Tobacco Documents Online

I I i I I 1 1 I I I I I I I I I I I Recidivism and Self-Cure of Smoking and Obesity STANLEY SCHACHTER Columbia University ABSTRACT: There is probably overwhelming profes- sional consensus that addictive-appetitive disorders such as obesity, opiate use, and cigarette smoking are markedly resistant to long-term modifiea' tion. Data are presented which indicate that this is not the case, and it is suggested that this is a conclusion based largely on the results of numerous studies of single therapeutic interventions with populations of self-selected subjeets who had actively sought help. Studies of nontherapeutic populations indicate that long-term self-cure of smok- ing, obesity, and drug use are relatively common events. It is generally accepted that smoking and over- eating are extraordinarily difficult conditions to correct. About obesity, Stunkard (1958) has con- cluded: "Most obese persons will not stay in treat- ment of obesity. Of those who stay in treatment most will not lose weight and of those who do lose weight, most will regain it" (p. 79). This depressing overview is based on Stunkard and McLaren-Hume's (1959) analysis of numerous studies of treatment of obesity and of the weight histories of 100 obese patients. After two years only two people had managed to maintain a weight loss of 20 pounds. This dismal state of affairs appears still to be the case 20 years after Stunkard's review, f or though there have been occasional reports of success, Wing and Jeffery (1979) summarized the results of 145 studies of treatment of obesity pub- lished between 1966 and 1977 as follows: "There has been little improvement in the clinical effec- tiveness of weight reduction therapy since Stunkard and McLaren-Hume's (1959) review" (p. 261). This pessimism is shared by those who have re- viewed the literature on treatment of cigarette of 87 studies (Hunt, Barnett, & Branch, 1971), they noted the startling resemblance of studies of smok- ing to studies of treatment of heroin addiction and alcoholism. In all three conditions, roughly 65% of successfully treated patients relapse within three months of termination of therapy, and within one year 80% of all patients are recidivists. Leventhal and Cleary (Note 1), in their more recent review, noted that after behavior therapy, "while most backsliding occurs within six months, it continues with inexorable force for 12 months till we arrive at a residual of quitters and smoking reducers of 10 to 25 percent of the pretherapy base level" (p. 22). Obviously there is much basis for pessimism, and there is probably overwhelming professional con- sensus that addictive-appetitive behaviors are markedly resistant to long-term modification. Yet despite this general consensus, something is curi- ously awry. Though therapists may find smoking notoriously difficult to cure, Horn's (1972) data 'indicate that literally millions of Americans have dropped the habit. Even the cachet of scientifically collected evidence seems unnecessary to mike the' ~ point, for virtually everyone knows large numberi of people who have quit smoking, apparently per-: manently. For narcotics addiction, too, there is evidenee: that it may not be the intractable condition that' has long been believed. Robins's (1974) study o( returned Vietnam veteran drug users indicate3 "a surprisingly high remission rate for heroin ad=i diction" (p. viii) without benefit of treatment. ,1 For both nicotine and heroin addiction, then;' there are indications that successful cure may be' smoking. Thus Leventhal and Cleary (Note 1) sum- ? maiized the situation in smoking therapy as fol- lows: "That so many people who are motivated to seek therapy drop out of treatment, and that so many people eventually return to the habit un- derscores the scope of the task that one is faced with in dealing with the smoking problem" (p. 23). Hunt and Matarazzo (1973) drew these same somber conclusions, and on the basis of an analysis This resarch was supported by National Institutes of HaUf<r+ Grant 1 RO1 HD 12910-Q2. Albert Stunkard was particul~ generous with his time and ideas in discussions of the interprel tation of these data. I am grateful to Don Hood and Robeet ICrauss for critical readings of early versions of the manuscd$ and to William Ger{n and Michael Rennert for their ae10, en.ble help in the preparation of this paper. Requests for reprints should be wnt to Stanley Schacktrr 400 ~ Department of Prychology Columbia University , , rnerhorn Hall, New York, New York 10027. 436 • APRIL 1982 • AMERICAN PSYCHOLOGIST - . Vol. 37, No. 4, Copyri& 1982 by t6e xp, abebdol A~so 'a.timY! _ .1 C'7 ~ ~ ~

