Technicalappendix

PrioritizingBrainHealth

April2025

Contents

2

Primaryandassociatedburdenofmentalhealthconditions

3

Mentalhealthimprovementthroughscalinginterventions

9

Impactofhealthimprovementsontheeconomy

11

Costanalysisandeconomicreturncalculationmethodology

15

Bibliography

Technicalappendix

Thisappendixoutlinesthemethodologyandkey

assumptionsunderlyingthePrioritizingBrain

Healthmodel,whichestimatestheprimaryand

associateddiseaseburdenofmentalhealth

conditionsandmodelsthepotentialimpactof

scalingprovenmentalhealthinterventionsintermsofpopulation-levelhealthimprovementandthe

globaleconomyboost.Inthisstudy,mentalhealthconditionsaredefinedasincludingbothmentalandsubstanceusedisorders.Theseconditionsoften

co-occurwithothernoncommunicablediseases(NCDs),exacerbatingtheoveralldiseaseburden.

ThisstudyisfocusedonshowingtherelationshipbetweenmentalhealthconditionsandNCDsandtheadvantagesofreducingdiseaseburdenby

expandingaccesstoestablishedinterventions.

Thisanalysisrepresentsan“artofthepossible”

approach,aimingtoestimatethepotentialbenefitsofexpandingaccesstoprovenbrainhealth

interventionsonaglobalscale.Whileitprovides

ahigh-levelperspectiveontheopportunitiesandpotentialimpact,itisimportanttoacknowledge

thatthereareinherentlimitationsinthedataand

assumptionsapplied.Furtherresearchinthisareawouldbevaluabletorefinetheestimates.

Primaryandassociatedburdenofmentalhealthconditions

Primarydiseaseburdenofmentalhealthconditions

DatafromtheInstituteforHealthMetricsand

Evaluation(IHME)GlobalBurdenofDisease(GBD)2021datasuitewasusedtoestimatetheprimary

diseaseburdenformentalhealthconditions,whichincludesbothmentalandsubstanceusedisorders.SpecificconditionsintheIHMEhierarchythatwereusedaspartofthisdefinitionareoutlinedasfollows:

—Mentaldisorders:

?anxietydisorders

?attention-deficit/hyperactivitydisorder

?autismspectrumdisorders

?bipolardisorder

?conductdisorder

?depressivedisorders

?eatingdisorders

?idiopathicdevelopmentalintellectualdisability

?othermentaldisorders

?schizophrenia

—Substanceusedisorders:

?alcoholusedisorders

?amphetamineusedisorders

?cannabisusedisorders

?cocaineusedisorders

?opioidusedisorders

?otherdrugusedisorders

Thetotalprimaryburdenofmentalhealthconditionswascalculatedbasedonthesumofthedisease

burdenforeachconditionoutlinedabove.The

modelquantifieddiseaseburdenintermsof

disability-adjustedlifeyears(DALYs)usingthe

IHMEdataset,whichincorporatesadjustmentsforcomorbiditiesandexcludesoverlappingimpacts

initscalculations.Thisensuresthattheburdenisaccuratelymeasuredwithoutdoublecountingacrossconditions.

Wedidnotincludeself-harmandneurological

disordersinthearticlefiguresandexhibits,but

weestimatedtheirdiseaseburdenandreduction

potentialseparatelyusingthesamemethodologyasmentalandsubstanceusedisorders.

AssociatedburdenofmentalhealthconditionsThemodelconsideredtwotypesofassociated

mentalhealthburden:

1.burdenfromotherNCDswheresubstanceuseisariskfactor

2.additionalriskofdevelopingotherNCDsifapersonhasapriormentalhealthdiagnosis

Associatedburdenfromsubstanceuserisks

Toestimatetheburdenassociatedwithsubstance

useriskfactors,themodelleveragedtheIHMEGBDdataset,whichquantifiesthediseaseburdenacrossalldiseasesattributabletoanymodifiableriskfactor.Fromthisdataset,themodelextractedthenon–mentalhealthNCDburdenlinkedtoalcoholordruguse.

