Psychological Medicine, 1992, 22, 55-60 Printed in Great Britain
Pyridostigmine-induced growth hormone responses in healthy and depressed subjects: evidence for cholinergic supersensitivity in depression V. O'KEANE, K. O'FLYNN, J. LUCEY AND T. G. DINAN 1 From Trinity College Medical School, St James's Hospital, Dublin, Ireland
Theorists have extrapolated the cholinergic supersensitivity theory of affective disorder from a convincing and broad spectrum of clinical observation and research. This hypothesis is tested using a neuroendocrine probe approach with the challenge drug pyridostigmine, an indirect cholinergic agent thought to release growth hormone (GH) by decreasing inhibitory somatostatin tone. The consequent increments in plasma GH were considered to reflect central acetylcholine responsivity. Fifty-four volunteers were tested: 27 DSM-III-R major depressives (18 women and 9 men) and 27 age- and sex-matched healthy controls. Subjects were cannulated at 9.00 h following an overnight fast and two baseline samples were taken at 15 min intervals. Pyridostigmine 120 mg was administered orally and thereafter samples were taken at the time points +60, +90, + 120 and + 180 min. GH responses were significantly greater in depressives than controls and this effect was more marked for men than women. These results support the proposal that muscarinic upregulation and/or supersensitivity is associated with depression. SYNOPSIS
Physiological studies exploring possible cholinergic overdrive, rather than its relationship to the monoamine system, are more convincing. Arecoline, a direct cholinergic muscarinic agonist, when administered to bipolar patients in remission induces the onset of REM sleep significantly more rapidly than in normal controls suggesting a muscarinic supersensitivity in this disorder that is state independent (Sitaram et al. 1982). Another group found this abnormality was not reproducible in depressed patients who had been in remission for longer periods and has questioned its validity as a state marker (Berger et al. 1989). This abnormality in cholinergic-induced REM sleep is also found in atypical depressives without anxiety symptoms (Wager et al. 1990). Centrally acting cholinergic drugs produce sleep changes characteristic of major depression (Sitaram et al. 1977) and induce dexamethasone nonsuppression in normal volunteers (Carroll et al. 1980). One previous study has applied a neuroendocrine challenge approach. Risch (1982) found that physostigmine caused significantly greater increases in /?-endorphin in depressed compared
INTRODUCTION The dysphoric and insomnious effects of cholinomimetic agents were first noted by Grob and colleagues as far back as 1947. Later Gershon & Shaw (1961) documented 16 cases of severe psychiatric sequelae resulting from chronic exposure to these organophosphorus insecticides and in 1972 Janowsky et al. incorporated these and other clinical observations into the monoamine theory of affective disorder by proposing a cholinergic-adrenergic balance hypothesis. Briefly, this states that disturbances may arise from an imbalance between central catecholamine and acetylcholine (ACh) activity: depression arising from a relative overactivity in ACh function associated with decreases in catecholamines and mania arising from the reverse. In support of this theory they found that physostigmine (an anticholinesterase)-induced adrenaline release only tended towards blunting in depressives compared to healthy controls (Janowsky et al. 1986). ' Address for correspondence: Dr Ted Dinan, Psychiatric Unit, St James's Hospital, Dublin 8, Ireland.
55
V. O 'Keane and others
56 •
• • »
§ 20
-
O
rowth
o
O
•
10
•
0
• t
• • f
Depressed males
Healthy controls
FIG. I. A growth hormone (difference between baseline and maximum increase post-pyridostigmine administration) responses in 9 depressed and 9 healthy males.
to normal subjects. We used the cholinesterase inhibitor pyridostigmine as it is well tolerated; its mechanism of action in the release of GH has been extensively studied and it has been used in diverse endocrine studies, e.g. diabetes mellitus (Giustina et al. 1990) and obesity (Cordido et al. 1989). It seems that pyridostigmine, by stimulating ACh neurotransmission, causes a reduction in the tonic inhibitory somatostatin control of GH release and an increased secretion of GH from the somatotrophs (Ross et al. 1987). We used this challenge regime to compare ACh function in depressed and healthy states to test the hypothesis that ACh supersensitivity is associated with this disorder. METHOD Fifty-four subjects gave fully informed consent to participate in this study - 27 depressives and 27 healthy controls. The depressed group consisted of 18 women and 9 men (mean age + SEM: 35-4 + 3-4 and 35-3 + 2-9 years respectively) who fulfilled DSM-III-R (American Psychiatric Association, 1987) criteria for major depression;
unipolar type. Twenty had a past history of depression. Six were drug naive and the remainder were drug-free for a minimum of four weeks. Of the 21 who had previous exposure to psychotropics 20 had been on tricyclic antidepressants; 2 on neuroleptics and 4 on both. Fifteen patients in total had been exposed to medication with muscarinic affinity in the preceeding three months. Severity was assessed using the 17-item Hamilton Rating Scale for Depression (HAMD) (Hamilton, 1960). The mean score for males was 21-8 ±1-5 and for females was 21-5 + 0-7. The degree of endogenicity was assessed using the Newcastle Scale (Carney et al. 1965) with males yielding a mean score of 4-4 + 0-5 and women a score of 4-3 + 0-3. In addition to full physical examination all had routine haematology, biochemistry and thyroid function screening. Those aged above 40 years had an ECG and chest X-ray. Our control group, 18 men and 9 women (mean age 33-8 + 9-9 and 32-3 + 2-3 years respectively) were physically healthy and had no personal or first-degree relative history of psychiatric disorder. All volunteers weighted within normal limits for sex and age. Premenopausal female subjects were tested in the follicular phase of the menstrual cycle. Subjects presented at 9.00 h following an overnight fast. They relaxed for 15 min following insertion of a cannula and heparin bung in a forearm vein. Baseline blood samples were then taken at —15 and 0 min and thereafter at +60, + 90, +120, +180 min. Pyridostigmine 120 mg orally was administered at 0 min. Volunteers remained in a supine position and peripheral signs of muscarinic hyperactivity such as hypersalivation, sweating and lacrimation were monitored throughout the procedure. Blood samples were collected in lithium heparin bottles and analysed blind to subject status using a standard radioimmunoassay technique with a sensitivity of 0-45 mU/1, an intrassay precision of 1-6% and an interassay precision of 2-3% at about 10mU/l. Hormonal responses over time between groups were compared by two-way ANOVA and post hoc tests were applied. Delta GH (A GH: difference between baseline values and the maximum increase post-pyridostigmine administration) measures were also used to compare
Pyridostigmine-induced growth hormone responses
responses. Student's / tests and correlational analysis were used where appropriate. Results are expressed as mean + SEM.
