Skills-Building Methods to Prevent Smoking by Adolescents

JOURNAL OF ADOLESCENT HEALTH CARE 1985;6:439- 444 Skills-Building Methods to Prevent Smoking by Adolescents STEVEN PAUL SCHINKE, Ph .D., LEWAYNE D. G ILCHRIST, Ph.D ., WILLIAM H. SNOW, M.A . , AND ROBERT F. SCHILLING II, M.S.W. This study evaluated cigarette smoking prevention methods with a sample of adolescent females and males. Arranged by school, 331 informed and consenting sixth graders were randomly divided into four grou ps: 1) pretest, skills-building methods, posttest; 2) pretest, discussion methods, posttest; 3) pretest and posttest; and 4) posttest only. All s ubjects were followed for 6 and 15 months after the posttest. Outcome results on measures of nonsmoking intentions, attitudes, predictions, problem-solving abilities, and peer interactions favored subjects in the skills-building group when compared with subjects in the discussion and pretest-posttest control groups. Smoking rates at posttest and at both follow-ups were lower in the skills-building group than in the other three groups. Results from pos ttest-only subjects did not support pretest reactivity. The stud y' s strengths and limits are discussed along with directions for future smoking prevention research . KEY worms: Cigarette smoking Prevention Cigarette smoking increases risk of cancer, cardfovascular disease, and chronic obstructive lung disorders (1- 3). Smoking-related illnesses account for significant health care costs (4). For the latest reporting period, economic costs of morbidity and mortali ty from habitual cigarette use in the United States were over $47 billion (5). Methods to help reg- From the Scl100/ of Social Work, University of Waslii11gto11, Seattle. Waslii11gto11 . Address rotpr111t requests to: Steven Paul Scl1111kc. Ph.D., School of Social Work. Umi-ersity of Washi11gto11. Seattle, WA 98195. Ma1111scripl accepted December 17, 1984. ular smokers quit the habit reach too few, too late. Cessation methods are further marred by unacceptable rates of relapse (6-7). Prevention methods with adolescents are preferable for reducing health risks and costs from habitual cigarette smoking (8). Viable, effective smoking prevention programs with youths must be based on empirical research. Prospective studies need to separate attentionplacebo effects from active primary prevention methods. Data from such research can assist clinicians and researchers in making decisions about smoking-prevention methods (9,10). Pretest reactivity should be isolated in smoking-prevention studies. Results on pretest effects will add precision to explanations of behavior change from prevention methods. Postintervention follow-up is also needed in smoking-prevention research (11,12). To place confidence in the efficacy of prevention methods used with early adolescents, follow-up must span the high-risk transition from elementary school to junior high. The present study evaluates smoking-prevention methods with adolescents. Methods Subjects and design. Subjects were recruited from sixth-grade classes in four suburban middle schools. All enrolled sixth graders were invited to participate; no incentives were provided. The sample of 331 sixth graders represented 94% of all eligible students. Subjects gave their informed consent, and they obtained parental consent. The sample had a mean age of 11.64 years, and was 52% female. Chi-square comparisons across schools indicated no differences for the age and sex distribution of subjects. Cigarette smoking rates among subjects' parents-overall, re- C> Society for Adolescent Medicine. 1985 Published by Elsevier Science Publis hing Co., Inc., 52 Va nderbilt Ave .