I I I I I I I I I I I I I I I I I far more common than heretofore anticipated. Though one can devise more exotic hypotheses, explanations for the discrepancy between profes- sional opinion and apparent fact may be embar- rassingly simple. First, people who cure themselves do not go to the therapists. Our view of the in- tractability of the addictive states has been molded largely by that self-selected, hard-core group of people who, unable or unwilling to help them- selves, go to therapists for help, thereby becoming the only easily available subjects for studies of re- cidivism and addiction. Second, the inferences drawn from studies of therapeutic effectiveness are based on single at- tempts to cure some addictive state or other. In fact, people do try to quit repeatedly. It may be that with or without professional help, success rates with multiple attempts to quit are greater than with single attempts to do so. Though I know of no related evidence on obe- sity, common sense suggests that the same in- terpretive problems may well hold for this con- dition. People who cure themselves without therapeutic intervention do not become part of the data, and the reputation of this condition for in- tractability may also be grossly exaggerated be- cause of this fact. The study described here was designed to ex- amine' the intractability of cigarette smoking and obesity in a non-self-selected population. Method The study involved interviews with 161 people on their smoking and weight histories. I conducted the interviews myself, and using a strategy similar to that of Kinsey, Pomeroy, and Martin (1948), 1 attempted to interview the entire membership of a number of carefully selected groups, rather than attempting to interview a randomly selected pop- ulation. Given the rationale of this study, it seemed crucial to minimize subject self-selection. Possibly those who try and fail to lose weight or to quit cigarettes are less willing to talk about their ex- periences or are the sort of people who do not care for interviews. The current best estimate from the National Opinion Research Center (NORC) is that some 25%-3095 of a national probability sam- ple with predesignated respondents will not yield completed interviews. The groups interviewed were: 1. The Psychology Department at Columbia University. The membership of this group was defined as everyone listed under Psychology in the university catalog and everyone listed in the de- partment's own directory in the fall of 1977. There were 89 such people. Interviews were conducted in the spring of 1978, by which time terminating and quitting school had reduced this number to 84. Of these, 83 agreed to be interviewed. Twenty- eight were faculty members, 43 were graduate stu- dents, and 12 were secretaries and technicians. There were 46 men and 37 women, and they ranged in age from 20 to 64 years old. 2. A substantial portion of the entrepreneurial and working population of Amagansett, a small town in eastern Long Island, New York. As with the Columbia population, this is a group I know well, for I have summered in this town for 20 years. Amagansett is a seaside resort community with a permanent population of about 1,500 people and an additional summer population estimated to be between 1,500 and 2,000 people. The group to be interviewed was chosen by designating all shops and enterprises fronting on a 750-foot stretch of the main street of the town as the source of the sample. There are 19 shops and miscellaneous en- terprises, such as hardware stores, liquor stores, and barber shops, located along this stretch. Since I meant this sample to consist largely of lifelong, year-round residents of the area, I interviewed all of the personnel working in only those enterprises that remained open throughout the year and that had been in business for at least a year. This elim- inated four enterprises designed to serve the sum- mer people and staffed largely by outsiders or new- comers to the area. All told, in midtown Amagansett 48 people worked in the 14 enterprises that made up the sample. Of these, 47 agreed to be inter- viewed during the late spring of 1980. An additional group of subjects came from the outskirts of the town, along which strings an ad- ditional nine enterprises. Of these, I interviewed the personnel of the two largest shops satisfying the open-all-year criterion, a supermarket with 26 employees and an automobile agency with 5 em- ployees. These 31 people were interviewed during the late spring of 1979 and 1980. All told, 78 of a possible 79 people in this small-town sample agreed to be interviewed. There were 44 men and 34 women, ranging in age from 16 to 79. There are, then, two distinct populations of sub- jects-the Columbia group, largely academic, completely urban, ethnically diverse with a large number of Jews and Hispanics; the Amzgansett group, largely entrepreneurial and blue collar, mostly small-town born and bred. To the extent AMERICAN PSYCHOLOGIST • APRIL 1982 • 437 ~ . ..__ ~ ,- - - - -- - - - - _. _

