Addiction (1993) 88, 1041-1054

THE SIXTH THOMAS JAMES OKEY MEMORIAL LECTURE

Vietnam veterans' rapid recovery from heroinaddiction: a fluke or normal expectation?

LEE N. ROBINS

Department of Psychiatry, Washington University School of Medicine, St Louis, Missouri,USA

AbstractBetween 1972 and 1974, the outcomes of army enlisted men who had served in Vietnam during 1970-71were evaluated and compared with that of a matched group. This paper reports the major findings of thatstudy with respect to frequency of narcotic addiction in and after Vietnam, and the major risk factors forVietnam addiction and later relapse. Extraordinary access to records facilitated drawing the sample, locatingit, and verifying interview responses. The surprisingly low levels of readdiction and the rarity of addiction tonarcotics alone as compared with poly-substance dependence are findings still not entirely incorporated intopublic and scientific views of heroin addiction. Some defenses against that incorporation are examined.

IntroductionWhen planning this paper, I was asked to lookback at my study of Vietnam veterans (Robins,1973, 1974b) to answer whether at the time Ihad thought it would be important, and to sayhow the results have fared over the last 20 years.I am grateful for the opportunity to indulge inthis exercise of nostalgia.

The paper will begin by saying somethingabout how the study happened to be done. Thenit will summarize how the study turned out bothtechnically and substantively. It will end by say-ing whether the findings seem to have beencorrect and generalizable, and if so, whether theyhave been integrated into the expert or publicview of heroin addiction.

Why the Vietnam study was doneVietnam was linked to the general concern in theUnited States about drug use from the timesoldiers were sent there as 'advisers' in the early

1960s. First the concern was about marijuana. Itwas not until 1969 that heroin appeared in Viet-nam. Rumors of its use by servicemen wereconfirmed dramatically in 1971 when two con-gressmen who travelled to Vietnam announcedthat many servicemen were addicted. Nixon thendeclared heroin addiction to be the nation's "No.1 Public Health problem". He created the Spe-cial Action Office on Drug Abuse Prevention(SAODAP) in the White House, and tapped DrJerome Jaffe to head it. President Nixon per-suaded Dr Jaffe to give up his academic life byarguing that wars had stimulated scientific dis-coveries in the past. Dr Jaffe accepted theargument, and agreed to take up the challenge ofcontrolling heroin use in Vietnam, because if hefailed, at least his efforts would yield a goodscientific study. In June 1971, he went to Viet-nam and established a urine screening program,which required that any man due to leave Viet-nam be tested and have a 'clean' urine before

1041

1042 Lee N. Robins

boarding the airplane. Urine initially positive bythe FRAT test was to be re-tested with gaschromatography. If the test was again positive,the man was to be sent to detoxification forabout a week, and his urine re-tested before hewas allowed to leave the country. As soon as DrJaffe returned from Vietnam, he invited me todesign a study to estimate the size of the problemboth in Vietnam and after return.

My selection for the job was based on ourhaving served together as members of the Nar-cotic Addiction and Drug Abuse Grant ReviewCommittee of NIMH. I had been appointed tothis committee based on a single publishedpaper. My first major research effort was afollow-up of a child guidance clinic populationand a matched normal control group (Robins,1966). That study showed an impressive conti-nuity between childhood antisocial behavior andadult outcomes, but I worried about the study'sgeneralizability, because the subjects were pa-tients, all white, and all born between 1907 and1919. I decided to replicate it in a non-patient,later-bom, non-white sample. In 1965-66, I in-terviewed 235 St Louis-bom black men in theirearly thirties, selected from records of the StLouis elementary schools, which had beenracially segregated in their childhoods. In bothstudies I looked at drug use and abuse as oneoutcome. While drug use was rare in the firststudy, the black schoolboys reached adolescencejust after World War II, when drugs had enteredthe black ghettos but had not yet reached mid-dle-class white America. By the time I followedthem, illicit drug use had reached the whitemiddle class and was a raging concern through-out the country. Half my sample said that theyhad used an illicit drug, and 10% said they hadbeen addicted to narcotics. Although that wasonly 22 narcotics addicts, this was the first studyof heroin addiction in a 'non-captive' adult pop-ulation, i.e. neither in college nor in treatment.The paper reporting on their drug experience(Robins & Murphy, 1967) ted to an invitation toserve on the Narcotic Addiction and Drug AbuseGrant Review Committee.

When Dr Jaffe invited me to do the Vietnamstudy, I was captivated. When I went to discussit with him at the SAODAP office and foundhim working all hours, I began to realize thestudy might be important politically, as well asscientifically.

Why the study was a technical successThe study went well, helped by the government'sconcern about having exposed so many youngmen to what was viewed as a scourge.

It was possible to draw a sample representativeof all enlisted Army men who left Vietnam inSeptember 1971, the first month urine screeningwas in operation throughout the country. Weselected two samples, one from a tape of 22 000men made for us by the Department of Defense(DoD) from the computerized Active Duty Ros-ter of November, 1971. That roster was purgedevery month of those discharged 4 months be-fore. Thus, the November 1971 tape stillcontained all men in service in September. Menwere selected if they were listed as leaving Viet-nam that month. We were able to enrich thenumber of heroin users in the sample by sam-pling as well from records of those who testedpositive at departure, records held in the Sur-geon General's office. The military records ofboth groups were then checked to verify theirdeparture dates. From Selective Service records,we were able to draw a sample of draft-eligiblecivilians who did not serve but matched theveterans with respect to age, region lived in, andeducation at the time the veterans entered ser-vice. The high level of cooperation from thesubjects made possible a remarkably high inter-view completion rate, 96% when the veteranshad been back 8-12 months, of whom 94% werethen re-interviewed 3 years after their departurefrom Vietnam.

