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 obtained at intervals of 14 days and 3 months. The ages of the volunteers ranged from 19 to 25 years, and heights and weights were within normal limits. Blood samples were taken at 10-min intervals and plasma luteinizing hormone (LH) was determined in the same immunoradiometric assay using monoclonal antibodies. Conventional pulse detection was carried out with PULSAR and CLUSTER programs. In addition, a simultaneous multiple parameter DECONVOLUTION was applied. As a group, no significant differences between the three profile series were found for any of the calculated parameters of LH concentration or LH secretion. However, most parameters showed low correlation coefficients between the three study periods, suggesting that substantial individual variations might contribute to the more reliable group results. Median coefficients of variation (cv) for the individual subject ranged from 9.7% (interpulse interval and endogenous half-life) to 37.7% (mass per burst). However, the maximal individual cv observed was 78.4%. Intra-individual variability was lower than the variability between subjects for quantitative properties of LH concentration and secretion, although not significantly so for all parameters. In conventional pulse detection, the highest individual reliability was found for mean and integrated LH concentrations (median cv 10.2 and 13.7%, respectively), number of pulses per 24h (CLUSTER, median cv 12.2%), mean pulse amplitude (PULSAR, median cv 10%) and interpulse interval (CLUSTER, median cv 9.7%). In DECONVOLUTION analysis, the endogenous LH half-life (median cv 9.7%), secretory burst amplitude (median cv 14.8%) and interburst interval (median cv 14.5%) 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 correlated highly (p < 0.01) across all three sessions. The relative frequencies of the LH peak amplitudes/heights and peak widths (durations) showed almost identical distribution curves for all three sampling periods. In conclusion, a high reproducibility of group results for both integrative parameters and pulse characteristics of LH concentrations and secretion were found in normal men. However, intra-individual reliability was variable and at times considerable, depending on the parameter chosen. These observations suggest caution in the interpretation of single LH profiles from individual subjects or patients unless the variation reported herein is considered.