PrioritizingBrainHealth2

PrioritizingBrainHealth3

Themodelincludedonlytheburdenfromother

NCDs,excludingtheburdenfrommentalhealth

conditions,asthesearealreadyconsideredpartoftheprimaryburden.Itisimportanttonotethattheremaybeasubstantialtimelagbetweensubstance

useandtheonsetofrelatedhealthissues,andsubstanceusedoesnotnecessarilyindicateasubstanceusedisorder.

AssociatedburdenfrompreexistingmentalhealthconditionsexacerbatingotherNCDsPeoplelivingwithmentalhealthconditions

experienceahigherprevalenceofotherNCDscomparedtopeoplewithoutamentalhealth

conditiondiagnosis.Additionalburdenwasestimatedfollowingathree-stepprocess:

1.Identifyanestimateoftheadditionalrelativeriskforpeoplewithapriormentalhealthcondition

diagnosis.

2.Calculatethepopulationattributablefraction(PAF)forthatconditionpair.

3.ApplythePAFtothediseaseburdenfor

therelevantnon–mentalhealthNCDs(pertheIHMEGBDprojectionsinthereferenceforecastscenario).1

Anevidencereviewwasconductedtoidentify

estimatesofadditionalriskforallmentalhealthandnon–mentalhealthNCDconditionpairs,looking

forstudiesthatmeasuretheadditionalriskof

developingnon–mentalhealthNCDsfollowingapriormentalhealthconditiondiagnosiscomparedtothepopulationwithoutapriormentalhealth

conditiondiagnosis.2

Tomaximizeconsistency,themodelreliedon

estimatesfromarecent,large-cohortlongitudinal

studywhereverpossible.Thisstudyisbasedon

apopulation-basedcohortof5.9millionpeople

borninDenmarkbetween1900and2015and

followedduringtheperiod2000to2016(83.9

millionperson-years).3Conditionpairhazardratiosadjustedforage,sex,calendartime,andprevious

mentalhealthorsubstanceusedisorders(identified

asModelBestimatesinthestudy)wereextracted

foruseinthisanalysis.Thissourcewasusedfor76percentoftheestimatesinthemodel(267individualdatapoints).WhereconditionpairswithinthescopeofthemodelwerenotcapturedintheDanishstudy,alternativeestimatesfrompeer-reviewed,publishedstudiesfromEurope,theAmericas,andAsiawere

identified.Wheremultipleestimateswereavailable,thelargestandmostrecentstudywasselectedforinclusioninthemodel.Intotal,18alternativestudieswereusedtoidentifytheremaining24percentof

estimatesincludedinthemodeling(86datapoints).Themodelassumednoadditionalriskwhereno

estimatesinthepublishedliteraturecouldbefound.

Conditioncategoriesusedintheselectedstudies

weremappedtotheconditionhierarchyusedintheIHMEGBDdataset,andthePAFwascalculated

foreveryconditionpairbyusingtheestimateof

additionalriskfromtheliteratureandcountry-,sex-,andageband–specificprevalenceestimatesfromthesamedataset.Thesevalueswereaddedup

toestimatetheassociateddiseaseburdenfrommentalhealthconditionsonotherNCDs.

ThisapproachisoutlinedinExhibit1.

Mentalhealthimprovement

throughscalinginterventions

Themodelestimatesthepotentialtoreducethe

burdenofmentalhealthconditionsbyimproving

accesstoproven,effectiveinterventions.Clinical

practiceguidelineswereappraisedtoidentify

themostappropriateinterventionstoscaleand

reviewedwithclinicalexperts.Foreachintervention,recentsystematicreviewsandmeta-analyseswereidentified.Ifthesewerenotavailable,high-quality

individualstudieswereusedtoextractthebest

availableestimatesofeffectivenessfordisease

burdenreduction,lookingseparatelyatimpacton

morbidityandmortality.Themodelconsidered100+condition-interventionpairsusingevidencefrom

acomprehensivereviewofabout100individualpapers,someofwhichcoveredmorethanoneinterventionorhealthcondition.

