A Statistical Method for Detection of Hormone Secretory Spikes

Abstract

A method for evaluation of hormonal time series is proposed. The method consists of partitioning blood plasma hormone concentrations for each experimental unit (or animal) into an inlier subgroup and an outlier subgroup using the skewness coefficient test after correction for serial correlation. The inlier subgroup contains the baseline sample, while the outlier subgroup contains those responses which constitute hormone secretory spikes. Frequencies of secretory spikes are found by inspection of the outlier subgroup and the mean magnitude of the secretory spikes is calculated by averaging the maximum value of each secretory spike. Mean values of the inlier subgroup may be compared. Two examples using the described method are cited. The first example, which is examined in detail, was taken from a study of plasma growth hormone (GH) concentrations in bulls and steers. The results demonstrate a difference between bulls and steers for mean magnitude but no difference in frequency of secretory spikes or baseline mean plasma growth hormone. The second example was taken from a study examining GH secretory patterns in immature rams and ewes. There was no significant difference between immature rams and ewes for baseline mean plasma GH or frequency and mean magnitude of GH secretory spikes. Copyright © 1978. American Society of Animal Science . Copyright 1978 by American Society of Animal Science.