Experimenta an Clinica Psych pharmacolog  Vol  No   th publi domai Comparativ Epidemiolog o Dependenc o Tobacco Alcohol Controlle Substances an Inhalants Basi Finding Fro th Nationa Comorbidit Surve
199K - views

Experimenta an Clinica Psych pharmacolog Vol No th publi domai Comparativ Epidemiolog o Dependenc o Tobacco Alcohol Controlle Substances an Inhalants Basi Finding Fro th Nationa Comorbidit Surve

rev America Psychiatri Association 1987 dru dependenc amon American 155 year old w foun abou 1 i 4 24 ha a histor o tobacc dependence abou 1 i 7 14 ha a histor o alcoho dependence an abou 1 i 1 75 ha a histor o dependenc o a inhalan o controlle drug

Download Pdf

Experimenta an Clinica Psych pharmacolog Vol No th publi domai Comparativ Epidemiolog o Dependenc o Tobacco Alcohol Controlle Substances an Inhalants Basi Finding Fro th Nationa Comorbidit Surve




Download Pdf - The PPT/PDF document "Experimenta an Clinica Psych pharmacolog..." is the property of its rightful owner. Permission is granted to download and print the materials on this web site for personal, non-commercial use only, and to display it on your personal computer provided you do not modify the materials and that you retain all copyright notices contained in the materials. By downloading content from our website, you accept the terms of this agreement.



Presentation on theme: "Experimenta an Clinica Psych pharmacolog Vol No th publi domai Comparativ Epidemiolog o Dependenc o Tobacco Alcohol Controlle Substances an Inhalants Basi Finding Fro th Nationa Comorbidit Surve"— Presentation transcript:


Page 1
Experimenta an Clinica Psych pharmacolog 1994 Vol 2 No 3 244-26 th publi domai Comparativ Epidemiolog o Dependenc o Tobacco Alcohol Controlle Substances an Inhalants Basi Finding Fro th Nationa Comorbidit Surve Jame C Anthony Lyn A Warner an Ronal C Kessle Studyin prevalenc o Diagnostic an Statistical Manual (3r ed. rev. America Psychiatri Association 1987 dru dependenc amon American 15-5 year old w foun abou 1 i 4 (24% ha a histor o tobacc dependence abou 1 i 7 (14% ha a histor o alcoho dependence an abou 1 i 1 (7.5% ha a histor o dependenc o a inhalan o controlle drug Abou

on thir o tobacc smoker ha develope tobacc dependenc an abou 15 o drinker ha becom alcoho dependent Amon user o th othe drugs abou 15 ha becom dependent Man mor American ag 15-5 hav bee affecte b dependenc o psychoactiv substance tha othe psychiatri disturbance no accorde a highe priorit i menta healt servic deliver systems prevention an sponsore researc programs Th ai o thi articl i t repor basi descrip tiv finding fro ne researc o th epidemiol og o dru dependenc syndromes conducte a par o th Nationa Comorbidit Surve (NCS) I thi study ou researc tea secure a nationall representativ sampl an

applie standardize diagnosti assessment i a wa tha allow direc comparison acros prevalenc estimate an cor Jame C Anthony Etiolog Branch Addictio Re searc Center Nationa Institut o Dru Abus an John Hopkin University Lyn A Warne an Ronal Kessler Institut fo Socia Researc an Depart men o Sociology Universit o Michigan Th Nationa Comorbidit Surve (NCS i a collabo rativ epidemiologi investigatio o th prevalence causes an consequence o psychiatri morbidit an Comorbidit i th Unite States Th NC i supporte U.S Publi Healt Servic Grant M 4637 an 4909 wit supplementa suppor fro th Nationa Institut o Dru

Abus an W T Gran Foundatio Gran 90135190 Preparatio o thi articl wa supporte b th Nationa Institut o Dru Abus Addictio Researc Center W acknowledg H Chilcoa fo valuabl re searc assistance Correspondenc concernin thi articl ma b ad dresse t Jame C Anthony P O Bo 5180 Addictio Researc Center Nationa Institut o Dru Abuse Baltimore Marylan 21224 Electroni mai ma b sen anthony@jhuhyg.sph.jhu.edu relate o tobacc dependence alcoho depen dence an dependenc o othe psychoactiv drug (Kessle e al. 1994) Fo thi overvie o th survey' findings a primar goa ha bee t answe tw basi epide miologi question abou

dru dependenc involv in tobacco alcohol controlle drug suc a cocaine an inhalants First i th populatio unde study wha proportio o person no qualifie a a currentl activ o forme cas o dru dependence Second wher ar th affecte case mor likel t b foun withi th sociodemo graphi structur o th stud population addition populatio estimate presente i thi articl she ligh o th epidemiolog o dependenc o tobacco alcohol an th followin individua drug an dru groups cannabis heroin cocaine psychostimulant othe tha cocaine an algesi drugs a dru grou consistin o anxiolytic sedative an hypnoti drugs psychedeli

drugs an inhalan drugs Th followin populatio esti mate ar presente fo eac o thes liste drugs includin tobacc an alcohol (a lifetim preva lenc o dru dependence evaluate i relatio t criteri publishe i th Diagnostic an Statistical Manual o Mental Disorders, Thir Edition Re vise (DSM-III-R America Psychiatri Associa tion 1987) (b lifetim prevalenc o extramedica dru use define t encompas illici dru us a wel a patient takin prescribe medicine t ge 24
Page 2
COMPARATIV EPIDEMIOLOG O DRUG 24 high takin mor tha wa prescribed o takin medicine fo othe reason no intende b th doctor an (c th

proportio o extramedica user wh ha becom dru dependent Usin estimate suc a these w see t de scrib th broa populatio experienc wit form psychoactiv dru us tha generall occu withou scrutin o contro b prescribers pharma cists o othe healt practitioners Althoug con cedin man reason peopl migh den o under repor thei illici dru us o dru problems w dra attentio t ho ofte illici dru us an symptom o dru us disorder ar acknowledge surve researc o thi type Fo example, o th basi o confidentia interview conducte fo th Epidemiologi Catchmen Are (ECA surve mor tha 1 year ago w foun tha on i thre adul

American (30.5% reporte a histor o recen o pas illici dru use O th basi o self-repor alone 20 o thes illici dru user ha a history of dependenc on controlle substance or a relate dru disorder No countin tobacc depen dence abou on i si adul American (17% me diagnosti criteri fo eithe a alcoho o dru disorder o bot (Anthon & Helzer 1991) Thes ar substantia estimate tha conve th publi healt significanc o dru us an dru dependenc i th Unite States an the ar fa to larg t b du t th typ o exaggeratio an overreportin sometime foun i survey o dru us i earl adolescenc (Johnston O'Malley & Bachman 1992) I a

correctio coul b mad fo underreporting thes substantia estimate woul eve larger th 1 year sinc th star o th EC surveys th population' dru experienc ha change i importan ways wit passag throug a now-subsidin epidemi o crac smokin an othe cocain us (Harrison 1992 Kandel 1991) Th NC chronicle result o thes change an draw strengt fro som methodologica refine ment tha wer no par o th EC researc plan (a a nationall representativ sampl o 15-54-year-olds (b a mor complet assessmen extramedica dru use, applyin measuremen strategie develope fo th Nationa Househol Surve o Dru Abus (NHSDA U.S Depart men o

Healt an Huma Service [USDHHS] 1993) (c deliberat alignmen o diagnosti crite ri an th measuremen strategie use t asses dependenc o alcohol tobacco an othe drugs a t allo comparison acros dru groups an (d mor thoroug adjustmen fo nonrespons biase introduce b designate respondent wh decline t b assessed perhap fo reason con necte t alcoho o dru dependence Th descriptiv estimate presente i thi over vie se th stag fo ongoin researc i whic w ar testin hypothese abou suspecte determi nant an consequence o dru dependence in cludin link betwee dru dependenc an othe psychiatri condition suc a anxiet an

moo disorder (Kessler i press) Thes finding ma interes pharmacologist an othe scientist wh ar concerne abou th population' dru experi enc outsid th boundarie o laborator an clinica researc an practice Thos wh stud th reinforcin function o dru us an dependenc liabilit o individua drug ma gai usefu insight considerin comparativ aspect o th epidemi olog o tobacco alcohol an othe dru depen dence includin epidemiologi evidenc o th transitio fro a single occasio o dru us towar th developmen o dru dependence a topi o considerabl interes withi th clinica an researc communit (e.g. se Anthony 1991 Glant

& Pickens 1992 Henningfield 1992) Inves tigator als wil fin thes populatio estimate usefu a the see t substantiat th potentia publi healt significanc o thei pharmacologi studie o t analyz publi policies Finally thes estimate ma hav valu fo primar car practitio ner an famil doctors a wel a psychologists psychiatrists o othe specialist wh prescrib psychoactiv drug o wh nee t anticipat ho frequentl th healt statu o thei patient migh complicate b a histor o dependenc o tobacco alcohol o othe drugs Metho an Material Th NC wa base o a stratified multistag are probabilit sampl o person 1 t 5 year ol

i th noninstitutionalize civilia populatio th 4 coterminou Unite States includin a representativ sampl o student livin i campu grou housing Fieldwor wa carrie ou b th professiona fiel staf o th Surve Researc Cente a th Universit o Michiga betwee Septembe 14,199 an Februar 6,1992 T allo midcours adjustment an adaptatio t unantici pate problems th fieldwor wa organize i
Page 3
24 ANTHONY L WARNER AN R KESSLE relatio t time releas o si replicat sub samples eac designe t b representativ o th stud population Overal respons rat wa 82.4% wit a tota o 8,09 participants A mor detaile

discussio o th NC samplin desig an it implementatio ha bee give b Kessle e al (1994) Afte sampling 1 o th 15 speciall traine surve interviewer me wit eac designate re sponden t administe th Composit Interna tiona Diagnosti Intervie (CIDI) a adapte fo th NC t yiel detaile informatio abou a broa rang o psychiatri disorders includin dru dependenc (Cottie e al. 1991 Robin e al. 1988 Wittchen i press) Thes interviewer participate i a 7-da study-specifi trainin pro gra i th us o th CID befor beginnin fieldwork The wer traine t follo a protoco intende t engag th designate respondents interes i th

surve an t reinforc surve participation t secur a privat locatio fo th interview whic mos ofte wa withi th plac o residence t develo trus an rappor wit th responden an t obtai informe consen befor startin th interview t administe th surve questions a worded i a fixe sequence an t recor eac subject' response i a precede respons booklet Th interviewer wer no give specia trainin i psychopatholog o psychophar macology an the wer no mad awar o an ke hypothese unde stud o o ou researc team' interes i th reinforcin function serve dru use Th highl structure an standard ize intervie schedul als wa use

t gathe informatio o suspecte correlate an conse quence o psychiatri disorders includin educa tiona attainment occupation an othe character istic o designate respondent o thei households Becaus previou survey hav provide som evidenc tha surve nonrespondent hav mor psychiatri disorder tha respondents a supple menta nonrespons surve wa conducte i tan de wit th mai NC survey Thi wa don b selectin a rando subsampl o designate respon dent wh initiall wer no interviewe eithe becaus o refusa o (i rar cases inabilit t contac afte man attempts Thes person wer aske t complet a short-for versio o th

