Relationship between ethnicity and sleep patterns in normalcontrols: implications for psychopathology and treatment

Uma Raoa, b, c,*, Russell E. Polanda, b, c, Preetam Lutchmansingha, Geo�rey E. Otta,James T. McCrackena, b, c, Keh-Ming Lina, b

aDepartment of Psychiatry, Harbor-UCLA Medical Center, Torrance, CA, USAbDepartment of Psychiatry UCLA School of Medicine, Los Angeles, CA, USAcBrain Research Institute, UCLA School of Medicine, Los Angeles, CA, USA

Received 12 November 1998; received in revised form 15 April 1999; accepted 26 April 1999

Abstract

The purpose of the study was to examine ethnic in¯uences on sleep regulation. Seventy-three normal volunteers from fourethnic groups (17 African-Americans, 10 Asians, 30 Caucasians and 16 Hispanics) were studied for two consecutive nights withsleep polysomnography recordings in the laboratory. The subjects were in good physical and psychological health, and wereasymptomatic with respect to sleep/wake complaints or sleep disorders. With the exception of minor di�erences, sleep continuity,

sleep architecture and rapid eye movement (REM) sleep patterns were comparable among the four groups. African-Americanshad evidence of more stages 1 and 2 and diminished stage 4 sleep, whereas the Hispanics had higher REM density. Thesepreliminary ®ndings suggest that sleep patterns are remarkably similar across cultures. There are, however, important cross-

ethnic di�erences, speci®cally in the depth of sleep and in phasic REM measures. Because sleep disturbances are commonsymptoms of emotional disorders and since many psychoactive agents a�ect sleep, cross-ethnic di�erences in sleep patterns mayhave potential implications for the treatment and prevention of psychiatric disorders. # 1999 Elsevier Science Ltd. All rights

reserved.

Keywords: Ethnicity; Normal; Psychopathology; Sleep regulation; Socio-cultural; Treatment

1. Introduction

Di�culties with sleep are among the most common

symptoms of emotional disorders. For example, per-

sistent insomnia is a harbinger of anxiety, mood and

substance use disorders (Ford and Kamerow, 1989;

Breslau et al., 1996; Perlis et al., 1997). In an attempt

to identify biological markers for major psychiatric

disorders, numerous electroencephalographic (EEG)

sleep studies have been performed in the past three

decades. In 1992, Benca and colleagues published a

meta-analytic review of EEG sleep investigations from

177 controlled studies in 7151 patients and controls.

The psychiatric conditions included alcoholism, anxiety

disorders, borderline personality disorder, dementia,

eating disorders, insomnia, mood disorders, narcolepsy

and schizophrenia. Benca et al. (1992) concluded that

most psychiatric groups showed evidence of reduced

sleep e�ciency and total sleep time, speci®cally with

decrements in non-rapid eye movement sleep.

Although rapid eye movement (REM) sleep was rela-

tively well preserved in all groups, altered REM pat-

terns were observed as well, most frequently in mood

disorders.

With the identi®cation of objective sleep changes in

association with emotional disorders, investigators

have begun to use more sophisticated designs, particu-

Journal of Psychiatric Research 33 (1999) 419±426

0022-3956/99/$ - see front matter # 1999 Elsevier Science Ltd. All rights reserved.

PII: S0022-3956(99 )00019 -9

www.elsevier.com/locate/jpsychires

* Corresponding author. Present address: UCLA Neuropsychiatric

Institute, 760 Westwood Plaza, Room 68-237, Los Angeles, CA

90024-1759, USA. Tel.: +1-310-206-9478; fax: +1-310-825-2682.

E-mail address: urao@mednet.ucla.edu (U. Rao)

larly in relation to research with mood disorders. Asreviewed by Giles et al. (1998b), several potentiallyconfounding variables, including age, gender, diagnos-tic category, severity of illness, episode history, episodecourse, psychotropic medication, clinical course, stress,family history, sleep schedules, reproductive phase andseason, have been explicitly evaluated. Despite hun-dreds of reports on di�erent aspects of EEG sleep dys-regulation in psychiatric illness, ranging frominvestigations of pathophysiologic underpinnings todiagnostic implications and treatment response, veryfew studies have examined whether ethnicity a�ectsEEG sleep patterns and if sleep changes associatedwith a speci®c psychiatric condition are `universal' intheir appearance.

There is growing evidence that ethnic processesmarkedly in¯uence disease states (see Kleinman, 1981;Cruickshank and Beevers, 1989; Polednak, 1989; Berryet al., 1992). Cross-ethnic studies have shown that eth-nic groups di�er with regard to prevalence rates ofpsychiatric disorders, symptom presentation, and treat-ment response (Karno et al., 1987; Hwu et al., 1989;Sommervell et al., 1989; Kalow and Bertilsson, 1994;Kessler et al., 1994; Lin and Poland, 1995; Brown etal., 1996; Vega et al., 1998). Because sleep disturbancesincrease vulnerability to many psychiatric disorders(Ford and Kamerow, 1989; Benca et al., 1992; Breslauet al., 1996; Rao et al., 1996; Perlis et al., 1997; Gileset al., 1998a), it is important to examine the natureand extent of ethnic in¯uence(s) on sleep regulation,and the interaction(s) among ethnicity, psychopathol-ogy and sleep disturbances.

