NEUROPSYCHOPHARMACOLOGY 1993-VOL. 8, NO.4 345

Effects of Catecholamine Depletion on Alertness and Mood in Rested and Sleep Deprived Normal Volunteers Una D. McCann, M.D., David M. Penetar, Ph.D., Yavin Shaham, Ph.D., DruidR. Thorne, Ph.D., Helen C. Sing, B.S., Maria L. Thomas, B.A., ,. Christian Gillin, M.D., and Gregory Belenky, M.D.

Alpha·methyl-para-tyrosine (AMPT), a tyrosine In contrast, although combined treatment with AMPT kydroxylase inhibitor, was used to evaluate the and sleep deprivation led to large increases in negative ylrysiologic role of central nervous system catecholamines mood, neither treatment alone produced consistent mood • modulating alertness and mood. Forty healthy males changes. These findings are consistent with the view that 1m randomized to one of four conditions: AMPT in a sleep deprivation is associated with decreased functional sed condition; AMPT plus 40.5 hours of total sleep catecholamine neurotransmission. Furthermore, mood Itprivation; placebo plus sleep deprivation; or placebo in effects following sleep deprivation plus AMPT suggest • rrsted condition. Repeated measures of alertness and that catecholamines may be involved in mood changes mxl rroealed that treatment with AMPT or sleep during sleep deprivation. £Neuropsychopharmacology. Itprir1(ltion increased sleepiness, and combined treatment 8:345-356, 1993J "muced greater sleepiness than either treatment alone.

DY WORDS: Sleep deprivation; Alertness; Mood; The restorative actions of amphetamine are thought to CIltcholamines; Monoamines be secondary to functional increases in catecholamine neurotransrnission (Glowinski et al. 1966). Conversely, Thereis considerable pharmacologic evidence that cen­ drugs that decrease catecholamine activity also influ­ Ira!n ervous system (CNS) catecholamines are impor­ ence arousal. In general, drugs that inhibit noradrener­ tant modulators of arousal in man. For example, gic or dopaminergic activity increase daytime sleepi­ llimulants such as amphetamine effectively reverse ness. For example, a variety of catecholamine receptor decrements in alertness and cognitive performance as­ blockers cause sedation, fatigue, and sleepiness (Betts IlCiatedwith sleep deprivation (Weiss and Laties 1962). 1981) and decrease sleep latency times (Nicholson and Pascoe 1990). From the Department of Behavioral Biology; Walter Reed Army Further evidence for catecholaminergic involve­ .atuteofResearch (UDM, DMP, DRT, HCS, MLT, GB), Washing­ ment in arousal stems from clinical observations made .... DC; Section on Anxiety and Affective Disorders, Biological Plydtiatry Branch (UDM), National Institute of Mental Health, in a variety of patient populations treated with alpha­ SiIionallnstitutes of Health, Bethesda, Maryland; Department of methyl-para-tyrosine (AMPT);AMPT is a selective cat­ PIydlology, Concordia University (YS), Montreal, Quebec, Canada; echolamine synthesis inhibitor that can be utilized to alDepartmentof Psychiatry, University of California,and Psychiatry SInn,San Diego V.A. Medical Center (JCG), San Diego, California. determine the involvement of catecholamines in be­ Addresscorrespondence to: Una D. McCann, Section on Anxiety havior (Nagatsu et al. 1964; Spector et al. 1965). In pa­ alAffectiveDisorders, National Institute of Mental Health, Building tient populations, AMPT has been noted to produce II. Room 35·239, 9000 Rockville Pike, Bethesda, Maryland 20892. RtreivedJuly 1, 1992; revised October 28, 1992; accepted November sedation early in the course of treatment (Engelman et 111992. a1. 1968; Jonsson et a1. 1971; Ahlenius et al. 1973;

Pabbshed 1993 by Elsevier Science Publishing Co., Inc. 0893-133X/93/$0.00 346 U. D. McCann et a1. NEUROPSYCHOPHARMACOLOGY 1993-VOL. 8, NO.4

