Surgeon General's Report

  Other variations in intervention approaches have
used media supplements and involved students' par-
ents. Flay et al. (1987), for example, used a five-day
smoking-prevention curriculum in junior high school
classrooms and coordinated it with five different five-
minute video segments aired on a local television station.
The focus of these television segments was smoking pre-
vention, and they were followed the next week by five
more segments dealing- with smoking cessation (Flay et
al. 1987).

  Pentz et al. have trained health, science, and social
studies teachers to deliver a social influences program
that was reinforced by 10 homework activity sessions
involving parents and other family members in role play-
ing and other forms of behavioral rehearsal (Penk, Dwyer,
et al. 1989). In a related project, this group has developed
a component that asks parents to attend organizational
meetings, support school activities, and participate in an
educational workshop (Penk, MacKinnon, Flay, et al.
1989). The results of these studies are discussed later in
this chapter, along with other community programs.
  Biglan, Glasgow, et al. (1987), have also designed a
component that tries to enlist direct parental support of
their standard classroom curriculum. The component
relies on a set of four mailed messages for parents of
participating students. These messages reinforce class-
room activities, encourage family discussions of smoking
in general, and urge parents to establish family policies
regarding smoking.

  Walter, Vaughan, and Wynder (1989) embedded
smoking education in a comprehensive school health
education program, the Know Your Body Program, with
fourth- through eighth-grade students in New York. This
more comprehensive program had a significant impact
on multiple risk-related behaviors, including cigarette
smoking.

  Finally, Cain, Dudley, and Wilkerson's (1992)
"Tar Wars" program has used health professionals
to deliver antitobacco messages with the help of
fourth-, fifth-, and sixth-grade children. The students
participate in a poster contest to counter the messages of
tobacco advertising, and a communitywide media cam-
paign complements the school program. Originating in
1977, this program is based on the DOC program
Superhealth 2000, which similarly emphasized
counteradvertising skills among 7th- through 10th~grade
students (Blum 1980).

  A number of recent reviews have closely examined
issues related to program design and content (Botvin and
Wills 1985; Flay 1985; Glasgow and McCaull985; Hansen
1992). Rather than replicate these efforts here, the next
section will provide examples of the range of programs
that can teach adolescents the skills needed to resist social
influences to smoke.

Exemplary Programs for Resisting Social
Influences

Social Inoculation

  In the mid-1970s Evans et al. developed the first
prevention program that instilled adolescent skills to
resist social influences to smoke. The program, described
as "social inoculation," taught students methods for rec-
ognizing and coping with pressures to smoke from peers,
family, and the media fMcGui.re 1964). The program's
hypothesis was that if young adolescents received class-
room "inoculations" of "peer pressure," for example,
and learned how to deal with it, they would be more
prepared to resist actual social pressure from peers.
Additional emphasis was placed on the immediate physi-
ological impairments that smoking produces, rather than
on long-term consequences (Evans et al. 1979). The
program used videotapes of nonsmoking peers to im-
part information and to teach skills needed to resist social
influences. In the pilot study involving 750 seventh-
grade students, the proportion of nonsmokers in the
experimental group who 10 weeks earlier had reported
smoking at least one cigarette was approximately half
that of those in the control group.

  This research group introduced a notable proce-
dure for enhancing the validity of self-reported smoking
behavior among study subjects. Students were shown a
film indicating that their smoking status could be veri-
fied biochemically by analyzing a sample of their saliva.
The perception that the samples could be examined led
to more truthful reporting by students and thereby de-
creased misclassification bias due to inaccurate self-re-
ports (see "Validity of Measures of Smoking," Appendix
2, in Chapter 3).

  Although interpretations of results from this early
work were complicated by a variety of methodological
flaws (Flay 19851, Evans' work provided the foundation
for much of the smoking-prevention research that fol-
lowed over the next decade.

Project CLASP

  Later in the 197Os, McAlister et al. (1980) developed
anintervention&ledCounselingLeadershipAboutSmok-
ing Pressure CLASP), during which peer leaders from
high school were trained to help junior high school stu-
dents develop the skills needed to resist social pressures to
smoke. The students learned to identify social pressures
and then rehearsed and modeled strategies for coping
with them McAlister et al. 1980).

  Besides this use of older students as peer leaders,
the use of behavioral rehearsal methods and strategies to
enhance commitment to nonsmoking was an innovation

126 Prevention


preventing Tobacco Use Among Young People

that has been incorporated into many of the prevention
programs developed later. The intervention consisted of
three sessions delivered on consecutive days, followed
by four booster sessions delivered over the remainder of
the seventh-grade school year. Nine months after pre-
test, 5.6 percent of the treatment group and 9.9 percent of
the control group reported smoking during the previous
week-a statistically significant 56 percent difference be
tween the groups. These reductions in smoking preva-
lence were observed up to the 10th grade.

Life Skills Training

  botvin (1986) has developed another variation of
the social influences approach that includes resistance
skills, behavioral rehearsal, role playing, self-control, de-
cision making, problem solving, and self-reward, as well
as components devoted to increasing self-esteem, self-
confidence, autonomy, and assertiveness. The program,'
called Life Skills Training, includes various aspects of
cognitive-behavioral psychological training. The pro-
gram consists of 15 to 20 sessions for seventh-grade
students; booster sessions are given in the eighth and
ninth grades. The specific objectives of the program are
to teach skills that help students resist direct pressures to
smoke; to enhance students' self-esteem, self-mastery,

and self-confidence in order to decrease their susceptibil-
ity to indirect social pressures to smoke; to prepare stu-
dents to cope with anxiety induced by social situations;
to enhance students' knowledge of the actual prevalence
of smoking among adolescents and adults; and to pro-
mote attitudes and beliefs consistent with nonsmoking.

  This program has been evaluated extensively in
progressively larger studies over the past decade; the
encouraging results have ranged from 40 to 80 percent
reductions in smoking prevalence, and long-term effects
havelasteduptofouryears(BotvinandDusenbury1989).
In the most comprehensive evaluation of the Life Skills
Training program to date, 56 schools in three different
geographic regions were randomly assigned to three
study conditions: Life Skills plus one-day teacher train-
ing, Life Skills plus video training for teachers, and a
control condition. Significant positive effects were re-
ported for cigarette use (see Table 5) and for smoking-
relatedknowledge,attitudes,andnormativeexpectations.
In most cases, the two treatment conditions had similar
results; students in both groups demonstrated more posi-
tive effects than students in the control group (Botvin et al.
1990). The effects of the Life Skills Training program have
been demonstrated when the program hasbeen delivered
by project staff, older peers, or regular classroom teachers.
These effects have also been demonstrated on inner-city

Table 5.  Outcomes of the Life Skills Training (LST) program: adjusted third-year follow-up mean for
   smoking-related knowledge, expectations, personality measures, and behavior

Adjusted mean scores*

                            LST             LST
                         (with          (with
Smoking variable           teacher training)    video training)
Knowledge
Smoking prevalence               1.10*            1.16+
Smoking consequences              4.W             4.60A
Smoking acceptability              1.495            1 .52A
Normative expectations
Adult smoking                  3.92+            3.95+
Peer smoking                   3.80'            3.77+
Personality measures
Self-esteem                    34.25+            34.07
Self-efficacy                   19.27            19.20
Social anxiety                 28.71t           29.36
Smoking behavior                 1.46"            1.50$

Source: Botvin et al. (1990).
`Means for LST groups differ from control group at + p < .05, fp -C .Ol, _p < .OOl, and "p < .OOOl.

Control





  .93
4.13
1.37


4.22
3.92


33.65
19.26
29.92
1.63

Prevention 127

161-235 95-5


Surgeon General's Report

populations of predominantly Hispanic (Botvin et al.
1992) and black (Botvin et al. 1989; Botvin and Cardwell
1992) adolescents.

The SODAS Model

  Several researchers have developed a variation of
the social skills training approach that adds to the basic
components of resistance skills, behavioral rehearsal, and
role playing. The additional components focus on self-
control, decision making, problem solving, and self-
reward. Using a problem-solving approach called Stop,
Options, Decide, Act, and Self-Praise (SODAS), students
are taught self-control skills for smoking prevention
coupled with self-reward for personal successes (S&i&e
et al. 1986; Gilchrist et al. 1986).

  This research group has conducted a variety of
studies evaluating this intervention model in different
settings and using varied delivery modalities. The results
of these studies have consistently demonstrated that treat-
ment students reduce their smoking prevalence more
than control students and that treatment students have
greater positive changes in smoking-related know-
ledge and attitudinal factors (Schinke and Gilchrist 1984,
1985,1986).

The Waterloo Smoking-Prevention Program

  Investigators at the University of Waterloo (Ontario,
Canada) have carried out a series of largescale, long&
dinal studies evaluating the efficacy of an intervention
that teaches sixth-grade students the skills they need to
resist social influences to smoke. This intervention is
based on an integrative model of attitude and behavior
changes surrounding health issues that suggests that if
information is attended to, comprehended, and accepted,
it may lead to changes in beliefs. Beliefs, however, will
not necessarily lead to changes in attitudes, and attitudes
will not necessarily lead to changes in intentions unless
values, expectancies, and social influences are consid-
ered. Lastly, intentions will not necessarily lead to changes
in behavior unless the individual has the requisite con-
trol and coping skills (Flay 1986).

  The intervention program has three main compo-
nents that are delivered to sixth graders in six one-hour
weekly sessions. The first component provides informa-
tion on the consequences of smoking and the reasons
that adolescents smoke. The second component exam-
ines social influences-including family, friends, other
peers, and the media-that promote smoking; students
then learn specific skills to resist these pressures. In the
third component, the students are asked to integrate
information learned in all previous sessions in order to
make a decision about their future smoking behavior

and to publicly commit to nonsmoking, if that is their
decision.

  In the first large-scale randomized trial of this pro
gram, 22 schools were randomly assigned to treatment
and control conditions. Sixth-grade students in the 11
treatment schools received the curriculum plus booster
sessions in seventh and eighth grade. Initial evaluation
results indicated that although the intervention did not
reduce levels of regular smoking or significantly increase
the probability of remaining a nonsmoker, it prevented
the onset of experimental smoking through the end of the
eighth grade. The results were particularly encouraging
for students who were at highest risk of becoming regular
smokers because they had tried smoking in grade six or
because their parents, siblings,`or friends were smokers
(Best et al. 1988).

  The University of Waterloo research group has
reported six-year follow-up data for the same cohort of
students studied earlier through the eighth grade. Ninety
percent of the students were located for this follow-up
study, and data were obtained from over 80 percent of
them. These students had not received any additional
intervention after the eighth grade. The significant inter-
vention effects observed in this cohort after the eighth
grade had begun to disappear by the fifth year after the
intervention; by the sixth year, there was no longer a
significant difference between treatment and control stu-
dents (Flay et al. 1989). These results (see Figure 2)
suggest that the initial positive impacts of such interven-
tions may dissipate over time (Kozlowski et al. 1989),
particularly if intervention activities and booster sessions
do not extend throughout middle school, junior high,
and high school (botvin and Botvin 1992). School-based
programs may also be strengthened by supplementary
intervention activities that extend beyond the school con-
text into the community (Perry, Klepp, Shulk 1988; Perry
et al. 1992).

The Minnesota Smoking-Prevention Program

  The Minnesota Heart Health Program is a
community-based cardiovascular disease prevention pro-
gram that has been carried out in selected Minnesota
study communities during the past decade (Blackburn et
al. 1984). As a part of this program, the Minnesota
Smoking-Prevention Program (MSPP) has addressed the
prevention of tobacco use by influencing the social and
psychological factors known to promote the onset
of smoking.

  The activities in MSPP are often led by peer (sarne-
age) leaders who are trained to communicate the social
and psychological messages embodied in the program.
The students first form small groups to discuss
the short-term, social consequences of smoking. By

128 Prevention


prtwrz ting Tobacco Use Among Young People

Pxarninirtg actual data and discussing young people's
tendency to overestimate smoking prevalence, students
lenm that smoking is not a normative behavior in our
++ety. After exploring why adolescentssmoke, students
J~SS positive alternatives to smoking. Students then
learn how these misperceptions about smoking are estab-
lish& in our culture through advertising and role model-
ing by peers and adults. Students practice the skills to
resist the social influences that promote smoking, in-
&ding peer influences and advertising techniques.
gear the end of the program, students state a goal to
remain nonsmokers

  In evaluating the effects of the MSPP in eight junior
high schools, Murray et al. (1988) reported that after four
irears, the peer-led social influences intervention reduced
;he incidence of daily and weekly smoking by 35 to 50
percent. In contrast, no reduction was observed in an

adult-led group that was taught the health consequences
of smoking or in a comparison group enrolled in an
existing curricuhun covering general health topics. These
differences, however, were no longer statistically sig-
nificant at the five- and six-year follow-ups (Murray
et al. 1988).

