Contextual Factors Associated With Temptations and Lapses Among Smokers Trying to Quit
An Ecological Momentary Assessment Study
Abstract
Abstract: People who try to quit smoking might relapse depending on certain contextual factors (e.g., drinking coffee). Ecological momentary assessment (EMA) was used to investigate contextual factors during temptation and lapse episodes. Contextual factors included what participants did (activities), who they were with (social environment), and where they were (location). Participants (N = 103) were smokers motivated to quit within three months. The EMA protocol (NEMA = 6,457) began on a self-determined quit day and lasted 14 days. Generalized linear mixed models were performed to assess the association between each contextual factor and temptations and lapses. Various contextual factors were positively associated with temptations (e.g., sexual intercourse) and lapses (e.g., being outdoors) relative to random assessments. Lapses relative to temptations were mainly associated with social contextual factors (e.g., being with friends). How contextual factors can be addressed in (just-in-time) relapse prevention interventions should be the focus of future studies.
People who smoke tobacco typically have poorer health and die earlier than people who do not smoke (U.S. Department of Health and Human Services, 2014). Tobacco smoking causes cancer, heart disease, lung diseases, stroke, chronic obstructive pulmonary disease (COPD), and various other diseases (U.S. Department of Health and Human Services, 2014). Public health efforts have led to an enormous decrease in smokers over time, but still 20.6% of the adult Dutch population smokes (Trimbos-instituut, 2022). Globally in 2019, about 1.14 billion individuals were current smokers (Reitsma et al., 2021). About three quarters of the adult Dutch smokers smoke daily, which is 15.2% of the adult Dutch population (Trimbos-instituut, 2022). In 2021, 30.9% of smokers made a serious quit attempt (quit for ≥24 h; Trimbos-instituut, 2022). Unfortunately, most attempts to quit smoking fail. Only about 5% of smokers achieve prolonged abstinence for 6–12 months after an unassisted quit attempt (Hughes et al., 2004). Following an assisted quit attempt, about 17% of smokers achieve prolonged abstinence for 12 months (Robinson et al., 2019). Many smokers may need multiple quit attempts to achieve prolonged abstinence (Greenhalgh et al., 2016). In fact, the term attempt might be misleading because smoking cessation can poorly be characterized as a behavioral act that either succeeds or fails. Rather, smoking cessation is a behavioral process that almost inevitably involves temptations (i.e., urges or desires to smoke, which are resolved without smoking), lapses (i.e., smoking temporarily, typically in a single instance), and oftentimes relapse (i.e., return to regular smoking). Notably, some temptations may be successfully resolved without smoking, but others may result in lapses, thereby increasing the risk of relapse. Therefore, studying temptations offers valuable insights into understanding the onset and the dynamics of smoking relapse.
Exploring the determinants of relapse is fundamental to explain, predict, and change relapse within the behavioral process of smoking cessation. Sociodemographic (e.g., socioeconomic position), physiological (e.g., nicotine dependence), and motivational factors (e.g., self-efficacy) can, to some extent, explain and predict who relapses (Vangeli et al., 2011). Yet, these factors are less helpful in explaining and predicting when (in which situations) relapse may occur (Shiffman, 2005). To answer this question, it is necessary to examine which contextual factors increase the risk of relapse. Earlier studies suggest that people usually relapse in the presence of specific contextual factors, such as being at a party (Shiffman, 2006). These contextual factors were classically conditioned to be associated with smoking during the daily life as a smoker (Benowitz, 2010; Winkler et al., 2011). Through conditioning, such factors can be strongly associated with the rewarding effects of smoking and can trigger relapse, even when the deliberate decision has been made to quit smoking. Shiffman, Gnys, et al. (1996) found that negative affect, arguing, discussion, restlessness, attention disturbance, others smoking, smoking permitted, and consumption of food, coffee, and alcohol were associated with increased risk of temptation relative to random assessments among participants in nonpharmacological group treatment. Negative affect, alcohol consumption, others smoking, cigarettes available, and smoking permitted were associated with increased risk of initial lapsing relative to temptations among participants in nonpharmacological group treatment (Shiffman, Paty, et al., 1996).
