Melatonin Suppression by Light in Euthymic Bipolar and Unipolar Patients

ORIGINAL ARTICLE Melatonin Suppression by Light in Euthymic Bipolar and Unipolar Patients

John I. Nurnberger, Jr, MD, PhD; Sherril Adkins, RN; Debomoy K. Lahiri, PhD; Aimee Mayeda, MD; Kuolung Hu, MS; Alfred Lewy, MD, PhD; Aaron Miller, MD; Elizabeth S. Bowman, MD; Marvin J. Miller, MD; N. Leela Rau, MD; Carrie Smiley, RN; Dawn Davis-Singh, BS

Background: Previous studies have suggested that bi- Results: No group differences were seen in light sup- polar patients are supersensitive to light suppression of pression among bipolar patients, unipolar patients, and melatonin and that this may be a trait marker for ge- controls; an analysis of the whole group did not reveal netic vulnerability. The present study was an attempt to differences in propranolol effect, differences in cortico- replicate and extend this observation. Propranolol hy- tropin or cortisol levels, or evidence for a lithium effect. drochloride effects were compared with light effects be- However, patients with bipolar I affective disorder showed cause of the documented influence of ␤-adrenergic re- the following: (1) significantly lower melatonin levels on ceptors on melatonin production. Nighttime levels of the light night, at baseline and following light exposure; corticotropin and cortisol were also examined as poten- and (2) a later peak time for melatonin on the dark night. tial trait vulnerability markers. Conclusions: The general hypothesis of increased light Methods: Melatonin levels in euthymic bipolar pa- sensitivity in bipolar patients was not supported. How- tients (n=29) were tested before and after 500-lux light ever, melatonin secretion abnormalities were con- was administered between 2 and 4 AM and on a separate firmed in the subgroup with bipolar I disorder. Further night in the dark. Results were compared with those of assessments of circadian rhythm disruption as a vulner- a group of patients with unipolar depression (n=24) and ability marker in bipolar illness are indicated. with those of a group of non–psychiatrically ill control subjects (n=50). Lithium effects and propranolol effects were tested in subgroups. Arch Gen Psychiatry. 2000;57:572-579

IPOLAR AFFECTIVE disorder Bright light can shift the phase of involves disruption of in- melatonin secretion in humans.24 The pho- trinsic biological rhythms, toperiodic control of melatonin makes it including the sleep-wake a potential marker of circadian rhythms. cycle, hormonal rhythms, Lewy25 has suggested use of the dim-light Band temperature regulation. A phase ad- melatonin onset as a marker of circadian vance in rhythms explains some of these rhythm phase and period. observations; a general phase instability has Melatonin is an indolamine derivative From the Department of also been hypothesized.1-7 of serotonin and is produced by the action Psychiatry, Indiana University The hormone melatonin has a time- of the enzymes N-acetyltransferase and School of Medicine keeping function in many mammals and hydroxyindole-O-methyltransferase,withthe (Drs Nurnberger, Lahiri, Mayeda, A. Miller, and appears to adjust the timing of circadian first step being rate limiting. Production is Bowman and Mss Adkins, Hu, rhythm information transmitted from the primarily within the pineal body; the hor- Smiley, and Davis-Singh), suprachiasmatic nucleus of the hypothala- mone is secreted into the blood and into the Department of Psychiatry, mus to entrain physiologic rhythms.8-16 A cerebrospinal fluid.23(pp40-66) Specific recep- Roudebush Veterans Affairs physiologic role for melatonin in hu- tors for melatonin are found in multiple Medical Center (Dr Mayeda), mans has yet to be clearly established. brain areas, including the hypothalamus, the Institute of Psychiatric Some apparently normal humans have low the cerebellum, and the pineal body itself. Research, Larue Carter to undetectable levels of melatonin.17 Pa- Melatonin is secreted into plasma at high Hospital (Dr M. J. Miller), and tients undergoing long-term treatment levels (60-100 pg/mL) at night and at low Department of Psychiatry, with ␤-blockers show suppressed mela- levels (3-10 pg/mL) during the daytime. Wishard Hospital (Dr Rau), 18,19 Indianapolis, Ind; and the tonin secretion without ill effects. Mela- The anatomic pathway mediating Department of Psychiatry, tonin levels decrease with chronological melatonin secretion is complex. Retinal 20,21 Oregon Health Sciences age. Administered melatonin is known stimulation by light produces a tonic in- University, Portland to lower body temperature at night22 and hibitory signal to the suprachiasmatic (Dr Lewy). may induce sleepiness.23(pp192-208) nucleus of the hypothalamus via the reti-