We also had access to all the record informa-tion we needed from the military and theVeterans Administration. These records, plustesting the urines we collected at the end of eachinterview, allowed us to verify what the men toldus in interviews. These checks showed remark-able honesty: 97% of those whose military recordshowed narcotics use told us about their usewhile in service, and tests of urine samples col-lected at the end of the interview showed nohigher rates positive for current use than didtheir self-reports given before they knew theywould be asked for a urine sample.

We were able to complete the first interviewswithin a 6-month period, which both enabled usto provide information to the Governmentpromptly, and meant that the men all had hadapproximately the same length of time at risk ofrelapse between return and interview. Wereinterviewed the veterans and the comparison

Sixth Thomas James Okey Memorial Lecture 1043

sample when veterans had been back 3 years tolearn whether the initial results would hold overtime.

We began to feel confident about our samplingstrategy when the DoD was not upset by ourestimate that the number of men who returnedin September 1971 was only 13 280, rather thanthe 22 000 on the tape given us. Our lowerestimate was based on finding that many of themen we initially selected from the tape had to bedropped because their military records showedeither that they never served in Vietnam or thatthey had left before September. While we werein the field, the DoD revised their estimate ofmen who returned in September based on ladingdocuments from the airplanes that brought themen home, and we had hit within 3% of therevised number!

The study's assets(A) A number of factors led to the high co-

operation rate and the accuracy of our estimates.First, Dr Jaffe had involved multiple agencies—the National Institute of Mental Health, theDepartment of Defense, the Veterans Adminis-tration, and the Department of Labor—in itsfunding. This opened all the files we needed forsampling and follow-up.

(B) Urine testing in Vietnam had begun inJuly, 1971, 2 months before we designed thesample. This 2-month experience with testinghad informed the DoD that the vast majority ofheroin users were Army men below the officerlevel, and they advised us to restrict the study tothis group.

(C) Being able to sample from the SurgeonGeneral's list of men who had tested positiveallowed us to over-sample heroin users so thatwe interviewed over 600 men who used narcoticsin Vietnam. It also enabled us to estimate theproportion detected as drug-positive by the mili-tary at departure. Duplications between those wesampled from the Surgeon General's list andthose we sampled from the Active Duty Rostergave us the needed data. Knowing the propor-tion positive then allowed us to weight theover-sample of positives back to their true repre-sentation in the population.

(D) All service records of those dischargedhappened to be stored in St Louis, making itconvenient for us to review the complete militaryrecord to check whether the man had actually

served in Vietnam and left in September, 1971.This turned out to be very important because theArmy was in a 'stand-down' phase, during whichmanpower in Vietnam was being reduced byshortening the scheduled I year tour of dutythere. Many of the names on our 'September'departure tape belonged to men who had actu-ally left earlier. The military record also providedname and addresses of the next of kin, whichhelped us locate the man, as well as informationabout disciplinary actions and reponed drug use,against which we could check honesty of theinterview.

(E) Even though the men selected lived allover the US, and a few were stationed overseas,they were not difficult to find. Those who en-tered via the draft were drafted at 18 for 2 years.They typically arrived in Vietnam at age 19 afteralmost a year Stateside, served in Vietnam for 1year, and were discharged soon after return, at20 or 21. At this age, most could still be reachedat their parents' home. The men who enlistedvoluntarily had a 3-year obligation, and more ofthem signed up for another term of service. Themilitary's Worldwide Locator made it possible tolocate them at their current station.

(F) We had excellent support from the Na-tional Opinion Research Center, the groupselected to carry out the interviewing. This wasmy first experience with employing an interview-ing organization; on previous studies, I had hiredand supervised interviewers myself. However,this study required rapid completion. An excel-lent study director and field director werewilling to do extraordinary things, such as ship-ping interviews to Canada to preserveconfidentiality through a double-link file, andcollecting urines and shipping them to Canadafor confidential testing. These extraordinary pre-cautions to preserve confidentiality were vital tothe enthusiastic cooperation of the interviewers,who were very concerned about possibly harm-ing the veterans. (The men themselves seemedremarkably unconcerned.)

(G) A delay in funding, that looked ominousinitially, turned out to be helpful in the end. DrJaffe had originally asked that we interview themen when they had been back 2-3 months. TTiefunding delay meant that they had been back 8months before we could enter the field. TTiisdelay gave them time to be discharged and re-turn home, making them easier to find. It alsomade the findings more meaningful. If they

1044 Lee N. Robins

100 r-

80 -

60

o 40o

20

Alcohol Marijuana Opium Heroin Amphet. Barbit.

SubstancesFigure 1. Subsiance use in Vietnam.

wanted to use heroin in the States, they had hadample time to locate a supply and begin to use it.

