PreMedline22zizi.com

ResearchEfficacyandsafetyofexogenousmelatoninforsecondarysleepdisordersandsleepdisordersaccompanyingsleeprestriction:meta-analysisNinaBuscemi,BenVandermeer,NicolaHooton,RenaPandya,LisaTjosvold,LisaHartling,SunitaVohra,TerryPKlassen,GlenBakerAbstractObjectiveToconductasystematicreviewoftheefficacyandsafetyofexogenousmelatonininmanagingsecondarysleepdisordersandsleepdisordersaccompanyingsleeprestriction,suchasjetlagandshiftworkdisorder.
Datasources13electronicdatabasesandreferencelistsofrelevantreviewsandincludedstudies;AssociatedProfessionalSleepSocietyabstracts(1999to2003).
StudyselectionTheefficacyreviewincludedrandomisedcontrolledtrials;thesafetyreviewincludedrandomisedandnon-randomisedcontrolledtrials.
QualityassessmentRandomisedcontrolledtrialswereassessedbyusingtheJadadScaleandcriteriabySchulzetal,andnon-randomisedcontrolledtrialsbytheDownsandBlackchecklist.
DataextractionandsynthesisOnereviewerextracteddataandanotherreviewerverifiedthedataextracted.
Theinversevariancemethodwasusedtoweightstudiesandtherandomeffectsmodelwasusedtoanalysedata.
MainresultsSixrandomisedcontrolledtrialswith97participantsshowednoevidencethatmelatoninhadaneffectonsleeponsetlatencyinpeoplewithsecondarysleepdisorders(weightedmeandifference13.
2(95%confidenceinterval27.
3to0.
9)min).
Ninerandomisedcontrolledtrialswith427participantsshowednoevidencethatmelatoninhadaneffectonsleeponsetlatencyinpeoplewhohadsleepdisordersaccompanyingsleeprestriction(1.
0(2.
3to0.
3)min).
17randomisedcontrolledtrialswith651participantsshowednoevidenceofadverseeffectsofmelatoninwithshorttermuse(threemonthsorless).
ConclusionsThereisnoevidencethatmelatoniniseffectiveintreatingsecondarysleepdisordersorsleepdisordersaccompanyingsleeprestriction,suchasjetlagandshiftworkdisorder.
Thereisevidencethatmelatoninissafewithshorttermuse.
IntroductionSleepdisordersaffectapproximately20%oftheAmericanpopulation.
1Asleepdisorderexistswheneveralowerqualityofsleepleadstoimpairedfunctioningorexcessivesleepiness.
2Sleepdisordersplaceaburdenonsocietyduetotheirnegativeimpactonqualityoflife,safety,productivity,andhealthcareutili-sation.
Onecategoryofsleepdisordersissecondarysleepdisorders,sleepproblemsthatareassociatedwithmedical,neurological,orsubstancemisusedisorders.
Anothercategoryofsleepdisordersarisesfromsleeprestriction:inadequatesleepresultsfromimposedorselfimposedlifestyleandworkschedules,suchasairtravelandshiftwork.
1Complementaryandalternativemedicinehasbeenusedincreasinglytomanagesleepdisorders.
Oneofthemostpopulartreatmentsofthistypeismelatonin,ahormonethatissecretedbythepinealglandandislinkedtothecircadianrhythm.
3Weconductedasystematicreviewoftheefficacyandsafetyofexogenousmelatonininmanagingsecondarysleepdisordersandsleepdisordersaccompanyingsleeprestriction,suchasjetlagandshiftworkdisorder.
Ourfindingscanhelptoguideclini-ciansandpatientsintreatmentdecisionsregardingtheuseofexogenousmelatonininthemanagementoftheseconditions.
MethodsSearchstrategyAhealthscienceslibrarianconductedacomprehensivesearchtoidentifyrelevantEnglish-languagestudies.
Wesearched13elec-tronicdatabases(table1;seebmj.
comforsearchterms).
Theref-erencelistsofrelevantreviews,aswellasarandomsampleofincludedstudies,werereviewedtoidentifyotherpotentiallyrel-evantstudies.
WehandsearchedabstractsofmeetingsoftheAssociatedProfessionalSleepSocietyfrom1999to2003.
Finally,wesearchedMedlineandEmbaseagaininearly2004toidentifyrecentlypublishedstudies.
StudyselectionThefulltextofallarticlesdeemedpotentiallyrelevantwasretrievedandreviewedindependentlybytworeviewers.
Toassesstheefficacyofexogenousmelatonin,weincludedrandomisedcontrolledtrialsthatinvolvedhumanparticipantswhohadasecondarysleepdisorderorasleepdisorderaccompanyingsleeprestriction;comparedmelatonintoplacebo;andreportedononeormoreof:sleeponsetlatency(amountoftimebetweenlyingdowntosleepandonsetofsleep),sleepefficiency(amountoftimespentasleepasapercentageoftotaltimespentinbed),sleepquality(perceivedqualityofsleep),wakefulnessaftersleeponset(amountoftimespentawakeinbedafterfirstattainmentSearchtermsusedareonbmj.
comCitethisarticleas:BMJ,doi:10.
1136/bmj.
38731.
532766.
F6(published10February2006)BMJOnlineFirstbmj.
compage1of9on16March2021byguest.
Protectedbycopyright.
http://www.
bmj.
com/BMJ:firstpublishedas10.
1136/bmj.
38731.
532766.
F6on10February2006.
Downloadedfromofsleep),totalsleeptime(totaltimespentasleepwhileinbed),orpercentageoftimeinrapideyemovement(REM)sleep.
