A prospective study of growth was undertaken in two groups of hyperkinetic children: a d-amphetamine-treated (index) group and a phenothiazine-treated (control) group. This study was coupled with longitudinal overnight sleep-endocrine assessments in the index group. At the end of the first year seven index subjects have been studied for growth and sleep-endocrine measures and eight control subjects have been studied for growth changes only. The significant findings are: (1) growth retardation in the index group; (2) inhibition of mean sleep-related prolactin secretion; (3) a highly significant correlation between loss of expected height percentile during the first year and decrease of mean sleep related prolactin at 6 months, as compared to baseline. These findings suggest that growth inhibition secondary to chronic d-amphetamine administration in hyperkinetic children might be mediated by inhibition of prolactin secretion.

Equimolar doses of D-amphetamine and L-amphetamine, and a water placebo were injected intravenously on different days into rhesus monkeys, and their plasma cortisol and GH responses were determined over a 3-hour period. D- and L-amphetamine (but not the placebo) equally suppressed plasma cortisol concentration and equally increased plasma GH concentration. Pretreatment of the monkeys with large doses of pimozide, a specific dopamine receptor blocker, did not appear to block the hormonal responses to either isomer of amphetamine. The results suggest that the hormonal responses to both D- and L-amphetamine were mediated by noradrenergic neurones stimulatory to GH and inhibitory to ACTH, and that D- and L-amphetamine exert equipotent noradrenergic effects in the neuroendocrine system.

Interest in possible neuroendocrine disturbances in endogenous depression is prompted by two lines of evidence: (1) clinical features of the illness suggest hypothalamic dysfunction; (2) the brain neurotransmitters implicated in depression also regulate neuroendocrine function. Our research reveals a marked, sustained hypersecretion in cortisol in severe depressive illness, which is apparently unrelated to stress and sleep disturbance, and which is associated with a distortion of the 24-hour cortisol secretory pattern. The hypersecretion is manifested primarily in the late afternoon, evening, and early morning hours, when cortisol secretion is normally inhibited. Growth hormone responses to hypoglycemia (but not to L-dopa) are also significantly reduced in endogenous depression, even when factors of age and the menopause are controlled. Postmenopausal depressed women appear to secret significantly less LH than normal postmenopausal women. Since all of these hormonal abnormalities can be reproduced by depletion of brain noradrenalin, the findings provide support for the the hypothesis of reduced functional noradrenergic activity in certain forms of depression.

After ingestion of 500 mg of levodopa, postmenopausal women had significantly diminished human growth hormone (HGH) responses (mean, 4.6 ng/ml), as compared with those of age-matched men (mean, 9.1 ng/ml; P smaller than .05). The differences between the groups were not related to plasma dopa concentrations. The HGH responses to levodopa of age-matched unipolar and bipolar depressed men, and of unipolar depressed postmenopausal women, did not differ significantly from their respective normal control groups. Depressive illness of these types does not appear to affect the HGH response to levodopa, once the effect of the menopause is taken into account.

Human growth hormone (HGH) responses to insulin-induced hypoglycemia were measured in ten postmenopausal women suffering from primary unipolar depressive illness, and in ten age-matched normal postmenopausal women. The mean maximal HGH response in the depressed patients was 4.6 plus or minus 4.4 ng/ml, and in the normals 13.3 plus or minus 9.8 ng/ml (P less than .05). All of the normal subjects had clinically adequate HGH responses, in contrast to only four of the depressed patients (P less than .01). The blood glucose responses were virtually the same in the two groups. Since brain catecholamines play a major role in mediating HGH responses to hypoglycemia, the findings are consistent with the hypothesis of diminished functional catecholaminergic activity in the depressed patients.

After 1 week of a normal control (baseline) period, 7 healthy young adult subjects were subjected to a 3-hr sleep-wake schedule, (ultradian) which was adhered to for 10 days. They were allowed eight 1 hr sleep times, equally spaced throughout each 24 hr period. They were then allowed a normal nocturnal 8-hr sleep time for 7 days. During all lights-out sleep periods, poly-graphic definition of sleep stages and waking time was made. On the sixth 24-hr period of the first week (baseline) and on the eighth 24-hr period of the ultradian period, sequential 20-min plasma samples were obtained by means of an indwelling intravenous catheter. Rectal temperature was obtained at regularly spaced frequent intervals. Despite significant sleep deprivation, a circadian pattern of total sleep time persisted throughout the 10-day ultradian condition. The distribution and amount of REM sleep time was most affected with stages 3–4 sleep least affected. The time of maximum sleep was delayed by approximately 6 hr. The temporal pattern of the secretory episodes of cortisol and the body temperature curves demonstrated a persistence of the 24-hr (circadian) periodicity for all subjects during the ultradian condition. A 3-hr cortisol cycle was superimposed on the 24-hr pattern. This 3-hr cycle was entrained to the 3-hr sleep-waking cycle such that low plasma concentrations of cortisol were associated with the dark (sleep) period and high concentrations with the first hour after "lights on." No correlation could be demonstrated between a specific sleep stage and the subsequent release of hormone even though a correlation was present for total sleep. The mean 24-hr output of GH was not different for the baseline and ultradian conditions. However, the sharp peak of GH secretion found between 11 PM to 1 AM in the baseline condition was not present in the ultradian condition. The persistence of a 24-hr temperature curve, sleep-waking cycle and cortisol pattern in spite of the attempt to disrupt these functions for 10 days demonstrates the highly resistant nature of these systems.

Serial plasma growth hormone (GH) concentrations were measured every 20 min for 24 hr before and after the administration of clomiphene citrate (100 mg/day for 7 days) to four healthy young adult male subjects. The number of GH secretory episodes and the magnitude of the peak plasma concentrations during both wakefulness and sleep were decreased after the clomiphene treatment periods.

In response to both insulin-induced hypoglycemia and the ingestion of 500 mg. of L-dopa, unipolar depressed patients (most of whom were postmenopausal women) secreted significantly less growth hormone (GH) than did normal subjects or bipolar depressed patients in the same age range (45 to 70 years). Prolactin responses to L-dopa were normal for nearly all subjects. Since it is believed that prolactin responses to L-dopa are mediated by brain dopamine and GH responses by brain norepinephrine, the data tend to support the hypothesis that there is a disturbance of brain norepinephrine metabolism in unipolar depressive illness. However, further study is needed to determine the effects of the menopause on GH secretion.

Plasma human growth hormone responses to oral administration of 500 milligrams of L-dopa were analyzed in three groups of subjects: normals, age 20 to 32; normals, age 48 to 68; and unipolar depressed patients, age 45 to 68. While only 7 percent of the young normals had deficient human growth hormone responses to this stimulus, 36 percent of the older normals and 77 percent of the depressed patients failed to have adequate responses, suggesting an effect of age and a further effect of depressive illness. Because the release of human growth hormone appears to be closely related to brain catecholamine metabolism, the deficient responses in the depressed patients may provide further support to the concept of a neurochemical defect in depressive illness.