Most past research has focused on affect and other cognitive variables, while contextual factors concerning the environment, with the exception of smoking cues, have received limited attention. A recent systematic review and meta-analysis found a positive relationship between environmental and social cues and lapse incidence (Perski et al., 2023). It is important to note that the meta-analysis exclusively included smoking-specific factors, such as the influence of others smoking or the availability of cigarettes. This indicates a gap in research regarding broader contextual factors beyond smoking-specific cues. In addition, previous studies have been conducted primarily among treatment-receiving (ex-)smokers, whereas the use of evidence-based assistance among smokers who are trying to quit is low (Borland et al., 2012). In this study, we extended current research by examining how contextual factors were related to temptations and lapses among (ex-)smokers in the first 14 days of a quit attempt. We examined what participants did (activities), who they were with (social environment), where they were (location), and how these factors related to temptations and lapses. The two research questions were:
- 1.Which contextual factors are associated with risk of experiencing temptations (i.e., resolved without smoking) and lapses (i.e., smoking) relative to random assessments?
- 2.Which contextual factors are associated with risk of developing temptations into lapses compared to being resolved without smoking?
The results may serve as input for the development of (just-in-time) interventions that can assist former smokers in preventing and/or coping with situations that trigger temptations and lapses.
Method
Design
Data were collected between October 5, 2021, and March 4, 2022. The methodology was ecological momentary assessment (EMA), also known as experience sampling method (ESM), in which data are repeatedly collected in participants’ natural environment and in real time (Shiffman et al., 2008). An advantage of EMA is that it reduces recall bias compared to traditional questionnaire studies, which require participants to recall past behaviors and experiences. The EMAs were collected by a mobile application (hereafter: app) on participants’ smartphones (Movisens, 2022).
Participants
We included adult smokers (18 years or older) who reported planning to quit smoking within three months and who owned an Android smartphone, as the app was only available in the Google Play Store for the Android mobile operating system (Movisens, 2022). We included smokers regardless of whether they used or planned to use a cessation method for their quit attempt to follow the natural cessation process without intervening. This was considered important because most previous research involved intervention studies among treatment-seeking smokers, whereas the actual use of evidence-based cessation methods among smokers trying to quit smoking is low, and therefore, the results cannot be generalized to all smokers (Borland et al., 2012).
The preregistered sample size calculation (N = 100) with assumed attrition and noncompliance did not correspond to the actual circumstances found during the recruitment process. Thus, substantially more potential participants had to be recruited than originally assumed to obtain enough participants at the end of the study whose data could be included in the data analysis. Of the 739 participants who met the inclusion criteria and were invited by mail, 241/739 (32.6%) downloaded the app, 194/739 (26.3%) answered the baseline questionnaire, and 103/739 (13.9%) answered sufficient assessments to be included in data analysis. Participants who answered fewer than 42 assessments (an average of 3 per day) or answered only event-contingent or only signal-contingent assessments were excluded from the analysis and did not receive incentives.
Participants were recruited through two Dutch research agencies (Flycatcher Internet Research B.V., Maastricht, NL and PanelClix, Amsterdam, NL). Participants stemming from Flycatcher Internet Research B.V. collected points in the research agency’s system that could be exchanged for gift vouchers or donations. Participants stemming from PanelClix received a gift voucher to the value of €15 for a popular Dutch webshop.
Procedure
After completing a screening questionnaire on the inclusion criteria and providing online informed consent, eligible participants received an email with instructions on how to download and connect the app to the participant. The app was available in Dutch. Participants then completed the baseline questionnaire in the app. Thereafter, no further action was required until the day participants attempted to quit. On the day of the quit attempt, participants had to open the app again and indicate that they were attempting to quit, at which point the EMA protocol began. A delay of one hour was programmed between answering the baseline questionnaire and the possibility to indicate to attempt to quit, to prevent participants from rushing through the app.
