.. ... .. ... . ... -

Brain Research. 323 (1984) 201-207Elsevier

BRE 10404

201

Research Reports

Effectsof Melatonin on Human Mood and Performance

HARRIS R. LlEBERMANI.2, FRANZ WALDHAUSERz*, GAIL GARFIELDI, HARRY J. LYNCH2andRICHARDJ. WURTMANZ

IDepartment 0/ Psychology £10-130 and 2Department 0/ Nutrilion £15-604, Massachusetts InstilUte o/Technology,CAmbridge, MA 02139 (U.S.A.)

(Accepted March 20th, 1984)

Key words: melatonin - pineal- sleepiness- performance- human

The function of melatonin, a hormone secreted by the pineal gland primarily at night, has not been definitively established in hu-'mans. To determine if pharmacologic doses of melatonin had any behavioral effects it was administered acutely to 14 healthy men.Their mood, performance, memory and visual sensitivity were assessed. Plasma melatonin concentration was assayed as well. Melato-nin significantly decreased self-reported alertness and increased sleepiness as measured by the Profile of Mood States and the StanfordSleepiness Scale self-report mood questionnaires. The effects were brief. Melatonin also affected performance, slowing choice-reac-tion time but concurrently decreasing errors of commission. Sustained fine motor performance was not impaired after melatonin ad-ministration nor were the tests of memory and visual sensitivity that were administered. It is concluded that melatonin, administeredorally in pharmacological quantities, has significant but short acting sedative-like properties.

INTRODUCTION

Little definitive information is available concern-

ing the function of the hormone melatQnin in hu-mans32.Evidence from a variety of sources does how-~ver suggest that this hormone could affect certainbehavioral functions. In both man and other species,melatonin is secreted by the pineal organ primarilyduring the night. This nocturnal release of melatonincan be suppressed by sufficiently intense Iight16.18.Itcan be hypothesized that the physiological role ofmelatonin in humans might thus have some relation-,ship to behaviors, like sleep, that are associated inhumans with darkness. That endogenous melatoninmight influence human behavior is also suggested byreports describing effects of this endogenous indoleon brain levels of serotonin1, a neurotransmitter

known to participate in the regulation of sleep and onthe ability of melatonin to reduce anxiety-like behav-iors in animals. For example, melatonin has been'shown to'decrease saccharin neophobia, increase ex-ploratory activity and decrease postural freezing inrats7,8.A few studies have also described induction of

sleep among human subjects receiving melato-nin2,s.26.

We have examined the ability of oral melatonin,given during the daylight hours in pharmacologicalquantities, to modify various aspects of behavior inhealthy young males, and have rel~ted such effects toplasma melatonin levels sampled concurrently. Weselected a battery of behavioral tests including manythat have been shown to be sensitive to substances

acting as sedatives.

* Present address: Univ. -Kinderklinik, Waehringer Geurtel 74-76 A-1090 Vienna, Austria.Correspondence: H. R. Lieberman, Dept. of Psychology E10-130, MassachusetU Institute of Technology, Cambridge, MA 02139,U.S.A.

0006-8993/84/$03.00@1984ElsevierSciencePublishers B.V.

202

MATERIALS AND METHODS

SubjectsFourteen paid male volunteers between the ages of

18 and 45 participated in this crossover study. Priorto enrollment in the study each subject had a physicalexamination and gave his informed consent to theprotocol.

Procedures

Melatonin and placebo were administered to eachsubject in a double blind, counterbalanced manner.Each subject received a total of 240 mg of oral mela-tonin or placebo (divided into three 80 mg doses)over a 2-h period (at 12.00, 13.00 and 14.00 h). Thedose selected was based on pilot studies conducted inour laboratory and previously published studies ofmelatonin's behavioral effects in humans2.5and other

species7.8. Since melatonin is rapidly metabo-lized14,27,it was administered in multiple doses tomaintain high plasma levels throughout the testingperiod. For every subject a minimum washout periodof two weeks separated the first from the second ses-sion.

