Attempts to quit smoking and relapse: Factors associated with success or failure from the ATTEMPT cohort study
Introduction
Smoking cessation is a dynamic process that often involves a sequence of unsuccessful attempts to quit before long-term abstinence is achieved. Although a significant number of smokers express a desire to quit smoking, only about one-third of them actually attempt to quit each year and few are successful (Rigotti, 2002, U.S. Department of Health and Human Services (USDHHS), 1990). Approximately 75% to 80% of smokers who attempt to quit relapse before achieving 6 months of abstinence. Of the remainder, relapses may occur years after a smoker initially quits (USDHHS, 1990). Making a quit attempt and maintaining abstinence after it are the two main components of the process of quitting (Hyland et al., 2006, West et al., 2001).
Many studies have addressed factors associated with smoking cessation. However, most of the studies are either cross-sectional or limited to specific populations (e.g., patients with specific disease, or receiving specific treatments). To our knowledge there are only two longitudinal studies that have looked at the effect of a wide range of variables both on predicting quit attempts and the success of those attempts in general populations (Hyland et al., 2006, West et al., 2001). In both cases, respondents were followed 1 year after the initial measurements and asked to report on any quitting in the intervening year. Attempts to stop smoking in the year between the two assessments were consistently predicted by past quitting behavior, intention to quit, concerns about the harmful effects of smoking, and lower enjoyment of smoking. Partner pressure to stop in the one study that examined it was predictive of quit attempts (West et al., 2001). Based on these studies, it appears that factors that predict the success of quit attempts are different from those that predict making an attempt.
These studies were limited by the long recall period; subjects may have forgotten attempts made during the year (Gilpin & Pierce, 1994). A second limitation is that several possible predictors were not canvassed (e.g., weight gain, physical and mental health, use of cessation aids). Another limitation of the Hyland and West studies was that they could not assess the influence of dynamic factors that change over time.
The ATTEMPT cohort study (West et al., 2006) followed smokers every quarter (3 months), allowing (1) more accurate estimates of the quit attempt and relapse rates and (2) exploration of predictors of a quit attempt in the near future and the subsequent relapse or abstinence. Measures were taken of a wide range of variables, including motivation, prior quit history, weight and weight concerns, use of cessation aids, and physical/mental conditions. Because of frequent measurements, the study also provided a unique opportunity to examine the role of time-varying factors in promoting quit attempts or relapse. The original ATTEMPT publication describes the overall methods for recruitment and data collection in more detail (West et al., 2006). In this original publication, the authors report results from the initial 12 months of the study, which demonstrate the feasibility of conducting a longitudinal survey among smokers via the Internet (West et al., 2006). A second publication using the ATTEMPT data (West & Zhou, 2007) reported that nicotine replacement therapy without formal behavioral support was associated with improved long-term abstinence rates (West & Zhou, 2007).
In this paper, we focus on the respondents who were enrolled after the first year of the study and completed at least one follow-up assessment during the 18-month follow-up period. We first estimated the cumulative attempt rates and relapse rates at 3 months, 6 months, and up to 18 months. Then, we used a discrete time survival analysis (DTSA) model to examine the association between potential predictor variables and the two smoking cessation outcomes: (1) attempts to stop smoking and (2) relapse among those who made quit attempts.
Section snippets
Data source
Briefly, the ATTEMPT study was a longitudinal, prospective, Web-based cohort study of smokers aged 35–65 years who smoked at least five cigarettes daily. To obtain a reasonable number of quit attempts during the follow-up period, we restricted the sample to those who expressed a willingness to quit smoking in the next 3 months. Participants were recruited from existing large Internet panels maintained by Harris Interactive, Inc., in the United States (US), Canada, the United Kingdom (UK),
Cohort description and retention
The multinational cohort for this analysis includes 2431 subjects from Canada (3%), France (35%), Spain (12%), the UK (30%), and the US (20%) who were followed for 3–18 months, i.e., completed at least the month 3 assessment, excluding those who were pregnant during the study period or provided inconsistent responses to their smoking status questions. Table 1 summarizes the characteristics of these subjects at baseline. Of these, 75% (n = 1835) completed the second follow-up assessment (month 6),
Discussion
This study suggests that two different sets of factors predict different components of the cessation process: quit attempt and subsequent relapse. Higher motivation and intent to quit predicts making a quit attempt but is not associated with subsequent relapse. Smokers who have recently made a failed attempt to quit are more likely to try again but also more likely to relapse than those who have not tried recently. These findings are consistent with findings from a previous prospective study (
Acknowledgments
Data for the online assessments were collected by the Harris Interactive Inc. on behalf of RTI Health Solutions and Sanofi-Aventis Recherche et Developpement.
References (17)
- et al.
The role of weight concern and self-efficacy in smoking cessation and weight gain among smokers in a clinic-based cessation program
Addictive Behaviors
(1998) - et al.
Weight concerns and smoking in a general population: The Inter99 study
Preventive Medicine
(2007) - et al.
A comparison of four self-report smoking cessation outcome measures
Addictive Behaviors
(2004) Survival analysis using the SAS® system: A practical guide
(1995)- et al.
Measuring smoking cessation: Problems with recall in the 1990 California Tobacco Survey
Cancer Epidemiology, Biomarkers & Prevention
(1994) - et al.
The Fagerstrom Test for nicotine dependence: A revision of the Fagerstrom tolerance questionnaires
British Journal of Addictions
(1991) - et al.
Applied logistic regression
(1989) - et al.
Signs and symptoms of tobacco withdrawal
Archives of General Psychiatry
(1986)
Cited by (402)
Associations between nicotine product use and craving among stable daily and non-daily users
2023, Addictive BehaviorsNegative affect and craving during abstinence from smoking are both linked to default mode network connectivity
2023, Drug and Alcohol Dependence