  As part of this overall research program, the Class
of 1989 Study was established to test the efficacy of the
MSPP approach when introduced as part of a broader,
community-based health promotion effort (Perry et al.
1992). Researchers hypothesized that the school-based
intervention program would have longer-lasting effects
if it was introduced in communities where adults were
involved in communitywide smoking-cessation pro-
grams, where antismoking ordinances in the schools and
public community spaces were being considered, and
where integrated school and community intervention

Figure 2. Six-year follow-up of the first Waterloo School Smoking Prevention Trial: proportion of
   subjects smoking regularly and experimentally at each wave of the study

45

40

35

O Experimental

* Regular

-  r

Control -Program Control Program Control Program Control Program Control Program Control Program Control Program
(End of Trial)     Wear 1       (Year 2      (Year 3      (Year 4       (Year 5      (Year 6
       follow-up)    follow-up)    follow-up)    follow-up)    follow-up)   follow-up)

Study condition

Sauce: Flay et al. (1989).

Prmention 129


Surgeon General's Report

activities were offered. Throughout junior and senior
high school, smoking prevalence was significantly lower
among students in the intervention community than
among students in the control community. The results of
this study are discussed later in this chapter, along with
other communitywide programs.

International Research on Smoking-
Prevention Programs

  Intervention studies reported in the English-
language literature outside the United States concentrate
primarily on school-based interventions directed at sec-
ondary school students (persons aged 11 years or older).
In many cases, these intervention programs have adopted
some elements of U.S. school programs in order to reflect
different local conditions. This section reviews several of
the more rigorously evaluated programs and pays par-
ticular attention to programs that have been followed up
for two or more years after intervention.

Western Australia

  Armstrong et al. (1990) conducted a large random-
ized trial evaluating peer- and teacher-led social influ-
ence programs among 12- and 13-year-old students in
Western Australia. The authors used the MSPP program
(Arkin et al. 1981) and resurveyed the students one year
and two years after the intervention. Although the ef-
fects of the program were not strong, at the two-year
follow-up, the smoking prevalence in the control group
was 6.6 percent higher than in the teacher-led interven-
tion group and 8.1 percent higher than in the peer-led
intervention group.

North Karelia Youth Project

  The North Karelia Youth Project in Finland (part of
the International Know Your Body study) was a two-
year controlled trial that targeted schoolchildren in grade
seven (12 and 13 years old) and included components on
smoking prevention, physical activity, and reduction of
dietary fat and alcohol consumption (Puska et al. 1981,
1982). The smoking intervention program was peer-led
and involved three 45-minute sessions for grade seven;
these students received seven shorter sessions the fol-
lowing year (a schedule similar to that of Project CLASP).
The program included sessions on social pressures to
smoke, ways to resist such pressures, ways to cope with
social anxiety, the short- and long-term health effects of
both active and passive smoking, and the impact tobacco
growing has on the environment.

  Health educators from the project team delivered a
direct, intensive intervention (intervention A) in two
schools (one urban and one rural). A less intensive,
countywide intervention (intervention B) provided ma-
terials and training to local youth and temperance work-
ers. The evaluation involved the two intervention A
schools, tie matched intervention B schools selectd
from the county, and two matched reference schools
selected from another county that did not receive an
organized intervention. h&a et al. (1982) found that
among boys, the prevalence of occasional smoking (one
or two times per month) had increased by 30 percent in
the reference group, by 8 percent in the A group, and by
13 percent in the B group. Among girls, the prevalence of
occasional smoking had increased by 20 percent in the
reference group, by 18 percent in the A group, and by 9
percent in the B group. Vartiainen et al. (1990) reported
the results of an eight-year follow-up and found that the
prevalence of "any smoking" in the reference group was
10 percent higher than in the A group and 16 percent
higher than in the B group.

United Kingdom

  In the United Kingdom, Nutbeam et al. (1993) con-
ducted a controlled trial of two school-based interven-
tions. The Family Smoking Education Project was derived
from a program first developed in Norway (Aam et al.
1983). Directed toward lO- through 12-year-olds, the
project consisted of five lessons on the immediate health
effects of smoking and on the wider environmental im-
pact of tobacco growing and use. A notable feature was
a leaflet sent to parents to encourage their support for
school-based smoking education. The Smoking and Me
project was the United Kingdom adaptation of the MSPP.
Directed toward lO- through 12-year-olds, the program
consisted of six sessions highlighting a range of social
influences and equipping students with skills to manage
these social pressures. At the first-year and second-year
follow-ups, no differences were observed between the
intervention population and the control population for
either smoking uptake or personal skills.

  Overall, school-based smoking education programs
that have been evaluated internationally have met with
limited success in the past decade. In general, these
programs were brief and were not continued through the
high school years. Many countries are taking more com-
prehensive approaches to smoking control among young
people; such approaches include community action, fur-
ther restrictions on tobacco advertising and promotion,
and substantially higher tobacco tax rates than are found
in the United States.

130 Prevention


preventing Tobacco Use Among Young People

Meta-Analyses of School-Based Smoking
. prevention

  Extensive discussions of the methodological issues
inherent in research on smoking prevention have been
thoroughly discussed elsewhere (Cook and Campbell
1979; Flay 1985; Biglan, Severson, et al. 1987; Murray and
Hannan 1990). The primary issues have included ques-
tions of mixed units of analysis, attrition of the subject
(student) population, integrity of implementation, and
homogeneity of the subject population. These issues
have been partly accounted for in four important meta-
analytic studies published since 1980.

  Tobler (1986) examined 143 studies of drug-use
prevention programs for 6th- through 12thgrade stu-
dents and found that these programs had an overall
significant impact on behavior, skills, and knowledge.
The study also found that peer-led programs and pro
grams dealing with social influences were more effective
than other modalities. Tobler (1992) later confirmed
these findings with more rigorous analytic methods. The
Rundall and Bruvold (1988) meta-analysis of 40 studies
of school-based programs to prevent smoking examined
knowledge, attitude, and behavioral outcomes of social
influence programs versus traditional programs; the so
cial influence programs were more likely to affect
attitudes and behavior. Rooney (1992) examined 90
school-based tobacco-use prevention programs con-
ducted from 1974 through 1989 that sought to develop
skills to resist social influences. The meta-analysis took
into account the clustering of students in schools and
used the school as the unit of analysis. Results mdicated
that smoking prevalence was 4.5 percent lower among
students in the social influence programs than among
students in control conditions. The social influence pro-
grams that were most effective at one-year follow-up
were those that were delivered to sixth-grade students,
that used booster sessions, that concentrated the pro-
gram in a short time period, and that used an untrained
peer to present the program. Under these more optimal
conditions, long-term smoking prevalence was reduced
by about 25 percent.

  Bruvold's meta-analysis (1993) included 94 sepa-
rate interventions from the 1970s and 1980s. The inter-
vention programs were categorized as rational Qxoviding
factual information), developmental (increasing self-
esteem and decision-making skills), social-norms-
oriented (providing alternatives and reducing alienation),
and social-reinforcement-oriented (developing skills to
deal with social pressures to smoke). The me&analysis
showed that the rational approach had very little impact
on smoking behavior, that the developmental and social
norms approaches had equivalent and intermediate

impact on smoking behavior, and that the social rein-
forcement approach had the greatest impact on smoking
behavior (Bruvold 1993).

Discussion

  In retrospect, research on smoking prevention has
by its very nature had to contend with various threats to
validity posed by factors such as mixed units of analysis,
differential attrition, and inconsistent implementation.
To a large extent, the most recent research studies have
been designed to deal with these methodological ob-
stacles and have still found moderately strong preven-
tion effects (Rooney 1992; Bruvold 1993). Therefore,mast
reviews of the smoking-prevention reseamh literature cons&
tidy have come to the same conclusions, which can be sum-
marized under three general findings.

  First, a variety of individual msearch reports @c&in
and Dusenbury 1989; Flay et al. 1989), severai comprehen-
sive literature reviews @lay 1985; Best et al. 19881, and four
meta-analyses Crobler 1986; Rundall and Bruvold 1988;
Rooney 1992; Bruvold 1993) have all reported lower
prevalences of smoking among students in social influence
programs than among students in equivalent comparison
groups or randomly assigned control groups. The differ-
ence between treatment and nontreatment groups ranges
from 25 to 60 percent and persists from one to four years,

  Second, as Best et al. (1988) have underscored, given
thenumberofresearchstudies,thevariabilityinprogramfor-
mat and scope, the various cofnmunties and cultures in
whichthesestudiesw~undertaken,andthepotential~~
to internal and external validity in school-based research,
the consistency of overall findings and reductions in
smoking prevalence across all these studies is rather
remarkable.

  Third, it has been observed repeatedly that the
positive shorter-term intervention effects reported in
adolescent smoking-prevention studies tend to dissi-
pate over time (Murray et al. 1989; Pen& MacKinnon,
Dwyer, et al. 1989; Flay et al. 1989; Ellickson, Bell,
McGuigan 1993). This general trend has been particu-
larly evident among school-based intervention studies
that included little or no emphasis on booster sessions,
few (if any) communitywide activities, or few (if any)
mass-media-based components (Botvin, Renick, Baker
1983; Perry, Klepp, Shultz 1988; Botvin and Botvin 1992).
These interventions may be enhanced if they are em-
bedded in a more comprehensive school health educa-
tion program (AlIensworth and Kolbe 1987; Walter,
Vaughan, Wynder 1989). The comprehensive school
health approach needs further evaluation but is promis-
ing as an effective prevention tool.

  Only the social influence approaches have been
scientifically demonstrated (through replicated research

Prevention   131


Surgeon General's Report

studies) to reduce or delay adolescent smoking. Still,
the effects of these programs have not been sustained
without additional educational interventions or commu-
nity components. This experience suggests that pro-
grams grounded in school-based skills training are indeed
important for preventing smoking, although more sus-
tained and comprehensive efforts may be needed for
long-term success.

  The concept of reciprocal determinism (Bandura
1986) would argue that these complementary compo-
nents should target the elements of the dynamic person-
environment interaction that school-based interventions
may not be capable of reaching, much less influencing.
These components would include the types of commu-
nity, environmental, legislative, policy-based, and soci-
etal interventions described later in this chapter.

Preventing Smokeless Tobacco Use

Introduction

  The 1986 publication of the Advisory Committee`s
Report to the Surgeon General (USDHHS 1986b) on the
health consequences of using smokeless tobacco (chew-
ing tobacco and snuff) and subsequent reports of wide-
spread use of smokeless tobacco among children and
adolescents (Boyd et al. 1987; USDHHS 1992b) have
called forth a wide range of written and media materials
(including films, pamphlets, and video programs) on the
risks of using smokeless tobacco (Wilson and Wilson
1987; Laflin, Glover, McKenzie 1987). These materials,
made available to school personnel and parents, have
aimed at countering the perception that smokeless to-
bacco is a safe alternative to cigarettes. Materials have
been produced by federal agencies (such as the NC1 and
the National Institute of Dental Research), voluntary
nonprofit groups (such as the ACS), and professional
organizations (such as the American Dental Association
and the American Academy of Otolaryngology). These
materials have been distributed widely, but the degree of
their diffusion has not been evaluated, nor has their effect
on young people's use of smokeless tobacco.

Evaluation of School-Based Efforts

or regular use and to promote or assist cessation for
adolescent and young adult users. Nine research grams
on smokeless tobacco use have been funded by the NQ
since 1987; most are focused on adolescent populations
(USDHHS 199Ob), and results are pending. Although
most of these projects have been school-based preven-
tion activities, some programs have targeted youth in
non-school settings (e.g., 4-H clubs, Little League base-
bail clubs, and Native American community centers).