(REPRINTED) ARCH GEN PSYCHIATRY/ VOL 57, JUNE 2000 WWW.ARCHGENPSYCHIATRY.COM 572

©2000 American Medical Association. All rights reserved. Downloaded From: https://jamanetwork.com/ on 09/25/2021 SUBJECTS AND METHODS not be pregnant (determined by a urine screen). Premeno- pausal women were not tested during their menstrual period. Subjects were stratified by age, race, sex, and season of SUBJECTS testing. Each bipolar and unipolar patient was matched with a control of the same sex, race, age (within 5 years), and sea- Subjects consisted of 29 patients with bipolar I (n=21) or bi- son of testing (within 6 weeks). Women were also matched polar II (n=8) affective disorder, 24 patients with unipolar for menstrual status (premenopausal or postmenopausal). The affective disorder, and 50 controls. Patients were recruited clinical characteristics of patients and matched controls are from an Indiana University, Indianapolis, outpatient clinic, summarized in Table 1 and Table 2. Three additional un- and additional patients and controls were recruited by ad- matched controls were included in the total control sample. vertisement (primarily on campus). A Structured Clinical In- terview for DSM-III-R was administered by a nurse-clinician CLINICAL PROCEDURE (S.A.). Two psychiatrists ( J.I.N., M.J.M., E.S.B., N.L.R. and A.M.) then made the diagnosis in patients who met the cri- Subjects sitting up in bed and looking at a 500-lux light po- teria by DSM-III-R40 and by modified Research Diagnostic Cri- sitioned at the foot of the bed can vary the amount of light teria41 for the disorder indicated. Patients with bipolar II they receive by 20% by leaning forward or backward ( J.I.N., disorder (bipolar disorder, not otherwise specified by unpublished data, 1986). Nurnberger et al35 reduced this varia- DSM-III-R) and unipolar patients had recurrent episodes of tion by placing a seated subject in the center of an arc of lights. depression. Controls had no ongoing psychiatric illness and In the present study, we further modified the procedure by no history of major affective disorder, psychosis, or any other placing a bank of full-spectrum shielded fluorescent lights psychiatric impairment by Structured Clinical Interview for behind the subject in a small room, such that light was re- DSM-III-R. All subjects gave informed consent in writing. flected from the wall, floor, and ceiling. The participant was Patients were tested in the euthymic state, ie, function- thus exposed to a dispersed source of light, the intensity of ing at their usual interepisode level for several months and which was calibrated by handheld photometer to be 500 lux, scoring less than 3 on the Raskin Mood Scale.42 Twenty-two ±50 lux. Subjects were carefully monitored to ensure that they bipolar patients and 24 unipolar patients were tested after re- were awake, with eyes open, gazing in an appropriate direc- maining free from psychoactive medication for at least 2 weeks tion, during the 2-hour period of the testing. (5 weeks for fluoxetine and 3 weeks for carbamazepine and Subjects arrived at the Indiana University General Clini- for antidepressants other than fluoxetine). No patient took cal Research Center, Indianapolis, on the evening of test- other medications known to affect melatonin levels. Alco- ing, and retired between 11 PM and midnight. For the dark hol was not permitted for 3 days before the test. A urine drug night condition, subjects were allowed to sleep through the and alcohol screen was performed on the day of testing. Smok- night; blood samples were drawn through an indwelling cath- ing and eating were not permitted after admission to the re- eter with the aid of a small flashlight with a red filter. For search unit. A subgroup of bipolar patients was tested on their the light night condition, subjects were awakened at 2 AM ongoing therapeutic regimen of lithium (n=20), including and asked to sit in a cardiac chair between 2 and 4 AM. Blood 13 patients who were also tested while not taking lithium and was drawn at 1, 1:30, 2, 2:20, 2:40, 3, 3:20, 3:40, 4, 4:30, 5, 7 additional patients tested while taking lithium only. A sub- and 6 AM. After the 4 AM sample, the subject was allowed to group of bipolar patients not taking lithium (n=13) and con- return to bed; thus, the first 3 samples and the last 3 samples trols (n=41) were tested with a single 10-mg oral dose of were drawn from subjects recumbent in a dark room. Pilot propranolol or placebo given at 11 PM on the night of test- studies43 showed that posture and sleep per se did not sig- ing. Twenty-two controls were also retested on a second oc- nificantly affect melatonin values. Each subject was invited casion to provide a measure of reliability. to participate in a dark night and a light night study; study Female subjects were required to be using a reliable method of birth control (but not oral contraceptives) and to Continued on next page

nohypothalamic tract. The signal in the suprachiasmatic ported that light suppression of melatonin (percentage re- nucleus of the hypothalamus appears to be mediated by duction in melatonin following light exposure) was greater glutamate, although prior studies26-28 implicated acetyl- in bipolar patients than in control subjects. Lewy et al34 choline. From the suprachiasmatic nucleus of the hypo- tested the effects of 500-lux light on melatonin secretion thalamus, processes travel to the paraventricular nucleus between 2 and 4 AM in 35 subjects. Eleven euthymic bi- of the hypothalamus, then, via the median forebrain bundle, polar patients not taking medication were found to have to the intermediolateral cell column of the upper thoracic 61.5%±1.6% suppression; 24 age- and sex-matched con- spinal cord, and then to the superior cervical ganglion. From trols were found to have 28.0%±5.8% suppression. Nurn- the superior cervical ganglion, sympathetic fibers inner- berger and colleagues35 found increased melatonin sup- vate the pineal body through ␣- and ␤-noradrenergic re- pression in the offspring of bipolar patients compared with ceptors. ␥-Aminobutyric acid and peptide neurotransmit- the offspring of psychiatrically unaffected controls. Thus, ters may also be involved.28,29 melatonin suppression by light is a candidate genetic vul- Wetterberg and colleagues23(pp69-95),30,31 described a nerability marker for bipolar illness. Two experiments by subgroup of depressed patients with persistent low mela- other groups36,37 were nonconfirmatory of the hypothesis tonin production; in their studies, a low melatonin level of increased light sensitivity in bipolar patients. Lam et al36 appears to be related to high cortisol secretion. reported a decreased baseline melatonin level in 8 acutely Light of 500 lux or greater usually suppresses mela- ill bipolar patients but reduced light suppression com- tonin secretion in humans.32 Lewy and collaborators33 re- pared with controls. Whalley et al37 reported no differ-