diagnosti interview Assessment of Alcohol and Other Drug Use Respondent wer aske separat question o thei us o alcoholi beverages tobacco an th othe individua drug an dru group liste below Th surve question o th frequenc an recenc o takin controlle substance an inhal ant wer nearl identica t standardize assess ment develope fo th NHSDA no sponsore th Offic o Applie Studie withi th Sub stanc Abus an Menta Healt Service Admin istration Thes question clarifie ou focu o illici us o Schedul I drug suc a marijuana heroin an LSD a wel a extramedica us o cocain an othe drug tha ca b obtaine throug

legitimat medica channels The wer phrase t encompas us o thes drug an medicine "o you own eithe withou you ow prescriptio fro a doctor o i greate amount mor ofte tha prescribed o fo an reaso othe tha a docto sai yo shoul tak them (USDHHS 1993) I addition respondent wer aske whethe the ha starte t fee dependen a dru whil takin i i accor wit a doctor' prescription Ensuin surve question covere topic suc a ag o onset frequency an recenc extramedica dru us fo eac o th followin individua drug an dru groups whic wer adapte fro NHSD conventions heroin othe opioid an analgesic tha ca b obtaine throug

medica channel (e.g. morphine propoxyphene codeine) cannabi (marijuana hashish o both) psychedeli drug (e.g. LSD peyote mescaline) inhalan drug (e.g. gasolin o lighte fluids spra paints amy nitrite nitrou oxide) cocain (includin crac cocain an freebase) psy chostimulant othe tha cocain (e.g. dextroam phetamine methamphetamine) an anxiolytic sedative an hypnoti drug (e.g. secobarbita an diazepam a wel a mor recentl introduce compound suc a flurazepam alprazolam an triazolam) Consisten wit th NHSD an EG surveys th NC assessmen strateg include a detaile verba descriptio o eac dru grou an list o

qualifyin drug tha wer rea t eac participant bu i di no includ th NHSD colore pil car wit picture o differen pharmaceutica products Furthermore th NC interviewe rea th ques tion an recorde eac participant' answer o a precede respons form Thi approac wa consis
Page 4
COMPARATIV EPIDEMIOLOG O DRUG 24 ten wit prio EC survey o dru dependenc bu wa i contras wit th NHSD approac o allowin th responden t self-administe surve question o dru use Whe interviewer adminis te th questions i i possibl t reduc th impac o lo level o literac an readin achieve men amon participant an t us intervie ski

out an branchin pattern tha ca shorte th intervie an distribut it coverag t othe importan topic suc a suspecte ris factor an us o menta healt an othe medica services Althoug w acknowledg tha som populatio group ma repor mor dru us whe the self-administe NHSD questionnaire (Schober Caces Pergamit & Branden 1992) w hav mad direc compariso o NC an 199 NHSD estimates generall findin tha th estimate wer ver clos t on another an tha th NHSD estimat alway wa locate withi th 95 confidenc interva fo th correspondin NC estimate W retur t thi topi i th Discussio section NC question o th us o alcoholi

beverage followe a simila NHSD forma an elicite informatio abou ag o onset frequency re cency an quantit o drinking Respondent als wer aske whethe the ha consume a leas 1 drink i an singl yea o thei lives Diagnostic Assessment of Alcohol and Other Drug Dependence Th CID diagnosti assessmen o alcoho an othe dru dependenc fo th NC wa base o DSM-III- criteria translate int standardize surve question fo administratio b a traine la interviewer A i th EC progra metho use fo Diagnosti Intervie Schedul diagnose (Rob ins Helzer Croughan & Radcliff 1985) eac participant' answer t th surve question hav

bee recorde an converte t a machine readabl format an a compute progra ha bee use t determin whethe th diagnosti criteri hav bee met A summarize recentl b Wittche (i press) Worl Healt Organizatio fiel trial an othe methodologica studie hav provide evidenc tha th CID assessment fo alcoho an othe dru dependenc hav accept abl level o interrate reliabilit an test-retes reliability an generall ar congruen wit inde pendentl mad standardize clinica diagnoses DSM-III- criteri requir evidenc concern in nin manifestation o alcoho o othe dru dependenc groupe unde th headin o Crite rio A modele loosel

afte th origina Edwards Gros concep fo a alcoho dependenc syn drom (Edward & Gross 1976) Th lis o nin manifestation cover a rang o sign o symp toms suc a thos tha occu i th contex o a dru withdrawa syndrome a wel a behaviora manifestation o dru dependenc suc a unsuc cessfu attempt t sto o cu dow o dru use an sustaine us despit recognitio tha i i relate t social psychological o physica prob lems T qualif fo a DSM-III- dru depen denc diagnosis a leas thre o thes nin Crite rio A manifestation mus b met I addition Criterio B require tha th disturbanc ha persiste fo a leas on mont o tha presentin

feature o dru dependenc hav appeare repeat edl ove a longe perio o time Th CID include tw o mor surve item designe t ta th domain represente b eac o th nin Criterio A manifestations a wel a question concernin Criterio B Th CID lifetim diagno si fo alcoho o othe dru dependenc i no mad unles ther i positiv evidenc tha th responden meet bot Criterio A an Criterio B Thi assessmen o alcoho o othe dru depen denc wa administere wheneve participant reporte occasion o extramedica us o con trolle substance o inhalant i thei lifetimes o whe the reporte consumin 1 o mor drink an on year Fo th assessment

identica standardize question wer aske fo alcohol controlle substances an inhalants B holdin constan bot th diagnosti criteri an th man ne i whic th criteri wer assessed w sough reduc methodologi variatio tha otherwis migh distor comparison betwee alcoho an th othe drugs I wa no possibl t contro fo thes difference i th EC survey (Anthony 1991 Anthon & Helzer 1991) Tw othe methodologi contrast betwee th EC survey an th NC als shoul b men tione i relatio t controlle substances First i contras wit th NCS th EC survey di no chec fo dru dependenc (Diagnostic an Statisti- ca Manual o Mental

Disorders, 3r ed. DSM-III America Psychiatri Association 1980 whe a participan reporte us o a medicin i accor wit a doctor' prescription eve i tha us ha
Page 5
24 ANTHONY L WARNER AN R KESSLE le t feeling o dependence Second th NC include inhalant whe assessin dru depen denc (DSM-III-R) wherea th EG survey di not A i th EG surveys dependenc wa assesse wheneve participant reporte a leas severa occasion o extramedica dru use unde th assumptio tha eve a fe a si occasion migh b sufficien fo developmen o dru depen dence bu tha dru dependenc woul b ex tremel rar o improbabl amon person wh

ha use th dru n mor tha severa times Assessment of Tobacco Use and Dependence Whe th NC fieldwor started ther wer insufficien fund t allo NC assessmen o tobacc us an tobacc dependenc durin a alread length interview Midwa throug field work supplementa fundin an intervie tim becam availabl fo inclusio o a CID sectio DSM-III- tobacc dependenc designe t paralle t th CID assessmen fo DSM-III- dependenc o alcoho an othe drugs bu als includin a fe standardize question o behav ior specifi t tobacc smoking Becaus o th NC replicat samplin plan i wa possibl t administe th tobacc assessmen t a representa

tiv subsampl o NC participants consistin o 4,41 persons o 55 o th tota NC sample Thi assessmen o tobacc dependenc in clude question abou dail tobacc smokin bu no abou mor infrequen o irregula smoking Specia analyse o th 199 NHSD dat hav bee complete t fil thi ga o information Th 199 NHSD wa conducte midwa throug th NC fieldwork wit a nationwid probabilit sampl o person 1 year o ag an olde an wit surve assessment o dru us alread de scribe i thi article Th NHSD surve ques tion ascertai whethe tobacc cigarette wer smoked eve o a singl occasion s tha th resultin estimate fo tobacc us correspon t

th NC estimate o alcoho us o a leas on occasio an extramedica us o othe drug o a leas on occasion T confor wit th NCS th NHSD analyse wer restricte t 15-54-year olds. Quality Control Measures During Fieldwork Th interviewer wer monitore b 1 regiona supervisor responsibl fo editin complete inter view befor the wer forwarde t th nationa fiel office I addition centra fiel offic staf reviewe interview a soo a the wer receive fro supervisors Wheneve error wer foun o importan informatio wa missing th intervie assignment wer sen bac t th fiel fo resolu tion an th respondent wer recontacte t clarif

thei answers Analysis Procedures presen survey-base populatio estimate fo th lifetim prevalenc o extramedica dru us an th lifetim prevalenc o dru depen denc i relatio t alcohol tobacco an th othe individua drug o dru group previousl listed Eac prevalenc estimat fo dru us i a propor tio i whic th numerato consist o th esti mate numbe o person wh hav ha a leas on occasio o extramedica dru us i thei lifetimes wherea th denominato i th tota stud population Eac populatio prevalenc esti mat fo dru dependenc ha th sam denomina tor bu th numerato i th estimate numbe o person wh qualif fo th CID

lifetim diagno si fo dru dependenc accordin t DSM-III- criteria I addition w repor fo eac individua dru o dru grou estimate proportion o dru user i th stud populatio wh ha develope dru dependence I concept thes proportion ma b regarde a estimate o th lifetim prevalenc o dru dependenc amon user i th stud population Algebr ca b use t sho tha eac proportio i equa t th lifetim prevalenc dru dependenc i th stud population di vide b th lifetim prevalenc o dru us i th stud population Becaus w ha t rel o NHSD estimate fo tobacc use i wa neces sar t appl th algebrai metho whe w estimate th proportio o

tobacc smoker wh ha develope tobacc dependence an standar error hav no bee estimate fo thes tobacc estimates als presen estimate fo th strengt o associatio betwee dru dependenc an plau sibl determinant o dru dependence includin fixe characteristic suc a birt yea (age an Thes NHSD analyse wer conducte b Howar Chilcoa a th Etiolog Branc o th Nationa Institut Dru Abuse Addictio Researc Center
Page 6
COMPARATIV EPIDEMIOLOG O DRUG 24 sex a wel a potentiall modifiabl characteris tic suc a employmen status Tabl 1 give a frequenc distributio fo eac variabl consid ere i thi analysi o th basi

o unweighte NC sampl data T inde th strengt o associa tio betwee dru dependenc an eac variabl liste i Tabl 1 w hav produce a estimat fo th odd ratio Whe a particula characteristi ha n associatio wit bein a currentl o for merl activ cas o dru dependence th odd rati estimat wil b 1.0 o indistinguishabl fro 1. withi th limit o surve precision A odd rati abov 1. signal a positiv association wherea a odd rati betwee 0. an 1. signal invers associatio (Fleiss 1981) Th odd ratio fo thi analysi hav bee estimate usin logisti regressio model an th Statistica Analy si System' PRO LOGISTI (SA Institute