I I I I I I I 1 I I I I I I ~ I that these two populations are similar in the vari- ables under study, we can begin to make guesses about the generalizability of the phenomena under study. The Interview The instrument for data collection was a stan- dardized, open-ended interview designed to get the individual's smoking and weight history. After initial explanations and obtaining the subject's es- timates of his or her weight and height, the ques- tioning proceeded by determining, first, whether the interviewee had ever been a smoker. If so, the interview continued as follows: "Now I'd like to go from the time of your first cigarette to the pres- ent. As well as you're able, I'd like you to tell me your smoking history. From the time of your first serious cigarettes to now, I'd like to know roughly how much you smoked per day and what brand. If at any point along the way you stopped smoking or cut down, I'd like to know about that." During the subject's narrative, I probed constantly to make sure that throughout his or her life, I knew his or her smoking habits including detailed descriptions of any attempts to quit. Smoking history completed, the interview turned to weight, beginning: 1. "How would you have described your build as a child?" If the answer was chubby or any syn- onym for overweight, I attempted to get a rough estimate of height and weight during various times of •childhood, and the interview continued. 2. "At what age roughly did you reach full height?" 3. "What was your weight at that time?" 4. "Now let's go from your age of full height to the present. Again tell me your weight history in such a way that throughout your life I know what you weighed. At any point along the way that you tried to lose weight or dieted, we'll stop and talk about that." Then just as with smoking, I probed throughout the subject's narrative to make sure that I knew exactly what he or she weighed TABLE 1 CharGctertSttcs of the SfrtolCi7tg Population when and was familiar with the details of any at ~ tempts to lose weight. • SMOKINC There were 94 people in these two popuiation= with a history of cigarette smoking. Some of their characteristics as smokers are presented in Table 1, in which smokers are characterized as either (1) heavy smokers-those who presently or for- merly smoked at least three quarters of a pack a day and had smoked for at least a year-or (2) light smokers-those who smoked less than three quarters of a pack a day and had smoked for at least a year. It is obvious that these are or were smokers of substance. Heavy smokers smoked over 1V2 packs a day and, on the average, were smokers for more than 17 years. Even the light smokers had been regular smokers for an average of more than 8 years. The categorization of every person in these groups according to present and former smoking status is presented in Table 2. The various column headings are defined as follows: 1. Cured smoker: Those who tried to quit smok- ing one or more times and are presently non- smokers. These people describe themselves as suc- cessful quitters, and the large majority are complete abstainers. A few admit to taking an occasional cigarette at a party or some such occasion. 2. Failed smoker: Those who have tried to quit smoking one or more times and are presently smokers. 3. Indifferent smoker: People who have never tried to stop. Most say that they enjoy smoking and have no intention of quitting. 4. Switched to cigar or pipe: Former cigarette, smokers who now smoke only cigars or a pipe, oc' both. 1 5. Cigar or pipe smokers: Those who have never smoked cigarettes but do or have smoked cigars or pipes. Table 3 summarizes the success of those who attempted to quit in these two populations. In tb~ Type aF' - smoker N M aa yr. smoked Raote of yr. smoked M no. ut~cduly Raaaa of cigarettes amoicdtily Heavy 73 17.4 2-90 32.9 17-90 Light 21 8.2 1.5-32 6.8 1-12 438 • APRIL 1982 • AMERICAN PSYCHOLOGISr ~ .-~ .;~--~ - - -- - . - ~

