Twenty-four hour plasma levels of growth hormone and prolactin in Huntington's disease.

Since hypothalamic neural degeneration is thought to occur in individuals with Huntington's disease, anterior pituitary hormone secretion which is in part regulated by the hypothalamus, was postulated to be altered in patients with this disease. To test this proposal, nine females with Huntington's disease were matched with controls to participate in a 24-hour basal level study of growth hormone and prolactin concentration in plasma. Patients who were free from all centrally active medication for at least six months and normal volunteers had blood sampled at 30-minute intervals over 24 hours in a minimal stress environment. The results demonstrated that plasma levels of growth hormone were elevated throughout the 24-hour time period in Huntington's disease individuals. Despite the elevation, the mean growth hormone curve of the Huntington's disease group retained characteristics similar to the control curve throughout the 24-hour time. Basal 24-hour plasma prolactin concentrations in Huntington's disease patients showed no difference from those in control individuals.

Huntington's disease is an autosomal dominant neurological disorder characterised by ganglion cell degeneration throughout the brain. The well known pathological changes in basal ganglia and, to a lesser extent, in cerebral cortex have been associated with the involuntary movements and cognitive deterioration which occur in the illness. ' Neuronal loss and degenerative changes associated with selective biochemical abnormalities have also been reported to involve the hypothalamus.2 4 These observations suggest that alterations in hypothalamic-modulated function may occur in Huntington's disease. Since the hypothalamus influences the release of hormones from the anterior pituitary, characteristic abnormalities in the secretion of growth hormone (GH), and prolactin (PRL), might be expected.
Numerous previous investigations have attempted to evaluate basal growth hormone and prolactin plasma levels in patients with Huntington's disease (reviewed in ref 5). However, these studies have often drawn conflicting conclusions about whether plasma GH and PRL values are increased, decreased, or remain within the normal range in Huntington's disease individuals. GH and PRL are both episodically released in a periodic pattern; thus, single or even multiple closely spaced measurements can be insufficient to fully describe basal levels. Because implications important to our understanding of the pathophysiology of Huntington's disease might be drawn from alterations in anterior pituitary hormone control, plasma GH and PRL levels were studied with a 30-minute repeated sampling technique over a 24-hour period in Huntington's disease subjects.

Methods
Nine females with Huntington's disease and an equal number of normal controls were hospitalised for study. Six of the Huntington's disease patients had the classical adultonset form of this disorder; the three remaining individuals had the rigid-akinetic variant. All had a positive family history. Mean (± SD) age of the Huntington's disease group was 43 ± 12 years and of controls was 45 ± 14 years. Five of the nine Huntington's disease subjects had never been treated with antipsychotic drugs; furthermore, none of the patients or controls received any regularly prescribed centrally active drugs for at least six months prior to admission to this study. Menstruating women were tested 1134 Twenty-four hour plasma levels ofgrowth hormone and prolactin in Huntington's disease during the luteal phase of their cycle. All study subjects gave their informed consent before participating in this investigation.
The 24-hour sampling procedure began by the placement of an indwelling heparinised intravenous catheter in the forearm. Following a 45-60 minute stabilisation period, blood sampling began (5 ml/sample) and was repeated at 30 minute intervals for 24 hours. Immediately after collection, specimens were placed on ice; plasma was separated from whole blood within one hour and frozen at -70°C until assay. At night, the sleep of each subject was monitored by electroencephalography to document the duration of total sleep and the proportion of slow wave sleep. All subjects were maintained on regular hospital diet given on a fixed schedule (0800 h, 1230 h, 1700 h) and allowed only to ambulate quietly within the confines of their own rooms.
A double antibody radioimmunoassay was used to measure levels of GH and PRL in plasma. Materials from the Hormone Distribution Program of the National Insti- Sleep EEG recordings (mean ± SD) showed a reduced amount of total sleep in patients compared with control individuals (332 ± 29 min and 394 ± 13 min respectively), but a similar duration of slow wave sleep (78 ± 14.5 min and 92.7 ± 18*5 min respectively).
The possible effect of stress as a factor contributing to elevations in the plasma content of GH or PRL was evaluated by comparing mean AM plasma cortisol levels (sampled at 0730 h, 0800 h, 0830 h) and mean PM plasma cortisol levels (sampled at 2200 h, 2230 h, 2300 h) between the Huntington's disease patients and normal control groups. The mean (±SD) morning and evening plasma cortisol levels (ng/ml) in the Huntington's disease patients was 205 ± 10 and 99 ± 14, respectively, and in the control subjects was 151 ± 13 and 60 ± 15, respectively. The differences between Huntington's disease and normal subjects for either morning or evening samples failed to reach statistical significance.
The possibility that the involuntary movements in some of the Huntington's disease patients influenced plasma GH and PRL concentrations was evaluated by comparing results from three of the nine-member Huntington's disease cohort who had the nonchoreatic, rigid-akinetic form of this disorder. Mean (±SD) 24 hour levels of GH (4-03 + 0*28 ng/ml) or PRL (16 ± 1*7 ng/ml) in these individuals did not differ significantly from the values found in the remaining six patients with the classical, choreatic form of Huntington's disease. Moreover, the plasma GH elevations in the Huntington's disease group continued during the evening and night in both choreatic and nonchoreatic subjects when normal sleep patterns were demonstrated by EEG but no involuntary movements could be observed.

Discussion
The present data may help clarify the conflicting reports on basal levels of plasma GH and PRL in Huntington's disease. Prior studies have found basal GH levels either elevated7-9 or unchanged""-2 from controls, and basal PRL levels either depressed,'3 elevated," 14 or unchanged.9 1'-17 The results of this study indicate that the basal plasma levels of GH are significantly increased in Huntington's disease subjects while basal PRL levels remain within the normal range. The GH findings in the Huntington's disease population could reflect either an increase in GH secretion from the anterior pituitary, or an alteration in its metabolism. Histopathologic changes already identified in the hypothalamus of Huntington's disease individuals provide the anatomical basis for Durso, Tamminga, Ruggeri, Denaro, Kuo, Chase postulating an alteration in the regulation of GH secretion to account for the abnormality. Since PRL levels remained normal, the present results do not support previous conclusions"- '3 that a generalised abnormality in the dopamine system, which contributes to the regulation of both GH and PRL, occurs in Huntington's disease.
The potential neural basis of the observed GH abnormality could be a disorder of either of the two hypothalamic factors regulating pituitary GH synthesis and release. Growth Hormone Releasing Factor (GRF) phasically stimulates GH secretion; '8 while Somatotropin-Release-Inhibiting Factor, somatostatin, tonically inhibits GH release."' Because somatostatin appears to mediate tonic and inhibitory control of GH,20 a loss of this peptide could disinhibit GH secretion, thus elevating plasma GH. If the somatostatin effect is diminished in Huntington's disease, due to a loss of inputs to the hypothalamic somatostatin system, to a degeneration of somatostatin-containing neurons, or to a hyporesponsivity of relevant somatostatin receptors, resultant plasma GH levels would be increased. Indeed, levels of somatostatin in spinal fluid of Huntington's disease subjects have been reported to be substantially reduced.2' However, diminished somatostatin levels in suprachiasmatic areas of hypothalamus have not been confirmed. 22 Further evaluation of the somatostatin system in Huntington's disease may therefore be of interest. Necropsy brain tissue was received from Dr Edward D Bird for analysis of hypothalamic somatostatin levels, supported by #1-RO1-MH/NS 31862 (Reference 22).
Ms Marjorie Gillespie contributed substantially to the clinical care and testing of the subjects.