(H) Before the 3-year follow-up was due, thedraft ended and selective service registrationswere moved to a few Federal Records Centers.This made the location of civilian control sub-jects a great deal more efficient than if we hadhad to approach the Draft Board in each sub-ject's home town to find a civilian who matchedhim. The DoD even provided us with membersof Army Reserve Units, who spent their requiredyearly 6 weeks of active duty in going to theFederal Records Office to select civilian matchesfor our second follow-up. This overcame theproblem that as civilians we could not legallyhave access to Selective Service records (al-though we accompanied the Reservists to be surethey selected cases properly.)

I believe these were the reasons that the studywent well, along with the veterans' wish to beheard. But a technically successful study is excit-ing only if the results are interesting. Let mesummarize what I see as the five major findings.

The major findings(1) First, the DoD had greatly underesti-

mated the size of the narcotics problem inVietnam. Almost half (45%) of Army enlistedmen there in 1970-71 tried narcotics; 34% triedheroin and 38% tried opium. While fewer usedheroin or opium than alcohol or marijuana (Fig.1), narcotics were used at a truly astounding

rate. These figures reflect the fact that thesedrugs were cheap— even an addict need spendonly about S6.00 per day—and pure—they neednot be injected, a method which repelled manysoldiers. Instead heroin could be mixed withtobacco and smoked. That opium was used sowidely was a particular surprise; the press hadreported only on heroin use.

Addiction was also more common than re-ported—20% claimed that while in Vietnam theyhad felt strung out or addicted to narcotics. Tosee whether that claim really meant physiologicaladdiction, we asked for frequency and length ofuse and withdrawal symptoms. Those whoclaimed addiction had almost all used narcoticsheavily for a considerable time and suffered theclassic symptoms of withdrawal for at least sev-eral days. We concluded that their claims ofaddiction were correct.

Almost 11% of Army enlisted men's urinestested positive at departure. If the urine screencaptured only those so addicted they could notstop their use when warned that they would bescreened, as Dr Jaffe intended, this would meanthat about 1400 recently or still addicted enlistedsoldiers were arriving in the US per month (plusa small number of addicted officers and enlistedpersonnel in other branches of the military).

Serving in Vietnam was associated with greatlyincreased use of marijuana as well as narcotics;close to 80% of enlisted men used marijuanawhile in Vietnam. Other drugs were used too,but it was use of marijuana and narcotics that

Sixth Thomas James Okey Memorial Lecture 1045

l O O r -

None Occasional Regular

Narcotic use in Vietnam

Addicted

Figure 2. In Vietnam, narcotic use and drinking were inversely related (Drinking frequency:than daily; D daily).

less than monthly; S less

was distinctively greater in veterans than in thecomparison sample.

(2) A second finding was that the orderly se-quence of substance use described as a Guttmanscale, in which no one progresses to an illicitdrug without having used the legal drugs, alcoholand tobacco, and no one progresses to a 'hard'illicit drug without having used a 'softer' illicitdrug (Kandel, 1975), had been turned topsy-turvy in Vietnam. Before and after Vietnam,heavy alcohol use was almost a prerequisite fornarcotic use; in Vietnam, they were inverselycorrelated (Fig. 2). (This may have been ex-plained by military rules against selling alcohol tosoldiers under 21; for the average enlisted man,who arrived in Vietnam at 19, heroin was moreavailable than alcohol.) This displacement fromthe expected sequence also occurred for am-phetamines. Before service, amphetamines wereoften used without narcotics or barbiturates, butnarcotic use virtually never occurred withoutamphetamine use. In Vietnam, amphetamineswere essentially used only by users of barbituratesand narcotics. These sequence reversals showthat a drug's position in the sequence has moreto do with its availability than either with itsintrinsic 'hardness', by which I think we gener-ally mean its addiction liability, or with popularbeliefs about its dangerousness.

(3) In my study of young black men, I hadfound a history of deviant behavior strongly asso-ciated with drug use (Robins & Murphy, 1967).One might have expected the much greateravailability of drugs in Vietnam and stresses ofserving in a war to break the link with deviance.But they did not. The same relationship between

early behavior problems and drug use held inVietnam. We counted occurrence and severity offive pre-service behaviors: fighting; truanting;drunkenness; arrest; and school expulsion. Eachwas scored as not present, present mildly, orpresent at a serious level. Both a larger numberand greater seriousness increased the risk of us-ing narcotics in Vietnam (Fig. 3).

(4) In the first year after return, only 5% ofthose who had been addicted in Vietnam wereaddicted in the US (Fig. 4; Robins, Davis &Nurco, 1974b). TTiis finding was totally unlikethe outcomes of young men treated in Lexing-ton, the Federal Narcotics Hospital at the time.When those young men were followed 6 monthslater, two-thirds were found to already be re-addicted (Stephens & Cottrell, 1972). Even vet-erans still on narcotics at departure fromVietnam were doing very well 8-12 months afterreturn. The curve showing the likelihood of anyuse, heavy use, and addiction for veterans wasthe mirror image of that for treated civilians. Norwas this good result transient. When we followedveterans at 3 years, only 12% of those addictedin Vietnam had been addicted at any time in the3 years since return, and for those re-addicted,the addiction had usually been very brief.