I I I I I I I ~ I I i I I I I I I I TABLE 2 Current Smoking Status of All Past and Present Cigarette Smokers n cured n failed n indifferent n switched to cigars or pipe Type of smoker Male Female Male Female Male Female Male Female prychology department' Heavy 13 5 6 2 1 3 3 0 Light 0 7 2 3 3 0 1 0 Amagansett" Heavy 13 7 7 7 3 6 1 0 Light 2 2 1 0 1 0 0 0 Combined populations Heavy 26 12 13 9 4 9 4 0 Light 2 9 3 3 4 0 1 0 ' No. of nonsmokers in department: 13 males. 17 females. No. of lifelong cigar or pipe smokers: 4 males, ' No. of nonsmokers in Amagansett: 15 males. 12 females. No. of lifelong cigar smokers: 1 male. respect the two populations are similar, and over- all, 63.6% of those who attempted to quit smoking had succeeded at the time of the interview. On the average, successful quitters had been off smok- ing for 7.4 years. Some 87.8% of these quitters had been nonsmokers for a year or more and 98.0% for three or more months. Comparing these figures to the roughly 10%-25% of the therapeutic popula- tion who remain quitters one year after therapy, it is clear that in two populations (one urban and academic, the other small town and largely entre- preneurial and working class), those who at- tempted to quit were at least two to three times more successful than were those self-selected sub- jects who in other studies went for professional help. The fact that the successful quitting rate is similar in these two very different populations does indicate that a high success rate is not a charac- teristic of one particular kind of group. Do these data indicate that giving up smoking is less difficult than heretofore suspected? Unsur- prisingly, the answer appears to be-for some peo- ple, yes, and for others, no. Answers to the ques- tion, "Was it hard (to give up)?" and related probes were coded to assess the difficulty of quitting. For light smokers, whether they succeeded or failed, giving up appears to have been virtually painless. Only two people reported any difficulties, and 88.235 of all light smokers who attempted to quit insisted that there was nothing to it-no with- drawal symptoms, no craving, no problems. For heavy smokers giving up was considerably more difficult: 45.8% of heavy quitters reported major difficulties such as marked irritability, sleep- lessness, intense cravings, fevers, cold sweats, and the like; 25.4% noted minor difficulties; and 28.8% of heavy quitters reported no problems. Despite these differences it can be seen in Table 2 that light smokers are no more successful at giving up than are heavy smokers, for 64.7% of all light smokers who tried to quit succeeded and 63.3% of heavy smokers did so. To summarize, these data suggest that a consid- erably larger portion of the general population may have given up smoking for long periods of time than has heretofore been suspected from es- timates based on studies of therapeutic success. The process of giving up appears to be relatively pain- less for light smokers and painful for many heavy smokers, but the two groups are equally successful at quitting. OBESITY Since weight data are self-reports, a brief note on the probable accuracy of these reports is in order. Interviewees were asked their weight early in the interview. In the Psychology Department popu- lation, it was possible to weigh 59.5% of the group after the interview. Self-report was close to scale weight, the two differing by an algebraic average of only 3.8 pounds, or 2.5% of scale weight (range 0-8.77c). This replicates Stunkard and Albaum's TABLE 3 Proportion of Successful Quitters in the Amagansett and Psychology Department Populations % oi Group n wbo ttiod to quit smoiting staccessful quitters Amagansett 39 61,5 Prychology department 38 65.8 AMERICAN PSYCHOLOCIST • APRIL 1982 9 439

I I I 1 I ~ I I I I I I I I I I t:~ . TABLE 4 Current Weight Status of People With a History of O6esity Who Attempted to Lose Weight Cured Partial Partial• Failed Population N fat cured fat failed fat fat Psychology department 18 72.0 0 16.7 11.1 Amagansett 22 54.5 18.2 9.1 18.2 Note. Current weight status is given as percentage, (in press) finding that self-reports of weight are extremely accurate. As for the reports of weight history, there is no simple way of checking their accuracy. I have known roughly 30% of these peo- ple for 10 or more years, however, and nothing reported about body size (or smoking behavior) jarred with memory. Of the 161 people in these two populations, 46 had a history of obesity and 40 of these had made an active attempt to lose weight. Table 4 presents the distribution of these 40 according to weight status at the time of the interview. Obesity is de- fined as 15% or more overweight, as calculated from the table of average weight, published by the Soeiety of Actuaries (1959). The categories in Ta- bles 4 and 5 are defined as follows: 1. Cured fat: people who at some time in their lives had been obese and (a) whose present weight was at least 10% less than their weight immediately before dieting, and (b) who were no longer obese, that is, were now less than 10% overweight. Such people can be considered complete cures. They have lost substantial amounts of weight and are no longer fat. 2. Partially cured fat: people who had been obese and (a) whose present absolute weight was at least 109ro less than their weight before dieting, but (b) who were still obese, that is, were presently 15% or more overweight. Such people are consid- ered only partial