(REPRINTED) ARCH GEN PSYCHIATRY/ VOL 57, JUNE 2000 WWW.ARCHGENPSYCHIATRY.COM 573

©2000 American Medical Association. All rights reserved. Downloaded From: https://jamanetwork.com/ on 09/25/2021 nights were separated by at least 3 days. The order of study zero with a 95% confidence limit was 0.22 pmol/L. The es- nights was randomized. Sleep deprivation did not gener- timated concentration of melatonin at 20% binding was ally affect mood in these euthymic patients. 105.7 pg/mL; the estimated concentration of melatonin at 50% binding, 281.8 pg/mL; and the estimated concentra- HORMONE ASSAYS tion of melatonin at 80% binding, 556.6 pg/mL of plasma. The intra-assay and interassay coefficients of variation were A radioimmunoassay (RIA) for melatonin was developed fol- 4.67% and 8.48%, respectively. lowing previously published methods.44-46 Samples of plasma were drawn into heparinized tubes, held on ice before cen- DATA ANALYSIS trifugation (Յ2 hours), and stored at −70°C until testing. For primary antibody, the sheep antimelatonin antiserum Melatonin suppression was calculated as in Lewy et al34: was obtained from Guildhay Antisera Co, Surrey, England, 100%−[(average melatonin level between 3 and 4 AM/ and supplied as a freeze-dried sample. The antiserum was average melatonin level between 1 and 2 AM)ϫ100%]. A developed against N-acetyl-5-methoxytryptophan or bo- dark-adjusted melatonin suppression index was then cal- vine thyroglobulin. Cross-reactivity with N-acetyltrypt- culated by subtracting the value of the suppression mea- amine, 6-hydroxy-melatonin, and N-acetyltryptophan was sure on the dark night from that on the light night. Sup- less than 1%. The assay was linear to 60 pg/mL and sensi- pression was compared between patient groups and their tive from 5 to 500 pg/mL. Intra-assay variability, based on matched controls using matched-pair 2-tailed t tests. The duplicate measurement of 28 samples, was 2.14%. Interas- power was 93% to detect an effect of size 0.8 and 63% to say variability averaged 7.3% (n=28). The correlation of val- detect an effect of size 0.5. Each group of patients was also ues between the RIA and a chromatographic–mass spectro- compared with the entire control group using indepen- photometric assay47 was 0.92 (n=14). dent 2-tailed t tests and Wilcoxon 2-sample tests. The ␣ Cortisol was assayed using a double-antibody iodine level of significance was .05. 125 (125I) RIA (Diagnostic Products Corporation, Los An- A 3-way analysis of variance was applied to test the geles, Calif ). Intra-assay variability was 2.5% when the value effects of diagnosis, treatment (light or dark), and drug (pro- was between 140 and 1400 nmol/L. Interassay variability pranolol or no drug), with 3 to 4 AM average melatonin level was 6.3%, and the detection limit was 5.6 nmol/L. as the dependent variable (as in a previous analysis).43 The Plasma content of ACTH was determined using an im- power was 83% to detect an effect of size 0.4 and 45% to munoradiometric assay (Nichols Institute, San Juan Cap- detect an effect of size 0.25. A separate analysis of vari- istrano, Calif ). The antibody is specific for ACTH1-39 and ance was used to test lithium effects in bipolar patients. Sepa- exhibits less than 0.001% cross-reactivity with peptides of rate analyses were performed to assess the effects of age similar structure, including ACTH fragments 1 to 24, 11 (Pearson product moment correlation) and sex and sea- to 24, and 18 to 39. A 200-µL aliquot of sample or stan- son of testing (3-way analysis of variance). In each case, dard was combined with 100 µL of the antibody complex possible interactions were also considered. Paired and in- (125I-labeled monoclonal ACTH antibody and biotin- dependent t tests were also used for secondary assessment coupled polyclonal ACTH antibody in HEPES buffer con- of lithium and propranolol effects. taining protein stabilizers and 0.1% sodium azide) in poly- For each subject, the amplitude of variation of the mela- propylene tubes and vortex mixed. An avidin-coated tonin signal was assessed using the average deviation from polystyrene bead was added to each tube, and the tubes were the total testing period mean at each point; the sum of the incubated at room temperature for 20±2 hours. Beads were absolute value of these deviations was then divided by the washed twice, and ACTH content was quantified48,49 by total testing period mean.50 Peak time was compared as a counting samples concurrently with standards containing surrogate for phase. All computations were calculated by known quantities of human ACTH1-39. The minimum de- SAS statistical software, version 6.12 (SAS Institute Inc, Cary, tectable quantity of ACTH that could be distinguished from NC). Data are given as mean±SE unless otherwise indicated.

ence in light suppression of melatonin between 15 euthy- most unipolar patients would not. We anticipated that bi- mic bipolar patients and age- and sex-matched controls. polar patients might be supersensitive to propranolol sup- Nathan and colleagues,38 however, did report increased suppression of melatonin as well. We did not expect a sig- pression in 8 lithium-treated bipolar patients compared nificant effect of lithium (based on pilot data), although with controls. lithium has been reported to decrease retinal sensitivity The present study was undertaken to further test the to light.39 Corticotropin (ACTH) and cortisol levels were hypothesis of increased light sensitivity in a group of bi- also measured as potential trait vulnerability markers. In polar patients compared with unipolar patients and con- light of previous data,23(pp69-95),30,31 the melatonin-cortisol trols, to assess the effects of propranolol hydrochloride on ratio was also calculated. melatonin production in patients with bipolar illness, and to study the effects of lithium on light suppression of mela- RESULTS tonin. A larger group of patients is included, with stan- dardized diagnostic assessments (including bipolar I and Mean suppression levels on the light night were II disorder subtyping and a requirement for recurrent uni- 29.8%±5.5%, 32.2%±6.2%, and 34.6%±2.6% in bipo- polar episodes), careful attention to medication effects and lar patients, unipolar patients, and controls, respec- present mood state, and improvements in clinical meth- tively (Figure 1). Matched-pair comparisons did not ods for light delivery. It was anticipated that bipolar pa- show differences for the light night alone or for the dark- tients would show increased sensitivity to light but that adjusted melatonin suppression value (data not shown).

(REPRINTED) ARCH GEN PSYCHIATRY/ VOL 57, JUNE 2000 WWW.ARCHGENPSYCHIATRY.COM 574

©2000 American Medical Association. All rights reserved. Downloaded From: https://jamanetwork.com/ on 09/25/2021 Table 1. Clinical Demographic Features in Bipolar Study Patients and Matched Control Subjects*