Pozuelo 1976; Nasrallah et al. 1977}. In addition, it has tained, subjects completed a Beck Depression Inven· been noted that discontinuation of AMPT treatment is tory ([BOI] Beck et al. 1961) and underwent a thorough associated with insomnia and decreased total sleep physical and laboratory examination, including an elec· (Bunney et al. 1971, 1977). This observed rebound in­ trocardiogram, psychiatric screening, SMA-20, thyroid somnia has been postulated to be secondary to an in­ function tests, hepatitis screen, human immunoden· crease in catecholaminergic activity. Further evidence ciency virus testing, complete blood count, urinalysis, for catecholaminergic involvement in the sleep-wake and urine drug screen for drugs of abuse. Exclusion cycle has been reviewed elsewhere (Mendelson 1987). criteria included self-reported past or present major In addition to their importance in sleep, there iscon­ medical or psychiatric illness, positive urine drug siderable pharmacologic evidence that brain catechol­ screens, a score of 10 or greater on the BDI, regular amines are involved in the regulation of mood (Schild­ tobacco use, ingestion of more than three cups of coffee kraut 1965; Bunney and Davis 1965; Greenspan et al. (or the equivalent) per day, or atypical sleeping patterns 1969; Iqbal et al. 1989; Schildkraut et al. 1973a, 1973b; (including unusually delayed or advanced sleep phases, Goodwin and Post 1975), although the relationship be­ history of insomnia or hypersomnia, or habitual sleep tween sleep and mood is complex. For example, major hours not occurring between 2000and 1000hours). Fol· depression is associated with objective electroen­ lowing completion of screening procedures, subjects cephalographic sleep abnormalities (Kupfer and Thase were randomlyassigned to one of four treatment con· 1983; Gillin 1983), and manic patients typicallyhave de­ ditions: 1) treatment with AMPT in a rested state creased sleep, particularly during the period when they (AMPT/R); 2) treatment with AMPT plus sleep depri· switch into mania (Bunney et al. 1972). Further suggest­ vation (AMPTISD); 3) treatment with placebo plus sleep ing a link between the regulation of mood and arousal deprivation (P/SD); or 4) treatment with placebo ina is the observation that the majority of depressed pa­ rested state (P/R). A schematic time line of study proce­ tients exhibit symptomatic improvement following one dures is depicted in Figure 1. night of total sleep deprivation (Vogel et al. 1975). The mechanism by which sleep deprivation improves mood Drug Administration in depressed individuals remains unclear, but may in­ volve catecholamines (Siegel and Rogawski 1988; Hart­ Drugs were administered in a double-blind fashion, mann and Schildkraut 1973; Schildkraut et al. 1973b; with subjects from both drug conditions included in Post et al. 1976; Mogilnicka et al. 1986). each study run, beginning at 1430 hours on treatment Despite the growing clinical interest in the relation­ day 1 (TD1). Subsequent doses were administered at ship between sleep and mood, few studies have 1830 and 2330 hours on TD1, and at 0730, 1330, 1830, reported on the effects of one night of sleep depriva­ and 2330 hours on treatment day 2 (TD2). Each of the tion on mood in healthy populations (Gerneret al. 1979; seven doses contained 750 mg AMPT (or lactose pia· Gillberg and Akerstedt 1981), and of these, results have cebo), for a total of 5.25 g AMPT over a 33-hour period. varied. Moreover, no study that has investigated the Previous studies have shown that this regimen depletes effect of sleep deprivation on mood in normal volun­ cerebrospinal fluid catecholamines by 68% to 77%af· teers has used a rested control group. The purpose of ter 24 hours (Brodie et al. 1971). Subjects treated with this study was twofold: frrst, to systematically evalu­ AMPT in the present study had signiflcantelevations ate the effects of sleep deprivation on self-reported in serum prolactin concentrations on day 2, indicating mood in healthy volunteers; and second, to use AMPT depletion of eNS catecholamines at the level of the as a pharmacologic probe for assessing the role of cate­ cholamine neurotransmitters in changes of alertness Tlme Une forAMPT study mood associated with sleep deprivation. The pres­ and 0700 ent study was performed as part of a larger research Sleep·Deprivation project investigating the role of catecholamines in alert­ ness, sleep, performance, and mood. Adaptation Baseline Treatment T,elltment Recovery Night Night Day t Day 2 Day

METHODS

Subjects

Forty healthy men with a mean age of 25 years (range Figure 1. Generalized time line for study procedures. Fol· 21 to 36) were recruited through advertisements in lo­ lowing adaptation and baseline sleep nights, subjects were cal newspapers. After an initial telephone screen, in­ sleep deprived for 40.5 hours (treatment days 1 and 2). Our· terested volunteers came to the laboratory for a more ing the sleep-deprivation period, subjects received seven detailed evaluation. After informed consent was ob- doses of AMPT (750 mg/dose) or placebo. �OPSYCHOPHARMACOLOGY 1993-VOL. 8, NO.4 AMPT and Sleep Deprivation in Normal Volunteers 347