cures. Though they have lost con- siderable weight, they are still, in statistical terms (as well as to the interviewer's eye), fat. As will be seen in Table 5, however, in terms of sheer pound- age lost, they are even more successful than are cured fats. 3. Partially failed fat: people who had been obese and (a) who had lost less than 10% of their maximum predieting weight, but (b) who were no longer obese, that is, were now less than 15% over- - weight. People in this category should probably be considered failures. They had not lost very much weight over the years, but since average weight 440• • APRIL 1982 • AMERICAN PSYCHOLOGIST I increases with age, a weight that in youth would' be classified as obese could, by the actuarial tables, be classified as normal in later years. 4. Failed fat: people who had been fat and who, despite past or present efforts to lose weight, were at the time of the interview still fat by the follow- ing criteria: (a) They had lost less than 10% of their maximum prediet weight, and (b) in terms of de- viation from the norm, they were at present more than 15% overweight. 5. Indifferent fat: obese people who maintained that their weight was of no concern to them and that they had never undertaken either a diet or an exercise program. At most, they admitted to some token gesture such as substituting Sweet-N-Low for sugar. 6. Normal: people who had never in their adult lives been as much as 15% or more overweight. Included in this group are three people who re- ported marked juvenile obesity but who had been normal throughout their adult lives. The following conventions were used to cope with special problems of categorization: (a) Weights during or in the months immediately following pregnancy were never used in assigning subjects to category. (b) People who, when interviewed, were on medication known to affect weight (e.g., diuretics, cortisone, etc.) were eliminated from the data. There were three such subjects. By the cri- teria for categorization, two of them would have been classified as cured fat and one was a lifelong normal who, on cortisone, had, to his bewilder- ment, fattened up in the months preceding the interview. (c) One subject, an athlete, was elimi- nated on common sense grounds. During college, he had deliberately built up his weight by weight lifting and following college promptly returned to normal weight. Following our rules, he would have been classified as a cured fat. In Table 4 it can be seen that of those inter--s viewees with a history of obesity who had actively attempted to lose weight, 62.5% had succeeded. They had lost substantial amounts of weight and were no longer fat. An additional 10%, though still obese in terms of deviation f rom an actuarial norm, : had lost and kept off large amounts of weight. The , Psychology Department population was more suc-'; cessful than the Amagansett group, but the succesa: rate in both of these populations was so much greater than the success rates reported in the ther- apeutic literature that there seems no need to dwell• on possible reasons for the relatively trivial differ= ences between these urban and small-town popu-: lations. The medical literature (Stunkard ~ ~ l I

I I I I I I I I I I I I I I I I I I McLaren-Hume, 1959) indicates that only 2540 of obese people in treatment will lose as much as 20 pounds, and only 5% will lose as much as 40 pounds. Details of the weight histories of the erstwhile and currently obese members of the combined psychology Department and Amagansett popula- tions are presented in Table 5. It is evident that the majority of these people have lost formidable amounts of weight and have kept it off for many years. Among men, 67% of those who tried to lose weight are classified as cured fats and, after an average of 13.4 years, are 39.1 pounds lighter than when they began whatever weight-loss regimen they favored. Among women, members of the cured fat group make up some 58% of the popu- lation and are on the average 29.0 pounds lighter than when they started dieting some 8.3 years ear- lier. The extent to which the members of these two populations outdo therapeutic populations be- comes evident on examination of studies of ther- apeutic outcomes. Stunkard and Penick (1979) re- ported that behavior modification has proven to be by far the most effective of all nonsurgical treat- ments for weight loss. They reviewed 10 studies employing-behavior therapy, and their tables in- dicate that in :a one-year follow-up, the average weight loss in the behavior therapy conditions of these studies was 10.9 pounds. By contrast, in the entire population of Amagansett and Psychology Department obese (all of the people in Table 5, excepting the six indifferent fats), the males have kept off an average of 26.8 pounds