Patient and Matched Controls Bipolar Patients† Matched No. of Episodes Time Not Control Sex/ Study Age, Study Length of Age at Length of Lithium Maintenance Taking No. Age, y Conditions y ConditionsIllness, y Onset, y Mania Depression Euthymia Dose, mg Medication‡ Medications§ 1 M/42 L, D, and P 42 L, D, P, and lithium 24 18 2 2 7 y 900 NA B 2 M/60 L, D, and P 57 L, D, P, and lithium 14 38 7 6-12 5 y 1050 NA B 3 F/36 L, D, and P 39 L, D, and P 15 23 1 2 1 y 600 Thiothixene B hydrochloride, 2mg 4 F/50 L, D, and P 48 L, D, and P 28 17 2 Ͼ20 3 y 1500 NA A 5 M/50 L, D, and P 47 L, D, and lithium 32 15 Ͼ20 Ͼ20 8 mo 750 Trazodone B hydrochloride, 100 mg 6 F/45 L, D, and P 44 L, D, and lithium 9 34 1 2 9 mo 1200 NA B 7 F/34 L, D, and P 33 L, D, and lithium 7 26 2 1 5 y 1200 NA B 8 M/38 L, D, and P 37 L, D, and lithium 1 28 1 1 9 y 750 NA B 9 M/56 L, D, and P 54 L, D, and P 4 30 2 1 24 y . . . NA 8 y 10 M/54 L and D 52 L, D, and P 28 24 Ͼ20 Ͼ20 5 mo . . . Carbamazepine, A 600 mg 11 M/24 L, D, and P 22 L, D, and P 6 16 2 0 2 y . . . Perphenazine 2y (dose unknown) 12 F/42 L, D, and P 40 L and D 7 33 3 1 6 mo . . . Carbamazepine, B 600 mg 13 M/35 L, D, and P 38 L, D, and P 12 20 2 0 6 y . . . NA 6 y 14 F/44 L and D 45 L, D, P, and lithium 16 28 10-15 0 1 y 900 NA A 15 F/. . . NA 23 Lithium 11 11 1 4 1 y 1050 NA C 16 F/. . . NA 40 Lithium 36 4 Ͼ20 Ͼ20 4 mo 900 Carbamazepine, C 200 mg, or fluoxetine, 20 mg 17 M/. . . NA 34 Lithium 4 26 Ͼ2 . . . 4 y 1200 Trimipramine C maleate, 25 mg, or fluoxetine, 20 mg 18 F/. . . NA 42 Lithium 7 32 3 0 3 y 1200 NA C 19 M/. . . NA 60 Lithium 8 42 3 2 10 y 900 NA C 20 M/. . . NA 63 Lithium 20 33 1 3 10 y 900 NA C 21 M/. . . NA 43 L and lithium 5 34 2 2 4 y 900 NA C 22 F/60 L, D, and P 45 L, D, P, and lithium 17 28 1 2 3 mo 600 Imipramine B hydrochloride, 50 mg 23 F/30 L, D, and P 29 L, D, and lithium 12 17 Ͼ20 Ͼ20 5 mo 900 NA B 24 F/42 L, D, and P 42 L, D, and P 26 16 16 3 4 mo . . . Fluoxetine, 20 mg A 25 F/37 L, D, and P 36 L and D 22 14 3 4 1 mo 300 Fluoxetine, 20 mg A 26 M/32 L, D, and P 34 L, D, P, and lithium 8 26 2 7 5 mo 900 NA A 27 F/44 L, D, and P 47 L, D, and lithium 16 27 . . . 6 4 y 600 NA A 28 M/45 L, D, and P 43 L, D, and P 14 29 Ͼ20 1 6 mo 900 NA A 29 M/47 L, D, and P 44 L, D, P, and lithium 23 17 Ͼ20 3 1 y 900 NA B

*All bipolar patients and matched controls were white. L indicates light; D, dark; P, placebo; NA, data not applicable; and ellipses, data not available. †Patients 1 through 21 had bipolar I disorder; 22 through 29, bipolar II disorder. ‡Maintenance medications were discontinued 3 to 6 months before the study. §A indicates 5 weeks or longer; B, from 2 to 5 weeks; and C, continued to take lithium.

A test of light suppression in 22 controls on 2 separate P=.41). Patients with light suppression greater than the occasions showed modest but significant reproducibil- bipolar mean did not differ from those with suppression ity (r=0.46; 95% confidence interval, 0.01-0.91; P=.03, less than the bipolar mean on age at onset, number of Spearman rank correlation). Baseline (average 1-2 AM) episodes, recent medication regimen, longest medica- melatonin levels were similar among the 3 groups. Ef- tion regimen, history of hospitalization, or presence of fects of sex and season of testing on light suppression level comorbid psychiatric conditions. were not seen in controls or in patients. Controls showed Propranolol decreased melatonin levels signifi- a negative correlation between age and light suppres- cantly, but this effect did not differentiate bipolar patients sion level (r=−0.35; 95% confidence interval, −0.64 to from controls (Figure 2, top and bottom). A significant −0.06; n=50; P=.01), whereas bipolar patients did not relation between light effect and propranolol effect was seen (r=0.05; 95% confidence interval, −0.48 to 0.38; n=23; in the controls (r=0.32; 95% confidence interval, 0.01-

(REPRINTED) ARCH GEN PSYCHIATRY/ VOL 57, JUNE 2000 WWW.ARCHGENPSYCHIATRY.COM 575

©2000 American Medical Association. All rights reserved. Downloaded From: https://jamanetwork.com/ on 09/25/2021 Table 2. Clinical Demographic Features in Unipolar Study Patients and Matched Control Subjects*

Patient and Matched Control Unipolar Patients Matched Length of No. of Episodes Age at Onset Length of Control No.† Sex/Age, y Study Conditions Age, y Study Conditions Illness, y of Depression of Depression, y Euthymia 1 F/48 L, D, and P 49 L and D 28 5 19 2 y 2 M/51 L, D, and P 47 L and D 24 3 18 4 y 3 M/46 L, D, and P 46 L and D 5 3 39 2 y 4 M/45 L, D, and P 43 L and D 15 4 26 2 y 5 F/32 L, D, and P 30 L and D 17 Ͼ20 11 2 y 6 M/64 L, D, and P 66 L and D 8 3 56 2 y 7 F/24 L, D, and P 22 L and D 5 3 13 3 y 8 M/39 L, D, and P 39 L and D 26 Ͼ20 13 3 mo 9 F/45 L, D, and P 44 L and D 12 3 27 3 y 10 F/46 L, D, and P 48 L and D 7 4 41 4 mo 11 M/21 L, D, and P 22 L and D 4 2 17 3 mo 12 F/43 L and D 42 L and D 28 Ͼ20 14 5 mo 13 M/31 L, D, and P 32 L and D 23 7 7 2 y 14 M/40 L, D, and P 40 L and D 27 2 12 8 mo 15 F/43 L, D, and P 43 L and D 36 Ͼ20 7 5 mo 16 F/30 L and D 32 L and D 15 4 17 3 mo 17 M/35 L, D, and P 33 L and D 11 5 22 6 mo 18 F/37 L, D, and P 34 L and D 13 2 21 4 mo 19 F/25 L and D 27 L and D 11 2 15 1 y 20 M/27 L, D, and P 36 L and D 6 2 20 9 y 21 F/28 L, D, and P 29 L and D 12 4 16 1 y 22 F/30 L and D 27 L and D 11 3 15 1 y 23 M/36 L and D 38 L and D 22 2 14 2 y 24 F/31 L and D 34 L and D 4 2 30 3 mo