To our knowledge, three investigations have beenundertaken to determine if EEG sleep changes occurconsistently across ethnic groups (Mendlewicz andKerkhofs, 1991; Giles et al., 1998b; Poland et al.,1999). The ®rst study, a collaborative cross-ethnic con-trolled study sponsored by the World HealthOrganization (WHO), reported on EEG sleep changesin subjects with major depression at eight di�erentsites across Asia, Europe and North America(Mendlewicz and Kerkhofs, 1991). As a group, the de-pressed patients showed sleep-continuity disturbances,diminished stages 2 and 3, reduced latency to the onsetof REM sleep and increased REM density comparedto control subjects. However, some sleep measures dif-fered across sites. For example, sleep continuity dis-turbances were not prominent in patients fromSapporo, diminished stage 2 sleep was not observed insubjects from Athens and Zagreb, and reduction instage 3 sleep was not seen in patients from Brussels.Also, REM latency was not reduced in subjects fromBrussels and Tokyo. Ethnic distribution across sites,and between patients and controls within each site wasnot reported in this investigation.

The second study was a historical case-control design,

comparing EEG sleep data of African-American (AA)and Caucasian (CC) depressed patients collected in thesame laboratory (Giles et al., 1998b). These datarevealed that AA patients did not show the depression-related REM sleep stigmata to the same extent as CCsubjects despite having consonant clinical pro®le.Compared to CC, the AA demonstrated more stage 2sleep but diminished slow-wave sleep (SWS), longerREM latency, less REM sleep and lower REM density.

In a recently completed project at our center, 106 de-pressed patients from four ethnic groups (namely AA,Asian (AS), CC and Hispanic (HS)) were examined forinter-ethnic di�erences in sleep pro®les at baseline andin response to a cholinergic antagonist, scopolamine(Poland et al., 1999). Similar to the observations madeby Giles et al. (1998b), AA had less REM sleep at base-line compared to the CC. AA males, in particular, haddecreased stage 4 sleep. CC tended to have the shortestREM latency. AS had decrements in sleep e�ciency andREM sleep. The sleep patterns of HS were comparableto the CC. These ®ndings, again, support the variabilityin sleep patterns among di�erent ethnic groups of de-pressed patients.

Inter-ethnic di�erences in sleep patterns observedamong depressed patients may have methodologicaland theoretical implications. In order to better under-stand the in¯uence(s) of ethnicity on sleep changes re-lated to depressive or other psychiatric illnesses, it isimportant investigate ethnic in¯uences on normal sleepregulation (Gillin, 1994). Knowledge of normativesleep patterns among di�erent ethnic groups will helpus to control for variations in sleep measures whencomparing psychiatric patients from diverse ethnicbackgrounds, thus allowing us to make more validconclusions regarding the interactions among ethnicity,psychopathology and sleep regulation. Furthermore,knowledge of cross-ethnic di�erences in sleep patternswill help us to identify ethnic groups that may bespeci®cally at high risk for psychopathology. Also,because most psychotropic agents a�ect sleep, knowl-edge of cross-ethnic variations in sleep patterns will behelpful in developing treatment guidelines for speci®cethnic groups, and thereby improve treatment compli-ance. In order to examine the in¯uence of ethnicity onsleep regulation, the present pilot investigation wasundertaken whereby EEG sleep pro®les of normal vol-unteers from four major ethnic groups were studied atthe same site under identical conditions.

2. Materials and methods

2.1. Participants

Seventy-three normal volunteers from four ethnicgroups (17 AA, 10 AS, 30 CC and 16 HS) were stu-

U. Rao et al. / Journal of Psychiatric Research 33 (1999) 419±426420

died as part of a control group for ongoing investi-gations. The purpose of these studies is to examine thee�ect of ethnicity on clinical characteristics, psychobio-logical correlates and treatment of various psychiatricdisorders. All of these investigations were approved bythe Institutional Review Board at the Harbor-UCLAResearch and Education Institute. The AS were pri-marily of Chinese origin, and the HS of Mexican back-ground. Membership in a particular ethnic grouprequired that all four biological grandparents were ofthe same ethnicity.

