Making, built on previous research by incorporating a broader range of data and by evaluating children who were appropriate for gestational age (AGA) as well as those who were small for gestational age (SGA).

Using growth data from 415 short prepubertal children, researchers at the University of Gothenburg, Sweden, constructed mathematical models to predict growth response during the initial years of GH therapy. Endocrine investigations performed prior to GH treatment included a GH stimulation test (the arginin insulin tolerance test). In addition, a spontaneous 24-hour GH profile was obtained for 188 children, and serum leptin and insulin-like growth factor I (IGF-I) levels were measured in samples from 200.

The children in the study group, treated between 1986 and 1997, received a daily dose of GH ranging from 25 to 66 g/kg. Of the 415 children, 271 had been born AGA and 144 had been SGA. Additionally, isolated idiopathic GH deficiency was present in 280.

The predictive model using only growth data yielded a standard error of the residuals (SDres) of 0.23. The models adding endocrine data, however, brought the SDres down, to 0.15 in the case of the model incorporating 24h GH profile, IGF-I and leptin. This corresponds to a prediction interval of 1.2 cm in a 4-year-old boy.

The researchers noted that if a spontaneous GH profile is not possible, IGF-I and leptin data are still valuable.

Validation of the models with a separate cohort of 112 children showed results similar to those in the study group. These children, subjected to the same inclusion criteria as the study group, received GH therapy from 1998 to 2001, after recruitment for the study had closed.

"The models presented here are independent of birth size and provide the highest prediction accuracy available," the researchers write. "They serve as a tool to identify those children who may benefit from GH treatment, and to help choose the optimal GH dose during the first years of treatment in order to optimize the individual catch-up growth response."

BMC Medical Informatics and Decision Making 2007;7:40.