Toassessthesafetyofexogenousmelatonin,weincludedrandomisedandnon-randomisedtrialsmeetingthefirsttwocri-teriaaboveandreportingonadverseevents.
Astudypopulationwasconsideredtohaveasecondarysleepdisorderifthepartici-pants,asagroup,weredefinedbyaspecificchronicmedicalorpsychiatricdisorderandthisdisorderwaslikelytobethecauseofthesleepdisorder.
Astudypopulationwasconsideredtohavebeenexposedtosleeprestrictionifparticipantshadbeenexposedtotransmeridianairtravel,shiftwork,orotherformsofsleepschedulealteration.
Disagreementsregardinginclusionofstudieswereresolvedthroughdiscussion.
QualityassessmentTworeviewersassessedstudyqualityindependently.
Fortheeffi-cacyreview,randomisedcontrolledtrialswereassessedformethodologicalqualitywiththevalidatedJadadscale.
4Inaddition,concealmentoftreatmentallocationwasassessedusingthecriteriaofSchulzetal.
5Allocationconcealmentwasconsid-eredtobeadequateifgroupallocationwasaccomplishedbyusingsuchmethodsascentralrandomisation,numberedorcodedcontainers,drugspreparedbyapharmacy,orseriallynumbered,opaque,sealedenvelopes.
Forthesafetyreview,whichreliedonevidencefromrandomisedandnon-randomisedtrials,theDownsandBlackchecklistwasused.
6Disagreementsregard-ingqualityassessmentwereresolvedthroughdiscussion.
DataextractionDatawereextractedbyusingastandardiseddataextractionformthatcaptureddetailsofstudydesign,population,intervention,andoutcomes.
Atrainedreviewerextracteddataandasecondreviewerverifiedtheextracteddata.
Disagreementswereresolvedthroughdiscussion.
DataanalysisWelistedouroutcomesinorderofimportance,withsleeponsetlatencyasmostimportant(primaryoutcome),followedbysleepefficiency,sleepquality,wakefulnessaftersleeponset,totalsleeptime,andpercentageoftimeinREMsleep.
Continuousoutcomeswerecombined,usingaweightedmeandifference,withtheexceptionofsleepquality,forwhichstudieswerecom-binedbyusingastandardisedmeandifference.
Dichotomousoutcomeswerecombinedbyusingariskdifference.
Theinversevariancemethodwasusedtoweightthestudies.
7Allmeta-analysesusedarandomeffectsmodel.
Apointestimatewithcorresponding95%confidenceintervalwascomputedforeachoutcome,usingthegenericinversevariancefunctioninRevMan4.
2.
5(UpdateSoftware,2004).
Inmostcases,wewereabletocalculatetheefficacyestimateforeachstudyexactly,butoccasionallyestimateshadtobemadebyextractingfromgraphsorusingmedians.
Standarderrorsofthedifferenceswerecalculatedfromavailabledata(individualpatientdataorexactPvalues)wheneverpossible.
Forstudieswithacrossoverdesign,weusedthemethodsofElbourneetaltocomputestandarderrorsofdifferences,8andacorrelationof0.
5wasimputedwhenitcouldnotbecalculatedfromavailabledata.
Allpooledestimateswereassessedforheterogeneity,usingtheI2statistic.
9Forourprimaryoutcome,weplannedtoexploreheterogeneityinsubgroupandsensitivityanalyses.
Wealsocon-ductedaposthocsensitivityanalysis.
Deeks'2statistic10wasusedtotestforsignificantheterogeneityreductioninpartitionedsub-groups(age,comorbidity,typeofsleepdisorder,dosage,treatmentduration,outcomemeasurementmethod,studydesign,studyquality,andallocationconcealment).
Wetestedforpublicationbiasvisuallyusingthefunnelplotandquantitativelyusingtherankcorrelationtest,11thegraphicaltest,12andthetrimandfillmethod.
13PublicationbiasgraphsandcalculationswereproducedwithSTATA7.
0(StataCorporation,2001).
ResultsFigure1showstheflowofstudiesthroughtheselectionprocess.
SecondarysleepdisordersEfficacyTable2describestheninetrials(279participants)includedintheefficacyanalysisforsecondarysleepdisorders.
14–25Themedianqualityscore,basedontheJadadscale,was4outof5(interquartilerange2-4).
Concealmentofallocationwasunclearinallstudiesexceptone,23whichhadadequateallocationconcealment.
Onlyfivestudiesdescribedafundingsource;forallofthesestudies,fundingwasreceivedfrompublicsponsors.
18–23SleeponsetlatencyTable3outlinesthemeansandstandarddeviationsforsleeponsetlatencyforplaceboandmelatoningroupsforthesixtrialsprovidingdataonthisoutcome.
15181921–23Thestudiesproducedacombinedestimatethatfavouredmelatoninbutwasnotsignificant(weightedmeandifference13.
2(95%confidenceinterval27.
3to0.
9)min)(fig2).
Heterogeneityamongthestudieswassubstantial(I2=79.
2%)dueprimarilytoonestudy23thathadaverysmallstandarddeviationandanestimatethatfavouredplacebo,whereastheotherstudieshadpointestimatesthatfavouredmelatonin.
Theresultsforplannedsubgroupandsensitivityanalysesareshownintable4.
Intheonlytwocategoriesforwhichtheconfi-denceintervalsacrosssubgroupsdidnotoverlap,astudybyShamiretalseemedtobehighlyinfluential.