1SteinEmilVollset,“Burdenofdiseasescenariosfor204countriesandterritories,2022–2050:AforecastinganalysisfortheGlobalBurdenofDiseaseStudy2021,”TheLancet,May2024,Volume403,Number10440.

2AlthoughtheunderlyingbiologyandcausalpathwaysbetweenmentalhealthconditionsandotherNCDsarenotwellunderstoodinmanycases,temporalassociationshavebeenidentifiedinmultiplewell-designedstudies.

3“WorldBankcountryandlendinggroups,”WorldBank,accessedApril1,2025.

PrioritizingBrainHealth4

Exhibit1

Methodologytodetermineburdeninnon-mentalhealthconditionswherecomorbidmentalhealthconditionsdrivediseaseburden

Prevalenceofmentalhealthcondition1

Bycountry,sex,andagegroupfromIHME

Populationattributablefraction

p(RR–1)

1+p(RR–1)

1

PAF2=

Quanti?cationofefectofriskfactorbycomparingburdenassociatedto

Relativeriskofdevelopinganon-mentalhealthconditioninthosewithamentalhealthcondition

outcomewithamount

CalculatePAF

expectedinhypotheticalsituationof‘ideal’(eg,no)riskfactorexposure

Pooledacrosspopulationfromliteraturesearch

2

Translatetoburden

PAF×

DALYsfromIHMEfor

mentalhealthcondition

×

NumberofDALYsfor

non-mentalhealth

conditionattributabletomentalhealthcondition

3

DALYsfornon-mentalhealthcondition

Identify

addressable

burden

attributabletomentalhealthcondition

×

=

Percentburdenaddressable

viamentalhealthcondition

interventions

Addressableburdeninnon-mentalhealthconditionattributabletomentalhealth

condition

Note:Burdeniscalculatedindisability-adjustedlifeyears(DALYs).1Includesmentalhealthdisordersandsubstanceusedisorders.

2Synonymouswithpopulationattributablerisk(PAR).

Source:FionaJ.Charlsonetal.,“Thecontributionofmajordepressiontotheglobalburdenofischemicheartdisease:Acomparativeriskassessment,”BMCMedicine,November2013,Volume11,Number250

McKinsey&Company

Themodelusedthebestavailablesurveydata,statusreports,andevidencefromexpertstoestimate

currentadoptionlevelsforeachinterventionby

countryincomearchetype,usingWorldBank

categoriesofhighincome,upper-middleincome,

lower-middleincome,andlowincome.4Foreach

interventiontype,aramp-upcurvewasassignedto

considertheimplementationtimeneededtoincreaseaccessand,whererelevant,anygapbetween

interventiondeliveryandhealthimpact.

Themodelthencalculatedthepotentialdisease

reductionthatcouldbeachievedbyincreasing

adoptionoftheinterventionfromthecurrentlevelto

90percent(inotherwords,if90percentofeligiblepatientswereabletoaccesstheintervention),

applyingeachinterventioninsequence.

Sequencingwasbasedonthetypeofintervention,withbehavioralandpreventioninterventions

appliedbeforetreatmentforestablisheddiseaseandtreatmentforearlydiseasesequenced

beforemanagementoflaterdisease.Impactwasmeasuredannuallyuntil2050.Exhibit2laysouttheoverallapproach.

Toillustratehowhealthimprovementisscaledovertimeforonedisease,Exhibit3highlightsthestepsfollowedintheexampleofanxietydisorders.

4“WorldBankcountryandlendinggroups,”WorldBank,accessedApril1,2025.

Exhibit2

PrioritizingBrainHealth5

1

2

3

4

5

6

7

Approachtocalculatehealthimprovementthroughinterventions

Analytical

step

Description

Identifyandcategorize

relevanthealthinterventions

Reviewclinicalliteraturetoidentifyscalable,cost-efectiveinterventions,withthehighestpotentialtopreventandtreatdiseaseburden.