Ipothalamus(McCann et al. 1992). Subjects received Alertness and Mood Measures 6eirla st dose of AMPT immediately before recovery Simultaneous measures of alertness and mood were *ep.at the end of the sleep-deprivation period. taken during the two sleep-deprivation and drug­ treatment days (TD1 and TD2) and the day following recovery sleep (recovery day). Measurements included IerpDeprivation a pencil and paper VAS designed to measure global PI!viouslydescribed methods for sleep-deprivation and vigor and affect (Monk 1989), and a computerized ver­ paformance assessment were used, with minor mod­ sion of the Prohle of Mood States ([POMS] McNair et btion(Newhouse et al. 1989). Briefly, following an al. 1971) which is incorporated into the WRAIR PAB. lllapta tion sleep night in the laboratory and training These measures were chosen because they have been .. acomp uterized performance assessment battery validated in studies assessing mood changes in (tt'a1terReed Army Institute of Research Performance depressed patients and normal controls (Gillin et al. h!essmentBa ttery, WRAIR PAB [Thome et al. 1985]), 1989; Wehr et al. 1985), and because unlike the MSLT IIbjects were monitored polysomnographically for and SSS, these measures assess both alertness and _line sleep using an Oxford Medilog Recorder. mood scales using identical testing methods. Adaptationand baseline sleep hours were between 2330 Visual Analogue Scale. All subjects completed three lid 0700. The next morning (TD1), subjects were in­ VAS questionnaires on TD1, four VAS questionnaires inned as to whether they had been randomized to the during T02, and four VAS questionnaires duringrecov­ �deprived (no sleep for 43 hours) or rested (7 1/2 ery day. Subjects were presented with eight questions, lwrs of sleep nightly) condition. Rested and sleep­ each pertaining to a particular mood. Below each ques­ deprived study runs were conducted separately, both tion was a 100-mm line on which they placed a mark because of the logistics involved in running both sleep indicating their self-rating on the mood in question at mnditionssimu ltaneously, and because of possible con­ the time the questionnaire was administered. Labels binding or adverse psychologic effects secondary to at the two ends of the line indicated the opposite ex­ nmning both sleep conditions simultaneously (e.g., it tremes of the mood. Mood scales on the VAS included I possible that a subject who was required to remain alertness, sleepiness, weariness, effortrequired to com­ ..ake while others were allowed to sleep would be plete tasks, happiness, sadness, tension, and calmness. Illrelik ely to report negative mood states). Sleep depri­ Possible scores ranged from 0 to 100. filion took place from 0700 hours on T01 until 2330 hours on TD2. All subjects slept from 2330 on TD2 un­ Profile of Mood States. All subjects took fIVePOMS IiITlOOon recovery day (recovery sleep). During sleep questionnaires during TD1, seven POMS question­ deprivati on and during the day after recovery sleep, naires on TD2, and eight POMS questionnaires on subjectstook repeated measures of performance and recovery day. The POMS requires a subject to rate a mod. Additionally, subjects took modifIed Multiple level of agreement or disagreement to a list of 65 ad­ SleepLate ncy Tests (MSLTs; Carskadon and Dement jectives or phrases. Mood scales on the POMS in­ 1982)as an objective measure of sleepiness. During each clude Vigor-Activity (Vigor), Fatigue-Inertia (Fatigue), lIOdif1ed MSLT, subjects were awakened as soon as Depression-Dejection (Depression), Anger-Hostility they fell asleep (within 30 seconds), to maximize the (Anger), Tension-Anxiety (Tension), and Confusion­ effects of sleep deprivation. This study routine con­ Bewilderment (Confusion). tinued for sleep-deprived subjects at a slower pace dur­ iIgthe time that rested subjects slept, to keep subjects Data Analysis IIsimilar activity levels as during the day, and at the same time minimize the possibility that testing proce­ A 2 (sleep condition: sleep-deprived/rested) x 2 (drug: duresthem selves would confound the data. During the AMPT/placebo) repeated-measures analysis of covari­ twotreatment days and recovery day, sleep-deprived ance (ANCOVA) was used to evaluate the effects of subjects completed 25 WRAIR PABs (rested subjects AMPT administration and sleep deprivation on the completed 21), 17 modifIedSLTs (rested subjects com­ alertness and mood scales of the POMS and VAS tests. pleted15), and 13 visual analogue scales (VAS) (rested Test scores in the morning (0800 to 1200 hours) of TD1, subjects completed 12). Rested subjects were not prior to drug administration, were used as covariates lWakened to collect data, and the data collected from for the analyses of the above mentioned variables. sleep-deprived subjects during the time that rested sub­ Repeated-measures analyses were conducted sepa­ jectsslept was not used in statistical comparisons. Data rately for the treatment phase of the experiment (T01 &om the cognitive tasks, sleep latency tests, and the and TD2), and for recovery day. In cases where there Stanford Sleepiness Scale (SSS) (Hoddes et al. 1973) was a signifIcantmain effectof sleep condition or drug has been reported elsewhere (McCann et al. 1992). treatment, a sleep condition x drug treatment interac- 348 U.D. McCann et al. NEUROPSYCHOPHARMACOLOGY 1993-VOL. 8, NO.1

tion or time x main effect(s) interaction(s), simple RESULTS ANCOVAs were conducted within each individual time point. When simple main effectswere observed within Results from the overall repeated-measures ANCOVA a particular time point, post hoc analyses (Duncan tests) during treatment days 1 to 2 and during recovery day were performed to determine which of the four treat­ are presented in Tables 1 and 2 respectively. Results ment groups differed signiftcantly. However, since from the simple ANCOVA and post hoc Duncan tests there are six possible post hoc comparisons at each time are reported below. point, only results of the post hoc analyses pertinent to experimental hypotheses are mentioned. Further­ Measures of Alertness and Fatigue more, due to a large number of statistical comparisons, results were considered signiftcantat a level of p < .01, Measures of Alertness. As shown in Fig. 2, scoresof and only signifIcant and near signiftcant results perti­ "alertness" from the VAS and "vigor" from the POMS nent to the discussion are reported. Finally, although indicate that: 1) during TD2, decreased alertness was more measurements of the POMS were taken during observed due to both sleep deprivation and AMPTad· the experiment, the data analyses include only data­ ministration; 2) the effectof AMPT on alertness is more points when both the POMS and the VAS were admin­ pronounced than the effect of sleep deprivation (e.g., istered (Of note, analysis including all time points the AMPT/R group differedsignifIcantly from the P/SD yielded the same results as those reported here). All group between 1200 of TD2 and 1200 of the recovery statistical analyses were performed using Statistical day in the VAS alertness scale); 3) AMPT plus sleep Analysis Systems (SAS, Cary, NC). deprivation led to greater decreases in alertness than