On each testing day for 2 h before (10~00-12.oo)and 2 h after (14.00-16.00 h) treatment, the subjectswere tested on a battery of behavioral tests thatmeasured various aspects of performance, memoryand visual sensitivity. In addition, two self-reportmood questionnaires were administered at hourly in-tervals.

Also, 10ml blood samples were drawn hourly froman intravenous line and used for the determination of

plasma melatonin concentration.

Mood scalesadministered

Profile of mood states (POMS). The POMS is aself-report mood questionnaire which, when ana-lyzed, yields 6 factors: Tension-Anxiety, Depression-Dejection, Anger-Hostility, Vigor-Activity, Fatigue-Inertia, and Confusion-Bewilderment21. The testconsists of 65 adjectives each of which is rated on a 5-point scale. The POMS has been employed in manypsychopharmacological studies and is sensitive to theeffects of many different classes of psychoactivedrugs, includinghypnotics12and stimulants4.

Stanford Sleepiness Scale (SSS). This self-rated 7-point scale was designed to quantify the progressive

stages of the alertness-sleepiness continuuml1. It hasbeen used in a number of psychopharmacologicalstudies and is sensitive to the effects of hypnotics24.

Performance, visualsensitivity and memory testsWe selected a battery of tasks that sampled a vari-

ety of perceptual and motor capabilities as well as vis-ual sensitivity and memory. Each test or similar testshave previously been shown to be sensitive to the be-havioral effects of various drugs.

Simple auditory reaction time. In this microcom-puter administered test, the subject responded, asrapidly as possible, to the onset of a 75 dB (SPL),1900Hz tone. After 5 warm-up trials, 125 test trialswere presented in rapid succession. A visual cue,presented on a cathode-ray tube (CRT), indicatedthe start of a trial. Both commission errors (respond-ing prior to the termination of the stimulus tone) anderrors of omission (response latency greater thanone s) were recorded.

Four-choice visual reaction time. This test closelyresembles the Wilkinson four-choice RT task30and is

a measure of sustained visual vigilance. Subjects arepresented a series of visual stimuli at one of four dif-ferent spatial locations on a CRT screen. The subjectmust correctly indicate, by striking one of four adja-cent keys on a microcomputer keyboard, the correctlocation of each stimulus. Four hundred trials wereadministered and errors of omission (response laten-cy greater than 1s) and commission wer~recorded.

Groovedpegboardtest. This modified version ofthe grooved pegboard test (Lafayette Inst., La-fayette, IN) was designed to measure sustained,complex motor performance. In the standard versionof the test the subject is required to insert, as .rapidlyas possible, a series of 25 pegs into randomly orientedholes on a board. Since both the pegs and board aregrooved, each peg must be properly oriented to be in-serted. To measure sustained performance, the taskwas repeated 8 times in rapid succession.

Digit symbol substitution test (DSST). This test,taken from the Weschler IQ test28,has been reportedto be sensitive to drug-induced decrements in per-formance20.The subject is presented with a series ofdigits and a code which identifies 'each digit with aparticular symbol. The subject must correctly copy,in boxes directly below the digits, as many of the ap-propriate symbols as possible in the 90-s time period

allotted.

Critical flicker fusion (CFF). CFF is the visualthreshold at which a flickering light is perceived to besteadily on. This test was selected to indicate whethermelatonin altered temporal visual sensitivity, It hasalso been reported that this simple sensory task issensitive to drugs that impair brain function13.CFFwas measured with two yellow (585 nm dominantwavelength) light emitting diodes (LED) as the lightsource. The light from the LEOs was diffused andpresented through a circular opening which sub-'tended 40 of visual angle. A parameter estimationstaircase technique17was used to rapidly determineeach observer's threshold.

Recall and recognition memory. Since benzodiaze-pines impair certain memory functions, and melato-nin has been rePorted to have some benzodiazepine-like properties7.8.19, we selected a memory testknown to be sensitive to the effects of these drugs.The test, adapted from Brown et aJ.3, presented lists

POMSVIGORSCALE

12 I . 2 3 4 5

ClockHours(p.m.)Fig. 1. The effect of melatonin or placebo on the Vigor scale ofthe POMS self-report mood questionnaire. Melatonin was ad-ministered in 3 doses of 80 mg each at 12.00. 13.00 and 14.00 has indicated by the arrows.

