Observed and predicted growth responses in prepubertal children with growth disorders: guidance of growth hormone treatment by empirical variables.

CONTEXT: Information about the expected growth response of children to GH therapy is currently inadequate.
OBJECTIVES: The aim of the study was to compare observed and expected growth in response to GH in prepubertal children and to propose how these parameters can be used to optimize GH therapy. Indices considered were observed growth, observed growth relative to reference data [height sd score (Ht SDS), change in (Delta) Ht SDS, height velocity (HV)], and observed growth relative to growth predicted from prediction models [Studentized residual (SR)]. Design/Setting/Patients/Intervention: Growth data from KIGS-Pfizer International Growth Database-on prepubertal children aged 1-13 yr with severe GH deficiency (GHD; maxGH <5 microg/liter; n = 2129), with less-severe GHD (maxGH of 5-10 microg/liter; n = 3075), and with Turner syndrome (n = 2350), and short children born small for gestational age (n = 993) were analyzed before and during 2 yr of GH treatment. MAIN OUTCOME MEASURE: For each patient group, growth responses during the first 2 yr of GH treatment were established. The relationships of HV and DeltaHt SDS with SR were determined.
RESULTS: Reference data were generated for assessing adequate individual responses. Responses to GH in terms of HV and DeltaHt SDS were greatest in children with severe GHD. HV and DeltaHt SDS were highly correlated with SR during only the first year of GH treatment (R approximately 0.7; P < 0.001).
CONCLUSIONS: Decisions on GH therapy regimens should be made using both traditional (HV or DeltaHt SDS) and prediction model-derived (SR) indices of growth response.