Association of tobacco dependence and quit attempt duration with Rasch-modeled withdrawal sensitivity using retrospective measures

Abstract

Aim: To examine whether Rasch modeling would yield a unidimensional withdrawal sensitivity measure correlating with factors associated with successful smoking cessation.

Design: The psychometric Rasch modeling approach was applied to estimate an underlying latent construct (withdrawal sensitivity) in retrospective responses from 1644 smokers who reported quitting for 3 or more months at least once.

Setting: Web-based, passcode-controlled self-administered computerized questionnaire.

Participants: Randomly selected convenience sample of 1644 adult members of an e-mail invitation-only web panel drawn from consumer databases.

Measurements: Lifetime Tobacco Use Questionnaire, assessing tobacco use across the life-span, including demographics and respondent ratings of the severity of withdrawal symptoms experienced in respondents' first and most recent quit attempts lasting 3 or more months.

Findings: Rasch-modeled withdrawal sensitivity was generally unidimensional and was associated with longer periods of smoking cessation. One latent variable accounted for 74% of the variability in symptom scores. Rasch modeling with a single latent factor fitted withdrawal symptoms well, except for increased appetite, for which the fit was marginal. Demographic variables of education, gender and ethnicity were not related to changes in sensitivity. Correlates of greater withdrawal sensitivity in cessation attempts of at least 3 months included younger age at first quit attempt and indicators of tobacco dependence.

Conclusion: The relationship between tobacco dependence symptoms and Rasch-model withdrawal sensitivity defines further the relationship between sensitivity and dependence. The findings demonstrate the utility of modeling to create an individual-specific sensitivity measure as a tool for exploring the relationships among sensitivity, dependence and cessation.

Figure 1

Estimated sensitivity was generated from the Rasch model, which also generated the standard errors for those estimates. Sensitivity was relatively precise for those points plotted at a standard error below 0.5 and lower. Quit attempts on the lower arc were rated on all sensitivity withdrawal questions. Points along the bottom arc represent 2,569 quit attempts for which responses of 1 to 5 were available for all symptoms. Points above the bottom arc represent 131 quit attempts for which one or more responses were decline/unsure or not provided (missing)

Figure 2

Fit of responses of 1 (not at all), 3, and 5 (intense) to the withdrawal symptom increased appetite.^† †Fit of responses of 1 (not at all), 3, and 5 (intense) to the withdrawal symptom increased appetite. This figure compares the empirical probabilities of ratings of 1, 3, or 5 (dotted lines) for individuals with modeled withdrawal sensitivity in a series of ranges with the theoretical probabilities (solid lines) of corresponding ratings for the withdrawal symptom of increased appetite. Vertical bars represent 95% confidence intervals for populations for which withdrawal sensitivity was completely known. This withdrawal symptom was selected because the infit and outfit statistics showed the least satisfactory fit with the Rasch model. However, considering that the modeled withdrawal sensitivity is measured with error, the empirical and theoretical probability curves appear to be reasonably similar.