in the mean 9.7 years since inception of their own weight loss procedures and females an average of 24.8 pounds in 7.5 years. It does appear that the obese members of these non-self-selected populations are markedly more able to lose and to keep off weight than are the obese patients in studies of therapeutic effec- tiveness. Since the patients in studies of therapy tend to be heavier than the obese of Amagansett and Co- lumbia University, it is conceivable that this fact could account for the relative success of these two populations. Therefore, a separate analysis was made of the heaviest members of these groups- those exceeding 30% overweight. There were 11 such people; six men, five women. They ranged from 32.6% to 75.2% overweight and averaged 56.1% overweight immediately preceding what- ever weight-loss regimen they favored. Obviously these were formidably heavy people. Of this group, 63.6% were classified as cured fat, 27.3% as partially cured fat, and 9.9% as failed fat. They had lost an average of 46.7 pounds and had been within a few pounds of their present weight for an average of 8.6 years. If anything, the once grossly obese members of these non-self-selected populations were even more successful at losing weight than were the modestly obese. Discussirm. It is the case that in nontherapeutic populations, the rates of successful self-cure of cigarette smok- TABLE 5 Weight Histories of All Erstwhile and Currently Obese Members of the Combined Psychology Department and Amagansett Populations Pat Pre~eat Category N Ase at iAftview Prediet wViok orer- ti at weisht iaterview cver- weia4k No. yr. since beoncift weiaht las rcEimeo Ma1C Cured fat 14 43.1 214.4 29.4 175.3 1.2 13.4 Partially cured fat 0 - Partially failed fat 2 51 ~ 179.5 11.9 168.5 4.6 3.6 Failed fat 5 23.0 190.4 31.8 191.9 30.2 1.6 Indifferent fat 6 39.7 - 200.7 21.9 - Femak ~ Cured fat 11 34.0 158.5 27.3 129.5 -.6 8.3 ~ Partially cured fat 4• 26.8 191.5 57.8 153.3 24.1 2.3 Partially failed fat 3 27.0 156 16.8 148.3 8.6 5.7 Failed fat 1 55.0 185 23.3 198 26.9 25.0 Indifferent fat 0 - - ~ AMEx1cAN PsYcxoLoclsr • AP[tll. 1982 • 441' ~ -~.

I I I I I I I I I I I I I I I I I ing and of obesity are considerably higher than anything yet reported in the therapeutic literature. This conclusion is based on virtual-100% samples of two different populations-an urban university psychology department and a geographically de- fined portion of the entrepreneurial and working population of a very small town. The fact that the rates of self-cure are so similar in these two pop- ulations is taken as evidence that these findings are generalizable beyond any single demographic group. Obviously the obtained rates of cure are at best gross estimates of national trends and, for ci- garette smoking at least, are probably somewhat high. In the two populations interviewed, the well- educated are overrepresented, laborers underrep- resented, and farmers not represented at all. Since it is known (Hammond & Garfinkel, 1971) that for smoking, quit rates tend to be relatively high for the educated and low for the working classes, it seems likely that the obtained rate of 63.6% suc- cessful quitters is on the high side. Evidence that this is the case can be found in the results of a Public Health Service (1979) study of a national sample of approximately 12,000 people. Of this group, 54% were present or former cigarette smok- ers. Some 74.8% of present and former smokers hzd tried to quit and of these 50.4% had suc- ceeded--a figure lower than that obtained in the Amagansett and Columbia University populations but still markedly higher than almost anything reported in the therapeutic literature. Those who quit had been off cigarettes for an average (roughly estimated) of seven years, a figure closely in line with the 7.4 years of the Amagansett-Columbia University populations. The apparent fact that cure rates in non-self- selected populations are markedly higher than in therapeutic populations appears also to be the case for heroin and opiate addiction. Lee Robins (1974) examined the histories of a randomly selected group of present or former narcotics users among the population of Vietnam veterans. As standard procedure when a soldier completed his tour of duty, before being shipped home he submitted a urine sample. If narcotics were detected, he was detoxified and then sent back to the States. Thanks in part to this procedure, Robins was able to select a random sample of 495 known drug users. Eight to 12 months after these soldiers returned to the States, Robins's staff, in a remarkable effort, was able to locate and interview 95% of this group and to collect urine specimens from 92% of these men. From this sample, Robins identified a group of 134 men who were heavily