2.2. Clinical assessments

All participants were evaluated for psychopathologywith the Structured Clinical Interview for DSM-III-R(SCID; Spitzer et al., 1986). Also, psychopatholgy inthe ®rst-degree relatives was assessed, using the subjectas an informant. In addition to the SCID interview,the clinician rated the severity of depressive symptomsusing the Hamilton Depression Rating Scale (HDRS;Hamilton, 1960). Subjects were included in the study ifthey and their family members were free from currentor previous psychiatric illness, and had a total HDRSscore of 5 or less. All subjects were free from medi-cation use for at least two weeks. Participants weremedically healthy, as determined by physical examin-ation, electrocardiogram, full chemistry panel, thyroidfunction tests and urine drug screens. In order to ruleout known sleep disorders, a sleep questionnaire was®lled-out and a sleep log was maintained for at leastone week prior to the study. Shift workers wereexcluded from the study. Subjects with a personal his-tory of a major sleep disorder or a family history ofnarcolepsy were excluded from the study. Participantswere screened also for the presence of sleep disorder(s)on the ®rst night of the sleep protocol. None of thesubjects was aware of the inclusion/exclusion criteria.

2.3. Sleep protocol and scoring of sleep records

EEG sleep was recorded for two consecutive nightsin the laboratory. Although most subjects normallyretired between 10.00 p.m. and midnight and awokebetween 5.30 and 8.00 a.m., participants wererequested to go to bed between 10.30 and 11.30 p.m.

and awake between 6.30 and 7.30 a.m. for at least oneweek prior to the study. On both nights, conventionalEEG electrodes were attached by 9.00 p.m., and sleeprecordings were made from 11.00 p.m. (lights out) to7.00 a.m. Examination of sleep logs for the week priorto the study revealed that there were no signi®cantdi�erences between home and laboratory sleep/wakeschedules.

The International 10±20 system was utilized for EEGelectrode placement, electromyogram (EMG), electroo-culogram (EOG) and electrocardiogram (EKG). Inorder to rule out the presence of sleep disorders, a fullsleep polysomnography was performed on the ®rstnight, including respiratory, oximetry and leg move-ment measurements. Data were digitized and were col-lected at a base sampling rate of 200 Hz using theSANDMAN system. The EEG, EMG, EOG and EKGdata were saved at 200, 50, 20 and 100 Hz, respectively.Bilateral EEG recordings were obtained from left (C3)and right (C4) central leads referenced to the oppositemastoid, A2 and A1, as well as to a linked reference(A1+A2). Ampli®er sensitivity settings were set at 5,low ®lter at 0.3 Hz and high ®lter at 70 Hz, with 60 Hz®lter in the `on' position.

Sleep records were coded and scored blindly accord-ing to standard criteria (Rechtscha�en and Kales,1968). REM latency was de®ned as the time betweensleep onset (®rst minute of stage 2 or deeper sleep, fol-lowed by at least 9 min of stage 2 or deeper sleep,interrupted by no more than 1 min of waking or stage1) and the ®rst REM period r3 min in length. REMlatency is reported with and without intervening waketime subtracted. Other REM measures, includingREM density and REM activity, and additional sleepvariables were scored according to the criteria ofKupfer (1976), as was done previously (Poland et al.,1989, 1997).

2.4. Statistical analysis

Statistical analyses were performed by analysis ofcovariance (ANCOVA), with age and gender as thecovariates. When there was a signi®cant e�ect of ethni-city, a least-squares means test was performed tolocate inter-ethnic group di�erences. Where necessary,the data were transformed prior to statistical analysis.

Table 1

Demographic characteristics of the four ethnic groups

African-American (N= 17) Asian (N= 10) Caucasian (N= 30) Hispanic (N= 16) Statistic P

Full sample (age) (years) 30.927.4 28.4212.4 42.2214.9 27.726.9 F= 7.33 0.0003

Females (n=38) (age) 28.725.7 34.8214.6 42.0216.7 27.727.3 F= 3.53 0.02

Males (n=35) (age) 35.028.9 24.229.8 42.0213.3 27.126.6 F= 5.22 0.005

Female/male 11/6 4/6 14/16 9/7 w 2=2.13 0.55

U. Rao et al. / Journal of Psychiatric Research 33 (1999) 419±426 421

Relationship between continuous variables was exam-ined using the Pearson product±moment correlationcoe�cient. Only polysomnography data from the sec-ond night were used in the statistical analyses, the ®rstnight being considered as an adaptation night.

3. Results

3.1. Demographic information

Demographic characteristics of the four ethnicgroups are given in Table 1. CC were signi®cantlyolder than the remaining three groups. The four ethnicgroups did not di�er signi®cantly with respect to gen-der distribution. AA males tended to be older than thefemales, whereas AS females were older than theirmale counterparts (see Table 1).

3.2. Sleep continuity measures

Sleep continuity variables for the four ethnic groupson the second night are shown in Table 2. The totalrecording time was comparable across the four groups

and, on average, it was approximately 7.5 h. Therewas marked variability in the sleep continuitymeasures. Age, however, accounted for majority ofthis variance. Sleep e�ciency, arousals and time awakeafter sleep onset were signi®cantly a�ected by age.Sleep latency was in¯uenced by gender, with femaleshaving longer sleep latencies. After controlling for ageand gender e�ects, none of the sleep continuitymeasures discriminated the four ethnic groups.

