Effects of sleep deprivation on sleep homeostasis and restoration during methadone-maintenance: A [31]P MRS brain imaging study

Em

GMa

b

c

d

e

f

a

ARRAA

KMSMBSS

1

aeapL1wd

MT

0d

Drug and Alcohol Dependence 106 (2010) 79–91

Contents lists available at ScienceDirect

Drug and Alcohol Dependence

journa l homepage: www.e lsev ier .com/ locate /drugalcdep

ffects of sleep deprivation on sleep homeostasis and restoration duringethadone-maintenance: A [31]P MRS brain imaging study

eorge H. Trksaka,d,∗, J. Eric Jensenc,d, David T. Planted, David M. Penetara,d, Wendy L. Tartarinib,elissa A. Maywalte, Michael Brendelb, Cynthia M. Dorseyd,e, Perry F. Renshawf, Scott E. Lukasa,b,c,d

Behavioral Pharmacology Research Laboratory, McLean Hospital, Belmont, MA, USASleep Research Program, McLean Hospital, Belmont, MA, USABrain Imaging Center, McLean Hospital, Belmont, MA, USAHarvard Medical School, McLean Hospital, Belmont, MA, USASleep Health Centers, Newton, MA, USAThe Brain Institute, Department of Psychiatry, University of Utah School of Medicine, Salt Lake City, UT, USA

r t i c l e i n f o

rticle history:eceived 22 October 2008eceived in revised form 14 July 2009ccepted 17 July 2009vailable online 22 September 2009

eywords:ethadone

leep deprivationagnetic resonance spectroscopy

eta-NTPleep homeostasisleep restoration

a b s t r a c t

Insomnia afflicts many individuals, but particularly those in chronic methadone treatment. Studiesexamining sleep deprivation (SD) have begun to identify sleep restoration processes involving brainbioenergetics. The technique [31]P magnetic resonance spectroscopy (MRS) can measure brain changesin the high-energy phosphates: alpha-, beta-, and gamma-nucleoside triphosphate (NTP). In the presentstudy, 21 methadone-maintained (MM) and 16 control participants underwent baseline (BL), SD (40wakeful hours), recovery1 (RE1), and recovery2 (RE2) study nights. Polysomnographic sleep was recordedeach night and [31]P MRS brain scanning conducted each morning using a 4T MR scanner (dual-tuned pro-ton/phosphorus head-coil). Interestingly, increases in total sleep time (TST) and sleep efficiency index(SEI) commonly associated with RE sleep were not apparent in MM participants. Analysis of methadonetreatment duration revealed that the lack of RE sleep increases in TST and SEI was primarily exhibited byshort-term MM participants (methadone <12 months), while RE sleep in long-term MM (methadone >12
months) participants was more comparable to control participants. Slow wave sleep increased duringRE1, but there was no difference between MM and control participants. Spectral power analysis revealedthat compared to control participants; MM participants had greater delta, theta, and alpha spectral powerduring BL and RE sleep. [31]P MRS revealed that elevations in brain beta-NTP (a direct measure of ATP) fol-lowing RE sleep were greater in MM compared to control participants. Results suggest that differences insleep and brain chemistry during RE in MM participants may be reflective of a disruption in homeostatic sleep function.
. Introduction

Due to public health concern surrounding widespread opioidbuse, a strong focus on treatment and rehabilitation has beenstablished. The development of pharmacological therapies toddress withdrawal and relapse in those seeking treatment hasroven to be of great importance (Connell, 1967; Senay, 1985;
eukefeld and Tims, 1990; Ling and Wesson, 1990; Weddington,990; Zweben and Payte, 1990). Methadone-maintenance isidely used and a standard pharmacotherapy for treating opioid-ependent persons as it results in cross-tolerance that attenuates
∗ Corresponding author at: Behavioral Pharmacology Research Laboratory,cLean Hospital, Harvard Medical School, 115 Mill Street, Belmont, MA 02472, USA.

el.: +1 617 855 2718; fax: +1 617 855 3784.E-mail address: [email protected] (G.H. Trksak).

376-8716/$ – see front matter © 2009 Elsevier Ireland Ltd. All rights reserved.oi:10.1016/j.drugalcdep.2009.07.022

© 2009 Elsevier Ireland Ltd. All rights reserved.

the effects of illicitly administered opioids (Zaks et al., 1971;Volavka et al., 1978; Doverty et al., 2001; Eugenio, 2004; Athanasoset al., 2006). Data suggests that compared to untreated opioid-dependent persons, methadone-maintained (MM) patients havehigher nutritional status, fewer emergency room visits, and a reduc-tion in violent crimes (Szpanowska-Wohn et al., 2004; Gossopet al., 2005; Friedmann et al., 2006). Elucidating the effects ofmethadone early in treatment and through the duration of therapymay be useful in extracting factors leading to failed treatment andrelapse. One notable example is that insomnia particularly afflictsmethadone patients who regularly self-report difficulties initiat-ing and/or maintaining sleep and poor sleep quality (Oyefeso et al.,
1997; Stein et al., 2004; Peles et al., 2006).
Clinical studies have noted that the chronic use of opiates includ-ing methadone results in marked disruptions in sleep (Gossop andBradley, 1984; Oyefeso et al., 1997; Stein et al., 2004). It has beenshown that a single administration of an opioid drug can alter nor-

http://www.sciencedirect.com/science/journal/03768716

http://www.elsevier.com/locate/drugalcdep

mailto:[email protected]

dx.doi.org/10.1016/j.drugalcdep.2009.07.022

80 G.H. Trksak et al. / Drug and Alcohol Dependence 106 (2010) 79–91

Table 1Subject demographics.

Age (years) N Dose (mg) Methadone duration (months) Opiate use (years)

Control 34.5 ± 2.91 16 total – – –50% male

Methadone-maintained 40.1 ± 2.19 22 total 84.2 ± 6.73 17.4 ± 2.90 14.1 ± 1.1657% male

Short-term methadone 40.8 ± 3.07 11 total 94.7 ± 6.93 7.7 ± 1.58 15.6 ± 3.0755% male

Long-term methadone 39.3 ± 3.30 11 total 73.2 ± 12.42 28.2 ± 3.91 13.3 ± 1.4955% male

S hadonp mainto

mRaiaaH1irai

rtacareisoameea(nest(bSpteiiCmtotteodp

ubject demographics—age, sex, and drug use demographics for control and metarticipants by methadone treatment duration as either short-term methadone-piate use values are represented as mean ± SEM.

al sleep architecture in healthy adults (Dimsdale et al., 2007).esearch examining the effects of opioid medications on sleeprchitecture has demonstrated some common nocturnal abnormal-ties with a range of opiate drugs. For example, morphine, heroin,nd methadone have been shown to result in a reduction in SWSnd a suppression of REM sleep (Roubicek et al., 1969; Kay, 1975;owe et al., 1980b,a, 1981; Pickworth et al., 1981; Staedt et al.,996; Dimsdale et al., 2007). Although the identification of dimin-

shed SWS and REM sleep during methadone treatment has beeneported by several studies, data indicate that sleep architecturebnormalities are variable during periods of methadone treatmentnitiation, stabilization, and long-term maintenance (Kay, 1975).

Sleep studies have established that sleep deprivation (SD)esults in marked behavioral challenge evidenced by increased day-ime sleepiness and task-specific functional impairment (Pilchernd Huffcutt, 1996; Van Dongen et al., 2003). Sleep electroen-ephalogram (EEG) studies have established that the sleep deficitssociated with SD results in alterations to the following nightsecovery (RE) sleep indicative of sleep homeostatic function (Tilleyt al., 1987; Akerstedt et al., 2009). The prolonged wakeful periodnvolved in total SD reliably results in a physiologic challenge toleep homeostasis and SD has been an effective tool to study home-static sleep mechanisms in animal and human studies (Johnson etl., 2004; Everson et al., 2005; Zeitzer et al., 2006). The involve-ent of sleep restoration processes in sleep homeostasis becomes

vident when an individual experiences RE sleep following SD. Asxpected, RE sleep from an extended wakeful period is commonlyssociated with reduced WASO, increased TST, and enhanced SEIDe Gennaro et al., 2002). The activity of sleep homeostatic mecha-isms during RE sleep is further exemplified by increased non-rapidye movement (NREM) slow wave sleep (SWS). In addition to thetandard visual criteria used for the analysis of SWS, the examina-ion of slow wave activity (SWA) using of spectral power analysisSPA) allows for a quantitative measure of frequency in the deltaand that does not require a specific amplitude criterion be met.WA in sleep has been shown to be a marker of sleep homeostaticrocesses, as enhancements in SWA have been strongly correlatedo the duration of extended wakefulness (Besset et al., 1998; Finellit al., 2000). While SWS and SWA correlate in healthy individuals,t has been demonstrated that SWS and SWA may be uncoupledn patients with neuropsychiatric illness (Armitage et al., 1995).ollectively, the study of RE sleep following SD provides objectiveeasures of compensatory sleep homeostatic changes in response

o an acute sleep deficit. While there have been no previous studiesf sleep homeostasis in MM participants, a novel study examinedhe effects of partial SD in alcohol-dependent participants in order
o probe potential differences in sleep homeostatic function (Irwint al., 2002). In part, the findings identified abnormal sleep physiol-gy in alcoholic participants, reflected in diminished SWS and SWAuring RE sleep when compared to age-matched healthy controlarticipants.
e-maintained subjects. Secondary analysis dichotomized methadone-maintainedained or long-term methadone-maintained. Age, dose, methadone duration, and

Recent brain imaging studies have begun to identify thebrain processes and mechanisms involved in sleep homeostaticprocesses and the subsequent involvement of sleep restorationfunction during RE sleep (Zeitzer et al., 2006; Dang-Vu et al., 2008).Examinations of the effects of SD on the brain have demonstratedchanges in the function of brain bioenergetics due to increasedwakefulness such as a reduction of brain glycogen stores (Konget al., 2002) as well as a potential interaction with glucocorti-coids (Gip et al., 2004). A report from our laboratory demonstratedin healthy participants that RE sleep following SD results in ele-vations in brain high-energy phosphates such as beta-nucleosidetriphosphate (beta-NTP) (Dorsey et al., 2003). These brain energychanges after RE sleep support theories that wake time activity issustained by brain bioenergetic restoration during sleep (Ticho andRadulovacki, 1991).

Based upon consistent reports of sleep disturbances in MMpatients and to ascertain a greater understanding of methadone-induced physiological changes, we were particularly interestedin how SD may differentially affect MM participants comparedto healthy controls. The premise of the current study was thatMM participants would exhibit a greater susceptibility to distur-bances in sleep architecture and continuity as a result of SD, whichwould be accompanied by more pronounced enhancements inbrain high-energy phosphates compared to control participants.We hypothesized that the enhanced sleep homeostatic responsein MM participants would be evidenced by elevated objectivesleep measures of sleep efficiency and SWS in MM participantswhen compared to that of control participants. It was predictedthat if greater increases in RE sleep measures of SWS were notevident in MM participants, that the additional examination ofSWA may reflect greater delta spectral power during RE sleepin MM participants. Based upon previous brain MRS imagingfindings demonstrating that SD results in RE sleep elevations ofbrain ATP levels, it was hypothesized that subsequent RE sleepelevations in brain ATP would be more pronounced in MM par-ticipants when compared to control participants. A critical aspectof methadone-maintenance is the relationship between treatmentretention/duration and positive treatment outcomes (Condelli andDunteman, 1993; Gottheil et al., 1993; Cacciola et al., 1998; Peleset al., 2008). Clinical data and practices stress the importance ofmethadone stabilization early in treatment related to a number offactors, such as physiological adjustment to methadone, stabilizingmethadone dosing, and providing appropriate medical and socialservices (Kay, 1975; Brown et al., 1982; Hoffman and Moolchan,1994; Eap et al., 2002; King and Brooner, 2008). Based upon apotentially greater physiological instability in patients early in
methadone treatment, it was also hypothesized that the effects ofmethadone-maintenance may be a function of treatment duration,in that MM participants who are early in treatment may exhibit agreater impact of SD on sleep homeostasis when compared to MMparticipants who have been in methadone treatment for more than

ohol D

asosoin

2

2

v2Pafs

bocicwdcpd(ot

owsodatofpus

Bppr

2

l4tet

2

aliatwofpaso

G.H. Trksak et al. / Drug and Alc

year. The present study took a novel approach in investigating theleep homeostatic response to SD in MM participants by examiningbjective EEG measures of sleep and utilizing magnetic resonancepectroscopy (MRS) brain imaging to obtain in vivo determinationsf high-energy brain phosphates, such as beta-NTP in the morn-ngs following baseline sleep, a night of SD, and two subsequent REights.