The EMA protocol consisted of event-contingent and signal-contingent assessments. The event-contingent assessments included temptations and lapses reported by participants in the app. Hereafter, the term temptation assessment is used for the assessments that participants were required to answer when they reported in the app that they were tempted to smoke. The term lapse assessment is used for the assessments that participants had to answer when they reported that they had smoked. The signal-contingent assessments were triggered by notifications (also called beeps) at four semirandom times during the day between 08:00 and 20:45 h, with at least one hour between beeps. After the initial notification, the questionnaire was available in the app for 30 min. The goal of the signal-contingent assessment was to obtain a representative and unbiased estimate of participants’ typical state to serve as a comparison for the temptation and lapse assessments in the statistical analysis. The signal-contingent assessments are referred to as random assessments hereafter. It took participants an average of 18 (SD = 35) seconds to complete an assessment.
Measures
All measures were self-reported. The baseline questionnaire consisted of sociodemographics and smoking characteristics. Sociodemographics collected were age (in years), gender (1 = man, 2 = woman, 3 = other), and education level (1 = low, 2 = intermediate, 3 = high). Smoking characteristics included daily smoking (1 = yes, 2 = no), use or planned use of a smoking cessation method (1 = no method, 2 = face-to-face counseling, 3 = eHealth interventions, 4 = telephone counseling, 5 = group-based programs, 6 = nicotine replacement therapy, 7 = prescription medication, 8 = other, namely…), use of an e-cigarette (1 = never done/past use, 2 = current use), and nicotine dependence. Participants could indicate whether they use or plan to use several smoking cessation methods, namely face-to-face counseling, eHealth interventions, telephone counseling, group-based programs, nicotine replacement therapy, prescription medication, or other methods. Nicotine dependence was assessed by the Fagerström Test for Nicotine Dependence (Heatherton et al., 1991). The items of the Fagerström Test for Nicotine Dependence were summed to yield a total score of 0–10, with higher scores indicating greater dependence. The constructs (activities, social environment, location) and items of the questionnaire used in the EMA protocol for the random assessments, temptation assessments, and lapse assessments are listed in Appendix A.
Statistical Analysis
Separate generalized linear mixed models with random intercepts were performed to assess the association between each contextual factor and temptations and lapses relative to random assessments and lapses relative to temptation assessments. To answer the first research question, which contextual factors are associated with temptations and lapses, two comparisons were conducted: temptation vs. random assessments and lapse vs. random assessments. The second research question was addressed in one comparison: lapse vs. temptation assessments. In this comparison, temptations were used as comparator instead of random assessments because the aim was to investigate which of the temptation situations posed a particular risk of leading to lapses.
Analyses were conducted using the function glmer of the package lme4 in R (Bates et al., 2014). Contextual factors were set as binary independent variables. Models were fitted by maximum likelihood (Laplace approximation). If the model failed to converge with Laplace approximation, it was fitted by the adaptive Gauss–Hermite approximation with 20 nodes. Odds ratios (OR) with 95% confidence intervals (CI) were calculated. The R script can be found in Appendix B.
Results
Sociodemographics and Smoking Characteristics
Participants (N = 103) were on average 39.7 (SD = 12.8, range = 20–70) years of age, and the majority was female (68.9%). The representation of education levels in the sample was uneven, with the low education level comprising 12.6%, while the intermediate and high education levels accounted for 44.7% and 42.7%, respectively. Overall, participants were moderately dependent according to the Fagerström Test for Nicotine Dependence (M = 4.1, SD = 2.4). Participants smoked on average 12.5 (SD = 8.3) cigarettes per day, and almost all participants were daily smokers (94.2%). Most participants did not use or plan to use any smoking cessation method (no method: 64.1%, face-to-face counseling: 2.9%, eHealth interventions: 3.9%, telephone counseling: 0.0%, group-based programs: 1.9%, nicotine replacement therapy: 27.2%, prescription medication: 4.9%, other: 4.9%). E-cigarettes were currently used by 9.7% of participants. A post hoc attrition analyses between the sample of this study and participants who responded to the baseline questionnaire but did not answer sufficient assessments (n = 91) revealed higher attrition among male (31.1% of the sample included in the analyses was male vs. 49.5% in those that dropped out) and older participants (M = 39.7 years vs. M = 44.7 years) but no differences in nicotine dependence-related variables (e.g., number of cigarettes smoked per day or scores on the Fagerström Test for Nicotine Dependence).