3.3. Sleep architecture variables

The various sleep stages, both in minutes and in per-centages, are presented in Table 3. Stage 1 sleep wassigni®cantly in¯uenced by age as well as by gender.Stage 4 sleep also was in¯uenced by gender. After cov-arying for age and gender, AA had more stage 1 sleepin minutes than all the other ethnic groups. Also, AAhad signi®cantly more stage 2 sleep and less stage 4sleep compared to the CC and HS. The AS scored in-termediate on these measures. On further evaluation ofcross-ethnic di�erences in non-REM sleep, it was pri-marily the AA males who showed these di�erences.

Table 2

Sleep continuity measures (mean2S.D.) in the four ethnic groups. �=p R 0.05, ��=p R 0.01, ���=p R 0.001

African-American (N= 17) Asian (N= 10) Caucasian (N= 30) Hispanic (N= 16) ANCOVA

Age Gender Ethnicity

Total study time (min) 452.6241.9 448.2219.9 458.3224.2 450.4221.0 0.54 0.80 0.17

Sleep latency (min) 20.5228.1 18.6211.4 41.3256.1 20.0217.2 0.28 4.87� 1.26

Sleep e�ciency (%) 0.92205 0.90204 0.88209 0.92204 5.21� 1.23 0.33

Number of arousals 35.8222.2 18.5210.2 52.5269.1 37.4281.1 7.48�� 0.43 0.22

Awake time (min) 29.6220.4 25.6220.6 46.7240.9 18.6210.7 11.97��� 2.03 0.65

Total sleep time (min) 416.9245.7 404.3229.2 406.4252.5 417.0226.6 1.41 0.10 0.21

Table 3

Sleep architecture variables (mean 2S.D.) in the four ethnic groups. Di�erent alphabetical superscripts depict signi®cant di�erences between

(among) groups ( p R 0.05). �=p R 0.05; ��=p R 0.0001

African-American (N= 17) Asian (N= 10) Caucasian (N= 30) Hispanic (N= 16) ANCOVA

Age Gender Ethnicity

Minutes

Stage 1 32.9216.7a 22.4211.0b 31.6217.0b 21.528.8b 26.74�� 20.13�� 3.62�

Stage 2 232.5248.7a 217.7247.0a,b 187.8241.3b 201.2236.6b 1.71 0.97 3.17�

Stage 3 21.7214.4 36.5212.3 29.2215.1 31.8218.6 1.56 0.86 2.46

Stage 4 37.5234.1a 42.3232.7a,b 60.1230.8b 59.9232.8b 0.84 6.38� 3.24�

Stage REM 92.3231.3 80.7224.5 97.6233.4 102.6219.5 0.01 0.08 1.22

Percent

Stage 1 7.924.0 5.623.0 8.124.8 5.222.1 27.84�� 16.72�� 2.49

Stage 2 55.729.1a 53.7210.2a,b 46.228.3b 48.529.8b 0.99 1.70 3.80�

Stage 3 5.423.8 9.023.0 7.324.0 7.524.4 1.30 0.87 2.24

Stage 4 9.128.2a 10.728.4a,b 14.727.7b 14.227.9b 0.40 6.35� 2.91�

Stage REM 21.826.8 20.825.2 23.526.1 24.624.3 0.08 0.07 1.09

U. Rao et al. / Journal of Psychiatric Research 33 (1999) 419±426422

Stage 4 sleep in all four ethnic groups by gender isdepicted in Fig. 1.

3.4. REM sleep measures

REM sleep variables for the four ethnic groups areshown in Table 4. With the exception of total REMdensity, which was higher in the HS, the four groupswere not signi®cantly di�erent on REM sleep vari-ables. Both HS males and females showed evidence ofhigher REM density than AA and CC. Again, AS sub-jects fell in the middle on REM density. REM latencytended to be the shortest in CC, particularly when theintervening wake time was subtracted ( p R 0.06).

3.5. Relationship between age and sleep variables

As noted above, sleep continuity measures and stage1 sleep were signi®cantly in¯uenced by age. In order tocon®rm the ®ndings that the other sleep measures were

not e�ected by age, the older CC subjects wereremoved and the sleep variables reanalyzed, and theresults were essentially the same. Moreover, examin-ation of Pearson correlation coe�cients between ageand other sleep variables revealed only modest corre-lations (r values ranging from 0.004 to ÿ0.25).

3.6. Relationship between depressive symptoms and sleepvariables

None of the sleep measures were in¯uenced by theHDRS score in the total sample or in individual ethnicgroups.

4. Discussion

To our knowledge, this is the ®rst study to examineEEG sleep patterns in normal volunteers from diverseethnic groups at a single site under identical con-ditions. In the present investigation, comparison ofEEG sleep pro®les of four major ethnic groupsrevealed that, overall, the sleep patterns were remark-ably similar. There were, however, important minordi�erences. Speci®cally, AA males di�ered from othermales as well as from AA females in non-REM sleeppatterns, with increases in stages 1 and 2, and dimin-ished stage 4 sleep. HS had higher REM density.REM latency tended to be the shortest in CC.