23Subgroupsthatomittedthisstudy(actigraphyandquestionnairemethodsofmeasuringsleepoutcomesandunclearallocationconcealment)showedasignificantresultinfavourofmelatoninwithminimalheterogeneity,whilethepointestimateforthisstudyshowedasignificanteffectinfavourofplacebo.
WeconductedaposthocsensitivityanalysisexcludingthestudybyShamiretalfromtheprimaryanalysis.
Whenthestudywasincludedintheanalysis,thepointestimatewas13.
2Table1BiomedicaldatabasessearchedforarticlesonexogenousmelatonininmanagingsecondarysleepdisordersandsleepdisordersaccompanyingsleeprestrictionDatabasePlatformDatescoveredbysearchMedlineOvid1966to30June2003PreMedlineOvid1970to30Juneand4July2003EmbaseOvid1988to30June2003PubMedNA1950to9July2003CABHealthSilverPlatterversion4.
31973to8July2003CINAHLOvid1982to30June2003CochraneCentralRegisterofControlledTrialsOvid3rdquarter2003;2July2003ScienceCitationIndexISIWebofKnowledge1945to4July2003BiologicalAbstractsSilverPlatterversion4.
31969to4July2003InternationalPharmaceuticalAbstractsOvid1970to12August2003NLMGatewayhttp://gateway.
nlm.
nih.
gov/gw/Cmd1950to13August2003OCLCPapersFirstandProceedingsFirstOCLCFirstSearch1993to11July2003ToxlineCSAInternetDatabaseService1965to4July2003Researchpage2of9BMJOnlineFirstbmj.
comon16March2021byguest.
Protectedbycopyright.
http://www.
bmj.
com/BMJ:firstpublishedas10.
1136/bmj.
38731.
532766.
F6on10February2006.
Downloadedfrom(27.
3to0.
9)min;whenitwasexcluded,thepointestimatewas17.
4(26.
4to8.
4)min.
Althoughthepointestimatedidnotchangesubstantially,theconfidenceintervalnarrowed,render-ingtheresultsignificant.
Notenoughstudiesexaminedsleeponsetlatencyforpublicationbiastobetestedonthebasisofthisoutcome.
OtherefficacyoutcomesSixtrialsreportingdataforsleepefficiencyshowedasignificanteffectthatfavouredmelatonin(weightedmeandifference1.
9%(0.
5to3.
3);I2=0%)1820–24;however,theeffectseemsnottobeclinicallyimportant.
Theresultsforotherefficacyoutcomesareshownintable4.
Referencesidentifiedfromdatabasesearches(n=1884)ScreeningoftitlesandabstractsStudiestoassessforpotentialinclusioninthereview(n=935)Studiesincludedinthereview(n=25)Studiesofsleepdisordersaccompanyingsleeprestriction(n=13)Studiesofsecondarysleepdisorders(n=12)ApplicationofinclusioncriteriaSafetyreview(n=10)Efficacyreview(n=9)Safetyreview(n=7)Efficacyreview(n=9)Fig1RetrievalandselectionofstudiesofexogenousmelatonininmanagementofsecondarysleepdisordersandsleepdisordersaccompanyingsleeprestrictionTable2CharacteristicsoftrialsofmelatonininpeoplewithsecondarysleepdisordersStudyandyearNoenrolled(analysed)Mean(SDorrange)age(years)%maleDisorderInterventionDesignFormulationDosageandtiming(route)FrequencyanddurationCamfieldetal,199614*68.
8(3to13)67DevelopmentaldisabilityNS0.
5or1.
0mgat1800(oral)Foreach2weekintervalof10weektrial,childrenreceivedmelatoninorplacebonightlyduringfirstweekwithalternativeagentgivenonsecondweekN-of-1RCTDodgeandWilson,20011536(17)7.
4(1to15)NSDevelopmentaldisabilityNS5mgat2000(oral)5mg/dayforweeks2-3and5-6of6weekstudyRCT;crossoverJanetal,19941615NS(0.
5-14)87NeurologicalimpairmentNS2-5mgatbedtime(NS)2-5mg/dayforupto12monthsN-of-1RCTJean-Louisetal,1998171068.
8(15.
8)40MildcognitiveimpairmentNS6mg2hbeforebedtime(NS)6mg/dayfor10daysNon-RCT;crossoverMcArthurandBudden,199818910.
1(1.
5)0RettsyndromeImmediaterelease2.
5-7.
5mg(dependingonbodyweight)given1hbeforebedtime(oralorgastrostomytube)1capsule/dayfor4weeksRCT;crossoverO'Callaghanetal,1999197Median=11(2-28)43TuberoussclerosisNS5mg20minbeforebedtime(oral)1capsule/dayfor2weeksRCT;crossoverSerfatyetal,20022044(25)84.
2(7.
6)64DementiaSlowrelease6mgtabletatusualbedtime(oral)1tablet/dayfor2weeksRCT;crossoverSerfatyetal,200321*33(31)39.
9(11.
8)45MajordepressionSlowrelease6mgtabletatbedtime(oral)1tablet/dayfor4weeksRCT;parallelShamiretal,200022*27(19)42(5)63SchizophreniaControlledrelease2mg2hbeforebedtime(NS)2mg/dayfor3weeksRCT;crossoverShamiretal,200023*1442.
3(13.
1)79SchizophreniaControlledrelease2mg2hbeforebedtime(NS)2mg/dayfor3weeksRCT;crossoverSingeretal,200324*157(151)77.
4(8.
9)44Alzheimer'sdiseaseSlowreleaseandimmediaterelease2.
5mg(SR)or10mg(IR)1hbeforebedtime(oral)1capsule/dayfor8weeksRCT;parallelVanWieringenetal,2001258133.