1988) Asid fro th unweighte sampl dat give i Tabl 1 al result reporte i thi articl ar base o conventiona procedure fo analysi o comple sampl surve data W hav use weight compensat fo variatio i sampl selectio probabilitie a wel a poststratificatio adjust men factor tha compensat fo surve nonre spons a wel a othe potentia source o surve error Correspondin weight an poststratifica tio adjustmen factor als hav bee take int accoun i th 199 NHSD estimate fo tobacc us reporte hereinafter Becaus o th comple sampl desig an weighting standar error o proportion wer estimate usin th Taylo serie

linearizatio metho (Woodruf & Causey 1976) Th PSRA TI progra in the OSIRI IV statistica analysi an dat managemen packag wa use t mak thes calculation (Universit o Michigan 1981) Standar error o odd ratio wer estimate usin th metho o Balance Repeate Replica tio i 4 design-base balance subsample (Kis Frankel 1970) Thes analyti procedure hav bee describe i mor detai b Kessle e al (1994) Result How Many 15-54- Year-Old Americans Have Developed Drug Dependence? Tabl 2 show lifetim prevalenc estimate an standar erro fo eac estimat o th basi o CID interview administere t th 15-54-year ol NC stud

populatio betwee lat 199 an earl 1992 Accordin t Tabl 2 (se colum 2) estimate 24.1 o thi stud populatio ha develope tobacc dependenc (±1.0%) wherea 14.1 ha develope alcoho dependenc (±0.7%) an 7.5 (±0.4% ha develope depen denc o a leas on o th controlle substance inhalan drug liste i Tabl 2 Thus i ran order a histor o tobacc dependenc appeare mos frequentl i thi stud population affectin abou 1 i 4 persons Alcoho dependenc wa nex mos prevalent havin affecte abou 1 i 7 persons A histor o dependenc o othe drug followed i aggregat havin affecte abou 1 i persons No countin tobacc an alcohol cannabi

accounte fo mor dependenc tha an othe dru o dru group I th NC stud population 4.2 qualifie fo th lifetim diagnosi o canna bi dependenc (Tabl 2 ro 4 colum 2) Depen denc o cocaine includin crac cocaine wa nex i rank A estimate 2.7 o th 15-54-year ol stud populatio ha develope cocain depen dence Prevalenc estimate fo onl tw othe dru categorie wer abov 1.0% Dependenc o psychostimulant othe tha cocain (e.g. amphet amines wa 1.7% an dependenc upo anxio lytic sedative o hypnoti drug wa 1.2 (Tabl ro 7 colum 2) Within the Study Population, Where Was Drug Dependence Found? Dru dependenc wa no distribute

randoml withi th stud population som group wer affecte mor tha others Thi ca b see i Tabl 3 i whic logisti regressio wa use t produc odd rati estimate tha sho th strengt associatio betwee dru dependenc an variou selecte prevalenc correlate suc a ag an sex Sociodemographic Variation: Sex, Age, and Race-Ethnicity Me wer somewha mor likel tha wome t hav bee affecte b dependenc o alcoho an dependenc o controlle substance o inhal ants bu no b tobacc dependence Whe w compare th odd o dependenc fo me versu
Page 7
Tabl 1 Description of National Comorbidity Survey Sample in Relation to

Selected Characteristics Characteristi Se Mal Femal Ag 15-2 25-3 35^ + Rac Whit Blac Hispani Othe Employmen Workin Studen Homemake Othe Educatio (i years 0-1 13-1 16 Incom (i thousand o dollars 00-1 20-3 35-6 70 Househol compositio Liv alon Liv wit spous Liv wit othe Liv wit paren Marita statu Marrie o cohabitin Separated widowed o divorce Neve marrie Religio Protestan Catholi preferenc Othe Regio Northeas Midwes Sout Wes Urbanicit Metropolita Othe urba Nonurba Me 86 1,21 1,12 64 2,93 42 36 12 3,02 52 28 73 1,22 94 93 96 99 1,36 52 68 2,02 81 32 2,05 50 1,29 2,00 1,06 30 47 72 99 1,35 78 1,70

1,27 86 Wome 90 1,41 1,11 82 3,15 58 37 14 3,01 55 48 20 73 1,45 1,18 88 1,38 1,03 1,35 47 51 2,31 58 83 2,35 75 1,14 2,46 1,18 29 30 83 1,08 1,52 80 1,88 1,45 91 Tota 3,84 4,25 1,76 2,62 2,24 1,46 6,08 1,01 73 27 6,03 1,07 48 49 1,47 2,67 2,13 1,81 2,34 2,03 2,72 1,00 1,19 4,34 1,40 1,15 4,41 1,25 2,43 4,47 2,25 59 78 1,55 2,07 2,88 1,58 3,58 2,72 1,78 Note. Unweighte sampl dat fro th Nationa Comorbidit Surve i th coterminou Unite States 1990-1992
Page 8
COMPARATIV EPIDEMIOLOG O DRUG 25 Tabl 2 Estimated Prevalence of Extramedical Use and Dependence in Total Study Population and

Lifetime Dependence Among Users Proportio wit a histor o dependenc Dru categorie Tobacco Alcoho Othe drug Cannabi Cocain Stimulan Anxiolytics etc. Analgesic Psychedelic Heroi Inhalant 24. 14. 7. 4. 2. 1. 1. 0. 0. 0. 0. SE 1. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. Proportio wit a histor o extramedica us 75.6 91. 51. 46. 16. 15. 12. 9. 10. 1. 6. SE 0. 0. 1. 1. 0. 0. 0. 0. 0. 0. 0. Dependenc amon extramedica user 31. 15. 14. 9. 16. 11. 9. 7. 4. 23. 3. SE 0. 0. 0. 1. 1. 1. 1. 0. 5. 1. Note. Weighte estimate fro th Nationa Comorbidit Surve dat gathere i 1990-199 fo person 15-5 year ol ( = 8,098) Das

indicate dat no estimated P = Estimate prevalenc proportion = 4,414 b Anxiolytics sedatives an hypnoti drugs grouped th odd o dependenc fo women th odd rati wa 2.8 fo alcoho dependenc (95 confidenc interva [CI = 2.29 3.44 an 1.6 fo controlle substance o inhalant (95 C = 1.24 2.13) I contrast th observe odd rati wa 1.1 fo tobacc dependence n mor tha slightl greate tha th odd rati valu (1.0 tha i expecte unde th nul hypothesi o n association More over th 95 C fo th associatio betwee tobacc dependenc an se ha a spa fro 0.9 1.40 trappin th nul valu o 1.0 Thus th evidenc i balance towar a mal exces

i th prevalenc o dependenc o alcohol controlle substances o inhalant withi thi stud popula tion bu no towar a mal exces i prevalenc o tobacc dependence Wit respec t age a histor o tobacc depen denc wa leas commo amon 15-24-year-olds wherea a histor o dependenc o alcohol con trolle substance o inhalant wa leas commo amon 45-54-year-olds A show i Tabl 3 th odd o tobacc dependenc amon 15-24-year old wer abou one-hal th odd o tobacc dependenc amon 45-54-year-old (odd rati [OR = 0.48 95 C = 0.35,0.64) However com pare wit 45-54-year-olds tobacc dependenc wa n mor commo amon 25-34-year-old (O = 0.96

o 35-44-year-old (O = 1.0) contrast a histor o alcoho dependenc wa leas commo amon 45-54-year-olds B compari so wit thes olde adults th 15-24-year-old wer a estimate 1.4 time mor likel t qualif fo a lifetim alcoho dependenc diagnosi (O = 1.41 95 C = 1.08,1.84) th correspond in odd rati wa 1.6 fo th 25-34-year-old (95 C = 1.27 2.16 an fo th 35-44-year-old (95%C = 1.35,2.02) history o dependenc o controlle sub stance o inhalant wa mos likel t b foun amon youn adults an wa leas likel t b foun amon 45-54-year-olds B compariso wit 45-54-year-olds th estimate odd o depen denc o thes drug wer 2.6

time greate amon 15-24-year-olds 3.5 time greate amon 25-34-year-olds an 3.0 time greate amon 35-44-year-olds Compare wit White Americans th African America segmen o th stud populatio wa les likel t hav a histor o tobacc dependenc (O = 0.59) alcoho dependenc (O = 0.35) o dependenc o othe drug (O = 0.54) Hispani American als wer les likel tha Whit Ameri can t hav a histor o tobacc dependence bu thi wa no th cas fo alcoho dependenc (p > .05 o dependenc o othe drug (p > .05) Thes invers association betwee dru depen
Page 9
Tabl 3 Demographic Correlates of Tobacco, Alcohol, and Other Drug

Dependence Tobacc Characteristi Se Mal Femal Ag 15-2 25-3 35-4 + Rac Whit Blac Hispani Othe Employmen Workin Studen Homemake Othe Educatio (i years 0-1 13-1 16 Incom (i thousand o dollars 00-1 20-3 35-6 70 Househol compositio Liv alon Liv wit spous Liv wit othe Liv wit paren Marita statu Marrie o cohabitin Separated widowed divorce Neve marrie Religio Protestan Catholi preferenc Othe Regio Northeas Midwes Sout Wes Urbanicit Metropolita Othe urba Nonurba 1.1 1.0 0.48 0.9 1.0 1.0 1.0 0.45 0.59 0.6 1.0 0.45 1.53 1.81 1.68 1.85 1.47 1.0 1.3 1.44 1.2 1.0 1.0 1.0 0.8 0.39 2.14 2.12 1.0 0.7 0.70 1.0

0.7 1.0 1.0 1.0 0.9 0.7 0.8 1.0 95 C 0.99 0.35 0.70 0.80 0.33 0.38 0.36 0.30 1.08 1.31 1.14 1.34 1.04 0.99 1.05 0.91 0.80 0.59 0.28 1.79 1.59 0.57 0.50 0.46 0.79 0.76 0.69 0.58 0.59 1.4 0.6 1.3 1.2 0.6 0.9 1.0 0.6 2.1 2.4 2.4 2.5 2.0 1.8 1.9 1.6 1.3 1.2 0.5 2.5 2.8 1.0 0.9 1.2 1.4 1.4 1.2 1.0 1.2 Alcoho 2.81 1.0 1.41 1.65 1.65 1.0 1.0 0.35 1.0 0.5 1.0 0.72 0.72 2.39 1.53 1.45 1.36 1.0 1.59 1.2 1.2 1.0 1.0 0.49 0.52 0.43 0.9 1.3 1.0 0.60 0.56 1.0 0.7 1.36 1.34 1.0 1.51 1.0 1.0 1.0 95 C 2.29 1.08 1.27 1.35 0.25 0.72 0.24 0.51 0.56 1.72 1.23 1.14 1.05 1.17 0.91 0.87 0.38 0.36 0.32 0.80 0.98 0.45

0.42 0.54 1.01 1.03 1.07 0.74 0.82 3.4 1.8 2.1 2.0 0.5 1.3 1.4 0.9 0.9 3.3 1.9 1.8 1.7 2.1 1.7 1.7 0.6 0.7 0.5 1.2 1.7 0.8 0.7 1.1 1.8 1.7 2.1 1.3 1.4 Othe Drug 1.62 1.0 2.64 3.50 3.08 1.0 1.0 0.54 0.8 0.5 1.0 0.6 1.59 3.31 1.50 1.47 1.3 1.0 2.11 1.4 1.2 1.0 1.0 0.59 0.6 0.44 1.0 1.3 1.0 0.51 0.46 1.0 0.8 1.2 0.8 1.0 1.79 1.92 1.72 1.0 95 1.24 1.28 1.92 1.64 0.37 0.56 0.20 0.46 1.00 2.10 1.04 1.08 0.89 1.35 0.85 0.72 0.41 0.44 0.28 0.75 0.86 0.36 0.33 0.48 0.88 0.63 1.31 1.33 1.18 2.1 5.4 6.3 5.7 0.7 1.3 1.4 1.0 2.5 5.2 2.1 1.9 1.9 3.3 2.4 2.0 0.8 1.0 0.6 1.5 2.1 0.7 0.6 1.3 1.7 1.1 2.4 2.7

