Because of recent interest in the effects of physical exercise on immunologic function, we decided to use state-of-the-art methods to evaluate cytokines in the peripheral blood leukocytes (PBLs) of 7 men before and after a maximal treadmill stress test. Change in cytokine gene expression was quantified from PBLs using a reverse transcriptase polymerase chain reaction assay (RT-PCR). In contrast to reports on serum levels or using in vitro testing, direct gene expression of TNF-alpha decreased after the stress test (p < 0.008). However, the 47% decrease was relatively small and of questionable biological significance. Levels of IL-1a, IL-1b, IL-2, IL-4, IL-6, IL-10, and INF-gamma did not change.

Plasma concentrations of adrenal, thyroid, and testicular hormones were measured at 4-h intervals around the clock in male hamsters on long (14:10-h light-dark cycle) and short (10:14-h light-dark cycle) days. Plasma corticosterone, cortisol, thyroxine (T4), triiodothyronine (T3), and testosterone rhythms were present on long days. The only one of these hormones to have a significant rhythm on short days was cortisol, but even its amplitude was suppressed compared with the cortisol rhythm on long days. Short days also lowered mean plasma levels of cortisol, T4, T3, and testosterone. Finally, short days raised the ratio of corticosterone to cortisol and lowered the ratio of T4 to T3. Both ratios had significant rhythms on long days but not on short days. Because of the many interactions among adrenal, thyroid, and testicular hormone axes, it is unclear whether the primary effect of short days is on one of these endocrine systems or on another factor that has separate effects on each of the hormone rhythms that was measured. Nonetheless, it is clear that a major effect of short day lengths in hamsters is to suppress hormone rhythms. Explanations of photoperiodic effects that depend on endocrine mediation should take this into account.

Previously we showed that rhesus monkeys have ultradian cortisol rhythms with periods of about 90 min that persist during the infusion of large amounts of adrenocorticotropin. The experiments reported here showed that rats exhibit ultradian corticosterone rhythms with a similar period (median period 61 min). However, rat ultradian corticosterone rhythms were distorted by large slow trends that had to be removed before the rhythm could be adequately assessed. The need for sophisticated trend removal suggests that physiological variables that have been described as exhibiting episodic fluctuations (e.g., human cortisol) may actually be periodic and ultradian. The ultradian glucocorticoid rhythm persisted during stress both in monkeys (median period 80 min) and rats (median period 52 min). However, the glucocorticoid stress response appeared to produce large slow trends in both monkeys and rats, indicating that environmental influences may introduce slow trends that make raw data difficult to evaluate. More importantly, the fact that ultradian glucocorticoid rhythms persist during stress provides further evidence of limits on the classical concept that glucocorticoid secretion is tightly and inseparably linked to hormonal events in the hypothalamic-pituitary unit.

For 5 days, rats were exposed to shocks that were signalled by a light 0, 33, 66, or 100% of the time. Basal hormone levels and responses to a light-shock pair were measured daily. Greater predictability was associated with higher basal plasma corticosterone and norepinephrine levels indicative of chronic stress. Habituation of the corticosterone response was also less in the groups with greater predictability. However, predictability did not affect plasma prolactin or epinephrine responses. Because the endocrine systems responded differently, it is unlikely that the changes were due to a unitary process. Greater predictability appeared to be more stressful in this paradigm. Both associative and nonassociative factors have major roles in determining the hormonal responses to repeated presentation of stressors.

We have developed a rhesus monkey model that enables us to investigate physiological rhythms and circadian effects on performance in an integrated framework. Monkeys worked for 8 h/day on a two-component task (a vigilance trial followed by a discrimination trial) for their daily food aliquot. Concurrently, we recorded activity and temperature rhythms around the clock. To test the model, we studied rhythms and performance during entrainment to a 24-h light cycle and after a 6-h phase advance. Results from this animal model displayed many of the essential characteristics seen in similar human experiments. During stable entrainment, temperature rhythms reached their maximum amplitude in late afternoon, with activity rhythms reaching their maximum amplitude several hours earlier. Performance exhibited consistent task-dependent variations over the course of daily sessions. Speed of discrimination performance was fastest at the beginning of the session, and speed of vigilance performance was fastest several hours later. After a 6-h phase advance, monkeys exhibited transient internal desynchrony with activity resynchronizing faster than temperature. Both vigilance and discrimination were impaired after the phase shift, with vigilance exhibiting larger-magnitude and longer-lasting impairments than discrimination. A second drop in performance was seen 10-14 days after the phase shift. These data replicate and extend earlier work in humans and show that this model can be used in the study of chronobiological questions that would be too expensive or too impractical to do with humans.