Table 1. Summary of AMPT Effects, Sleep Deprivation Effects, and AMPT by Sleep Deprivation Interaction Effects Across Time During Treatment Days 1 and 2 AMPT by AMPT Time SD by Time Time Type of Effect SD Effect Effect Effect Effect Effect df = 1,35 df = 1,35 df = 5, 175 df = 5, 175 df = 5, 175 Alertness scale VAS F = 36.1** NS F = 13.5** F = 3.9** NS Vigor scale POMS F = 22.2** F = 11.1** F = 6.9** NS NS Sleepiness scale VAS F= 23.1** F = 7.4** F = 11.5** F = 8.2** F = 13.9** Weariness scale VAS F = 21.6** F = 16.5** F = 8.1** F = 8.3** F = 10.1** Effort scale VAS F = 5.6* NS F = 9.6** F = 6.9** F = 16.1** Fatigue scale POMS F = 12.8** F = 14.6** F = 6.6** F = 5.8** F = 5.3** Sadness scale VAS NS NS F = 2.4* F = 5.2** F = 5.2** Tension scale VAS NS F = 5.1* F = 2.3* F = 3.1** F = 9.1** Depression scale POMS F = 8.7** 9.8** 7.0** 5.8** NS Tension scale POMS F = 5.8* F = 7.8** F = 4.7** F = 4.9** F = 4.8** Anger scale POMS F = 8.5** F = 7.5** F = 2.9* NS NS Happiness scale VAS F = 13.2** F = 5.1* F = 2.7* NS F = 9.3** Calmness scale VAS NS F = 7.0* F = 2.4* F = 2.7* NS

• Approaching signifIcant effects, p < .05. *. SignifIcant effects, p < .01. Abbreviation: SD, sleep deprivation. Note: No signifIcant AMPT by SD or AMPT by SD by time effect was observed. NEUROPSYCHOPHARMACOLOGY 1993-VOL. 8, NO.4 AMPT and Sleep Deprivation in Normal Volunteers 349

Table 2. Summary of AMPT Effects, Sleep Deprivation Effects, and AMPT by Sleep Deprivation Interaction Effects Across Time During Recovery Days AMPT by AMPT Time SD by Time Time Type of Effect SD Effect Effect Effect Effect Effect df = 1,35 df = 1,35 df = 3,105 df = 3,105 df = 3,105 Alertness scale VAS F = 8.0** NS F = 2.8** NS NS Vigor scale POMS F = 11.6** NS NS NS NS Sleepiness scale VAS NS NS F = 5.2 NS NS Weariness scale VAS NS NS F = 5.3** NS NS Effort scale VAS F = 11.3** NS NS NS NS Fatigue scale POMS F = 8.2** NS NS NS NS Sadness scale VAS NS NS NS NS NS Tension scale VAS NS NS NS NS NS Depression scale POMS F = 5.9* NS NS NS NS Tension scale POMS NS NS F = 3.7* NS F = 4.9** Anger scale POMS NS NS NS NS NS Happiness scale VAS F = 4.1* NS F = 2.7* NS NS Calmness scale VAS NS NS NS NS NS

* Approaching signifIcant effects, p < .05. ** SignifIcant effects, p < .01. Abbreviation: SO, sleep deprivation. Note: No signifIcant AMPT by SO or AMPT by SO by time effect was observed.

either treatment alone; 4) the AMPT effect persisted Mood Measurements during the recovery day; and 5) the effects of AMPT Measures of Negative Mood. As shown in Fig. 4 and (II alertness are not observed until T02. 5, scores of "sadness" and "tension" from the VAS, and Mtuuresof Sleepiness and Fatigue. As shown in Fig. scores of "tension," "depression," and "anger" from the 3, scoresof "sleepiness," "weariness," and "effort" from POMS indicate that: 1) signihcant main effectsof AMPT !he VASas well as scores of "fatigue" from the POMS administration and sleep deprivation on negative mood (notshown in Fig. 3) indicate the following: 1) an in­ scales observed in TD2 and recovery day are largely a crease in fatigue was observed following both sleep reflection of high scores reported by the AMPT/SO deprivationand AMPT administration; 2) during T02, group at most time points; 2) in both the AMPTIR group subjectsin the AMPT/SOgroup had the highest scores and the P/SO group, there was a trend for increases onall of the sleepiness scales used; 3) on some scales in negative mood that did not reach statistical [Le., effo rt and sleepiness scales of the VAS) the effect signihcance, suggesting that for the short duration of rI AMPTper se was more pronounced than the effect this study, neither treatment alone was sufficientto al­ rlsleepdeprivation per se (i.e., signi&cant differences ter negative mood states in the subjects; and 3) drug wereobse rved between the AMPT/R and P/SO groups effects on negative mood states may persist after the duringTD2 and recovery day on many of the data termination of the drug administration. points); 4) the effects of AMPT on fatigue persisted af­ Ierthe termination of drug administration; and 5) no Measures ofPositive Mood. As shown in Fig. 6, scores difference among the groups was observed during TD1. on the "happiness" and "calmness" scales of the VAS 350 U. D. McCann et al. NEUROPSYCHOPHARMACOLOGY 1993-VOL. 8, NO.4