Determinehealthinterventione代cacyandadoptionrates

Literaturereviewforeachinterventionineachdiseaseareatoidentifytheefectivenessestimateinrelationtomortalityandmorbidityreduction.

Estimatetimetoseeimpactfromscalinginterventions

Estimateapprox.timerequiredforimplementationrampupandtimelagfrominterventionimplementationtoseeimpactondiseaseburden.

Establishsequencetoapplyhealthinterventions

Environmentalandbehavioralinterventionsapplied?rst,followedbymedicalprevention,andthentherapeuticinterventions.

Calculatediseaseburdenreductionpotentialbyscalinginterventions

Estimateimpactofapplyinghealthinterventionsforeverydisease,country,agegroupandgendersub-groupovertime.

Estimateimpactonlifeexpectancyandhealth-adjustedlifeexpectancy

Estimateimpactinhealth-adjustedlifeexpectancyyearsusingdeathsandYLDvaluesestimatedaspartofearlierstepsinthemodel.

Reviewoutputswithexperts

Inputs/outputstestedandre?nedfollowingreviewbyrelevantexperts.

McKinsey&Company

Calculationsofhealthimprovementthrough

interventionsreliedonsevensteps,outlinedbelow.

1.Identifyandcategorizerelevantinterventions

Healthinterventionswerecategorizedintofourgroups.

—environmental:interventionsrelatedtopolicyandregulation(forexample,alcoholtaxation)andplace-basedinterventions(forexample,

school-basedprogramsfordruguse,needleandsyringeprograms,orworkplaceprogramsforhigh-riskalcoholuse)

—behavioral:interventionsrelatedtoindividualbehavioralchange(forexample,supportforsmokingcessationorweightmanagement

throughlifestylechange)

—healthpromotionandprevention:including

screeningandearlydetection,primarycare,

andmedicinesforprevention(forexample,

antihypertensivesorGLP-1sforobesity)

—therapeutic:interventionssuchasspecialized

care(forexample,nonsurgicalbrainstimulation),

medicinesfortreatment(forexample,

antidepressants),caremanagement,counselingandtalkingtherapies(forexample,peersupportprogramsandpsychotherapy),anddigitaltoolsandtherapies

Theobjectivewastoidentifyhigh-impact,scalableinterventionsthatcouldhavethemostimpacton

reducingdiseaseburdenifscaledmoreeffectivelyandifaccessgapswerebridged.Itdoesnot

representacompletesetofinterventionsthatmightbeavailableinawell-resourcedandcomprehensivehealthsystem.

2.Determinehealthinterventionefficacyandadoptionrates

Interventioneffectiveness.Estimatesof

interventioneffectivenesswereextractedfromsystematicreviewsand,ifnosystematicreviewwasidentified,fromotherclinicalliterature.

Effectivenesswasestimatedseparatelyfor

morbidityandmortality.Formorbidityreduction,themostappropriateavailableoutcomemeasurewas

selected—forexample,changeinsymptomseverity.

Exhibit3

PrioritizingBrainHealth6

Components

Exampleofhealthimprovementthroughinterventions:Anxietydisorders

Step

1·2·3·4·5·6

Description

Clinicalpracticeguidelinesusedtoidentifycorehigh-impact,

scale-ableinterventions

Efectivenessestimates

from

systematicreviews

Adoption

Timerequired

Orderto

Disease

Other

estimates

forimplemen-

applyinter-

burden

impact

takenfrom

WHOsurvey

of21countries

tation(ramp-up)

ventions

reduction

estimates

Interventioncategory

InterventionIntervention

sub-description

category

Therapeutic

Medicines

usedin

Psychiatricmedicines

generalizedanxiety

disorders,eg,SSRIs/SNRIs

Estimateof

efectiveness

Current&additional1

adoptionrates

61%reductionafecting

morbidity

severityonly(YLDdisabilityweight)

HICs:22%68%

UMICs:13%77%

LMICs:9%81%

LICs:9%81%

Timeframe

HICs:5years

UMICs:10years

LMICs:15years

LICs:15years

Sequence

1

HALE/LEimpact

Reduction(2050)