4(10.
7)27ChronicwhiplashsyndromeNS5mg5hbeforeindividualdimlightmelatoninonsettime(oral)1tablet/dayfor4weeksRCT;parallelNS=notspecified;RCT=randomisedcontrolledtrial.
*Includedinefficacyreviewonly;includedinsafetyreviewonly.
ResearchBMJOnlineFirstbmj.
compage3of9on16March2021byguest.
Protectedbycopyright.
http://www.
bmj.
com/BMJ:firstpublishedas10.
1136/bmj.
38731.
532766.
F6on10February2006.
DownloadedfromSafetySevenstudieswereincludedinthesafetyanalysis15–2025;onewasnon-randomisedandsixwererandomised(table2).
Thestudiesincluded164participants.
Thequalityofthesestudieswasgood(medianqualityindex21(outof29);range20-22).
Themostcommonlyreportedadverseeventswereheadaches,dizziness,nausea,anddrowsiness.
Theoccurrenceoftheseoutcomeswassimilarformelatoninandplacebo(table4).
SleeprestrictionEfficacyTable5describestheninetrialsincludedintheefficacyanalysisforsleeprestriction.
2629–3235–38Thetrialsencompassed427participants.
Themedianqualityscorewas4outof5(interquar-tilerange3-4).
Concealmentofallocationwasunclearinallstud-iesexceptthree,293238whichhadadequateallocationconceal-ment.
Noneofthestudiesdescribedafundingsource.
SleeponsetlatencyTable6outlinesthemeanandstandarddeviationsforsleeponsetlatencyforplaceboandmelatoningroupsfortheninetri-alsthatprovideddataonthisoutcome.
26293235–37Thestudiespro-ducedacombinedestimatethatfavouredmelatoninbutwasnotsignificant(weightedmeandifference1.
0(2.
3to0.
3)min;I2=4.
0%)(fig3).
Theresultsforplannedsubgroupandsensitivityanalysesareintable7.
Thesubgroupsdidnotdiffersignificantlyinanyofthecategories(allconfidenceintervalswereoverlapping,andinallbuttwocases(422524WMD13.
1(24.
8to1.
5)Measurementmethod*:Polysomnography11414WMD5.
8(2.
5to9.
1)Actigraphy34443WMD14.
5(25.
0to4.
1)Questionnaire22424WMD25.
7(43.
3to8.
0)Studydesign:Parallel11615WMD13.
5(32.
5to5.
5)Crossover56666WMD13.
5(29.
7to2.
8)Allocationconcealment*:Unclear56867WMD17.
4(26.
4to8.
4)Adequate11414WMD5.
8(2.
5to9.
1)WMD=weightedmeandifference;RD=riskdifference*P<0.
001,Deeks2test.
ResearchBMJOnlineFirstbmj.
compage5of9on16March2021byguest.
Protectedbycopyright.
http://www.
bmj.
com/BMJ:firstpublishedas10.
1136/bmj.
38731.
532766.
F6on10February2006.
Downloadedfromfast,arangeofdoseswasused,andthedurationofadministrationvariedfromdaystoweeks.
Indeed,ourresultsshowthatdosageanddurationofmelatoninadministrationexplainaconsiderableamountofheterogeneityacrossstudies.
Twoothersystematicreviewsexaminingtheuseofmelatoninforjetlagconcludedthatmelatoniniseffectiveinalleviatingthesymptomsofjetlag.
4142Thesereviewsexaminedtheeffectofmelatoninonboththedaytimefatigueandthesleepdisturbanceaspectsofjetlag.
Ourreviewshowsthatmelatonindoesnotaffecteithersleeponsetlatencyorsleepefficiencyinpeoplewithjetlagorpeoplewithshiftworkdisorder.
Ourresultsdonotpro-videevidencethatmelatoniniseffectiveinalleviatingsleepdisturbanceinjetlag,butwedidnotdeterminetheeffectofmelatoninonmeasuresofdaytimefatigue.
OtherlimitationsTheobservationsofthisreviewarebasedmostlyonstudieswithrelativelyshortdurations,sotheefficacyandsafetyofmelatoninreportedheremayreflectonlyitsshorttermeffects.
Secondly,severalstudiesdidnotreportadequatelyondetailsoftheinter-vention,suchascontent,quality,andformulationofthemelatoninproductunderstudy,noronmethodsofallocationconcealmentorsourceoffunding,whichcastsdoubtonthemethodologicalqualityofthesestudies,despiteagoodmedianJadadscoreorDownsandBlackqualityindex.
Thirdly,non-Englishlanguagereportswereexcludedfromthereview;however,wedidnotfindstrongevidenceofpublicationbias,soitisunlikelythattheinclusionofthesereportswouldhavealteredourfindingssubstantially.
WethanktheNationalCentreforComplementaryandAlternativeMedicine,NationalInstitutesofHealthforsponsoringthisresearch,throughtheAgencyforHealthcareResearchandQuality.
Wearegratefultomembersofourtechnicalexpertpanelforprovidinginputonthedirec-tionandscopeofthereview.
WeareespeciallygratefultoManishaWitmansforherinputonthemanuscript.
Contributors:NBplanned,oversaw,andparticipatedinallstepsofthesys-tematicreviewprocessandinwritingandeditingthemanuscript.
BVper-formedallstatisticalanalysesandparticipatedinwritingandeditingtheTable5CharacteristicsoftrialsofpeoplewithsleepdisordersaccompanyingsleeprestrictionStudyandyearNoenrolled(analysed)Mean(SDorrange)age(years)%maleDisorderInterventionDesignFormulationDosageandtiming(route)FrequencyanddurationBeaumontetal,200426*27(18)35.