57
15
12
RESULTS All but 1 healthy and 1 depressed subject responded to pyridostigmine with an increase in plasma GH. There was a significant sex-based GH response difference in healthy controls (mean A GH (men) = 41 + 10 mU/1; mean A GH (women) = 14-75 + 2-6 mU/1; / =2-6, df = 52, P = 0-012) and therefore results for males and females were analysed separately. There were no differences in basal GH levels between depressed and control groups, either males or females.
E
o U
0 0
60
120
180
Time (min)
50
•
FIG. 3. Plasma growth hormone levels following pyridostigmine administration at 0 min in 9 depressed male patients (O O) and 9 healthy male controls ( • • ) . Results are expressed ±SI;M.
•
40 -
• •
••
•
• •
30
•
• • ••
o .c
1 20
• i
• •
•
• ••
• • *
10 -
•
•
•
0 Depressed females
• • Healthy controls
FIG. 2 A Growth hormone (difference between baseline and maximum increase post-pyridostigmine administration) responses in 18 healthy females.
A mean A GH in depressed men of 16-1 ± 3-4 mU/1 was significantly elevated compared to a mean A GH of 41 + 1 mU/1 in healthy men (/ = 3-4, df = 16, P = 0-003) (see Fig. 1). Responses in depressed females were likewise significantly enhanced (mean A GH = 23-3 + 3 mU/1) compared to control females (mean A GH = 14-8 ±2-6 mU/1: t = 2\, df = 34, P = 004) (see Fig. 2). A repeated-measures 2-way ANOVA used to assess differences over time between depressed males and healthy controls demonstrates a significant effect for group (F(\, 88) = 9-43, P = 0002); time (F(4, 85) = 3-43, P = 001) and the group by time interaction (F(4, 85) = 2-63, P = 004). Post hoc Tukey comparisons show these responses to be significantly different at +90 (P < 0-05) and +120 min (P < 005) (see Fig. 3) A two-way ANOVA comparing hormonal responses over time between the depressed and control female group again demonstrates significant effect for group (F (1, 178) = 3-99; P = 0047), time (F(4, 175) = 7-85; P < 0-0001) and the group by time interaction (F (4, 175) = 2-6; P < 005) with post hoc Tukey comparisons showing significant effect at + 60 min (P < 0-05) only (see Fig. 4).
V. O'Keane and others
58 18
15
D 12 E
o
6
0
0
60
120
180
Time (min) FIG. 4. Plasma growth hormone levels following pyridostigmine administration at Omin in 18 depressed female patients ( O 1 - - © ) and 18 healthy female controls ( • #). Results are expressed as + S1-M.
There was no correlation between the severity of the depression as rated by the Hamilton Rating Scale and the GH response for either women (r = 0-4, df = 18, P = 009) or men (r = 0-38, df = 9, P = 0-3). Neither was there a significant correlation between scores on the Newcastle Scale (males: P = 0-34; females P = 0-16); age (males: P = 0-22; females: P = 0-56); length of the illness (males: P = 018; females: P = 0-23); presence of a positive family history (males: /> = 0-36; females: P = 0-36), weight loss (males: P = 0-12; females: P = 019) or a history of a previous episode of depression (males: P = 0-78; females: P = 0-4) and the endocrine response. There was no difference in A GH responses between those exposed to psychotropics with cholinergic affinity in the preceeding three months (N = 15) and those who were not (N = 12: / = 1-6, df = 25, P = 0\) or between the drug naive and drug-exposed group (t = 0-8, df = 25, P = 0-4). DISCUSSION This study demonstrates a significantly augmented GH response to pyridostigmine in depressed patients compared to healthy controls.
Pyridostigmine is a cholinesterase inhibitor and it has become increasingly apparent that ACh is an important neuromodulator of growth hormone (GH) release. Physiological and pharmacological stimuli of GH secretion are blocked by ACh antagonists (Casanueva et al. 1984) and indirect cholinergic agonists enhance the GH response to stimuli such as GHRH (Massara et al. 1986), clonidine and arginine (Cordido et al. 1990). Anticholinesterases alone also induces increases in GH in non-obese human subjects (Casanueva et al. 1989). GH release is regulated by a combined stimulatory GHRH and an inhibitory somatostatin input from the hypothalamus to the somatotroph cells of the anterior pituitary, with a negative feedback effect exerted by GH on its own secretion. It is suggested that GH effects this negative feedback by increasing hypothalamic somatostatin release and this somatostatin secretion is under inhibitory cholinergic control (Ross et al. 1987). Pyridostigmine impairs the GH feedback effect by decreasing somatostatin tone (Massara et al. 1986; Ross et al. 1987). That ACh effects on GH secretion are mediated through somatostatin is borne out by rat studies demonstrating a complete inhibition of GH responses to cholinomimetic drugs if hypothalamic somatostatin is selectively depleted (Locatelli et al. 1986). The increased GH secretion seen in our depressed sample may reflect an alteration at a supra-hypothalamic level or at the hypothalamic-pituitary axis. Studies of the somatotroph axis in depression show a wide variability with two indicating that GH response to GHRH are reduced (Lesch et al. 1987; Peabody et al. 1990); one suggesting a tendency towards increased responses (Thomas et al. 1989) and one demonstrating an exaggerated response (Krishnan et al. 1988). GH secretion has been found to be reduced (Jarrett et al. 1990), normal (Rubin et al. 1990) and increased in depressed states (Mendlewitz et al. 1985). Although inconclusive these studies all conclude that the abnormal GH secretory patterns associated with depression are likely to occur primarily at a suprapituitary site. It is therefore more probable that our findings reflect an abnormality at the hypothalamic level of the somatotroph axis in the depressed group: ACh hyperresponsivity resulting in a greater decrease in somatostatin tone and a bigger surge
Pyridostigmine-induced growth hormone responses