2.5 Note. Dashe indicat dat no estimated O = odd ratio C = confidenc interval *p < .05
Page 10
COMPARATIV EPIDEMIOLOG O DRUG 25 denc an bein Africa America o Hispani America als wer foun i multipl logisti regressio analyse tha hel constan employmen an tw primar indicator o socioeconomi sta tus educationa achievemen an incom (dat no show i a table) Employment and Socioeconomic Status stud variatio i occurrenc o dru depen denc b employmen statu a th tim o assess ment individual wh primaril wer workin fo pa hav bee compare wit students homemak ers an other (e.g. thos wh ha bee recentl

lai of o terminate an no ye reemployed othe person n longe i th activ labo force) compariso wit employe workers student generall wer a lowe lifetim prevalenc grou fo dependenc o tobacc (O = 0.45 an alco ho (O = 0.72) bu no fo othe drug suc a marijuana cocaine o inhalant (O = 0.69 > .05) A histor o alcoho dependenc wa observe les frequentl amon homemaker ver su employe worker (O = 0.72) bu homemak er wer somewha mor likel t hav bee affecte b dependenc o tobacc (O = 1.53 an b dependenc o othe drug (O = 1.59) Prevalenc o dependenc o alcohol tobacco an othe drug wa especiall commo amon person

recentl lai of bu no ye reemploye an othe individual no workin i th pai labo forc a th tim o th assessment Compare wit employe workers thes unemploye person wer a estimate 1.8 time mor likel t hav a histor o tobacc dependence 2.3 time mor likel t hav a histor o alcoho dependence an 3.3 time mor likel t hav a histor o depen denc o controlle substance o inhalant (Tabl 3) Thes moderatel stron association betwee unemploymen an dependenc o alcohol to bacco o othe drug als wer foun i multipl logisti regressio analyse tha hel constan age sex education income an a selectio o othe potentiall confoundin

variable (dat no pre sente i a table) Lo educationa achievemen als ha a moder atel stron associatio wit a histor o depen denc o tobacco alcohol o othe drugs wit o withou statistica adjustmen usin th multipl logisti regressio model Fo person wit 0-1 year o schoolin compare wit person wh wen t schoo fo mor tha 1 years th histor tobacc dependenc wa associate wit lowe educationa achievemen (O = 1.68) a wa a histor o alcoho dependenc (O = 1.53 an als a histor o dependenc o othe drug (O = 1.50) A simila profil o modes bu statisticall significan association wa observe whe comparin person wit 1 year o

schoolin thos wit mor tha 1 year (Tabl 3) i interestin t not tha lowe educationa achievemen wa associate wit dependenc o tobacc an alcoho eve amon person wh ha complete mor tha 1 year o schooling Thi ca b see i th odd ratio tha contras person wit 13-1 year o educatio wit thos wh attende schoo fo 1 year o mor (fo 13-1 year vs 16 years O = 1.4 fo tobacc depen dence O = 1.3 fo alcoho dependence) general a lowe annua incom wa associ ate wit havin bee affecte b dru depen denc (Tabl 3) Fo example person earnin les tha $20,00 pe yea wer 2.1 time mor likel hav a histor o dependenc o controlle

substance o inhalant compare wit person whos annua incom wa $70,00 o mor (95 = 1.35 3.31) Fo tobacc dependence th stronges associatio wa observe i th contras betwee person wit income o $20,00 t $34,00 pe yea versu thos wit annua incom o $70,00 o mor (O = 1.43 95 C = 1.05,1.98) Household Composition Mos o th stud populatio ( = 4,344 con siste o marrie person livin wit a spous (wit withou othe famil members) bu a consider abl numbe o respondent ( = 1,198 wer livin alon i thei household (se Tabl 1) Compare wit thos livin alone individual livin wit thei spouse wer jus a likel t hav lifetim tobacc

dependenc (O = 0.87) bu wer les likel t qualif a recen o pas case o alcoho dependenc (O = 0.49 o dependenc controlle drug o inhalant (O = 0.59) a presente i Tabl 3 Invers association als wer observe fo person livin wit thei parent relatio t tobacco alcohol an othe drugs I part thes invers association shoul b under stoo i relatio t sex Withi th stud popula tion a larg majorit o person livin wit parent wer women a show i Tabl 1
Page 11
25 ANTHONY L WARNER AN R KESSLE Marital Status and Religious Preference thi stud population ther wa a tendenc fo a histor o tobacc dependenc t b mor

commo amon person wh wer currentl mar rie o livin wit partner a i marrie (se "married i Tabl 3 an amon formerl mar rie individual (separated divorced o widowed) an les commo amon person wh wer neve married thi wa no th cas fo dependenc o alcoho o othe drug (Tabl 3) O th othe hand ther wer fairl consisten association involvin religiou preference wit a histor o dependenc foun mor frequentl amon per son wh professe n religiou preference an b comparison leas frequentl amon Catholic (O = 0.7 fo tobacco O = 0.5 fo alcohol = 0.4 fo othe drugs) I addition Protes tant ha lowe prevalenc o dependenc o

alcoho (O = 0.60 an controlle substance o inhalant (O = 0.51 i a contras t person wit n religiou preference bu th odd rati estimat fo tobacc dependenc amon Protes tant wa close t th nul valu o 1. an th associatio wa no statisticall significan b con ventiona standard (O = 0.77; > .05) Location of Residence Fo dependenc o controlle substance o inhalants ther wer nonrando distribution i relatio t bot regio o th countr an urban nonurba locatio o residence Compare wit resident o state i th South individual livin th Wes wer foun t b a estimate 1.7 time mor likel t hav develope dependenc thes drugs

Resident o metropolita area wer mos likel t hav a histor o dependenc controlle substance o inhalant (O = 1.92) followe b resident o othe urba area (O = 1.72) an resident o nonurba area wer leas likel t hav bee dru dependent Alcoho dependenc als wa associate wit regio o th country bu no wit metropolita o othe urba environments I a compariso wit resident o th South th odd o alcoho depen denc wer abou 50 greate amon person livin i th Wes (O =1.51 an abou 35 greate amon person livin i th Northeas (O = 1.36 o i th Midwes (O = 1.34) contras wit dependenc o alcoho o othe drugs tobacc dependenc wa

distribute essen tiall a rando i relatio t regio o th countr an urbanicity Al o th tobacc odd ratio correspondin t regio an urba an nonurba residenc wer clos t th nul valu o 1. an th associate confidenc interval ha span fro belo 1. t abov 1. (Tabl 3) Drug Use and the Transition to Drug Dependence Lifetim prevalenc proportion fo dru depen denc i th stud populatio ar determine i par b ho man person hav trie eac typ o dru a leas onc an survive t b interviewed an i par b ho man o thes survivin dru user ha proceede t becom dru dependent Fo example considerabl fewe member o th stud populatio ha trie

tobacc tha alcoho (75.6 lifetim prevalenc fo tobacc us vs 91.5 fo alcoho use a show i Tabl 2 colum 4) However dependenc wa mor likel t occu amon tobacc smoker tha amon alcoho drink ers O th tobacc smokers 31.9 ha develope tobacc dependence wherea onl 15.4 o th alcoho drinker ha develope alcoho depen denc (Tabl 2 colum 6) I consequence withi th tota stud population th lifetim prevalenc tobacc dependenc (24.1% wa greate tha th lifetim prevalenc o alcoho dependenc (14.1%) eve thoug mor person ha consume alcoho (91.5% tha ha smoke tobacc (75.6%) Whe controlle substance an inhalant wer considere a

a single group th dat showe tha slightl mor tha on hal o th stud populatio ha take on o mor o thes drug fo extramedi ca reason (51.0% an 14.7 o th user ha develope dependenc o a leas on o th liste drug (Tabl 2 ro 3 column 4 an 6) Nonethe less on migh expec considerabl variatio i th prevalenc o extramedica dru us an i th transitio fro dru us t dru dependence acros individua drug an dru groups Afte alcoho an tobacco cannabi wa th nex mos frequentl use dru liste i Tabl 2 bu i ranke lo i relatio t ou inde o dependenc amon user (Tabl 2 colum 6) Withi th stud population a estimate 46.3 ha use

cannabi a leas once bu onl 9.1 o th user ha develope cannabi dependence Fo ever use wit a histor o cannabi depen
Page 12
COMPARATIV EPIDEMIOLOG O DRUG 25 dence ther wer 1 user wh ha no becom dependent B comparison a estimate 16.2 o th stud populatio ha trie cocain a leas once an 16.7 o the ha qualifie a cocain dependent Fo eac cocain use wit a histor o cocain dependence ther wer 5 user wh ha no becom dependen (Tabl 2 column 4 an 6) estimate 15.3 ha use psychostimulant othe tha cocain (e.g. amphetamines fo extra medica reasons an 11.2 o thes user ha progresse t develo dependenc o thes

drugs Correspondin estimate fo th anxiolytic seda tive o hypnoti drug wer 12.7 an 9.2% respectively A estimate 9.7 o th stud popu latio ha use analgesi drug fo extramedica reasons 7.5 o thes analgesi user ha becom dependen (Tabl 3 column 4 an 6) estimate 10.6 o th 15-54-year-ol stud populatio reporte usin psychedeli drug a leas once 6.8 reporte us o inhalants an 1.5 reporte usin heroin A estimate 4.9 o th psychedelic user qualifie fo th depen denc diagnosis Amon inhalan users a esti mate 3.7 qualifie a dependent B compari son amon heroi user i thi sample 23.1 ha becom dependen (Tabl 2) Age and

Drug Dependence by Drug Group som extent age-associate variatio i dru dependenc tha wa observe i ou logisti regressio analyse shoul b understoo i rela tio t difference i prevalenc o dru use, i additio t othe factors Fo example a histor o tobacc us wa les commo amon 15-24-year old a compare wit olde ag group (se Tabl an Figur 1) thi b itsel migh b sufficien t accoun fo lowe lifetim prevalenc o tobacc dependenc i th compariso o youn versu olde persons However th NC dat als high lighte th importanc o age-relate difference th transitio fro tobacc us t tobacc dependence I analyse tha considere

smoker only w foun tha 15-24-year-ol smoker wer les likel tha olde smoker t hav develope tobacc dependenc (Tabl 4 an Figur 1) Fo tw group o medicall prescribe drugs namely th analgesic an th anxiolytic sedative an hypnoti drugs th survey-base estimate fo lifetim prevalenc o dependenc wer highe fo olde ag group versu th younges ag group despit generall lowe prevalenc o extramedica us amon olde adults Thi wa tru fo lifetim prevalenc o dependenc amon extramedica user o thes drug a wel a fo lifetim preva lenc o dependenc amon al person (Tabl 4 an Figur 1) substantiall differen patter wa observe fo

marijuana cocaine psychostimulant othe tha cocaine an th psychedeli dru group whic showe comparativel lowe lifetim preva lenc o dependenc i th oldes ag group alon wit generall lowe lifetim prevalenc o extra medica dru us (Tabl 4 an Figur 1) Alcoho offere a uniqu profile wit 45-54 year-old havin a relativel hig lifetim preva lenc o alcoho us (93.1% bu a relativel lo prevalenc o alcoho dependenc (10.1% a com pare wit th othe ag groups Alcoho als i noteworth becaus th lifetim prevalenc o dependenc amon drinker wa abou 10 fo person 45-5 year o age considerabl les tha estimate o abou 16 tha wer

observe fo al thre younge ag group (Tabl 4 an Figur 1) Drug Dependence by Drug Group and by Sex relatio t lifetim prevalenc o dru depen denc an lifetim prevalenc o extramedica use th ran orderin o individua drug an dru group wa generall identica fo me an wome (Tabl 5 an Figure 2 an 3) Th exceptio ca see i relatio t a se differenc i th rankin o psychedeli drugs whic ha bee take b 14.1 o me a compare wit 7.2 o women respec t 6 ou o 1 individua drug o dru groups th lifetim prevalenc estimate fo depen denc amon user wer roughl simila fo me an women Fo example slightl mor tha 30 th tobacc smoker ha

develope tobacc dependence an slightl mor tha 22 o th heroi user ha develope heroi dependence regardles o sex Minima male-femal difference th prevalenc o dependenc amon user wer observe fo cocaine analgesics hallucinogens an inhalant (Tabl 5 an Figur 4) contrast fo alcoho an cannabis mal user wer somewha mor likel tha femal user t hav develope dependenc (Tabl 5 an Figur 4) Furthermore wit respec t th dru grou tha include anxiolytics sedatives an hypnotics
Page 13
25 ANTHONY L WARNER AN R KESSLER Tabl 4 Estimated Prevalence Proportion (P) of Extramedical Use and Dependence in Total