. Materials and methods

.1. Participants

The study participants were between the ages of 21 and 53 and were recruitedia newspaper, radio, and web-based advertisements. 16 control participants and1 MM participants completed the study. See Table 1 for participant demographics.SG data from 3 control and 5 MM participants was not submitted to power spectralnalysis (PSA) due to inadequate data or insufficient calibration. [31]P MRS scan datarom 3 MM participants was excluded due to inadequate data resulting from MRcanner malfunction.

Participants were excluded during initial telephone screening if they had unsta-le primary medical or psychiatric illness, any current Axis I psychiatric diagnosis,r history of psychotic disorders. Control participants were excluded if they met anyriteria for abuse or any current use of any drugs. MM participants were includedn this study if they met the DSM-IV criteria for opioid dependence, and wereurrently enrolled in a stable methadone-maintenance program. MM participantsere excluded based upon any of the following criteria: (1) diagnosis of substanceependence (other than opioids or nicotine); (2) current primary medical or psy-hiatric illness; (3) and any contraindications to the MR scanning procedures. Allarticipants were additionally screened for psychiatric, medical, and primary sleepisorders during an onsite-screening visit. Structured Clinical Interviews for DSM-IVSCID) were conducted to further rule out primary psychiatric disorders. A physicianbtained a detailed health and medical history and performed a physical examina-ion and routine laboratory tests prior to selection.

All participants who provided informed consent participated in a screening nightf PSG measures to objectively rule out primary sleep disorders. Three participantsere disqualified from the study due to clinically significant obstructive sleep apnea

yndrome (>10 respiratory events per hour of sleep) or periodic limb movement dis-rder (>10 periodic leg movements per hour of sleep). Urine screens were performeduring the initial screening visit and on each of the laboratory nights to ensure thebsence of drug use in control participants and to verify the absence of drugs otherhan methadone and its respective metabolites in MM participants. Due to the highccurrence of polydrug use in MM participants, MM participants were not excludedor positive drug screens for tobacco, low-level marijuana, or past cocaine use. Sixarticipants were excluded from the study due to the presence of other drugs inrine screens prior to sleep study nights and three participants for positive drugcreens other than methadone during the study procedure.

The research protocol was reviewed and approved by the Institutional Reviewoard of McLean Hospital and written informed consent was obtained from allarticipants with a complete description of the study procedures. Participants com-leting segments of the experimental protocol received monetary and/or voucheredeemable at local retail shops as compensation for their voluntary participation.

.2. Sleep deprivation paradigm

Following the initial screening night all participants underwent a night of base-ine sleep, which was followed by a night of total sleep deprivation (SD) involving0 total consecutive wakeful hours. Following SD, participants completed two addi-ional nights of recovery sleep (RE). PSG recordings were collected each nightxcluding the SD night and between 7:00 and 8:00 each study morning the par-icipants underwent brain phosphorous [31]P MRS scanning.

.3. Sleep physiologic monitoring

Electroencephalogram (EEG), electrooculogram (EOG), electromyogram (EMG),nd electrocardiograph (ECG) measurements were collected during screening, base-ine, and RE nights. For participant screening full respiratory monitoring was appliedncluding respiratory flow, effort, and oximetry. Electrode placement was done inccordance with standard PSG procedure (Rechtschaffen and Kales, 1968). Objec-ive sleep measures obtained from physiological recording were defined as follows:akefulness after sleep onset (WASO)—the amount of awake time in bed that

ccurred after sleep onset, sleep onset latency (SOL)—the total time accumulatedrom wakefulness to sleep, sleep efficiency index (SEI)—the total time asleep as aroportion of the total time in bed, total sleep time (TST)—total minutes of sleepccumulated after sleep onset, slow wave sleep (SWS)—the total minutes of SWScored by standard scoring criteria, rapid eye movement (REM) sleep—total minutesf REM sleep scored by standard criteria.

ependence 106 (2010) 79–91 81

2.4. Spectral power analysis (SPA)

Electroencephalogram spectral power analysis (SPA) was calculated fromobtained PSG data using SomnologicaTM Science (Embla®) software. Thirty-secondepochs of non-rapid eye movement (NREM) sleep from stages 2, 3, and 4 obtainedwith a data sampling rate of 220 Hz and a 70 Hz high frequency filter from the elec-trode pairs C3–A2, which were submitted to SPA and absolute values were averagedacross the night. Specific frequency bands were defined as delta (0.5–4 Hz), theta(4–8 Hz), alpha (8–12 Hz), sigma (12–16 Hz), beta 1 (16–24 Hz), beta 2 (24–32 Hz)and gamma (32–48 Hz). PSA data units were obtained for these frequency bands inmicrovolts/hertz.

2.5. Brain magnetic resonance spectroscopy imaging

All MRI/[31]P MRS imaging experiments were performed on a 4 Tesla whole-body (Varian/Unity-INOVA; Varian, Palo Alto, CA, USA), magnetic resonance scanneroperating at 170.31 MHz and 68.95 MHz for proton and phosphorus measurements,respectively. A dual-frequency, transverse electromagnetic (TEM)-design volumehead-coil tuned to both proton and phosphorus frequencies was used for all imagingand spectroscopy experiments (Bioengineering Inc., Minneapolis, MN, USA).

2.6. Proton MRI

High-contrast, T1-weighted sagittal images of the entire brain were firstacquired using a three-dimensional, magnetization-prepared FLASH imagingsequence (3D-FLASH), allowing for clear differentiation between grey-matter andwhite-matter, as well as clearly delineating between the different anatomicalregions of interest. The acquisition parameters for the sagittal images were:TE/TR = 6.2/11.4 ms, field-of-view (FOV) = 24 cm × 24 cm, readout-duration = 4 ms,receive bandwidth = ±32 kHz, in-plane matrix size = 256 × 256, in-plane reso-lution = 0.94 mm × 0.94 mm, readout points = 512, axial-plane matrix size = 16,axial-plane resolution = 2.5 mm sagittal, scan time = 1 min 15 s. Then, high-resolution T1-weighted images were acquired in the transverse plane, lasting 2 m30 s each with the same 3D-FLASH imaging sequence, but instead 32 slices werecollected (phase-encodes) of 4 mm nominal thickness (TE/TR = 6.2/11.4 ms, field-of-view (FOV) = 24 cm × 24 cm, readout-duration = 4 ms, receive bandwidth = ±32 kHz,in-plane matrix size = 256 × 256, in-plane resolution = 0.94 mm × 0.94 mm, readoutpoints = 512).

2.7. Phosphorus MRS

Phosphorus MRS was performed using the phosphorus channel of the dual-tuned proton-phosphorus head-coil. Initially, eight control participants werescanned using a three-dimensional (3D) [31]P MRS sequence for the phosphorusacquisition. It was recognized that MM participants were unable to complete therequired 1-h scan duration of the 3D [31]P MRS scan. In order to maintain valuablescan data and to maintain comparability of additional scan data, a two-dimensional(2D) [31]P MRS version of the original 3D [31]P MRS sequence was used for the remain-ing twenty-nine participants. Limiting the acquisition of voxels from a 2D slab,allowed for a markedly shorter 2D [31]P MRS scan duration of 9 min compared tothe 46-min 3D [31]P MRS sequence.

The 2D-MRSI version phase encoded over a 6 cm thick excitation slab placedin the exact same mid-sagittal position as the 3 slices involved in the 3D [31]P MRSsequence. Aside from this, all other parameters were identical between the 2D- and3D-[31]P MRS protocols, including FOV (33 cm × 33 cm), TR (500 ms), matrix (16 × 16,sparse sampling scheme using the same SINC-lobe-modulated, weighted-average k-space filter). Transmit/receive frequency was first centered on the PCr resonance,as measured with a global free-induction decay (FID). The 2D-CSI sequence useda reduced phase-encoding scheme based on prior work (Ponder and Twieg, 1994).This scheme allows for the inclusion of circularly bound, reduced-point, weightedk-space acquisition, providing approximately 35% more signal-to-noise for a givenscan time and effective voxel volume over conventional methods. All viable vox-els from three mid-sagittal slices over the entire brain were analyzed from 3D[31]P MRS data. This was essentially equivalent to the 2D axial-plane consistingof 2 cm × 2 cm × 6 cm voxels (slices are effectively 2 cm thick × 3 slices equals6 cm thick slab-of-interest). The 2D-PSF (actual voxel signal distribution) as wellas signal-to-noise, were analogous between the 2D and 3D [31]P MRS acquisitions.Also, since the tip-angle (32◦) and TR were the same between sequences, the resul-tant spectra were virtually identical between sequences when tested back-to-backon a healthy control in terms of metabolite T1-weighting. In addition to match-ing scan parameters, fundamental differences exist between the 2D- and 3D-[31]PMRS sequences that primarily emanate from diminutive differences in the tip-anglebetween sequences (global square pulse for 3D vs. selective SINC pulse for 2 d), lead-ing to minute T1-weighted differences in derived peak areas. There is an inherentchemical-shift displacement artifact using 2D [31]P MRS, which could affect mea-
sures of any off-resonance metabolites due to spatial-shifts in slab excitation. Torectify these potential influences, correction-factors for each measured metabolitefrom an in vivo healthy control participant were derived and applied to the 3D[31]P MRS data. The resultant metabolite measures were extremely in-line with theacquired 2D [31]P MRS data and there were no significant differences for any of theMRS metabolites.