It was not treatment that explained this re-markable rate of recovery. Only a third of themen addicted in Vietnam received even simpledetoxification while in service, and only a tinypercetitage of Vietnam enlisted men went intodrug abuse treatment after return—less than 2%of those who used narcotics in Vietnam, 6% ofthose who were positive at depanure, and 14%of those positive at departure who continued to

1046 Lee N. Robins

100,-

80 -

60

40

20

0None or one Two or three Four or five Six or more

Score on pre-Vietnam deviance

(Fight, Truant, Drunk, Arrest, Expelled)

Figure 3. Behavior before Vietnam and narcotic use in Vietnam.

use after return (Robins, 1975). Yet, those whodid enter treatment had relapse rates as high asthe young civilian men in Lexington—two-thirdshad relapsed by the time we interviewed them.Relapse often occurred the very day they left thehospital (Fig. 5).

Nor did recovery require abstention. Althoughnearly half the men addicted in Vietnam triednarcotics again after return, only 6% overall gotre-addicted (Fig. 6). Some were spared by usingonly narcotics other than heroin; some by notinjecting, some by using only occasionally. Buteven regular heroin users became re-addicted inonly half the cases (Robins et al., 1980).

This surprising rate of recovery even whenre-exposed to narcotic drugs ran counter to theconventional wisdom that heroin is a drug whichcauses addicts to suffer intolerable craving thatrapidly leads to re-addiction if re-exposed to thedrug.

(5) In the US, many drug programs have beenset up especially for heroin addicts, for whommethadone, experimental narcotic antagonists,or group therapy is provided. We found littleevidence for a group of veterans whose mainproblem is heroin addiction (Robins et al.,1978). Those who used heroin after return werealso using a great variety of other drugs. Morethan 80% used amphetamines, more than 70%used barbiturates, and almost all used marijuana.Those addicted to heroin often had multiple

dependencies. When heroin addicts were askedwhat their 'main' drug was, more than halfnamed alcohol or marijuana rather than heroin.

We found little to justify the view of heroin asan especially dangerous drug. Heroin was associ-ated with more adverse outcomes (crime,unemployment, illegal employment, divorce orseparation, violence, transiency, credit problems,alcohol abuse, drug problems) than barbituratesand amphetamines, but not if pre-service behaviorproblems and the number of other types of drugs usedwas held constant. When they were, barbituratesand amphetamines were associated with as manyproblems. Indeed, the variety of drugs used wasa better predictor of adverse consequences thanwhich drugs were used.

These then were the major conclusions: nar-cotic use and narcotic addiction were extremelycommon in Vietnam (although not as commonas use of alcohol and marijuana); availability wasthe main explanation; those with a history ofdeviant behaviour before Vietnam were particu-larly at risk; addiction was rare and brief afterreturn, even when men continued to use nar-cotics; veterans re-addicted and entering treatmenthad as high a relapse rate as civilians; andheroin's adverse effects were no greater thanthose of amphetamines or barbiturates whenjuvenile behavior and concomitant use of otherdrugs was taken into account.

Sixth Thomas James Okey Memorial Lecture 1047

MEN WITH NARCOTICEXPERIENCE BEFORE

N = 106)

u

40

20

0No Light Heavy Add"Use Use Use icMon

General Sample,N = 90

Drug Positive Sarr^ple,N = 341

122 NARA Pahents 6months after release toaftercare (Stephens &Conrell, 1972)

NoUse

LightUse

Heavy AddictionUse

Figure 4. Narcotic use in the 8-10 months post Vietnam by men first addicted in Vietnam.

The response to the study by the DoD, thepress, and the research communityEven though our study showed that the amountof heroin use in Vietnam bad been underesti-mated, the DoD was pleased with the findings,because they showed that Vietnam veteranshad not been consigned to a life of unrelentingdependence on drugs. (Our only problem waswith one general who wanted to take credit forhaving cured the addicts with a little Valiumduring detoxification, and claimed that ourstudy proved that methadotie maintenance wasunnecessary.)

The press and the research community weremore skeptical. They resisted giving up the be-liefs that heroin was a uniquely dangerous drug,to which a user became addicted very quickly,and addiction to which was virtually incurable.

TTiey maintained these beliefs by offering threeinterpretations of our findings:

(1) The results could be wrong. Perhaps thefindings were tailored to exonerate theDoD; the men might have been addictedafter return more often than we reported.(The suspicion that the results were politi-cally manipulated may have beenencouraged by the DoD's insistence onsponsoring the press conference whichannounced initial results.)

(2) Addiction in Vietnam was explained bythe extraordinary setting—the misery ofwar that made addiction a 'normal' reac-tion; so the relatively benign outcome ofaddiction in Vietnam was irrelevant toaddiction in the US.

(3) The drop in addiction on return was

1048 Lee N. Robins

1001-

80

60

M 40

20

1 4 12Weeks since release

Figure 5. Relapse to drug use following VA in-patient treatment.

caused by a change in setting. In the USthese men would not know where to getheroin, and the settings in which theylived and worked would not be associatedin their minds with use or withdrawalsymptoms, and therefore would not serveas stimuli to relapse. Thus again, the studyfindings would be irrelevant to addictionbeginning in the US.

I will argue that each of these views fails to fitthe facts.

(1) The idea of a whitewash by the DoD wasextensively investigated by a New York Times(NYT) reporter who had recently returned fromthe newspaper's Saigon office. He was followingup on a NYT news story in which a number ofscientists had commented on the study, somewith doubts about the results. During the 2months' preparation of his article, he went overthe study in great detail, often calling me tochallenge what he took to be discrepancies be-tween statements in my report. I spent a lot oftime explaining how the results had beenreached, and that percents should difFer when thegroup being described shifts from the wholepopulation of enlisted men to sub-populations,such as men who had already used drugs beforearriving in Vietnam. When the article finallyappeared, the Vietnam study was barely men-tioned in it. To get only one line in the two-pagestory that had been intended as an expose

showed that he could not support suspicions of aDoD whitewash.