Study Population and Lifetime Dependence Among Users by Age Proportio wit histor o extramedica us Dru an ag (years Tobacco 15-2 25-3 35-4 + Tota Alcoho 15-2 25-3 35-4 45 Tota Cannabi 15-2 25-3 35-4 + Tota Cocain 15-2 25-3 35-4 + Tota Stimulant 15-2 25-3 35-4 + Tota Anxiolytics etc. 15-2 25-3 35-4 + Tota Analgesic 15-2 25-3 35-4 + Tota Psychedelic 15-2 25-3 35-4 + Tota 64. 76. 80. 82. 75. 82. 95. 94. 93. 91. 36. 61. 52. 25. 46. 10. 26. 17. 4. 16. 11. 20. 18. 7. 15. 8. 16. 16. 8. 12. 10. 12. 9. 5. 9. 8. 14. 12. 3. 10. SE 1. 0. 1. 1. 0. 1. 0. 0. 0. 0. 2. 1. 1. 1. 1. 1. 1. 1. 0. 0. 1. 1. 1. 0. 0.

0. 1. 0. 1. 0. 1. 0. 0. 0. 0. 0. 1. 1. 0. 0. Proportio wit histor o dependenc 15. 26. 27. 27. 24. 13. 15. 15. 10. 14. 5. 5. 4. 0. 4. 2. 4. 2. 0. 2. 1. 2. 1. 0. 1. 0. 1. 1. 1. 1. 0. 0. 1. 0. 0. 0. 0. 0. 0.0 0. SE 1. 2. 1. 2. 1. 1. 1. 1. 1. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0.0 0. Dependenc amon user 23. 34. 34. 32. 31. 16. 16. 16. 10. 15. 15. 8. 8. 3. 9. 24. 15. 15. 11. 16. 13. 13. 7. 6. 11. 2. 6. 11. 20. 9. 1. 6. 11. 16. 7. 8. 4. 3. 0. 4. SE 1. 1. 1. 1. 0. 2. 0. 1. 1. 0. 4. 1. 2. 6. 1. 2. 3. 1. 2. 1. 1. 2. 2. 6. 1. 0. 1. 2. 7. 1. 2. 1. 1. 0.

0.
Page 14
COMPARATIV EPIDEMIOLOG O DRUG 25 Tabl 4 (continued) Proportio wit histor o extramedica us Dru an ag (years Heroi 15-2 25-3 35-4 + Total Inhalant 15-2 25-3 35-4 + Total An dru group 15-2 25-3 35-4 + Tota 0. 1. 2. 0. 1. 8. 9. 5. 1. 6. 42. 64. 56. 31. 51. SE 0. 0. 0. 0. 0. 1. 0. 0. 0. 0. 1. 1. 1. 1. 1. Proportio wit histor o dependenc 0. 0. 0. 0. 0. 0. 0. 0. 0.0 0. 7. 9. 8. 2. 7. SE 0. 0. 0. 0.0 0. 0. 0. 0. 0.0 0. 1. 0. 0. 0. 0. Dependenc amon users 20. 15. 31. 10. 23. 7. 1. 2. 2. 3. 17. 14. 14. 9. 14. SE 10. 6. 10. 7. 5. 3. 0. 1. 2. 1. 2. 1. 1. 2. 0. Note. Weighte estimate

fro th Nationa Comorbidit Surve dat gathere i 1990-199 fo person 15-54-years-ol (n = 8,098) Dashe indicat dat no estimated Tobacc us estimate fro th 199 Nationa Househol Surve o Dru Use a describe i th Assessment o Alcohol an Other Drug Us section n = 4,414 Dependenc amon user estimate th algebrai metho describe i th Analysis Procedures section standar error no estimated Anxiolytics sedatives an hypnoti drugs grouped "An dru group refer t th aggregat categor comprisin th controlle substance an inhalan drugs bu no alcoho o tobacco femal user wer somewha mor likel tha mal user t hav develope

dependence Fo ex ample a estimate 14 o me (±0.8% an 11.5 o wome (±0.8% reporte extramedica us o a leas on anxiolytic sedative o hypnoti drug Amon thes extramedica users a esti mate 6.6 o th me (±1.0% an 12.3 o th wome (±2.2% ha develope dependenc o thi clas o drug (Tabl 5) Discussio Comparison of Prevalence Estimates th basi o genera consensus a DSM III- tas panel o dru dependenc decide t modif th Edwards-Gros alcoho dependenc concep an t adop thi modificatio a a unifie construc tha woul appl t al psychoactiv drug (Koste & Kosten 1991 Kosten Rounsa ville Babour Spitzer & Williams 1987) A a

resul o thi development fo th firs tim i a epidemiologi fiel surve w hav bee abl t hol constan th diagnosti criteri fo dru dependenc a the ar applie t user o to bacco alcohol controlle substances an inhal ant an t reduc marke between-dru differ ence i diagnosti assessment fo dru dependence I doin so w foun tha a substan tia proportio o th 15-54-year-ol populatio i th Unite States—abou 1 i 1 person o 7.5%—ha bee affecte b dependenc o con trolle substance o inhalants B comparison almos twic a man ha develope alcoho depen dence abou 14% an abou thre time a many 24.1% ha develope tobacc dependenc

a som tim i lif u t th tim o th survey Th exten t whic th nation' healt i compromise thi histor o dru dependenc i likel t b a topi o investigatio fo man years plac th potentia healt an socia burden dru dependenc i contex wit th burde o othe psychiatri morbidity i ma b usefu t
Page 15
25 ANTHONY L WARNER AN R KESSLE compar lifetim prevalenc estimate fo depen denc o individua drug wit th lifetim preva lenc o selecte DSM-III- anxiet an moo disorder ascertaine b th CID method whic ou researc grou ha reporte i a separat publicatio (Kessle e al. 1994) T illustrate fo thi articl w hav

estimate tha cannabi depen denc ha affecte 4.2 o th stud population Thi place th prevalenc o cannabi depen denc i ran betwee pani disorde (3.5 preva lence an generalize anxiet disorde (5.1 prevalence o agoraphobi withou pani disorde (5.3 prevalence) Comparatively a histor o cocain dependenc wa foun i 2.7 o th stud population Co cain dependenc wa almos a prevalen a antisocia personalit disorde (3.5 prevalence) an wa abou 70 mor commo tha bipola disorde (1.6 prevalence) Alcoho dependenc (14.1% wa somewha mor commo tha simpl phobi (11.3% an socia phobi (13.3%) wherea tobacc dependenc a 24.1 wa

mor prevalen tha majo depressiv disorde a 17.1 (Kessle al. 1994) Checke agains EC estimate fo lifetim prevalenc o dru dependence th NC value generall wer similar T illustrate EC versu NC estimate fo cannabi disorder bot wer clos t 4 (4.4 vs 4.2% respectively) fo psychostimulant othe tha cocaine bot esti mate wer 1.7% fo th sedativ dru group bot estimate wer 1.2 (Anthon & Helzer 1991) On noteworth exceptio involve cocaine i tha les tha 1 o th EC stud populatio qualifie fo a cocain disorder wherea 2.7 o th NC stud populatio di so I part thi differenc ca b attribute t a chang i diagnos ti

criteria DSM-II criteri use fo th EC covere cocain abuse bu di no provid fo th diagnosi o cocain dependenc (Anthon & Trinkoff 1989) I addition th mos recen epi demi o cocain dependenc i th Unite State als ca accoun fo som o th observe increase tes whethe th NC estimate fo lifetim prevalenc o alcoho us an illici dru us wer consisten wit thos fro th NHSD conducte 1991 w reanalyze th NHSD dat t con for wit th ag an se group create fo th NC analyse withi th ag rang fro 1 t 5 years Thi seeme t b especiall necessar i ligh o observation abou th importanc o privac an self-administratio strategie i

surve researc abou illici dru us (e.g. se Schobe e al. 1992) Nonetheless i man instances th NC observe estimate wer slightl highe tha correspondin NHSD estimates particularl amon olde adults i others the wer n mor tha slightl lower particularl amon 15-24-year olds However fo eac comparison th NHSD estimat wa wel withi th rang define b 95 fo th correspondin NC estimat (dat no show i a table) Ther ar n prio nationa surve estimate fo tobacc dependence bu a recen populatio sur ve o youn adult i th Detroit Michiga are foun tha 74 o 21-30-year-old ha smoke tobacc a leas once amon thes smokers 27 ha

develope DSM-III- tobacc dependenc (Breslau Fenn & Peterson 1993) B comparison estimate fro th nationall representativ NC fo roughl th sam ag grou wer no to different O th 25-34-year-olds 76.4 ha smoke an 34.7 ha develope DSM-III- tobacc dependence Comparison With Prevalence Estimates for Other Nations b sure ther ar som countrie i whic th lifetim prevalenc o alcoho dependenc o tobacc dependenc appea t equa o excee th value observe i th NCS mos notabl Sout Kore (e.g. se Le e al. 1990a 1990b) How ever w ar awar o n epidemiologi evidenc countrie i whic prevalenc o dependenc controlle substance

exceed th value w foun fo 15-54-year-ol noninstitutionalize resi dent o th coterminou Unite State (e.g. se Bland Orn & Newman 1988 Canin e al. 1987 Compto e al. 1991 Hwu Yeh & Chang 1989 Le e al. 1990a Wells Bushnell Hornblow Joyce & Oakley-Browne 1989 Wittchen Essau vo Zerssen Krieg & Zaudig 1992) Comparative Epidemiology of Tobacco, Alcohol, and Other Drug Dependence Epidemiologi analyse o populatio dat ofte begi wit a consideratio o birt yea o age sex an race Thi approac draw o a lon traditio includin thes variable o birt record an
Page 16
COMPARATIV EPIDEMIOLOG O DRUG 25 Tabl 5