  The prevention programs that have been evaluated
have targeted both smoking and smokeless tobacco use
among middle and high school students. The primary
focus has been on middle school (grades 6-8, ages 12-14).
Smokeless tobacco prevention has also been included as
part of more comprehensive`curricula to prevent drug
use, such as Here's Looking at You,2ooO (Roberts, Fitzmahan
& Associates, Inc., and Comprehensive Health Educa-
tion Foundation 1986), or as part of community-based
interventions to reduce drug use. Seldom have pro-
grams to prevent smokeless tobacco use been instituted
independent of other substance-use prevention or of a
more general tobaccouse prevention effort. Since smoke-
less tobacco products are used primarily by males, the
overall prevalence of use is lower than that of smoking.
There is also less concern about the health effects of
smokeless tobacco than about those of illegal drugs and
cigarettes. This logical inclusion, however, of smokeless
tobacco prevention in the context of other prevention
efforts makes the evaluation of the smokeless tobacco
component problematic.

  A factor that more directly obscures the impor-
tance of smokeless tobacco prevention is the widespread
acceptance of use by both young people and parents.
Youth generally perceive that smokeless tobacco use is a
safe alternative to cigarette smoking. For example, in one
study, 77 percent of school-aged children believed that
cigarette smoking was very harmful to one's health, yet
only 40 percent believed the same of smokeless tobacco
use (Schaefer et al. 1985). Parents are also more likely to
accept smokeless tobacco use than smoking among teens
(Chassin, Presson, Sherman 1985; see `Tarental Reaction
to Smokeless Tobacco Use" in Chapter 4).

  because the increased use of smokeless tobacco
among youth is a relatively recent phenomenon, few pro-
grams for preventing adolescent use of these products
have been evaluated for either short- or long-term efficacy.
Those that have been evaluated have been but one compo-
nent of a broad tobacco-prevention program.

The Oregon Research Institute Program

  In several studies, young adolescents have received
a preventive curriculum that targeted both smoking and
smokeless tobacco use. In one such study (Severson et al.
1991), a social influences program conducted by the Or-
egon Research Institute was delivered by regular class-
room teachers and by same-age peer leaders to entire
classrooms in randomly assigned schools. The brief
seven-session program significantly reduced smokeless
tobacco use among males in both seventh and (to a lesser

  In response to the emerging concern about the
health risks of regular smokeless tobacco use, the Na-
tional Institutes of Health has funded numerous research
grants to develop interventions to prevent initiation

132 Prevention


Preventing Tobacco Use Among Young People

extent) ninth grades. Parallel analysis failed to show that
the intervention had any positive effect on cigarette
smoking. The results for smokeless tobacco use, how-
ever, were particuIarly encouraging, since only two of
the seven class periods of the intervention were devoted
to smokeless tobacco.

  The intervention used in the Severson et al. (1991)
study sought to make students sensitive to overt and
covert pressures to use tobacco and taught effective ways
to respond to these pr&sures. The students practiced
how to refuse offers of tobacco. Besides using a struc-
tured curriculum with role-play activities, the teacher
used videotapes to standardize instruction and maintain
student interest. The program was taught by regular
classroom teachers; same-age peer leaders assisted in
role-playing activities for the seventh-grade students. A
videotape titled Big Diper (Oregon Research Institute
1986) was developed to highlight the physical and social
consequences of smokeless tobacco. To involve parents,
brief brochures were mailed to students' homes.

Toward No Tobacco Use

  A study by Sussman et al. (1993) reports positive
results in their Toward No Tobacco Use (TNT) project for
reducing smokeless tobacco use. The study compared
four different prevention curricula developed to coun-
teract three types of factors related to the onset of tobacco
use that are typically addressed within a comprehensive
social-skills program. These include peer approval for
using tobacco, incorrect social information provided about
tobacco use, and lack of knowledge about physical con-
sequences of tobacco use. The development of these
curricula is detailed in previous reports (Sussman 1991).

  Smokeless tobacco use was significantly less preva-
lent among students who had received the TNT inter-
vention than among those who had not (Sussman et al.
1993). The results of the evaluation of this lo-lesson
curriculum intervention suggest that learning about the
physical consequences of smokeless tobacco use can be
as successful as a social influences program and that a
combination of both is probably best for deterring use of
smokeless tobacco. The Sussman et al. (1993) study in
southern California and the Severson et al (1991) study
in Oregon suggest that smokeless tobacco use can be
reduced through school-based programs that try to pre
vent all types of tobacco use among seventh- and ninth-
grade studen&

Project SHOUT

  Elder et al. (1993) developed Project SHOUT, a
social influences program that has been evaluated in
22 junior high schools in San Diego County, Califor-
nia. Based on an operant conditioning model of

tobacco use (Elder and Stem 19861, the intervention
was delivered in randomly assigned schools to
seventh-grade students. Intervention and assessment
continued for three years (through seventh, eighth,
and ninth grades). Because of multiple school changes
at the end of the eighth grade, Project SHOUT used
telephone calls and programnewsletters for the ninth-
grade intervention.

  At the three-year follow-up, the intervention
had a significant effect on cigarette use, smokeless
tobacco use, and combined cigarette and smokeless
tobacco use. The intervention effect was particularly
strong during the ninth grade (Elder et al. 19931. The
three-year intervention and follow-up is a strength of
this study; previous studies have been limited to a
single intervention year and one-year follow-up.

Programs for Native American Populations

  Smokeless tobacco use by Native American youth
on reservations is higher than that of other groups (Schinke
et al. 1989). There is evidence of early, frequent, and heavy
use of snuff and chewing tobacco by Native American
children and Alaskan Natives (S&i&e et al. 1987X Young
people in these populations begin using smokeless to-
bacco at an early age, and girls use it at levels almost equal
to boys &hinke et al. 1987). Current reservation-based
interventions aimed at reducing this pattern of smokeless
tobacco use have not yet been evaluated. These ongoing
programs are sensitive to the unique aspects of tobacco
use by Native Americans, since tobacco has traditionally
played a role in sacred rites. The programs make extant
materials appropriate for Native American children by
creating a specific curriculum for the tribal group and
having Native Americans provide the intervention in
schools or other settings on their reservation.

Smoking Cessation
Introduction

  Few studies have examined adolescent smoking
cessation. The four primary sources of information on
adolescent cessation are national probability surveys on
patterns of adolescent attempts to quit (see "Attempts to
Quit Smoking" and "Self-Reported Indicators of
Nicotine Addiction Among Smokers" in Chapter 3), con-
venience sample surveys of adolescents who have tried
to quit on their own, reports from prevention projects on
effects of treatment on youth who were smokers at
baseline, and programs that explicitly try to recruit
adolescent smokers into cessation programs. The rela-
tively few intervention studies vary considerably in sci-
entific quality; many are anecdotal or descriptive accounts
of programs.

Prevention   133


Surgeon General's Report

Convenience Samples of Adolescents Who Try to
Quit Smoking

  Although national surveys ask a great many re-
spondents a few questions about quitting smoking, some
smaller studies have more deeply probed the experience.
The role of nicotine's pharmacologic effects has received
increasing attention, culminating in the 1988 Surgeon
General's report on nicotine addiction. The report dem-
onstrated that cigarette smoking is characterized by the
same addictive processes that have been observed with
other drugs that are abused (USDHHS 1988). Recent
observations of adolescents who have tried to quit smok-
ing suggest that dependency or addiction has developed
in many adolescent smokers and may play an important
role in their attempts to quit. Data from both Great
Britain (McNeill et al. 1986; McNeill1991) and the United
States (Hansen 1983; Hansen et al. 1985; Brshler et al.
1989) show that many adolescents who try to quit have
withdrawal symptoms that parallel those reported by
adult smokers (see "Nicotine Addiction in Adolescence"
in Chapter 2).

  Inasurveyof116Britishschoolgirls(aged 11 through
17) who had tried to quit smoking, 63 percent reported
withdrawal effects. The degree of withdrawal effects was
related positively to both self-report and biochemical mea-
sures of nicotine intake (McNeill et al. 1986). These f%d-
ings were replicated, although without biochemical
measures, in a study of American 6th through 12th-
graders of both sexes (Brshler et al. 1989). Over half of the
smokers in both of these studies reported attempts to quit,
and most were unsuccessful. These observations, along
with other data summarized in Chapters 2, 3, and 4,
strongly suggest that adolescent smoking is more than
socially driven and that addictive processes in adolescents
are similar to those that characterize adult smoking.

Effect of Smoking-Prevention Programs on Cessation

  Smoking-prevention programs have typically, and
appropriately, targeted younger adolescents. In these
populations, prevalence rates tend to be low, and those
who smoke are mostly doing so infrequently. These
studies, reviewed earlier in this chapter, focus on pre-
venting onset or on preventing the progression from
experimentation to regular smoking. The impact of
smoking-prevention programs on students who are ex-
perimental or regular smokers appears to be small and
inconsistent (Best et al. 1984; Johnson et al. 1986; Biglan,
Severson, et al. 1987). However, the small number of
regular smokers (that is, those who smoke every week)
tends to preclude meaningful analyses of cessation re-
sulting from these programs (Best et al. 1984).

Cessation Interventions in the School

  Young people who smoke have been a persistent
concern of both educators and voluntary health agencies.
A number of materials and programs for adolescent
smoking cessation have been developed and imple-
mented, but evaluation typically has been anecdotal or
descriptive (Hulbert 1978; Patterson 1984; Brink et al.
1988). Many of the older programs are described by
Thompson (19781, USDHEW (1979), and Seffrin and
Bailey (1985). Cessation programs are sometimes led by
peers, sometimes by teachers or volunteers. Participants
are recruited through school channels such as newslet-
ters, classes, and public address announcements. Evi-
dence from these descriptive reports, as well as from
some of the formal research programs described below,
indicates that recruitment is difficult; adolescent smokers
are hesitant to come forth. In some instances, the par-
ticipants in the school cessation programs are referred
by school authorities for infractions of school smoking
policies and are thus not coming to these programs
voluntarily.

  These issues are illustrated by a program evalua-
tion reported by the American Lung Association (un-
published data). The program, developed by a Minnesota
affiliate of the American Lung Association, was evalu-
ated in 22 schools in four states. A total of 241 students
(mean age = 16 years old) participated in eight 5Ominute
sessions during school hours over a four-week period.
Over half the students, however, were required to par-
ticipate as a consequence of being caught smoking on
school grounds. This inclusion of nonvoluntary partici-
pants may partly explain the program's low success rate:
at the end of the sessions, only 30 students (14 percent)
reported that they were abstinent (program dropouts
were counted as smokers). Low cessation rates like
these, coupled with recent legislation such as the Oregon
law forcing school authorities to take action against stu-
dents caught smoking on school grounds, signal the
need for more effective cessation approaches for student
smokers.

  Lotecka and MacWhinney (1983) compared an in-
tervention group focusing on cognitive behavioral skills
(N = 53) with a group only receiving health information
(N = 54). Less than 50 percent of the students in each
group participated in the three-month follow-up. Of
those assessed at that time, 78 percent of the students in
the cognitive behavior group reported a decrease in smok-
ing, and only 4 percent reported an increase; the compa-
rable figures for the information-only group were 46
percent and 31 percent. No information was provided
on complete abstinence. Given that reported rates of
smoking are relatively unreliable and that the program

134 Prevention


praenting Tobacco Use Among Young People

did not report cessation rates, this study cannot be con-
sidered conclusive.

Perry et al. (1980, 1983) conducted two school-
based cessation interventions in California schools. hi
the first, IOth-grade classes in three high schools
(N = 477) received a special program that focused on im-
mediate physiological effects of smoking and on social
cues that influence the adoption of smoking. Classes in
two control schools (N = 394) received standard infor-
mation on long-term health effects. The program con-
sisted of four consecutive 45minute sessions in regular
health classes conducted in the fall. Posttest outcome
data were obtained approximately five months later and
included carbon monoxide measures of smoking. At the
posttest, the experimental group, compared with the
control group, had a significantly greater percentage of
subjects who reported abstinence in the previous week
(22 vs. 16 percent) and month (30 vs. 24 percent). Parallel
significant differences were also found for carbon mon-
oxide measures.

  In their second study, the Perry group (1983) tried
to sort out the specific efficacious components within the
intervention program by analyzing three kinds of pro-
grams-those that discussed long-term health effects (the
control group), those that discussed immediate and long-
term physiological effects, and those that discussed so-
cial consequences--and comparing programs taught by
either teachers or college students. Twenty health classes
and four high schools were randomized by using a facto-
rial design. The study obtained three-month follow-up
data that included self-reports and carbon monoxide
breath tests. Using entire 10th~grade health classes solved
the recruitment problem but yielded a limited number of
current smokers; the relatively small number of pretest
smokers in this study (N = 82) precluded finding any
significant difference between the groups. Overall, 23
percent of the pretest smokers reported not smoking at
the three-month follow-up. Teachers tended to be more
effective with the traditional curriculum covering long-
term health effects, and college students seemed more
effective with the social influences curriculum.