A. AMPT Adminls�ation 20 I Sleep· Deprivation

• 4) 15 U)u 4) '" :> U) 10 � 0 IL 5

I-- Treatment Day t ---1 r--- Treatrooot Day 2 ----l f-- RacovDry Day ----1

B. AMPT Administration ------I 100 Sleep·Deprivation ------1

.. 8 80 U) ...... 60 Figure2. Self-rated alertness S • using a VAS (Monk 1989) and Ci U) 40 vigor using the POMS (McNair « et al. 1971). Thenumbers 1 and > 2 denote signifIcant simple 20 main effects (p < .01) of drug and sleep condition, respec­ 2210 tively, within an individual B. .. lln. 1800 2200 0100 1200 IUD 2200 0800 1200 1600

time point. The numbers 1* I- Treatment Day 1 ---l I-- TrealroootDay2 ----l f-- Recovery Day ----1 and 2* denote relevant near Time signifIcant simple main effects ---tr- Ploubo-R• • tod (p < .05) of drug and sleep con­ ---.- AMPT·Re.tod dition, respectively, with an in­ --0- Plocebo-Sleep-Oepri.ed dividual time point. ---.- AMPT·SIHp-Ooprived

indicate the following: 1) signifIcant main effects of tions leads to increased subjective sleepiness, and are AMPT administration and sleep deprivation on posi­ consistent with the view that prolonged wakefulness tive mood measures observed during TD2 mainly stem is associated with a functional decrease in catechol­ from the fact that the AMPTfSDgroup had lower scores aminergic neurotransmission. The current results con­ than the other groups at most time points; 2) the lower, fumprevious informal observations in patient popula­ although not signifIcant, scores in the happiness scale tions, where AMPT was observed to produce sedation of the VAS in the AMPTfR and PfSD versus the PIR (Engelman et a1. 1968; Jonsson et a1. 1971; Ahleniuset during TD2, and the approaching significant AMPT al. 1973; Pozuelo 1976; Nasrallah et al. 1977), and ex­ effects during the recovery day on this scale may indi­ tend the observation to healthy populations. Thepres­ cate that AMPT administration and sleep deprivation ent results are also in concordance with data collected per se decrease positive mood; and 3) AMPT effectson using the modifIed MSLT and the SSS (McCann et aI. positive mood may persist after the termination of the 1992). The ftndingthat combined treatment with AMP!' drug administration. and sleep deprivation produced greater changes insu� jective measures of sleepiness than either treatment DISCUSSION alone supports the notion that both treatments inter­ act with CNS catecholamines. The ftndings of this study indicate that, in healthy in­ The present results further indicates that AMPr­ dividuals, a decline in CNS catecholamine concentra- induced catecholamine depletion leads to negative NEUROPSYCHOPHARMACOLOGY 1993-VOL. 8, NO.4 AMPT and Sleep Deprivation in Normal Volunteers 351

AMPT Administralion Sktep-DopnvaTion

U •• , . lUI ....., .. , . .. 220' .... UIO '"' 22 ...... I-- 1IoawWy!loy --I t- T_""" ----i t--- Tr"'menI�2 -----i

B. AMPTAdmioisiraillOn I.

1 • I • • : •

•

2201 _11M 11" 22" IIU 121. , ... 21" .... 12 .. UIO

1IoawWy!loy --l I-- r_Dey' � t----- T,_!loy2 --I I--

C. ANPTA(fminislraleon .1 SIeep·Deprwalon 1.2 1.2

j •

• I Figure3. Self-ratedsleepiness, wear­ • iness, and effortrequired to complete • S a task using a V AS (Monk 1989). The • numbers 1 and 2 denote significant • simple main effects (p < .01) of drug and sleep condition, respectively, 1211 "" 22 ...... n .. lU' lU' ..- "" 2211 .... within an individual time point. The T_Doy2 --I I-- 1IaaMIy!loy --l I-- r_Doy' --l t----- numbers 1* and 2* denote relevant near significant simple main effects Time -t!r- PlM;illlaA ___ ---.- AM'1'_ (p < .05) of drug and sleep condition, -0-- _Sto.pDoplwod respectively, with an individual time _ AM'1'.sIoop-I)opriMd point.

IIKlOdhealthy in sleep-deprived individuals. In partic­ ing that the two treatments interact with a common neu­ aIir,normal subjects treated with AMPTand subjected ral substrate (i.e., eNS catecholamines). �40.5hours of total sleep deprivation report increased Although the effects of AMPT on mood in healthy dtpression, anger, confusion, tension, and anxiety, as individuals have not been previously reported, there wellas decreased calmness and happiness. Interest­ have been observations on AMPT's effects in patient q1y,treat ment with AMPT or sleep deprivation alone populations. In a study of 52 patients with a variety of led onlyto mild and inconsistent changes in mood. It medical conditions, it was noted that some patients de­ wasonly when the two treatments were combined that veloped an "anxiety state" (Engleman et al. 1968b). ignibcant mood effects were observed, further suggest- When given to unmedicated patients with bipolar affec- 352 D.D. McCann et aI. NEUROPSYCHOPHARMACOLOGY 1993-VOL. 8, NO.1

A. 1------AMPTAdminiSY"ioo ------l

50 I�------S'eep-Depov,'ion ------l

.. 2 o 40 u I/)

...... 30 I: ." .. U) U) 20 c( >

10

- 0 �=.�.8 o�, i n�o��t �O� �--O t6� --H0 -O.�O- -�t2� �-t6�O- �tI

I-- TrealmenlDay t ---1 I----- TrealmenlDay 2 ----I I--- RecoveryDay ----l

B. t----- AMPT Adminislration ------l 60 t------Sleep-Deprivalion ------

.. 5 o 0 u U) I: 40 o ii I: � 30 I/) c( > 20 Figure4. Self-rated sadness and tension over time using a V AS 10 (Monk 1989). The numbers 1 and 2 denote signifIcant simple main � 0 8=.:"�lI �t�2 --�2� �-1&�O-D� -22 effects (p < .01) of drug and sleep TrealmenlDay 1 Trealmenl Day condition, respectively, within an /- ---1 I-- 2 ----I I---- RecoveryDay -I individual time point. The num­ Time bers 1* and 2* denote relevant ----tr- PI.cebo-Resled ----.- AMPT·Rosl.d near signifIcant simple main ef­ --0- Plocobo-Sleep-D.p'ivod fects (p < .05) of drug and sleep ____ AMPT·SI.ep-Oeprivod condition, respectively.