Estimated Prevalence Proportion (P) of Extramedical Use and Dependence in Total Study Population and Lifetime Dependence Among Users by Sex Proportio wit histor o extramedica us Dru Tobacco Alcoho Cannabi Cocain Stimulant Anxiolytics etc. Analgesic Psychedelic Heroi Inhalant Dru group 78. 93. 51. 19. 18. 14. 11. 14. 2. 9. 55. SE 0. 0. 1. 0. 0. 0. 0. 0. 0. 0. 1. Me Wome Proportio wit histor o dependenc 25. 20. 6. 3. 1. 1. 0. 0. 0. 0. 9. SE 1. 1. 0. 0. 0. 0. 0. 0. 0. 0. 0. Dependenc amon mal user 32. 21. 12. 18. 9. 6. 6. 5. 22. 4. 16. SE 1. 1. 1. 1. 1. 1. 1. 6. 1. 1. Proportio wit histor o

extramedica us 73. 89. 41. 12. 12. 11. 7. 7. 0. 4. 46. SE 0. 0. 1. 0. 0. 0. 0. 0. 0. 0. 1. Proportio wit histor o dependenc 22. 8. 2. 1. 1. 1. 0. 0. 0. 0. 5. SE 1. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. Dependenc amon femal user 30. 9. 5. 14. 13. 12. 8. 4. 25. 2. 12. SE 0. 0. 2. 2. 2. 1. 1. 12. 2. 1. Note. Weighte estimate fro th Nationa Comorbidit Surve dat gathere i 1990-199 fo person 15-54-years-ol (n = 8,098) Dashe indicat dat no estimated Tobacc us estimate fro th 199 Nationa Househol Surve o Dru Use a describe i th Assessment o Alcohol an Other Drug Us section n = 4,414 Dependenc amon user

estimate b th algebrai metho describe i th Analysis Procedures section standar error no estimated Anxio!ytics sedatives an hypnoti drugs grouped "Dru group refer t th aggregat categor comprisin th controlle substance an inhalan drugs bu no alcoho o tobacco deat certificates whic i man place hav bee th mai source o epidemiologi data I addi tion i generall i necessar t tak thes thre inbor variable int accoun befor assessin evidenc abou potentiall modifiabl ris factor condition tha ca b leverage fo preventio an contro o disease Moreover birt year sex an rac ar fixe characteristic tha canno b

modifie b change i condition suc a diseas experienc o toxi exposures contrast whe w tur t th epidemiologi stud o potentiall modifiabl condition suc a educationa attainment employmen status an acces t materia wealth w fac th possibilit tha thes characteristic migh influenc th occur renc o diseas an als migh b influence b disease I i i thi contex tha th limitation o cross-sectiona surve dat an th advantage o prospectiv o longitudina dat becom mos apparent Reasonin alon thes lines w cautio agains th causa interpretatio o cross-sectiona surve results an w not tha lifetim preva lenc ratio o odd

ratio fro lifetim prevalenc analyse typicall canno b interprete a ris ratio o relativ ris estimate (Kramer Vo Korff & Kessler 1980) Eve thoug the shoul no b give causa interpretation th cross-sectiona association an odd ratio o th typ estimate fo thi stud ca poin ou segment o th populatio wher cur rentl activ o forme case o dru dependenc ar mor o les likel t b found I additio t highlightin th locatio o thes case withi th population cross-sectiona odd ratio serv t inde th strengt o eac observe association whethe causa o noncausal Togethe wit ou basi prevalenc estimates th observe pattern

associatio (o lac thereof represen th beginnin step fo a comparativ epidemiolog o tobacc dependence alcoho dependence an dependenc o controlle substances Scannin acros odd rati estimate fro ou analyses on ca fin man commonalitie an som noteworth difference i th association wit dependenc o tobacco alcohol an othe drugs Man o th observe association ar consisten wit prio research recentl summa rize b u an b other (e.g. Anthony 1991
Page 17
26 ANTHONY L WARNER AN R KESSLE ° * ° ^ - ° •14--9 •frE-9 •fS-91 •^3-91 tz-s t-g-gt w-g t-E-9 •tt-g 9S 1 90Ue|BA8J (% sjes Buoiu aauepuede j

eou9|BA9J
Page 18
COMPARATIV EPIDEMIOLOG O DRUG 26 Anthon & Helzer 1991 Anthon & Helzer i press Hawkins Catalano & Miller 1992 Kandel 1991) Fo example se (bein a man ha a moderat degre o associatio wit alcoho depen denc an othe dru dependenc bu no wit tobacc dependence Thi findin converge wit othe evidenc showin a increasin healt bur de o tobacc smokin amon women an per hap a declinin burde amon me (e.g. Adams Gfroerer & Rouse 1989 Johnsto e al. 1992) Anothe comparativ differenc wa observe i th association wit age especiall i th contras 45-54-year-old wit 15-24-year-olds Th 45-54

year-old wer mor likel t hav a histor o curren o pas tobacc dependence the wer les likel t hav a histor o dependenc o alcoho or on othe drugs However age-specifi preva lenc estimate fo individua controlle sub stance promp a slightl differen conclusio abou tw classe o drug tha ar availabl b prescrip tion Namely compare wit 15-24-year-olds th oldes adul extramedica dru user unde stud wer mor likel t hav becom dependen o analgesi drug an o anxiolytic sedative o hypnoti drugs whic ar use widel fo legiti mat medica reasons Thi aspec o th 45-54 year-ol population' dru experienc differ fro it

experienc wit alcohol Schedul I drug suc heroin an inhalan drugs Thes relationship betwee ag an dru depen dence generall consisten wit prio surve find ing i th Unite States ma b understandabl a reflectio o variatio i dru availabilit o leve o dru involvemen acros differen birt cohorts periods o segment o th lif spa (Kandel 1991) Alternately ther ar tw genera epidemiologic observation tha ma clarif under lyin issue o interpretation Th firs genera observatio involve th forc drug-relate mortality whic ca hav a impor tan bu nearl hidde impac o age-specifi estimate fro cross-sectiona assessments A

th tim of assessment eac age grou unde stud consist o survivor fro on o mor origina birt cohort (e.g. liv birth i a give year o i a give spa o years) A som point mortalit associate wit dru dependenc start t contrib ut t th los o drug-dependen person fro eac birt cohort Althoug th forc o drug relate mortalit ca b i operatio fo al age o dru users accumulate attritio du t depen dence-relate death become a increasingl im portan facto i successiv year o adulthood thi respect alcoho provide a goo illustra tion A show i Figur 1 th lowe lifetim prevalenc o dependenc observe amon th oldes drinker migh

b du partl t a prematur mortalit tha i secondar t alcoho dependence Ther als wa a shar dro i prevalenc o heroi dependenc acros th tw oldes ag group surveyed 2.7 fo 35-44-year-old versu 0.7 fo 45-54-year-olds Amon 2,24 partici pant ag 35-4 year i th NC sample ther wer 6 heroi users bu amon 1,46 participant ag 45-5 years ther wer onl 1 heroi users Base o th experienc o thes heroi users w wer abl t estimat tha almos on thir o th 35-44-year-ol heroi user i th stud popula Figure 1 (opposite). Drug-specifi estimate fo th lifetim prevalenc o dru dependenc (firs ro o graphs) lifetim prevalenc o

extramedica dru us (secon ro o graphs) an lifetim prevalenc o dru dependenc amon extramedica dru user (thir ro o graphs) base o estimate presente i Tabl 4 Weighte estimate base o standardize assessmen o 8,09 American 15-54-years-old wh wer selecte b probabilit samplin an interviewe fo th Nationa Comorbidit Survey 1990-1992 Th scal fo individua graph ha bee tailore fo eac se o estimates wit variatio acros row an withi rows Erro bar hav bee use t sho th precisio (standar error o eac estimat wheneve possible N erro ba ha bee draw whe th standar erro wa quit smal (i.e. hal th heigh o th diamon

symbo use t depic th poin estimate o i th cas o tobacc dependenc amon user (se Analysis Procedures section) AS = anxiolytic sedative an hypnoti drugs groupe (e.g. secobarbita an diazepam a wel a mor recentl introduce compound suc a flurazepam alprazolam an triazolam) "An dru group = aggregat categor comprisin th controlle substance an inhalan drugs bu no alcoho o tobacco t = chang i scale
Page 19
26 ANTHONY L WARNER AN R KESSLE Tobacc Alcoho Cannabi Cocaine includin crac Stimulant othe tha cocain Anxiolytics sedatives hypnotic Analgesi drug Psychedeli drug Heroi Inhalant Male ain ^

pnotic Q 0 Female . . . o 3 Percentag Figure 2 Estimate lifetim prevalenc o dru depen dence b sex base o estimate presente i Tabl 5 tio ha a histor o heroi dependence bu w ha to fe user t produc a correspondin estimat fo th 45-54-year-ol heroi users I th contex o thi discussion i i informativ tha withi thi grou o 1 peopl wh wer 45-54 years-ol an wh ha use heroi a som poin thei lives ther wa onl on use wit a histor heroi dependence W speculat tha prema tur mortalit associate wit heroi dependenc migh accoun fo thes shar difference i th observe heroi experienc o thes tw olde adul ag group withi

ou stud population Th secon genera epidemiologi observatio tha merit consideratio whe studyin ag an dru dependenc relate curren ag t th transi tion fro adolescenc throug ag 34 whic recen evidenc peg a th mai perio o ris fo initiatin dru us an dru dependenc (e.g. se Anthony 1991 Kandel Murphy & Karus 1985) th tim o assessmen i 1990-1992 th 15-24-year-old ha jus starte t pas throug thi mai ris period Th importanc o thi fac migh b see mos clearl i relatio t th analgesic an th anxiolytic-sedative-hypnoti dru group Fo thes drugs th proportio o extra medica user ag 15-2 year wh ha develope histor o

dependenc wa low relativ t th othe ag group (Figur 1) Thi ma reflec tha thes birt cohort hav jus entere th high-ris period an wit passin time thei experienc ma prov t b mor lik th experienc o thei elders relatio t thes NC findings tw othe association deserv specia comment African Americans an drug use. Ther i a widesprea popula belie tha Africa American ar especiall vulnerabl t dru dependence perhap becaus o thei overrepresentatio i certai clinica samples However consisten wit evidenc fro othe recen epidemiologi surveys th NC estimate indicat tha tobacc depen dence alcoho dependence an

dependenc o othe drug ar mor commo amon Whit non Hispani American tha amon Africa Ameri can (e.g. se Anthon & Helzer 1991 Kandel 1991 Lillie-Blanton Anthony & Schuster 1993) Residence i a nonurban area. Alcoho an tobacc ar widel availabl throughou th Unite States bu th illici availabilit o controlle substance seem t var considerably Perhap reflectin thes differen pattern o availability resident o nonurba area wer a lowe preva lenc grou fo dependenc o controlle sub stance bu no fo dependenc o tobacc o alcoho (se Tabl 3) Analogously th EC surve i th Durham-Piedmon populatio o Nort Carolin foun

somewha lowe prevalenc dru disorder amon inhabitant o rura area compare wit thos livin i urba area (An thon & Helzer 1991) Male Female Alcoho O Cannabi < Cocaine includin crac Stimulant othe tha cocain Anxiolytics sedatives hypnotic Analgesi drug Psychedeli drug Heroi Inhalant 10 O 10 Percentag Figure 3 Estimate lifetim prevalenc o extramedi ca dru use b sex base o estimate presente i Tabl 5
Page 20
COMPARATIV EPIDEMIOLOG O DRUG 26 Male Tobacc O Heroi Alcoho Cocaine includin crac Cannabi Stimulant othe tha cocain Anxiolytics sedatives hypnotic Analgesi drug Psychedeli drug Inhalant

