Analysis of trough serum growth hormone concentrations: comparison of an immunoradiometric assay and a sensitive ELISA for growth hormone.

OBJECTIVES: We compared a sensitive assay for GH (ELISA) with a conventional immunoradiometric (IRMA) assay with particular reference to the oscillatory activity detected by Fourier transformation and the estimation of trough concentrations using occupancy analysis. DESIGN: Eight healthy adult male volunteers underwent 24-hour profiles during which samples were drawn at 20-minute intervals. Samples were analysed by an ELISA and an IRMA system. MEASUREMENTS: The 24-hour serum GH concentration profiles were subjected to Fourier transformation and to occupancy analysis. RESULTS: No additional GH periodicities could be determined in the ELISA data other than the well documented 180-200-minute periodicity. Median observed concentrations (OC) at 5% occupancy were 0.035 mU/l (range 0.004-0.22) for the ELISA and 0.035 mU/l (range 0.001-0.50) for the IRMA. For all OC parameters, 5, 50 and 95%, there was a good correlation between the ELISA and IRMA systems. The mean difference (bias) between the ELISA and IRMA were -0.05, -0.28 and -1.40 mU/l at OC values of 5, 50 and 95% respectively and the standard deviations of the difference at the same OC values were 0.10, 0.50 and 1.61 mU/l. CONCLUSION: Although there is a qualitative improvement on visual inspection of individual 24-hour serum GH profiles obtained using the ELISA system, there is little additional information gained in terms of pulse periodicity or occupancy analysis.