The largest and most systematic school-based ado
lescent cessation study has not yet been published. Bur-
ton et al. (unpublished data) worked with rural and
suburban high schools in two states. Within each of the
16 treatment schools, students volunteering to partici-
pate in a cessation clinic were randomly assigned to a
clinic or to a control group of students told they were on
a waiting list. Clinic students were further randomly
assigned either to a clinic designed to address addiction
or to one designed around psychosocial dependency.
Clinics consisted of five sessions spaced over one month.
A follow-up session was held three months after the fifth

session. The control participants were also invited to the
follow-up session, where smoking status was assessed
both by self-report and measurement of saliva co&-tine.

  At the three-month follow-up, 8.4 percent of clinic
participants and 10.5 percent of controls were abstinent.
When corrected for biochemical verification, these figures
become 6.8 and 7.9 percent, respectively. There was con-
siderable attrition; students lost to follow-up were as-
sumed to be smokers. The negative results in the study are
especially sobering because the investigators had previ-
ously conducted 31 focus groups with adolescents to help
inform the intervention's reuuitment strategies and con-
tent (Sussman et al. 1991).

Difficulty in recruiting adolescent smokers in school
programs has been a pervasive problem for investigators.
Adolescents may be concerned about parents or teachers
learning that they smoke (since parental consent could be
required for participation). Adolescents may also be less
motivated thanadults to quit, since long-term health con-
sequences carry less weight with the young. A simpler
explanation of low recruitment is that prevalence rates are
low; schools do not provide large populations of smokers
from which to recruit. Multisite trials that pool subjects
may be needed before rigorous and meaningful evalua-
tions can take place.

Cessation Interventions Based Outside the School

Hollis et al. (in press) tried an unusual approach to
recruit young smokers. Adolescents between 14 and 17
years of age who were members of a large health mainte-
nance organization (HMO) were mailed a screening ques-
tionnaire that asked about "health habits." Those who
reported that they had smoked in the past week were
asked if they would participate in a twoyear study of
adolescent health and were randomly assigned to either
an intervention group that received help to quit smoking
or a control group that received no such help.

  The focus of the intervention was an office visit
with a nurse practitioner at a conveniently located HMO
clinic. Incentives were offered for attending these ses-
sions, each of which lasted about 60 minutes. The partici-
pants reviewed their health history, watched and
discussed a video on adolescent smoking cessation, were
encouraged to set a quit date, and were given tips and
strategies for successful quitting. Those who wanted to
quit smoking received a follow-up call one week later;
additional calls were also made, depending on the
adolescent's continued interest in quitting. Participants
who had quit smoking were eligible to participate in a
lottery with chances to win $100.

All participants were followed up at one year, at
which time both self-report and biochemical (saliva
cotinine, carbon monoxide) data were obtained. The

Proven tion  135


Surgeon General's Report

intervention and control groups had similar self-report
measures of smoking (i.e., measured in number of ciga-
rettes in the last month, week, day) and similar biochemi-
cal indicators of smoking. No relationship was found
between the number of contacts with the HMO interven-
tionist and either quitting rates or the number of ciga-
rettes smoked. Similar interventions in health care settings
with adult smokers have usually yielded positive results
(e.g., Hohis et al. 1991), but this was clearly not the case
for adolescent smokers.

Discussion

  The data reviewed indicate consistently that ado-
lescent smokers frequently try to quit but are usually
unsuccessful, often have withdrawal reactions much like
adult smokers, are difficult to recruit and retain in formal
cessation programs, and are not responsive to programs
thus far developed. Further basic research and new.
directions for intervention are clearly needed. Data pre
sented in Chapter 3 (see "Adult Implications of Adoles-
cent Smoking") from the Monitoring the Future Project
show that well over 80 percent of adolescents who smoked
half a pack a day or more as seniors in high school (over
15 percent of the sample) were smoking five to six years
later as young adults; over half of these were smoking a
pack or more a day at follow-up. In the absence of
intervention, adolescent smokers will most likely be
come adult smokers.

Smokeless Tobacco Cessation

Introduction

  Of the estimated six million people who regularly
use smokeless tobacco, half are under age 21 (USDHHS
1986b). Data from several national surveys show an
increase in the prevalence of smokeless tobacco use, spe-
cifically in the use of moist snuff among young males
(Boyd and Glover 1989; Marcus et al. 1989; Novotny et al.
1989; Rouse 1989; see "Current Use of Smokeless To-
bacco" in Chapter 3). The high prevalence of smokeless
tobacco use underscores the growing need to help young
people quit.

  To date, there are few published studies of smoke-
less tobacco cessation. The withdrawal symptoms for
smokeless tobacco are the same as those for smoking-
cravings for the substance, irritability, distractibility, and
hunger (Hats&an&Gust, Keenan 1987)-althoughthese
symptoms may be less intense and felt less frequently.
Because of these similarities, most cessation programs for
smokeless tobacco users are multicomponent treatments
that use key elements from smoking-cessation programs
that have been extensively evaluated in largescale studies
(Sever-son 1993).

Clinical Studies

  Clinical studies of smokeless tobacco cessation have
been done with both adolescents and adults. The first
published study of smokeless tobacco cessation was re
ported by Glover (1986), who adapted the ACS' Fresh
Start Adult Smoking Cessation Program for use with 41
adults who used smokeless tobacco. This pilot study
resulted in a six-month self-reported abstinence rate of
only 2 percent. However, these subjects had not vohm-
tarily sought assistance in quitting; they had been re-
quired to attend the program for violating school rules at
a college that prohibited the use of tobacco products.
Low success rates are not surprising in a nonvoluntary
cessation program.

  Eakin, Severson, and Glasgow (1989) reported an
intervention with adolescent male daily users, aged 14
through 18, who were recruited from high schools in
Eugene, Oregon. The study recruited 25 students, five of
whom also smoked cigarettes concurrently. The program
consisted of three small group meetings with counselors,
each lasting approximately one hour, during which the
focus was on developing coping skills for cessation. Of the
21 subjects who completed treatment, two s&j&s had
quit using smokeless tobacco by the end of treatment, and
three subjects were abstinent at the six-month follow-up.
Compared with the other students, however, these suc-
cessful quitters had consumed a smaller amount of smoke
lesstobaccoatbaselineandwerelessaddicted,asmeasured
by an adapted Fagerstriim Tolerance Questionnaire
(Fagerstrom 1978). They were also more involved
in school athletics than those who did not succeed
at quitting.

School-Based Efforts

  Three recent studies of smokeless tobacco cessation
are informative about school-based cessation and self-help
approaches. Burton et aL (unpublished data) report msulk
from a school-based cessation clinic model tested in 16 high
schools in Illinois and California. Within each school, ciga-
rette and smokeless tobacco users were recruited and either
randomly (and vohmtarily) assigned to a cessation clinic or
told the clinics were filled. Clinics consisted of five sessions
over a one-month period. A sixth session was held three
months later to assess the intervention and control groups.
The attrition rate for the clinic group was higlx almost half
the students did not complete the treatment. Of the 16
smokeless tobacco users who completed five sessions, seven
reported quitting at the end of the treatment; none of the
five students in the control group reported quitting.
However, when the clinic dropouts were included as
the denominator and the results corrected for biochemi-
cal verification, the quit rate for students in the smoke-
less tobacco clinic was 15 percent; none of the control

136 Prevention


Proven ting Tobacco Use Among Young People

subjects had quit at the three-month follow-up. The
study suggests that a school-based multisession clinic
can achieve small cessation rates for adolescent subjects
who volunteer, although the volunteer rates for the study
were notably low.

  Persons going through treatment for smokeless
tobacco addiction often request an oral substitute to help
them through withdrawal. Smokeless tobacco users re
port using cinnamon sticks, gum, sunflower seeds, finely
ground mint leaves, or other chewed foodstuffs to lessen
the effects of withdrawal (Severson 1992). To evaluate
the use of nonnicotine substitutes as aids for smokeless
tobacco cessation, a recent study compared the use of a
ground-up mint product, chewing gum, and no substi-
tute (Chakravorty 1992). Subjects were recruited from
six high schools in rural Illinois. Two schools each were
randomly assigned to either the treatment group (mint
snuff substitute), gum group, or lecture-only control
group. Within schools, smokeless tobacco users were
invited to volunteer for a two-session school-based ces-
sation program. Eighty-three males were recruited to
participate. Of the 70 students who completed the treat-
ment, 30 were in the mint group, 15 in the gum group,
and 25 in the lecture-0 nly group. At the end of the
treatment period, all three groups had about the same
quit rates. Eleven students reported quitting smokeless
tobacco, but nine of these quitters also smoked cigarettes.
The author reports that students using the mint snuff
substitute significantly reduced their frequency and in-
tensity of smokeless tobacco use, but the study had rio
biochemical verification of use. The results suggest that
adolescent males who use smokeless tobatico can be
recruited to attend sessions at school and that nontobacco
oral substitutes may be a helpful adjunct to quitting.

  Research with adults suggests that health care pro-
viders can motivate some adult users of smokeless to-
bacco to quit (Stevens et al., in press). The clinical
opportunity to provide advice on quitting in the context
of health care delivery has been referred to as a. "teach-
able moment" (vogt et al. 1989; Morosco 1986). The
results are modest in terms of overall quit rates, but
having dentists, hygienists, nurses, and physicians coun-
se1 their patients to quit using smokeless tobacco could
have a significant effect on prevalence. The Stevens et al.
(in press) study provided the first examination of a large-
scale, low-cost intervention to encourage smokeless to-
bacco users to quit. This program, which was conducted
in the context of regular hygiene visits, provided strong
evidence of the effect of smokeless tobacco use on oral
health: 73 percent of the adult users in this study had
identifiable oral lesions (Little, Stevens, La Chance, et al.
1992). Parallel studies with youth or studies of programs
using physicians or other health care providers have not
been conducted.

Smokeless Tobacco and Cigarettes

Young people who use smokeless tobacco may
also smoke cigarettes. Studies have reported that from
12 to 30 percent of all regular users of smokeless tobacco
also use cigarettes (E&in, Severson, Glasgow 1989; Wil-
liams 1992; Stevens et al., in press; see "Use of Smokeless
Tobacco and Cigarettes" in Chapter 3). This relationship
is critical, since cessation programs may motivate smoke
less tobacco users to quit using snuff or chewing tobacco,
yet not affect their use of cigarettes-and thus not affect
their addiction to nicotine. Moreover, deprivation of one
substance may lead to a direct increase in the use of the
other (Biglan, La Chance, Benowitz, unpublished data).
Cessation rates among men who use both tobacco prod-
ucts are significantly lower than those among men who
use smokeless tobacco exclusively (Stevens et al., in press).

Research and Programmatic Challenges

  Certain peculiar aspects of smokeless tobacco use
may present problems to those who plan or study cessa-
tion programs. The lack of public data on the nicotine
content of smok&ss tobacco products is not only a
research problem but a challenge to cessation efforts that
might reduce the severity of nicotine withdrawal by
gradually cutting back on nicotine ingestion. Such ef-
forts are further hampered, as are studies or programs
depending on self-monitoring of product consumption,
by the nonuniform (bulk) packaging of most smokeless
products and by the variation in the amount of product
that constitutes a "pinch" (of chewing tobacco) or a
"dip" (of moist snuff) (Severson et al. 1990.) External
monitoring of use also has inherent limitations, since
snuff (and to a lesser extent, chewing tobacco) can be
used surreptitiously. On the other hand, the oral lesions
frequently experienced by smokeless tobacco users
readily indicate smokeless use--and provide direct physi-
cal evidence to the user that this behavior has detrimen-
tal health effects (Little, Stevens, Severson, et al. 1992).