Female - Percentag Figure 4 Estimate lifetim prevalenc o dru depen denc amon extramedica dru users b sex base o estimate presente i Tabl 5 Transition From Drug Use to Drug Dependence Whe EG result wer published som read er wer surprise t fin tha man o th individu al wh ha use heroi o othe controlle substance o mor tha fiv occasion ha no develope dru dependenc o othe dru disor der eve thoug prio researc o Vietna veter an ha indicate a much Fo example amon heroi user i th EG sample onl 44 ha becom a cas o heroi dependenc o heroi abus a define b DSM-II criteria Th corre spondin estimate wer clos

t 20 fo extra medica user o cannabis psychostimulant othe tha cocaine an anxiolytic o sedative-hypnoti drug (Anthon & Trinkoff 1989) Similarly Hel ze (1985 an Robin (1993 hav reporte tha mos Vietna veteran wh use heroi an othe opioi drug i Vietna di no becom depen den takin thes drug whil overseas Althoug no directl comparabl t thes EG estimate base o th experienc o individual wh ha use drug o mor tha fiv occasions th NC estimate als sugges tha a larg major it o person wh hav initiate extramedica dru us hav no proceede t develo dru depen dence Eve fo drug know fo thei dependenc liabilit (e.g.

tobacco cocaine heroin) th propor tio o dru user wh wer foun t hav becom dependen wa i a rang fro 20-40% Consider in thes prevalenc estimate o th transitio fro dru us t dru dependence i i noteworth tha NC interviewer wer abl t develo trus an rappor sufficien t elici report o illega drug-takin behavio fro man participants W speculat tha man participant wh acknowledg pas illici behavio als ma b willin t repor persona problem an othe symptom o dru dependenc tha the hav experienced A th sam time w acknowledg a stron possibilit tha som drug-dependen participant di no repor o thes problem wit

completenes o tota accuracy whic woul ten t attenuat th proportio o dependen person amon extramedi ca dru users conceptua grounds i ca b argue tha th transitio fro dru us t dru dependenc i th populatio i determine partl b th reinforc in function serve b drug-takin an associate behaviors wit a linkag bac t profile o dru activit discovere throug laborator researc (e.g. se Schuster 1989 Thompson 1981) I addition thi transitio seem t b determine i par b othe factors som linke t th reinforc in function o dru use an som no s readil studie insid th laborator (e.g. se Brady 1989 Glant & Pickens 1992

Schuster 1989) I theory th arra o thes interrelate factor include relativ dru availabilit an opportunitie fo us differen drug a wel a thei cost pattern an frequencie o dru us tha diffe acros drugs differen profile o vulnerabilitie o individual whos extramedica us start wit on dru versu another a wel a bot forma an informa socia control an sanction agains dru us o i it favor whic migh b exercise eithe withi intimat socia field suc a th famil o work plac o b large unit o socia organization Considere al together th arra o theoreticall plausibl determinant o th transitio fro dru us t dru dependenc

run a spa fro th microscopi (e.g. th dopamin receptor throug th macroscopi (e.g. socia norm fo o agains dru use internationa drug-contro policies) bette understandin o thes source o variation a wel a attentio t methodologica feature o cross-sectiona surve research woul hel u accoun fo th ran orderin o individua drug an dru group i relatio t th prevalenc extramedica dru us an i relatio t th proportio o extramedica user wh wer foun hav develope dependence I i o consider
Page 21
26 ANTHONY L WARNER AN R KESSLE abl interes tha th ran orderin b prevalenc extramedica dru us wa quit simila

fo bot me an women differin onl b th highe prevalenc o psychedeli dru us amon men Th ran orderin i relatio t transition fro dru us t dru dependenc wa no th sam a tha see fo prevalenc o extramedica dru use addition th rankin o drug i relatio t th use-to-dependenc transitio showe som varia tio acros mal an femal dru user an acros ag groups Fo bot me an women an fo al bu th oldes ag grou o dru users tobacc an heroi wer to ranked psychedeli drug an inhalant wer a th bottom Ther wer male femal difference i lifetim prevalenc o depen denc amon extramedica dru user onl fo alcoho an cannabis A estimate

21.4 o th mal alcoho drinker ha develope dependence compare wit a estimate 9.2 o femal drink ers Correspondin mal an femal estimate fo cannabi wer 12 an 5.5% respectively I i noteworth tha alcoho an cannabi ha highe ran amon men wherea th anxiolytics-seda tives-hypnoti dru grou wa highe ranke fo wome (Tabl 5) Notwithstandin thes genera observations som attentio shoul b give t th fac tha 15-24-year-ol dru user ha a comparativel hig lifetim prevalenc o dru dependenc i connectio wit cocain an alcoho an wit Schedul I drug suc a marijuana T illustrate fo cocaine almos 25 o th 15-24-year-ol user ha

develope dependence B comparison onl 15 o th 25^4-year-old cocain user ha develope dependence T th exten tha th 15-24-year-old hav man remainin year a risk thei alread hig valu ma becom eve higher bu fo a possibl compensatin influence Namely ha bee observe tha earl onse illici dru user (e.g. thos wh initiat illici dru us befor age 17) see to be at increase ris for developin dru problem compare wit dru user wit a late star (e.g. afte ag 17) eve whe statistica adjustment ar mad fo difference i duratio dru us (e.g. se Anthon & Petronis 1993 Robin & Przybeck 1985) I follow tha th cumulativ

occurrenc o dru dependenc a firs migh b especiall hig amon 15-24-year-ol dru users bu the migh declin a th lowe ris experienc o late onse dru user i adde t thi birt cohort' tota dru experience Thi i a empirica questio tha deserve continue study includin futur analyse o th NC dat o th experienc o individua birt cohort bor sinc 1935 Future Directions for Research and Other Implications Futur direction fo researc o thes topic ca b guide b carefu consideratio o th presen study' deficiencies Foremos amon thes limitation i a cross-sectiona stud desig tha ha place heav relianc o retrospectiv self

repor methods constrainin scientifi inferenc abou ris an ris factor i relatio t dru dependence Give unbounde resources w woul hav like t mak a prospectiv investiga tio o eac drug' users startin wel befor dru us ha begun wit periodi observation sus taine throug period o ris fo dru depen denc an othe drug-relate hazards includin th ris o drug-induce death Instead fo cos containment a i th EC program th annua NHSDA an othe large-scal epidemiologi sur veys w hav sample th population' experienc cross-sectionall an hav measure retrospec tively Thi approac leave ou th experience o peopl wh hav died a

wel a thos wh faile recal an repor thei dru involvemen wit accuracy Thus i i usefu t remembe tha o on sid lifetim prevalenc estimate base o self-report ar hemme i b th seriousnes o death o th othe sid b long-forgotte o casua dru involvemen tha doesn' see wort mention in a th tim o assessment T th exten tha som condition ma b associate wit consider abl ris o deat (e.g. heroi dependenc i th Unite States) thi commonl use stud desig actuall ca yiel a undercoun o morbidity Man seriousl affecte person hav died T th exten tha othe condition ma b regarde a inconsequentia an pointles t mentio (e.g.

takin a puf o someon else' marijuan ciga rett withou inhaling) th sam approac ca yiel a overcoun o morbidity Man mildl affecte person ar neglected note previousl i thi section cross sectiona surve design als ca lea t misinter prete tim relationships Fo example i thi study i appear tha lo educationa achievemen migh signa a increase ris o alcoho depen
Page 22
COMPARATIV EPIDEMIOLOG O DRUG 26 dence I fact th cross-sectiona evidenc o thi stud doe no clarif whethe alcoho dependenc a ris facto fo lo educationa achievement whethe lo educatio i a ris facto fo alcoho dependence whethe th

relationshi i recipro cal o whethe unmeasure antecedent migh explai th observe association betwee educa tio an alcoho dependence I thi instance w ar fortunat t hav recentl publishe evidenc fro a prospectiv study whic foun exces ris alcoho disorder amon adult wh ha no receive hig schoo o colleg diploma compare wit colleg graduate (Crum Bucholz Helzer & Anthony 1992 Crum Helzer & Anthony 1993) course whethe cross-sectiona o prospective th evidenc fro a singl observationa stud doe no alway lea t clea inference abou caus an effect ou stud i n exception A i experimenta research systemati replicatio

i essential an ultimatel causa inference mus b base o judgment abou th availabl evidence Limitation o mor secondar importanc in clud a restrictio o th samplin fram t nonin stitutionalize resident wh coul b sample fro identifie dwellin units Anthon an Trinkof (1989 an Anthon an Helze (1991 hav discusse an demonstrate ho overal prevalenc rate fo dru dependenc an relate condition chang ver littl whe institutiona resident (an b extension th homeless ar include withi th surve samplin frame How ever, i shoul b acknowledge tha thi overal generalizatio migh no hol fo populatio sub group wit especiall

hig rate o drug-relate incarceratio (e.g. youn me o African-America heritage) Curren interes i hai analysi an othe bioas say fo illici dru us raise a legitimat questio abou intervie assessment o dru dependenc (e.g. se Kidwell 1992) Denne i term o DSM-III- diagnosti criteri an cas defini tions dru dependenc i a psychiatri disturbanc fo whic recen dru us i bu on indicator Excep i unusua circumstances thes diagnosti criteri fo dru dependenc ca b assesse onl mean o interview o examinations eithe wit designate respondent themselve o wit informant fo thes respondents Give th siz o th NC sampl an

restricte resources i wa no possibl t intervie informants Agains thi backgroun o stud limitations i i importan t focu o th tw epidemiologi ques tion fo whic thi typ o cross-sectiona fiel stud i indispensable (a I th population wha proportio o person ha bee affecte b dru dependence an (b Comparin subgroups wher th populatio ar case o dru dependenc mor likel t b found Althoug th answer t thes question d no constitut definitiv evi denc wit regar t caus an effec relationship mechanism o action thes answer hav a importan publi healt value A mentione i ou introduction thes answer ca b o us t member

o th scientifi communit an ca serv guid progra an polic decisions thi respect th mos importan implication thi study' result ma concer it quantitativ finding abou th prevalenc an locatio o dru dependenc withi th population an th transi tio fro dru us t dru dependence no assesse wit epidemiologi estimate o th basi a NC sampl designe t generaliz t a larg segmen o th America population Th NC draw attentio t th relativ frequenc o depen denc o tobacco alcohol controlle substance an inhalants disclosin tha man mor Ameri can ag 15-5 hav bee affecte b dru depen denc tha b othe psychiatri disturbance no

accorde a highe priorit i menta healt servic deliver systems prevention an government sponsore researc programs NC finding o prevalenc correlate ad t a growin bod o evidenc tha Africa American d no see t mor vulnerabl t dru dependenc b virtu thei race an thes finding als poin towar som populatio segment suc a homemakers i whic exces dru dependenc ha bee suspecte bu neve demonstrate clearly Finally th NC result highligh a increasingl well-documente observatio tha man dru users perhap a vas majority d no see t mak a transitio t dru dependence Fo them instead dru us lack th majo complication

associate wit clinicall de fine syndrome o dru dependence I a tim o increasin concer abou governmen expendi ture fo healt an healt car reform th distinc tio betwee dru us an dru dependenc de serve greate consideration wit commensurat allocatio o resource i th directio o dru dependenc syndrome tha affec public healt an societ al to commonly
Page 23
26 ANTHONY L WARNER AN R KESSLE Reference Adams E H. Gfroerer J C. & Rouse B A (1989) Epidemiolog o substanc abus includin alcoho an cigarett smoking Annals o th Ne York Acad- em o Sciences, 562, 14-20 America Psychiatri Association (1980)