These ®ndings should be interpreted in the contextof the study limitations. The samples, speci®cally inthe ethnic-minority groups, were of modest sizes, thuslimiting statistical power in detecting other potentialcross-ethnic di�erences in EEG sleep patterns.Particularly, the ethnicity � gender interactions shouldbe considered a very preliminary ®nding in view of therelatively modest samples. The samples varied signi®-cantly with respect to age distribution. It is to be

Fig. 1. Stage 4 sleep (mean2SEM) in African-American, Asian,

Caucasian and Hispanic male and female normal volunteers.

Table 4

REM sleep measures (mean2S.D.) in the four ethnic groups. Di�erent alphabetical superscripts depict signi®cant di�erences between (among)

groups ( p R 0.05). �=p R 0.05

African-American (N= 17) Asian (N= 10) Caucasian (N= 30) Hispanic (N= 16) ANCOVA

Age Gender Ethnicity

REM latency (min) 86.6247.7 79.5238.5 65.2227.8 94.9230.2 0.07 0.32 1.99

REM latency-W (min) 83.8245.6 77.1237.9 60.2223.6 93.6230.7 0.29 0.52 2.60

First REM period

REM activity (units) 23.7221.7 25.8225.4 33.8228.0 36.9233.2 2.20 1.65 0.89

REM density (units/min) 1.420.7 1.621.2 1.620.6 1.820.9 1.35 2.13 0.94

REM duration (min) 13.228.5 13.227.1 20.3212.6 17.8210.1 3.00 1.11 1.26

All REM episodes

REM activity (units) 193.52125.4 219.52122.4 200.6286.3 286.42110.7 0.89 0.19 2.07

REM density (units/min) 1.921.0a 2.520.8a,b 2.020.6a 2.720.8b 1.47 0.29 2.99�

REM duration (min) 95.2233.8 84.7222.9 100.1235.1 103.8221.2 0.13 0.66 0.89

# REM episodes 4.221.0 4.020.7 4.420.9 4.120.6 0.46 0.87 0.82

U. Rao et al. / Journal of Psychiatric Research 33 (1999) 419±426 423

noted that the participants in this study were notmatched on important demographic characteristics,such as age, gender, education and socioeconomic sta-tus. As described above, the primary focus of the orig-inal investigations was on ethnic in¯uence(s) onpsychobiological processes. Attempts were made tomatch these participants as controls for the experimen-tal groups, and not with each other. As these studiesprogressed, we became aware of considerable variabil-ity in sleep measures across ethnic groups. These ana-lyses were undertaken in order to better understandthe ethnic di�erences observed among depressedpatients by us and by others. These ®ndings should beconsidered as an initial step for a more rigorous inves-tigation of ethnic in¯uences on normal sleep regu-lation. Also, information on ethnic heritage and familyhistory was obtained only from the subject. These datado not help us to di�erentiate between genetic/biologi-cal and socio-cultural factors associated with inter-eth-nic di�erences in EEG sleep pro®les; it is likely thatboth may interplay. Finally, the clinical signi®cance ofcross-ethnic di�erences in EEG sleep, includingreduced stage 4 sleep and higher REM density in oneor more ethnic groups in comparison with others, isnot clear.

The two studies examining ethnic in¯uences on EEGsleep have reported increased stage 2 and diminishedSWS in AA patients with depression, and relativelynormal REM sleep patterns (Giles et al., 1998b;Poland et al., 1999). In contrast, reduced REM latencyand other REM sleep changes were notable in the CC(Giles et al., 1998b). In the study by Poland et al.(1999), both AA and AS had signi®cantly less REMsleep compared to the CC and HS. In the present in-vestigation, although the AS were not signi®cantlydi�erent from the CC on any of the EEG sleepmeasures, their sleep pro®le more closely resembled theAA. However, it should be noted that, while not sig-ni®cant, both AA and AS tended to have less REMsleep than CC and HS. This tendency was magni®edin depressed patients, such that REM sleep was signi®-cantly less in AA and AS compared to the CC (Polandet al., 1999). With the exception of REM latency andREM density, the CC and HS had similar EEG sleeppro®les.

Kupfer and Ehlers (1989) have hypothesized twoseparate mechanisms for reduced REM latency in de-pression. In one group, there may be a weakened SWSprocess that is manifested especially during the ®rstnon-REM period. Here, the REM latency appears tobe altered by the impact of developmental changes inSWS in the ®rst sleep cycle. In the second case,reduced REM latency may be more closely related tochanges in arousal and REM sleep physiology itself.Increased REM sleep pressure may lead to advance-ment of REM sleep as well as other altered REM pat-

terns. It is speculated that AA have a weakened SWSgenerator, whereas the CC have increased REM sleeppressure. Reduction in SWS in the AA may be com-pensated by increases in stages 1 and 2. Longitudinalstudies of depressed and high risk populations fromthese two ethnic backgrounds will be helpful in testingthese hypotheses. REM latency tended to be the short-est in CC, whereas REM density was signi®cantlyhigher in the HS. The di�erences in tonic and phasicREM sleep measures in the two groups suggest thatthese two processes may be controlled by di�erentneuro-regulatory systems (Siegel, 1994; Cabeza et al.,1994; Poland et al., 1997).