F e corro (C) Rem for dis

nssicssblgcbl

F(ds

ig. 1. (A) Numbered 1–6 phosphorus [31]P MRS spectrum from example voxels on thf [31]P MRS acquisition with voxel locations 1–6 corresponding to spectral output.odeled fit and residual. All spectra are displayed with 15 Hz exponential filtering

All in vivo CSI/image data was processed and viewed using Varian Nuclear Mag-etic Resonance (VNMR) software, Version 6.1b (Varian, Palo Alto, CA, USA) andoftware designed and written on site. Prior to Fast Fourier Transform (FFT) recon-truction to spatially resolve the CSI spectra, the collected k-space data was centeredn a 16 × 16 square matrix. Each time-domain FID was then zero-filled out to 2048omplex points and left-shifted five points to remove residual bone/rigid membraneignal. Using the MRI images, the 2D-CSI data grid was shifted in the x and y dimen-ions in order to position the sampling grid such that it was centered inside the
rain according to anatomical landmarks. The peak areas of the following metabo-
ites: phosphoethanolamine (Pe), phosphocholine (PCh), inorganic phosphate (Pi),lycerophosphoethanolamine (GPE) and glycerophosphocholine (GPC), phospho-reatine (PCr), and three peaks for adenosine triphosphate (alpha-, gamma-, andeta-NTP) (Fig. 1), were fitted to an in vivo spectral model through the use of a non-

inear, iterative (Potwarka et al., 1999; Jensen et al., 2002). The fitting routine is based

ig. 2. Polysomnogram sleep measures recorded for baseline, recovery1, and recovery2 slSWS)-participants exhibited a reliable increase in SWS during recovery1 sleep compareecrease SOL during recovery1 sleep compared to baseline sleep; (C) Rapid Eye Movemeleep compared to baseline sleep. All values are represented as mean ± SEM with an alph

esponding mid-axial slice. (B) Representative mid-axial slice indicating the locationpresentative [31]P MRS spectrum indicating labeled phosphorus containing peaks,play.

on a Marquardt–Levenberg algorithm, utilizing prior spectral knowledge for the rel-ative amplitudes, linewidths, lineshapes, peak positions and J-coupling constants tomodel the in vivo [31]P brain spectrum. Each fitted spectral peak area was expressedas a ratio to the total [31]P signal per voxel. The fitted metabolite amplitudes are notT2-weighted since the fitting algorithm back-extrapolates to time zero.

2.8. Statistical analysis

In order to identify differences under baseline conditions, PSG, SPA, and [31]PMRS data measures were subjected to ANOVA analysis. All obtained data were ini-tially analyzed for effects of age and sex in order to identify potential treatmentinteractions. Significant interactions with age or sex with methadone treatmentwere not detected for any obtained dependent measures. Statistical linear mixedmodel analyses were conducted to determine effects of treatment for PSG and SPA

eep nights in control and methadone-maintained participants. (A) Slow Wave Sleepd to baseline sleep; (B) Sleep Onset Latency (SOL)-participants exhibited a markednt (REM) Sleep-participants exhibited a reliable increase in REM during recovery1a of p < 0.05. *Denotes significantly different than baseline sleep data.

G.H. Trksak et al. / Drug and Alcohol Dependence 106 (2010) 79–91 83

F covert overym h an a

dtaa

iBctaTw(todmmc

3

3

3bgSiit(indtmo

3oittippd[

ig. 3. (A) Total sleep time (TST)-measures recorded for baseline, recovery1, and reime (TST)-secondary analysis of methadone treatment duration for baseline, rec

ethadone-maintained participants. All values are represented as mean ± SEM wit

ata obtained on baseline, RE1, and RE2 study nights and [31]P MRS data collectedhe mornings following baseline, SD, RE1, and RE2 study nights. When appropri-te, post hoc analyses of pair-wise comparisons were performed using Fischer’s LSDnalysis. Alpha was set to p < 0.05 for all statistical testing.

Although the identification of treatment effects between control and MM partic-pant data was of primary interest, additional treatment variables were considered.ased upon the possibility that significant treatment differences between MM andontrol participants may be further explained by either the duration of methadonereatment, the duration of opioid use, and/or methadone dose were submitted asdditional variables for further analysis for all PSG, SPA, and [31]P MRS measures.o determine effects of the duration of methadone treatment, participant dataas dichotomized as short-term MM (methadone <12 months) and long-term MM

methadone ≥12 months) or the data was examined as a continuous covariate byhe number of months enrolled in methadone treatment. Similarly, potential effectsf methadone dose were determined with participant data dichotomized as low-ose MM (daily dose <80 mg methadone) and high-dose MM (daily dose ≥80 mgethadone) or the data was examined as a continuous covariate by mg dosage ofethadone. The effects of the duration of opioid use were examined as a continuous

ovariate.

. Results

.1. Polysomnogram (PSG) sleep measures

.1.1. Slow wave sleep (SWS). The overall amount of SWS duringaseline sleep was not significantly different between treatmentroups as both control and MM participants had comparable totalWS time. Statistical analysis across the sleep deprivation paradigmdentified strong effect of study day [F[1,36] = 15.542, p < 0.0001],ndicative of a reliable elevation in SWS observed across all par-icipants during RE1 sleep compared to SWS during baseline sleepFig. 2A). Despite lower mean values of SWS across sleep nightsn MM participants compared to control participants, there wereo significant treatment differences in SWS with respect to studyay. Additionally, there were no significant treatment effects inhe secondary analysis that differentiated effects by duration of

ethadone treatment, methadone treatment dose, or duration ofpioid use on SWS.

.1.2. Sleep onset latency (SOL). There was a significant main effectf treatment [F[1,36] = 4.761, p < 0.05], as MM participants exhib-ted decreased overall SOL during baseline sleep when comparedo that of control participants. Similarly, statistical analysis acrosshe sleep deprivation paradigm identified a statistically signif-
cant main effect of treatment [F[1,36] = 5.268, p < 0.05], as MMarticipants had lower mean SOL across sleep nights when com-ared to control participants. Statistical analysis across the sleepeprivation paradigm identified a significant effect of study dayF[1,36] = 23.514 p < 0.0001], exemplifying a reliable reduction in SOL
y2 sleep nights in control and methadone-maintained participants. (B) Total sleep1, and recovery2 for control, short-term methadone-maintained, and long-termlpha of p < 0.05. *Denotes significantly different from control participant data.

across all participants during RE1 sleep compared to baseline sleep(Fig. 2B). Additionally, there were no significant treatment effectsin the secondary analysis that differentiated effects by methadonetreatment, methadone treatment dose, or duration of opioid use onSOL.

3.1.3. Rapid eye movement (REM) sleep. There were no significanttreatment differences in baseline REM sleep and although MM par-ticipants had lower mean REM across sleep nights compared tocontrol participants analysis of treatment effects across the sleepdeprivation paradigm were not statistically significant. Statisticalanalysis across the sleep deprivation paradigm identified strongeffect of study day [F[1,36] = 10.64, p < 0.01], exemplifying increasesin REM sleep across participants during RE1 sleep compared tobaseline sleep (Fig. 2C). Additionally, there were no significanttreatment effects in the secondary analyses that differentiatedeffects by methadone treatment, methadone treatment dose, orduration of opioid use on REM sleep. Additionally, there were notreatment differences in REM latency across SD study nights.

3.1.4. Total sleep time (TST). There were no significant treatmenteffects in baseline sleep TST. Statistical analysis across the sleepdeprivation paradigm identified a statistically significant treat-ment × sleep night interaction [F[1,36] = 7.922, p < 0.01] (Fig. 3A).Post hoc analysis revealed that there was a significant difference inTST between treatment groups during both RE1 [p < 0.05] and RE2[p < 0.05] sleep nights, as MM participants exhibited significantlyreduced TST compared to control participants.

Secondary analysis of treatment effects specific to the partici-pants’ duration of methadone treatment revealed a significant maineffect of treatment duration [F[2,31] = 5.937, p < 0.05] and a treat-ment × study day interaction [F[2,35] = 3.910, p < 0.05] (Fig. 3B). Posthoc revealed that the original treatment difference (control vs.MM) was primarily driven by increases in TST in control and long-term MM participants during RE1 night, that were not exhibitedby short-term MM participants [p < 0.5]. There were no signifi-cant treatment effects in the secondary analysis that differentiatedeffects by methadone treatment dose or duration of opioid use onTST data.

3.1.5. Wake after sleep onset (WASO). There were no significant
effects in WASO during baseline sleep. Statistical analysis across thesleep deprivation paradigm identified a significant main effect oftreatment [F[1,34] = 4.386, p < 0.05], as WASO amounts were greaterin MM participants across sleep nights when compared to con-trol participants (Fig. 4A). There was a significant treatment × sleep


Fig. 4. (A) Wake after sleep onset (WASO)-measures recorded for baseline, recovery1, and recovery2 sleep nights in control and methadone-maintained participants. (B) Wakeafter sleep onset (WASO)-secondary analysis of methadone treatment duration for baseline, recovery1, and recovery2 for control, short-term methadone-maintained, andlong-term methadone-maintained participants. All values are represented as mean ± SEM with an alpha of p < 0.05. *Denotes significantly different from control participantdata.

F nd ree line, rt witha .

nst

tdt

F(p

ig. 5. (A) Sleep efficiency Index (SEI)-measures recorded for baseline, recovery1, afficiency index (SEI)-secondary analysis of methadone treatment duration for baseerm methadone-maintained participants. All values are represented as mean ± SEMnd denotes significantly different from control and long-term MM participant data

ight interaction [F[1,36] = 5.239, p < 0.05], which emanated fromignificantly lower WASO in control participants when compared
o MM participants on RE1 [p < 0.05] and RE2 [p < 0.05].
Secondary analysis of treatment effects specific to the par-icipants’ duration of methadone treatment revealed markedifferences in WASO between short-term and long-term MM par-icipants (Fig. 4B). When treatment groups were specified as control

ig. 6. Spectral power analysis (SPA) of polysomnogram sleep measures for baseline, recovA) Delta (0.5–4 Hz); (B) theta (4–8 Hz); (C) alpha (8–12 Hz). All values are represented asarticipant SPA data.

covery2 sleep nights in control and methadone-maintained participants. (B) Sleepecovery1, and recovery2 for control, short-term methadone-maintained, and long-

an alpha of p < 0.05. *Denotes significantly different from control participant data

vs. short-term MM vs. long-term MM, analysis revealed a sig-nificant main effect of treatment [F[2,32] = 5.122, p < 0.05] as well
as a significant treatment × study day interaction [F[2,36] = 6.983,p < 0.01]. Post hoc analysis determined that these effects weredriven by significantly lower WASO in both control and long-termMM participants in comparison to WASO from short-term MM par-ticipants on both RE1 [p < 0.05] and RE2 [p < 0.05] nights. There were
ery1, and recovery2 sleep nights in control and methadone-maintained participants.mean ± SEM with an alpha of p < 0.05. *Denotes significantly different than control


F ollowi( d mes signib

nfo

3omabsc

immaRp(od

3

3spwaepacimo

3spbSfiibwc

ig. 7. Global brain phosphorus [31]P MRS measurements recorded the mornings fA) Brain beta-NTP across study nights in all participants. (B) Beta-NTP-in control anubjects. Values are represented as mean ± SEM with an alpha of p < 0.05. *Denotesaseline and denotes significantly different than controls subjects.

o significant treatment effects in the secondary analysis that dif-erentiated effects by methadone treatment dose or duration ofpioid use on WASO data.

.1.6. Sleep efficiency index (SEI). There were no significant effectsf treatment in baseline sleep SEI. There was a significant treat-ent × sleep night interaction [F[1,32] = 7.390, p < 0.01]. Post hoc

nalysis revealed that there was a significant difference in SEIetween treatment groups during RE1 [p < 0.05] and RE2 [p < 0.05]leep nights, as MM participants exhibited significantly reduced SEIompared to control participants (Fig. 5A).

Secondary analysis of treatment effects specific to the partic-pants’ duration of methadone treatment identified a significant

ain effect of treatment [F[2,32] = 4.166, p < 0.05] and a treat-ent × study day interaction [F[2,36] = 5.941, p < 0.05]. Post hoc

nalyses identified statistically significant increases in SEI duringE1 and RE2 nights in control [p < 0.05] and long-term MM [p < 0.05]articipants were not exhibited by short-term MM participantsFig. 5B). There were no significant treatment effects in the sec-ndary analysis that differentiated effects by methadone treatmentose or duration of opioid use on SEI data.

