Tobacco Documents Online

I I I I I I I I I I I I I I I I I Contents Rzewnicki and Forgays on Schachter .................. 97 Wolpe on Lazarus ............. 100 Lazarus Replies ............... 101 Alvarado on Danziger .......... 101 ,viountjoy on Kelly ............ 102 Johnson and Morris on Public Information on Research With Nonhumans ................ 103 Gibbs and Schnell on Eichorn .... 104 Franzini on Anonymous Review .. 104 Riordan and Marlin on Unethical Research Practices ........... 104 Recidivism and Self-Cure of Smoking and Obz'sity: An Attempt to Replicate Randy Rzewnicla and Donald G. Forgays University of Verntont In 1982 Stanley Schachter published an article that questioned the prevailing ste- reotypes concerning the difficulty of smoking cessation and weight control. Schachter presented evidence obtained from two small samples in the New York metropolitan area that suggested quite clearly that, contrary to expectations, many people are able to stap smoking or to reduce weights to non-obese levels, and, tpparently can do these things without ti- tanic effort. Reviews of interventional programs dealing with smoking oessation and weight control present a rather disappointing picture. For example, Leventhal and Cleary (1980) stated that intervention for smoking is effective in heiping people to stap smoking but faiis at maintenance of the stoppage: Fewer than 20% of smokers M"e greatly reduced their smolang or quit a Yeat' after intervention. This low degree Of effectrveaess appears frequeatly in the Smoking management literature. The pia- tae appeats to be as bleak for weight re- duction in the obese. Extensive literature m'icvvs of relevant interventional attempts Comment have been published by Stunkard and McLaren-Hume (1959) and Wing and Jef- frey (1979), among others. According to these summaries, only 20% of the obese can expect to lose as much as 20 pounds as a result of the interventional process. The dismal results in both types of therapy do not support the notion that we know how to manage these behaviors. All the more startling, then, are Schachter's results. He reported that as many as 60% of heavy smokers and about 60% of the obese he studied are able to quit or greatly reduce smoking or to lower their weights to nonobese standards-and, in most cases, without any interventional help. Schachter offered as possible expla- nations of the differences in results (a) the self-selective nature of the persons seeking intervention,(b)thelikelihoodthatforin- terveation seekers it is a single attempt to modify behavior whereas the interview respondents may have made several at- tempts to stop smoking or control weight, and (c) the interventional processes may have been completely ineffective (Schach- ter called them pervase). It is impossible at this time to eval- uate which of these explanations is the most tenable. Nor is it important to do so before a replication of the study is accom- plished. It is to this end that the present study was directed. Method Following the procedure of Schachter to interview the entire membership of se- lected groups, rather than a randomly se- lected population, this study entailed in- terviewing most of the members of the Psychology Department of the University of Vermont. This membership was defined as those having a mailbox in the depart- meat in the fall of 1984 and whose mail was picked up at least once a week during October. Nine members did not pick up their mail in the time allocated, and four members refused to take part in the study. Thus, we interviewed 96% of the 105 de- partmental members, as defined above. Of the 92 interviewed, 51 were graduate stu- dents, 26 were Faculty, and 15 were sec- retaries, technicians, and support staE Forty-nine were females and forty-three January 1987 • American Psychologist were males. Their ages ranged from 22 to 80 (an emeritus professor!), averaging 34.89 years (the age range for Schachter's study was 16 to 79). This group is largely academic, most are from urban back- grounds, and there is some ethnic diversity among them. The interviewing was con- ducted by the first author from January to March, 1985. The interview followed Schachter's procedures as closely as possible. Infor- mation that was not complete enough in Schachter's article (1982) to allow repli- cation was obtained directly from him. The interview began by obtaining the sub- ject's estimate of height and weight and ended by obtaining the actual weight of each willing subject on a digital scale. In between, the entire smoking