in GH release from the anterior pituitary in this group. The mechanism of this ACh hyperresponsivity is not suggested by this challenge drug and could be caused by postsynaptic receptor alterations, either hypersensitivity or upregulation; presynaptic autoreceptor subsensitivity or by abnormalities in intracellular secondmessenger systems coupled to the ACh receptor, possibly the phosphatidylinositol cycle (Ellis & Lenox, 1990). The possible pathophysiological mechanisms of this cholinergic supersensitivity have been excellently set out by Dilsaver (1986). Our finding of ACh hyperresponsivity in association with depression is in keeping with a wealth of clinical observation: cholinomimetic drugs such as insecticides (Gershon & Shaw, 1961) and physostigmine (Risch et al. 1981) induce depressive-type symptoms; anticholinergic drugs such as antiparkinsonian agents (Crawshaw & Mullen, 1984; Pullen et al. 1984), frequently used in the management of schizophrenic patients, can cause euphoria and are liable to abuse; cannabinoids (Layman & Milton, 1971), barbiturates (Walhstrom & Norberg, 1979), opiates (Nordberg & Sundwall, 1977) and ethanol (Tabakoff et al. 1979), all drugs of abuse, interfere with the release of ACh; withdrawal of antidepressants can lead to a distressing syndrome (Kramer et al. 1961) which Dilsaver et al. (1983a; Dilsaver & Greden, 1984) has been attributed to rebound cholinergic overdrive and treated sucessfully with anticholinergic agents (Dilsaver et al. 19836). All females were tested during the first ten days of the menstrual cycle as the sex steroids oestrogen and progesterone have been shown to alter endocrine responses to a variety of stimuli (Dinan et al. 1990; O'Keane et al. 1991) and GH response to clonidine (Merimee & Fineberg, 1971) and to affect central ACh turnover in rat studies (for review see McEwan, 1988). In our own preliminary studies we found the luteal phase GH response to pyridostigmine to be about twice the follicular phase response. Pyridostigmine produced no severe sideeffects and was in general very well tolerated. Six subjects reported mild abdominal cramps and two reported muscle fasciculations. In this respect pyridostigmine has clear advantages over cholinergic challenge drugs previously used in psychiatric disorders, which required the administration of peripheral anticholinergics to prevent
59
the onset of distressing ill-effects. Also, it is administered orally in a standard dose. To summarize, we found that plasma GH responses to pyridostigmine were significantly enhanced in depressed compared to healthy volunteers and that this probably reflects a supersensivity of ACh neurotransmitter function. Future studies looking at treatment and state effects on this response would be valuable.
REFERENCES American Psychiatric Association (1987). Diagnostic and Statistical Manual of Mental Disorders, 3rd edn, revised. American Psychiatric Association: Washington, DC. Berger, M., Riemann, D., Hochli, D. & Spiegel, R. (1989). The cholinergic rapid eye movement sleep induction test with RS 86. Archives of General Psychiatry 46, 421-428. Carney, M. W., Heninger, G. R. & Steinberg, D. E. (1965). The diagnosis of depressive symptoms and the prediction of ECT response. British Journal of Psychiatry 140, 659-674. Carroll, B. J., Greden, J. & Haskett, R. (1980). Neurotransmitter studies of neuroendocrine pathology in depression: Biogenic Amines Symposium, London, England. Ada Psychiatrica Scandinavica 61 (suppl), 183-200. Casanueva, F. F., Villanueva, L., Cabranes, J. A., Cabezas, J. & Fernandez, A. (1984). Cholinergic mediation of growth hormone secretion elicited by arginine, clonidine and physical exercise in man. Journal of Clinical Endocrinology and Metabolism 59, 526-530. Casanueva, F. F., Villanueva, L., Cordido, Y., Diaz, J. A., Cabranes, R., Spiegel, R. & Dieguez, C. (1989). Differential effects of direct and indirectly acting cholinergic agonists on growth hormone release in man, and lack of effect on cortisol secretion. Journal of Neuroendocrinology 1, 379-382. Cordido, F., Casanueva, F. F. & Dieguez, C. (1989). Cholinergic receptor activation by pyridostigmine restores GH responsiveness to GH-releasing hormone administration in obese subjects: evidence for hypothalamic somalostalinergic participation in the blunled GH release of obesity. Journal of Clinical Endocrinology and Metabolism 68, 290-293. Cordido, F., Dieguez, C. & Casanueva, F. F. (1990). Effect of central cholinergic neurotransmission enhancement by pyridosligmine on the growth hormone secretion elicited by clonidine, arginine or hypoglycaemia in normal and obese subjects. Journal of Clinical Endocrinology and Metabolism 70, 1361-1370. Crawshaw, J. A. & Mullen, P. E. (1984). A study of benzhexol addiction. British Journal of Psychiatry 145, 300-303. Dilsaver, S. C. (1986). Pathophysiology of 'cholinoceptor supersensitivity ' in affective disorders. Biological Psychiatry 21,813-829. Dilsaver, S. C. & Greden, F. G. (1984). Antidepressant withdrawalinduced activation (hypomania and mania): mechanism and theoretical significance. Brain Research Reviews 7, 29-48. Dilsaver, S. C , Feinberg, M. & Greden, J. F. (1983 a). Antidepressant withdrawal symptoms treated with anticholinergic agents. American Journal of Psychiatry 140, 249-251. Dilsaver, S. C , Kronfol, Z., Greden, J. F. & Sackellares, J. C. (19836). Antidepressant withdrawal syndromes: evidence supporting the cholinergic overdrive hypothesis. Journal of Clinical Psychopharmacology 3, 157-164. Dinan, T. G., Barry, S., Yatham, L. N., Mobayed, M. & O'Hanlon, M. (1990). The reproducibility of the prolactin response to buspirone: relationship to the menstrual cycle. International Clinical Psychopharmacology 5, 119-123. Ellis, J. & Lenox, R. H. (1990). Chronic lithium treatment prevents 3-2
60
V. O 'Keane and others