3(8.
1)67JetlagNS5mgonday-1at1700;onday0at1600;onday1today3at2300(NS)5mg/dayfor5daysRCT;parallelClaustratetal,19922737(15)Melatonin36.
3(8.
9);placebo:35.
7(6.
4)Melatonin53;placebo67JetlagNS8mgat2200(oral)1capsule/dayfor4daysRCT;parallelEdwardsetal,20002831Melatonin40(13);placebo41(12)Melatonin93;placebo88JetlagNS5mgtakenonplanebetween1800to1900andbetween2200to2300,accordingtolocaltimeatdestinationandfornext3evenings2capsules/dayforfirstdayandthen1capsule/dayfor3daysnon-RCT;parallelFolkardetal,199329*17(7)29(7)88ShiftworkdisorderNS5mgat0642±7.
6min(oral)1capsule/dayfor6successivedaysleepstakenbetweennightshiftsRCT;crossoverJamesetal,19983024(22)29(8)77ShiftworkdisorderNS6mg0.
5hbeforeeachconsecutivedaysleep(oral)6mg/dayfor4treatmentcycleslasting4to6consecutivenightshiftsRCT;crossoverJockovichetal,2000311928.
2(NS)21ShiftworkdisorderNS1mg0.
5to1hbeforedaytimesleep(oral)1caplet/dayfor3consecutivedaysRCT;crossoverJorgensenandWitting,19983220(18)32(25to40)89ShiftworkdisorderNS10mgmorningaftereachnightshift(oral)1tablet/dayforvariedamountoftimeRCT;crossoverPetrieetal,19893320(15)NS(28to68)60JetlagNS5mgtakenbetween1000and1200localtime;alsotakenatthesametimeduringtheflightandbetween2200and2400(destinationtime)afterarrival(NS)1dosefor3daysbeforeflight,1doseduringflight,and1dose/dayfor3daysafterarrivalRCT;crossoverPetrieetal,19933452(44)34.
9(7.
7)50JetlagNS5mgtakenbetween0700to08005mgearlymelatoninfor8days,5mglatemelatoninfor5daysRCT;parallelSuhneretal,199835320(234)20to6554JetlagFastreleaseandcontrolledrelease0.
5mgfastrelease,5mgfastrelease,or2mgcontrolledreleasemelatoninonfirstdayafterflightat2310andonsubsequentdaysat2329(NS)1dose/dayfor4daysaftereastwardflightRCT;parallelSuhneretal,200136160(74)41.
3(18to68)51JetlagNS5mgtakenonreturnflight(eastbound)between1700and2100localtimeattheplaceofdeparturedependingonflightschedule(NS)1dose/dayonreturnflightandfor4consecutivedaysafterflightRCT;parallelWaldhauseretal,199037*20(20)26.
4(4.
8)50InducedinsomniaNS80mgat2100(oral)SingledoseRCT;parallelWrightetal,19983820(15)38.
6(32to45)80ShiftworkdisorderNS5mg30minbeforebedtimeintheevening(oral)5mg/nightfor3nightsfollowingshiftworkRCT;crossoverNS=notspecified.
*Includedinefficacyreviewonly;includedinsafetyreviewonly.
Researchpage6of9BMJOnlineFirstbmj.
comon16March2021byguest.
Protectedbycopyright.
http://www.
bmj.
com/BMJ:firstpublishedas10.
1136/bmj.
38731.
532766.
F6on10February2006.
Downloadedfrommanuscript.
NHparticipatedinmoststepsofthesystematicreviewprocessandinwritingandeditingthemanuscript.
RPparticipatedinallstepsofthesystematicreviewprocessandreviewedthemanuscript.
LTconductedtheliteraturesearch,providedtechnologicalexpertisefortheinclusionprocess,andparticipatedineditingthemanuscript.
LHparticipatedinwritingtheproposal,providedmethodologicalexpertise,andparticipatedinwritingandeditingthemanuscript.
SVparticipatedinwritingtheproposal,providedmethodologicalandcontentexpertise,andparticipatedineditingthemanuscript.
TKparticipatedinwritingtheproposal,providedmethodologicalexpertise,andprovidedfeedbackonthemanuscript.
GBparticipatedinwritingtheproposal,providedcontentexpertise,andparticipatedinwritingandeditingthemanuscript.
MichelleTubman,MiaLang,MariaOspina,VictorJuorio,andEllenCrumleywereinvolvedinstudyselection,qualityassessment,anddataextractionorentry.
TKisguar-antor.
Funding:ThisstudywasconductedundercontracttotheAgencyforHealthcareResearchandQuality(contractNo290-02-0023),Rockville,MD,andsupportfromtheNationalCenterforComplementaryandAlter-nativeMedicine,NationalInstitutesofHealth,Bethesda,MD.
SVissupportedbyAgencyforHealthcareResearchandQuality(AHRQ),USA;CanadianInstitutesofHealthResearch;ChangeFoundation;DepartmentofPediatrics;NationalHealthProductsDirectorate,HealthCanada;OntarioMentalHealthFoundation;StolleryChildren'sHospitalandFoun-dation;TheHospitalforSickChildrenFoundation;andtheUniversityofAlberta.
GBissupportedbyAHRQ;CanadianInstitutesofHealthResearch;CanadaResearchChairsProgram;StanleyFoundation;UniversityofAlbertaHospitalFoundation;BebenseeSchizophreniaResearchFund;DaveyEndowment;andZyprexaResearchFoundation.