history from the time of the first cigarette and the entire weight history from the time of full height was obtained. Probing was done through- out to assure completeness. Any attempts to quit smoking or reduce weight were ex- amined in detail. In a few cases additional information was obtained in follow-up visits or calls. Results Smokiug Scbachter had interviewed nearly the en- tire membership of the Columbia Univer- sity Psychology Department (N - 83) and 78 entrepreneurial and working persons from a small resort community in eastern Long Island. Because we had obtained in- formation only from the University of Vermont Psychology Department mem- bership, we tried to compare only the two departments. Because of the way that Schachter displayed his data in the 1982 article, this was not always possible. We have attempted in each case to make clear which samples are being compared. When the two departments are being compared, they are referred to as CU (Columbia Univezsity) and UVM (University of Ver- mont). When Schachter's combined two samples are compared to UVM, they are referred to as NY (New York). Schachter defined a heavy smoker as one who has smoked 15 or more cigarettes per day for at least one year and a light 97 I

I I I I I I I I I I I I I I I I smoker as one who has smoked less than 15 cigarettes per day for at least one year. In the NY population, 94 of 161 inter- viewed were classified as smokers; in the UVM population. 52 of 92 were so clas- sified. The percentage of smokers in the NY and UVM populations are essentially the same at 58.4% and 56.5%, respectively. Following Schachter's definitions of heavy and light smokers, characteristics of the NY and the UVM smoking popula- tions are given in Table 1. As seen there, the two populations are similar in per- centage of heavy to light smokers, number ears e of y r g an d ng, an i k o years o smo f f smoked. Although the two populations are similar for light smokers, the heavy smok- ers in NY smoke more than do those in the UVM population. Thus, although the two populations are similar enough to compare in other features, it should be noted that NY heavy smokers smoke more than UVM heavy smokers. Table 2 presents data on the present and past smoking status of the two pop- ulations, employing Schachter's classifi- cations of cured smoker (former smokers who are presently not smoking although they may have a cigarette on infrequent oaxsions), failed smoker (present smokers who have tried to quit), indifferent smoker (present smokers who have never tried to stop), switched smoker (former cigarette smokers who now smoke cigars or pipes), and cigar or pipe smoker (those who smoke cigars or pipes at present and/or in the past but have not smoked cigarettes). As seen there, CU members were more successful at quitting smoking than were the UVM membets. About 70% of the CU male and female heavy smokers success- fully quit whereas only about half of the UVM male and female heavy smokers did so. All of the CU male light smokers who tried to quit were unsuccessful whereas 70% of the CU female light smokers were successful. For UVM 80% of the male and 70% of the female light smokers quit suc- cessfully. Perhaps more striking is the finding that every smoker at UVM tried to quit at some time whereas about 80% of the CU members did so. Schachter re- ported that the overall percentage of sue- ce.ssfiil quitters for the CU population is 65.8 and that for his Long Lsland group 61.5. We found the corresponding per- centage for the UVM population to be 57.7. The NY samples of heavy smokers were better able to quit than the UVM heavy smokers, whereas the UVM light smokers were better able to quit than the NY samples. We subjected the success/ failure data for the CU and the UVM groups to several two by two chi-square analyses, treating sex and amount of Table 1 Characteristics of the NY and the UVM Smoking Populations N Mean number of years smoked Rang. of years trtbMcetl Mean rnxnber of agarettes smoKed detly Range of ogareKea srraketl aalh/ Type of smoKer NY UVM NY UVM NY UVM NY UVM NY UVM Heavy 73 37 17.4 18.9 2-50 4-48 32.9 20.6 17-90 15-60 Light 21 15 8.2 9.6 1.5-32 1-33 6.8 7.3 1-12 1-14 Note. NY - CambrW aamp/e from a l.ang lalrid rsaort cortxrasrty and Cokrnbba UrWersrty PsycP1o/ogy Deprtrnent. UVM . UMVersrty of Vermortt Psyefl0b~nr OeperaneMt smoking separately, and none of these four analyses was statistically significant at the p < .05 level. On the average the sucttssful quitters at UVM had not smoked for 9.2 years, close to