VARIABILITY OF PULSATILE LUTEINIZING-HORMONE SECRETION IN YOUNG MALE-VOLUNTEERS

Characteristics of spontaneous pulsatile luteinizing hormone secretion were compared in ten young healthy men in three 24-h profiles obtaine d at intervals of 14 days and 3 months. The ages of the volunteers ran ged from 19 to 25 years, and heights and weights were within normal li mits. Blood samples were taken at 10-min intervals and plasma luteiniz ing hormone (LH) was determined in the same immunoradiometric assay us ing monoclonal antibodies. Conventional pulse detection was carried ou t with PULSAR and CLUSTER programs. In addition, a simultaneous multip le parameter DECONVOLUTION was applied. As a group, no significant dif ferences between the three profile series were found for any of the ca lculated parameters of LH concentration or LH secretion. However, most parameters showed low correlation coefficients between the three stud y periods, suggesting that substantial individual variations might con tribute to the more reliable group results. Median coefficients of var iation (cv) for the individual subject ranged from 9.7% (interpulse in terval and endogenous half-life) to 37.7% (mass per burst). However, t he maximal individual cv observed was 78.4%. Intra-individual variabil ity was lower than the variability between subjects for quantitative p roperties of LH concentration and secretion, although not significantl y so for all parameters. In conventional pulse detection, the highest individual reliability was found for mean and integrated LK concentrat ions (median cv 10.2 and 13.7%, respectively), number of pulses per 24 h (CLUSTER, median cv 12.2%), mean pulse amplitude (PULSAR, median cv 10%) and interpulse interval (CLUSTER, median cv 9.7%). In DECONVOLUT ION analysis, the endogenous LH half-life (median cv 9.7%), secretory burst amplitude (median cv 14.8%) and interburst interval (median cv 1 4.5%) revealed the lowest intra-individual variation. In contrast, the half-duration of a secretory episode and the mass of LH secreted per burst proved to be the least reliable measures (median cv 32.7% and 37 .7%, respectively). Calculated endogenous LH production rates correlat ed highly (p < 0.01) across all three sessions. The relative frequenci es of the LH peak amplitudes/heights and peak widths (durations) showe d almost identical distribution curves for all three sampling periods. In conclusion, a high reproducibility of group results for both integ rative parameters and pulse characteristics of LH concentrations and s ecretion were found in normal men. However, intra-individual reliabili ty was variable and at times considerable, depending on the parameter chosen. These observations suggest caution in the interpretation of si ngle LH profiles from individual subjects or patients unless the varia tion reported herein is considered.