Influence of gender on the correlation between plasma growth hormone profiles and urinary growth hormone excretion

A lot of interest has been directed towards the measurement of urinary growth hormone (GH) excretion instead of plasma GH profiles or provocation tests. We investigated the factors influencing the relationship between 24- and 3-hour plasma GH profiles and urinary GH excretion in a cohort of 113 pediatric patients with growth disorders and healthy volunteers. Plasma and urinary GH were measured by polyclonal immunoassays differing in cross-reactivity with 20 kD GH (100 versus 46%), but not with 22-kD, dimer, deamidated and pituitary GH. In the 24-hour urine samples, only urinary GH excretion expressed as nanograms per 24 h correlated with plasma GH parameters, whereas the correlations for short-term samples were strongest if urinary GH excretion was expressed as nanograms per gram creatinine (r = 0.70-79, p < 0.00005-0.0001). In short-term samples urinary GH excretion depends on urinary volume and should thus be expressed in nanograms per gram creatinine, whereas 24-hour samples correlate best when urinary GH is expressed as nanograms per period. We found a significant sex difference (p < 0.02) in the correlation between 24-hour plasma GH profiles and urinary GH excretion with strong correlations in the female group (r = 0.63-0.78, p < 0.00005-0.0002) and a lack of correlation in the male group. The sex difference in the correlations between serum and urinary GH may reflect sex differences in GH profiles and metabolism, with urinary GH better reflecting the basal and slowly clearing portion of plasma GH than spontaneous GH peaks. The difference in cross-reactivities of molecular GH forms in polyclonal assays may have an impact on the correlation between plasma and urinary GH. Thus, the diagnostic value of urinary GH measurement as compared to serum GH profiles needs to be further evaluated.