.2. Spectral power analysis (SPA)

.2.1. Delta. Analysis of Power during baseline sleep revealed aignificant main effect of treatment [F[1,26] = 5.61, p < 0.05], as MMarticipants exhibited elevated delta power during baseline sleephen compared to control participants (Fig. 6A). Statistical analysis

cross the sleep deprivation paradigm identified a significant mainffect of treatment [F[1,28] = 7.28, p < 0.01], illustrating that deltaower levels were elevated in MM participants at baseline [p < 0.05]nd continuing during recovery sleep [p < 0.05], when compared toontrol participants. There were no significant treatment effectsn the secondary analysis that differentiated effects by duration of

ethadone treatment, methadone treatment dose, or duration ofpioid use on delta power.

.2.2. Theta. Analysis of theta during baseline sleep revealed atatistically significant main effect of treatment [F[1,26] = 11.66,< 0.01], as MM participants exhibited elevated theta duringaseline sleep when compared to control participants (Fig. 6B).tatistical analysis across the sleep deprivation paradigm identi-
ed a significant main effect of treatment [F[1,28] = 12.24, p < 0.01],
llustrating that theta levels were elevated in MM participants ataseline [p < 0.01] and continuing during recovery sleep [p < 0.01],hen compared to control participants. There were no signifi-

ant treatment effects in the secondary analysis that differentiated

ng baseline, sleep deprivation (Sleep Dep), recovery1, and recovery2 sleep nights.thadone-maintained subjects. (C) Total NTP-in control and methadone-maintainedficant difference between beta-NTP levels following recovery1 when compared to

effects by duration of methadone treatment, methadone treatmentdose, or duration of opioid use on theta measures.

3.2.3. Alpha. Analysis of alpha power during baseline sleeprevealed a statistically significant main effect of treatment[F[1,26] = 9.38, p < 0.01], as MM participants exhibited elevated alphaduring baseline sleep when compared to control participants(Fig. 6C). Statistical analysis across the sleep deprivation paradigmidentified a significant main effect of treatment [F[1,28] = 7.61,p < 0.05], illustrating that alpha levels were elevated in MM partic-ipants at baseline [p < 0.05] and continuing during recovery sleep[p < 0.05], compared to control participants. There were no signifi-cant treatment effects in the secondary analysis that differentiatedeffects by duration of methadone treatment, methadone treatmentdose, or duration of opioid use on alpha power.

3.3. Brain phosphorus [31]P magnetic resonance spectroscopy

3.3.1. Beta-nucleoside triphosphate (beta-NTP). There were notreatment differences in baseline levels of brain beta-NTP betweenMM and control participants. Analysis of changes in beta-NTP overthe course of the sleep deprivation paradigm revealed notable dif-ferences with respect to treatment. There was a significant maineffect of treatment [F[1,32] = 26.19, p < 0.001], as MM participantshad greater levels of beta-NTP compared to levels in control partic-ipants when collapsed over the study days. There was a significanteffect of study day [F[1,68] = 11.872, p < 0.01] (Fig. 7A). Similar toprevious findings (Dorsey et al., 2003), total sleep deprivationenhanced brain beta-NTP levels which was most evident in [31]PMRS following a night of recovery sleep. Post hoc analysis verifiedsignificantly greater levels of beta-NTP following recovery sleepwhen compared to baseline beta-NTP levels [p < 0.05]. These resultsprovide confirmation of a build up of brain beta-NTP followingrecovery sleep from sleep deprivation, indicating marked changesin brain ATP as a result of sleep deprivation.

Most notably, there was a significant interaction of treat-ment × study day [F[1,68] = 27.925, p < 0.001], identifying significantdifferences in brain beta-NTP levels between control and MM par-ticipants over the course of the sleep deprivation paradigm (Fig. 7B).Post hoc analyses determined that brain beta-NTP levels follow-ing SD [p < 0.05], RE1 [p < 0.05], and RE2 [p < 0.05] nights weresignificantly greater in MM participants than in control partici-
pants. These data demonstrate that elevations in beta-NTP levelswere greatest in MM participants following RE1 sleep, indicating adifferential influence of sleep deprivation on metabolic processesinvolving the regulation and/or restoration of brain ATP in thesedrug-dependent participants.

8 cohol D

dmcpapsefowgd

3ArsPaflv[

rNa(pMsiNt

amwdatfott

3wpclo

dvPnwpt(tmr

6 G.H. Trksak et al. / Drug and Al

There were significant treatment and a treatment × studyay interaction effects in the secondary analysis of duration ofethadone treatment on baseline beta-NTP levels and over the

ourse of the sleep deprivation paradigm. Despite these effects,ost hoc analysis revealed that beta-NTP levels in short-term MMnd long-term MM participants were different than control partici-ants, but there were no differences between beta-NTP levels fromhort-term and long-term MM participants. Similarly, secondaryxamination of treatment effects on beta-NTP levels emanatingrom participant differences in methadone dose or the durationf opioid dependence did not reveal informative significant effectshen each was considered a continuous covariate within the MM

roup or when methadone dose was dichotomized as high or low-ose MM.

.3.2. Alpha-, gamma-, and total nucleoside triphosphate (NTP).lthough, changes in brain beta-NTP were of primary interest ineflecting changes in brain ATP with sleep deprivation and recoveryleep, results are reported for alpha-NTP, gamma-NTP, total NTP,Cr, and Pi. There were no significant treatment differences in brainlpha- or gamma-NTP levels between MM and control participantsollowing a normal baseline night of sleep. Examination of baselineevels of total NTP revealed that brain total NTP levels were ele-ated in MM participants when compared to controls participantsF[1,34] = 10.31, p < 0.01].

Analysis over the course of the sleep deprivation paradigmevealed notable differences with respect to gamma- and totalTP levels. There were no significant differences detected in brainlpha-NTP levels over the course of the sleep deprivation paradigmTable 2). There was a main effect of treatment [F[1,32] = 8.12,< 0.01], as brain gamma-NTP levels were significantly greater inM participants compared to control participants collapsed over

leep nights. Analysis of brain total NTP levels revealed a signif-cant main effect of treatment [F[1,32] = 30.81, p < 0.0001], as totalTP levels were significantly greater in MM participants compared

o control participants collapsed over sleep nights (Fig. 7C).Additionally, there were no significant effects in the secondary

nalysis of duration of methadone treatment, methadone treat-ent dose, or duration of opioid use on alpha-NTP levels. Thereere statistically significant treatment and/or treatment × studyay interaction effects on gamma- and total NTP in the secondarynalyses of duration of methadone treatment, methadone dose, orhe duration of opioid use, but these treatment effects did not dif-erentiate effects between either short-term MM vs. long-term MMr high vs. low-dose MM. There were no significant effects whenreatment variables were considered a continuous covariate withinhe MM group.

.3.3. Phosphocreatine (PCr) and inorganic phosphate (Pi). Thereere no treatment differences in baseline levels of brain phos-hocreatine (PCr) or inorganic phosphate (Pi) between MM andontrol participants. There was significantly reduced baseline Pievels [F[1,34] = 4.39, p < 0.05] in MM participants compared to thatf control participants.

Brain PCr levels markedly increased over the course of the sleepeprivation, which was reflected in a significant main effect of studyisit [F[1,68] = 5.18, p < 0.05]. Post hoc analysis revealed that brainCr levels were significantly elevated following recovery1 sleepight when compared to after baseline sleep night [p < 0.05]. Thereas a significant interaction of treatment × study day [F[1,68] = 6.62,< 0.01] as brain PCr levels following recovery1 sleep in MM par-

icipants were greater than that of control participants [p < 0.05]Table 2). In contrast, brain Pi levels markedly decreased overhe course of the sleep deprivation and there was a significant

ain effect of study visit [F[1,68] = 7.98, p < 0.01]. Post hoc analysisevealed that brain Pi levels following recovery sleep in MM par-

ependence 106 (2010) 79–91

ticipants were greater than that of control participants [p < 0.05](Table 2). There was a main effect of treatment across study days[F[1,32] = 17.52, p < 0.001] as MM participants had lower Pi levelswhen compared to control participants.

There were statistically significant treatment and/or treat-ment × study day interaction effects on Pi and PCR in the secondaryanalyses of duration of methadone treatment, methadone dose, orthe duration of opioid use, but these treatment effects did not dif-ferentiate effects between either short-term MM vs. long-term MMor high vs. low-dose MM. There were no significant effects whentreatment variables were considered a continuous covariate withinthe MM group.

4. Discussion

The current study aimed to further understand sleep distur-bances in MM participants by examining the response to 40 hof total SD followed by two RE sleep nights while measuringneurophysiologic PSG and [31]P MRS neuroimaging. The presentinvestigation is the first study of SD in MM participants to inves-tigate the integrity of sleep homeostatic and brain restorationprocesses. We presently hypothesized that MM participants wouldexhibit greater reductions in WASO and enhancements of TST andSEI during RE sleep when compared to control participants. Con-trary to our predictions, although control participants exhibitedthe typical enhancements in RE sleep, MM participants not onlydid not appear to experience the expected increases in TST andSEI and reductions in WASO during RE sleep. Analyses examiningpotential effects of the duration of methadone treatment providedgreater insight into the mediation of the treatment effect, further-more indicating that short-term MM participants did not displaythe expected changes in RE sleep such as decreased WASO andgreater TST and SEI, while long-term MM participants were moresimilar to controls. These data indicate that RE sleep is altered inMM participants, but particularly in short-term MM participantswho have been enrolled in methadone treatment for less than 12months. The RE sleep differences observed in short-term MM par-ticipants may be reflective of a greater disruptive impact of SD toRE-related sleep restoration processes when methadone patientsare not yet stabilized to methadone treatment. Overall, the reducedRE sleep efficiency in short-term MM participants may result in apartial loss of the expected restorative qualities of sleep, requirea longer duration of RE sleep, and contribute to greater and/orprolonged consequences to behavioral performance.

Additionally, based upon the notion that SD would potentiallyhave a greater impact in MM participants and the strong relation-ship of homeostatic control and measures of SWS and SWA, it washypothesized that MM participants would have greater levels ofSWS and/or SWA during RE sleep when compared to that of con-trol participants. Although we did not detect treatment effects inSWS across sleep nights, SWS did increase during RE sleep com-pared to baseline with a subsequent return to baseline becomingevident during RE2 sleep in both MM and control participants.The lack of an effect of MM on SWS is interesting considering theconsistent treatment differences presently obtained and in lieu ofreports implicating a primary role of SWS in sleep homeostaticand restoration processes (Horne and Staff, 1983; Dorsey et al.,1996, 2003). One interpretation may be that although a strong rela-tionship exists between SWS and sleep restoration; other sleepprocesses may also contribute to restorative sleep. Alternatively,
it is also plausible that examination of SWA as determined by deltaSPA may better characterize differences in MM participants, other-wise unidentifiable using standard SWS criteria, due to a potentialuncoupling of SWS and SWA measures in MM participants. In orderto further investigate changes in restorative processes in MM par-


Table 2Global Brain levels of alpha-NTP, gamma-NTP, Pi, and PCr.