Cross-ethnic variations in sleep pro®les have poten-tial implications for the treatment of depressive illnessin patients from di�erent ethnic backgrounds (Thase,1998). Di�erent antidepressants have distinctly di�er-ent e�ects on sleep (Ware and Pittard, 1990; Vogel etal., 1990, 1998; Nofzinger et al., 1995; Sharpley andCowen, 1995; Sharpley et al., 1996; Rush et al., 1998).For instance, selective serotonin reuptake inhibitors(SSRIs) strongly suppress REM sleep, and also mayworsen sleep maintenance and reduce SWS (Sharpleyet al., 1996; Rush et al., 1998). In contrast, nefazodoneand bupropion tend to preserve REM sleep and mayimprove sleep e�ciency (Nofzinger et al., 1995;Sharpley et al., 1996; Rush et al., 1998; Vogel et al.,1998). Other antidepressants, including tricyclic agentsand trazodone, may enhance SWS (Montgomery et al.,1983; Ware and Pittard, 1990; Reynolds et al., 1991).Di�erential e�ects of antidepressants on sleep architec-ture may have implications for di�erent treatmentstrategies based on EEG sleep pro®les. According tothe available data, AA and AS have less REM sleepcompared to the CC and HS. It is possible that thesetwo ethnic groups may be more sensitive to anti-depressants that robustly suppress REM sleep. AAalso may be sensitive to agents which reduce SWS.There is evidence that AA and AS are more sensitiveto antidepressants (Kleinman, 1981; Lin and Poland,1995; Lesser et al., 1997;Raskin and Crook, 1975;Rosenblat and Tang, 1987). Controlled studies shouldaddress this clinically relevant question in the future.Ethnically-sensitive treatment guidelines based onEEG sleep pro®les may reduce treatment-related sleepdi�culties and enhance treatment compliance(Kleinman, 1981; Marcos and Cancro, 1982; Lin andPoland, 1995; Lesser et al., 1997).

Cross-ethnic di�erences in sleep regulation may berelevant for understanding vulnerability and protectivefactors for speci®c psychiatric illnesses. Both sleepcomplaints and objective sleep measures have beenshown to increase vulnerability to psychiatric illness,depression in particular (Giles et al., 1988a; Ford andKamerow, 1989; Benca et al., 1992; Breslau et al.,1996; Rao et al., 1996; Perlis et al., 1997).

U. Rao et al. / Journal of Psychiatric Research 33 (1999) 419±426424

Epidemiological studies have indicated that rates ofselected psychiatric disorders di�er across ethnicgroups. For example, AA, Chinese and Mexicans havereduced prevalence of major depression (Karno et al.,1987; Sommervell et al., 1989; Hwu et al., 1989;Kessler et al., 1994; Takeuchi et al., 1998; Vega et al.,1998). In contrast, AA and Chinese may be more vul-nerable to anxiety disorders (Hwu et al., 1989; Kessleret al., 1994). It is possible that genetic factors in con-cert with socio-cultural forces may determine sleepregulation in di�erent ethnic groups (Partinen et al.,1983; Linkowski et al., 1991) and speci®c ethnic groupsmay be susceptible to speci®c illnesses (Karno et al.,1987; Vega et al., 1988; Sommervell et al., 1989;Kessler et al., 1994; Takeuchi et al., 1998). In order todevelop more e�ective treatment and preventive inter-ventions, it is important to examine the role of ethnicprocesses on sleep regulation and psychopathology.

In summary, the EEG sleep pro®les of AA, AS, CCand HS in this preliminary investigation were, for thelarge part, similar. However, AA showed evidence ofmore stages 1 and 2 and less stage 4 sleep, whereas theHS had higher REM density. CC tended to have theshortest REM latency. Because sleep disturbancesincrease vulnerability for commonly occurring psychia-tric disorders, cross-ethnic di�erences in sleep patternsmay have potential implications for the treatment andprevention of such conditions. In order to betterunderstand the role of sleep dysregulation in psycho-pathology, future investigations should focus on theinterplay between genetic/biological and socio-culturalfactors Among di�erent ethnic groups.

Acknowledgements

This study was supported in part by the NIMHResearch Center on the Psychobiology of EthnicityMH47193, MH34471, NIMH Scientist DevelopmentAward MH01419 (to U.R.), NIMH Research ScientistDevelopment Award MH00534 (to R.E.P.), NIHGeneral Clinical Research Center Grant RR00425 andGlaxo-Wellcome. The authors thank MannyMadamba for technical assistance, Gwen Gordon forstatistical analysis and Debbie Hanaya for administra-tive support.

References

Benca RM, Obermeyer WH, Thisted RA, Gillin JC. Sleep and psy-

chiatric disorders: a meta-analysis. Arch Gen Psychiatry

1992;49:651±68.

