Effects of "jet lag" on hormonal patterns. IV. Time shifts increase growth hormone release

Twenty-four-hour GH profiles were obtained in five normal male volunteers before travel, 1, 11, and 21 days after the Brussels-Chicago flight (time shift, 7 h); and 1, 11, and 21 days after the return flight. The westward and eastward travels involved, respectively, periods of 23 and 33 h of sleep deprivation. One year later, two of the five volunteers were submitted, in the laboratory, to an investigation mimicking the conditions of sleep deprivation undergone in the course of the eastward travel and involving two 24-h periods of blood sampling. Blood samples were drawn every 15 min, and sleep was polygraphically monitored. The amounts of GH secreted were quantified, and their relationship with the different sleep stages was analyzed. Time shifts, whether caused by "jet lag" or by sleep deprivation in the laboratory, had two effects on GH secretory patterns. First, a marked increase in GH release, due to an augmentation of the magnitude, rather than the number, of secretory spikes was observed, independently of sleep disturbances. Return to basal levels was slower after westward than after eastward travel and took at least 11 days. Second, 1 day after the eastward transportation as well as immediately after 33 h of sleep deprivation, the major GH spike, which occurred in early sleep in the other studies, was shifted to late sleep. In these investigations, the only consistent alteration of sleep was a reduction in the amount of rapid eye movement (REM) stage. The occurrence of GH spikes in sleep was significantly associated with slow wave (SW) stage. However, total amounts of GH secreted during sleep were negatively correlated with the total duration of REM stages rather than positively correlated with the total duration of SW stages. A spike by spike analysis showed that the amount of GH secreted correlates best with the ratio (SW - REM) to (SW + REM), which relates the amount of REM preceding the spike to the amount of SW during the spike and thus constitutes an indicator of the status of the REM-non-REM oscillation.