Theauthorsofthisarticleareresponsibleforitscontents,includinganyclinicalortreatmentrecommendations.
NostatementinthisarticleshouldbeconstruedasanofficialpositionoftheAgencyforHealthcareResearchandQuality,theNationalCenterforComplementaryandAlternativeMedicineortheUSDepartmentofHealthandHumanServices.
Competinginterests:Nonedeclared.
Ethicalapproval:Notrequired.
1NationalInstitutesofHealth.
NationalCenteronSleepDisordersResearchPlan.
2003.
www.
nhlbi.
nih.
gov/health/prof/sleep/res_plan/sleep-rplan.
pdf(accessed14Jan2006).
2RollerL.
Treatingsleepdisorders.
AustralJPharm2002;83:443-7.
3ReiterRJ.
Melatonin:clinicalrelevance.
BestPractResClinEndocrinolMetab2003;17:273-85.
4JadadAR,MooreRA,CarrollD,JenkinsonC,ReynoldsDJ,GavaghanDJ,etal.
.
Assess-ingthequalityofreportsofrandomizedclinicaltrials:isblindingnecessaryControlledClinTrials1996;17:1-12.
5SchulzKF,ChalmersI,HayesRJ,AltmanD.
Empiricalevidenceofbias.
Dimensionsofmethodologicalqualityassociatedwithestimatesoftreatmenteffectsincontrolledtri-als.
JAMA1995;273:408-12.
6DownsSH,BlackN.
Thefeasibilityofcreatingachecklistfortheassessmentofthemethodologicalqualitybothofrandomizedandnon-randomizedstudiesofhealthcareinterventions.
JEpidemiolCommunHealth1998;52:377-84.
7SuttonAJ,AbramsKR,JonesDR,SheldonTA,SongF.
Methodsformeta-analysisinmedi-calresearch.
Chichester:Wiley,2000:58.
8ElbourneDR,AltmanDG,HigginsJPT,CurtinF,WorthingtonHV,VailA.
Meta-analysesinvolvingcross-overtrials.
IntJtEpidemiolt2002;31:140-9.
9HigginsJPT,ThompsonSG.
Quantifyingheterogeneityinameta-analysis.
StatMed2002;21:1539-58.
10DeeksJJ,AltmanDG,BradburnMJ.
Statisticalmethodsforexaminingheterogeneityandcombiningresultsfromseveralstudiesinmeta-analysis.
In:EggarM,SmithGD,AltmanDG,eds.
Systematicreviewsinhealthcare:meta-analysisincontext.
3rded.
London:BMJPublishingGroup,2001:285-312.
11BeggCB,MazumdarM.
Operatingcharacteristicsofarankcorrelationtestforpublica-tionbias.
Biometrics1994;50:1088-101.
Table6Sleeponsetlatency(minutes)inpeoplewithsleepdisordersaccompanyingsleeprestrictionStudyDesignMethodofcomputingSEofdifferencebetweenmelatoninandplaceboMelatoninPlaceboNoMean(SD)NoMean(SD)Mean(95%CI)differenceBeaumontetal,200426ParallelFromSDs929.
7(14.
0)932.
2(14.
3)2.
6(15.
7to10.
5)Folkardetal,199329CrossoverSDsusingestimatedcorrelationof0.
574.
5(1.
6)75.
6(1.
8)1.
1(2.
3to0.
2)Jamesetal,199830CrossoverFromexactPvalueofdifference2215(7.
0)2216(9.
0)1.
0(5.
7to3.
7)Jockovichetal,200031CrossoverFromexactPvalueofdifference197.
6(notprovided)196.
8(notprovided)0.
8(2.
7to4.
3)JorgensenandWitting,199832CrossoverFromconfidenceintervalofdifference1813.
6(notprovided)1815.
6(notprovided)2.
0(7.
5to3.
5)Suhneretal,199835ParallelEstimatedfromupperboundofPvalue17419.
4(notprovided)6032.
1(notprovided)12.
6(25.
2to0.
1)Suhneretal,200136ParallelFromSDs3521.
7(23.
7)3921.
2(27.
3)0.
5(11.
1to12.
1)Waldhauseretal,199037ParallelEstimatedfromupperboundofPvalue1014.
7(notprovided)1023.
7(notprovided)9.
0(19.
2to1.
2)Wrightetal,199838CrossoverSDsusingestimatedcorrelationof0.
5.
1522.
3(14.
2)1519.
0(13.
7)3.
3(3.
8to10.
4)Waldhauser199037Folkard199329James199830Jorgensen199832Suhner199835Wright199838Jockovich200031Suhner200136Beaumont200426Total(95%CI)Testforheterogeneity:χ2=8.
33,df=8,P=0.
4,I2=4.
0%Testforoveralleffect:z=1.
47,P=0.
14Study10722181741519359309Melatonin(N)1072218601519399199Placebo(N)1.
6066.
627.
275.
401.
063.
3112.
531.
250.
97100.
00Weight(%)-9.
0(5.
2)-1.
05(0.
64)-1.
0(2.
4)-2.
0(2.
8)-12.
62(6.
4)3.
3(3.
6)0.
79(1.
8)0.
5(5.
9)-2.
6(6.
7)Meandifference(SE)-9.
00(-19.
19to1.
19)-1.
05(-2.
30to0.
20)-1.
00(-5.
70to3.
70)-2.
00(-7.
49to3.
49)-12.
62(-25.
16to-0.
08)3.
30(-3.
76to10.
36)0.
79(-2.
74to4.
32)0.
50(-11.
06to12.
06)-2.
60(-15.
73to10.