Of course, the idea that there was a high rateof addiction after return was even moredefinitively disproved by the failure of veterans toapply for treatment. One of the original motiva-tions for the study was the VA's concern thatreturning addicts would overtax their services;but the anticipated large demand never oc-curred. A small increase in cases occurred onlybecause of a change in DoD regulations a fewmonths after the men's return, allowing theArmy to delay discharge by 30 days, which drug-positive soldiers would spend at a VA drugtreatment program. The DoD wanted men tobecome familiar with the VA system so that theywould return to it if they relapsed. However,they arrived at the hospital already detoxified,and in no obvious need of treatment. The VAgenerally transferred them to the hospital nearesttheir homes and allowed them to go home onleave, while remaining on the hospital books aspatients, as required by the regulation.

(2) The argument that addiction in Vietnamwas a response to war stress, and therefore remit-ted on exit from the Vietnam war theatre, is stillfrequently cited as though it were self-evident,because it sounds so plausible. Yet accepting thisargument is difficult in the face of the facts:Heroin was so readily available in Vietnam thatmore than 80% were offered it, and usuallywithin the week following arrival. Those whobecame addicted had typically begun use early in

Sixth Thomas James Okey Memorial Lecture 1049

100 n

8 0 -

•E 6 0 1 -

oo^ 40

20

Addicted in VN + Narc in US + Heroin in US + Regular use

Use by addicts after Vietnam

Figure 6. Risk of re-addiction depending on degree of exposure (M re-addicted; D not re-addicted).

their Vietnam tour, before they were exposed tocombat. Further, the dose-response curve that issuch a powerful causal argument did not apply:those who saw more active combat were notmore likely to use than veterans who saw less,once one took into account their pre-servicehistories. (Those with pre-service antisocial be-havior both used more drugs and saw morecombat. Their greater exposure to combat waspresumably because they had none of the skillsthat kept cooks, typists, and construction work-ers behind the lines.)

While combat had little effect, deviant behav-ior before service was a powerful predictor, as wehave seen. So was pre-service drug experience(Fig. 7); the greater the variety of drugs usedbefore entering seivice, the greater the likelihoodthat narcotics would be used in Vietnam. Likethe St Louis-bom young black men we studiedearlier, the best predictors for use of narcotics inVietnam were earlier antisocial behavior and ex-perience with precursor drugs.

Finally, when we asked men why they usedheroin, they did not tell us that they were over-come by fear or stress. Rather, they said it wasenjoyable and made life in service bearable(Table 1).

(3) TTie argument that men addicted in Viet-nam would not relapse because they would notbe re-exposed in the US to stimuli associatedwith their drug use and withdrawal symptomsexperienced abroad is based on Abraham Wik-Ier's conditioned response theory of relapse(Wikler, 1973). This theory has been the basisfor successful treatment programs, which gives it

empirical support. However, it fails to accountfor the fact that among the one-half of the Viet-nam addicts who did use heroin after theirreturn, most did not become re-addicted, andthose who did become addicted were addictedonly briefly. Only one-fourth of those re-addicted by the time of their first interview at8-10 months after return were addicted at anytime in the next 2 years, even though, havingbeen addicted in the US, they presumably hadthe opportunity to develop conditioned re-sponses to the US stimuli associated withaddiction and withdrawal (Fig. 8). Their claimnot to be still addicted was validated by findingfew of them to have positive urines at either thefirst or the second follow-up interview.

Where to the Vietnam study's findings standnow?What then are my views of heroin addictiontwenty years after the Vietnam study? I have stillnot found a serious fiaw in the study, and I thinkthe findings are consistent with those of otherstudies. For example,

(1) Charles Winick (1962), using the FederalBureau of Narcotics records and the New YorkNarcotics Register, found that names tended todrop off at ages 35-40, or after 5 years. Becausehe carried out no personal interviews, he couldnot prove that his critics were wrong when theyargued that the addicts had left town or died orsomehow managed not to get caught again.However, he believed that they had 'maturedout'—i.e. recovered spontaneously.

1050 Lee N. Robins

Preservice illicitdrugs

No preservice use

One drug only

Two types of drugs

Three types of drugs

Four Of more

27%

46%

76%

91%

100%

0 •) 0 20 30 40 50 60 70 80 90 100

Percent using narcotics in Vietnam

Figure 7. Preservice drug experience and use of narcotics in Vietnam.

(2) Waldorf & Biemacki (1979) did an ethno-graphic study of 50 untreated heroin addicts whohad been free of addiction for at least 2 years,and reported that the median duration of theiraddiction had been only 5.7 years.

(3) Our own small study of young black men(Robins & Murphy, 1967) included 22 who re-ported having been addicted to heroin. Only fourof these addicts had used any heroin at all in theyear prior to interview at age 31-35. And onlyfour had had any treatment. Heroin addictionfor them, as for Vietnam veterans, was associatedwith use of a large number of other drugs. Likethe Vietnam veterans, they experienced an accre-tion of drug types, not a 'stepping stone' patternof moving from milder to stronger drugs. Indeed,heroin use preserved, rather than replacing, theuse of other drugs.