Berry JW, Poortinga YH, Segall MH, Dasen PR. Cross-cultural psy-

chology: research and applications. Cambridge: Cambridge

University Press, 1992.

Breslau N, Roth T, Rosenthal L, Andreski. Sleep disturbance and

psychiatric disorders: a longitudinal epidemiological study of

young adults. Biol Psychiatry 1996;39:411±8.

Brown C, Schulberg HC, Madonia MJ. Clinical presentations of

major depression by African-Americans and Whites in primary

medical care practice. J A�ect Disord 1996;41:181±91.

Cabeza RDJ, Zoltoski RK, Gillin JC. Biochemical pharmacology of

sleep. In: Chokroverty S, editor. Sleep disorders medicine.

Boston: Butterworth-Heinemann, 1994. p. 37±56.

Cruickshank JK, Beevers DG. Ethnic factors in health and disease.

London: Wright, 1989.

Ford DE, Kamerow DB. Epidemiologic study of sleep disturbances

and psychiatric disorders: an opportunity for prevention? J Am

Med Assoc 1989;262:1479±84.

Giles DE, Kupfer DJ, Rush AJ, Ro�warg HP. Controlled compari-

son of electrophysiological sleep in families of probands with uni-

polar depression. Am J Psychiatry 1998a;155:192±8.

Giles DE, Perlis ML, Reynolds CF, Kupfer DJ. EEG sleep in

African-American patients with major depression: a historical

case control study. Depression Anxiety 1998b;8:58±64.

Gillin JC. Sleep: a royal road to pathophysiology? J Psychiatr Res

1994;28:189±94.

Hamilton M. A rating scale for depression. J Neurol Neurosurg

Psychiatry 1960;25:56±62.

Hwu HG, Yeh EK, Chang LY. Prevalence of psychiatric disorders

in Taiwan as de®ned by the Chinese Diagnostic Interview

Schedule. Acta Psychiatr Scand 1989;79:136±47.

Kalow W, Bertilsson L. Inter-ethnic factors a�ecting drug response.

Adv Drug Res 1994;25:2±52.

Karno M, Hough RL, Burnam A, Escobar JI, Timbers DM,

Santana F, Boyd JH. Lifetime prevalence of speci®c psychiatric

disorders among Mexican-Americans and non-Hispanic Whites in

Los Angeles. Arch Gen Psychiatry 1987;44:695±701.

Kessler RC, McGonagle KA, Zhao S, Nelson CB, Hughes M,

Eshleman S, Wittchen HU, Kendler KS. Lifetime and 12-month

prevalence of DSM-III-R psychiatric disorders in the United

States. Arch Gen Psychiatry 1994;51:8±19.

Kleinman A. Culture and patient care: psychiatry among the

Chinese. Drug Ther 1981;11:134±40.

Kupfer D. REM latency: a psychobiologic marker for primary de-

pressive disease. Biol Psychiatry 1976;11:159±74.

Kupfer DJ, Ehlers CL. Two roads to rapid eye movement latency.

Arch Gen Psychiatry 1989;46:945±8.

Lesser IM, Smith M, Poland RE, Lin K-M. Psychopharmacology

and ethnicity. In: Friedman S, editor. Cultural issues in the treat-

ment of anxiety disorders. New York: Guilford Press, 1997. p.

199±225.

Lin K-M, Poland RE. Ethnicity, culture and psychopharmacology.

In: Bloom FE, Kupfer DJ, editors. Psychopharmacology: the

fourth generation of progress. New York: Raven Press, 1995. p.

1907±17.

Linkowski P, Kerkhofs M, Hauspie R, Mendlewicz J. Genetic deter-

minants of EEG sleep: a study in twins living apart.

Electroencephalogr Clin Neurophysiol 1991;79:114±8.

Marcos LR, Cancro R. Pharmacotherapy of Hispanic depressed

patients: clinical observations. Am J Psychother 1982;36:505±12.

Mendlewicz J, Kerkhofs M. Sleep electroencephalography in depress-

ive illness: a collaborative study by World Health Organization.

Br J Psychiatry 1991;159:505±9.

Montgomery I, Oswald I, Morgan K, Adam K. Trazodone enhances

sleep in subjective quality but not in objective duration. Br J Clin

Pharmacol 1983;16:139±44.

Nofzinger EA, Reynolds CF, Thase ME, Frank E, Jennings JR,

Fasiczka AL, Sullivan RL, Kupfer DJ. REM sleep enhancement

by bupropion in depressed men. Am J Psychiatry 1995;152:274±6.

Partinen M, Kaprio J, Koskenvuo M, Putkonen P, Langinvainio H.

Genetic and environmental determination of human sleep. Sleep

1983;6:179±85.

U. Rao et al. / Journal of Psychiatric Research 33 (1999) 419±426 425

Perlis ML, Giles DE, Buysse DB, Tu X, Kupfer DJ. Self-reported

sleep disturbance in recurrent depression: preliminary evidence

that insomnia may be a prodromal symptom of depression. J

A�ect Disord 1997;42:209±12.