tive disorder, AMPTdecreased manic symptoms in fIve where norepinephrine levels reach control values 24 of seven patients, and increased depression in three of hours after the last AMPT dose, and dopamine levels three depressed patients (Bunney et al. 1971). These reach control values 48 hours after the last AMPTdose reports, taken with the fIndings of the present study, (Rech et al. 1966). Second, it is possible that the combi­ lend support to the view that increases in functional nation of AMPT and sleep deprivation produced a catecholamine activity are associated with improved general dysphoria, leading to a negative response bias mood, and decreases in catecholamine activity with w here all negative mood states were given high scores. negative mood states. Nevertheless, the fact that only the combination treat­ Three aspects of the present fIndingsdeserve com­ ment leads to dysphoria and the fInding that dysphoric ment. First, it could be that the effectsof AMPT are due mood parallels the time course of catecholamine deple­ to nonspecifIcside effects of the drug. Arguing against tion, implicates catecholamines in the dysphoria pro­ this possibility is the fInding that the time course of duced during sleep deprivation. Third, although the AMPT -related decrements in alertness and mood par­ actions of AMPT are specifIc to tyrosine hydroxylase allel the time course of catecholamine depletion (En­ (Nagatsu et al. 1964; Spector et al. 1965), and thus do gelman et al. 1968; Rech et al. 1966). In addition, the not directly influence metabolism of other neurotrans­ return of mood and alertness measures to baseline lev­ mitters, AMPT'sactions on catecholaminergic systems els paralleled the time course of brain catecholamine could indirectly influence other neurotransmitters or recovery, as demonstrated in preclinical animal models, neuropeptides (e.g., serotonin, acetylcholine, cholecys- NEIJIOPSYCHOPHARMACOLOGY 1993-VOL. 8, NO.4 AMPT and Sleep Deprivation in Normal Volunteers 353

A. AMPT AdmlniSlration IOrfl------SI99p·09prlvatIGn

".2'

, ,2' 1,2 I r c

1200

I--- RocovO<)' [)o, ---I

A.t.tPT�mlnlstrallon B.

10 1------$198p·OeptlvatiOl'l ------i 1

11kinin,neuro peptide Y). As such, secondary changes sant effect of sleep deprivation in depressed patients lJotherneurotransmitters and neuromodulators may is related to changes in catecholamine neurotransmis­ i1Dbeinvolvedin the AMP'T-inducedchanges in mood. sion (Siegel and Rogawski 1988). For example, it has Althoughthe relationship between mood and alert­ been postulated that downregulation of catecholamine lU5iscomplex, evaluation of the effectsof sleep depri­ neurotransmission caused by sleep deprivation (Stern mand AMPTin depressed patients could shed light and Morgane 1974; Mogilnicka et al. 1986; Radulovacki tnlheextentof catecholamine involvement in the ther­ and Micovic 1983) leads to increased catecholamine !plIdiceffects of sleep deprivation. SpecifIcally, AMPT turnover (Siegel and Rogawski 1988; Reisine 1981; aid usedbe to test the hypothesis that the antidepres- Schildkraut and Hartmann 1972), which in depressed 354 U.D. McCann et al. NEUROPSYCHOPHARMACOLOGY 1993-VOL. 8, NO.1

A. 1------AMPT Adminis�allon ------I 90 Sleep·o.privallon ------1

.. 80 (; u rn ., OJ 70 .. c: E iii 60 0 VI « > 50

40

30 all.llne 1800 2200 OaDO 1200 noD 2200 oaoo 1200 IUD 2/11

r Trealmeol Day 1 -----l I--- Trealmenl Day 2 ---l I--- RocoveryDay --1

1------AMPT Adminislralion ------I B. 90 Sleep-Deprivalion ------1

.. (; 80 u VI OJ 70 OJ .. I: 'ii 60 CL .. Figure6. Self-rated calmness J: 50 VI and happiness using a VAS « (Monk 1989). The numbers 1 > 40 and 2 denote Significant simple 30 1,2 1 ' main effects (p < .01) of drug 1.2' and sleep condition, respec­ 20 BneUne 1800 2200 0800 1200 1600 2200 0800 1200 1600 2111 tively, within an individual

time point. The numbers 1* r Trealmonl Day 1 -----l I--- Trealmenl Day 2 ---l I--- RecoveryDay ------1 and 2* denote relevant near Time signifIcantsimple main effects ---tr- Plocebo·R.sled (p < .05) of drug and sleep con­ ------.- AMPT·Reslod dition, respectively, with an in­ -0- Pllcobo.Sloop·D.ptlvtd -...- AMPT-SI.ep-Dop,lvod dividual time point.

