Circadian and circannual rhythms of hormonal variables in elderly men and women.

A group of fourteen men (73 +/- 5 yr of age), and eighteen women (77 +/- 7 yr of age) institutionalized at the Berceni Clinical Hospital, Bucharest, Romania, were studied over a 24-hr span once during each season (winter, spring, summer and fall). All subjects followed a diurnal activity pattern with rest at night and ate three meals per day with breakfast at about 0830, lunch at about 1300 and dinner at about 1830. The meals were similar, although not identical for all subjects during all seasons. On each day of sampling blood was collected at 4-hr intervals over a 24-hr span. Seventeen hormonal variables were determined by radioimmunoassay. Statistically significant circadian rhythms were detected and quantitated by population mean cosinor analysis in pooled data from all four seasons in both sexes for ACTH, aldosterone, cortisol, C-peptide, dehydroepiandrosterone-sulfate (DHEA-S), immunoreactive insulin, prolactin, 17-OH progesterone, testosterone, total T4 and TSH. In women, estradiol and progesterone also were determined and showed a circadian rhythm during all seasons. Total T3 and FSH showed circadian rhythm detection by cosinor analysis in the men only; LH showed no consistent circadian rhythm as group phenomenon in men or women. A circannual rhythm was detected using the circadian means of each subject at each season as input for the population mean cosinor in the women for ACTH, C-peptide, DHEA-S, FSH, LH, progesterone, 17-OH progesterone and TSH. In the men, a circannual rhythm was detected for ACTH, FSH, insulin, LH, testosterone and T3. There were phase differences between men and women in ACTH, FSH and LH. In those functions in which both the circadian and circannual rhythms were statistically significant, a comparison of the amplitudes showed in the women a higher circannual rather than circadian amplitude for DHEA-S. In 17-OH progesterone, TSH and C-peptide, the circadian amplitude in women was larger. In men, the circannual amplitude of T3 was larger than the circadian amplitude and in insulin the circadian amplitude was larger than the circannual amplitude. There was no statistically significant difference between the circadian and circannual amplitudes in the women in ACTH and progesterone and in the men in ACTH and testosterone.