Contextual Factors Associated With Temptations and Lapses Relative to Random Assessments
Of the total number of assessments (NEMA = 6,457), there were 1,781 temptation assessments, 802 lapse assessments, and 3,784 background assessments. Table 1 shows the associations between the contextual factors and temptations and lapses relative to random assessments. The descriptive statistics of random, temptation, and lapse assessments can be found in Appendix C. The activities participants engaged in influenced the risk of temptation and lapsing. Work, more specifically having a job, was associated with lower risk of temptation and lapsing. Leisure was associated with increased risk of temptation and lapsing. Regarding the type of leisure activity, several activities were associated with increased risk. Sexual intercourse was associated with a 5-fold increase in the odds of temptation. This was the strongest association between a contextual factor and temptations relative to random assessments in this study. Doing nothing was associated with increased risk of temptation and lapsing. Being social and having slept were associated with increased risk of lapsing. Doing nothing, being social, and having slept were particularly strongly associated with lapsing: Doing nothing was associated with a 218%, being social with a 133%, and having slept with a 234% increase in the odds of lapsing.
Eating and drinking were generally not associated with risk of temptation and lapsing. However, regarding the type of food or drink, coffee and alcohol consumption were risk factors. Coffee drinking and alcohol consumption were associated with increased risk of lapsing.
The social environment affected the risk of temptation and lapsing in various ways. Being with the family and being with colleagues were associated with lower risk of temptation. Being alone was associated with increased risk of temptation. Being with friends was particularly strongly associated with lapses relative to random assessments, with the odds of lapsing increasing by 117%. The presence of people in view smoking was associated with increased risk of temptation and lapsing. The association was stronger for lapses than for temptations: 277% increase and 36% increase in odds, respectively.
The location of participants was also associated with the odds of temptation and lapsing in various ways. Being in the workplace was associated with lower risk of temptation and lapsing. Being outdoors was associated with increased risk of temptation and lapsing. For lapsing, being outdoors was associated with a 5-fold increase in odds.
Contextual Factors Associated With Lapses Relative to Temptations
Table 2 shows the associations between the contextual factors and lapses relative to temptations. None of the activity categories (work, chores, leisure, eating, or drinking) were related to the odds of lapses relative to temptations. Regarding the type of leisure activity, being social was particularly strongly associated with increased risk of lapses relative to temptations, increasing the odds of lapsing by 136%. Regarding the social environment, being with friends and being in the presence of people who were smoking were associated with increased risk. People in view smoking increased the odds of lapsing more than 3-fold. Being with strangers was associated with decreased risk. Regarding location, being outdoors was associated with increased risk.
Discussion
The present study adds to a long history and robust body of literature. Yet, it is one of the few studies to provide an assessment of the role of contextual factors on smoking relapse in a sample of primarily daily smokers who were not provided an intervention. The first research question sought to examine which contextual factors were associated with risk of temptation and lapsing relative to random assessments in the first 14 days after a quit attempt. The results of this study show that contextual factors play a key role in smoking relapse, as multiple factors were associated with either lower or increased risk of temptation and lapsing.
Certain activities appeared to be important contextual factors in smoking relapse. A clear finding of the analysis was that working was consistently associated with lower risk, whereas being engaged in leisure was consistently associated with increased risk. These results are in line with those of previous studies (Shiffman et al., 2009, 2014; Stennett et al., 2018) and suggest that relapse prevention interventions should focus on contextual factors during leisure time. The lower risk of temptations and lapses during work might be related to workplace smoking bans being in place. In the Netherlands, workplaces have been required to be smoke-free since 2004. During the study period, also designated indoor smoking areas were prohibited (Rijksoverheid, 2020).