  The relationship between smokeless tobacco use
and cigarette smoking also presents problems for re-
search and intervention. Because many adolescents per-
ceive smokeless tobacco use to be a safe alternative to
smoking, motivation to quit using smokeless tobacco
products may be low. On the other hand, because as
many as one-third of all smokeless tobacco users also
smoke cigarettes, the possibility exists (as was discussed
previously) that persons trying to quit using smokeless
tobacco may continue to smoke-or even increase their
smoking-to minimize nicotine cravings.
Although the preliminary evidence is that cessa-

tion rates for smokeless tobacco are similar to those for
smoking, the difficulty in recruitment, the small sample

Prevention 137


Surgeon General's Report

sizes, the limited number of studies, the lack of control
groups, and the lack of long-term follow-up necessitate
cautious interpretation. Further research on cessation
must consider the effects of usage frequency and
intensity and must focus on relapse rates, use of nico-
tine replacement in cessation, self-help attempts at
quitting, effects of advice by physicians and other health
professionals, and effects of taxation and environmen-
tal restrictions.

Clinical Interventions to Prevent Tobacco Use

Introduction

  Physicians, dentists, and other health care provid-
ers who take care of children are in a unique position to
help their patients avoid the use of tobacco (Perry and
Silvis 1987). Children perceive these professionals as
credible health experts and thus may attend more to
what they say than to what parents and other adults say.
Health care providers can serve as powerful role models
who can positively influence the health behavior of their
young patients, especially where a long-term relation-
ship has been formed with the child and the family.
Lastly, health care providers should know when to pro-
vide specific health information at critical times in a
child's development.

  The medical office provides an important opportu-
nity for physicians, dentists, and staff to communicate
attitudes about smoking and smokeless tobacco use
(Kottke et al. 1989; Richards 1992). By not smoking,
health professionals can serve as positive role models,
as the American Academy of Pediatrics @AI?) and the
American Academy of Family Physicians (AAFP) have
recommended. Smoking by physicians, other staff,
adolescents, or parents should not be allowed in
the physician's office or reception area (AAP 1987;
AAFr 1992).

  The AAP recommends that between birth and 21
years of age, a child should make a minimum of 20 visits
to the physician (AAP 1988). These visits offer opportu-
nities to prevent and deter tobacco use. To be successful
at preventing tobacco use, physicians and other health
professionals must know what the risk factors are, how
to identify children who are most vulnerable, and how to
intervene effectively.

Recommendations to Clinicians Who Care for
Children and Adolescents

  Education about tobacco should begin in child-
hood, when family standards and values are developing
(AAl' and Center for Advanced Health Studies 1988).
The child's visit may also afford the opportunity for a
health professional to advise young parents who smoke
to stop (Perry, Griffin, Murray 1985). During infancy and

early childhood, clinicians should emphasize to parents
the relationship between environmental tobacco smoke
and the infant's health, particularly the association be-
tween environmental tobacco smoke and children's pneu-
monia, bronchitis, asthma, middle ear disease, and sudden
infant death syndrome WSDHHS 1986a, 1990a; U.S. En-
vironmental Protection Agency [USEPA] 1992). Advice
from a child's physician can reinforce advice that parents
may have received from their own doctors. Clinicians
thus need to learn skills to promote antismoking behav-
ior and encourage parents to stop smoking.
  The NC1 and the AAP have developed recommen-
dations for health professionals to prevent their preadult
patients from trying smoking /Epps and Manley 1991bI.
These brief activities can be carried out during the peri-
odic visits that the AAP recommends between birth and
21 years of age, as well as at other visits. Five steps that
begin with the letter "a'`-anticipate, ask, advise, assist,
and arrange follow-up-are recommended:
o Anticipafe the risks for tobacco use associated with the
child's development stage. These risks include expo
sure to environmental tobacco smoke, experimenta-
tion with tobacco, and nicotine addiction (Kandell975;
Hawkins, Lishner, Catalan0 1985; Dent et al. 1987;
AAP 1988). Children and adolescents are more likely
to use tobacco if their siblings and friends use it and if
tobacco use is perceived as normative or functional
(IJSDHI-IS 1986a; see "Interpersonal Factors" and `Ter-
ceived Environmental Factors," both for smoking and
for smokeless tobacco use, in Chapter 4). Adolescents
are vulnerable to tobacco use-especially those with
fewer coping skilIs (Doueck et al. 1988), those suscep-
tible to cigarette advertising (Blum 19801, and adoles-
cent females concerned about their body weight. (Gritz
1986).

o Ask at each visit, about tobacco exposures and tobacco
use (Richards 1992). Ask about tobacco use by the
patient and by the patient's friends and family. When
seeing infants and young children, ask parents whether
the patient has regular contact with anyone who
smokes. Ask if tobacco use is being discussed among
the child's friends or in school and, if so, in what
classes. Ask about the child's school health education
program. Ask the child about participation in sports
and extracurricular activities that may be incompat-
ible with smoking. In dental examinations, inspect
the intraoral soft tissue. If changes are noted in the
mucosa, ask about smokeless tobacco use.

o Advise tobacco users to stop. Advise women of the
adverse effects of smoking during pregnancy. Inform
smoking parents of the health consequences that envi-
ronmental tobacco smoke can have on their children.
Advise children and adolescents who are using (or

138 Prevention


Preventing Tobacco Use Among Young People

even trying) tobacco to stop. Advise smokers of the
short-term adverse consequences of tobacco use, such
as bad breath, other odors, and the cost of cigarettes.
Advise smokeless tobacco users of the potential con-
sequences of use, such as discoloration of teeth, de
struction of soft tissue in the mouth, and potential
early development of oral lesions and cancers.
o Assist tobacco users in stopping. Encourage parents
who are trying to quit smoking and help them choose
effective strategies to help them quit (Richards 1991,
1992). Assistance for parents or adolescents can in-
clude selecting a quit date, providing self-help materi-
aIs, and in some cases counseling on the use of nicotine
replacement (transdennal nicotine patch or nicotine
gum) (Glynn and Manley 1989). Help children and
adolescents take additional responsibility for their
health behaviors. Encourage participation in pro-
grams that develop skills for solving problems, setting
goals, making decisions, and countering peer pres-
sure (Bingham, Edmondson, Stryker 1984a, b).

o Arrange follow-up visits as appropriate. Arrange more
frequent follow-up visits for an adolescent who is
experimenting with tobacco products. At the first
follow-up visit, one to two weeks after a scheduled
quit date, discuss progress and problems. Arrange a
second visit in one to two months.
  The five steps described above should be common-
place in the medical setting. Richards (1992) notes that
"the words that a physician chooses to discuss smoking
with a patient should be considered no less a therapeutic
agent than the pharmacologic agent that the physician
prescribes" (p. 687). Yet Frank et al. (1991) found that
only 14 percent of smokers aged 12 through 17 years who
had seen a physician in the previous year had been
advised to quit smoking. In contrast, over 50 percent of
smokers aged 25 years and older were advised to quit.
Clearly, more consistent advice, concern, and counsel
from the medical profession is warranted.

Role of Health Professionals in the School, in the
Community, and in Policy Formation

  Physicians and other health professionals are often
considered leaders in their communities and have the
opportunity to mobilize schools and communities to
develop tobacco-use prevention, cessation, and policy
change strategies. Health professionals who have exam-
ined their roles in this larger context should encourage
their colleagues to act as advocates for such programs
and, if possible, participate in their development or imple-
mentation (Shank 1985; AAP 1987; Blum 1992).

  Health professionals play a powerful role as
sources for nonsmoking advice and assistance, as role

models of nonsmoking adults, as providers and sup-
porters of a nonsmoking health care environment, and
as agents who deliver nonsmoking programs in schools
and communities (USDHHS 1991). Several medical
organizations have adopted policies and developed
programs to encourage member concern and involve-
ment in preventing adolescent tobacco use. The AMA
House of Delegates has adopted numerous policy reso-
lutions that support local tobacco-control activities on
behalf of children and others (AMA 1992b). The AAF'P
(1987) has also published policies and a manual on
how to encourage patients of all ages to stop smoking.
The AMA Guidelines for Adolescent Preventive Ser-
vices recently recommended that physicians actively
screen and counsel adolescentpatients about tobacco
use (AMA 1992a). The AAP, with the NCI, has drafted
a set of age-specific recommendations for pediatric
practice as part of their Tobacco Free Generation pro-
gram to prevent adolescent tobacco use (Epps and
Manley 1991a).  The AAP also distributes Healthy
Beginning kits developed by the American Lung As-
sociation for counseling parents on the harmful effects
of smoking around children and distributes pamphlets
for parents and adolescents regarding tobacco use (AAP
1988, 1990a, b). The American Academy of Oto-
laryngology-Head and Neck Surgery, Inc., launched
a major public service campaign titled Through with
Chew in response to the problem of smokeless tobacco
use by youth. The campaign includes a video, a physi-
cian volunteer kit to encourage and assist members in
community outreach, and a variety of educational aids
designed to persuade young men, especially athletes,
not to use smokeless tobacco (American Academy of
Otolaryngology-Head and Neck Surgery 1992).

Community Programs to Discourage
Tobacco Use

Introduction

Community-based strategies to prevent smoking
are important adjuncts to school-based programs. Some
studies have shown that classroom-based smoking-
prevention programs, by themselves, have produced only
short-term effects (Lichtenstein et al. 1990; Pentz,
MacKinnon, Flay, et al. 1989; Best et al. 1988). These
limited outcomes suggest the need to mobilize parents
and elements of the community outside the schools to
produce lasting behavior change.

Young people who have the highest rates of to-
bacco use are those least likely to be reached through
school programs (Glynn, Anderson, Schwarz 1991).
Messages concerning tobacco use will be more accept-
able to high-risk adolescents if they are embedded in
groups or programs to which these youth already

Prevention   139


belong, rather than in tobacco-use prevention programs
that stand conspicuously apart (Glynn, Anderson,
Schwarz 1991). Community organizations and groups,
on the other hand, are associated with particular social
networks and social groupings of adolescents-potential
avenues of program entry to the various social contexts
of adolescents' lives.

  Such contacts with and through these groups are
important, since a strong correlation has been observed
between smoking behavior and social group member-
ship among youth (Novick et al. 1985; La Greta and
Fisher 1992). The social environment of youth may in-
clude strong cues to use tobacco, such as adult role
models who smoke or social groups where tobacco use is
viewed positively. Community programs can effectively
address these environmental elements and disperse mes-
sages against tobacco use (Becker et al. 1989; USDHHS
1991). Concerted use of multiple school and community
channels for affecting adolescent tobacco-use behavior
can produce a synergistic effect on the risk factors associ-
ated with adolescent tobacco use (USDHHS 1991).

  Information about the programs described in the
following sections was obtained through national and
regional organizations and published literature. Many
other locally initiated programs have been carried out in
individual communities throughout the United States,
but information on them was not readily available.

Communitywide Research Trials on Smoking
Prevention

  In the last 15 years, several major community-based
prevention trials that target youth smoking have been
undertaken. Three of these, the Stanford Heart Disease
Prevention Program, the Pawtucket Heart Health Pro
gram,andtheMinnesotaHeartHealthProgram,address4
several cardiovascular risk factors for all age groups and
used a variety of community strategies and channels,
including school-based programs for youth (Farquhar et
al. 1985;Mittelmarketal. 1986;Carletonetal. 1987). Young
people therefore received these interventions directly-
through school and home-based programs-and indi-
rectly-through a communitywide attempt to structure
the overall social and physical environment to support
smoking cessation and to discourage young people from
starting to smoke. In the Class of 1989 Study, which was
part of the Minnesota Heart Health Program, all of the
2,400 students in the graduation class of 1989 in two of
the state program's six communities took part in a longi-
tudinal study of health behaviors from 1983 through
1989. In one community, the students also participated
in five years of school-based health education, including
a peer-led prevention program that addressed social
influences to smoke (Perry, Klepp, Sillers 1989). At each

140 Prevention

Surgeon General's Report

of the annual follow-up surveys from 1984 through 1989,
youth from the intervention communities had signifi-
cantly lower smoking prevalences and smoking intensi-
ties than youth from the reference communities (Figure
3); at the end of 12th grade, the intervention group had
reduced its smoking prevalence by 40 percent (Perry et
al. 1992).

  Similar results a172 anticipated from COMMIT, which
isacomprrhensive,,community-basedappIoachtosmoldng
cessation Though COMMITS adolescent component is
largely limited to the school-based efforts, the program is
designed tochangethecommunityenvinmmentbymaking
smoking a major public health issue and strengthening the
social norms and values that support nonsmoking (T'homp
son et al 1990-91).

  The Richmond Quits Smoking Program tested the
communitywide approach in a predominantly black com-
munity.  Program components, including youth
programs, were integrated into existing communication
channels and social structures, and the smoking issue
was presented in ways relevant to the black community
(Hunkeler et al. 1990).

