Psychoneuroendocrinology,Vol. 17, No. 2/3, pp. 243-248,

1992

0306-4530/92 $5.00+0.00 ©1992 Pergamon Press Ltd.

Printed in Great Britain

SHORT C O M M U N I C A T I O N MELATONIN, CORTISOL A N D PROLACTIN RESPONSE TO ACUTE NOCTURNAL LIGHT EXPOSURE IN HEALTHY VOLUNTEERS IAIN M. MCINTYRE, 1,3 TREVOR R. NORMAN, 2 GRAHAM D. BURROWS,2 and STUART M. ARMSTRONG3 1Victorian Institute of Forensic Pathology, South Melbourne, 2Department of Psychiatry, University of Melbourne, Austin Hospital, Heidelberg, and 3Department of Psychology and Brain Behaviour Research Institute, La Trobe University, Bundoora, Victoria, Australia (Received 31 May 1990; infinal form 4 May 1991)

SUMMARY An investigation of the cortisol and prolactin responses accompanying acute melatonin suppression by light (600 lux) in humans is described. Light given from midnight to 0300h suppressed nocturnal plasma melatonin concentrations by 65%. Despite this significant suppression of melatonin, no significant effect on plasma cortisol or prolactin concentrations was observed. These data support recent studies which argue that, if there is a relationship between melatonin, the hypothalamo-pituitary, and the hypothalamo-pituitary-adrenal axis in humans, it is neither direct nor simple. INTRODUCTION ALTHOUGHTHE PHYSIOLOGICALFUNCTIONof the pineal gland is largely unknown, accumulated data suggest that it m a y be an important regulator of adrenal function (cf. Johnson, 1982; Vaughan, 1984; Touitou, 1989, for reviews). Despite evidence that pineal hormones have an ability to act at almost every site of the adrenal axis, much research is needed to define the precise mechanism of action. A substantial amount of work has involved animal and in-vitro incubation studies, and investigations in humans are few. An interaction between the human pineal gland hormone, melatonin, and the hypothalamopituitary-adrenal (HPA) axis has been proposed (Wetterberg et al., 1981; Beck-Friis et al., 1985). Melatonin is suggested to be an inhibiting factor of corticotrophin releasing factor (CRF), so that when melatonin levels are reduced, the lack of melatonin produces disinhibition of the HPA axis and an increased cortisol secretion as a consequence of CRF hypersecretion. HPA axis abnormalities often have been observed in patients with depressive illness (Sachar et al., 1973; Carroll et al., 1981; Beck-Friis et al., 1985). Address correspondence and reprint requests to: Dr. Iain M. Mclntyre, Victorian Institute of Forensic Pathology, 57-83 Kavanagh Street, South Melbourne, Victoria 3205, AUSTRALIA. 243

244

I.M. MclNI~'REet al.

Other studies, however, have produced conflicting results. Some authors have administered exogenous melatonin and failed to show any effect on cortisol (Waldhauser et al., 1987; Nickelsen et al., 1989). Furthermore, no alteration of plasma melatonin was found in a bilaterally adrenalectomized patient, with or without glucocorticoid replacement (Vaughan et al., 1979), or in children with congenital adrenal hyperplasia due to 21-hydroxylase deficiency (Waldhauser et al., 1986). Unlike the association between altered melatonin levels and cortisol concentrations found in depressive illness (Beck-Friis et al., 1985), Hariharasubramanian et al. (1985) could not find significant changes in plasma cortisol in women during the menstrual cycle, even though melatonin was significantly higher premenstrually (days 2 5 - 3 0 ) compared to the mid-menstrual period (days 13-17). Nevertheless, in patients with Cushing's syndrome, an inverse relationship between plasma concentrations of cortisol and melatonin has been reported (Werner et al., 1980; Wetterberg et al., 1980). During acute inhibition of cortisol production induced by metyrapone in patients with pituitary disease and in control subjects, melatonin excretion was increased (Brismar et al., 1982), while high ACTH and cortisol levels were accompanied by reduced melatonin levels (Brismar et al., 1985). An approach to assess directly the effect of melatonin on the HPA axis without possible interference by drugs is to use artificial light to suppress acutely human nocturnal melatonln production. Because a brief exposure to bright artificial light is known to suppress human melatonin significantly (Lewy et al., 1980; Mclntyre et al., 1989b), any inhibitory effect on CRF would be reduced and an increased cortisol secretion should be evident. The present investigation therefore was undertaken to determine the effect of acute melatonin suppression by light (midnight-0300h) on cortisol in normal healthy volunteers. In addition, since there is an interrelationship between the pineal, melatonin and prolactin in humans, hamsters and rats (Petterborg et al., 1984; Chik et al., 1985; Blask et al., 1986; Esquifino et al., 1989) and since excess plasma was available after our cortisol and melatonin assays, the effect of light on prolactin levels was assessed. SUBJECTS AND METHODS Details of the procedure have been described previously (Mclntyre et al., 1989a). In brief, six normal healthy subjects (mean age = 32.3 + 6.2 yr; four men, two women) without a family history of psychiatric disorders were examined. Plasma melatonin, cortisol and prolactin concentrations were determined on an hourly basis (from 2300h-0500h) on a night where light of 600 lux intensity was given for 3 hr duration from midnight to 0300h, and on a control night where the background light intensity was less than 10 lux throughout. Control and experimental nights were separated by 1 wk, and subjects were counterbalanced so that half were exposed to light on the first occasion and half on the second. Blood was collected via an indwelling needle with a heparin lock system. Subjects were encouraged to remain awake throughout the period of investigation by talking and watching television. Artificial light was produced with eight fluorescent tubes (Vita-Lite, Interlight Tru-Lite Australia Pty.) as previously described (Mclntyre et al., 1989a). The desired intensity was achieved by sitting subjects at the appropriate distance from the light source. Light intensities were measured by a Topcon IM-3 digital light meter at eye level. The study was performed in Melbourne between May and June (late autumn-early winter), at which time sunset was about 1710h and sunrise about 0730h. Plasma melatonin concentrations were measured by radioimmunoassay (Fraser et al., 1983; Mclntyre et al., 1986). Plasma cortisol and prolactin concentrations were determined by commercially available radioimmunoassay kits (Amerlex, Amersham, U.K.).

