Independent and interactive effects of real-time risk factors on later temptations and lapses among smokers trying to quit

Abstract

Purpose: The current study sought to expand our understanding of relapse mechanisms by identifying the independent and interactive effects of real-time risk factors on temptations and the ability to resist temptations in smokers during a quit attempt.

Procedures: This study was a secondary analysis of data from 109 adult, treatment-seeking daily smokers. Ecological momentary assessment data was collected 4 times a day for 21 days following a quit attempt and was used to assess affect, urge, impulsiveness, recent cigarette exposure, and alcohol use as predictors of temptations to smoke and smoking up to 8h later. All smokers received nicotine replacement therapy and smoking cessation counseling.

Findings: In multinomial hierarchical linear models, there were significant main (agitation odds ratio (OR)=1.22, 95% CI=1.02-1.48; urge OR=1.60, 95% CI=1.35-1.92; nicotine dependence measured by WISDM OR=1.04, 95% CI=1.01-1.08) and interactive effects (agitation×urge OR=1.12, 95% CI=1.01-1.27; urge×cigarette exposure OR=1.38, 95% CI=1.10-1.76; positive affect×impulsiveness OR=2.44, 95% CI=1.02-5.86) on the odds of temptations occurring, relative to abstinence without temptation. In contrast, prior smoking (OR=3.46, 95% CI=2.58-4.63), higher distress (OR=1.30, 95% CI=1.06-1.60), and recent alcohol use (OR=3.71, 95% CI=1.40-9.89) predicted smoking versus resisting temptation, and momentary impulsiveness was related to smoking for individuals with higher baseline impulsiveness (OR=1.12, 95% CI=1.04-1.22).

Conclusions: The risk factors and combinations of factors associated with temptations and smoking lapses differ, suggesting a need for separate models of temptation and lapse.

Keywords: Ecological momentary assessment; Relapse; Smoking cessation; Temptation; Tobacco.

Figure 1. Time since quit day interactions with agitation, distress, and urge separated by recent smoking status

Moderating effects of time since quit day on relations between agitation, distress, and urge and experiencing temptations or smoking separated by whether recent smoking occurred before the report (yes/no). Each line represents the regression coefficient for the predictor at each time point: week 1, week 2, week 3. The coefficients for most of these effects (with the exception of distress and smoking vs. temptation) were most divergent from zero (indicating stronger effects) in the first week post-quit compared to the following one to two weeks. Figure Legend: ■Agitation ▲Distress ●Urge Solid line=temptation vs. abstinent, Dashed line=smoked vs. temptation.

Figure 2a-c. Level-one interactions predicting log odds of temptation vs. abstinence

Interaction effects of urge x agitation, urge x cigarette exposure, and positive affect x impulsiveness on the log odds of experiencing a temptation. To plot these effects for the continuously scaled constructs, log odds were calculated from low (1 SD below mean), mid (mean), and high (1 SD above mean) range values. Impulsiveness values reflect square-root percent commission errors on a momentary behavioral disinhibition task. Log odds can be exponentiated to derive odds ratios of experiencing a temptation (e^x). Figure Legend: 2a. Solid line=high agitation. Dashed line=low agitation. 2b. Solid line=easy access to cigarettes. Dashed line=no cigarettes available. 2c. Solid line=high impulsiveness. Dashed line=low impulsiveness.

Figure 3. Cross-level interaction between state and trait impulsiveness predicting log odds of smoking vs. temptation

Interaction between state impulsiveness (measured by momentary behavioral disinhibition) x trait impulsiveness (measured by the Barratt Impulsiveness Scale, BIS-11). To plot these effects, log odds were calculated from low (1 SD below mean), mid (mean), and high (1 SD above mean) range values for the momentary behavioral disinhibition task. Log odds can be exponentiated to derive odds ratio of smoking (e^x). Figure Legend: Solid line=high trait impulsiveness measured by BIS. Dashed line=low trait impulsiveness measured by BIS.