Baseline Sleep deprivation Recovery1 Recovery2

Alpha-NTP Control 10.14 ± 3.2 6.88 ± 0.1 6.86 ± 0.1 6.96 ± 0.1Methadone 7.57 ± 0.2 10.9 ± 3.0 7.97 ± 0.1 7.71 ± 0.1

Gamma-NTP Control 8.70 ± 0.1 8.78 ± 0.1 8.98 ± 0.1 8.93 ± 0.1Methadone 9.59 ± 0.4 8.95 ± 0.1 9.12 ± 0.1 9.11 ± 0.1

Pi Control 7.84 ± 0.1 7.67 ± 0.1 7.66 ± 0.1 7.63 ± 0.1Methadone 7.40 ± 0.2 7.01 ± 0.2 6.77 ± 0.2 7.21 ± 0.2

PCr Control 15.75 ± 0.2 15.92 ± 0.3 15.87 ± 0.2 15.71 ± 0.3Methadone 16.29 ± 0.4 14.97 ± 0.2* 14.98 ± 0.2* 15.54 ± 0.4

Results summary of alpha-nucleoside triphosphate (NTP), gamma-NTP, inorganic phosphate (Pi), and phosphocreatine (PCr), obtained from global brain phosphorus [31]PM deprm e resp

twsdsta

trMiSathiaSPwdn(AiilStMdtRsdcitIfies

btpfF

RS measurements. The data was recorded the mornings following baseline, sleepaintained (n = 19) participants. All values are a mean metabolite ratio ×10−2 of th* Denotes significantly different than controls participants.

icipants that were not observed in SWS measures during RE sleep,e conducted power spectral analysis (PSA) to characterize delta

pectral power related SWA for all study sleep nights. PSA findingsemonstrated that MM participants had significantly greater deltapectral power during RE1 and RE2 sleep compared to control par-icipants, supporting the notion that RE sleep restoration processesre elevated during methadone treatment.

In contrast to the findings of Irwin et al., which demonstratedhat alcoholics have diminished plasticity in the sleep homeostaticegulation of SWS and SWA (2002), the current study found thatM participants did not display a loss in sleep homeostatic plastic-

ty. Alternatively, MM participants exhibited RE sleep increases inWS that were comparable to control participants and intriguingly,hyper-responsive sleep homeostatic response in SWA. The direc-

ional disparity between alcoholic and MM participants in sleepomeostatic function may in part be driven by underlying changes

n sleep related aspects of central nervous system physiology. Thecute effects of alcohol have been shown to result in increases inWS and reductions in REM sleep (Salamy and Williams, 1973;rinz et al., 1980; Roehrs et al., 1999). Similar to sleep findingsith MM, alcohol dependence has been shown to result in sleepifficulties and disruptions in sleep architecture, with the mostotable findings identifying a reduction in SWS and REM sleepWagman et al., 1978; Gillin et al., 1990; Nicholas et al., 2002).nother comparison exists from findings of congruent alterations

n sleep measures of SWS in MM participants and those observedn depression patients. A strong relationship between sleep regu-ation and depression has been established and interestingly, acuteD has been shown to have therapeutic effects on depressive symp-omology (van den Burg and van den Hoofdakker, 1975; Reyero and

uller, 1977; Duncan et al., 1980). The therapeutic effect of SD onepressive symptoms has been shown to have a direct relationshipo the occurrence of SWS recovery or plasticity during subsequentE sleep (Reynolds et al., 1987; Nissen et al., 2001). Despite similarleep architecture findings such as reduced SWS between alcoholependence, depression, and methadone treatment, there are somelear differences when comparing the response to SD. The dimin-shed SWS plasticity in response to SD in alcoholics is in contrast tohe elevated SWS in response to SD observed in depressed patients.n comparison, the present findings demonstrate a differential pro-le of SWS/SWA plasticity in response to SD, as MM participantsxhibited RE sleep plasticity in SWS comparable to that of controlubjects and additionally displayed markedly enhanced SWA.

Several studies have demonstrated marked fluctuations in brain
ioenergetics as a result of SD and provide support to theorieshat fluctuations in brain bioenergetics are likely attributable to therolonged metabolic demands encountered with extended wake-ulness (Adam and Oswald, 1977; Benington and Heller, 1995;rank, 2006). For example, the noted depletion of brain glycogen
ivation, recovery1, and recovery2 study nights in control (n = 16) and methadone-ective metabolite/total 31P ± SEM with an alpha of p < 0.05.

energy stores resulting from SD may reflect prolonged neuronalenergy demands depleting glycogen fuel sources (Kong et al.,2002; Gip et al., 2004). Furthermore, evidence of SD-induced ele-vations in brain adenosine may reflect the wakeful metabolismof brain ATP into adenosine diphosphate (ADP) and adenosinebyproducts, which may underlie an enhancement of sleep pres-sure experienced with prolonged wakefulness (Radulovacki, 1985;Porkka-Heiskanen et al., 1997; Zeitzer et al., 2006). Collectively,these data support the notion that under normal conditions, brainglycogen driven by metabolic activity may be utilized as ATP andfollowing ATP utilization adenosine is derived promoting sleepdrive. Both a depletion of brain glycogen stores and elevated brainadenosine levels with SD implicate corresponding fluctuations inbrain ATP. Subsequently, Dorsey et al., provided confirmation of SD-induced changes in brain ATP levels, specifically demonstrating abuild up of brain ATP levels the morning following RE sleep (2003).

Given the previous [31]P MRS findings of SD-induced build upof brain beta-NTP following RE (Dorsey et al., 2003), the presentstudy sought to examine changes in brain bioenergetics follow-ing SD, but aimed to investigate how the build up of brain ATPfollowing RE sleep may be altered in MM participants. [31]P MRSis the most studied method to examine brain high-energy phos-phate metabolites and it provides a direct measure of NTP levels(Renshaw et al., 2001). Beta-NTP is of primary interest as beta-NTPis unique to ATP, whereas alpha- and gamma-NTP also contributeto nucleoside diphosphates (NDP). The present study did not iden-tify methadone-maintenance-related differences in baseline levelsof brain beta-NTP, which is consistent with previous findings of nodifferences in baseline brain NTP levels between early treatmentMM participants and control participants (Silveri et al., 2004). Thepresent study in accordance with the findings of Dorsey et al., par-ticipants of both treatment groups exhibited a significant increasein brain beta-NTP, providing confirmation that brain ATP levelsfluctuate over the course of a SD paradigm, with peak ATP levelsoccurring following a first night of RE sleep (2003). Furthermore,the present global brain [31]P MRS results identified a significantdifference in the RE1 associated elevation in brain beta-NTP in MMparticipants. Specifically, it was shown that the increases in beta-NTP levels following RE1 sleep were significantly larger in MMparticipants than those in control participants. This greater changein beta-NTP in MM participants was also significant during activeSD and following the RE2 sleep night, but the largest change in beta-NTP was following RE1 sleep. Additionally, the [31]P MRS imagingresults demonstrate that fluctuations in brain ATP (beta-NTP) were
associated with fluctuations in brain Pi and PCr levels with SD andRE sleep. These data are consistent with previous findings that PCris an energy carrier between sites of ATP production and utilization(Chance et al., 1988; Nishijima et al., 1989; Yoshizaki et al., 1989).Additionally, the larger brain beta-NTP increases in MM partici-

8 cohol D

pripia

doldbwilbpsitfabtb

Smsic1tcfilsIppntScaiowltl

pSaadWnplmrtp


ants over the SD paradigm were inversely linked to a trend foreductions in brain Pi levels and significant reductions in PCr levelsn MM participants over the course of the SD paradigm when com-ared to Pi and PCr levels in control participants. The reductions

n PCr and Pi levels in MM participants were most evident duringctive SD and following RE sleep.

The current findings may result in several interpretations. Theseata may suggest that metabolic sleep restoration processes whichccur during RE sleep are characterized by a build up of brain ATPevels in compensation for greater metabolic demands requireduring the prolonged wake period of SD. One perspective maye that the prolonged metabolic demands necessary to sustainake-time function with SD initiates a sleep homeostatic process

nvolving enhanced restoration of brain ATP following RE sleep. Fol-owing this view, the present findings of greater enhancements inrain ATP following RE in MM participants when compared controlarticipants may reflect a greater impact of SD to sleep homeosta-is. A greater impact of SD may imply a greater metabolic challengen MM participants when compared to healthy individuals, suchhat the physiological demands experienced with extended wake-ulness may be larger in these drug-dependent individuals. It maylso be the case that the disruption to sleep homeostasis inducedy SD results in sleep restoration processes which may be unableo properly regulate the brain phospholipid balance and maintainrain high-energy equilibrium.

An alternative view may be that the homeostatic response toD is characterized by a compensatory slowing of bioenergeticetabolism. It is plausible that during SD metabolic processes are

lowed down resulting in a build up of unutilized brain ATP. Thisnterpretation corresponds with numerous reports of region spe-ific reductions in metabolic activity during active SD (Wu et al.,991; Braun et al., 1997; Thomas et al., 2000), possibly indicatinghat one aspect of the homeostatic response to SD is to initiate aompensatory mode of metabolic conservation. Additional supportor this view is provided by the present [31]P MRS brain imag-ng results demonstrating that in MM participants that while theargest elevation in brain ATP levels was observed following REleep, brain ATP levels appear to begin to rise during active SD.nterestingly, the findings of Dorsey et al., in conjunction with theresent results illustrate that the elevations in brain ATP in controlarticipants were most evident following RE sleep and there wereo apparent changes in brain ATP levels during active SD. Underhese conditions it is conceivable that the homeostatic response toD in MM participants may be altered in that the level of metaboliconservation may be larger in methadone-maintained individu-ls. In contrast, the present results demonstrating correspondingncreases in brain PCr levels with increases in brain ATP levelsver the course of the SD paradigm indicate that with prolongedakefulness there is an enhancement in the utilization of ATP. Col-

ectively, it may be the case that both conditions take place, suchhat with SD a decreased metabolic state is initiated, but the pro-onged energy demands still require greater ATP utilization.

In summary, it was hypothesized that short-term MM partici-ants may exhibit the greatest differences over the course of theD paradigm, as stabilization in methadone treatment is associ-ted with numerous complications, such as alterations in sleeprchitecture and neuroendocrine function, as well as psychiatricysfunction (Judson and Goldstein, 1982; Woody et al., 1984;illenbring et al., 1989; Neshin, 1993). Sleep PSG results identified

otable differences in RE sleep in short-term MM participants com-ared to control participants, highlighting consequences of sleep

oss with SD and the importance of monitoring sleep early in treat-ent. Although long-term MM participants did not exhibit the

obust changes in PSG measures observed in short-term MM par-icipants, similar to short-term MM participants, long-term MMarticipants did exhibit greater elevations in beta-NTP measures of

ependence 106 (2010) 79–91

brain ATP levels following SD than control participants. The datahighlights the notion that although individuals who have been sta-bilized to methadone-maintenance may not exhibit differences insleep PSG measures in response to SD, long-term MM individualsmay have persistent underlying changes in brain metabolic and/orsleep restorative processes.

Collectively, The present study identified a differential responseto SD in MM participants compared to control participants, as mea-sured by PSG and [31]P MRS brain imaging. The lack of increasesin TST and SEI during RE sleep in MM participants (particularlyshort-term MM) indicate disrupted RE sleep, which may result in aprolonging the RE sleep restoration processes into subsequent REsleep nights. It is likely that MM participants may require moresleep nights to recover from the impact of sleep loss. This the-ory may be congruent with the notion that SD results in a greaterimpact to sleep homeostasis in MM individuals. The results iden-tifying differences in brain ATP levels in MM participants mayreflect an inability to properly regulate brain bioenergetics inresponse to sleep loss. Particularly with methadone-maintenance,the SD challenge may result in an over compensation in therestoration of brain ATP, indicative of an increase in brain ATP pro-duction or a reduction in ATP metabolism. A relationship may existbetween SD-induced metabolic conservation and SD-induced cog-nitive performance deficits, potentially leading to greater cognitiveperformance deficits in MM individuals. Previous brain imagingstudies examining the effects of SD have implicated cognitive task-specific changes in frontal brain activations (Drummond et al.,2005; Zeitzer et al., 2006) and regional decreases in frontal brainmetabolism, which were also observed after RE sleep (Wu et al.,2006). The alterations in sleep homeostatic and sleep restorationprocesses in MM participants in response to SD may have con-sequences in other physiological processes that require a highmetabolic demand. Sleep loss in MM participants may disruptrestorative sleep processes, subsequent sleep patterns, and impactdaytime wake function. Sleep in MM patients is important asinsomnia is associated with psychiatric disorders, depression, anda decline in executive function (Chuah et al., 2006).