atropine-induced supersensitivity of the muscarinic phosphoinositide response in rat hippocampus. Biological Psychiatry 28, 609-619. Gershon, S. & Shaw, F. H. (1961). Psychiatric sequelae of chronic exposure to organophosphorus insecticides. Lancet i, 1371-1374. Giustina, A., Bossoni, S., Cimino, A., Pizzocolo, G., Romanelli, G. & Wehrenberg, W. B. (1990). Impaired growth hormone (GH) responses to pyridostigmine in type 1 diabetic patients with exaggerated GHRH-GH-stimulated GH secretion. Journal of Clinical Endocrinology and Metabolism 71, 1486-1490. Grob, D., Harvey, A. M., Langworthy, O. R. & Lilenthal, J. L. (1947). The administration of diisopropylfluorophosphate (DFP)) to man. Johns Hopkins Hospital Bulletin 81, 257-266. Hamilton, M. (1960). A rating scale for depression. Journal of Neurology, Neurosurgery & Psychiatry 23, 56-62. Janowsky, D. S., Davis, J. M., El-Yousef, H. & Sekerke, H. J., (1972). A cholinergic-adrenegic hypothesis of mania and depression. Lancet ii, 632-635. Janowsky, D. S., Risch, S. C , Ziegler, M. G., Gillan, J. C , Huey, L. & Rausch, J. (1986). Physostigmine-induced epinephrine release in patients with affective disorder. American Journal of Psychiatry
143,919-921. Jarrett, D. B., Miewald, J. M. & Kupfer, D. J. (1990). Recurrent depression is associated with a persistent reduction in sleep-related growth hormone secretion. Archives of General Psychiatry 47, 113-118. Kramer, J. C , Klein, D. F. & Fink, M. (1961). Withdrawal symptoms following discontinuation of imipramine therapy. American Journal of Psychiatry 118, 549-550. Krishnan, K. R. R., Manepalli, A. N., Ritchie, J. C , Rayasam, K., Melville, M. L., Daughtry, G., Thorner, M. O., Rivier, J. E., Vale, W. W., Nemeroff, C. & Carroll, B. J. (1988). Growth-hormone releasing factor stimulation test in depression. American Journal of Psychiatry 145, 90-92. Layman, J. M. & Milton, A. S. (1971). Some actions of Atetrahydrocannabinol and cannabionoids at cholinergic functions. British Journal of Pharmacology 41, 379p-380p. Lesch, K. P., Laux, G., Pfuller, H., Erb, A. & Beckmann, H., (1987). Growth hormone response to GH-releasing hormone in depression. Journal of Clinical Endocrinology and Metabolism 65, 1278-1281. Locatelli, V., Torsello, A., Redaelli, E., Ghigo, E., Massara, F. & Muller, E. E. (1986). Cholinergic agonist and antagonist drugs modulate the growth hormone response to growth hormone releasing hormone in the rat: evidence for mediation by somatostatin. Journal of Endocrinology 111, 271-278. McEwan, B. S. (1988). Basic research perspective: ovarian hormone influence on brain neurochemical functions. In Contemporary Issues in Obstetrics and Gynaecology, vol. 2 (ed. L. H. Gise), pp. 21-33. Churchill Livingstone: New York. Massara, F., Ghigo, E., Molinatti, P., Mazza, E., Locatelli V. & Muller, E. E. (1986). Potentiation of cholinergic tone by pyridostigmine bromide re-instates and potentiates the growth hormone responsiveness to intermittent administration of growth hormonereleasing factor in man. Ada Endocrinologica 113, 12-16.
Mendlewicz, J., Linkowski, P., Kerkhofs, M., Desmedt, D., Golstein, J., Copinschi, G. & Van Canter, E. (1985). Diurnal hypersecretion of growth hormone in depression. Journal of Clinical Endocrinology and Metabolism 60, 505-512. Merimee, T. J. & Fineberg, S. E. (1971). Studies of the sex based variation of human growth hormone secretion. Journal of Clinical Endocrinology and Metabolism 33, 896-902. Nordberg, A. & Sundwall, A. (1977). The effect of sodium pentobarbitol on the apparent turnover of acetylcholine in different brain regions. Ada Physiological Scandinavia 99, 336 344. O'Keane, V., O'Hanlon, M., Webb, M. & Dinan, T. G. (1991). dFenfluramine/prolactin response throughout the menstrual: evidence for an oestrogen-induced alteration. Clinical Endocrinology 34, 289-292. Peabody, C. A., Warner, M. D. & Markoff, E. (1990). Growth hormone to growth releasing hormone in depression and schizophrenia. Psychiatric Research 33, 269-276. Pullen, G. P., Best, N. R. & Maguire, J. (1984). Anticholinergic drug abuse. A common problem ? British Medical Journal 289,612 613. Risch, S. C. (1982). /?-endorphin hypersecretion in depression: possible cholinergic mechanisms. Biological Psychiatry 17, 10711079. Risch, S. C , Cohen, P. M. & Janowsky, D. S. (1981). Physostigmine induction of depressive symptomatology in normal human subjects. Journal of Psychiatric Research 4, 89 94. Ross, R. J., Tsagarakis, S., Grossman, A., Nhagafoong, S., Touzel, R. J., Rees, L. H. & Besser, G. M. (1987). GH feedback occurs through modulation of hypothalamic somatostatin under cholinergic control: studies with pyridostigmine and GHRH. Clinical Endocrinology 27, 727-733. Rubin, R. T., Poland, R. E. & Lesser, I. M. (1990). Neuroendocrine aspects of primary endogenous depression X: serum growth hormone measures in patients and matched control subjects. Biological Psychiatry 27, 1065-1082. Sitaram, N., Mendelson, W. B., Wyatt, R. J. & Gillin, J. C. (1977). The time dependent induction of REM sleep and arousal by physostigmine infusion during normal human sleep. Brain Research 122, 562-567. Sitaram, N., Nurnberger, J. I., Gershon, E. S. & Gillin, J. C. (1982). Cholinergic regulation of mood and REM sleep: potential marker of vulnerability to affective disorder. American Journal of Psychiatry 139, 571-576. Tabakoff, B., Muroz-Marcus, M. & Fields, J. Z. (1979). Chronic ethanol feeding produces an increase in muscarinic cholinergic receptors in mouse brain. Life Sciences 25, 2173-2180. Thomas, R., Beer, R., Harris, B., John, R. & Scanlon, M. (1989). GH responses to growth hormone releasing factor in depression. Journal of Affective Disorders 16, 133-137. Wager, S., Robinson, D., Goetz, R., Nunes, E., Gully, R. & Quitkin, F. (1990). Cholinergic REM sleep induction in atypical depression. Biological Psychiatry 27, 441-446. Wahlstrom, G. & Nordberg, A. (1979). Sensitivity to an active synaptosomal uptake of choline in the abstinence after chronic barbital treatments. Ada Physiologica Scandinavica (suppl) 473, 165-203.