*Maintenance medications were discontinued 3 to 6 months before the study. Three patients (patients 12, 15, and 16) withdrew from treatment with fluoxetine, 20 mg, 5 weeks before the study. L indicates light; D, dark; and P, placebo. †Unipolar patients 14 and 17 and their matched controls were African American; all other subjects were white.

creased dark-adjusted melatonin suppression com- 80 70 pared with matched controls (62.7%±15.6% vs 60 40.0%±9.1%). The patients with bipolar I disorder dem- 50 onstrated significantly lower baseline melatonin levels and 40 nadir (the 3-4 AM average) melatonin levels on the light 30 20 night. Patients with bipolar I disorder also showed a trend 10 to greater amplitude of variation in melatonin secretion 0 than controls (20.70%±12.09% vs 13.80%±8.30%; –10 t=1.98; P=.07) on the dark night. Peak time was noted –20 –30 to be significantly later in patients with bipolar I disor- Melatonin Suppression, % –40 der compared with matched controls. –50 Bipolar Patients (n = 23) Among patients with bipolar I disorder, there was a –60 Control Subjects (n = 50) significant interaction between medication status and dark- –70 Unipolar Patients (n = 24) Figure 8 –80 adjusted melatonin suppression ( ), with the greatest suppression scores among those not taking any psy- Figure 1. Melatonin suppression by light in 3 subject groups. Suppression is chotropic medications for 5 weeks or longer (98.8%±24.1%) calculated as follows: 100%−[(average level between 3 and 4 AM/average level between 1 and 2 AM)ϫ100%]. Distributions in the 3 groups do not compared with modest suppression among those taking differ (F2,94=0.33, P=.72). lithium alone (36.6%±43.8%) or not taking any medications for less than 5 weeks (35.7%±43.1%). 0.63; n=41; P=.04) but not in bipolar patients (r=−0.19; 95% confidence interval, −0.82 to 0.44; n=13; P=.38). COMMENT No effect of lithium treatment was seen in bipolar patients tested while not taking and while taking medi- The present results provide some support for the hypoth- cation (Figure 3). esis of melatonin secretion abnormalities among patients Cortisol and ACTH levels increased during the light with bipolar I illness. Patients showed decreased melato- administration procedure. Cortisol values did not dis- nin levels before and after light administration and a later tinguish the diagnostic groups (Figure 4 and Figure 5), peak on the night without light. Patients with bipolar I dis- and neither did ACTH values (Figure 6) or the mela- order showed 62% dark-adjusted melatonin suppression tonin-cortisol ratio (data not shown). compared with 40% in the matched controls. The greatest When bipolar patients were subdivided, patients with suppression scores were noted in the patients not taking bipolar I disorder (Figure 7) showed a trend to in- medication for the longest period. This study is similar to

(REPRINTED) ARCH GEN PSYCHIATRY/ VOL 57, JUNE 2000 WWW.ARCHGENPSYCHIATRY.COM 576

©2000 American Medical Association. All rights reserved. Downloaded From: https://jamanetwork.com/ on 09/25/2021 100 550 Placebo and Dark Bipolar Patients 90 Propranolol and Dark Control Subjects Placebo and Light Unipolar Patients 80 Propranolol and Light 410 70 60 50 280 40 Cortisol, nmol/L Melatonin, pg/mL 30 140 20 10 0 0

1 1:30 2 2:20 2:40 3 3:20 3:40 4 4:30 5 6 100 Time, AM 90 Figure 4. Cortisol levels during the light condition in 3 subject groups 80 (bipolar, n=23; unipolar, n=24; and control, n=50). A time and condition interaction (F11,2057=7.83, PϽ.001) was seen, but there was no effect of 70 group and no interaction including group. The gap between the dark bars on 60 the x-axis indicates the period of light exposure. 50 40 550 Melatonin, pg/mL 30 Bipolar Patients Control Subjects 20 Unipolar Patients 10 410 0

1 1:30 2 2:20 2:40 3 3:20 3:40 4 4:30 5 6 280 Time, AM

Figure 2. Melatonin levels throughout the night in bipolar patients (n=13) Cortisol, nmol/L (top) and control subjects (n=41) (bottom) under conditions of light and 140 dark with placebo administration and following oral propranolol hydrochloride administration, 10 mg, given at 11 PM. The gap between the dark bars on the x-axis indicates the period of light exposure on the nights 0 when light was administered. Data are presented as mean and SE. There is

an effect of condition (F1,156=129.75, PϽ.001) and drug (F1,156=4.27, P=.04) 1 1:30 2 2:20 2:40 3 3:20 3:40 4 4:30 5 6 onthe3to4AM average melatonin level, but there is no group effect and no Time, AM significant interaction including group. Figure 5. Cortisol levels during the dark condition in 3 subject groups (bipolar, n=23; unipolar, n=24; and control, n=50). A time and condition 100 interaction (F11,2057=7.83, PϽ.001) was seen, but there was no effect of group and no interaction including group. 90 80 11.0 70 Bipolar Patients 9.9 60 Control Subjects 8.8 Unipolar Patients 50 7.7 40

Melatonin, pg/mL 6.6 30 Dark 5.5 20 Lithium and Dark 4.4

10 Light ACTH, pmol/L Lithium and Light 3.3 0 2.2 1 1:30 2 2:20 2:40 3 3:20 3:40 4 4:30 5 6 Time, AM 1.1 0 Figure 3. Melatonin levels throughout the night in bipolar patients taking and

not taking lithium (n=12). A time and condition interaction (F11,484=3.11, 1 1:30 2 2:20 2:40 3 3:20 3:40 4 4:30 5 6 PϽ.001) is seen, but there is no effect of drug. One additional patient was Time, AM tested on a light night only while taking lithium; these data are not included in the figure. The gap between the dark bars on the x-axis indicates the Figure 6. Corticotropin (ACTH) values during the light condition in 3 subject period of light exposure on the nights when light was administered. groups (bipolar, n=10; unipolar, n=3; and control, n=21). An effect of time (F11,355=2.71, PϽ.003) was seen, but there was no effect of group and no groupϫtime interaction. The gap between the dark bars on the x-axis that of Lam et al36 in the low baseline melatonin levels noted indicates the period of light exposure. in patients. Our results tend to agree with those of Lewy et al34 and Nathan et al38 when the subgroup with bipolar Baseline melatonin levels were in the 40 pg/mL range I disorder is considered. However, a general difference be- for subjects in the Lewy et al34 study and in the 60 pg/mL tween bipolar patients and controls is not seen. range in the present study. This may be due to nonspe-