In consideration of neurophysiological mechanisms under-lying the present objective sleep and brain MRS imagingfindings, one plausible factor may be the function of thehypothalamic–pituitary–adrenal (HPA) axis. The HPA axis is aprimary component of the neuroendocrine system that regu-lates physiological functions such as immunity, mood, and energymetabolism and the response to stress (Hauger and Datzenberg,2000). The HPA axis also function to maintain alertness and insleep regulation. Dysfunction in the HPA axis such as alterationsto corticotrophin-releasing hormone (CRH) or adrenocorticotropichormone (ACTH) can disrupt sleep (Lesch et al., 1988; Garcia-Borreguero et al., 2000; Rodenbeck et al., 2002; Buckley andSchatzberg, 2005). Cortisol is secreted during the stress responseand stress tends to worsen sleep leading to theories that HPAaxis abnormalities may directly contribute to some sleep disor-ders, such as insomnia (Buckley and Schatzberg, 2005). A studydemonstrated that SD results in a significant reduction of cortisolsecretion which is under the control of increase of slow wave sleepduring the recovery night (Vgontzas et al., 1999). One intriguingexample is the relationship of HPA axis function and depres-sion, as depression is often associated with sleep disturbances andhypercortisolemia (Arborelius et al., 1999). Investigations seek-ing to elucidate the neurobiological mechanisms associated withthe therapeutic effects of SD in depressed patients have demon-
strated that the response to SD in depressed patients results indecreased cortisol levels highlighting decreased HPA axis activ-ity during RE sleep (Arborelius et al., 1999; Vgontzas et al., 1999;Murck et al., 2006). Similarly, opioid addiction and methadonetreatment have been shown to result in alterations to noradren-

ohol D

e(ow(btriSitf1tahth(AtacmMr

tittieoiimmamttdtupmorataettdtntIetSmp


rgic regulation and the function of the stress-responsive HPA axisStine et al., 2001). For example, findings suggest that chronic opi-id dependence results in reduced HPA axis function, while opioidithdrawal may decrease pituitary and increase adrenal responses

Zhang et al., 2008). Although it was initially thought that pertur-ations in HPA axis function become normalized with methadonereatment stabilization (Kreek et al., 1983; Kreek et al., 1984), moreecent studies have observed HPA axis abnormalities, which persistnto stable long-term methadone-maintenance (Stine et al., 2002;chluger et al., 2003). It has been postulated that hypoadrenalismn methadone addicts might explain some of the negative symp-oms frequently associated with methadone treatment, includingatigue, weakness, and depression (Dackis et al., 1982; Pullan et al.,983). It has also been shown that alcohol consumption results inhe activation of the hypothalamus–pituitary–adrenal (HPA) axis,nd with more persistent alterations of the HPA axis existing ineavy drinkers (Badrick et al., 2008; Wong et al., 2008). In contrasto the hypoadrenalism noted with methadone-maintenance, dataas shown that chronic heavy drinking results in hypercortisolismBeresford et al., 2006; Badrick et al., 2008; Wong et al., 2008).lthough the precise neurobiology remains uncertain, based upon

he relationship of the HPA axis to sleep and energy metabolismnd the identification of methadone-induced hypoadrenalism, it isonceivable that HPA axis dysfunction in MM participants is a pri-ary mechanism driving the current objective sleep measures andRS brain imaging findings of a differential RE sleep homeostatic

esponse to SD.While the current findings highlight the importance of long-

erm treatment stabilization in methadone-maintenance, anothermportant and related aspect is that potential functional disrup-ions such as reduced cognitive and/or executive function(s) dueo drug abuse-related insomnia may directly contribute to anncrease drug relapse potential (Ford and Kamerow, 1989; Staedtt al., 1996; Beswick et al., 2003). For example, a greater impactf acute sleep loss early in methadone treatment may directlyncrease the propensity for opioid relapse furthermore imped-ng the patients’ ability to achieve prolonged participation in

ethadone-maintenance. In other words, MM patients who areore susceptible to certain conditions such as acute sleep loss

nd are unable to become stabilized to methadone-maintenanceay become unable to reach long-term MM status. Limitations of

his study include the limited sample size, which may contributeo less than ideal statistical power in examining treatment × sexifferences, and potential differences in participant drug use his-ory. While the present study excluded MM participants basedpon dependence to drugs other than opiates or nicotine and forrimary psychiatric disorders, a common feature of methadone-aintenance is polydrug abuse, drug chipping, and the occurrence

f depression. Studies have shown that methadone-maintenanceesults in a greater occurrence of central sleep apnea (Teichtahl etl., 2004; Teichtahl and Wang, 2007; Webster et al., 2008). Addi-ionally, while participants exceeding criteria for respiratory eventsnd periodic leg movements (PLMs) during study screening werexcluded from the study, a potential limitation may exist fromhe contribution of fluctuations in respiratory and/or PLMs overhe course of the study sleep nights. The contribution of poly-rug abuse, depressive symptomology, or central sleep apnea tohe effects of SD during methadone-maintenance treatment areot examined by the current study and the potential interaction ofhese factors may be important considerations with MM patients.n lieu of the current findings, methadone-maintenance remains an
ffective and affordable treatment for opioid-dependence. Overall,he results highlight the biological sleep homeostatic response toD in a drug-dependent population, underscoring the need for closeonitoring of sleep loss or disturbances in MM patients, and sup-
ort the notion that SD-induced cognitive deficits may be more

ependence 106 (2010) 79–91 89

prevalent or of a larger magnitude in drug-dependent individu-als. Further characterization is necessary to fully understand theimpact of sleep disturbances in MM patients on sleep behavior andcognition as well as on the propensity for drug relapse.

Role of funding source

This research was financially supported through the generosityof the National Institute on Drug Abuse (NIDA) grants DA016542,T32DA15036, KO2DA017324, K24DA15116, and KO5DA000343,and in part by the Counter-Drug Technology Assessment Center(CTAC), an office within the Office of National Drug Control Policy(ONDCP), via Contract Number DABT63-99-C awarded by the ArmyContracting Agency. The content of the information does not nec-essarily reflect the position or the policy of the Government andno official endorsement should be inferred. This project also wassponsored in part by National Institutes of Health (NIH) grant S10RR13938 NIDA, CTAC, and NIH had no further role in study design;in the collection, analysis and interpretation of data; in the writingof the report; or in the decision to submit the paper for publication.

Contributors

Authors Scott E. Lukas, Cynthia M. Dorsey, and Perry F. Ren-shaw, designed the study and wrote the protocol. Author GeorgeH. Trksak managed the literature searches and summaries of previ-ous related work. Authors George H. Trksak, J. Eric Jensen, David M.Penetar, David T. Plante, Michael Brendel, Melissa A. Maywalt, andWendy Tartarini carried out study procedures and collected studydata. Author George H. Trksak undertook the statistical analysis.Author George H. Trksak wrote the first draft of the manuscript andprepared the present revised manuscript. All authors contributed toand have approved the final manuscript and the current subsequentrevised manuscript submission.

Conflict of interest

There are no conflicts of interest related to George H. Trksak,J. Eric Jensen, David T. Plante, David Penetar, Wendy L. Tartarini,Melissa A. Maywalt, Michael Brendel, Cynthia M. Dorsey, Perry F.Renshaw, Scott E. Lukas.

References

Adam, K., Oswald, I., 1977. Sleep is for tissue restoration. Journal of the Royal Collegeof Physicians of London 11, 376–388.

Akerstedt, T., Kecklund, G., Ingre, M., Lekander, M., Axelsson, J., 2009. Sleep home-ostasis during repeated sleep restriction and recovery: support from EEGdynamics. Sleep 32, 217–222.

Arborelius, L., Owens, M.J., Plotsky, P.M., Nemeroff, C.B., 1999. The role ofcorticotropin-releasing factor in depression and anxiety disorders. The Journalof Endocrinology 160, 1–12.

Armitage, R., Hoffmann, R., Fitch, T., Morel, C., Bonato, R., 1995. A comparison ofperiod amplitude and power spectral analysis of sleep EEG in normal adults anddepressed outpatients. Psychiatry Research 56, 245–256.

Athanasos, P., Smith, C.S., White, J.M., Somogyi, A.A., Bochner, F., Ling, W., 2006.Methadone maintenance patients are cross-tolerant to the antinociceptiveeffects of very high plasma morphine concentrations. Pain 120, 267–275.

Badrick, E., Bobak, M., Britton, A., Kirschbaum, C., Marmot, M., Kumari, M., 2008. Therelationship between alcohol consumption and cortisol secretion in an agingcohort. Journal of Clinical Endocrinology and Metabolism 93, 750–757.

Benington, J.H., Heller, H.C., 1995. Restoration of brain energy metabolism as thefunction of sleep. Progress in Neurobiology 45, 347–360.

Beresford, T.P., Arciniegas, D.B., Alfers, J., Clapp, L., Martin, B., Beresford, H.F., Du,Y., Liu, D., Shen, D., Davatzikos, C., Laudenslager, M.L., 2006. Hypercortisolism
in alcohol dependence and its relation to hippocampal volume loss. Journal ofStudies on Alcohol 67, 861–867.
Besset, A., Villemin, E., Tafti, M., Billiard, M., 1998. Homeostatic process and sleepspindles in patients with sleep-maintenance insomnia: effect of partial (21 h)sleep deprivation. Electroencephalography and Clinical Neurophysiology 107,122–132.

9 cohol D

B

B

B

B

C

C

C

C

C

D

D

D

D

D

D

D

D

D

E

E

E

F

F

F

F

G

G

G


eswick, T., Best, D., Rees, S., Bearn, J., Gossop, M., Strang, J., 2003. Major disrup-tions of sleep during treatment of the opiate withdrawal syndrome: differencesbetween methadone and lofexidine detoxification treatments. Addiction Biol-ogy 8, 49–57.

raun, A.R., Balkin, T.J., Wesenten, N.J., Carson, R.E., Varga, M., Baldwin, P., Selbie,S., Belenky, G., Herscovitch, P., 1997. Regional cerebral blood flow through-out the sleep–wake cycle. An H2(15)O PET study. Brain 120 (Pt 7), 1173–1197.

rown, B.S., Watters, J.K., Iglehart, A.S., 1982. Methadone maintenance dosage lev-els and program retention. The American Journal of Drug and Alcohol Abuse 9,129–139.

uckley, T.M., Schatzberg, A.F., 2005. On the interactions of thehypothalamic–pituitary–adrenal (HPA) axis and sleep: normal HPA axisactivity and circadian rhythm, exemplary sleep disorders. The Journal ofClinical Endocrinology and Metabolism 90, 3106–3114.

acciola, J.S., Alterman, A.I., Rutherford, M.J., McKay, J.R., McLellan, A.T., 1998. Theearly course of change in methadone maintenance. Addiction (Abingdon, Eng-land) 93, 41–49.

hance, B., Waterland, R.A., Tanaka, A., Poyton, R.O., 1988. Mitochondrial functionin normal and genetically altered cells and tissues. Annals of the New YorkAcademy of Sciences 550, 360–373.

huah, Y.M., Venkatraman, V., Dinges, D.F., Chee, M.W., 2006. The neural basis ofinterindividual variability in inhibitory efficiency after sleep deprivation. Journalof Neuroscience 26, 7156–7162.