the figure of 7.4 years for CU. The percentage of former smokers at UVM who had not smoked for over one year was 89.3, and 94.4% had not smoked for three or more months. These data are compa- rable to the figures of 87.8% and 98%, re- spectively, reported by Schachter. Thus, the data for smoking at the two institutions appears to be quite compa- rable. However, at least small differences between the two populations exist, in that NY has a somewhat greater success rate for quitting, more indifferent smokers and more who switch to other tobacco forms, a somewhat greater success rate for quit- ting for heavy-smoking males and females, and less of a success rate for light-smoking males. Schachter had asked his cured sub- jects how difficult it was to quit and whether they had experienced any craving or withdrawal symptoms, including irri- tability, sleeplessness, cold sweats, weight gain, and so on. He reported that light smokers had little difficulty in quittiag. Only two of the NY light smokers reported problems whereas the other 15 (88.2%) reported none. For his heavy smokers, 45.8% experienced major difficulties, 25.4% had minor difficulties, and 28.8% had none. UVM light smokers reported more difficulty quitting than the NY groups with 37.5% describing minor and 12.5% indi- cating major problems. Heavy smokers at UVM similarly had more difficulty quit- ting than the NY group, with only 7.9% reporting no difficulty, 50% reporting mi- nor difficulty, and 42.1 % indicating major difficulty. We wished to compare statistically the ease of quitting smoking between the UVM and the NY data. Because Schach- ter had not reported all three levels of his difficulty scale for light smokers, only easy veisus the remaindet; we collapsed our moderate and difficult categories to make this comparison, X2 (1) = 13.09, p < .01, between the proportions of subjects in these two categories between the two sam- ples. Light smokers found it easier to quit inNYthanatUVM. For heavy smokers, comparing the NY and WM samples, the difficulty srale yielded the following: x2 (2) = 21.58, p< .01. More NY than UVM smokers re- ported that their attempt to quit was dif- Tab{e 2 Present Smoking Status of Smokers in the Two Psychology Department Populations Curs d Fd.d MaM Fertr/G Ma le Fntile Mw FartrM Mw FemaN Typa a srnoker CU UVM CU wM CU trvM CU uVM cu UvM CU wM CU wM CU : UVM Heavy 13 12 5 7 6 10 2 8 1 0 3 0 3 0 0 0 ~ Lght 0 4 7 7 2 1 3 3 3 0 0 0 1 0 0 0 i , nia.. cu - caurnaa uv." Psycrro+oqy oqrane,t. UVM - Ur*Mrs:y of V«mont Psycho+oQy o. Narwno/c.n at CU: 13 mds and 1 7 fertWN. Nornmokers at UVM: 13, Waa and 26 kmaks. Ciqar a pip. amokux 4 at CU. 1 at UVM, ri mda. J a 98 January 1987 • American Psychologi9

I I I I I I I I I i I I I I I ficult. Even though the NY and UVM samples reported about as much severe craving and withdrawal symptoms. 45.8% and 42.1% of heavy smokers, respectively, the full comparison still yielded a signifi- cant X= between the two samples. XZ (2) = ;0.21, p < .01, with NY having many more persons than WM reporting no ,a,ithdrawal symptoms and no craving. tain the following: X2 (1, N= 28) = 11.50, p<.01. Clearly, the success rate for losing weight for UVM obese individuals is muoh much more successful at losing weight than that at UVM. less than that for CU obese individuals. Discussion The weight histuries of obese mem- What we reported above is a partial rep- bers of the NY and UVM populations by lication of the Schachter findings on sex are given in Table 3. Data were not smoking and obesity. The NY and the available by sex for the two NY samples, UVM populations are similar in the dis- and thus, the CU and LTVM groups could tribution of heavy and light smokers, not be compared directly. As seen in Table number of years and the range of years of Obaity 3, there appear to be important differenas smoking, and the percentage of each pop- As in Schachter's study, the weight data in the data representing the two popula- ulation that smokes. However, heavy were obtained by self-report. To assess the tions. A substantial portion of the NY smokers in NY smoke more than those at validity of these reports, we weighed con- obese males were sucxessful in reducing UVM. Both studies report a smoking quit senting interviewees at the end of the in- their weight to within nonobese levels, rate that is about three times greater than terview. Of the total of 92 subjects, 72 or whereas only one obese UVM male was that reported for most therapeutic inter- 78.396 agreed to be weighed. This is some- able