MELATONIN, CORTISOL AND PROLACTIN RESPONSE TO NOCTURNAL LIGHT

245

RESULTS

The mean response of nocturnal melatonin to light of 600 lux intensity is shown in Fig. 1A. On the control night, concentrations of melatonin remained relatively stable throughout the period of investigation, with a maximum observed at 0200h. On the night where light was given from midnight to 0300h, a significant suppression of melatonin was observed (p0.05; Mann-Whitney U-test). A similar pattem occurred for the prolactin concentrations, except that levels on the control and light-exposure nights overlapped, so that no formal statistical testing was deemed necessary (Fig. 1C). DISCUSSION The majority of data from animal studies support an inhibitory effect of melatonin on the HPA axis (e.g., Johnson, 1982). However, interpretation of data from animal studies in this area is difficult, which makes generalization to humans unreliable. There must be consideration of the species used and the experimental design, particularly with regard to acute versus chronic studies and the time of day that studies are undertaken. Research examining the relationship between melatonin and the HPA axis in humans has produced conflicting data. Although an inhibitory effect of melatonin on cortisol secretion through inhibition of CRF has been proposed (Wetterberg et al., 1981; Beck-Friis et al., 1985), overnight or stress-induced studies attempting to correlate cortisol and melatonin generally have not demonstrated a relationship (Vaughan et al., 1978; 1979; Rao & Mager, 1987; Salin-Pascual et al., 1988). Two investigations utilizing light suppression of nocturnal melatonin also failed to find a change in cortisol secretion (Krieger et al., 1971; Strassman et al., 1988). Both these studies, however, employed constant light-exposure regimens. Strassman et al. (1988), for example, "functionally pinealectomized" their subjects by exposure to light of 3,000 lux intensity overnight from 2200h. In the present investigation, we used an acute light exposure which suppressed melatonin by 65% for only 3 hr beginning at midnight. It was hoped that this procedure would make apparent any acute effects of melatonin inhibition on cortisol levels. The nocturnal rhythm of cortisol in humans is well-established, with increased secretion generally beginning between 0100h and 0300h (Sachar et al., 1973). If the lack of melatonin is capable of disinhibiting the HPA axis, as previously suggested, an investigation at a time near the onset of cortisol secretion may be expected to produce earlier and/or greater cortisol secretory episode(s). In other words, by suppressing melatonin and thereby removing the inhibition on the HPA axis, cortisol secretion at a time sensitive to its secretory stimulus should be increased. This was not observed. Our data, therefore, support a number of human studies which show no simple, direct relationship between melatonin and the HPA axis (e.g., Touitou, 1989). On the basis of our study, a similar conclusion must be drawn for melatonin and prolactin. The use of acute melatonin suppression, as in the present study, does not necessarily

246

I.M. MclNTYREet al.

A 5O

4O

30 t-"E"

o

• ""

I

",

'-

::

Ii

20

10

+:~

I

i

i

I

,L"

I

• *

I i

i

I

'

B

1614 2

l

i

I

,y

20 18' •~.

i

I I

22

~

I

~°°°°°°

,~ 12-" 'E O L)

8-

6-

i

I

'

I

i

i

C 11 10 9

"%

I

i

I

i..I

E

vee--

g_

7 6 5 4 3 2 1 0

Light I

2300

j

~ :

I l l l i

2400

- i - control Night - 0 - Light Night i

0100

0200

Time

(hours)

I

0300 0400

l

I

0500

Melatonin {A), cortisol (B) and prolactin (C) responses on a control night (filled squares) and on a night of exposure to light of 600 lux intensity between midnight and 0300h (open circles) in six healthy volunteers, *p