Variation of bone acquisition during growth hormone treatment in children can be explained by proteomic biomarkers, bone formation markers, body composition and nutritional factors

Objective: Growth hormone (GH) regulates both longitudinal growth and bone acquisition in children, and has profound metabolic effects. The aim was to investigate the association between proteomic biomarkers, body fat, nutrition and bone formation markers, and longitudinal growth in response to GH during the first year of treatment. The degree to which changes in these factors could explain variations in GH-dependent longitudinal growth and bone mineralization was also assessed. Methods: The individualized GH dose trial included 128 short prepubertal children with either normal (non-GHdeficient) or reduced levels of GH secretion (GH-deficient) (mean age +/- SD, 8.6 +/- 2.6 years; 90 boys), i.e., with a broad range of GH-secretion and GH-responsiveness, receiving GH treatment (mean 43 mu g/kg/day). Blood samples were taken and dual-energy X-ray absorptiometry (DXA) measured at baseline and 1 year of treatment. Step-wise multiple regression models were constructed including three steps with different independent variables added at each step to explain the variance in outcome variables (height(SDS), bone mineral content (BMC) and bone mineral density (BMD). Independent variables included in Step I were previously identified proteomic markers related to GH treatment response, bone formation markers (intact PINP, bone -specific alkaline phosphatase and osteocalcin), variables at treatment start (GH dose mU/kg/clay, GH maximum secretion, and difference between child's current and mid-parental height(SDS)). Step II explored the added influence of body composition data (body mass index or DXA). Step HI explored the added influence of serum nutritional markers and hormones. Results: Step I variables explained 71% of the variation in first year heightsDs gain, median (minimum-maximum) 0.8 (0.24-1.67); and the proportion explained rose to 73% following inclusion of step H variables and 75% following step III. Corresponding values for total body BMC were 58%, 78%, and 80%, respectively. Proportions fell by approximately 20% when BMC was adjusted for height; 33%, 57%, and 57% for steps I, II, and III, respectively. Corresponding values for total body BMD were 29%, 39%, and 45%, respectively. Conclusion: For total BMC, as much as 80% of the variation during the first year of GH treatment could be explained by proteomic biomarkers, body fat, nutrition and bone formation markers, whereas for height -adjusted BMC 57% could be explained. The inclusion of information about either body composition (fat/lean mass) or nutritional markers contributed with approximately 20%. The variation in heightsps gain could be explained to 75%. Hence, information of fat or nutrition markers was needed for explaining the variation in bone acquisition to the same magnitude as explaining the variation in height response.