53)-0.
97(-2.
26to0.
33)Meandifference(random)(95%CI)-10-50510FavoursmelatoninFavoursplaceboFig3SleeponsetlatencyinpeoplewithsleepdisordersaccompanyingsleeprestrictionResearchBMJOnlineFirstbmj.
compage7of9on16March2021byguest.
Protectedbycopyright.
http://www.
bmj.
com/BMJ:firstpublishedas10.
1136/bmj.
38731.
532766.
F6on10February2006.
Downloadedfrom12EggerM,SmithGD,SchneiderM,MinderC.
Biasinmeta-analysisdetectedbyasinglegraphicaltest.
BMJ1997;315:629-34.
13DuvalS,TweedieR.
Anonparametric"trimandfill"methodofaccountingforpublica-tionbiasinmeta-analysis.
JAMA2000;95:89-98.
14CamfieldP,GordonK,DooleyJ,CamfieldC.
Melatoninappearsineffectiveinchildrenwithintellectualdeficitsandfragmentedsleep:six"Nof1"trials.
JChildNeurol1996;11:341-3.
15DodgeNN,WilsonGA.
Melatoninfortreatmentofsleepdisordersinchildrenwithdevelopmentaldisabilities.
JChildNeurol2001;16:581-4.
16JanJ,EspezelH,AppletonR.
Thetreatmentofsleepdisorderswithmelatonin.
DevelopMedChildNeurol1994;36:97-107.
17Jean-LouisG,vonGizyckiH,ZiziF.
Melatonineffectsonsleep,mood,andcognitioninelderlywithmildcognitiveimpairment.
JPinealRes1998;25:177-83.
18McArthurAJ,BuddenSS.
SleepdysfunctioninRettsyndrome:atrialofexogenousmelatonintreatment.
DevelopmentalMedChildNeurol1998;40:186-92.
19O'CallaghanFJK,ClarkeAA,HancockE,HuntA,OsborneJP.
Useofmelatonintotreatsleepdisordersintuberoussclerosis.
DevelopmentalMedChildNeurol1999;41:123-6.
20SerfatyM,Kennell-WebbS,WarnerJ,BlizardR,Raven,P.
Doubleblindrandomisedplacebo-controlledtrialoflowdosemelatoninforsleepdisordersindementia.
IntJGeriatrPsychiatry2002;17:1120-7.
21SerfatyMA,OsborneD,BuszewiczMJ,RavenPW.
Theeffectsofexogenousmelatonininmajordepression.
BiolMedRhythmRes2003:1191-3.
22ShamirE,LaudonM,BarakY,AnisY,RotenbergV,ElizurA,etal.
Melatoninimprovessleepqualityofpatientswithchronicschizophrenia.
JClinPsychiatry2000;61:373-7.
23ShamirE,RotenbergVS,LaudonM,ZisapelN,ElizurA.
First-nighteffectofmelatonintreatmentinpatientswithchronicschizophrenia.
JClinPsychopharmacol2000;20:691-4.
24SingerC,TractenbergRE,KayeJ,SchaferK,GamstA,GrundmanM,etal.
Amulticenter,placebo-controlledtrialofmelatoninforsleepdisturbanceinAlzheimer'sdisease.
Sleep2003;26:893-901.
25VanWieringenS,JansenT,SmitsMG,NagtegaalJE,CoenenAML.
Melatoninforchronicwhiplashsyndromewithdelayedmelatoninonset.
ClinDrugInvest2001;21:813-20.
26BeaumontM,BatejatD,PierardC,VanBeersP,DenisJB,CosteO,etal.
Caffeineormelatonineffectsonsleepandsleepinessafterrapideastwardtransmeridiantravel.
JApplPhysiol2004;96:50-8.
27ClaustratB,BrunJ,DavidM,SassolasG,ChazotG.
Melatoninandjetlag:confirmatoryresultusingasimplifiedprotocol.
BiolPsychiatry1992;32:705-11.
28EdwardsBJ,AtkinsonG,WaterhouseJ,ReillyT,GodfreyR,BudgettR.
Useofmelatonininrecoveryfromjet-lagfollowinganeastwardflightacross10time-zones.
Ergonomics2000;43:1501-3.
29FolkardS,ArendtJ,ClarkM.
Canmelatoninimproveshiftworkers'toleranceofthenightshiftSomepreliminaryfindings.
Chronobiol1993;10:315-20.
30JamesM,TremeaMO,JonesJS,KrohmerJR.
CanmelatoninimproveadaptationtonightshiftAmJEmergMed1998;16:1-7.
31JockovichM,CosentinoD,CosentinoL,WearsRL,SeabergDC.
Effectofexogenousmelatoninonmoodandsleepefficiencyinemergencymedicineresidentsworkingnightshifts.
AcademicEmergencyMed2000;7:955-8.
32JorgensenKM,WittingMD.
DoesexogenousmelatoninimprovedaysleepornightalertnessinemergencyphysiciansworkingnightshiftsAnnEmergencyMed1998;31:699-704.
33PetrieK,ConaglenJV,ThompsonL,ChamberlainK.
Effectofmelatoninonjetlagafterlonghaulflights.
BMJ1989;298:705-7.
Table7EfficacyandsafetyoutcomesandsubgroupandsensitivityanalysesfortrialsofpeoplewithsleepdisordersaccompanyingsleeprestrictionOutcomeNoofstudiesNoinmelatoningroupNoinplacebogroupSummarymeasurePointestimate(95%CI)EfficacySleeponsetlatency(min)9309199WMD1.
0(2.
3to0.
