Correlations between mean LH levels and LH pulse characteristics: differences between normal and anovulatory women.

Objective: Since LH secretion occurs as a series of pulses, relationships between mean LH levels and LH pulse characteristics are to be expected. The aim of this study was to determine whether such relationships are similar in normal women and anovulatory patients.
Design: We studied the correlations between mean LH levels and the products amplitude x frequency and area x frequency of LH pulses in normal women and in patients with disorders of ovulation. Blood samples were taken from each subject every 10 minutes during 6 hours on the 8th day after the last menses.
Patients: The patients were divided into three groups: patients with polycystic ovary syndrome (n = 11), patients with idiopathic anovulation (n = 14) and patients with short luteal phase (n = 13). Their results were compared to those of 12 normal women.
Measurements: LH was evaluated with an immunoradiometric assay. LH data were analysed with a 3 standard deviation threshold criterion for significant peaks, and with cluster analysis algorithm using 1, 2.5 and 5% false positive error rates and 'optimal parameters' (which give less than 5% false positive and false negative error rates in LH male data).
Results: Highly significant correlations between amplitude x frequency, area x frequency and mean LH were found in normals and patients with short luteal phase; no significant correlation was found in patients with polycystic ovary syndrome, while significant correlations were found in patients with idiopathic anovulation only with some of our criteria for peak detection.
Conclusion: The differences that we found between the groups suggest that when commonly used methods are employed to determine LH pulse characteristics, most of the significant LH pulses are taken into account in normals and patients with short luteal phase, but not in anovulatory patients, especially in patients with polycystic ovary syndrome. This method using two correlations appears to be a simple and useful way to show the differences in the mechanisms by which mean LH levels are achieved in normal subjects and patients.