(4) A study of young men 18-28 registeredwith Selective Service (O'Donnell et al., 1976),conducted at the same time as our 3-year follow-

Table 1. Why men used narcotics inVietnam

Reason volunteered

To feel highTolerate Army regulationsRelieve boredomRelieve depressionRelieve fearPass timeBe one of the group

%

401399853

up of veterans, showed that only 13% of thosereporting heroin use had had any treatment, andonly 27% of the untreated had used heroinwithin the last year, while 65% of the treatedhad. This suggested that treated cases are thosewho cannot stop on their own^and often do notstop when treated.

(5) A follow-up of Harlem youth (Brunswick,1979) showed that only 25% of young people18-23 who had used heroin twice or more hadused any in the current year; the rate of recentuse was particularly low among the untreated^—16%.

(6) Norman Zinberg identified a group of'chippers'—recreational users of heroin whonever became addicted (Zinberg & Jacobson,1976).

(7) The Epidemiologic Catchment Area studyof all major psychiatric disorders in random sam-ples of US adults in five geographic areas askedall affected the age of first symptom and the ageof most recent symptom, and whether they hadever discussed their illness with a physician(Robins, Locke & Regier, 1991). Subtracting ageof first problem from age at the last problemgives an estimate of duration. Substance abusehad the shortest mean duration of any of thepsychiatric disorders—only 2.7 years. And onlyone in five of those dependent on or abusing anillicit drug had had any treatment. Most of thesecases were not heroin addicts; marijuana was themost common drug of abuse. Still, this study

Sixth Thomas James Okey Memorial Lecture 1051

100

>•

-a

Addicted User only Abstainer

Narcotic experience in first year back

Figure 8. Fate of year I addicts, users, and abstainers in the next 2 years (U addicted year 2-3;abstainer, year 2 and 3).

user, non-addict; D

shows that drug addiction tends to remit quickly,even in the absence of treatment.

So, I am unrepentant. I think the results wereboth correct and not unique. TTiis is not to denythat the Vietnam situation was one quite un-known previously. Surely there has never been atime when so many young men were addicted toopiates in the absence of typical risk factors. Fewof the Vietnam addicts would have become ad-dicted if they had remained in the US. However,their history of brief addiction followed by spon-taneous recovery, both in Vietnam andafterwards, was not out of line with the Ameri-can experience; only with American beliefs. Norwas their ability to reuse without becoming re-addicted very unusual.

I think there are three important implicationsfor policy in this study, which still have not beenfully incorporated into our understanding ofheroin addiction. First, addiction looks very dif-ferent if you study it in a general population thanif you study it in treated cases. The small num-ber of Vietnam addicts who came to treatmentrelapsed as rapidly as do other young men intreatment. TTieir having become originallyaddicted as a response to stress or in an exoticsituation was of no help to them. TTiose who didnot come to treatment were able to get them-selves off heroin without help. The implicationsare first, that heroin addicts are not doomed forlife, and should be helped to remain in or re-

enter conventional society, and second, that drugusers who appear for treatment have specialproblems that will not be solved just by gettingthem off drugs.

Second, I think the evidence that heroin is notan exceptional drug has stood the test of time. Itsassociation with long-term addiction and crimeappears to be as much a function of its badreputation, that relegates its use chiefly to peoplewith little regard for social proprieties, as to itsintrinsic effects on brain and behavior.

Third, at least in the US, the 'heroin addict'for whom current treatment programs are de-signed is a mythological creature. Most users ofheroin use many substances, and may be as ormore dependent on one of these other drugs ason heroin. They are almost all (in the US) alsodependent on alcohol and tobacco. Yet treat-ment slots are still mainly available for heroin useand heroin users. In the Epidemiologic Catch-ment Area study conducted in the early 1980s,we found that persons with heroin problems inthe last year had a reasonable chance (35%) ofbeing seen in a drug treatment facility. Thosewith problems with other drugs only were veryunlikely to receive treatment in a drug program(less than 2% did). Yet they were only a little lesslikely than those with heroin problems to receivesome kind of psychiatric care (Fig. 9). The segre-gation of treatment by drug type seemsmisguided to me. It is no wonder that drug

1052 Lee N. Robins

100 p

80 -

60

40

20

0In drug facility Only other care

ECA sample with drug problem this year

Figure 9. Treatment this year,- heroin (IHJ vs. other illicit drugs only

treatment is not more successful when only asmall part of the total problem is beingaddressed.

As [ look back at the Vietnam heroin experi-ence, I conclude that soldiers had no specialreadiness to use narcotics or to recover fromaddiction to them. Their remarkable rate of usewas a response to market conditions—both thehigh availability of opiates and the lack of alter-native recreational substances, to the absence ofdisapproving friends and relatives, and to the factthat serving in Vietnam was not seen as part oftheir real-life career. Their readiness to recoverfrom addiction did not differ from that of otherusers. The reason that the press and scientistsalike were surprised was because studies of thegeneral population's drug dependence were andcontinue to be so rare. Their expectations werebased on the rates of relapse found in patientpopulations, made up of addicts who tried butfailed to get themselves off drugs.