Coveredseparately

24%

HICs:22%68%HICs:5years

UMICs:13%77%UMICs:10years

LMICs:9%81%LMICs:15years

LICs:9%81%LICs:15years

Therapeutic

Talking

therapies

and

counselling

Psycho-

therapeuticapproaches,eg,CBT

andRT

31%reductionafecting

2

13%

morbidity

severityonly(YLDdisabilityweight)

HICs:5years

UMICs:10years

LMICs:15years

LICs:15years

HICs:21%69%

UMICs:12%78%

LMICs:12%78%

LICs:5%85%

Therapeutic

Digital

mental

healthappsforanxiety

Talking

therapies

and

counselling

15%reductionafecting

3

6%

morbidity

severityonly(YLDdisabilityweight)

7Inputsandoutputsreviewedbyinternalexpertsandexternalexpertreviewer

Note:SSRI=selectiveserotoninreuptakeinhibitor;SNRI=serotoninandnorepinephrinereuptakeinhibitor;CBT=cognitivebehavioraltherapy;RT=relaxationtherapy;HIC/UMIC/LMIC/LIC=high-,uppermiddle-,lowermiddle-,andlow-incomecountries.

1Calculatedbytakingthediferencebetweenanaspirationaladoptionrateof90%minusthecurrentadoptionrate.

McKinsey&Company

Whereaninterventionwasonlyapplicabletoa

proportionofthediseaseburden,suchasaspecificagegroup,effectestimateswereappliedonlyto

appropriategroups.Forexample,aschools-basedcannabispreventionprogramwasappliedonlyto

theassociatedburdeninagegroupsfromtento

19years.Efficacywasassumedtobeconsistent

acrosscountryincomearchetypes.Theestimates

usedinthismodelwereintendedasaveragesacrossrelevantpatientpopulationsandmayvaryforspecificsubpopulationsnotconsideredinthemodel.

Interventionadoptionrates.Themodelaimstoestimatetheadditionalimpactofscalingmental

healthinterventionscomparedwiththecurrent

state.Theinterventionadoptionassumptionsusedinthemodelwerebasedonthedifferencebetweencurrentadoptionandaspirationaltargetadoption.

Currentadoptionrateswereestimatedforeach

interventionandcountryincomearchetypeusing

thebestavailableevidencereviewedbyexperts

inthefield.Theaspirationaltargetadoptionwas

assumedtobe90percentinallcases.ThisisbasedontheKennedyForum’sAlignmentforProgress

Goalsfor2033withavisiontoensureparityin

resources,access,quality,andoutcomesonmental

andsubstanceusedisorders,knownasthe90-90-90

PrioritizingBrainHealth7

framework.Theframeworksetsoutatargetfor90percentofindividualstobescreenedformental

healthconditionsorsubstanceusedisorders,90

percenttoreceivetheevidence-basedservicesandsupportstheyneed,and90percentofthosetreatedtomanagetheirsymptomsandachieverecovery.5

Thisdoesnotindicatethatallburdenisaddressed,

anditisnotthemaximumburdenacountrycould

aimtoaddress.Thereareinterventionsnotcapturedinthemodel,andtherewillbeinnovationsoverthe

timeframeofthismodelthatarenotincluded.

3.Estimatetimetoseeimpactfromscalinginterventions

Expandingaccesstointerventionstakestime.

Assumptionsaroundimplementationramp-up

timestoreachpeak(oraspirationaltarget)adoptionwerebuiltintothemodel,tailoredtodifferenttypesofinterventionandtoeachofthefourcountry

incomearchetypes.Theseestimateswerebased

onreal-worldexamplesoftimetoimplementationindifferenthealthsystemcontextsaswellasuniversalhealthcoveragetrends.Theanalysisusedan

S-shapedramp-upcurve,reflectingaslowerinitialadoptionratefollowedbyacceleratedadoptionovertime,tobettersimulatereal-lifescenarios.Ifthere

wasatimelagbetweenaccessinganinterventionandrealizingthehealthbenefitforaspecific

condition,thiswasalsoaccountedforthroughan

adjustmenttotheramp-upcurve.Delaysinseeinghealthbenefitsfromtreatmentarenottypical

formentalandsubstanceusedisordersbutmay

applytosomeoftheotherNCDscapturedinthe

additionalburden.Forexample,smokingcessationsupportnotonlyhasimmediatebenefitsforsomeconditionsbutalsoreducestheriskofdevelopingotherconditionsoversubsequentdecades.