to do so. A chi-square of the data re- ventions. The CU heavy smoking sample what higher than Schachter's 59.5%. The flecting all five categories for the two pop- was more successful at quitting smoking self-reports turned out to be reasonably ulations of males was significant, X2 (4, than was that of UVM. Howeven the accurate, on average being 4.4 pounds N= 35) = 10.80, p < .05. NY females nearly 50% quit level for this group is still (2.5%) discrepant from scale weight. This appear also to be more successful at weight very good indeed, and higher than what we finding replicates Schachter's results and management than the UVM females, and see so firequently in the reports of clinical Stunkard and Albaum's (1981) report on this difference is also significant, Xz (4, N= intervention studies of heavy smokers. The accuracy of weight self-reports. 28) = 11.01, p < .05. Another difference CU light smoking males are less successful Schachter defined obesity as 15% or between the two populations is the number at quitting smoking than are those at more over the average weights found in of indifferent male subjects; NY has six UVM. The females at both institutions are the tables of the Society of Actuaries such subjects and UVM none. The per- equally and highly successful. The years (1959). Cured fat individuals are those centages of overweight individuals both at of successful nonsmoking are quite com- who have been obese, have lost at least prediet and at interview time are not parable for the two populations. Howevet; 10% of their prediet weight, and are no markedly different for the two populations, both light and heavy smokers in NY report longer obese. Panially cured fcu individuals and so the sua.~ess-rate differences seem that it is easier to quit than do those at have been obese and have lost 1.0% or more to be due to other factors. UVM. Because virtually all of the com- of their prediet weight, but are still 15% Schachter reported that the obese in parisons made of the two populations are or more averweight. PartiaUy failed fat in- his study who attempted to lose weight not statistically signiflcantly different (only dividuals are persons who have been obese lost more weight than subjects in the clin- the quit rate difference for light male and have lost less than 10% of their prediet ical studies summarized in Stunkard and smokers approaches statistical signifi- weight but are no longer obese. Failed fat Penick (1979). His male subjects lost an cance), we can conclude that the data for individuals are obese persons at interview average of 26.8 pounds in the 9.7 years smoking results are similar for the two who had lost less than 10% of their prediet since they started dieting, and his female populations. weight. Indifferent fcu persons are those subjects 24.8 pounds in 7.5 years. Each of Both studies have found a quit rate obese who have never attempted to lose these averages is considerably higher than for all smokers of approximately 60%. Be- weight. Norntals are those who have never that reported from the clinical studies. AU cause both studies dealt with samples that as adults been 15% or more overweight. of the obese males in the UVM population were very close to the size of their popu- Of the 18 obese persons in the CU tried to lose weight and were able to lose lations, we can conclude that the 60% quit Population who attempted to lose weight, an average of 14.8 pounds in the 10.6 years rate is more representative of actual suc- 13 were classified as cured fat, 3 as par- since they began. For UVM females, the cess than the much lower figures usually tia11y failed fat, 2 as failed fat, and none figures are 23.2 pounds for an average of reported in the internntion literature. as putially cured fat. Eighteen of the 92 9.24 years. Thus, our female weight loss We have provided somewhat less UVM interviewees reported a history of data corresponds well to Schachter,s support for Schachter's obesity findings. obesity, and 17 of them had attempted to whereas our mak data does not. There were many more fat subjects in the ~ ge lose 2~t Three were classified as cured Schachter examined separately his ~~~tion who were able to mana partially cured fat, 2 as partially subjects wl~ were at least 30% overvva ght.. ~"'~~t than in the UVM poQttla- failed fat, and 10 as failed fat. Thus, per- He found that these 11 "grossly" over- tion-about two and one-half times as saos in the LJVM population did not enjoy maay. Both males and females in NY were the sucass at losing weight that the CU weight