ondelli, W.S., Dunteman, G.H., 1993. Exposure to methadone programs and heroinuse. The American Journal of Drug and Alcohol Abuse 19, 65–78.

onnell, P.H., 1967. Centres for treatment of drug addiction. Importance of research.British Medical Journal 2, 499–500.

ackis, C.A., Gurpegui, M., Pottash, A.L., Gold, M.S., 1982. Methadone induced hypoa-drenalism. Lancet 2, 1167.

ang-Vu, T.T., Schabus, M., Desseilles, M., Albouy, G., Boly, M., Darsaud, A., Gais,S., Rauchs, G., Sterpenich, V., Vandewalle, G., Carrier, J., Moonen, G., Balteau,E., Degueldre, C., Luxen, A., Phillips, C., Maquet, P., 2008. Spontaneous neuralactivity during human slow wave sleep. Proceedings of the National Academyof Sciences of the United States of America 105, 15160–15165.

e Gennaro, L., Ferrara, M., Spadini, V., Curcio, G., Cristiani, R., Bertini, M., 2002. Thecyclic alternating pattern decreases as a consequence of total sleep deprivationand correlates with EEG arousals. Neuropsychobiology 45, 95–98.

imsdale, J.E., Norman, D., DeJardin, D., Wallace, M.S., 2007. The effect of opioids onsleep architecture. Journal of Clinical Sleep Medicine 3, 33–36.

orsey, C.M., Lukas, S.E., Teicher, M.H., Harper, D., Winkelman, J.W., Cunningham,S.L., Satlin, A., 1996. Effects of passive body heating on the sleep of older femaleinsomniacs. Journal of Geriatric Psychiatry and Neurology 9, 83–90.

orsey, C.M., Lukas, S.E., Moore, C.M., Tartarini, W.L., Parow, A.M., Villafuerte, R.A.,Renshaw, P.F., 2003. Phosphorous31 magnetic resonance spectroscopy aftertotal sleep deprivation in healthy adult men. Sleep 26, 573–577.

overty, M., Somogyi, A.A., White, J.M., Bochner, F., Beare, C.H., Menelaou, A., Ling,W., 2001. Methadone maintenance patients are cross-tolerant to the antinoci-ceptive effects of morphine. Pain 93, 155–163.

rummond, S.P., Meloy, M.J., Yanagi, M.A., Orff, H.J., Brown, G.G., 2005. Com-pensatory recruitment after sleep deprivation and the relationship withperformance. Psychiatry Research 140, 211–223.

uncan Jr., W.C., Gillin, J.C., Post, R.M., Gerner, R.H., Wehr, T.A., 1980. Relationshipbetween EEG sleep patterns and clinical improvement in depressed patientstreated with sleep deprivation. Biological Psychiatry 15, 879–889.

ap, C.B., Buclin, T., Baumann, P., 2002. Interindividual variability of the clini-cal pharmacokinetics of methadone: implications for the treatment of opioiddependence. Clinical Pharmacokinetics 41, 1153–1193.

ugenio, K.R., 2004. Profound morphine tolerance following high-dose methadonetherapy. Journal of Pain and Palliative Care Pharmacotherapy 18, 47–54.

verson, C.A., Laatsch, C.D., Hogg, N., 2005. Antioxidant defense responses to sleeploss and sleep recovery. American Journal of Physiology. Regulatory, Integrativeand Comparative Physiology 288, R374–383.

inelli, L.A., Baumann, H., Borbely, A.A., Achermann, P., 2000. Dual electroencephalo-gram markers of human sleep homeostasis: correlation between theta activityin waking and slow-wave activity in sleep. Neuroscience 101, 523–529.

ord, D.E., Kamerow, D.B., 1989. Epidemiologic study of sleep disturbancesand psychiatric disorders. An opportunity for prevention? JAMA 262, 1479–1484.

rank, M.G., 2006. The mystery of sleep function: current perspectives and futuredirections. Reviews in the Neurosciences 17, 375–392.

riedmann, P.D., Hendrickson, J.C., Gerstein, D.R., Zhang, Z., Stein, M.D., 2006. Domechanisms that link addiction treatment patients to primary care influencesubsequent utilization of emergency and hospital care? Medical Care 44, 8–15.

arcia-Borreguero, D., Wehr, T.A., Larrosa, O., Granizo, J.J., Hardwick, D., Chrousos,G.P., Friedman, T.C., 2000. Glucocorticoid replacement is permissive for rapid eyemovement sleep and sleep consolidation in patients with adrenal insufficiency.The Journal of Clinical Endocrinology and Metabolism 85, 4201–4206.

illin, J.C., Smith, T.L., Irwin, M., Kripke, D.F., Schuckit, M., 1990. EEG sleep stud-ies in “pure” primary alcoholism during subacute withdrawal: relationships to
normal controls, age, and other clinical variables. Biological Psychiatry 27, 477–488.
ip, P., Hagiwara, G., Sapolsky, R.M., Cao, V.H., Heller, H.C., Ruby, N.F., 2004. Gluco-corticoids influence brain glycogen levels during sleep deprivation. AmericanJournal of Physiology. Regulatory, Integrative and Comparative Physiology 286,R1057–1062.

ependence 106 (2010) 79–91

Gossop, M., Bradley, B., 1984. Insomnia among addicts during supervised withdrawalfrom opiates: a comparison of oral methadone and electrostimulation. Drug andAlcohol Dependence 13, 191–198.

Gossop, M., Trakada, K., Stewart, D., Witton, J., 2005. Reductions in criminal convic-tions after addiction treatment: 5-year follow-up. Drug and Alcohol Dependence79, 295–302.

Gottheil, E., Sterling, R.C., Weinstein, S.P., 1993. Diminished illicit drug use as a con-sequence of long-term methadone maintenance. Journal of Addictive Diseases12, 45–57.

Hauger, R.L., Datzenberg, F.M., 2000. Regulation of the stress response bycorticotropin-releasing factor receptors. In: Conn, P.M., Freeman, M.E. (Eds.),Neuroendocrinology in physiology and medicine. Humana Press, Totowa.

Hoffman, J.A., Moolchan, E.T., 1994. The phases-of-treatment model for methadonemaintenance: implementation and evaluation. Journal of Psychoactive Drugs26, 181–197.

Horne, J.A., Staff, L.H., 1983. Exercise and sleep: body-heating effects. Sleep 6, 36–46.Howe, R.C., Phillips, J.L., Hegge, F.W., 1980a. Acute heroin abstinence in man. III. Effect

upon waking and slow wave sleep. Drug and Alcohol Dependence 6, 247–262.Howe, R.C., Hegge, F.W., Phillips, J.L., 1980b. Acute heroin abstinence in man. I.

Changes in behavior and sleep. Drug and Alcohol Dependence 5, 341–356.Howe, R.C., Phillips, J.L., Hegge, F.W., 1981. Acute heroin abstinence in man. IV.

Sleep–waking state contingencies. Drug and Alcohol Dependence 7, 163–176.Irwin, M., Gillin, J.C., Dang, J., Weissman, J., Phillips, E., Ehlers, C.L., 2002. Sleep

deprivation as a probe of homeostatic sleep regulation in primary alcoholics.Biological Psychiatry 51, 632–641.

Jensen, J.E., Drost, D.J., Menon, R.S., Williamson, P.C., 2002. In vivo brain (31)P-MRS: measuring the phospholipid resonances at 4 Tesla from small voxels. NMRBiomedicine 15, 338–347.

Johnson, M.L., Belenky, G., Redmond, D.P., Thorne, D.R., Williams, J.D., Hursh, S.R.,Balkin, T.J., 2004. Modulating the homeostatic process to predict performanceduring chronic sleep restriction. Aviation, Space, and Environmental Medicine75, A141–146.

Judson, B.A., Goldstein, A., 1982. Symptom complaints of patients maintained onmethadone, LAAM (methadyl acetate), and naltrexone at different times in theiraddiction careers. Drug and Alcohol Dependence 10, 269–282.

Kay, D.C., 1975. Human sleep and EEG through a cycle of methadone dependence.Electroencephalography and Clinical Neurophysiology 38, 35–43.

King, V.L., Brooner, R.K., 2008. Improving treatment engagement in opioid-dependent outpatients with a motivated stepped-care adaptive treatmentmodel. Joint Commission Journal of Quality and Patient Safety/Joint CommissionResources 34, 209–213.

Kong, J., Shepel, P.N., Holden, C.P., Mackiewicz, M., Pack, A.I., Geiger, J.D., 2002.Brain glycogen decreases with increased periods of wakefulness: implicationsfor homeostatic drive to sleep. Journal of Neuroscience 22, 5581–5587.

Kreek, M.J., Ragunath, J., Plevy, S., Hamer, D., Schneider, B., Hartman, N., 1984. ACTH,cortisol and beta-endorphin response to metyrapone testing during chronicmethadone maintenance treatment in humans. Neuropeptides 5, 277–278.

Kreek, M.J., Wardlaw, S.L., Hartman, N., Raghunath, J., Friedman, J., Schneider, B.,Frantz, A.G., 1983. Circadian rhythms and levels of beta-endorphin, ACTH, andcortisol during chronic methadone maintenance treatment in humans. Life sci-ences 33 (Suppl. 1), 409–411.

Lesch, K.P., Widerlov, E., Ekman, R., Laux, G., Schulte, H.M., Pfuller, H., Beckmann, H.,1988. Delta sleep-inducing peptide response to human corticotropin-releasinghormone (CRH) in major depressive disorder. Comparison with CRH-inducedcorticotropin and cortisol secretion. Biological Psychiatry 24, 162–172.

Leukefeld, C.G., Tims, F.M., 1990. Compulsory treatment for drug abuse. InternationalJournal of Addiction 25, 621–640.

Ling, W., Wesson, D.R., 1990. Drugs of abuse—opiates. The Western Journal ofMedicine 152, 565–572.

Murck, H., Uhr, M., Ziegenbein, M., Kunzel, H., Held, K., Antonijevic, I.A., Schus-sler, P., Steiger, A., 2006. Renin-angiotensin-aldosterone system, HPA-axis andsleep-EEG changes in unmedicated patients with depression after total sleepdeprivation. Pharmacopsychiatry 39, 23–29.

Neshin, S.F., 1993. Methadone maintenance treatment: dispelling myths and recov-ering truths. New Jersey Medicine 90, 850–852.

Nicholas, C.L., Sullivan, E.V., Pfefferbaum, A., Trinder, J., Colrain, I.M., 2002. Theeffects of alcoholism on auditory evoked potentials during sleep. Journal of SleepResearch 11, 247–253.

Nishijima, M.K., Koehler, R.C., Hurn, P.D., Eleff, S.M., Norris, S., Jacobus, W.E.,Traystman, R.J., 1989. Postischemic recovery rate of cerebral ATP, phosphocre-atine, pH, and evoked potentials. American Journal of Physiology 257, H1860–1870.

Nissen, C., Feige, B., Konig, A., Voderholzer, U., Berger, M., Riemann, D., 2001. Deltasleep ratio as a predictor of sleep deprivation response in major depression.Journal of Psychiatric Research 35, 155–163.

Oyefeso, A., Sedgwick, P., Ghodse, H., 1997. Subjective sleep–wake parameters intreatment-seeking opiate addicts. Drug and Alcohol Dependence 48, 9–16.

Peles, E., Schreiber, S., Adelson, M., 2006. Variables associated with perceived sleepdisorders in methadone maintenance treatment (MMT) patients. Drug and Alco-hol Dependence 82, 103–110.