4.Establishasequenceforapplyinghealthinterventions

Foreachincludeddisease,themodelquantified

theimpactofoneormorerelevantinterventions,

applyinganinterventiontomultipleconditions

whereappropriate.Tomoreaccuratelyreflectreal-worldimplementation,theimpactofinterventionswascalculatedsequentially.Theorderofthese

interventionswasdeterminedbytheirtype:

Environmentalandbehavioralinterventions

wereappliedfirst,followedbyhealthpromotion

andpreventivemeasuresandthentherapeutic

interventions.Eachsubsequentintervention’s

potentialimpactwasappliedonlytotheremainingdiseaseburdenafteraccountingforthereductionachievedbythepreviousinterventions.The

sequencingofinterventionswithineachcategory

wasdeterminedinconsultationwithclinicalexpertsinrelevantfields.Thissequencingapproachwas

alsousedtoavoidunintentionallydoublecountingpotentialimpactsanddoesnotreflectreal-life

clinicalpractice,inwhichmultipleinterventionsmaybedeployedsimultaneouslyandtreatmentorder

isbasedonindividualcircumstancesratherthanapredefinedsequence.

5.CalculatediseaseburdenreductionpotentialThemodelestimatedthepotentialreductionin

diseaseburdenforprimaryandassociatedmentalhealthconditionsthroughscalingprovenhealth

interventionsovertime.Theeffectsofapplying

healthinterventionswerecalculatedatthelevelofintervention,disease,country,agegroup(five-yeargroups),andsexfrom2025to2050.

Diseaseburdenreductionforprimarymental

healthconditions.Tocalculatetheaddressable

burdenfromprimarymentalhealthconditions,themodeluseddiseaseburdendatafromtheIHME

GBDdatasetandestimatedtherisk-attributable

burdenwhereapplicabletoensuretheattributableburdenwasmutuallyexclusiveacrossriskfactors.

Abaselinedatasetofdiseaseburden,including

risk-associatedburdenandcause-levelburdenforallmentalandsubstanceusedisordersinscope,

wasgeneratedfortheperiodfrom2025to2050byagegroup(five-yeargroups),sex,andcountry.

Measuresincludedinthemodelwereyearslived

withdisability(YLDs),yearsoflifelosttoprematuremortality(YLLs),meandisabilityweight,incidence,prevalence,anddisease-relateddeaths.

Healthinterventionsandtheireffectswere

implementedsequentiallyovertimeasoutlinedintheprevioussection,beginningwiththoselinkedtomodifiableriskfactors.Subsequently,eachfurther

5“Alignmentforprogressgoalsfor2033:90-90-90,”KennedyForum,accessedApril1,2025.

PrioritizingBrainHealth8

interventionwasappliedtotheresidualcondition-

levelburden.Theinterventionswereappliedto

theappropriateagegroupswhenapplicable.For

instance,school-basedprogramsforalcoholuse

wereimplementedonlyforindividualsunderage

20.Todeterminetheimpact,thediseaseburden

foreachrelevantpopulationwasmultipliedbythe

interventionefficacyrateadjustedfortheadditionalpotentialadoptionrateandtheramp-upfactorfor

theyear(andspecifictothecategoryofinterventionandincomearchetypeofthecountry).

