Multidimensional reference regions for IGF-I, IGFBP-2 and IGFBP-3 concentrations in serum of healthy adults.

CONTEXT: The insulin-like growth factor (IGF) system exerts many effects on the growth and differentiation of both normal and malignant cells. The serum concentrations of insulin-like growth factor I (S-IGF-I), insulin-like growth factor-binding protein 2 (S-IGFBP-2) and insulin-like growth factor-binding protein 3 (S-IGFBP-3) and their inter-relations may differ in certain disease states from those seen in healthy individuals. OBJECTIVE: To estimate age-, gender- and body mass index (BMI)-specific univariate, bivariate and trivariate 95% reference regions for these components in healthy adults and present indices that will facilitate interpretation of patient observations in relation to these reference regions. DESIGN: Blood samples were taken in the morning from 427 healthy, non-fasting German blood donors of both genders (age range, 18-79 years; BMI range, 16-44 kg/m(2)). Reference regions were developed with multivariate regression methods. RESULTS: Regression analyses showed that S-IGF-I and S-IGFBP-3 levels decrease with increasing age, whereas S-IGFBP-2 concentrations increase with age (P<0.0001). Females had significantly higher S-IGFBP-3 levels than males (P<0.0001) and similar S-IGF-I and S-IGFBP-2 concentrations. Increasing BMI was associated with decreasing S-IGFBP-2 (P<0.0001), but was not significantly associated with the concentrations of the other two analytes. Controlling for age, gender and BMI, S-IGF-I and S-IGFBP-3 were positively correlated (r=0.58), whereas S-IGF-I and S-IGFBP-2, and S-IGFBP-2 and S-IGFBP-3 were negatively correlated (r=-0.11 and r=-0.18, respectively). Based on the regression models, which were controlled for age, gender and BMI, two- and three-dimensional 95% reference regions with associated patient indices were described for each pair of analytes and for the trio of analytes, respectively. CONCLUSIONS: The multivariate reference regions developed in this study should be useful in identifying patients with an abnormal IGF-IGFBP axis. It is suggested that introducing multidimensional reference regions and the described patient indices into clinical practice may aid monitoring of the safety of GH therapy. These patient indices may also be useful in the assessment of cancer risk.