• New Yo rk, NY 10017 439 0197-0070/851$3.30 440 SCHINKE ET AL. ported for 31 % of mothers and for 28% of fathersdid not differ across schools and classrooms. By school, subjects were randomly divided into fom groups: skills-building, discussion, pretest-posttest, and posttest. Subjects in the first three groups were pretested. Subjects in the skills-buildi ng and discussion groups received smoking prevention intervention. All subjects were posttested and were then followed for 6 and 15 months. /11strunrents a11rl procedure. Pretests, posttests, and the 6-month follow-up measured s ubjects' smokingrelated knowledge, attitudes, specific and general problem-solving abilities, and interpersonal performance (13-16). The 15-month follow -up measured subjects' attitudes toward nonsmoking, general problem-solving abilities, intentions to smoke in junior high, and predictions of nonsmoking when adults (17- 19). Subjects gave 1-ml saliva samples and reported their cigarette use at every measurement period (20,21). Saliva sampling enhanced subjects' self-report accuracy; analyses of thiocyanate in the samp les revealed significant correlations between reported and biochemically measured smoking. Subjects additionally reported their noncigarette tobacco use at final foJlow-up (22,23). Intervention Pa irs of graduate socia l work students delivered eight SO-minute skills-building or discussion sessions with subjects in the respective groups. Sessions were held at school, during noninstructional school hours. Both interventions included films on the health consequences of cigarette smoking and peer testimonials on the merits of not smoking. Subjects in each group saw and analyzed slides on tobacco industry advertising. Both intervention groups ended with subjects overtly committing themselves to nonsmoking. Each subject's oral commitment to nonsmoking was sta ted in front of the group and videotaped. Skills-Building Methods Skills-building subjects additionally lea rned cognitive and behavioral steps to stop, think, decide, and act w hen tempted or urged to smoke cigarettes. The leaders introduced this four-step chain by observing that personal choices about smoking can be made on impulse, but are best when planned ahead. J\ccordingly, subjects described recent past and anticipated future personal choice situa tions. Leaders then demonstrated steps in the stop, think, decide, JOURNAL OF ADOLESCENT HEALTl-I CARE Vol. 6, No. 6 and act chain. ln subgroups and aided by leaders, subj ects practiced each step . Thus, the stopping step involved self-instructional cues when faced with tobacco use situations-e.g., "Uh oh. Rick has pulled out his pack of smokes and is going to pass them around." The second step, thinking, was similarly practiced, first aloud, then silently, as a step toward prevention- e.g., "All right, I've handled this kind of thing before. Let's see, I could split, [ coLild pretend I got a cold and can't smoke because it'll make me sick." The third step in the cognitive-behavioral chain, deciding, gave subjects a method of weighing the merits of their options and of choosing the most feasib le one-e.g., "Well, if I split, the guys will think I'm a squealer and won't want me around, saying I got a cold looks dumb and fake. OK, here's what I'U do-I'll just say 'Nah, I don't like to smoke .. .'and, if anyone hassles me, I'll teU them to leave me alone. I don' t have to smoke if I don't want to." Acting, the last step of the cognitive and behavioral chain, allowed subjects to prepare words, voice inflections, and nonverbal gestures for situations in which they would carry out their decisions. Subgroup practices enabled subjects to rehearse behavｩｯｲ｡ ｾ＠ actions while leaders provided feedback, coaching, and praise. Finally, skilJs-building subjects practiced all four steps in the cognitive-behavioral chain relative to ciga rette-use situations presented by group members . Discussion Methods Instead of learning the four-step chain covered in the skills-building group, discussion group subjects took pa rt in oral quizzes, debates, games, and skits. Quizzes patterned after television programs engaged teams of subjects in answering questions on the immediate and long-term consequences of smoking. Debates in this group were also tea m efforts as subjects argued the attraction and hazards of tobacco use_ Games involved subgroups of subjects who competed for points by identifying tobacco use health risks. Skits parodying cigarette industry advertisements were planned, performed, videotaped, and played back. Throughout discussion groups, leaders gave subjects individuaJ attention and encouraged each s ubject to participate in every activity. Results Subject attrition