persons (6.8%) were as sucoessful more suxesful at losing weight than their population did. at losing weight as those less obese. They UyM countetparts. These differences are Combining the cured and partially lost an average of 46.7 pounds; seven were all statistically significant. Because the de- cured fat and the failed and the partially class;ified as cured fat and only one as failed gree of overweightness is similar in the two 6iled fat data and comparing the CU and fat. For UVM, there were eight "gtOssly" populations, unknown factors must aa chi-square tests rc- obese (8.7%), 3 males and 5 females, who count for them UVM populations b It wuuld be informative . y ~ Kals a significant differencx between tbe averaged 47.6% overweight- Only one of and likely would contribute to theory t1"o groups, X2 (1, N= 35) = 6.41, p< the UVM obese individuals was in the building in this area to study further tIu .05. If we dr% the two partial classifica- gi'oup of cured fat at interview, whereas 5 interesting location and sex differences re- ao°s aad compare only the cures and fail- were in the failed fat group. Thus, the NY ported above. ~ ura for the CU and UVM groups, we ob- sample of "grossly" obese persons was The accuracy of the retrospective Jwuary 1987 • American Psychologist 99  - -

I I I I I I I I I I I I I I I I I I I Table 3 Weight Histories of Formerly and Currently Obese Members of the NY and UVM Populations N Predbt we0ght % oysrweught W®igM at fntervlew % overweght• No. yean snos ceqrr,ng we+qnc loas resurroa Category NY UVM NY UVM NY UVM NY UVM NY UVM NY UVM Male Cured fat 14 1 214.4 207.0 29.4 16.0 175.3 162.0 1.2 -9.0 13.4 19.0 Partially cured fat 0 1 308.0 - 77.0 - 211.0 - 21.0 - 16.0 Partially failed fat 2 1 179.5 210.0 11.9 18.0 168.5 203.0 4.6 14.0 3.6 7.0 Failed fat 5 5 190.4 240.8 31.8 25.0 191.9 241.8 30.2 23.4 1.6 6.5 Indifferent 6 0 - 200.7 - 21.9 - - - Female Cured fat 11 2 158.5 168.5 27.3 30.0 129.5 115.0 -0.6 -10.5 8.3 11.0 Partially cured fat 4 1 191.5 240.0 57.8 66.0 153.3 170.0 24.1 17.0 2.3 6.0 Partially failed fat 3 1 156.0 155.0 16.8 24.0 148.3 147.0 8.6 18.0 5.7 6.0 Failed fat 1 5 185.0 176.4 23.3 34.8 198.0 176.4 26.9 33.6 25.0 9.8 Indifferent 0 1 - 165.0 - 38.6 - - NoN. NY - Cmntxrtl sampa hom a rasort oornnxity n Long bNnd ano Cofunbia lhtirsnity Payofwbgy oepertrn«x. UVM = Univsnity of Vhmront PSyehdogy oep.rv,"rx. ' Thaea perarap. orerwaiplx fipLra ars a&usted 1or .pa. a«c. and hd¢x ot subjaa n th. two poptYtlornm data collection technique used by Schach- ter (=d lry us) has_beea questioned by Jef- frey and Wing (1983). They displayed both smoking and obesity data based on lacgc national samples, which show lower levels of sucass than Schachter had reported: Schachter (1983) has suggested that their obesity data are questionable but that their smoking data may be more accurate than his. Our own data are more in line with Schachter's than with those of Jeffrey and Wing, especially for smoking. Thus, Scfachter's @eaeral finding that smoking and being ovetwaght may be self- cured in many petsons has been confirmed in Iarge measure by the present results, which are based upon a nea*ly complete sample of a known population. Jeffre3, R. W & Wiag, R. R. (1983). Raidfvism and selfcure of smoking and obesity: Data from population snidiat. Amerfcmt PsyChal- ogist, 38,852. Leveathal, H., & Cleary, P. D. (1980). The smoking problem: A review of the reaarch and theocy in behaviocal risk modif.ation. Ps}riiological Bulletirt. 88, 370-405. Schacttte~ S. (1982). Rocidavism and seif-Gttts of smoking and obesty. Amerim Psyrhoi- ogist. 37, 436-444. Scitschtez S. (1983). More on raad'rrism. Americmt Psyrbologist. 38. 854-855. Society of Actuaries. (1959). New weight sttm- dards for men and women. staristtccl su!- letin. Meaepolitan Life Instamice Camatx 40. 1-4. Stunkard. A. J., & AIbaum, J. M. (1981). Tbe ; axuacy of sdf-reported weighm Anuricaa ~ Journal of Clinical Ntaritiat. 34, 1593-1599. ,' Stunkard, A.1., & McLarea-Hume, M. (1959~ ~ The results of treatment for obesity. ArchfMet Of Inttrnal Medicine, 103, 79-85. ~ Sr.tnkard, A. J., & Penick, S B. (1979). Behaviar' rGodiScation in the treatmeW of obesity. Ara ch:res of General Psyritiatryt 36, 801-806. I Wing, R R. & Jeffrcy, R W. (1979). Outpatimw treatments of obesity: A comparison or; metttodoiogy and ciinipl results. Interrar tional Jounwl of Obtsit;t .i, 261-279. I Coc:esponaeace 0womnin= this com>zxae should be addrested to Dooald G. Forgays, Dtr parttaeat of Psy+chology, Unixssity of 1k~ Btuli»gtcn, VT 05405