It is not likely that the lack of general popula-tion studies will be remedied. Even large studiesof general populations, such as the NationalInstitute of Drug Abusers regular NationalHousehold Survey of Drug Use and yearly Mon-itoring the Future, which follows high schoolseniors into adulthood, do not study addictionbecause there are too few cases. The Epidemio-logic Catchment Area study, committed to thestudy of disorder, not just use, reported on drugabuse and dependence as a whole because only

0.7% of the population met criteria for opiateabuse or dependence.

Beliefs about heroin based entirely on resultsin treated populations have created a self-fulfilling prophecy. Heroin is or was (perhapsnow surpassed by 'crack') thought of, by lawenforcement personnel and users alike, as the'worst' drug, virtually instantly and permanentlyaddictive and creating craving so extreme that itovercomes all normal ability to resist temptationsto theft and robbery to acquire it. Users whoshare that view show by their use that they areready to commit themselves to their concept ofan addictive life style. The public's ranking ofdrugs with respect to 'hardness' probably hasmore to do with the drug's legal status, thegovernment's commitment to discouraging itsuse, and its price than with any intrinsic addic-tive liability. This is best demonstrated bytobacco, a highly addictive drug ranked as 'soft'.

The fact that heroin addiction is not alwaysinterminable is still being 'discovered', even byscientists. Just last year the Neuroscience Insti-tute at Harvard University published an anicleentitled "Addiction: What Science Knows" (Hy-man, 1992). It reassured readers that cravingdoes not condemn the user to indefinite use: "Inthe past, it was thought that physical withdrawalsymptoms appearing with drug cessation werethe key indicators of drug addiction . .. O^ow,i)ndeed, despite the 'common wisdom', avoid-ance of physical withdrawal appears to be a

Sixth Thomas James Okey Memorial Lecture 1053

Table 2. Variables that failed la predict re-addictionamong veteran addicts who returned to narcotics

Behavior problems before serviceParents' substance abuse, arrest, psychiatric careSubstance use in Vietnam {types, duration, injection)IQDraftee or volunteerCombatMilitary discipline problemsDemographic characteristics

indeed be biological. We don't know whetherthat biological vulnerability is presetit from birthor created by action of the addictive substanceon the brain, as the Newsletter article suggested.What we do know is that different factors operateat each of the critical tratisitions: from the drugnaive state to use; from use to addiction; fromaddiction to remission; from remission to re-lapse.

relatively minor factor in maintaining addictioneven for. . . opiates . . . " But the article still sup-ported the view that addiction is largelyirreversible. Accepting this 'common wisdom'triggered straining after some biological basis foran irreversibility that may well not exist: "Whatcauses the compulsive use?. . . (O)piates tap into(a) critical circuit in the ventral tegmentalarea . . . that interacts with the nucleus accum-bens and . . . the cerebral cortex . . ., the 'brainreward pathway' . . . With repetitive drug use thebrain reward system . . . undergo(es) adaptivechanges involving . . . dopamine . . . When thedrug is taken away, these neurons now functionabnormally, producing dysphoria and intensedrug craving . . . for many years." The author istoo sophisticated to think that dopamine is thewhole story: "Factors relevant to individualvulnerability include . . . heredity, . . . unknowndevelopmental factors, . . . other psychiatricdisorders, . . . pain,. . . stress. Sociocultural

factors also . . . modify the risk of addiction,includ(ing) drug availability, and pressures for oragainst drug use in one's family, subculture orpeer group."

Such a complicated explanation for a beliefthat addiction liability is relatively permanent isvery difficult to test, panicularly when the factsindicate that such permanence is rare rather thanuniversal. The solution to achieving some sim-plification may lie in shifting our focus fromaddiction as a unitary phenomenon to the transi-tions between use, addiction, and recovery. Inthe Vietnam study, the social and behavioralvariables that had been so useful in explainingwho would use narcotics enough to become ad-dicted to them in Viemam and which ex-addictswould use them again after they returned to theUS were of no use at all in explaining whichrelapsers would become re-addicted (Table 2).Vulnerability to re-addiction if re-exposed may

ReferencesBibliography of principal references from the Vietnam studyGOODWIN, D . W . , DAVIS, D . H . & ROBINS, L . N .

(1975) Drinking amid abundant illicit drugs; dieVietnam case, Archives of General Psychiatry, 32, pp.230-233.

HEI-ZER, J. E. & ROBINS, [,. N. (1985) Specification ofpredictors of narcotic use vs addiction, in: ROBINS,L, N. (Ed.) Studying Drub Abuse, Rutgers UniversityPress Series in Psychosocial Epidemiology, A. E.Slaby, Series edn., pp. 173-197 (New Brunswick,University of Rutgers Press).

HEiy.ER, J. E., ROBINS, L . N . & DAVIS, D . H . (1975-1976) Antecedents of narcotic use and addiction: astudy of 898 Vietnam veterans. Journal of Drug andAlcohol Dependence, 1, pp. 183-190.

Hi-t-ZKR, J. E., ROBINS, L . N . & DAVIS, D . H . (1976)Depressive disorders in Vietnam retumees. Journal ofNervous and Mental Disorders, 163, pp. 177-185.

HEL.ZER, J. E., RoBrNS, L. N., WISH, E . D . & HESSEL-BROCK, M. (1979) Depression in Vietnam veteransand civilian controls, American Journal of Psychiatry,136, pp. 526-530.