  Trials that focus specifically on youth include the
Midwestern Prevention Project (MPP), which tested the
use of a home- and community-based program in addi-
tion to school curricula to prevent the onset of tobacco

Figure 3. Smoking prevalence of the cohort

sample, Class of 1989 Study

~1*~~*~ Intervention community

- Reference community

   6     7     a     9*    10     11     12
                 Grade
Source: Pen-y et al. (1992).
*Smoking prevalence adjusted for false negatives in ninth
grade.


Preventing Tobacco Use Among Young People

ug. The overall design of the h4PP included all commu-
&ies within metropolitan Kansas City (Kansas and Mis-
souri) and Indianapolis (Indiana). Within each of these
two areas, cohorts of adolescents were assigned by school
to intervention or delayed intervention (control) condi-
tions. The intervention programs initiaIly targeted sixth-
or seventh-grade students and consisted of a IO-session,
school-based social skills curriculum; 10 homework as-
signments to be completed with parents or guardians;
mass media coverage using television, radio, and print;
community organization; and policy change. In the first
two years of the project, 22,500 adolescents participated
in the school and community intervention. Analyses
from students in 42 schools (N = 5,008) indicated a lower
prevalence of past-month cigarette, alcohol, and mari-
juana use at one-year follow-up for those exposed to the
school intervention than for the control group (17 per-
cent vs. 24 percent for cigarette smoking, 11 percent vs. 16
percent for alcohol use, and 7 percent vs. 10 percent for
marijuana use) (Pen& Dwyer, et al. 1989).

  Similar results were observed after two years for a
longitudinal panel of students from eight schools in Kan-
sas City (N = 1,122) (Pentz, MacKinnon, Hay, et al. 1989)
(Table 6). Third-year results demonstrated sustained
impact only on tobacco and marijuana use, but reduc-
tions were equivalent for adolescents at lower or higher
risk (Johnson et al. 1990). The MPP is particularly
important because it demonstrates the feasibility of a
large-scale, communitywide effort focused exclusively
on youth. The program has also demonstrated impact
on those at high risk, and it has considerable method-
ological strength. The MPP's long-term impact on
tobacco is still to be determined.

  The New England Research Institute has developed
and tested a community program for smoking prevention

among Hispanic (Puerto Rican) adolescents. The program
includes a music video, buttons and T-shirts, a smoking
cessation booklet, information booths and a traveling
music show at area festivals, and a basketball tournament
that includes a discussion about pressures to smoke
(McGraw 1990). The prehminary results of the evaluation,
however, indicate no differences between the intervention
group (in Boston) and a comparison group (in Hartford) in
reported smoking rates, attitudes toward smoking, or in-
tentions to smoke.

  Currently under way is Project SixTeen, a commu-
nity trial being conducted by the Oregon Research Insti-
tute from 1990 to 1995. In this project, experimental
communities receive a school program combined with
community intervention that includes parental involve-
ment, media campaigns, efforts by health care providers,
and changes in policies and regulations (Ary and Biglan,
unpublished data).

State and Federal Tobacco-Control Efforts at the
Local Level

  A number of states have adopted tobacco-control
programs that include community-based adolescent
components. The Association of State and Territorial
Health Officials (ASTHO) has recommended the devel-
opment of statewide tobacco-control plans that include
both school and nonschool activities for youth (ASTHO
1989). At least 12 states have developed freestanding
statewide tobaccocontrol plans, and another 22 states
have incorporated them into plans for controlling chronic
disease (CDC 19!Xb). All but 15 states have a specific
budget devoted to tobacco-related activities. Examples
of state-funded nonschool activities to prevent tobacco
use include the K.I.D.S. Coalition, a Utah program that
encourages youth to work with community leaders to

Table 6.  Outcomes of the Midwestern Prevention Project: adjusted net differences in the percentage of
   smokers in program and control groups, from baseline to 6-month, l-year, and 2-year follow-up

                           Adjusted net difference*

Smoking variable               6 months

Lifetime use                     2.3
Past-month use                  -7.5$
Past-week use                   -6.43

1 year

1.2
-10.25
-7.9s

2 years

11.7+
-16.05
-1l.F

Source: Pentz, MacKinnon, Flay, et al. (1989).
`Analyses done with school as a unit of analysis, adjusted for race and grade.
`p c .lO (one-tailed test).
tp < .05 (one-tailed test).
"p < .Ol (one-tailed test).

Preuen t ion   141


Surgeon Generals Report

create social change around the tobacco issue (Utah De
partment of Health 1991), and the Body Guards cam-
paign, a program sponsored by the Minnesota De-
partment of Health that trains minority youth (aged 12
through 14 years) to involve their families and others in
the community in tobacco-free pledges and messages
(ASTHO 1992).

  The Federal Comprehensive Smokeless Tobacco
Health Education Act of 1986 (Public Law 99-2521, which
included a mandate for health education programs and
materials about risks of smokeless tobacco, coincided
with an increase in state-funded community programs
addressing smokeless tobacco. In Ohio, for example, the
Department of Health has involved American Lung As-
sociation affiliates, Boys and Girls Clubs of America, Little
League, the Cleveland Indians baseball team, 4-H Clubs,
and juvenile detention centers in efforts to reach youth at
high risk of using smokeless tobacco (Capwell 1990).

  The most comprehensive state tobacco-control pro-
gram operates in California. Administered by the state's
Department of Health Services and Department of Edu-
cation, the program has been funded since 1989 by a
cigarette excise tax increase of 25 cents per pack (as a result
of Proposition 991, one-fifth of which is dedicated to
antitobacco education (Bal et al 1990).  Community-
based prevention services are specifically directed to
high-risk youth (i.e., those who have parents who smoke,
those who have dropped out of school, or those who are
economically disadvantaged) (Tobacco Education Over-
sight Committee 1991). During its first two years, this
program created local tobacco-use prevention coalitions
in all 61 local health jurisdictions, organized a youth
summit called Kids Choose a Tobacco Free Future, held
training workshops for county staff of the Child Health
and Disability Prevention Program to introduce materials
and techniques for counseling children and parents about
tobacco use, and funded many projects targeting ethnic
minority youthand theircommunities. CalifomiaSmoke-
Free Cities is a joint project sponscmd by the California
Healthy Cities Project and funded by Proposition 99. This
program encourages cities to strengthen local tobacco-
control efforts through various activities, many of which
include youth (California Smoke-Free Cities 1992).

  A community-based program that embraces mul-
tiple states and communities is the Planned Approach to
Community Health (PATCH), a partnership of the CDC,
state health departments, and local communities to plan,
carry out, and evaluate programs to prevent chronic
disease (IJSDHHS 1992a). Many of the 19 states and the
more than 50 communities that have been involved in the
PATCH program have carried out communitywide
tobacco-use prevention efforts,

  The Center for Substance Abuse Prevention, part
of the Substance Abuse and Mental Health Services

Administration WMHSA), sponsors a program of Com-
munity Partnership Grants, in which communities ad-
dress local drug-use prevention issues.  Public Law
102-321, the ADAMHA Reorganization Act, Section 114,
provides that all projects funded as prevention, treat-
ment, and rehabilitation model projects for high-risk
youth are to include strategies for reducing both tobacco
and alcohol use among minors.

  The NC1 has supported nearly 100 controlled inter-
vention trials aimed at preventing young people from
taking up tobacco and herping adult users quit. These
trials have involved more than 10 miIlion people in 33
states and over 200 communities in North America; 24
trials specifically targeted adolescents, and 6 addressed
the prevention of adolescent use of smokeless tobacco
KJSDHHS 199Ob).

  The NCI's American Stop Smoking Intervention
Study for Cancer Prevention (ASSIST) is the largest
tobacco-control project attempted in the United States.
ASSIST is designed to demonstrate that a comprehen-
sive, coordinated intervention effort can significantly re-
duce smoking and tobacco use. The scientific rationale
for this approach was clearly detailed in Strategies to
Control Tobacco Use in the United States: A Blueprint for
Public Health Action in the 1990s (LJSDHHS 1991).

  ASSIST is predicated on a coalition model. During
the planning phase, nearly 1,000 community health agen-
cies, social service organizations, and voluntary heahh
groups have joined state and local tobacccFcontro1 coali-
tions. This number will grow as the project enters its
intervention phase, when these organizations are ex-
pected to begin carrying out interventions targeting youth
and other high-risk populations served by these groups.
A number of states, including Maine, Virginia, Michigan,
Massachusetts, Colorado, and Minnesota, have supple-
mented their broader statewide coalitions with separate
coalitions for controlling tobacco use among youth. Those
ASSIST states that have high rates of smokeless tobacco
use (West Virginia, Virginia, North Carolina, and South
Carolina) specifically address such behavior among both
adults and youth in their statewide comprehensive plans.

  ASSIST has the potential to save more than
1.2 rnilhon lives, including over 4CKl,OC0 deaths averted
from lung cancer alone. The majority of these lives saved
would be the direct results of ASSIST's primary
prevention efforts among children, adolescent, and young
adults.

Community Organizations for Preventing
Tobacco Use

  Many youth organizations include a program-
matic focus on substance use. These program activities
may or may not explicitly focus on tobacco separately

142 Preuention


Preventing Tobacco Use Among Young People

from other drugs. In most cases, little or no evaluation
has been done to measure the effect these programs
have on tobacco use.

  Project California 4-Health focuses specific~y on
tobacco and is a joint effort of the University of Califor-
nia at Davis and the University of California Coopera-
tive Extension 4-H programs. The program, which
teaches older teens to present a tobacco-use prevention
program to youth aged 9 through 12 in settings outside
of school, is currently being evaluated (Project Califor-
nia 4-Health 1992).

  Two programs are noteworthy because they have
been designed to reach high-risk youth. Girls Inc. (for-
merly Girls Clubs of America) is a nationwide (120-&y)
network of over 200 centers serving young girls aged 6
through 18; over half of these girls belong to racial and
ethnic minority groups. The organization's Friendly
PEERsuasion program focuses on avoiding substance
abuse (Girls inc. 1991). Developed under a grant from
the Office for Substance Abuse Prevention, Friendly
PEERsuasion uses an older-to-younger peer leadership
approach to encourage girls aged 11 through 14 to
choose healthy alternatives to using illegal drugs, alco-
hol, and tobacco. The Boys and Girls Clubs of America,
a nonprofit organization that provides programs in sev-
eral areas, including health and physical education, has
recently established clubs (built on the structures and
supports of the Boys and Girls Clubs of America) in
several housing developments around the country.
Dubbed the SMART Moves (Self-Management and I+-
sistance Training) program, these clubs aim to prevent
substance abuse (including tobacco use) among high-
risk youth by also targeting parents and the community
(S&i&e, Orlandi, Cole 1992).

   To counter the association between baseball and
smokeless tobacco use, Little League Baseball, Inc., with
the support of the NC1 and NIDA, has developed for
young players two pamphlets that emphasize the
negative social consequences of smokeless tobacco. A
more extensive program for preventing smokeless to-
bacco use among youth who are baseball players is
currently being evaluated among Little League and
Senior League teams in Harris and Galveston counties
in Texas (Evans, Raines, Getz 1992). This intervention
targets players and their parents and involves profes-
sional baseball players.

  In 1987, .a program developed and implemented
in 72 of the 4-H clubs in 24 California counties targeted
reduction of smoking and smokeless tobacco use
(D'Onofrio, Moskowik, Braverman, unpublished data).
Club members aged 10 through 14 years were involved
in the study; 68 percent of the sample were retained at
the tweyear follow-up. The program included five

tobacco-related outcome variables-knowledge, attitudes,
perceived social influences, intentions, and behaviors-
and involved five sessions of tobacco education provided
at the monthly club meetings by volunteers (41 adults and
26 teens) trained to deliver the program. At the first
follow-up (one year later), the program demonstrated a
significant impact on participants' knowledge of the harm-
ful effects of smokeless tobacco use and on participants'
intentions to smoke, but the program had no effect on
actual use of smokeless tobacco. The two-year follow-up
showed no difference between members of clubs receiv-
ing treatment and members of control clubs. The authors
concluded that providing a tobacco-prevention program
through 4-H clubs was difficult to manage because of time
constraints on club meetings, but the effort proved to be a
useful complement to school-based programs to change
social norms.