addicted before returning to the States. For this group 'she notes: Half of these men, all of whom were certainly psycho• logically dependent on narcotics and mcst of whom were probably physiologically dependent used no narcotics at all after their return to the States, and only 14% became readdicted. While 14% is a readdiction rate twice as high as that for all men detected as drug users in Vietnam, it is still remarkably low compared to remission rates in the States for men identified as actively addicted in hos- pitals and clinics. Not only did few become readdicted to narcotics after return, but 72% said they were having no problems at follow-up attributable to drug use. (p. 63) These appear to have been long-lasting cures, for in a follow-up study, Robins, Heizer, Hessel- brock, and Wish (1980) found that, "of all the men addicted in Vietnam, only 12 percent have re- lapsed to addiction at any time since their return, that is, at any time in the last three years" (p. 220). One could interpret this extraordinary cure rate, in sociopsychological terms, as attributable to the change in setting from the Vietnam war scene to a presumably less stressful home setting or, as I am inclined to do, in sampling terms as one more in- stance of the effect of evaluating cure rate in a non-self-selected population. Whichever explana- tion one favors, it is dramatically clear that as for obesity and cigarette smoking, the rate of recidi- vism in opiate and heroin addiction may be as- tonishingly lower than anything suggested in the literature on therapeutic effectiveness. It does appear that the generally accepted professional and public impression that nicotine addiction, heroin addiction, and obesity are almost hopelessly difficult conditions to correct is flatly ; wrong. People can and do cure themselves oF smoking, obesity, and heroin addiction. They do'4 so in large numbers and for long periods of time,': in many cases apparently permanently. The reputation of the addictive and appetitive+ disorders for intractability is undoubtedly tbe product of the huge number of studies that hav e demonstrated the ineffectiveness of therapy. H°~ is the fact that the rates of therapeutic success aaa~ generally so pitiable when compared to the ra of self-cure in a non-self-selected population to explained? Certainly the most obvious explana must be in terms of self-selection--Only the m difficult cases seek help; people who cure th selves do not go to therapists. Though this was original hypothesis, I suspect now that there is siderably more involved than one more em rassing demonstration that grossly distorted '1 ~ 442 • APau. 1982 • AMERtcAN Psircxoc.ocisr

I I. I I I I I I ~ I I I I I I ~• pies lead to grossly distorted conclusions. The inferences- that have been drawn from studies of therapeutic effectiveness are curiously misleading. They correctly describe the results of a single at- tempt to quit smoking or lose weight or what have you, but from such results nothing can or should be inferred about the probable success of a lifetime of effort to quit smoking or lose weight. Yet these are precisely the inferences that have been drawn again and again. Because literally hundreds of studies of single attempts to cure some addictive disorder or other have repeatedly reported pa- thetic rates of success, we have concluded that the addictive behaviors are unyielding, almost hopeless disorders. The logical difficulties involved in such a con- clusion become evident on consideration of the nature of the data gathered in interviews such as those conducted with the Amagansett-Psychology Department populations. These were retrospective self-reports-{iescriptions of a lifetime of effort to control weight or smoking, a lifetime that may have involved one or many attempts to lose weight or quit smoking. If it is assumed that the proportion of successful quitters cumulatively increases with successive attempts, it becomes clear why the gen- eralizations that have been made from the results of single therapeutic attempts to cure are probably unwarranted. For example, assume that for any single attempt to quit, 104b of those who try suc- ceed permanently-a figure in line with cure rates reported in most therapeutic studies. Assume fur- ther that at a later time, all of those who failed try again. Again, 10% succeed. Cumulatively, then, 19% of those who have tried twice will have suc- ceeded. A third attempt to quit, again assuming a 10% success rate, would yield a cumulative total of 27% success. And so on. Obviously, depending on the exact parameters assumed for rate of success per attempt, rate of recidivism, and the proportion of people who try repeatedly, one could derive virtually any cumulative success rate. However, the general line of reasoning