individuals, leads to improved mood. If the antidepres­ amines play a role in alertness, and suggests that cate­ sant effectof sleep deprivation requires increased cate­ cholamine systems are involved in mood changes as­ cholamine turnover (or indeed, requires normal levels sociated with sleep deprivation. of CNS catecholamines), it would be predicted that AMPTtreatment would block mood improvements sec­ ondary to sleep deprivation. ACKNOWLEDGMENT In summary, the present fmdings indicate that in healthy adults, treatment with sleep deprivation or The authors gratefully acknowledge the Student Employees AMPT produces sleepiness, and that combined treat­ of the Continuous Operations Branch for their technical as­ ment with sleep deprivation and AMPT produces sistance, and without whom this research could not have been conducted. greater sleepiness than either treatment alone. In con­ trast, only combined treatment with AMPT plus sleep deprivation leads to marked increases in anxiety, DEPARTMENT OF DEFENSE DISCLAIMER depression, and anger, with either treatment alone producing only moderate and inconsistent mood Human volunteers participated in this study after givingtheir changes. The observation that sleep deprivation and free and informed consent. Investigators adhered to AR� catecholamine depletion interact to influence alertness 25 and USAMRDC Reg 70-50 on the use of volunteers inre­ and mood is consistent with the view that catechol- search. The opinions or assertations contained herein are the NEUROPSYCHOPHARMACOLOGY 1993-VOL. 8, NO. 4 AMPT and Sleep Deprivation in Normal Volunteers 355

privateviews of the authors and are not to be construed as in affectiveillness and in schizophrenics: A comparative oIIiciaI or as reflecting the views of the Department of the analysis. In Freedman DX (ed), The Biology of the Major Anny or the Department of Defense. Psychoses. New York, Raven Press, pp 247-258 Greenspan K, Schildkraut JJ, Gordon EK, Levy B, Durell J (1969): Catecholamine metabolism in affectivedisorders. REFERENCES II. Norepinephrine, normetanephrine, epinephrine, metanephrine, and VMA excretion in hypomanic pa­ AhIeniusS, Carlsson A, Engel J, Svensson T, Sodersten FK tients. Arch Gen Psychiatry 21:710-716 (1973): Antagonism by alpha methyltyrosine of ethanol­ Hartmann E, Schildkraut JJ (1973): Desynchronized sleep and induced stimulation and euphoria in man. Clin Phar­ MHPG excretion: An inverse correlation. Brain Res 61: macol Ther 14:586-591 412-416 BeckA!, Ward CM, Mendelson M, Moch T, Erbaugh J (1961): Hoddes E, Zarcone V, SmytheH, Phillips R, Dement W (1973): An inventory for measuring depression. Arch Gen Psy­ QuantifIcation of sleepiness: A new approach. Psy­ chiatry 4:561-571 chophysiology 10:431-436 IeitsTA(198 1): Adrenoceptor drugs and sleep. In Wheatley Iqbal N, Bajwa W, Asnis G (1989): The role of norepineph­ o (ed), Psychopharmacology of Sleep. New York, Ra­ rine in depression. Psychiatry Ann 19:354-359 ven Press, pp 199-212 Jonsson LE, Anggard E, Gunne LM (1971): Blockade of intra­ Brodie HKH,Murphy DL, Goodwin FK, Bunney WE (1971): venous amphetamine euphoria in man. Clin Pharmacol Catecholamines and mania: The effect of alpha-methyl­ Ther 12:889-896 para-tyrosine on manic behavior and catecholamine me­ tabolism. Clin Pharmacol Exp Ther 12:218-224 Kupfer DJ, Thase ME (1983): The use of the sleep laboratory in the diagnosis of affective disorders. Psychiatr Clin Bunney WE Jr, Davis JM (1965): Norepinephrine in depres­ North Am 6:3-25 sive re actions. Arch Gen Psychiatry 13:483-494 McCann UD, Penetar DM, Shaham Y, Thorne D, GillinJC, Bunney WE, Brodie HK, Murphy DL, Goodwin FK (1971): SingHC, Thomas ML, Belenky G (1992): Sleep depriva­ Studiesof alpha-methyl-para-tyrosine, L-dopa, and L-tryp­ tion and impaired cognition: Possible role of brain tophan in depression and mania. Am J Psychiatry catecholamines. Bioi Psychiatry 31:1082-1097 127:872-881 McNairDM, Lorr M, Droppleman LF (1971): Prohle of Mood lunney WEJr, Goodwin FK, Murphy DL, House KM, Gor­ States. San Diego, Educational and Industrial Testing don EK (1972): The "switch process" in manic-depressive Service illness. II. Relationship to catecholamines, REM sleep, and drugs. Arch Gen Psychiatry 27:304-309 Mendelson WB (1987): Pharmacology and neurotransmitters in sleep. In Mendelson WB (ed), Human Sleep. New lunney WE, Kopanda RT, Murphy DL (1977): Sleep and be­ York, Plenum Medical Book Co, pp 33-78 havioral changes possibly reflecting central receptor hypersensitivity following catecholamine synthesis in­ MogilnickaE, Przewlocka B, van Luijtelaar ELJM, Klimek V, hibition in man. Acta Psychiatr Scand 56:189-203 Coenen AML (1986): Effectsof REM sleep deprivation on central alpha 1- and beta-adrenoceptors in rat brain. Cmkadon MA, Dement WC (1982): The multiple sleep Pharmacol Biochem Behav 25:329-332 latency test: What does it measure? Sleep 5:S67-72 Monk TH (1989): A visual analogue scale technique to mea­ qeIman K, Horwitz D, Jequier E, Sjoerdsma A (1968): Bio­ sure global vigor and affect. Psychiatry Res 27:89-99 chemical and pharmacologic effects of alpha-methyl­ tyrosine in man. J Clin Invest 47:577-594 Nagatsu T, Levitt M, Udenfriend S (1964): Tyrosine hydrox­ ylase, the initial step in norepinephrine biosynthesis. J Gerner RH, Post RM, Gillin JC, Bunney WE (1979): Biologi­ Bioi Chern 239:2910-2917 cal and behavioral effects of one nights sleep depriva­ tion in depressed patients and normals. J Psychiatry Res Nasrallah HA, Donnelly EF, Bigelow LB, Rivera-CalirnlimL, 15:21-40 Rogol S, Rauscher FP, Wyatt RJ, GillinJe (1977): Inhibi­ CiIbergM, Akerstedt T (1981): Sleep deprivation in normals­ tion of dopamine synthesis in chronic schizophrenia. some psychological and biochemical data from three Arch Gen Psychiatry 34:649-655 studies.In Koella WP (ed), Sleep 1980: Proceedings of Newhouse PA, Belenky G, ThomasML, ThorneD, SingHC, the5th European Congress on Sleep Research. Amster­ Fertig J (1989): The effectsof d-amphetamine on arousal, dam, Karger, pp 16-22 cognition, and mood after prolonged total sleep depri­ CilinJC(1983): Sleep studies in affectiveillness: Diagnostic, vation. Neuropsychopharmacology 2:153-164 therapeutic, and pathophysiological implications. Psy­ Nicholson AN, Pascoe PA (1990):Dopaminergic transmission chiatry Ann 13:367-382 and the sleep-wakefulness continuum in man. Neu­ CilinJC, Kripke OF, Janowsky DS, Risch SC (1989): Effects ropharmacology 4:411-417 of brief naps on mood and sleep in sleep-deprived Post R, Kotin J, Goodwin FK (1976): Effectsof sleep depriva­ depressed patients. Psychiatry Res 27:253-265 tion on mood and central amine metabolism in depressed GIowinskiJ, Axelrod J, Iversen LL (1966): Regional studies patients. Arch Gen Psychiatry 33:627-632 ofcat echolamines in the rat brain. IV. Effects of drugs Pozuelo J (1976): Suppression of craving and withdrawal in onthe disposition and metabolism of 3H-norepinephrine humans addicted to narcotics or amphetamines by ad­ and3H-dopamine. J Pharmacol Exp Ther 153:30-41 ministration of alpha-methyl-para-tyrosine (AMPT) and Goodwin FK, Post RM (1975): Studies of amine metabolites 5-butylpicolinic acid (fusaric acid). Cleve Clin Q 43:89-94 356 U.O. McCann et al. NEUROPSYCHOPHARMACOLOGY 1993-VOL. 8, NO.1