Pyridostigmine-induced growth hormone responses in healthy and depressed subjects: evidence for cholinergic supersensitivity in depression V. O'KEANE, K. O'FLYNN, J. LUCEY AND T. G. DINAN 1 From Trinity College Medical School, St James's Hospital, Dublin, Ireland
Theorists have extrapolated the cholinergic supersensitivity theory of affective disorder from a convincing and broad spectrum of clinical observation and research. This hypothesis is tested using a neuroendocrine probe approach with the challenge drug pyridostigmine, an indirect cholinergic agent thought to release growth hormone (GH) by decreasing inhibitory somatostatin tone. The consequent increments in plasma GH were considered to reflect central acetylcholine responsivity. Fifty-four volunteers were tested: 27 DSM-III-R major depressives (18 women and 9 men) and 27 age- and sex-matched healthy controls. Subjects were cannulated at 9.00 h following an overnight fast and two baseline samples were taken at 15 min intervals. Pyridostigmine 120 mg was administered orally and thereafter samples were taken at the time points +60, +90, + 120 and + 180 min. GH responses were significantly greater in depressives than controls and this effect was more marked for men than women. These results support the proposal that muscarinic upregulation and/or supersensitivity is associated with depression. SYNOPSIS
Physiological studies exploring possible cholinergic overdrive, rather than its relationship to the monoamine system, are more convincing. Arecoline, a direct cholinergic muscarinic agonist, when administered to bipolar patients in remission induces the onset of REM sleep significantly more rapidly than in normal controls suggesting a muscarinic supersensitivity in this disorder that is state independent (Sitaram et al. 1982). Another group found this abnormality was not reproducible in depressed patients who had been in remission for longer periods and has questioned its validity as a state marker (Berger et al. 1989). This abnormality in cholinergic-induced REM sleep is also found in atypical depressives without anxiety symptoms (Wager et al. 1990). Centrally acting cholinergic drugs produce sleep changes characteristic of major depression (Sitaram et al. 1977) and induce dexamethasone nonsuppression in normal volunteers (Carroll et al. 1980). One previous study has applied a neuroendocrine challenge approach. Risch (1982) found that physostigmine caused significantly greater increases in /?-endorphin in depressed compared
INTRODUCTION The dysphoric and insomnious effects of cholinomimetic agents were first noted by Grob and colleagues as far back as 1947. Later Gershon & Shaw (1961) documented 16 cases of severe psychiatric sequelae resulting from chronic exposure to these organophosphorus insecticides and in 1972 Janowsky et al. incorporated these and other clinical observations into the monoamine theory of affective disorder by proposing a cholinergic-adrenergic balance hypothesis. Briefly, this states that disturbances may arise from an imbalance between central catecholamine and acetylcholine (ACh) activity: depression arising from a relative overactivity in ACh function associated with decreases in catecholamines and mania arising from the reverse. In support of this theory they found that physostigmine (an anticholinesterase)-induced adrenaline release only tended towards blunting in depressives compared to healthy controls (Janowsky et al. 1986). ' Address for correspondence: Dr Ted Dinan, Psychiatric Unit, St James's Hospital, Dublin 8, Ireland.
55
V. O 'Keane and others
56 •
• • »
§ 20
-
O
rowth
o
O
•
10
•
0
• t
• • f
Depressed males
Healthy controls
FIG. I. A growth hormone (difference between baseline and maximum increase post-pyridostigmine administration) responses in 9 depressed and 9 healthy males.
to normal subjects. We used the cholinesterase inhibitor pyridostigmine as it is well tolerated; its mechanism of action in the release of GH has been extensively studied and it has been used in diverse endocrine studies, e.g. diabetes mellitus (Giustina et al. 1990) and obesity (Cordido et al. 1989). It seems that pyridostigmine, by stimulating ACh neurotransmission, causes a reduction in the tonic inhibitory somatostatin control of GH release and an increased secretion of GH from the somatotrophs (Ross et al. 1987). We used this challenge regime to compare ACh function in depressed and healthy states to test the hypothesis that ACh supersensitivity is associated with this disorder. METHOD Fifty-four subjects gave fully informed consent to participate in this study - 27 depressives and 27 healthy controls. The depressed group consisted of 18 women and 9 men (mean age + SEM: 35-4 + 3-4 and 35-3 + 2-9 years respectively) who fulfilled DSM-III-R (American Psychiatric Association, 1987) criteria for major depression;
unipolar type. Twenty had a past history of depression. Six were drug naive and the remainder were drug-free for a minimum of four weeks. Of the 21 who had previous exposure to psychotropics 20 had been on tricyclic antidepressants; 2 on neuroleptics and 4 on both. Fifteen patients in total had been exposed to medication with muscarinic affinity in the preceeding three months. Severity was assessed using the 17-item Hamilton Rating Scale for Depression (HAMD) (Hamilton, 1960). The mean score for males was 21-8 ±1-5 and for females was 21-5 + 0-7. The degree of endogenicity was assessed using the Newcastle Scale (Carney et al. 1965) with males yielding a mean score of 4-4 + 0-5 and women a score of 4-3 + 0-3. In addition to full physical examination all had routine haematology, biochemistry and thyroid function screening. Those aged above 40 years had an ECG and chest X-ray. Our control group, 18 men and 9 women (mean age 33-8 + 9-9 and 32-3 + 2-3 years respectively) were physically healthy and had no personal or first-degree relative history of psychiatric disorder. All volunteers weighted within normal limits for sex and age. Premenopausal female subjects were tested in the follicular phase of the menstrual cycle. Subjects presented at 9.00 h following an overnight fast. They relaxed for 15 min following insertion of a cannula and heparin bung in a forearm vein. Baseline blood samples were then taken at —15 and 0 min and thereafter at +60, + 90, +120, +180 min. Pyridostigmine 120 mg orally was administered at 0 min. Volunteers remained in a supine position and peripheral signs of muscarinic hyperactivity such as hypersalivation, sweating and lacrimation were monitored throughout the procedure. Blood samples were collected in lithium heparin bottles and analysed blind to subject status using a standard radioimmunoassay technique with a sensitivity of 0-45 mU/1, an intrassay precision of 1-6% and an interassay precision of 2-3% at about 10mU/l. Hormonal responses over time between groups were compared by two-way ANOVA and post hoc tests were applied. Delta GH (A GH: difference between baseline values and the maximum increase post-pyridostigmine administration) measures were also used to compare
Pyridostigmine-induced growth hormone responses
responses. Student's / tests and correlational analysis were used where appropriate. Results are expressed as mean + SEM.