3)Sleepefficiency(%)57878WMD0.
5(0.
6to1.
6)Sleepquality(SD)5248138SMD0.
2(0.
2to0.
6)Wakefulnessaftersleeponset(min)24448WMD10.
4(21.
0to0.
2)Totalsleeptime(min)7100100WMD18.
2(8.
1to28.
3)%REMsleep11010WMD3.
6(7.
3to0.
1)AdverseeventsHeadaches9342218RD0.
01(0.
05to0.
02)Dizziness9342218RD0.
00(0.
03to0.
03)Nausea10356235RD0.
00(0.
03to0.
02)Drowsiness10356235RD0.
00(0.
03to0.
03)SubgroupandsensitivityanalysesofsleeponsetlatencyDosage*:<1mg15860WMD11.
8(23.
6to0.
0)1-3mg27779WMD4.
5(17.
3to8.
3)4-5mg5124130WMD1.
0(4.
0to2.
1)10-20mg11818WMD2.
0(7.
5to3.
5)Typeofsleepdisorder:Jetlag3218108WMD4.
7(12.
6to3.
1)Shiftwork58181WMD0.
8(1.
9to0.
3)Deprivation11010WMD9.
0(19.
2to1.
2)Studydesign:Parallel4228118WMD6.
1(11.
9to0.
2)Crossover58181WMD0.
8(1.
9to0.
3)Quality:High(Jadadscore4-5)5264154WMD1.
2(4.
6to2.
3)Moderate(Jadadscore2-3)44545WMD0.
9(2.
7to0.
8)Allocationconcealment:Unclear6254144WMD1.
4(3.
8to1.
1)Adequate35555WMD0.
5(3.
7to2.
7)WMD=weightedmeandifference;SMD=standardisedmeandifference;RD=riskdifference.
*Nostudieswith6-9mgdoserange.
WhatisalreadyknownonthistopicSleepdisordersareawidespreadproblemandplaceaburdenonsocietythroughtheirnegativeimpactonqualityoflife,safety,productivity,andhealthcareutilisationComplementaryandalternativetherapies,suchasmelatonin,havebeenusedincreasinglytomanagesleepdisordersWhatthisstudyaddsThereisnoevidencethatmelatoniniseffectiveintreatingsecondarysleepdisordersorsleepdisordersaccompanyingsleeprestriction,suchasjetlagorshiftworkdisorderThereisevidencethatmelatoninissafewithshorttermuse,butadditionalstudiesareneededtodetermineitslongtermsafetyResearchpage8of9BMJOnlineFirstbmj.
comon16March2021byguest.
Protectedbycopyright.
http://www.
bmj.
com/BMJ:firstpublishedas10.
1136/bmj.
38731.
532766.
F6on10February2006.
Downloadedfrom34PetrieK,DawsonAG,ThompsonL,BrookR.
Adouble-blindtrialofmelatoninasatreatmentforjetlagininternationalcabincrew.
BiolPsychiatry1993;33:526-30.
35SuhnerA,SchlagenhaufP,JohnsonR,TschoppA,SteffenR.
Comparativestudytodeterminetheoptimalmelatonindosageformforthealleviationofjetlag.
Chronobiol1998;15:655-6.
36SuhnerA,SchlagenhaufP,HoferI,JohnsonR,TschoppA,SteffenR.
Effectivenessandtolerabilityofmelatoninandzolpidemforthealleviationofjetlag.
AviationSpaceEnvi-ronmentalMed2001;72:638-46.
37WaldhauserF,SaletuB,Trinchard-LuganI.
Sleeplaboratoryinvestigationsonhypnoticpropertiesofmelatonin.
Psychopharmacol1990;100:222-6.
38WrightSW,LawrenceLM,WrennKD,HaynesML,WelchLW,SchlackHM.
Randomizedclinicaltrialofmelatoninafternight-shiftwork:efficacyandneuropsychologiceffects.
AnnEmergencyMed1998;32:334-40.
39VallieresA,MorinCM.
Actigraphyintheassessmentofinsomnia.
Sleep2003;26:902-6.
40ChalmersTC,CelanoP,SacksHS,SmithHJr.
Biasintreatmentassignmentincontrol-ledclinicaltrials.
NEnglJMed1983;309:1358-61.
41HerxheimerA,PetrieKJ.
Melatoninforpreventingandtreatingjetlag.
CochraneData-baseSystRev2002;2:CD001520.
42ChaseJE,GidalBE.
Melatonin:therapeuticuseinsleepdisorders.
AnnPharmacother1997;31:1218-26.
(Accepted18November2005)doi10.
1136/bmj.
38731.
532766.
F6UniversityofAlberta/CapitalHealthEvidence-basedPracticeCentre,DepartmentofPediatrics,UniversityofAlberta,Edmonton,AB,CanadaT6G2J3NinaBuscemiresearchassociateBenVandermeerstatisticianNicolaHootonprojectcoordinatorRenaPandyaprojectmanagerLisaTjosvoldresearchlibrarianLisaHartlingadministrativedirectorTerryPKlassendirectorComplementaryandAlternativeResearchandEducationProgram,DepartmentofPediatrics,UniversityofAlbertaSunitaVohradirectorDepartmentofPsychiatry,UniversityofAlbertaGlenBakerprofessorandchairCorrespondenceto:NBusceminina.
buscemi@ualberta.
caResearchBMJOnlineFirstbmj.
compage9of9on16March2021byguest.
Protectedbycopyright.
http://www.
bmj.
com/BMJ:firstpublishedas10.
1136/bmj.
38731.
532766.
F6on10February2006.
Downloadedfrom