(REPRINTED) ARCH GEN PSYCHIATRY/ VOL 57, JUNE 2000 WWW.ARCHGENPSYCHIATRY.COM 577

©2000 American Medical Association. All rights reserved. Downloaded From: https://jamanetwork.com/ on 09/25/2021 100 Qualities of the light stimulus differed in the present 34 90 study compared with the study by Lewy et al and could 80 conceivably account for the disparate results. For instance, although full-spectrum fluorescent light was used in both 70 studies, the light in the present study was reflected from a 60 greenish beige wall and thus the incident light was not full 50 spectrum. Pursuing this explanation would require the hy- 40 pothesis that bipolar patients differ in neural systems linked Melatonin, pg/mL 30 to one or more sets of retinal cones but not all. 20 Other characteristics of the procedure may ac- 10 Patients With Bipolar I Disorder (n = 14) Dark Light count for observed differences. In the Lewy et al34 study, Matched Control Subjects (n = 14) Dark Light 0 subjects were not directly monitored during the proce-

1 1:30 2 2:20 2:40 3 3:20 3:40 4 4:30 5 6 dure. It is possible that bipolar patients may have been Time, AM more highly motivated than controls to look consis- tently and directly at the light source in front of them. Figure 7. Melatonin levels in patients with bipolar I disorder and control subjects matched for age, sex, and season of testing. Patients with bipolar I We attempted to minimize such behavioral differences disorder showed a trend to increased dark-adjusted melatonin suppression in the present study with close monitoring. The subject (light night suppression−dark night suppression) compared with matched was seated in a chair and was asked to be awake through- controls ( t13=1.78, P=.08). The patient group demonstrated significantly out the procedure, facing forward with eyes open. Nurs- lower baseline ( t13=2.40, P=.03) and nadir (the 3-4 AM average) ( t13=2.32, P=.04) melatonin levels on the light night. Patients with bipolar I disorder ing staff monitored compliance with these instructions. also showed a later peak in melatonin secretion (3:40 AM vs 2:50 AM; t=2.66, Another possible explanation involves random vari- P=.02) on the dark night. The gap between the dark bars on the x-axis ability in modest samples within a heterogeneous sub- indicates the period of light exposure on the nights when light was 34 administered. ject pool. The Lewy et al study involved observations of 11 bipolar patients. The present study includes 29 patients (21 with bipolar I disorder). It may be that in- 110 creased sensitivity is a shared characteristic of only a sub- 100 set of bipolar patients. This argument gains cogency by 90 recent studies in the genetics of complex disease. Re- 80 sults from a recent genomic survey by the NIMH Genet- 70 ics Initiative Bipolar Group52 reduce the likelihood that 60 any single locus accounts for 50% or greater variance in 50 a large data set. In fact, no single locus putatively iden- 40

Melatonin, pg/mL tified thus far appears to account for more than 15% of 30 the variance. The hypothesis of a unitary cause for bi- 20 Patients Not Taking Medications for ≥5 wk (n = 6) Dark Light polar illness has thus lost substantial support and no 10 Patients Not Taking Medications for 2-5 wk (n = 8) Dark Light 53 Patients Taking Lithium (n = 14) Dark Light longer seems reasonable as a premise. Suarez et al have 0 analyzed the statistical characteristics of such complex 1 1:30 2 2:20 2:40 3 3:20 3:40 4 4:30 5 6 inheritance. A true finding may not be replicated until Time, AM many subsequent groups of similar size have been tested. Figure 8. Dark-adjusted melatonin suppression in 3 groups of patients with A similar argument may apply to trait markers. bipolar I disorder separated by medication status. Patients not taking any The major limitations of this study are the limited psychotropic medications for 5 weeks or longer had greater suppression than those not taking any medication for 2 to 5 weeks or those tested while sample size and the age distribution of the subjects. We

taking lithium (F2,50=4.39, P=.02). The gap between the dark bars on the were also unable to control for the circadian phase of each x-axis indicates the period of light exposure on the nights when light was subject on the light-testing night (eg, by assessing dim- administered. light melatonin onset). The inclusion of a dark night did show that melatonin onset and offset times were similar cific immunoreactive substances detected by the RIA. in patients and controls (Figure 2, top and bottom). How- However, the values obtained from the 2 assays corre- ever, patients with bipolar I disorder show a delayed peak late highly (see the “Hormone Assay” subsection of the in melatonin level on the dark night. This suggests that “Subjects and Methods” section). Also, suppression lev- further investigation of phase delay in euthymic pa- els in controls were similar in the 2 studies (28% and 35%). tients with bipolar I disorder may be indicated. More detailed studies51 of the behavior of the RIA under There is reason for some optimism regarding future different clinical and laboratory conditions have been studies. Patients with bipolar I disorder showed melato- published. nin secretion abnormalities, as previously noted. This sub- We have noted a cortisol increase related to the light group is more severely ill than patients with bipolar II dis- stimulation procedure. Analysis of ACTH shows an ac- order and may have some genetic distinctions as well.52 companying peak. The 3 subject groups show generally These subgroups have generally not been separated in pre- similar effects, although a nonsignificant delay is noted vious melatonin suppression studies. It is also possible that in the bipolar cortisol peak. These appear to reflect a stress medication effects might mask a true difference between effect related to wakening since they are produced to the groups. Increased suppression in the subgroup not tak- same degree with a dim light (30 lux) and a bright light ing medication for 5 weeks or longer is consistent with (500 lux).43 this interpretation. However, it may be difficult to pursue