57
15
12
RESULTS All but 1 healthy and 1 depressed subject responded to pyridostigmine with an increase in plasma GH. There was a significant sex-based GH response difference in healthy controls (mean A GH (men) = 41 + 10 mU/1; mean A GH (women) = 14-75 + 2-6 mU/1; / =2-6, df = 52, P = 0-012) and therefore results for males and females were analysed separately. There were no differences in basal GH levels between depressed and control groups, either males or females.
E
o U
0 0
60
120
180
Time (min)
50
•
FIG. 3. Plasma growth hormone levels following pyridostigmine administration at 0 min in 9 depressed male patients (O O) and 9 healthy male controls ( • • ) . Results are expressed ±SI;M.
•
40 -
• •
••
•
• •
30
•
• • ••
o .c
1 20
• i
• •
•
• ••
• • *
10 -
•
•
•
0 Depressed females
• • Healthy controls
FIG. 2 A Growth hormone (difference between baseline and maximum increase post-pyridostigmine administration) responses in 18 healthy females.
A mean A GH in depressed men of 16-1 ± 3-4 mU/1 was significantly elevated compared to a mean A GH of 41 + 1 mU/1 in healthy men (/ = 3-4, df = 16, P = 0-003) (see Fig. 1). Responses in depressed females were likewise significantly enhanced (mean A GH = 23-3 + 3 mU/1) compared to control females (mean A GH = 14-8 ±2-6 mU/1: t = 2\, df = 34, P = 004) (see Fig. 2). A repeated-measures 2-way ANOVA used to assess differences over time between depressed males and healthy controls demonstrates a significant effect for group (F(\, 88) = 9-43, P = 0002); time (F(4, 85) = 3-43, P = 001) and the group by time interaction (F(4, 85) = 2-63, P = 004). Post hoc Tukey comparisons show these responses to be significantly different at +90 (P < 0-05) and +120 min (P < 005) (see Fig. 3) A two-way ANOVA comparing hormonal responses over time between the depressed and control female group again demonstrates significant effect for group (F (1, 178) = 3-99; P = 0047), time (F(4, 175) = 7-85; P < 0-0001) and the group by time interaction (F (4, 175) = 2-6; P < 005) with post hoc Tukey comparisons showing significant effect at + 60 min (P < 0-05) only (see Fig. 4).
V. O'Keane and others
58 18
15
D 12 E
o
6
0
0
60
120
180
Time (min) FIG. 4. Plasma growth hormone levels following pyridostigmine administration at Omin in 18 depressed female patients ( O 1 - - © ) and 18 healthy female controls ( • #). Results are expressed as + S1-M.
There was no correlation between the severity of the depression as rated by the Hamilton Rating Scale and the GH response for either women (r = 0-4, df = 18, P = 009) or men (r = 0-38, df = 9, P = 0-3). Neither was there a significant correlation between scores on the Newcastle Scale (males: P = 0-34; females P = 0-16); age (males: P = 0-22; females: P = 0-56); length of the illness (males: P = 018; females: P = 0-23); presence of a positive family history (males: /> = 0-36; females: P = 0-36), weight loss (males: P = 0-12; females: P = 019) or a history of a previous episode of depression (males: P = 0-78; females: P = 0-4) and the endocrine response. There was no difference in A GH responses between those exposed to psychotropics with cholinergic affinity in the preceeding three months (N = 15) and those who were not (N = 12: / = 1-6, df = 25, P = 0\) or between the drug naive and drug-exposed group (t = 0-8, df = 25, P = 0-4). DISCUSSION This study demonstrates a significantly augmented GH response to pyridostigmine in depressed patients compared to healthy controls.
Pyridostigmine is a cholinesterase inhibitor and it has become increasingly apparent that ACh is an important neuromodulator of growth hormone (GH) release. Physiological and pharmacological stimuli of GH secretion are blocked by ACh antagonists (Casanueva et al. 1984) and indirect cholinergic agonists enhance the GH response to stimuli such as GHRH (Massara et al. 1986), clonidine and arginine (Cordido et al. 1990). Anticholinesterases alone also induces increases in GH in non-obese human subjects (Casanueva et al. 1989). GH release is regulated by a combined stimulatory GHRH and an inhibitory somatostatin input from the hypothalamus to the somatotroph cells of the anterior pituitary, with a negative feedback effect exerted by GH on its own secretion. It is suggested that GH effects this negative feedback by increasing hypothalamic somatostatin release and this somatostatin secretion is under inhibitory cholinergic control (Ross et al. 1987). Pyridostigmine impairs the GH feedback effect by decreasing somatostatin tone (Massara et al. 1986; Ross et al. 1987). That ACh effects on GH secretion are mediated through somatostatin is borne out by rat studies demonstrating a complete inhibition of GH responses to cholinomimetic drugs if hypothalamic somatostatin is selectively depleted (Locatelli et al. 1986). The increased GH secretion seen in our depressed sample may reflect an alteration at a supra-hypothalamic level or at the hypothalamic-pituitary axis. Studies of the somatotroph axis in depression show a wide variability with two indicating that GH response to GHRH are reduced (Lesch et al. 1987; Peabody et al. 1990); one suggesting a tendency towards increased responses (Thomas et al. 1989) and one demonstrating an exaggerated response (Krishnan et al. 1988). GH secretion has been found to be reduced (Jarrett et al. 1990), normal (Rubin et al. 1990) and increased in depressed states (Mendlewitz et al. 1985). Although inconclusive these studies all conclude that the abnormal GH secretory patterns associated with depression are likely to occur primarily at a suprapituitary site. It is therefore more probable that our findings reflect an abnormality at the hypothalamic level of the somatotroph axis in the depressed group: ACh hyperresponsivity resulting in a greater decrease in somatostatin tone and a bigger surge
Pyridostigmine-induced growth hormone responses
