Article Text

Download PDFPDF

Intervention effects on youth tobacco use in the community intervention trial (COMMIT)
  1. D J Bowen1,
  2. M A Orlandi2,
  3. E Lichtenstein3,
  4. K M Cummings4,
  5. A Hyland4
  1. 1Fred Hutchinson Cancer Research Center, Seattle, Washington, USA
  2. 2Johnson & Johnson, Inc, Skillman, New Jersey, USA
  3. 3Oregon Research Institute, Eugene, Oregon, USA
  4. 4Roswell Park Cancer Center, Buffalo, New York, USA For the COMMIT Research Group
  1. Correspondence to:
 Deborah J Bowen, PhD, Fred Hutchinson Cancer Research Center, 1100 Fairview Ave N, MP-900, Seattle, WA 98109, USA;

Statistics from

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.

The Community Intervention Trial for Smoking Cessation (COMMIT) was an intervention trial funded by the National Cancer Institute to evaluate the effects of a multi-component, community based smoking control intervention on cessation in adult smokers.1,2 The primary (adult) outcomes of this trial have been published elsewhere.3,4 In this letter we test the hypothesis that a comprehensive, community based intervention aimed at adult smokers would have an ancillary impact on the prevalence of youth smoking.

The COMMIT intervention5 included youth oriented activities directed toward four principle areas: school based education programmes, smoking policies in schools, legislative activities related to youth smoking, and participation by students and teachers in other COMMIT activities. The evaluation involved a two group, pre-test/post-test, quasi-experimental design with community as the unit of assignment and ninth grade classroom (ages 14–15 years) as the unit of assessment. Overall classroom participation rates were 90% (8235) at time 1 and 86% (8945) at time 2.

Table 1 shows percentages and change scores (increases or decreases) in mean per cents comparing time 1 to time 2 for each study condition. None of these differences were significant.

Rank correlations were calculated contrasting pair wise differences in adolescent seven day smoking prevalence with pair wise (that is, same pair) differences in adult cohort quit rates from the 1993 COMMIT Endpoint survey.6 These adult rate differences for each community pair were correlated with youth smoking differences in the same community pair using current weekly smoking rates from the 1992 Youth Survey. The correlation was 0.2 (p < 0.001), indicating that higher quit rates are associated with higher youth smoking.

The data reported here do not support the hypothesis that the adult focused COMMIT intervention was efficacious in reducing the prevalence of regular youth smoking. Among ninth graders living in treatment communities as well as among their counterparts living in comparison communities, the general trend was toward little or no difference over the time interval assessed (1990 to 1992)—a levelling off in tobacco use rates that is consistent with national trends reported in other surveys conducted during this time period.

It is important to underscore that the COMMIT approach was without question and by design an adult focused intervention, and the design of the study was not set up to evaluate youth smoking changes. Other concerns that are relevant to the interpretation of these results include: implementation fidelity; the possibility that these activities may have been delivered inconsistently, or, at least, more effectively in some communities than in others; the age group selected for the evaluation (it is possible that the intervention had a greater effect on adolescents who were either older or younger than the ninth graders selected for our sample); and the time frame for the evaluation (that is, it is possible that the interim between 1990 and 1992 was not long enough for an intervention effect to have been demonstrated, especially given secular trends during that period).

It appears that the COMMIT intervention, which did target adult smokers, was not a cause of change in adolescent smoking behaviour. Changes in adolescent smoking rates are likely to come from other sources, such as exposure to tobacco product marketing, and broad based policies and programmes intended to discourage smoking such as cigarette taxes, limits on public smoking behaviour, and community based anti-tobacco education, and mass media messages about smoking. Targeting these influences certainly forms part of the national tobacco use reduction agenda for youth.7–10

Table 1

Percentage students by smoking status condition totals