203

of 10 words selected from a single category such as'animals', Each list was presented by tape recorder ata rate of one item per second and succeeded by a listof 6 digits. Six different lists were presented onthe two testing days at 10.10, 14.10 and 15.40 h.Immediately after the presentation of each list andthe accompanying digits, the subject was asked to re-call the digits in order of presentation and the wordlist in any order. At 16.20 h, recognition memory foreach word list presented during the day was tested.At that time, for each list, the subject was sequential-ly presented' with 20 cards in randomized order; 10with the words from that list and 10 with new words

selected from the same category. The subject ratedeach word on a four point scale.

Melatonin assayMelatonin in serum was measured by radioimmu-

noassay (RIA) using anti-melatonin serum providedby Dr. L. Levine of Brandeis University, Waltham,MA15.One ml samples of serum were extracted inchloroform. The organic extract was evaporated todryness under a stream of nitrogen and the residueredissolved in 0.5 ml of Tris buffer (pH 7). The bufferextract was then combined with 100.ul of antiserumsolution (diluted 1:5(00) and 100J.d(1750 CPM) ofpH]melatonin (New England Nuclear, Boston,MA). After the mixture was incubated for 1 h at35 °C, . saturated ammonium sulfate solution wasadded and antibody-bound pH]melatonin was col-lected as a precipitate by centrifugation. Radioactivi-ty was then measured in a liquid scintillation counterand melatonin concentrations were estimated bymeans of the logit-Iogplot23.The sensitivity of the as-say (B/Bo = 85%) varied between 5 and 10pglml se-rum. The recovery of authentic melatonin added tothe serum samples was 86.9 ::!:2.1%. In control sam-ples containing 38.5 and 188 pg melatonin per ml se-rum, the intraassay coefficients of variation were13.85 and 7.95%, respectively. The corresponding

. interassay coefficients of variation were 21.2% and12.5%.

RESULTS

Behavioral tests

Data from the POMS and SSSwere analysed withrepeated measures analyses of variance (ANOY As).

(IMELATONIN

201- T ToPLACEBO

1",1p<O.05

181 U

...........1\(\1;.

f'-r

Q) 16' o,Z\.5u

c.n 1c:: 140Q):E

12

10

8

204

POMS FATIGUESCALE

I MELATONIN

DPLACEBO

p<005

2 3 4 5

Clock Hours (p.rn.)

Fig. 2. Effect of melatonin or placebo on the Fatigue scales ofthe POMS self-report mood questionnaire. The arrows indicatethe.times when melatonin was administered (SOmg each time).

If a significant main effect was detected with theANOV A, then two-tailed Newman-Keuls a posteri-ori tests were performed on melatonin-versus-place-bo means for each hourly measurement.

Significant hypnotic-like properties of melatoninwere detected by the self-report mood question-naires. The scores from the Vigor scale of the POMS(Fig. 1) were significantly reduced by melatonin

.(F(1,13) = 9.48, P = 0.0088) at 15.00 and 16.00 h,and scores from the Fatigue scale were significantlyelevated (F(1,13) = 6.87, P = 0.021) at 15.00 b(Fig. 2). Melatonin also significantly increased thescores on the Confusion subscale of the POMS

(F(1,13) = 6.63, P = 0.023) but no significant differ-ences were isolated by the a posteriori tests at anyspecific time period. The SSSalso detected a main ef-fect of melatonin (F(1,13) = 11.29, P = 0.(05), withsubjects reporting significantly increased sleepinessat 15.00 and 16.00 h (Fig. 3). By 17.00 b melatoninno longer significantly modified any of these varia-bles. The other POMS subscaleswere not significant-ly altered by melatonin at any time.

Melatonin also significantly altered certain aspectsof performance. Since these tests were administered

on each test day both before. and after each subjectreceived melatonin or its placebo, difference scorescould be computed by subtracting the baseline per-

formance score from that obtained during the post-.drug session. Paired t-tests could then be performed:Melatonin significantly increased response latencyon the 4-choice visual RT task (Fig. 4; t (13) = 4.40,P < 0.(01). Although this aspect of performance wasimpaired, the numbers of errors (incorrect re-sponses) that subjects made were significantly de-creased (Fig. 5; t (13) = 3.10, P < 0.01).Melatoninalso significantly reduced the number of errors madeon the simple RT task (t (13) = 2.%, P < 0.02) al-though simple RT latency was not significantly in-creased. None of the other tests administered was

significantlyaffected by melatonin.