in GH release from the anterior pituitary in this group. The mechanism of this ACh hyperresponsivity is not suggested by this challenge drug and could be caused by postsynaptic receptor alterations, either hypersensitivity or upregulation; presynaptic autoreceptor subsensitivity or by abnormalities in intracellular secondmessenger systems coupled to the ACh receptor, possibly the phosphatidylinositol cycle (Ellis & Lenox, 1990). The possible pathophysiological mechanisms of this cholinergic supersensitivity have been excellently set out by Dilsaver (1986). Our finding of ACh hyperresponsivity in association with depression is in keeping with a wealth of clinical observation: cholinomimetic drugs such as insecticides (Gershon & Shaw, 1961) and physostigmine (Risch et al. 1981) induce depressive-type symptoms; anticholinergic drugs such as antiparkinsonian agents (Crawshaw & Mullen, 1984; Pullen et al. 1984), frequently used in the management of schizophrenic patients, can cause euphoria and are liable to abuse; cannabinoids (Layman & Milton, 1971), barbiturates (Walhstrom & Norberg, 1979), opiates (Nordberg & Sundwall, 1977) and ethanol (Tabakoff et al. 1979), all drugs of abuse, interfere with the release of ACh; withdrawal of antidepressants can lead to a distressing syndrome (Kramer et al. 1961) which Dilsaver et al. (1983a; Dilsaver & Greden, 1984) has been attributed to rebound cholinergic overdrive and treated sucessfully with anticholinergic agents (Dilsaver et al. 19836). All females were tested during the first ten days of the menstrual cycle as the sex steroids oestrogen and progesterone have been shown to alter endocrine responses to a variety of stimuli (Dinan et al. 1990; O'Keane et al. 1991) and GH response to clonidine (Merimee & Fineberg, 1971) and to affect central ACh turnover in rat studies (for review see McEwan, 1988). In our own preliminary studies we found the luteal phase GH response to pyridostigmine to be about twice the follicular phase response. Pyridostigmine produced no severe sideeffects and was in general very well tolerated. Six subjects reported mild abdominal cramps and two reported muscle fasciculations. In this respect pyridostigmine has clear advantages over cholinergic challenge drugs previously used in psychiatric disorders, which required the administration of peripheral anticholinergics to prevent
59
the onset of distressing ill-effects. Also, it is administered orally in a standard dose. To summarize, we found that plasma GH responses to pyridostigmine were significantly enhanced in depressed compared to healthy volunteers and that this probably reflects a supersensivity of ACh neurotransmitter function. Future studies looking at treatment and state effects on this response would be valuable.
REFERENCES American Psychiatric Association (1987). Diagnostic and Statistical Manual of Mental Disorders, 3rd edn, revised. American Psychiatric Association: Washington, DC. Berger, M., Riemann, D., Hochli, D. & Spiegel, R. (1989). The cholinergic rapid eye movement sleep induction test with RS 86. Archives of General Psychiatry 46, 421-428. Carney, M. W., Heninger, G. R. & Steinberg, D. E. (1965). The diagnosis of depressive symptoms and the prediction of ECT response. British Journal of Psychiatry 140, 659-674. Carroll, B. J., Greden, J. & Haskett, R. (1980). Neurotransmitter studies of neuroendocrine pathology in depression: Biogenic Amines Symposium, London, England. Ada Psychiatrica Scandinavica 61 (suppl), 183-200. Casanueva, F. F., Villanueva, L., Cabranes, J. A., Cabezas, J. & Fernandez, A. (1984). Cholinergic mediation of growth hormone secretion elicited by arginine, clonidine and physical exercise in man. Journal of Clinical Endocrinology and Metabolism 59, 526-530. Casanueva, F. F., Villanueva, L., Cordido, Y., Diaz, J. A., Cabranes, R., Spiegel, R. & Dieguez, C. (1989). Differential effects of direct and indirectly acting cholinergic agonists on growth hormone release in man, and lack of effect on cortisol secretion. Journal of Neuroendocrinology 1, 379-382. Cordido, F., Casanueva, F. F. & Dieguez, C. (1989). Cholinergic receptor activation by pyridostigmine restores GH responsiveness to GH-releasing hormone administration in obese subjects: evidence for hypothalamic somalostalinergic participation in the blunled GH release of obesity. Journal of Clinical Endocrinology and Metabolism 68, 290-293. Cordido, F., Dieguez, C. & Casanueva, F. F. (1990). Effect of central cholinergic neurotransmission enhancement by pyridosligmine on the growth hormone secretion elicited by clonidine, arginine or hypoglycaemia in normal and obese subjects. Journal of Clinical Endocrinology and Metabolism 70, 1361-1370. Crawshaw, J. A. & Mullen, P. E. (1984). A study of benzhexol addiction. British Journal of Psychiatry 145, 300-303. Dilsaver, S. C. (1986). Pathophysiology of 'cholinoceptor supersensitivity ' in affective disorders. Biological Psychiatry 21,813-829. Dilsaver, S. C. & Greden, F. G. (1984). Antidepressant withdrawalinduced activation (hypomania and mania): mechanism and theoretical significance. Brain Research Reviews 7, 29-48. Dilsaver, S. C , Feinberg, M. & Greden, J. F. (1983 a). Antidepressant withdrawal symptoms treated with anticholinergic agents. American Journal of Psychiatry 140, 249-251. Dilsaver, S. C , Kronfol, Z., Greden, J. F. & Sackellares, J. C. (19836). Antidepressant withdrawal syndromes: evidence supporting the cholinergic overdrive hypothesis. Journal of Clinical Psychopharmacology 3, 157-164. Dinan, T. G., Barry, S., Yatham, L. N., Mobayed, M. & O'Hanlon, M. (1990). The reproducibility of the prolactin response to buspirone: relationship to the menstrual cycle. International Clinical Psychopharmacology 5, 119-123. Ellis, J. & Lenox, R. H. (1990). Chronic lithium treatment prevents 3-2
60
V. O 'Keane and others