(REPRINTED) ARCH GEN PSYCHIATRY/ VOL 57, JUNE 2000 WWW.ARCHGENPSYCHIATRY.COM 578

©2000 American Medical Association. All rights reserved. Downloaded From: https://jamanetwork.com/ on 09/25/2021 hypotheses that require extensive periods of not taking 22. Dollins AB, Zhdanova IV, Wurtman RJ, Lynch HJ, Deng MH. Effect of inducing nocturnal serum melatonin concentrations in daytime on sleep, mood, body tem- medication and overnight challenge tests, if large popu- perature, and performance. Proc Natl Acad SciUSA.1994;91:1824-1828. lations must be tested. More economical and efficient meth- 23. Shafii M, Shafii SL, eds. Biological Rhythms, Mood Disorder, Light Therapy and ods might be devised to explore hypotheses related to cir- the Pineal Gland. Washington, DC: American Psychiatric Press; 1990. 24. Lewy AJ, Sack RL, Miller S, Hoban TM. Antidepressant and circadian phase- cadian rhythm disruption in bipolar affective disorder. shifting effects of light. Science. 1987;235:352-354. 25. Lewy AJ. Biochemistry and regulation of mammalian melatonin production. In: Rel- Accepted for publication January 24, 2000. kin R, ed. The Pineal Gland. New York, NY: Elsevier Science Inc; 1983:77-128. 26. Zatz M, Brownstein MJ. Injection of ␣-bungarotoxin near the suprachiasmatic This study was supported by grant R01 MH43325 from nucleus blocks the effects of light on nocturnal pineal enzyme activity. Brain Res. the National Institute of Mental Health, Rockville, Md; and 1981;213:438-442. 27. Zatz M, Brownstein MJ. Intraventricular carbachol mimics the effects of light on by a grant from the Indiana Division of Mental Health, In- the circadian rhythm in the rat pineal gland. Science. 1979;203:358-361. dianapolis. We acknowledge grant PHS MO1 RR750 from 28. Moore RY. Neural control of the pineal gland. Behav Brain Res. 1996;73:125-130. the Indiana University General Clinical Research Center, 29. Moore RY. The innervation of the mammalian pineal gland. In: Reiter RJ, ed. The Pineal and Reproduction. Basel, Switerland: S Karger AG; 1978:4:1-29. Progress Indianapolis.We thank Janice Froehlich, PhD, Department in Reproductive Biology; 4th ser. of Medicine, Indiana University School of Medicine, India- 30. Wetterberg L, Beck-Friis J, Aperia B, Petterson U. Melatonin/cortisol ratio in de- napolis, for determining the corticotropin levels. pression [letter]. Lancet. 1979;2:1361. 31. Wetterberg L, Aperia B, Beck-Friis J, Kjellman F, Ljunggren J-G, Petterson U, Sjo- Reprints: John I. Nurnberger, Jr, MD, PhD, Depart- lin A, Tham A, Unden F. Pineal-hypothalamic-pituitary function in patients with de- ment of Psychiatry, Indiana University School of Medi- pressive illness. In: Fuxe K, Gustafsson JA, Wetterberg L, eds. Steroid Hormone cine, 791 Union Dr, Indianapolis, IN 46202-4887 (e-mail: Regulation of the Brain. Oxford, England: Pergamon Press Inc; 1981:397-403. 32. Lewy AJ, Wehr TA, Goodwin FK, Newsome DA, Markey SP. Light suppresses [email protected]). melatonin secretion in humans. Science. 1980;210:1267-1269. 33. Lewy AJ, Wehr TA, Goodwin FK, Newsome DA. Manic depressive patients may be supersensitive to light [letter]. Lancet. 1981;1:383-384. REFERENCES 34. Lewy AJ, Nurnberger JI Jr, Wehr TA, Pack D, Becker LE, Powell RL, Newsome DA. Supersensitivity to light: possible trait marker for manic-depressive illness. Am J Psychiatry. 1985;142:725-727. 1. Goodwin FK, Jamison KR. Manic-Depressive Illness. New York, NY: Oxford Uni- 35. Nurnberger JI Jr, Berrettini W, Tamarkin L, Hamovit J, Norton J, Gershon E. Su- versity Press Inc; 1990. persensitivity to melatonin suppression by light in young people at high risk for 2. Benca R, Obermeyer W, Thisted A, Gillin C. Sleep and psychiatric disorders. Arch affective disorder: a preliminary report. Neuropsychopharmacology. 1988;1: Gen Psychiatry. 1992;49:651-668. 217-223. 3. Wehr TA, Goodwin FK. Biological rhythms and psychiatry. In: Arieti HK, Brodie 36. Lam RW, Berkowitz AL, Berga SL, Clark CM, Kripke DF. Melatonin suppression H, eds. American Handbook of Psychiatry. 2nd ed. New York, NY: Basic Books Inc Publishers; 1992:46-74. in bipolar and unipolar mood disorders. Psychiatry Res. 1990;33:129-134. 4. Carroll BJ, Feinberg M, Greden JF, Tarika J, Albala AA, Haskett RF, James NM, 37. Whalley LJ, Perini T, Shering A, Bennie J. Melatonin response to bright light in Kronfol Z, Lohr N, Steiner M, de Vigne JP, Young E. A specific test for the diag- recovered, drug-free bipolar patients. Psychiatry Res. 1991;38:13-19. nosis of melancholia. Arch Gen Psychiatry. 1981;38:15-23. 38. Nathan NJ, Burrows GD, Norman TR. Melatonin sensitivity to dim white light in 5. Checkley S. Neuroendocrinology. In: Paykel ES, ed. Handbook of Affective Dis- affective disorder. Neuropsychopharmacology. 1999;21:408-413. orders. 2nd ed. New York, NY: Churchill Livingstone Inc; 1992:255-266. 39. Seggie J. Lithium and the retina. Prog Neuropsychopharmacol Biol Psychiatry. 6. Whalley LJ, Kutcher S, Blackwood DH, Bennie J, Dick H, Fink G. Increased plasma 1988;12:241-253. LH in manic-depressive illness: evidence of a state-independent abnormality. Br 40. American Psychiatric Association. Diagnostic and Statistical Manual of Mental J Psychiatry. 1987;150:682-684. Disorders, Revised Third Edition. Washington, DC: American Psychiatric Asso- 7. Avery D, Wildschiodtz G, Rafaelsen OJ. Nocturnal temperature in affective dis- ciation; 1987. order. J Affect Disord. 1982;4:61-71. 41. Gershon ES, Hamovit J, Guroff JJ, Dibble E, Leckman J, Sceery W, Targum SD, 8. Lewy AJ. Human melatonin secretion (II): a marker for the circadian system Nurnberger JI Jr, Goldin LR, Bunney WE. A family study of schizoaffective, bi- and the effects of light. In: Post RN, Ballenger JC, eds. Neurobiology of Mood polar I, bipolar II, unipolar and normal control probands. Arch Gen Psychiatry. Disorders. Baltimore, Md: Williams & Wilkins; 1984:215-226. 1982;39:1157-1167. 9. Redman J, Armstrong S, Ng KT. Free-running activity rhythms in the rat: 42. Raskin A, Schulterbrandt J, Keating N, McKeon J. Replication of factors of psy- entrainment by melatonin. Science. 1983;219:1089-1091. chopathology in interview, ward behavior, and self-report ratings of hospital- 10. Murakami N, Hayafuji C, Sasaki Y, Yamazaki J, Takahashi K. Melatonin accel- ized depressives. J Nerv Ment Dis. 1969;148:87-98. erates the reentrainment of the circadian adrenocortical rhythm in inverted 43. Mayeda A, Mannon S, Hofstetter J, Adkins M, Baker R, Hu K, Nurnberger JI Jr. illumination cycle. Neuroendocrinology. 1983;36:385-391. Effects of light and propranolol on melatonin levels in normal human subjects. 11. Lewy AJ, Saeeduddin A, Sack RL. Phase shifting the human circadian clock Psychiatry Res. 1998;81:9-17. using melatonin. Behav Brain Res. 1996;73:131-134. 44. Arendt J, Pannier L, Sironenko PC. Melatonin radioimmunoassay. J Clin Endo- 12. Lerner AB, Nordlund JJ. Melatonin: clinical pharmacology. J Neural Transm. 1978; crinol Metab. 1975;40:347-350. 13(suppl):339-347. 45. Arendt J, Wirz-Justice A, Bradtke J, Kornemark M. Long-term studies on 13. Marczynski TJ, Yamaguchi N, Ling GM, Grodzinska L. Sleep induced by the ad- immunoreactive human melatonin. Ann Clin Biochem. 1979;16:307-312. ministration of melatonin (5-methoxy-N-acetyltryptamine) to the hypothalamus 46. Lahiri DK, Davis D, Adkins M, Nurnberger JI Jr. Factors that influence radio- in unrestrained cat. Experientia. 1964;20:435-437. immunoassay of human plasma melatonin: a modified column procedure to 14. Arendt J. The pineal hormone melatonin in seasonal and circadian rhythms. In: eliminate interference. Biochem Med Metab Biol. 1993;49:36-50. Redfern PH, Campbell IC, Davies JA, eds. Circadian Rhythms in the Central Ner- 47. Lewy AJ, Markey SP. Analysis of melatonin in human plasma by gas chroma- vous System. London, England: McMillan Publishing Co Inc; 1985:15-24. tography negative chemical ionization mass spectrometry. Science. 1978;201: 15. Brittman EL, Dempsey RJ, Karsch FJ. Pineal melatonin secretion drives the repro- 741-743. ductive response to daylength in the ewe. Endocrinology. 1983;113:2276-2283. 48. Rodbard D. Statistical quality control and routine data processing for radioim- 16. Bubenik GA, Smith PS, Schams D. The effect of orally administered melatonin munoassay and immunoradiometric assays. Clin Chem. 1974;20:1255-1270. on the seasonality of deer pelage exchange, antler development, LH, FSH, pro- 49. Rodbard D. Statistical analysis of radioimmunoassays and immunoradiometric (la-

