1. Introduction
Tobacco use constitutes the major cause of preventable death and disability worldwide and is considered a significant public health concern [
1]. Although smoking cessation interventions are shown to be efficacious [
2,
3], smoking relapse is a frequent and challenging phenomenon. Studies report that a high percentage of smokers receiving smoking cessation treatment and achieving abstinence relapse in the following weeks and months [
4,
5], with relapse rates ranging approximately from 30% to 80% [
6,
7].
Research has tried to identify variables predicting smoking relapse [
8], including, among others, cigarette consumption per day, years smoking, living with a smoker, tobacco dependence [
9], tobacco withdrawal [
10], craving [
11], stage of change [
12], smoking self-efficacy [
13], negative affect [
14], and even sociodemographic variables such as age, sex, or education [
15]. Among the most robust factors related to a higher risk of relapse are withdrawal symptoms [
10,
16]. Such symptoms are aversive and especially relevant during the first weeks of abstinence [
17]. Among them are included depressed mood/sadness, irritability, frustration, anxiety, anger, difficulty concentrating, restlessness/impatience, sleep-related problems, increased appetite, or weight gain [
17,
18], which are usually assessed together as a composite score. However, there is evidence that withdrawal symptomatology is highly variable between subjects [
19] and, therefore, the use of a composite measurement may overshadow the impact of individual withdrawal symptoms on smoking relapse. Negative affect has also been identified as a motivational factor of smoking behavior [
20], as individuals tend to experience increased smoking urgency in this type of situation. In fact, some studies have reported that negative affect increases during quit attempts [
21] and is associated with a higher likelihood of relapse [
14,
22].
Another significant variable involved in smoking relapse is self-efficacy, defined as confidence in one’s ability to quit or avoid smoking [
23]. This variable has been found to be associated with smoking lapse and relapse during unaided quit attempts [
24] and after receiving a smoking cessation intervention [
13]. Besides, it has been suggested that abstinence self-efficacy should be examined across different situations, as confidence in maintaining abstinence may differ in specific situations [
23]. For instance, in a cross-sectional study examining smoking cessation predictors in a sample of Chinese participants [
25], the authors found that self-efficacy related to negative-affect situations was significantly associated with successful smoking cessation, whereas positive affect and habit-related self-efficacy did not reach statistical significance. These results highlight the relevance of considering specific contextual aspects of the smoking self-efficacy construct.
In a smoking cessation treatment context, most studies have evaluated the aforementioned variables at pre-treatment. However, most smoking-related variables could be considered as dynamic [
26], which implies that they may change due to the fact of quitting smoking itself (i.e., withdrawal symptoms) or because of the cognitive-behavioral strategies used during the intervention (i.e., self-efficacy). Therefore, examining the effect of the variables that are modified by abstinence achievement and treatment components could contribute to a better understanding of which ones could hinder abstinence maintenance.
Although extensive research has been conducted on smoking relapse, to our knowledge, no studies have compared the differential predictive value between composite measures and their corresponding subscales or specific elements in the relapse of smokers who received cognitive-behavioral treatment for smoking cessation. Therefore, the aims of the study were (1) to examine the effect of post-treatment tobacco withdrawal symptomatology, smoking-related self-efficacy, and post-quit smoking urgency in negative-affect situations using a composite measure on smoking relapse; and (2) to determine whether the use of their specific subscales or items provides additional information compared with the composite scores.
4. Discussion
The present study compared the predictive value of the composite measures of post-treatment self-efficacy, smoking urgency in negative-affect situations, and tobacco withdrawal, as well as the specific predictive value of such variables, but using their subscales or their items, on smoking relapse at the 3 month follow-up in a sample of smokers who quit smoking after attending a cognitive-behavioral smoking cessation treatment.
Results showed that there were no significant differences between relapsers and abstinent participants in sociodemographic variables, pre-treatment smoking-related variables, or treatment condition received. When considering post-treatment smoking-related variables’ composite measures, none of the study variables were significantly related to smoking relapse. However, when examining the regression model including each specific subscale and item, our data showed that lower post-treatment smoking-related self-efficacy in negative-affect situations was a significant predictor of smoking relapse. Concretely, these findings are in line with previous research [
25], highlighting the importance of considering specific self-efficacy-related contextual aspects, as a global measure of confidence in one’s ability to abstain may not capture the relevance of situational factors that impact on smoking cessation outcomes [
23].
Concerning the withdrawal syndrome, only three specific symptoms (irritability/frustration/anger, restlessness, and craving) showed a significant relation with smoking relapse. Following the previous literature, our results are consistent with the idea that individual symptoms are meaningful in their own right [
40] and with previous research finding that such symptoms are strongly related to smoking relapse during the first months of abstinence [
41].
Contrary to our expectations, this study showed that post-treatment smoking urgency in negative-affect situations (anger and anxiety) did not predict smoking relapse at the 3 month follow-up. Some of the strategies trained during the smoking cessation treatment may be influencing this result, as participants could have learned to cope with smoking urgency in such situations. It is also plausible that smoking urgency itself is not enough to predict relapse, and that other variables may interact with it, such as, for example, individual impulsivity [
42] or cessation fatigue [
43]. Further research is needed to examine the impact of this variable on smoking relapse.
The findings of this study may have significant implications for smoking cessation treatments. Our data showed that the relapse predictors in this study are potentially modifiable, and therefore, could be considered intervention targets. For instance, lower self-efficacy in negative-affect situations at the end of treatment, as well as greater negative-affect withdrawal symptoms and craving, could be addressed by incorporating intermediate sessions between the final treatment session and the 3 month follow-up, in which participants could learn strategies to cope with these specific difficulties. Another possibility could be to incorporate technology-based strategies to support and maintain contact with participants after treatment completion during the follow-up period. Such technology-based strategies (i.e., tailored mobile phone messaging, social media support groups) could be used to aid participants to cope with the distress associated with quitting as well as with their confidence to be able to manage smoking urges in specific high-risk contexts or situations.
This study has several limitations that should be considered in the interpretation of results. First, the current sample comprised treatment-seeking smokers who attended a cognitive-behavioral treatment to quit smoking, and therefore, findings are not generalizable to smokers achieving unaided abstinence. Second, the current study focused exclusively on demographics and smoking-related variables. Because other psychological factors could be related to smoking relapse, such as individual cue reactivity or coping skills [
44,
45], future research should analyze the contribution of such variables in the complex process of smoking relapse. Third, study variables were measured through self-report instruments, and consequently, could be influenced by response bias. Lastly, we only considered withdrawal symptoms measured through the MNWS. Recent research has suggested that symptoms such as anhedonia [
46] or mood swings [
47] should be considered among withdrawal symptoms as well, and therefore, future studies should include such variables when examining the impact of withdrawal symptoms on relapse.