makes it seem likely that retrospective interviews with people who have at some time sought help will yield higher cure rates than those reported in studies of one-shot therapeutic intervention. There are too few such people in the Amagansett-Psychology Department populations to permit any firm conclusions, but for these few, the data indicate that this is the case. There were 14 people who at some time during their lives had sought help, 2 for smoking, 12 for obesity.l Their "helpers" included psychothera- pists, physicians, hypnotists, and groups such as Weight Watchers. Of these 14 people, 42.9% were categorized as either cured fat or cured smoker- a cure rate greater than almost anything reported in the therapeutic literature and, to me, an indi- cation that the reputation of these conditions for hopeless intractability may, in part, be an unfor- tunate by-product of the fact that almost all tests of the matter have involved evaluation of a single attempt to quit. Encouraging though these cure rates are, they still do not match the rates for those in the Ama- gansett-Psychology Department populations who had never sought help. Of these, 69.2% of the 26 such fat people were classified as cured fats and 65.3% of the 75 such smokers as cured smokers. Obviously there are still far too few cases to permit solidly based speculation, but the fact that the cure rates for those who had never had help are higher than for those who had, does suggest that there remain differences still to be explained. If in future research such differences persist, it seems that the explanation must be either, as I have repeatedly suggested, in terms of the perversity of those who seek help or, alternatively, in terms of the per- versity of the therapeutic process proper-an in- triguing guess certainly, for if correct it opens up the exciting possibility that clinical psychology and psychiatry are even now as capable of inadvertent mischief as are the proven medical specialties. We may yet take professional as well as etymological pride in the neologism-psychiatrogenics. Two men who were under medical supervision are elimi- nated from the data. They had both gone to physicians because of cardiovascular difficulties, not because of obesity, and both had been advised that if they enjoyed living, they had better lose weight. One of these men was classified as a cured fat and the other as a partially failed fat. REFERENCE NOTE 1. Leventhal, H., & Cleary, P. D. The smoking problevn: A review of the research, theory, and research policies in lie- haaioral risk modification (Research rep.). Madison: Uni- versity of Wisconsin, Center for Medical Sociology and Health Services Research, 1977. REFERENCES Hammond, E. C., & Carfinkel, L. Smoking habits in men and women. Jour?al of National Caruxr Ingtttute,1971, 27, 419- 432. Horn, D. Determinants of change. In R. C. Richardson (Ed.), The second world conference on smoking and health. Lon- don: Pitman Medical, 1972. Hunt, W. A., Barnett, L. W., & branch, L. C. Relapse rates in AMERICAN PSYCHOI.OCIST • APRIL 1982 • 443

I I I I I I I I I I I I I I I I I I I addiction programs. Journal of Clinical Psychology, 1971. 27, 455-459. Hunt, W. A., & Matarazzo, J. D. Recent developments in the expe-imental modification of smoking behavior. journal of Abnormal Psychology, 1973. 81, 107-114. Kinsey, A. C., Pomeroy, W. B., & Martin, C. E. Sexual behatnor in the human male. Philadelphia: Saunders. 1948. Public Health Service. Change in cigarette smoking and current smoking practices among adults: United States, 1978. Ad- vance Data, 1979, 52, 1-16. Robins, L. The Vietnam drug user returns. Special Action Of- fice Monograph, May 1974 (Series A, No. 2). Robins„L. N., Helzer, J. E., Hesselbrock, M., & Wish, E. Viet- nam veterans three years after Vietnam. In L. Brill & C. Winick (Eds.). The yearbook of subnance use and abuse (Vol. 11). New York: Human Sciences Press, 1980. I Society of Actuaries. New weight standards for men and women. Statistical Bulletin, Metropolitan Life Insurancs Company, 1959, 40, 1-4. Stunkard, A. J. The results of treatment for obesity. New Yorlr State Journal of Medicine, 1958, 58, 79-87. Stunkard, A. J., & Albaum, J. M. The accuracy of self-reported weights. American Journal of Clinical Nutrition, in press. Stunkard, A. J., & McLaren-Hume, M. The results of treatment for obesity. Archives of Internal Medicine, 1959. 103. 79-85. Stunkard, A. J., & Penick. S. B. Behavior modification in the treatment of obesity. Archives of Cerural Psychiatry, 1979, 36, 801-806. Wing, R. R., & Jeffery, R. W. Outpatient treatments of obesity: A comparison of methodology and clinical results. Interna. tional Journal of Obesity, 1979, 3, 261-279. - -, 444 - AML 1982 • MaEIttCArr PsYCHOr.OCtsr : --. -. = -----