lactin testosterone, T3 ,T4, cortisol and alkaline phosphatase. J Pineal Res. 1986; beled antibody) assays. In: Ericson A, ed. Symposium on Radioimmunoassays and 3:331-349. Related Procedures in Medicine. New York, NY: Unipub Inc; 1974:165-192. 17. Reiter RJ, Bitt JH, Armstrong JD. Absence of a nocturnal rise in either norepi- 50. Souetre E, Salvati E, Belugou JL, Pringuey D, Candito M, Krebs B, Ardisson JL, nephrine, N-acetyltransferase, hydroxyindole-O-methyltransferase or melato- Darcourt G. Circadian rhythms in depression and recovery: evidence for blunted nin in the pineal gland of the domestic pig kept under natural environmental amplitude as the main chronobiological abnormality. Psychiatry Res. 1989;28: photoperiods. Neurosci Lett. 1987;81:171-176. 263-278. 18. Connolly MS, Lynch CB. Classical genetic analysis of circadian body temperature 51. Lahiri DK, Davis D, Nurnberger JI Jr. Melatonin-like immunoreactivity in the pres- rhythms in mice. Behav Genet. 1983;13:491-500. ence of different chemicals as determined by the radioimmunoassay. Biochem 19. Cowen PJ, Bevan JS, Gosden B, Elliott SA. Treatment with ␤-adrenoceptor Med Metab Biol. 1994;51:51-54. blockers reduces plasma melatonin concentration. Br J Clin Pharmacol. 1985; 52. Nurnberger JI Jr, DePaulo JR, Gershon ES, et al for NIMH Genetics Initiative Bi- 19:258-260. polar Group. Genomic survey of bipolar illness in the NIMH genetics initiative 20. Iguchi H, Kato K, Ibayashi H. Age-dependent reductions in serum melatonin con- pedigrees: a preliminary report. Am J Med Genet. 1997;74:227-237. centrations in healthy human subjects. J Clin Endocrinol Metab. 1982;55:27-29. 53. Suarez BK, Hampe CL, Van Eerdewegh P. Problems of replicating linkage claims 21. Sack RL, Lewy AJ, Erb DL, Vollmer WM, Singer CM. Human melatonin produc- in psychiatry. In: Gershon ES, Cloninger CR, eds. Genetic Approaches to Mental tion decreases with age. J Pineal Res. 1986;3:379-388. Disorders. Washington, DC: American Psychiatric Press; 1994:23-46.

(REPRINTED) ARCH GEN PSYCHIATRY/ VOL 57, JUNE 2000 WWW.ARCHGENPSYCHIATRY.COM 579