Statistical evaluation of coincident prolactin and luteinizing hormone pulses during the normal menstrual cycle.

The purpose of this work was 2-fold. First, we sought to develop statistical criteria by which it could be established that the coincident occurrence of pulses of two different hormones exceeds that which would occur by chance alone, thereby suggesting that secretion of the two hormones is either coupled or controlled from a single source generator. Using computer simulations of uncoupled pulse generators operating at different frequencies, we were able to derive the appropriate statistical criteria and to apply them to achieve our second objective, to determine whether the occasional coincidence of plasma LH and serum PRL pulses that occurs throughout the menstrual cycle in normal women exceeds that which would happen by chance. The results of the computer simulations indicated that pulses emanating from two completely independent oscillators will occur coincidently at a predictable rate, despite the fact that the generator sources are not coupled; moreover, the rate of coincidence is increased when the pulse frequency of one of the source generators is increased. Using this knowledge and the statistical criteria we derived, we analyzed the coincidence of LH and PRL pulses in five normal women during their early follicular, late follicular, and midluteal phases and in another five women during their late luteal phase. We found that the number of PRL pulses that occurred coincidently with LH pulses consistently exceeded that which would be predicted if the two pulse generators were operating completely independently of one another; however, only during the late follicular and late luteal phases was the coincidence level between LH and PRL pulses sufficiently high in a sufficient number of women to conclude that there was coupling between the pulse sources. These studies suggest, first, that stringent and rigorous statistical criteria must be applied to the analysis of spontaneously coincident secretory phenomena before it can be deduced that two pulse generators are indeed coupled, and second, that the pulse generators governing the secretion of PRL and LH are probably coupled, at least during certain phases of the menstrual cycle.