was nonsignificantly different by condition, across measurements. Pretests were SKILLS-BUILDING METHODS TO PREVENT SMOKING Novembe r 1985 covariates for analyses of posttest and 6-month data from subjects in the skills-building, discussion, and pretest- posttest groups. Relative to pretest- posttest group subjects, skills-building and discussion group subjects knew more facts about cigarette use at posttest and 6-month follow-up (Table 1). Subjects in the skills-building group had more favorable attitudes towa rd nonsmoking at both measurement periods than subjects in discussion and pretest- posttest groups. Compared with the discussion and pretestposttest groups, the skHls-building group had higher posttest and 6-month scores for their solutions to tobacco use problems and for their anticipation of the consequences of nonsmoking deci.sions. Table 1. Pos ttest and 6-Month Follow-Up Results Posttcst Variable and group Mean Cigarette use knowledge Skills building 5.38 Discussion 5.37 Pretest- posttest 4.23 Attitudes towa rd nonsmoking Skills building 5.36 Discussion 5.08 Prctcst- posttest 4. 12 Solut ions to tobacco problems Skills building 6.69 Discussion 5.07 Pretest- posttest 4.49 Alternatives to smoking Skills building 3. 90 Discussion 2.72 2.53 Pretest- posttest Anticipation of consequences 1.64 Skills building Discu ssio n 1.38 Prctest- pos ttest 0.84 General proble m solving Skills building 7.14 Discussion 5.09 4.49 Pretest-posttest Provision of facts Skills building 2.31 1.41 Discussion Pretest-posttest 1.17 Refusal of cigarettes Skills building 5.35 Discussion 4.47 Pretest- posttest 5.01 Reward for nonsmoking Skills building 0.75 Discussion 0.36 Prctest- posttest 0.30 1 JI < 0.05. ' 11 < 0.01. ' 11 < 0.001. ' 6-Months F Mean F 41.346•" 5.28 5.24 4.-ll 22.301< 6.94 1•" 5.36 4.14 4.46 6.362' 20.782< 4.91 4.34 3.87 6.1751• 8.798' 3.Q.l 2.60 2.64 1.116 8. IS7r 1.39 0.98 0.85 3.367" 5.03 4.35 3.86 7.184C 22.289' 441 Posttest data for suggested alternatives to smoking favored the skills-building group when compared with discussion and pretest-posttest subjects. General problem-solving scores at posttest and at 6months were h igher in the skills-building group than in the other two groups. According to ratings from an in teractive analog measure, skills-building subjects, relative to discussion and pretest- posttest subjects, at posttest were able to provide more facts about tobacco use, refuse more ciga rettes, and more often reward nonsmoking. Measurement data at final follow-up were analyzed by dependent t-tests within each of the four groups. From 6to15 months, skills-building subjects reported more positive attitudes, and posttest subjects reported more negative attitudes toward nonsmoking (Table 2). General problem-solving scores 1 increased from 6 to 15 mon ths in the skills-building and d iscussion groups. Skills-building subjects, over the same follow-up period, reported fewer intentions to smoke in junior high and reported stronger predictions of nonsmoking when adults. Pretest to posttest percentages of subjects who reported ever smoking rose 1. 9% in the skills-building group, 5.0% in the discussion group, and 2.5% in the pretest- posttest group (Figure 1). Percen tages of subjects who reported ever smoking from posttes t to 6 months d id not change for skills-build ing subjects, but increased by 4. 9, 5.8, and 5.1 % for the discussion, pretest- posttest, and pos ttes t subjects, respectively. Ever-smoking rates from 6 to 15 months rose 2.4% in the skills-building group, 11.2% in the discussion group, 9.6% in the pretest-posttest group, T able 2. Changes from 6- to 15-Month Follow-up Within-group depe ndent I 8.594< 0.98 2.60 0.61 1.116 3.619•• 4. 96 4.38 4.90 1.493 9.80 I' 0.62 0.35 0.46 2.852 Variable Attitudes toward nonsmoking General problem solvi ng Inten tions to smoke in junior high Predictions o( nonsmoking when adu lts •11 < 0.05. ''11 < 0.01. 'I'< 0.001. Skills building Discussion 2.33" 0.51 4.44< 5.22c Pretestposttest - 1.06 Posttcst -2.0-lD 0.60 -0.67 -2.901• - 1.55 - 0.53 -0,07 2.16" 1.06 1.84 1.34 442 JOURNAL OF ADOLESCENT HEALTH CARE Vol. 6, N o. 6 SCHINKE ET AL. • 12 • 0- - -0 •-· - · e O·· ·· · · 0 Sk1lls· bu1ld1ng group 0 1scuss1on group Pr etes1 -pos ttes1 group Posttes 1 g roup 10 -g 8 .