Radulovacki M, Micovic N (1982): Effectsof REM sleep depri­ Siegel JM, Rogawski MA (1988): A function for REM sleey vation and desipramine on beta-adrenergic binding sites Regulation of noradrenergic receptor sensitivity. Brain in rat brain. Brain Res 235:393-396 Res Rev 13:213-233 Rech RH, Borys HK, Moore KE(1966): Alterations in behavior Spector S, Sjoerdsma A, Udenfriend S (1965): Blockadeil and brain catecholamine levels in rats treated with alpha­ endogenous norepinephrine synthesis by alpha-meth� methyl-para-tyrosine. J Pharmacol Exp Ther 153:412-419 para-tyrosine, an inhibitor of tyrosine hydroxylase. Reisine T (1981): Adaptive changes in catecholamine recep­ Pharmacol Exp Ther 147:86-95 tors in the central nervous system. Neuroscience 6:1471- Stern WC, Morgane PJ (1974): Theoretical view of REMsleef 1502 function: Maintenance of catecholamine systems in til Schildkraut JJ (1965): The catecholamine hypothesis of affec­ central nervous system. Behav BioI 11:1-32 tive disorders: A review of supporting evidence. Am J Thorne0, Genser S, SingH, Hegge F (1985): The Walter Reed Psychiatry 122:509-522 performance assessment battery. Neurobehav Tom Schildkraut JJ,Hartmann E (1972): Turnoverand metabolism Teratol 7:415-418 of norepinephrine in rat brain after 72 hours on a O-dep­ Vogel GW, Thurmond S, Gibbons R, Sloan K, WalkerN rivation island. Psychopharmacology 27:17-27 (1975): REM sleep reduction effects on depressionsyn­ dromes. Arch Gen Psychiatry 32:765-777 Schildkraut JJ,Keeler BA, Grab EL, Kantrowich 1, Hartmann E (1973a): MHPG excretion and clinical classi&cation in Wehr TA, Rosenthal NE, Sack DA, Gillin JC (1985):An­ depressive disorders. Lancet 1:1251-1252 tidepressant effectsof sleep deprivation in bright anddill SchildkrautJJ, Keeler BA, Papousek M,Hartmann E (1973b): light. Acta Psychiatr Scand 72:161-165 MHPGexcretion in depressive disorders: Relation to clin­ Weiss B, Laties VG (1962): Enhancement of human perf(J' ical subtypes and de synchronized sleep. Science 181: mance by caffeineand the amphetamines. PharmacolRel 762-764 14:1-36