Diagnostic and statistical manual of mental disorders (3rd. ed.). Washington DC Author America Psychiatri Association (1987) Diagnostic and statistical manual of mental disorders (3rded., rev.). Washington DC Author Anthony J C (1991) Th epidemiolog o dru addic tion I N S Mille (Ed.) Comprehensive handbook o drug an alcohol addiction (pp 55-86) Ne York Marce Dekker Anthony J C. & Helzer J E (1991) Syndrome o dru abus an dependence I L N Robin & D A Regie (Eds.) Psychiatric disorders i America (pp 116-154) Ne York Fre Press Anthony J C. & Helzer J E (i press) Epidemiolog dru dependency I M

Tsaung M Tohen & G Zahne (Eds.) Textbook o psychiatric epidemiology. Ne York Wiley Anthony J. C. & Petronis K. R. (1993) Early-onset drug us an ris o later drug problems. Manuscrip submit te fo publication Anthony J C. & Trinkoff A M (1989) Unite state epidemiologi dat o dru us an abuse Ho ar the relevan t testin abus liabilit o drugs I W Fischma & N K Mell (Eds.) Testing fo abuse liability o drugs i humans (NID Researc Monograp No 92 pp 241-266) Rockville MD Nationa Institut o Dru Abuse Bland R C. Orn H. & Newman S C (1988) Lifetim prevalenc o psychiatri disorder i Edmonton Acta Psychiatric

Scandinavica, 7 (Suppl 338) 24-32 Brady J V (1989 Issue i huma dru abus liabilit testing Overvie an prospect fo th future I W Fischma & N K Mell (Eds.) Testing fo abuse liability o drugs i humans (NID Researc Monograp No 92 pp 357-370) Rockville MD Nationa Institut o Dru Abuse Breslau N. Fenn N. & Peterson E L (1993) Earl smokin initiatio an nicotin dependenc i a cohor o youn adults Drug an Alcohol Dependence, 33 129-137 Canino G J. Bird H R. Shrout P E. Rubio-Stipec M. Bravo M. Martinez R Sesman M. & Guevara M (1987) Th prevalenc o specifi psychiatri disorder i Puert Rico Archives o General

Psychia- try, 44 727-735 Compton W M. Helzer J E. Hwu H.-G. Yeh E.-K. McEvoy L. Tipp J E. & Spitznagel E L (1991) Ne method i cross-cultura psychiatry Psychiatri illnes i Taiwa an th Unite States American Journal o Psychiatry, 148, 1697-1704 Cottier L. Robins L N. Grant B. Elaine J. Towle I. Wittchen H.-U. Sartorius N. Burke J. Regier D. Helzer J. & Janca A (1991) Th CIDI-cor sub stanc abus an dependenc questions Cros cul tura an nosologica issues British Journal o Psychia- try, 159, 653-658 Crum R M. Bucholz K K. Helzer J E. & Anthony C (1992) Th ris o alcoho abus an depen denc i adulthood Th

associatio wit educationa level. American Journal o Epidemiology, 135, 989-999 Crum R M. Helzer J E. & Anthony J C (1993) Leve o educatio an alcoho abus an dependenc adulthood A furthe inquiry American Journal o Public Health, 83 830-837 Edwards G. & Gross M M (1976) Alcoho depen dence Provisiona descriptio o a clinica syndrome British Medical Journal, I, 1058-1061 Fleiss J E (1981) Statistical methods fo rates an proportions. Ne York Wiley Glantz M. & Pickens R (Eds.) (1992) Vulnerability t drug abuse. Washington DC America Psychologica Association Harrison L D (1992) Trend i illici dru us i

th Unite States Conflictin result fro nationa sur veys Th International Journal o th Addictions, 27 817-847 Hawkins J D. Catalano R F. & Miller J Y (1992) Ris an protectiv factor fo alcoho an othe dru problem i adolescenc an earl adulthood Implica tion fo substanc abus prevention Psychological Bulletin, 112, 64-105 Helzer J E (1985) Specificatio o predictor o narcoti us versu addiction I L N Robin (Ed.) Studying drug abuse (pp 173-205) Ne Brunswick NJ Rutger Universit Press Henningfield J E (1992) Occasiona dru use Compar in nicotin wit othe addictiv drugs Tobacco Control, 7 161-162 Hwu H.-G.

Yeh E.-K. & Chang L.-Y (1989) Preva lenc o psychiatri disorder i Taiwa define b th Chines Diagnosti Intervie Schedule Acta Psychiat- ri Scandinavica, 79 136-147 Johnston L D. O'Malley P M. & Bachman J G (1992) Smoking, drinking, an illicit drug us among
Page 24
COMPARATIV EPIDEMIOLOG O DRUG 26 American secondary school students, college students and young adults, 1975-1991. Vol. 2. College students an young adults (Nationa Institute o Healt Publi catio No 93-3481) Washington DC U.S Govern men Printin Office Kandel D B (1991) Th socia demograph o dru use Th Milbank Quarterly, 69

365-402 Kandel D B (1992) Epidemiologica trend an impli cation fo understandin th natur o addiction I P O'Brie & J H Jaff (Eds.), Addictive states (pp. 23-40) Ne York Raven Kandel D B. Murphy D. & Karus D (1985) Cocain us i youn adulthood Pattern o us an psychoso cia correlates I N J Koze & E H Adam (Eds.) Cocaine use in America: Epidemiologic and clinical perspectives (NID Researc Monograp No 61 pp 76-110) Rockville MD Nationa Institut o Dru Abuse Kessler R C (i press) Th epidemiolog o psychiat ri comorbidity I M Tsaung M Tohen & G Zahne (Eds.) Textbook o psychiatric epidemiology. Ne York

Wiley Kessler R C. McGonagle K A. Zhao S. Nelson C. Hughes M. Eshleman S. Wittchen H.-U. & Ken dler K (1994) Lifetim an 12-mont prevalenc o DSM-III- psychiatri disorder i th Unite States Result fro th Nationa Comorbidit Survey Ar chives o General Psychiatry, 51 8-19 Kidwell D A (1992) Discussion: Caveats i testing fo drugs o abuse. I M M Kilbe an K Asgha (Eds.) Methodological issues in epidemiological, prevention, and treatment research on drug-exposed women and their children (NID Researc Monograp No 117 pp 98-120) Rockville MD Nationa Institut o Dru Abuse Kish L. & Frankel M R (1970) Balance

repeate replication fo standar errors Journal o th Ameri- ca Statistical Association, 65 1071-1094 Kosten T A. & Kosten T R (1991) Criteri fo diagnosis I N S Mille (Ed.) Comprehensive hand- book o drug an alcohol addiction (pp 263-283) Ne York Marce Dekker Kosten T R. Rounsaville B J. Babor T F. Spitzer L. & Williams J B (1987) Substance-us disor der i DSM-III-R Evidenc fo th dependenc syndrom acros differen psychoactiv substances British Journal o Psychiatry, 151, 834-843 Kramer M. Vo Korff M. & Kessler L (1980) Th lifetim prevalenc o menta disorders Estimation use an limitations

Psychological Medicine, 10 429 435 Lee C K. Kwak Y S. Yamamoto J. Rhee H. Kim, S. Han J H. Choi J O. & Lee Y H (1990a) Psychiatri epidemiolog i Korea Par I Gende an ag difference i Seoul Th Journal o Nervous an Mental Disease, 178, 242-246 Lee C K. Kwak Y S. Yamamoto J. Rhee H. Kim S. Han J H. Choi J O. & Lee Y H (1990b) Psychiatri epidemiolog i Korea Par II Urba an rura differences Th Journal o Nervous an Mental Disease, 178, 247-252 Lillie-Blanton M. Anthony J C. & Schuster C R (1993) Probin th meanin o racia ethni grou comparison i crac cocain smoking Journal o American Medical Association,

269, 993-997 Robins L N (1993) Vietna veterans rapi recover fro heroi addiction A fluk o norma expecta- tion! Addiction, 88 1041-1054 Robins L N. & Przybeck T R (1985) Ag o onse o dru us a a facto i dru an othe disorders I L Jone an R J Battje (Eds.) Etiology o drug abuse (NID Researc Monograp No 56 pp 178 192) Rockville MD Nationa Institut o Dru Abuse Robins L N. Helzer J E. Croughan J. & Radcliff S (1985) Nationa Institut o Menta Healt Diagnosti Intervie Schedule It history character istics an validity Archives o General Psychiatry, 38 381-389 Robins L N. Wing J. Wittchen H.-U. Helzer J E.

Babor T F. Burke J. Farmer A. Jablenski A. Pickens R. Regien D A. Sartorius N. & Towle H (1988) Th Composit Internationa Diagnosti Interview A epidemiologica instrumen suitabl fo us i conjunctio wit differen diagnosti system an i differen cultures Archives o General Psychia- try, 45 1069-1077 SA Institute (1988) £4 6.03 [Compute program] Cary NC Author Schober S. Caces M. Pergamit M. & Branden L (1992) Effec o mod o administratio o reportin dru us i th Nationa Longitudina Survey I C Turner J T Lessler & J C Gfroere (Eds.) Survey measurement of drug use: Methodological studies (pp 267-276 DHH

Publicatio No AD 92-1929) Washington DC U.S Governmen Printin Office Schuster C R (1989) Testin an abus liabilit o drug i humans I M W Fischma & N K Mell (Eds.) Testing fo abuse liability o drugs i humans (NID Researc Monograp No 92 pp 1-6) Rock ville MD Nationa Institut o Dru Abuse Thompson T (1981) Behaviora mechanism an loc
Page 25
26 ANTHONY L WARNER AN R KESSLE dru dependence A overview I T Thompso & E Johanso (Eds.) Behavioral pharmacology o human drug dependence (NID Researc Monograp No 37 pp 1-10) Rockville MD Nationa Institut Dru Abuse U.S Departmen o Healt an Huma Services

(1993) National household survey on drug abuse: Main findings 1991 (DHH Publicatio No SM 93-1980) Washing ton DC U.S Governmen Printin Office Universit o Michigan (1981) OSIRIS VI [Compute program] An Arbor MI Institut fo Socia Re search Th Universit o Michigan Wells J E. Bushnell J A. Hornblow A R. Joyce R. & Oakley-Browne M A (1989) Christchurc psychiatri epidemiolog study Par 1 Methodolog an lifetim prevalenc fo specifi psychiatri disor ders Australian an Ne Zealand Journal o Psychia- try, 23 315-326 Wittchen H.-U (i press) Reliabilit an validit studie o th WH Composit Internationa Diagnos

ti Intervie (CIDI) A critica review Journal o Psychiatric Research. Wittchen H.-U Essau C A. vo Zerssen D. Krieg J.-C. & Zaudig M (1992) Lifetim an six-mont prevalenc o menta disorder i th Munic Fol low-U Study European Archives o Psychiatry an Clinical Neumscience, 241, 247-258 Woodruff R. S., & Causey B. D. (1976) Computerize metho fo approximatin th varianc o a compli cate estimate Journal o th American Statistical Association, 71, 315-321 Receive Januar 6 199 Revisio receive Apri 14 199 Accepte Apri 23 199