  Other youth organizations that incorporate tobacco-
use prevention as part of a general emphasis on prevent-
ing substance abuse include the YWCA (Condas 1992),
Camp Fiie Boys and Girls (Emerson 19921, the Boy Scouts
of America (Grau 1992), and the Girl Scouts of the U.S.A.
(Eubanks 1992).

  The National Parent Teacher Association (PTA) has
adopted a number of resolutions that recognize the haz-
ards of tobacco use and support educational programs
and community policies to discourage tobacco use (Na-
tional PTA 1984). However, the organization's materials
for parents about drugs do not discuss tobacco use.

  "Just Say No" International is an organization
founded in the late 1980s to promote local clubs for youth
aged 7 through 14 years. These clubs give children infor-
mation, ski%, and support to help them resist drugs,
including tobacco ("Just Say No" International 1992). The
parent organization and the 11,000 local clubs are largely
funded through private sources and are based in schools
and community settings, including some public housing
sites. Activities include education, recreation, outreach
and peer-education, and community service. An evalua-
tion of 12 local clubs that had been active for at least one
year revealed that these clubs can offer young people a
meaningful role in improving the community, strengthen-
ing community ties, helping community members com-
mit to drug-use prevention, and coordinating other
prevention efforts (Duper 1992).

Prevention Programs Initiated by the Tobacco
Industry

  Since 1984, the Tobacco Institute has distributed a
series of publications intended to discoumge chilti fzvm
smoking (National Association of State Boards of Education
[NASBEI 1984,1987;TobaccoObserver1984). AlthoughalIof
thesepublicationsemphasiz.edecision-makingskills,onlythe

Praention   143


Surgeon General's Report

most recent, Tobacco: Helping Youth Say No, actually focuses
on tobacco use (Tobacco Institute 199Oa, b). The program's
cosponsor, The Family COURSE Consortium (Communi-
cation through Open minds, Understanding Respect and
Self Esteem) has approached schools and worked with
school districts in four major cities to determine the content
of their program (8launstein 1991). Although promotional
materials include testimonials and endorsements, no data
concerning the effect of these programs are available.

The first program sponsored by the Tobacco In-
stitute was Helping Youth Decide (NASBE 1984). The
program's focus is on parentxhild communication
skills and responsible decision making (NASBE 1984;
Coulson 1985). The program acknowledges that young
people should not smoke, but the program itself offers
no specific advice on preventing tobacco use (NASBE
1984).

  In 1987, Helping Youth Decide was supplanted by
Helping Youth Say No (NASBE 1987). Both programs
were published in conjunction with NASBE. Like its
predecessor, Helping Youth Say No focuses on parent-
child communication and on adolescents' decision-
making skiIls. NASBE was criticized by a number of
individuals and organizations for its involvement with
the Tobacco Institute and eventually ended its associa-
tion with the program.

  The current version of Helping Youth Say No con-
sists of a booklet entitled Tobacco: Helping Youth Say No-
A Parent's Guide to Helping Teenagers Cope with Peer
Pressure. Provided at no charge, these booklets are de-
signed "to increase communication between parents and
children and to raise levels of mutual trust and respect."
The text discusses the role of peer pressure in yotmg
peoples' lives, helps parents talk with their child about
not using tobacco, and includes practical exercises to
increase parent-child communication. The booklet is
likely to appeal to both smoking and nonsmoking par-
ents, since smoking is described as an adult choice
(DiFranza and McAfee 1992). This booklet would not
likely affect adolescent behaviors because it is directed at
parents, who rarely participate in such programs with-
out an incentive (Perry et al. 1989). The materials also do
not attempt to set new peer-group norms or encourage
peer leadership. Although the program does not specify
whether it is to be used as a school-based curriculum, it
would not meet the recommended criteria established by
the NC1 in conjunction with a panel of smoking preven-
tion experts CGlynn 1989; see Table 4).

Prevention Programs Sponsored by Health-Related
Organizations

  Most of the programs developed by voluntary
organizations to prevent smoking among youth are

offered as part of a school cunicuhu-n. An exception is
the American Cancer Society's preschool smoking-
prevention program Starting Free--Good Air for Me,
which includes various home activity sheets and group
activities for preschool settings (ACS 1987). This pro-
gram was tested among 86 families in four primary care
medical settings.  Results indicated that children ex-
posed to the program were almost three times as likely
as others to report that they intended to protect them-
selves from adult cigarette smoke W-riIips et al. 1990).

  The American Lung Association disseminates the
Unpuffables, a four-week, home-based program de-
signed to help parents and children aged 9 through 12
years discuss the issue of preventing tobacco use. Pilot
tests of the Unpuffables program in schools in Minne-
sota and Massachusetts and with Camp Fire and YWCA
youth groups in Oklahoma showed that parents were
aware of and approved of the program (Perry et al.
1990; American Lung Association of Green Country
Oklahoma, unpublished data).

  The American Lung Association has been active
in the area of adolescent smoking cessation. In 1988, a
technical advisory group on adolescent smoking cessa-
tion reported that demands in this area were unmet and
research questions unanswered (Hitchcock 1991). Lo-
cal affiliates of the American Lung Association have
developed one of the few available programs for smok-
ing cessation among adolescents-Tobacco Free Teens,
which is used by schools and other organizations in 25
states and 84 local affiliates (Terwedo 1992). A recent,
limited evaluation showed lower cessation rates and
higher dropout rates than were observed in American
Lung Association programs targeting smoking cessa-
tion among adults (American Lung Association 1991).

  The American Cancer Society, American Heart
Association, and American Lung Association joined
together in 1988 to launch the Smoke-Free Class of 2000
program. The goal of this education effort is to help the
cohort of young people who were first graders in 1988
remain tobacco-free when they graduate in the year
2000. The project reaches about 2 miIIion students and
135,000 teachers nationwide. As students enter junior
and senior high school, learning activities will shift
from information to community advocacy, creating
"youth ambassadors" for a smoke-free society.

Tobacco-Control Advocacy Organizations

  DOC, the organization for health professionals
that has more than 150 chapters in 23 countries, encour-
ages physicians to counteract the promotion of tobacco
to young people (Blum 1980; DOC 1992). Proactive and
prohealth strategies in the classroom, clinic, and commu-
nity use humor and ridicule of tobacco products and

144 Prevention


Preventing Tobacco Use Among Young People

tobacco industry messages to call attention to the mar-
keting of tobacco to children. DOC chapters sponsor
youth sports teams and leagues with an antitobacco
message, support local minority organizations and events
such as the Cincinnati Smoke-Free Jazz Festival, and
make "housecalls" (protests) at youth-appealing events
sponsored by tobacco companies. DOC has aIso estab-
lished a program whereby medical students can teach in
school-based smoking prevention efforts and become
special&s in school and community health promotion
(Shank 1985). DOC's leadership in innovative activities
has been noted nationally and internationally, and these
activities have been replicated or have been the basis for
many communitywide programs.

  Other tobacco-control advocacy organizations,
such as Stop Teenage Addiction to Tobacco (STAT),
SmokeFree Educational Services, Inc., and Americans
for Nonsmokers' Rights, sponsor many other creative
and effective community-based events, chapters, and
conferences. Although the results of these organiza-
tional efforts are not usually published in scientific jour-
nals, their contributions to smoking-prevention
programs and policies in the United States are widely
recognized.

  STAT, for example, is the only organization in
the United States dedicated solely to issues of teenage
access to tobacco. Public education and information
form a major part of STAT's activities. Central to this
are the STAT newsletter, the Tobacco Free Youth Re-
porter, which appears quarterly and is sent to over
100,000 persons worldwide. This newsletter, along
with STAT-authored journal articles and press adviso-
ries and a STAT-sponsored annual conference, has
been used to present and analyze the practices of the
tobacco industry. Statewide and community projects
to reduce sales of tobacco products to youth have also
been central to STAT's activities since its inception.
Currently, STAT has a major grant from the Robert
Wood Johnson Foundation to expand activities re-
lated to teenage access to tobacco in communities in
four states and to demonstrate how other communi-
ties can take similar actions.

  The Teens as Teachers program has been created
and disseminated by the American Nonsmokers' Rights
Foundation. Teens as Teachers reaches young people
most vulnerable to tobacco addiction. Although many
current smoking-prevention programs do a good job of
teaching adolescents how to resist peer influence, Teens
as Teachers also teaches them to think critically while
examining both the nature of the tobacco industry's strat-
egies and their right to be protected from primary and
secondhand smoke. Teens as Teachers has reached over
1,000 high school students, who in turn have reached
over 6,000 elementary and middle school students.

Role of the Mass Media in Reducing Tobacco
Use

Introduction

  Mass media are particularly appropriate prohealth
channels for tobacco education among yotig people,
who are heavily exposed to-and often greatly inter-
ested in-the media (Minnesota Department of Health
1989). However, although the general public has re-
ceived many antismoking messages in one form or an-
other since the 1964 Surgeon General's report on smoking
and health (Warner 1989), few messages have been de-
signed specifically to prevent young people from trying
tobacco.

Programmatic Use of Mass Media to Reduce
Adolescent Tobacco Use

  By the early 198Os, the Office on Smoking and
Health had responded to the lack of media messages
discouraging tobacco use among youth by developing
a series of national public service announcements (see
Table 7).  The major voluntary health agencies have
also produced a national broadcast message for youth.

  DOC began creating counteradvertising in 1977,
often involving young people in designing parodies of
tobacco advertisements. DOC purchased advertising
space, used counterpromotions (e.g.; the Emphysema
Slims Tennis Tournament) (Solberg 1992), and encoun-
tered occasional censorship (Fitzgerald 1990). DOC
has maintained visibility by enlisting medical profes-
sionals, youth, and parents for innovative media- and
community-based antismoking campaigns. The pro-
gram has not been formally evaluated.

  Young people have also been a major (but not
exclusive) target group of several important statewide
tobacco-use prevention and cessation campaigns. At
their onset in the late 1980s and early 199Os, campaigns in
Minnesota, Michigan, and California used funds from
dedicated cigarette taxes to fund multimedia promo-
tions. The programs have received funding for several
years. These states have employed sophisticated mar-
keting techniques (i.e., they have used marketing ex-
perts, focus groups, pretesting, pilot campaigns, and
ongoing evaluations) to increase their effectiveness and
have arranged for extensive paid and donated advertis-
ing to ensure adequate reach and frequency of statewide
coverage (Minnesota Department of Health 1991; Kizer
and Honig 1990). Each of these campaigns also included
an outdoor billboard or poster component that mirrored
themes in the broadcast media. In 1989, the Michigan
Legislature dedicated revenues from a tax on computer
software (about $9 million per year) to health promotion,
primarily for AIDS and smoking education (Moore &

Prevention   145


Surgeon General's Report

Table 7.  Major mass-media campaigns to prevent tobacco use among young people, United States, 1983-1992

Source and dates
Year of survey

Campaign description            Representative spots

Office on
Smoking and
Health
(1983-1990)

A series of W spots
with attractive images of
young people dancing or playing
sports; the general theme is
that living is positive and
smoking is out of fashion

Cigarette Mash
Nit (A Teen)

National Cancer Institute
(1987)


American Lung
Association (1988)

. . `-

Radio campaign `1' :   . :_,,.                    . _ _, ,. -.      .._,. ..".
                 1                 : i
feat$ngna&naIrac+ `:-:-. -:      .: Smoking's Out
                       .`I I      .i `.< .-  .- *-..v;; .~,
personality Casey K&em                         i-,.`,T ._ `4.. :. 4
                                         .-                 ., i
                     . ,_ ..e -._   .~  1 "_ . bi c_- 5 .*... I~_i -_    _ :-/... ___b ..t
TV spot with awareness           Cigarettes Are Drugs
message

1% Prevention


Preventing Tobacco Use Among Young People

Format and duration
(in seconds)

Content

TV (60)

Dancing girls stomp on cigarettes to model
quitting; viewers invited to write in for poster

TV (60)           Cartoon of a "butthead" getting shunned by peers

TV (30)

A boy in a rundown neighborhood appears to
be buying drugs, but it's a pack of cigarettes

TV 60)
TV (30)
TV (30)
TV (15)

Fast-paced music video: smoking's not cool
Cartoon: young kids are smart and don't smoke
Tobacco executives joke about "getting" smokers
Disgusting look of a cigarette butt in the mouth

TV (30)

Three boys show disgust for a girl's smoking

Prmenfion 147


Surgeon General's Report

Associates, Inc. 1990). The Michigan Department of Health
invited representatives from television stations and news-
papers to participate in the creative process; the multirne-
dia campaign has included paid and public-service
broadcast time, as welI as space on television, radio, bill-
boards, and buses.