Thepotentialdiseaseburdenreductionwas

estimatedformultiplemeasures,including

incidence,deaths,prevalence,YLDs,YLLs,and

DALYs.Reductionsinincidencewerecalculated

usingtheIHMEdiseaseburdenasthebaseline

foreachyearandapplyingthereductionimpact

(effectivenessadjustedforadditionalpotential

adoption)forpreventiveinterventionsasdescribedpreviously.Toestimatetheimpactondisease-

relateddeaths,thechangeindeathratewas

calculatedbyconsideringboththereductionin

mortalityfrominterventionsandthepreviously

calculatedreductioninincidence.Themodel

estimatedbaselinerecoveriesusingIHME

prevalence,incidence,anddeathvalues,andit

assessedtheimpactofanycurativeinterventions.

Baselinemeandisabilityweightwasdetermined

usingIHMEprevalenceandYLDvalues,withthe

impactestimatedbasedonthepotentialeffectof

interventionsonmorbidity(forexample,reduction

infrequency,duration,orseverityofsymptoms).Theimpactonprevalencewasestimatedbasedonthe

newlycalculatedincidence,deaths,andrecoveries.

Next,theimpactonYLDswascalculatedbased

ontheestimatedimpactonprevalenceandmean

disabilityweight(morbidity),whiletheimpacton

YLLswasderivedfromthedeathsestimatedinthe

previousstep.Finally,outputswereextrapolated

fordiseasesnotincludedinthedetailedanalysis,

assumingthesameaverageimpactratefordiseaseswithinthelevel2diseasecategoryascategorizedin

theIHMEGBDdataset.Therewasonlyonediseasegroupforwhichthisextrapolationwasperformed:

othermentaldisorders.

Estimatediseaseburdenreductionforassociated

mentalhealthconditionburden.Toestimatethe

potentialreductioninadditionaldiseaseburden

associatedwithapreexistingmentalorsubstance

usedisorder(asdescribedinthepreviousstep,

“Associatedburdenfrompreexistingmentalhealth

conditionsexacerbatingotherNCDs”),themodel

usesasimplifyingassumptionofadirectrelationshipbetweeneachconditionpair.Itisimplicitinthis

premisethattheconditionpairrelationshipisboth

linearandcausalthatis,thata10percentreductioninanxietydisorderdiseaseburdenwouldleadtoa

10percentreductioninanyadditionalassociated

burden(fromnon–mentalhealthNCDs).Thereis

insufficientevidencetotestthispremise,anditis

beyondthescopeofthisworktodoso.Thiscouldbeavaluableareaforfurtherresearch.

6.Estimateimpactonlifeexpectancy

andhealth-adjustedlifeexpectancy

Toestimatetheimpactofscalingmentalhealth

interventionsonlifeexpectancy(LE)andhealth-

adjustedlifeexpectancy(HALE),themodel

recalculatedtheunderlyinglifetablesusingthe

remainingdeathsandYLDpercapitaderivedfromthepreviousstepsafterscalingmentalhealth

interventions.Comparingpre-andpostinterventionvaluesforLEandHALEresultedinadeterminationoftheincreaseinLEandHALEthatwasduetotheappliedhealthinterventions.6

7.Reviewoutputswithexperts

Allmodelinputsgatheredbytheresearchteamandmodeloutputsfromthemodelwerereviewedby

clinicalexpertsinspecificdiseaseareasinmentalhealthandsubstanceusedisorders.Theseexpertsassessedthebasketofinterventionsidentifiedforeachdisease,thepotentialforincreaseduptake,

theorderofimplementation,andtheoverallhealthimpactacrossdifferentcountryincomegroups.

6HaidongWangetal.,“Globalage-sex-specificfertility,mortality,healthylifeexpectancy(HALE),andpopulationestimatesin204countries

andterritories,1950–2019:AcomprehensivedemographicanalysisfortheGlobalBurdenofDiseaseStudy2019,”TheLancet,October2020,Volume396,Number10258.AbridgedlifetabledefinitionsbyM.Greenwood,“Discussiononthevalueoflife-tablesinstatisticalresearch,”

JournaloftheRoyalStatisticalSociety,June1922,Volume85,Number4;andChinLongChiang,TheLifeTableandItsApplication,KriegerPublishingCompany,1984.

PrioritizingBrainHealth9

Thisprocessprovidedadditionalassuranceof

appropriatec