¥. 0 E (/) ｾ＠ Q) > w 0 .c 6 3: (/) ti Q) 'E ::J (/) / / 4 / 0 / Q) ｾ＠ O> c: Q) ｾ＠ Q) 2 Cl. 0 Pretest Postt est 10 10 10 Posltest 6 Months 15 Months 6 M onths Figure 1. Men11 clrnnges for percentnges of subjects wlro ever smoked from pretest to posttes/, from pastiest to 6 1110111/rs, nnd from 6 to 15 months. and 11.0% in the posttest group. Noncigarette tobacco use at 15-month follow-up was reported by 4.0% of skills-building group subjects, by 14.5% of discussion group subjects, by 4.6% of pretest-posttest group subjects, and by 21.4% of posttest group subjects . Discussion Data from this study support the viability and efficacy of skills-building methods to prevent cigarette smoking among early adolescents . Sixth graders who received skills-building methods of preven- tion had higher postintervention scores on smokingrelated attitudes, problem solving, and interpersonal measures than youths who participated in discussions or youths who received no intervention. Skillsbuilding and discussion group youths knew more facts about cigare tte use than d id control group youths a t posttest and at 6-month foll ow-up. Subjects in skills-building and discussion groups outscored subjects in both control groups on a general problem-solving measure at 15-month follow-up . Relative to discussion and control group subjects, skills-building subjects had more favorable attitudes towa rd nonsmoking at posttest a nd at 6- and 15month foUow-ups, greater specific problem-solving abilities at posttest and at 6-month follow- up, fewer intentions to smoke in junior high, and stronger predictions of nonsmoking when adults a t the 15-month November 1985 follow-up. Oisrussion and pretest- posttest group subjects more frequently reported ever smoking cigarettes from pretest to posttest than skills-building group subjects. Ever-smoking reports at 6 and 15 months were lower in the skills-building group with respect to the discussion, pretest- posttest, and posttest groups. Noncigarette tobacco use was reported more often at final follow-up by discussion and posttest subjects than by skills-building and pretesil:posttest subjects. Possibly, subjects in the former two groups were more inclined to experiment with tobacco than were subjects in the latter two groups. Our results suggest advantages of skills-building methods for preventing cigarette smoking among early adolescents. The multidimensional nature of smoking in the adolescent years demands multicomponent solutions. Cigarette smoking for young people appears to be the product of affective, interpersonal, physiological, and environmental stimuli (2426). SkHls-building methods can address these stimuli. As evidenced in findings on attitudinal, problem-solving, and interpersonal measures, skillsbuilding methods may allow youths to counter prosmoking urges from within and pressures from without. Youths in the discussion group may have performed less well on outcome measures because they did not learn cognitive and behavioral skills. The relative lack of success for discussion group procedures in altering youths' measured outcomes warrants elaboration. Perhaps debates for and against tobacco use in discussion groups increased subjects' prosmoking thoughts and behavior; or, discussion groups could have contained persuasive role models to encourage tobacco use (27,28). Alternatively, these groups may have had few powerful antismoking role models- so-called "cheerleaders, jocks, and preppies." The presence or absence of such pro- and anti-smoking role models, unfortunately, was not expressly measured. These explanations therefore remain heuristic. Similarly, positive changes for skills-building subjects cannot be attributed to distinct intervention components-e.g., feedback, coaching, praise- because data were not gathered on these components (29). Other limitations of the study are also worth reiterating. Study subjects may not be representative of the general population of early adolescents. Infrequent smoking across our groups begs the question of longitudinal changes from our prevention methods. Though saliva thiocyanate results