  Several other state health departments have devel-
oped smaller campaigns. In 1986, Arizona created a
smokeless-tobacco-prevention campaign that included a
short television message (or "spot"), a series of peer-
influence radio spots, a poster, and a ballplayer spokes-
person (Arizona Department of Health Services 1986).
Indiana created a television spot to discourage smokeless
tobacco use (Indiana State Board of Health 1992); a
smoking-prevention campaign with monthly broadcast
spots was conducted in Alabama (Alabama Department
of Public Health 1992); and in Tennessee, a local television
spot was used to support the Smoke-Free Class of 2000
school program (Tennessee Department of Health 1992).

  State health departments often use advertising
agencies and production companies to create their cam-
paign messages. The campaign in Vermont, however,
used materials developed previously by other states
and by a research grant from the University of Vermont
(Flynn et al. 1992). Using focus groups of Vermont
children, the Vermont Department of Health pretested
the existing materials (including 15-second messages
titled "Girl Mouth" and "Boy Mouth") borrowed from
Michigan and the "Smoking Is Real Gross" spot pro-
duced by the ACS. The spots that were rated highest by
the focus groups were included in Vermont's 1992 state-
wide campaign.

  Most of the major mass-media campaigns listed
in Table 7 employed social influence strategies similar
to those that were successful in school-based smoking-
prevention programs. The California campaign, how-
ever, focused more on information-based approaches
and most prominently on a strategy to deglamorize
tobacco use by exposing the business side of the to-
bacco industry and by repositioning tobacco marketers
as playing a significant role in the problem of adoles-
cent tobacco use (Kizer and Honig 1990).  Messages
alerting young people to the negative impact of tobacco
promotion were also included as a part of research-
oriented campaigns (discussed later in this section) in
Richmond, California (Hunkeler et al. 1990), and in
Vermont (Flynn et al. 1992), but results have not yet
been published about the effectiveness of these specific
messages. It has yet to be established that making
young people aware that they can be vulnerable targets
of tobacco advertising contributes to smoking preven-
tion (McKenna and Williams 1993).

Theory and Research on Using Mass Media
to Reduce Adolescent Tobacco Use

  During the past 20 years, various ideas have
emerged on using mass media effectively to prevent the
onset of tobacco use or bring about its cessation among
young peopie. An important article by Flay, DiTecco,
and Schlegel (1980) expanded previous information-
based models to include new elements that would in-
creasethelikelihoodofpromotingandmaintaininghealth
behaviors through the mass media. These elements in-
cluded techniques to ensure that messages are attended
to, comprehended, and accepted, as well as techniques
to convey skills, stimulate social interaction, and rein-
force behavior. Schilling and McAlister (1990) integrated
sociaI and behavioral research and theory into media-
based prevention strategies for tobacco and drug use.
Further, DeJong and Winsten (1990) incorporated more
developed principles of social marketing and experi-
ences of researchers and other practitioners in health
promotion and commercial marketing to present a de
tailed set of recommendations on the use of mass media
to prevent substance abuse.

  As in the case of national campaigns, research on
the use of mass media to bring about the prevention or
cessation of tobacco use among young people has been
sporadic and may warrant further commitment at the
national level (Bauman 1992). The best-organized re
search effort was coordinated in the mid-1980s through
the NCI's Smoking, Tobacco, and Cancer Program
(Bettinghaus 1988). Three research grants coordinated
by this program tested approaches for using mass media
for smoking prevention and cessation among young
people.

  The first of these studies, at the University of South-
em California (Hay et al. 19881, evaluated a strategy
developed in previous projects. In that strategy, school-
based programs that emphasized ski& to resist sociaI
influences to smoke were extended to include segments
on southern California's evening news broadcasts
(Sussman et al. 1987). Although school programs were
effectively carried out, the television segments were not
able to meet the objectives of the study, because the
commercial news organization and its Labor contracts
did not allow the newscast to include scripted demon-
strations of prevention skills. Researchers from the uni-
versity were not able to participate in the production
process, nor were they able to pilot-test the television
segments. The authors conclude that "the resulting pro-
gramming did not demonstrate social resistance skills in
the progressive and detailed way that is necessary for
adequate learning to take place" (p. 604).

  The second study, at the University of North
Carolina at Chapel Hill (Bauman et al. 1988), used

148 Prmention


Preventing Tobacco Use Among Young People

contemporary marketing techniques coupled with be
havioral science theory to develop three campaigns that
could be practical and inexpensive enough to be dissemi-
nated nationally if proven successfuI. A radio campaign
used eight messages about expected consequences of
smoking. Another radio campaign invited young people
and their friends to enter a sweepstakes by pledging not
to smoke. Lastly, a television campaign combined these
two approaches. These campaigns were conducted as
paid media, not as public service announcements. The
intervention, which involved 10 media markets in the
southeastern United States, was expected to reach 75
percent of its adolescent target audience during 1985 and
1986. Although none of these campaign approaches
resulted in reductions in the onset of smoking, improve-
ments were observed in two important psychosocial fac-
tors-the expected utility of smoking and friends'
approval of smoking (see "So&I Support for Smoking"
and "Subjective Expected Utility' in Chapter 4). The
authors also found that radio was as effective as televi-
sion for reaching the adolescent audience (Bauman,
Padgett, Koch 1989; Bauman et al. 1991).
  The thixd study, at the University of Vermont (worden
et al. 19881, tested the ability of mass media interventions to
inaease the efficacy of a school-based smoking-prevention
pm In this intervention strategy, media and school
programssharededucationalobpctivesbutwereotherwise
independent. A total of 36 television and 17 radio messages
were developed by using extensive diagnostic and forma-
tive resear& with students in grades 4 through 10. The
messages were broadcast in a four-year paid campaign in
cities in Montana and the northeastern United States from
1986 thmugh 1989. Results indicated that the smoking
prevalence for students who received both the media cam-
paign and the school program was 34 to 41 percent lower
than for students who received the school program only
(Figure 4). The study observed consistently positive results
for intervening measures (Flynn et al. 1992). An alternative
approach that used the community as the unit of analysis
ah showed a significant difference between treatment
groups over time (Flynn et aI. 1992). This campaign used
various message formats and production styles, including
nonauthoritarian appeals that avoided direct exhortations
not to smoke. The authors suggested that because the
media campaign was not expIicitly Iinked to the school
Program (e.g., $e two components did not share materials,
designs, or slogans), adolescent viewers may have perceived
that young people across the nation were receiving the same
nonsmoking messages-and that nonsmoking was indeed
the norm.

  Other than the three studies funded by the NCI,
little mass-media research has been directed at adolescent
smoking. The recent California mass media campaign
included young people as a major target audience; about

one-third of the television messages, one-quarter of the
radio messages, and over one-half of the outdoor adver-
tisements addressed young people as welI as other speci-
fied groups (e.g., pregnant women, young adults, adults)
(Kizer and Honig 1990).

  Although the goals of the California campaign in-
termingle youth and adult priorities, the goals that seem
to apply to youth are those that deglamorize the myths
about tobacco use, expose problems created by the to-
bacco industry, and provide information about the haz-
ards of smoking. A few spots touch on these topics (Table
7), but several others, said to be targeted to the youth
audiences in the California media plan, seem to be in-
tended for adults, such as spots about youth access to
cigarette vending machines and about spots that show
children worrying about their parents' smoking. Mea-
surements before and after campaign waves, however,
indicated significant changes in message awareness
(Popham et al. 1991), and a report by Glantz (1993) indi-
cates an association between the media campaign and a
decline in cigarette consumption throughout California.
Recently released data suggest, however, that this decline
is not being observed among youth (Pierce et al. 1993).

Figure 4.

20 -

  18 -
s
; 16-
.Ei 14-
E
k 12-
3 lo-
% 8-
2
z 6-
1 4-

Smoking prevalence in University of
Vermont program using mass media to
prevent adolescent smoking

o-
     1     2     3     4     5     6
            Study year
    School program only
- - School program and media campaign
Source: Adapted from Flynn et al. (1992).

Prevention   149


Surgeon General's Report

  Mass media were also used in the Midwestern
Prevention Project, a multicomponent community pro-
gram (Per&, MacKinnon, Dwyer, et al. 1989) in Kansas
City in 1987, but effects of the media were not assessed
separately. An evaluation of the statewide Minnesota
campaign indicated that youth were aware of the nega-
tive personal and social consequences of smoking and
could recall two campaign themes-that "smoking is
unnatural" and that "not many kids my age smoke"
(Minnesota Department of Health 1991). Mass media
were also an integral part of a community-based
smoking-cessation program for minorities in Richmond,
California, in which billboards, bus posters, direct
mail, television, coverage on a national evening televi-
sion news show, and rap music video presentations sup
ported community program activities. Both participation
and awareness were high among these minority youth,
although summary results have yet to be reported
(Hunkeler et al. 1990).

Effective Designs for Mass-Media Campaigns

  Although mass media in the United States have
been used to convey messages urging youth not to use
tobacco, efforts to use the media for this purpose have
been meager when compared with the highly coordi-
nated, well-funded campaigns of tobacco advertisers. In
the absence of a national campaign against tobacco use,
with coordinated themes and paid counteradvertisirtg,
state agencies and voluntary organizations have launched
short-term efforts that have had limited evaluations of
their impact. Research on the potential uses of the media
has been restricted to a few experimental studies using
divergent media strategies, and only one of the studies has
resulted in a significant reduction in smoking among ado-
lescents (Flynn et al 1992).

  Although a national commitment to using mass
media to prevent tobacco use among youth has been
limited, sufficient evidence now exists to examine this
tactic further. The effectiveness of a large-scale mass-
media and school-based program has been demon-
strated in the University of Vermont study (FIynn et al.
1992), albeit with largely white student populations in
northern states. In addition, several applicable prin-
ciples of effective campaign design have been identified
within the disciplines of marketing, advertising, health
education, and the social sciences (Flay, DiTecco,
Schlegell980; Flay 1986; Schilling and McAlister 1990;
DeJong and Winsten 1990; Flay and Burton 1990; Flynn
et al. 1992). These principles, which are discussed be-
low, can be applied to future mass media programs for
young people.

o In planning campaigns to prevent tobacco use, target
groups should be carefully differentiated. If a cam-
paign is aimed at youth only, it may be best to separate
it from community or school ties and to use media and
message formats that appeal to youth only (Flynn et
al. 1992). -Even within the youth population, segmen-
tation (e.g., by age, gender, racial/ethnic group) may
be necessary. If the campaign is community based,
either for youth or their parents, it should closely
connect with community resources and appeal spe
cifically to either the youth or the parent target group-
not to both (Hunkeler et al. 1990).

o The planning of prohealth campaigns for young people
should attend to the critical issues of message design
identified in the literature (Flay, DiTecco, Schlegel
1980; Flay 1986; SchiIling and McAlister 1990; DeJong
and Winsten 1990; FIynn et al. 1992). These issues
include appealing to the needs and interests of the
target group (e.g., peer approval, freedom, autonomy);
using peer models, image appeals, or lifestyle appeals
instead of cognitive appeals; providing novelty and
humor (Blum 1980); avoiding exhortation; using ce-
lebrity spokespersons cautiously; and demonstrating
preventive skills.

o Messages should be carefully scrutinized by knowl-
edgeable persons and by representatives of target
groups to ensure that these messages are not convey-
ing unintended effects that may eclipse their positive
value (Flay and Burton 1988). Antismoking messages
that show young people smoking or asking someone
for a cigarette may unintentionally employ powerful
images of the social functions of smoking, particularly
if the supposedly negative role model is in any way
attractive or appealing to the target audience. These
images may greatly outweigh the impact of a voice
over narrator's message-a message that could be
almost meaningless to the image-oriented target group
of young people.

o Diagnostic and formative research, including surveys
and focus groups, should be employed at appropriate
points throughout the creative process. Diagnostic
research can identify perceptions and needs in the
target audience that are critical for concept develop
ment (Worden et al. 1988). Formative research, at
both preliminary and advanced stages of message
execution, avoids potentially damaging, unintended
message effects (Flay and Burton 1988) and gives
producers confidence that the message wilI be ac-
cepted and appreciated by the target audience. Pre-
testing during the execution phase is critical for
messages aimed at youth, because much of the

150 Prevention