Poland RE, McCracken JT, Lutchmansingh P, Lesser IM, Tondo L,

Edwards C, Boone KB, Lin K-M. Di�erential response of rapid

eye movement sleep and cholinergic blockage by scopolamine in

currently depressed, remitted, and normal control subjects. Biol

Psychiatry 1997;41:929±38.

Poland RE, Rao U, Lutchmansingh P, McCracken JT, Lesser IM,

Edwards C, Lin K-M. REM sleep in depression is in¯uenced by

ethnicity. Submitted for publication.

Poland RE, Tondo L, Rubin RT, Trelease RB, Lesser IM.

Di�erential e�ects of scopolamine on nocturnal cortisol secretion,

sleep architecture and REM latency in normal volunteers: re-

lation to sleep and cortisol abnormalities in depression. Biol

Psychiatry 1989;25:403±12.

Polednak AP. Racial and ethnic di�erences in disease. New York:

Oxford University Press, 1989.

Rao U, Dahl RE, Ryan ND, Birmaher B, Williamson DE, Giles

DE, Rao R, Kaufman J, Nelson B. The relationship between

longitudinal clinical course and sleep and cortisol changes in ado-

lescent depression. Biol Psychiatry 1996;40:474±84.

Raskin TH, Crook MA. Antidepressants in Black and White inpati-

ents. Arch Gen Psychiatry 1975;32:643±9.

Rechtscha�en A, Kales A. A manual of standardized terminology,

techniques and scoring system for sleep stages of human subjects.

Los Angeles: Brain Information Service/Brain Research Institute,

UCLA, 1968.

Reynolds III CF, Hoch CC, Buysse DJ, George CJ, Houck PR,

Mazumdar S, Miller M, Pollock BG, Rifai H, Frank E. Sleep in

late-life recurrent depression: changes during early continuation

therapy with nortiptyline. Neuropsychopharmacology 1991;5:85±

96.

Rosenblat R, Tang SW. Do Oriental psychiatric patients receive

di�erent dosages of psychotropic medication when compared

with Occidentals? Can J Psychiatry 1987;32:270±4.

Rush AJ, Armitage R, Gillin JC, Yonkers KA, Winokur A,

Moldofsky H, Vogel GW, Kaplita SB, Fleming JB, Montplaisir

J, Erman MK, Albala BJ, McQuade RD. Comparative e�ects of

nefazodone and ¯uoxetine on sleep in outpatients with major de-

pressive disorder. Biol Psychiatry 1998;44:3±14.

Sharpley AL, Cowen PJ. E�ect of pharmacologic treatments on the

sleep of depressed patients. Biol Psychiatry 1995;37:85±98.

Sharpley AL, Williamson DJ, Attenburrow ME, Pearson G, Sargent

P, Cowen PJ. The e�ects of paroxetine and nefazodone on sleep:

a placebo controlled trial. Psychopharmacology 1996;126:50±4.

Siegel JM. Brainstem mechanisms generating REM sleep. In: Kryger

MH, Roth T, Dement WC, editors. Principles and practice of

sleep medicine. Philadelphia: W.B. Saunders Company, 1994. p.

125±44.

Sommervell PD, Lear PJ, Weissman MM, Blazer DG, Bruce ML.

The prevalence of major depression in Black and White adults in

®ve United States communities. Am J Epidemiol 1989;130:725±

35.

Spitzer RL, Williams JBW, Gibbons M. The structured clinical inter-

view for DMS-III-R (SCID. New York: New York State

Psychiatric Institute, 1986.

Takeuchi DT, Chung RC-Y, Lin K-M, Shen H, Kurasaki K, Chun

C-A, Sue S. Lifetime and 12-month prevalence rates of major de-

pressive episodes and dysthymia among Chinese Americans in

Los Angeles. Am J Psychiatry 1998;155:1407±14.

Thase ME. Depression, sleep and antidepressants. J Clin Psychiatry

1998;59:55±65.

Vega WA, Kolody B, Aguilar-Gaxiola S, Alderete E, Catalano R,

Caraveo-Anduaga J. Lifetime prevalence of DSM-III-R psychia-

tric disorders among urban and rural Mexican-Americans in

California. Arch Gen Psychiatry 1998;55:171±778.

Vogel GW, Bu�enstein A, Minter K, Hennessey A. Drug e�ects on

REM sleep and on endogenous depression. Neurosci Behav Rev

1990;14:49±63.

Vogel G, Cohen J, Mullis D, Kensler T, Kaplita S. Nefazodone and

REM sleep: how do antidepressant drugs decrease REM sleep?

Sleep 1998;21:70±7.

Ware JC, Pittard JT. Increased deep sleep after trazodone use: a

double-blind placebo-controlled study in healthy young adults. J

Clin Psychiatry 1990;51:18±22.

U. Rao et al. / Journal of Psychiatric Research 33 (1999) 419±426426