Ultra rapid blood sampling for the determination of short-term variations in the circulating concentration of oestradiol in man

Abstract

ABSTRACT Blood was taken from three healthy female and three healthy male volunteers every 20 s for 15 min. The serum level of oestradiol was measured and the pattern of variation assessed by a cusum plot of the sequential data, by autocorrelation of the detrended serial data, and by looking for pulses. In two cases the variation in oestradiol values was greater than that which could be attributed to variation in the assay. Both these subjects showed a significant overall change in values during the sampling period (an increase and a decrease). There was no trend in the remaining four subjects. In two of the six subjects there was significant autocorrelation of detrended sequential levels. Defining a 'pulse' as three times the assay coefficient of variation no more pulses were identified than was expected from random fluctuations. By frequency analysis the two subjects with significant autocorrelation showed periodic fluctuations of approximately 70/h and 9/h respectively. It is apparent that both the rate of sampling and the method of analysis greatly influence the evaluation of pulsatile release of oestradiol. J. Endocr. (1988) 118, 161–165