Melatonin assaysBy 13.00h, 1h after administration of the first mel-

atonin dose, a substantial increase in plasma melato-nin (over 3 orders of magnitude) was apparent(Fig. 6). Plasma levels remained high throughout the

STANFORD SLEEPINESS SCALE

5

12 51 2 3 4

Clock hours (p.rn.!Fig. 3. Effect of melatonin or placebo on the SSS self-reportmood questionnaire. Melatonin was administered in 3 doses' ofSOmg each as indicated by the arrows.

12L

10

Q,)....0 8u

Cf)

c:::cQ,):iE

012

IIMELATONIN

oPLACEBOp<0.05

p<0.05

I [ 1\1

1I1

lLi

N1",.

($,

,L- L- L - '--

4

Q,)....0u

Cf)

c::: 3cQ,):iE

2

afterno.on, an effect probably attributable to the ad-ditional dosesof melatonin administered at 13.00 and

14.00 h27. It is of interest that the greatest effect ofmelatonin administration on subjective mood was at15.00 and 16.00 h, well after the onset of elevated

plasma melatonin concentrations even though mela-tonin readily diffuses across the blood-brain bar-rier31. Melatonin plasma levels were still elevated at17.00 h, although mood state had returned to nearnormal by this time.

DISCUSSION

These data indicate that melatonin alters mood

FOUR-CHOICEREACTIONTIME

-16

SubstanceFig. 4.ChangeinFour-choicereactiontimelatencyafteringes-tionofmelatoninascompairedto placebo.

205

FOUR-CHOICEREACTIONTIME

ERRORS

_MELATONIN

4~ DPLACEBO

p<O.OI

2-o

o

-2Q)c:.-

Q)enC

.0 -4Eo~-Q) -6C"c::C

..t::.u -8

-10Substance

Fig. S.Changein number of errors on the Four-choice reactiontime taskafter ingestionof melatonin ascompaired to placebo.

state in a manner similar to drugs with sedative-likeproperties. For example, Johanson and Uhlenhuth12noted a similaracute effect of diazepam (5 and 10 mgadministered orally) on the Vigor and Fatigue scalesof the POMS. The increase in response latency de-tected by the 4-choice RT task is also similar to thatwhich can be induced by such drugs10, or by sleepdeprivation29,on RT tasks. The decreased number oferror:son the 4-choice and simple RT tasks noted af-ter melatonin administration is .somewhat surprisingsince one might expect errors to increase after ad-ministration of a substance with sedative-like prop-erties. In fact, sleepiness and fatigue do increase er-rors of omission and such errors are considered to be asensitive index of impaired performance29. Howeverthe types of error decreasedby melatonin were errors

_MELAl:ONIN

161 ID PLACEBO

u p<O.OOIQ)enE 12c:

Q)E 8+=c:0

+=4<..)

cQ)

Q)c: 0Q)enC

.0

E -40-

-8c:c

..t::.<....:>

-12

0.001I I I I I I

12 I 2 3 4 5

ClockHours(p.m.)Fig. 6. Mean plasmamelatonin plasmaconcentration (:t S.E.)after administration of 3 oral doses of melatonin (each arrowindicatesan 80 mgdose).

of commission, either incorrect responses (in the4-choice RT task) or anticipation errors (on the sim-ple RT task); such errors reportedly correlate negati-vely with R'f9, hence their reduction by melatonin

REFERENCES

1 Anton-Tay, F., Chou, C., Anton, S. and Wurtman, R. J.,Brain serotonin concentration:elevation followingintrape-ritoneal administration of melatonin, Science, 162 (1968)277-278.