atropine-induced supersensitivity of the muscarinic phosphoinositide response in rat hippocampus. Biological Psychiatry 28, 609-619. Gershon, S. & Shaw, F. H. (1961). Psychiatric sequelae of chronic exposure to organophosphorus insecticides. Lancet i, 1371-1374. Giustina, A., Bossoni, S., Cimino, A., Pizzocolo, G., Romanelli, G. & Wehrenberg, W. B. (1990). Impaired growth hormone (GH) responses to pyridostigmine in type 1 diabetic patients with exaggerated GHRH-GH-stimulated GH secretion. Journal of Clinical Endocrinology and Metabolism 71, 1486-1490. Grob, D., Harvey, A. M., Langworthy, O. R. & Lilenthal, J. L. (1947). The administration of diisopropylfluorophosphate (DFP)) to man. Johns Hopkins Hospital Bulletin 81, 257-266. Hamilton, M. (1960). A rating scale for depression. Journal of Neurology, Neurosurgery & Psychiatry 23, 56-62. Janowsky, D. S., Davis, J. M., El-Yousef, H. & Sekerke, H. J., (1972). A cholinergic-adrenegic hypothesis of mania and depression. Lancet ii, 632-635. Janowsky, D. S., Risch, S. C , Ziegler, M. G., Gillan, J. C , Huey, L. & Rausch, J. (1986). Physostigmine-induced epinephrine release in patients with affective disorder. American Journal of Psychiatry
143,919-921. Jarrett, D. B., Miewald, J. M. & Kupfer, D. J. (1990). Recurrent depression is associated with a persistent reduction in sleep-related growth hormone secretion. Archives of General Psychiatry 47, 113-118. Kramer, J. C , Klein, D. F. & Fink, M. (1961). Withdrawal symptoms following discontinuation of imipramine therapy. American Journal of Psychiatry 118, 549-550. Krishnan, K. R. R., Manepalli, A. N., Ritchie, J. C , Rayasam, K., Melville, M. L., Daughtry, G., Thorner, M. O., Rivier, J. E., Vale, W. W., Nemeroff, C. & Carroll, B. J. (1988). Growth-hormone releasing factor stimulation test in depression. American Journal of Psychiatry 145, 90-92. Layman, J. M. & Milton, A. S. (1971). Some actions of Atetrahydrocannabinol and cannabionoids at cholinergic functions. British Journal of Pharmacology 41, 379p-380p. Lesch, K. P., Laux, G., Pfuller, H., Erb, A. & Beckmann, H., (1987). Growth hormone response to GH-releasing hormone in depression. Journal of Clinical Endocrinology and Metabolism 65, 1278-1281. Locatelli, V., Torsello, A., Redaelli, E., Ghigo, E., Massara, F. & Muller, E. E. (1986). Cholinergic agonist and antagonist drugs modulate the growth hormone response to growth hormone releasing hormone in the rat: evidence for mediation by somatostatin. Journal of Endocrinology 111, 271-278. McEwan, B. S. (1988). Basic research perspective: ovarian hormone influence on brain neurochemical functions. In Contemporary Issues in Obstetrics and Gynaecology, vol. 2 (ed. L. H. Gise), pp. 21-33. Churchill Livingstone: New York. Massara, F., Ghigo, E., Molinatti, P., Mazza, E., Locatelli V. & Muller, E. E. (1986). Potentiation of cholinergic tone by pyridostigmine bromide re-instates and potentiates the growth hormone responsiveness to intermittent administration of growth hormonereleasing factor in man. Ada Endocrinologica 113, 12-16.
Mendlewicz, J., Linkowski, P., Kerkhofs, M., Desmedt, D., Golstein, J., Copinschi, G. & Van Canter, E. (1985). Diurnal hypersecretion of growth hormone in depression. Journal of Clinical Endocrinology and Metabolism 60, 505-512. Merimee, T. J. & Fineberg, S. E. (1971). Studies of the sex based variation of human growth hormone secretion. Journal of Clinical Endocrinology and Metabolism 33, 896-902. Nordberg, A. & Sundwall, A. (1977). The effect of sodium pentobarbitol on the apparent turnover of acetylcholine in different brain regions. Ada Physiological Scandinavia 99, 336 344. O'Keane, V., O'Hanlon, M., Webb, M. & Dinan, T. G. (1991). dFenfluramine/prolactin response throughout the menstrual: evidence for an oestrogen-induced alteration. Clinical Endocrinology 34, 289-292. Peabody, C. A., Warner, M. D. & Markoff, E. (1990). Growth hormone to growth releasing hormone in depression and schizophrenia. Psychiatric Research 33, 269-276. Pullen, G. P., Best, N. R. & Maguire, J. (1984). Anticholinergic drug abuse. A common problem ? British Medical Journal 289,612 613. Risch, S. C. (1982). /?-endorphin hypersecretion in depression: possible cholinergic mechanisms. Biological Psychiatry 17, 10711079. Risch, S. C , Cohen, P. M. & Janowsky, D. S. (1981). Physostigmine induction of depressive symptomatology in normal human subjects. Journal of Psychiatric Research 4, 89 94. Ross, R. J., Tsagarakis, S., Grossman, A., Nhagafoong, S., Touzel, R. J., Rees, L. H. & Besser, G. M. (1987). GH feedback occurs through modulation of hypothalamic somatostatin under cholinergic control: studies with pyridostigmine and GHRH. Clinical Endocrinology 27, 727-733. Rubin, R. T., Poland, R. E. & Lesser, I. M. (1990). Neuroendocrine aspects of primary endogenous depression X: serum growth hormone measures in patients and matched control subjects. Biological Psychiatry 27, 1065-1082. Sitaram, N., Mendelson, W. B., Wyatt, R. J. & Gillin, J. C. (1977). The time dependent induction of REM sleep and arousal by physostigmine infusion during normal human sleep. Brain Research 122, 562-567. Sitaram, N., Nurnberger, J. I., Gershon, E. S. & Gillin, J. C. (1982). Cholinergic regulation of mood and REM sleep: potential marker of vulnerability to affective disorder. American Journal of Psychiatry 139, 571-576. Tabakoff, B., Muroz-Marcus, M. & Fields, J. Z. (1979). Chronic ethanol feeding produces an increase in muscarinic cholinergic receptors in mouse brain. Life Sciences 25, 2173-2180. Thomas, R., Beer, R., Harris, B., John, R. & Scanlon, M. (1989). GH responses to growth hormone releasing factor in depression. Journal of Affective Disorders 16, 133-137. Wager, S., Robinson, D., Goetz, R., Nunes, E., Gully, R. & Quitkin, F. (1990). Cholinergic REM sleep induction in atypical depression. Biological Psychiatry 27, 441-446. Wahlstrom, G. & Nordberg, A. (1979). Sensitivity to an active synaptosomal uptake of choline in the abstinence after chronic barbital treatments. Ada Physiologica Scandinavica (suppl) 473, 165-203.