confirmed subjects' sell-reports, biochemical data were n ot used in hypothesis testing. These data were used solely to check self-reports, a nd could be misleading SKILLS-BUILDING METHODS TO PREVENT SMOKING 443 because thiocyanate can be present in foods and in ambient smoke (20,23,30-32). The study's methods, measures, and data fail to address the role and relative strength of potentially important predictors of adolescent tobacco use. These predictors include such factors as socioeconomic status, youths' academic success, athletic interests, and concerns over weight control. Moreover, group differences in ever-smoking rates from pretest to posttest could be an artifact of mutual peer support. Youths in the skills-building group may have reinforced one another for not smoking, thereby enhancing results seen at subsequent measurement periods. Finally, inconclusive differences between our two control groups shed no light on the issue of test reactivity (33,34). More smoking prevention research is needed. Studies are needed to evaluate prevention methods with youths at high risk for habitual cigarette use. These include children whose peers, siblings, or parents smoke; youths likely to quit school; those from working class, blue-collar homes; and young persons from demographic backgrounds where cigarette smoking is predominant (35- 37). The value of booster sessions in smoking-prevention programs also merits study. Regular boosters might combat ascending rates of cigarette use as adolescents go through school and into adulthood. Family members have been neglected in smoking-prevention programs. Mothers, fathers, and siblings are underused fo r nurturing and sustaining youth's prevention efforts (38-39). Perhaps the data presented here wilJ encourage refinements and replications of methods to prevent cigarette smoking among early adolescents. This work was supported by Grant CA29640 from the National Cancer Institute. The authors thank Virginia Senechal, Thomas Glynn, Catherine Bell, Cheryl Kelso, and Lois Holt. References 1. Office on Smokin g and Health. Th e health consequences of smoking: cancer: a report of the Surgeon General. Wash· ington. D.C., U.S. Government Printing Office, 1982. 2. Office on Smoking and Health. The health consequences of smoking: cardiovascula r disease: a re port of the Surgeon General. Washington, D.C., U.S. Government Printing Office, l983. 3. Office on Smoking and Health. The health consequences of smoking: chronic obstructive lung disease: a report of the Surgeon General. Washing ton, D.C., U.S. Government Printing O ffice, 1984. 4. Bell CS, Levy SM. Public policy and smoking prevention: Implications for research. In: Matarazzo JD, Weiss SM, Herd JA, ct al., eds. Behavioral health. New York. Wiley, 1984:775- 85. 444 JOURNAL OF ADOLESCENT HEALTH CARE Vol. 6, No. 6 SCHINKE ET AL. 5. Popescu CB. Smoking or health. Summit, NJ, American Council on Science and Health, 1984. 6. Cullen JW. Role of the social and behavioral sciences in cancer preve ntion. In: Cohen J, Cullen JW, Mullen LR, eds. Psycho· social aspects of cancer. New York, Raven, 1982:1946-53. 7. Lando HA. Data collection and questionnaire design: smok· ing cessation in adul ts. In: Grabowski J, Bell CS, eds. Measure· ment in the ana lysis and treatment of smoking behavior. Washington, DC, U.S. Government Printing Office, 1983:7489. 8. McCaul KO, Glasgow RE, Schafer LC, et al. Commitment and the prevention of adolescent cigarette smoking. Health Psy· chol 1983;2:353-65. 9. Crow CS. Smoking areas on school grounds: Are we encour· aging teenagers to smoke? J Ado! Health Care 198-J;5:117-9. 10. Best JA, Flay BR, Towson SMJ, et al. Smoking prevention and the concept of risk. J Applied Soc Psycho!, in press. 11. Kirscht JP. Preventive health behavior: a review o f research and issues. Health Psycho! 1983;2:277-301. 12. 131oom M. Primary prevention. Englewood Cliffs, NJ, Pren· lice-Hall, 1981. 13. Flay BR, d' Avernas JR, Best JA, et al. Cigarette smoking: Why young people do it and ways of preventing it. In: McGrath P, Firestone P, eds. Pediatric behavioral medicine. New York, Springer, 1983:132-83. 