2 Anton-Tay, F., Diaz, J. L. and Fernandez-Guardiola, G.,

. maybe consistent with the other hypnotic propertiesof this hormone. This negative correlation may rep-resent a trade-off between speed and accuracy. Oth-er hypnotic drugs, such as the benzodiazepines, have.also been reported to decrease commission errors incertain timed cognitive tasks6 while simultaneouslyincreasing the time required to complete the task25.

Melatonin's nocturnal secretion, the fact that this

secretion is under direct photic control, and the abilityof exogenous melatonin to induce sleepiness allsuggest that the hormone might mediate the syn-chronization of human circadian rhythms to thelight-dark cycle. No other hormone is known to havethis unique combination of properties. Of course, theserum melatonin concentrations we observed after

melatonin administration were much greater thanpeak nocturnal concentrations, hence the behavioraleffects produced by the exogenous hormone mayormay not mimic those of endogenous melatonin. Mela-tonin has recently been reported to synchronize cir-cadian rhythms in rats when it is administered at thesame time of day for several weeks22. Exogenousmelatonin should also be considered as a candidate

for evaluation as a clinical hypnotic, especially in sub-jects whose sleep disorders may be related to de-synchronized circadian rhythms.

ACKNOWLEDGEMENTS

This research was supported by National Instituteof Health Grant 2ROI-HDl1722 and National Aero-

nautics and Space Administration Grant NAG2-132.We wish to thank the staff of the Clinical Research

Center, Massachusetts Institute of Technology fortheir assistance. Dr. M. Potter provided valuable ad-vice in the development of the memory task and Dr.E. S. Lieberman provided helpful comments on anearlier version of this manuscript.

Portions of this work were presented at the Societyfor Neuroscience November 1983meeting in Boston,Massachusetts.

On the effect of melatonin upon human brains: its possible'therapeutic implications, Life Sci., 10(1971)841-850.

3 Brown, J., Lewis, V., Brown, M., Horn, G. and Bowes, J.B., A comparison between transient amnesiasinduced bytwo drugs (diazepamor lorazepam) and amnesiaof organicorigin, Neuropsychologia.20(1982)55-70.

4 Cole, J. 0., Pope, H. G., LaBrie, R. and Ionescu-Pioggia,

206

SerumMelatoninConcentration

1000I-j t t100

/!-t''-I1 1"'-1 f- I ,

E10I-..........

C'C-CCC> 1.0I--C(1)::2:E:::3 0.1 ....(1)

en

o.J

M., Assessing the subjective effects of stimulants in casualusers, Clin. Pharmacol. Ther., 24 (1978) 243-252.

5 Cramer, H., Rudolph, J., Consbruch, U. and Kendel, K.,On the effects of melatonin on sleep and behavior in man,Adv. Biochem. Psychopharmacol., 11 (1974) 187-191.

6 Fisch, H.-U., Groner, M., Groner, R. and Menz, C., Influ-ence of diazepam and methylphenidate on identification ofrapidly presented letter strings: diazepam enhances visualmasking, Psychopharmacology, 80 (1983) 61-66.

7 Golus, P. and King, M. G., The effects of melatonin onopen field behavior, Pharmacol. Biochem. Behav., 15(1981) 883-885.

8 Golus, P., McGee, R. and King, M. G., Attenuation of sac-charin neophobia by melatonin, Pharmacol. Bwchem. Be-hav., 11 (1979) 367-369.

9 Herscovitch, J. and Broughton, R., Performance deficitsfollowing short-term partial sleep dperivation and subse-quent recovery oversleeping, Canad. I. Psychol., 35 (1981)309-322.

10 Hindmarch, I., Psychomotor function and psychoactivedrugs, Brit. I. Clin. Pharmacol., 10 (1980) 189-209.

11 Hoddes, E., Zarcone, V., Smythe, H., PhiUips, R. and De-ment, W. C., Quantification ofsleepiness: a new approach,Psychophyswlogy, 10 (1973) 431-436.

12 Johanson, C. E. and Uhlenhuth, E. H., Drug preferenceand mood in humans: diazepam, Psychopharmacology, 71(1980) 269-273.

13 Kleinknecht, R. A. and Donaldson, D., A review of the ef-fects of diazepam on cognitive and psychomotor perform-ance,l. nerv. ment. Dis., 131 (1975) 399-411.