ROBINS, L . N . (1973) A Folhw-Up of Vietnam DrugUsers, Special Action Office Monograph, Series A,No. I (Washington, DC, Executive Office of thePresident).

ROBINS, L . N . (1974a) The Vietnam Drug User Returns,Special Action Office Monograph, Series A, No. 2(Washington, DC, US Government Printing Office).

ROBINS, L . N . (1974b) A follow-up study of Vietnamveterans' drug use. Journal of Drug Issues, 4, pp.61-63.

ROBINS, L . N . (1975) Drug treatment after recum inVietnam veterans. Highlights of the 20th Annual Con-ference (Perry Point, MD, Veterans AdministrationStudies in Mental Health and Behavioral Sciences,Central NP Research Laboratory).

ROBINS, L . N . (1977) Estimating addiction rates andlocating target populations: how decomposition intostages helps, in: RJTTENHOUSE, J. D. (Ed.) The Epi-demiology of Heroin and Other Narcotics, NIDAResearch Monograph Series 16 (DHEW Pub. No. 1(ADM) 78-559).

ROBINS, L . N . (1978) The interaction of setting andpredisposition in explaining novel behavior: druginitiations before, in, and after Vietnam, in: KANDEL,D. (Ed.) longitudinal Research in Drug Use: EmpiricalFindings and Methodological Issues (Washington,Hemisphere).

1054 Lee N. Robins

ROBINS, L . N . , DAVIS, D . H . & GOODWIN, D . W .(1974a) Drug use by US Army enlisted men inVietnam; a follow-up on their return home, AmericanJournal of Epidemiology, 99, pp. 235-249.

ROBINS, L . N . , DAVIS, D . H . & NURCO, D . N . (1974b)How permanent was Vietnam drug addiction? Amer-ican Journal oj Public Healthy Suppl. 64, pp. 38-43.

ROBINS, L . N . , HFI-ZER, J. E . & DAVIS, D . H . (1975)Narcotic use in Southeast Asia and afterward: aninterview study of 898 Vietnam returnees. Archives ofGeneral Psychiatry, 32, pp. 955-961.

ROBINS, L, N . , HESSEIJIKOCK, M . , WISH, E . & HELZI-R,J. E. (1978) Polydrug and alcohol use by veteransand nonveterans, in; SMITH, D , E . , ANDERSON, S.M., BuxTON, M., GoTTUEB, N,, HARVEY, W . &CHUNG, T. (Eds) A Multicultural View of Drug Abuse(Cambridge, Schenckman).

ROBINS, L . N . , HELZER, J. E., HESSELBROCK, M . &WISH, E . (1980) Vietnam veterans three years afterVietnam: how our study changed our view of heroin,

, L. & WINICK, C . (Eds) Yearbook of Sub-stance User and Abuse (New York, Human SciencePress).

WISH, E . D . , ROBINS, L, N . , HESSELBROCK, M . &HEt-ZliR, J. E. (1979) The course of alcohol problemsin Vietnam veterans, in: GAIANTER, M . (Ed.) Cur-rents in Alcoholism, Vol. VI (New York, Grune &Stratton).

References to other studies

BRUNSWICK, A. (1979) Black youths and drug-use be-havior, in: BESCHNER, G . M . & ERIHDMAN, A. S.(Eds) Youth Drug Abuse, pp. 433-492 (Lexington,MA, Heath).

KANDEL, D . (1975) Stages in adolescent involvementin drug use. Science, 190, pp. 912-914.

HYMAN, S. E . (1992) Addiction: what science knows.The Harvard Mahoney Neuroscience Institute LeUer, 1,pp. 1-3.

O'DoNNEtJ., J. A., Voss, H. L., CLAYTON, R . R . ,SlATiN,G. T. &RooM,R. G. W. (1976) Young Menand Drugs: A Nationwide Study, Research Mono-graph Series, Vol. 5 (Rockville, MD, NationalInstitute on Drug Abuse).

ROBINS, L . (1966) Deviant Children Grown Up: A Socio-logical and Psychiatric Study of Sociopathic Personality(Baltimore, Williams & Wilkins), Reprinted 1974 asa Behavioral Science Club Book selection (Hunting-ton, NY, Robert E. Krieger).

ROBINS, L . & MURPHY, G . E . (1967) Drug use in anormal population of young Negro men, AmericanJournal of Public Health, 57, pp. 1580-1596.

ROBINS, L . N . , U)CKE, B . & REGIER, D . (1991) Anoverview of psychiatric disorders in America, inROBINS, L . & RECIIER, D . (Eds) Psychiatric Disordersin America (New York, The Free Press).

STEPHENS, R. & COITREU.,, E . (1972) A follow-up of200 narcotic addicts conunitted for treatment underthe Narcotic Addict Rehabilitation Act (NARA),Bntish Journal of Addiction, 67, pp. 45-53.

WAtJJORF, D. & BiERNACKl, P. (1979) Natural recov-ery from opiate addiction: a progress report.Unpublished manuscript.

WlKLER, A. (1973) Dynamics of drug dependence:implications of a conditioning theory for researchand treatment. Archives of General Psychiatry, 28, pp.611-616.

WINICK, C. (1962) Maturing out of narcotic addiction.Bulletin on Narcotics, 14, pp. 1-7.

ZiNBERG, N. E. & JACOBSON, R. C. (1976) The naturalhistory of chipping, American Journal of Psychiatry,133, pp. 37^0.