14. Bolvin GJ, Renick NL, Baker E. The effects of scheduling for· mat and booster sessions on a broad-spectrum psychosocial approach to smoking prevention. J Behav Med 1983;6:359-79. 15. Glasgow RE, Ely RO, Besyner JK, ct al. Behavioral measures of asse rtiveness: A comparison of audio and audiovisual coding of struc tured interactions. J Behav Assessment I 980;2:273-85. 16. Biglan A, Severson H, Bavry ), et al. Social influence and ado· lescent s moking: A firs t look behind the barn. Health Ed 1983;September/ October:14-18. 17. Gilchrist LO, Schinke SP, Blythe BJ. Primary prevention ser· vices for children and youth. Child Youth Services Rev 1979;1 :379-91. 18. Schinke SP, Gilchrist LO, Lodish D, e l al. Strategies for pre· vention research in service environments. Eva! Rev 1983;7: 126-36. 19. Schinke SP, Gilchrist LO. Life skills counseling with adoles· cents. Baltimore, University Park Press, 1984-. 20. Luepker RV, Pechacek TF, Murray OM, ct al. Saliva thiocya· nate: A chemical indicator of cigarette smoking in adolescents. Am j Pub Health 1980;71:1320-24. 21. Vogt TM. Questionnaires vs biochemical measures of smoking exposure. Am J Pub Health 1982;72-93. 22. Hunter SM, Webber LS, Berenson GS. Cigarette smoking and tobacco usage behavior in children and adolescents, Bogalusa Heart Study. Prevent Med 1980;9:701- 12. 23. Schinke SP, Gilchrist LO. Survey and evaluation methods: smoking prevention among children and adolescents. In: Gra· bowski J. Bell CS, eds. Measurement in the analysis and treat· 24. 25. 26. 27. 28. 29. ment of smoking behavior. Washington, D.C., U.S. Government Printing Office, 1983:96-104. Johnson CA, Solis J. Comprehensive community programs for drug abuse prevention. In: Glynn TJ, Leukefeld CG, Ludford JP, eds. Preventing adolescent drug abuse. Washington, D.C., U.S. Government Printing Office, 1983:76-114. L.rndo HA. T reatment and prevention of smoking. In: Keller PA, ed. Innovations in clinical practice, volu me 3. Sarasota, FL, Professional Resource Exchange, in press. Lichtenstein E. The smoking problem: A behavioral perspective. J Consult Clin Psychol 1982;50:804-19. Kellam SG, Simon MB, Ensminger ME. Antecedents in first grade of teenage substance use and psychological well-being . In: Ricks OF, Dohrenwend BS, eds. Origins of psycho· pathology. Cambridge, MA, Cambridge University, 1983:1742. Murray OM, Johnson CA, Leupker RV, et al. The prevention o( ciga re tte smoking in children: A comparison of four strat· egies. J Applied Soc Psycho!, in press. Durlak JA, Jason LA. Preventive programs for sc.hool·aged children and adolescents. In: Roberts MC, Peterson L, eds. Prevention of problems in childhood. New York, Wiley, 1984:103-32. 30. Pechacek TF, Murray OM, Leupker RV, et al. Measurement of adolescent smoking behavior: Rationale and methods. J Behav Med 1984;7:123-40. 31. Schinke SP, Gilchrist LO. Primary prevention of tobacco smoking. J School Health 1983;53:416-9. 32. Gilchrist LO, Schinke SP, Self-control skills for smoking pre· vention. In: Engstrom PE, Anderson PN, Mortenson LE, eds. Advances in ca ncer con trol, 1983. New York, Alan R. Liss, 1984: 125-30. 33. Flay BR. What do we know about the social influences ap· proach to smoking prevention? In: Bell CS, ed. Prevention research: Deterring drug abuse among children and adolescents. Washington, D.C., U.S. Government Printing Office, in press. 34. Wills TA. Stress, coping, and substance use in early adoles· cencc. In: ShHfman S, Wills TA, eds. Coping and substance use. New York, Academic, in press. 35. Gordon NP. De-fusing the symbolic va lue of smoking. Am J Pub Health 1983;73:1102. 36. D'O nofrio CN, Thier HD, Schnur AE, ct al. The dynamics of adolescent smoking behavior. World Smoking and Health 1982;7: 18-34. 37. Perry C. Adolescent health: An educalional-ecolo,gical per· spective. In: Coates TJ, Peterson AC, Perry C, eds. Promoting adolescent health. New York, Academic, 1982:73-86. 38. Macdonald DI. Drugs, drinking, and adolescents. Chicago, Year Book Medical Publishers, 1984. 39. Glynn TJ. From family to peer: Transitions of influence among drug-using youth. In: Lettieri DJ, Ludford JP, eds. Drug abuse and the American adolescent. Washington, D.C., U.S. Government Printing Office, 1981:57- 81.