14 Kopin, I. J., Pierce, C. M., Axelrod, J. and Weissbacb, H.,The fate of melatonin in animals, I. Bioi. Chem., 236 (1961)3072-3075.

15 Levine, L. and Riceberg, L. R., Radioimmunoassay formelatonin, Res. Commun. Chem. Pathol. Pharmacol., 10(1975) 693.

16 Lewy, A. J., Wehr, T. A., Goodwin, F. K., Newsome, D.A. and Markey, S. P., Light suppresses melatonin secretionin humans, Science, 210 (1980) 1267-1269.

17 Lieberman, H. R. and Pentland, A. P., Microcomputer-based estimation of psychophysical thresholds: the bestPEST, Behav. Res. Meth. InstrumentatitJn, 14 (1982)21-25.

18 Lynch, H. J., Wurtman, R. J., Moskcowitz, M. A., Archer,M. C. and Ho, M. H.., Daily rhythm in human melatonin,Science, 187(1975)169-171.

19 Marangos, P. J., Patel, J. Hirata, F., Sandheim, D., Paul,S. M., Skolnick, P. and Goodwin, F. K., Inhibition of diaze-

207

pam binding by tryptophan derivatives including melatonin

and its brain metabolite N-acetyl-5 methoxy-kynuramine,LifeSci., 29 (1981) 259-267.

20 McNair, D. M., Antianxiety drugs and human perform-ance,Arch. Gen. Psychiat., 29 (1973) 611-617.

21 McNair, D. M., Lorr, M. and Droppleman, L. F., Profile ofMood States Manual, Educational and Industrial TestingService, San Diego, 1971:

22 Redman, J., Armstrong, S., and Ng, K. T., Free-runningactivity rhythms in the rat: entrainment by melatonin, Sci-ence, 219 (1983) 1089-1091.

23 Rodbard, D., Statistical quality control and routine dataprocessing for radioimmunoassay and immunoradiometricassays, Clin. Chem., 20 (1974) 1255.

24 Spinweber, C. L. and Johnson, L. C., Effects of triazolamon sleep, performance, memory, and arousal threshold,Psychopharmacology, 76 (1982) 5-12.

25 Uhlenhuth, E. H., Stem, M., Schuster, C. R. and Domizi,D., Diazepam: efficacy and toxicity as revealed by a smallsample research strategy. In A. Sudilovsky, S. Gershon andB. Beer (Eds.), PredicuzbUityin Psychopharmacology: Pre-clinical and Clinical Co"elations, Raven Press, New York,1975.

26 Vollrath, L., Semm, P. and Gammel, G., Sleep il1ductionby intranasal application of melatonin. In N. Birau and W.Schloot (Eds.), Melatonin Current Studies and Perspectives,Pergamon Press, London, 1975.

27 Waldhauser, F., Waldhauser, M., Lieberman, H. R.,Deng, M., Lynch, H. J. and Wurtman, R. J., Bioavailabili-ty of oral melatonin in humans, in press.

28 Wechsler, D., Wechsler Adult Intelligence Scale, Psycholo-gical Corporation, New York, 1955.

29 Wilkinson, R. T., Sleep deprivation: performance tests forpartial and selective sleep deprivation. In L. A. Abt and B.F. Reiss (Eds.), Progress in Clinical Psychology, Vol. 7,Grune & Stratton, New York, 1%9.

30 Wilkinson, R. T. and Haughton, D., Portable four-choicereaction time test with magnetic tape memory, Behav. Res.Meth. Instrumenuztion, 7 (1975) 441-446.

31 Wurtman, R. J., Axelrod, J. and Potter, L. T., The uptakeof W-melatonin in endocrine and nervous tissue and the ef-fects of constant light exposure, I. Pharmacol. expoTher.,143 (1964) 314-318.

32 Wurtman, R. J., Waldhauser, F. and Lieberman, H. R.,The secretion and effects of melatonin in humans. In J. Ax-elrod, F. Fraschini and G. P. Velo (Eds.), The Pineal Glandand UsEndocrine Role, Plenum Press, New York, 1983.