Peles, E., Linzy, S., Kreek, M., Adelson, M., 2008. One-year and cumulative retentionas predictors of success in methadone maintenance treatment: a comparisonof two clinics in the United States and Israel. Journal of Addictive Diseases 27,11–25.

Pickworth, W.B., Neidert, G.L., Kay, D.C., 1981. Morphinelike arousal by methadoneduring sleep. Clinical Pharmacology & Therapeutics 30, 796–804.

ohol D

P

P

P

P

P

P

R

R

R

R

R

R

R

R

S

S

S

S

S

S

S

S

S

T

T


ilcher, J.J., Huffcutt, A.I., 1996. Effects of sleep deprivation on performance: a meta-analysis. Sleep 19, 318–326.

onder, S.L., Twieg, D.B., 1994. A novel sampling method for 31P spectroscopic imag-ing with improved sensitivity, resolution, and sidelobe suppression. Journal ofMagnetic Resonance B 104, 85–88.

orkka-Heiskanen, T., Strecker, R.E., Thakkar, M., Bjorkum, A.A., Greene, R.W., McCar-ley, R.W., 1997. Adenosine: a mediator of the sleep-inducing effects of prolongedwakefulness. Science (New York, NY) 276, 1265–1268.

otwarka, J.J., Drost, D.J., Williamson, P.C., 1999. Quantifying 1H decoupled in vivo31P brain spectra. NMR Biomedicine 12, 8–14.

rinz, P.N., Roehrs, T.A., Vitaliano, P.P., Linnoila, M., Weitzman, E.D., 1980. Effectof alcohol on sleep and nighttime plasma growth hormone and cortisol con-centrations. The Journal of Clinical Endocrinology and Metabolism 51, 759–764.

ullan, P.T., Watson, F.E., Seow, S.S., Rappeport, W., 1983. Methadone-inducedhypoadrenalism. Lancet 1, 714.

adulovacki, M., 1985. Role of adenosine in sleep in rats. Reviews in Clinical & BasicPharmacology 5, 327–339.

echtschaffen, A., Kales, A., 1968. A Manual of Standardized Terminology Techniquesand Scoring System for Sleep Stages of Human Subjects. University of Californiaat Los Angeles, Brain Information Service/Brain Research Institute.

enshaw, P.F., Parow, A.M., Hirashima, F., Ke, Y., Moore, C.M., Frederick Bde, B.,Fava, M., Hennen, J., Cohen, B.M., 2001. Multinuclear magnetic resonance spec-troscopy studies of brain purines in major depression. American Journal ofPsychiatry 158, 2048–2055.

eyero, F., Muller, C., 1977. Repeated sessions of sleep deprivation in the treatmentof depression. L’Encephale 3, 755–761.

eynolds 3rd, C.F., Kupfer, D.J., Hoch, C.C., Stack, J.A., Houck, P.A., Berman, S.R., 1987.Sleep deprivation effects in older endogenous depressed patients. PsychiatryResearch 21, 95–109.

odenbeck, A., Huether, G., Ruther, E., Hajak, G., 2002. Interactions between eveningand nocturnal cortisol secretion and sleep parameters in patients with severechronic primary insomnia. Neuroscience Letters 324, 159–163.

oehrs, T., Papineau, K., Rosenthal, L., Roth, T., 1999. Ethanol as a hypnotic ininsomniacs: self-administration and effects on sleep and mood. Neuropsy-chopharmacology 20, 279–286.

oubicek, J., Zaks, A., Freedman, A.M., 1969. EEG changes produced by heroinand methadone. Electroencephalography and Clinical Neurophysiology 27,667–668.

alamy, A., Williams, H.L., 1973. The effects of alcohol on sensory evoked andspontaneous cerebral potentials in man. Electroencephalography and ClinicalNeurophysiology 35, 3–11.

chluger, J.H., Bart, G., Green, M., Ho, A., Kreek, M.J., 2003. Corticotropin-releasingfactor testing reveals a dose-dependent difference in methadone maintained vs.control subjects. Neuropsychopharmacology 28, 985–994.

enay, E.C., 1985. Methadone maintenance treatment. International Journal ofAddiction 20, 803–821.

ilveri, M.M., Pollack, M.H., Diaz, C.I., Nassar, L.E., Mendelson, J.H., Yurgelun-Todd,D.A., Renshaw, P.F., Kaufman, M.J., 2004. Cerebral phosphorus metabolite andtransverse relaxation time abnormalities in heroin-dependent subjects at onsetof methadone maintenance treatment. Psychiatry Research 131, 217–226.

taedt, J., Wassmuth, F., Stoppe, G., Hajak, G., Rodenbeck, A., Poser, W., Ruther, E.,1996. Effects of chronic treatment with methadone and naltrexone on sleepin addicts. European Archives of Psychiatry and Clinical Neuroscience 246,305–309.

tein, M.D., Herman, D.S., Bishop, S., Lassor, J.A., Weinstock, M., Anthony, J., Anderson,B.J., 2004. Sleep disturbances among methadone maintained patients. Journal ofSubstance Abuse Treatment 26, 175–180.

tine, S.M., Grillon, C.G., Morgan 3rd, C.A., Kosten, T.R., Charney, D.S., Krystal, J.H.,2001. Methadone patients exhibit increased startle and cortisol response afterintravenous yohimbine. Psychopharmacology 154, 274–281.

tine, S.M., Southwick, S.M., Petrakis, I.L., Kosten, T.R., Charney, D.S., Krystal,J.H., 2002. Yohimbine-induced withdrawal and anxiety symptoms in opioid-dependent patients. Biological Psychiatry 51, 642–651.

zpanowska-Wohn, A., Kolarzyk, E., Pach, D., Targosz, D., 2004. Nutritional status of
opiate-dependent persons before and during methadone maintenance therapy.Przeglad Lekarski 61, 339–344.
eichtahl, H., Wang, D., 2007. Sleep-disordered breathing with chronic opioid use.Expert Opinion on Drug Safety 6, 641–649.

eichtahl, H., Wang, D., Cunnington, D., Kronborg, I., Goodman, C., Pro-dromidis, A., Drummer, O., 2004. Cardiorespiratory function in stable

ependence 106 (2010) 79–91 91

methadone maintenance treatment (MMT) patients. Addiction Biology 9, 247–253.

Thomas, M., Sing, H., Belenky, G., Holcomb, H., Mayberg, H., Dannals, R., Wagner,H., Thorne, D., Popp, K., Rowland, L., Welsh, A., Balwinski, S., Redmond, D., 2000.Neural basis of alertness and cognitive performance impairments during sleepi-ness. I. Effects of 24h of sleep deprivation on waking human regional brainactivity. Journal of Sleep Research 9, 335–352.

Ticho, S.R., Radulovacki, M., 1991. Role of adenosine in sleep and temperature reg-ulation in the preoptic area of rats. Pharmacology, Biochemistry, and Behavior40, 33–40.

Tilley, A., Donohoe, F., Hensby, S., 1987. Homeostatic changes in slow wave sleepduring recovery sleep following restricted nocturnal sleep and partial slow wavesleep recovery during an afternoon nap. Sleep 10, 600–605.

van den Burg, W., van den Hoofdakker, R.H., 1975. Total sleep deprivation on endoge-nous depression. Archives of General Psychiatry 32, 1121–1125.

Van Dongen, H.P., Maislin, G., Mullington, J.M., Dinges, D.F., 2003. The cumula-tive cost of additional wakefulness: dose–response effects on neurobehavioralfunctions and sleep physiology from chronic sleep restriction and total sleepdeprivation. Sleep 26, 117–126.

Vgontzas, A.N., Mastorakos, G., Bixler, E.O., Kales, A., Gold, P.W., Chrousos, G.P., 1999.Sleep deprivation effects on the activity of the hypothalamic–pituitary–adrenaland growth axes: potential clinical implications. Clinical Endocrinology (Oxford)51, 205–215.

Volavka, J., Verebey, K., Resnick, R., Mule, S., 1978. Methadone dose, plasma level,and cross-tolerance to heroin in man. The Journal of Nervous and Mental Disease166, 104–109.

Wagman, A.M., Allen, R.P., Funderburk, F., Upright, D., 1978. EEG measures of func-tional tolerance to alcohol. Biological Psychiatry 13, 719–728.

Webster, L.R., Choi, Y., Desai, H., Webster, L., Grant, B.J., 2008. Sleep-disorderedbreathing and chronic opioid therapy. Pain Medicine (Malden, MA) 9, 425–432.

Weddington, W.W., 1990. Towards a rehabilitation of methadone maintenance:integration of relapse prevention and aftercare. International Journal of Addic-tion 25, 1201–1224.

Willenbring, M.L., Morley, J.E., Krahn, D.D., Carlson, G.A., Levine, A.S., Shafer, R.B.,1989. Psychoneuroendocrine effects of methadone maintenance. Psychoneu-roendocrinology 14, 371–391.

Wong, W.M., Hasemann, S., Schwarz, M., Zill, P., Koller, G., Soyka, M., Preuss,U.W., 2008. Citalopram neuropharmacological challenge in alcohol-dependentpatients and controls: pharmacogenetic, endocrine and psychobehavioralresults. Pharmacopsychiatry 41, 72–78.

Woody, G.E., McLellan, A.T., O’Brien, C.P., 1984. Treatment of behavioral and psychi-atric problems associated with opiate dependence. NIDA Research Monograph46, 23–35.

Wu, J.C., Gillin, J.C., Buchsbaum, M.S., Hershey, T., Hazlett, E., Sicotte, N., BunneyJr., W.E., 1991. The effect of sleep deprivation on cerebral glucose metabolicrate in normal humans assessed with positron emission tomography. Sleep 14,155–162.

Wu, J.C., Gillin, J.C., Buchsbaum, M.S., Chen, P., Keator, D.B., Khosla Wu, N., Darnall,L.A., Fallon, J.H., Bunney, W.E., 2006. Frontal lobe metabolic decreases with sleepdeprivation not totally reversed by recovery sleep. Neuropsychopharmacology31, 2783–2792.

Yoshizaki, K., Takanashi, Y., Motonaga, T., Naruse, S., Nishikawa, H., Hirakawa, K.,1989. Effect of cerebral ischemia and hypercapnia on cerebral pH studied with31P-NMR and electrical activity in rat brain. Japan Journal of Physiology 39,155–167.

Zaks, A., Fink, M., Freedman, A.M., 1971. Duration of methadone induced cross-tolerance to heroin. British Journal of Addiction of Alcohol and Other Drugs 66,205–208.

Zeitzer, J.M., Morales-Villagran, A., Maidment, N.T., Behnke, E.J., Ackerson, L.C.,Lopez-Rodriguez, F., Fried, I., Engel Jr., J., Wilson, C.L., 2006. Extracellularadenosine in the human brain during sleep and sleep deprivation: an in vivomicrodialysis study. Sleep 29, 455–461.

Zhang, G.F., Ren, Y.P., Sheng, L.X., Chi, Y., Du, W.J., Guo, S., Jiang, Z.N., Xiao, L., Luo,X.N., Tang, Y.L., Smith, A.K., Liu, Z.Q., Zhang, H.X., 2008. Dysfunction of the
hypothalamic–pituitary–adrenal axis in opioid dependent subjects: effects ofacute and protracted abstinence. The American Journal of Drug and AlcoholAbuse 34, 760–768.
Zweben, J.E., Payte, J.T., 1990. Methadone maintenance in the treatment of opi-oid dependence. A current perspective. The Western Journal of Medicine 152,588–599.

Documents

Effects of sleep deprivation on sleep homeostasis and restoration during methadone-maintenance: A [31]P MRS brain imaging study