The aim of the study was to test the hypothesis that in serial determinations of concentrations of LH and FSH involving blood samples taken every minute, the observed pulses of LH and FSH which last less than 3–4 min might not be a physiological phenomenon but part of the 'noise' of the radioimmunoassay or blood-sampling technique.
Blood was sampled every minute for a period of 90 min in six men. During the first 45 min, blood was sampled by means of vacuum tubes only. During the second 45 min, sampling took place with a syringe via a rubber stopper, either using a tourniquet (n = 3) or flushing the cannula with heparinized saline.
Three criteria were used to identify variations in the patterns of LH and FSH as true hormonal changes. First, a threshold was used which had to be exceeded by the difference between nadir and maximum values before a pulse could be identified. An average of approximately six pulses per 90 min was found in both the LH and FSH series. The majority of these pulses lasted less than 3–4 min. In two subjects, larger LH pulses of longer duration were measured. Secondly, differences between duplicate measurements of nadir and/or maximum values of more than one-third of the amplitude of a pulse were considered unacceptable. This involved about 75% of the pulses. Thirdly, the reproducibility of the hormone variations was estimated. In one subject, concentrations of LH were measured four times in four separate assays. Measurement of FSH concentrations in this subject and of LH and FSH in the samples from the other five subjects were repeated once again, but only in those parts of the series of samples which had shown hormone variations beyond the threshold composed of acceptable duplicate measurements. Only the larger variations of longer duration, as found earlier in two of the LH series, were reproducible.
The different blood-sampling techniques used had no significant influence on the frequency of pulses.
It was concluded that a rapid, small amplitude, pulsatile pattern of release of LH and FSH is probably not present in men or is obscured by limitations of current techniques of radioimmunoassay.
J. Endocr. (1987) 114, 153–160