To the best of our knowledge, sexual intercourse has not previously been included in studies on smoking relapse. In this study, sexual intercourse had the strongest association with temptations relative to random assessments of all contextual factors in this study. However, due to the infrequency with which sexual intercourse was reported (which is also reflected in the broad confidence interval for this factor), caution is warranted here and further studies on this topic are needed. This finding, while preliminary, suggests that having sex should be addressed as a risk factor for relapse in interventions. Individuals for whom smoking after sexual intercourse was common could receive information to either not have sex during the first days of their quit attempt and/or be informed about preparatory action and coping plans to prevent temptation and lapsing after sexual intercourse.
In this study, a positive association was found between coffee consumption and lapses. This finding confirms prior research on the relationship between coffee consumption and smoking (Marshall et al., 1980; Shiffman & Paty, 2006; Shiffman et al., 2009; Swanson et al., 1994) and coffee consumption and relapse (Shiffman, 1982). Moreover, the results of this study indicate that alcohol use was associated with increased lapse risk. This result matches those observed in earlier studies (Shiffman, 1982; Shiffman, Gnys, et al., 1996; Shiffman, Paty, et al., 1996).
The results on social environment and location are in line with those of previous studies (Shiffman et al., 2014, 2020). Two themes can be discerned. First, being with colleagues and being at the workplace was protective against relapse, confirming the finding that working was associated with less risk. Second, being with friends and being outdoors was detrimental to relapse, confirming the finding that leisure and being social were associated with greater risk. Moreover, this study confirms that the presence of other people smoking is associated with temptations and lapses (Bolman et al., 2018; Shiffman, Gnys, et al., 1996).
The second research question examined which contextual factors were associated with the risk of developing temptations into lapsing compared to being resolved without smoking. The results show that one theme characterized the differences between lapses and temptations, consisting of several social contextual factors: Lapses tended to occur more often when participants were social, when they were with friends, when they saw people smoking, and when they were outdoors. This theme is an accentuated version of the social theme that characterized the differences between temptations and random assessments and supports a model in which lapses are viewed as stronger versions of temptations (Shiffman, Paty, et al., 1996). In terms of intervention development, it emphasizes the need to inform people to avoid situations that provide opportunities to smoke in the early stages of the quit attempt.
The study was conducted during the COVID-19 pandemic, which was a period marked by changes in people's behaviors and location preferences. Throughout the study period, from October 5, 2021, to March 4, 2022, the Dutch government implemented various measures in response to the pandemic. Notably, a lockdown was enforced between December 18, 2021, and January 14, 2022 (Rijksoverheid, 2022). It is important to note that due to the unique circumstances of the pandemic, the results obtained in this study may differ from those of studies conducted during different time periods. This aspect is particularly pertinent when considering locations related to public life because of restrictions, such as limited access to bars. Under different circumstances, it is likely that individuals would have visited these places more frequently, making it crucial to acknowledge this influence on the study's findings.
An important observation was that more participants than originally assumed had to be recruited to reach the intended sample size. One probable explanation for this unexpected requirement is the inclusion of smokers who were motivated to quit within three months. In hindsight, it appears that implementing a stricter measure, such as targeting participants motivated to quit within one month, could have been more appropriate and decreased attrition. A limitation of this study is the high participation burden. First, the ubiquity and intensity of the EMA protocol may lead to reactivity. Second, individuals participating in such extensive research may not resemble the average smoker trying to quit, which may be indicated by the higher attrition among male and older participants. However, a strength of this study was that participants were not provided an intervention or treatment. In addition to the treatment itself affecting participants' responses, smokers seeking treatment are not representative of smokers trying to quit as a whole because treatment uptake is low (Borland et al., 2012). Another strength of this study was the EMA period of 14 days per participant, beginning with the day of the quit attempt. Studies show that this is an important period of relapse behavior. First, most relapse occurs within the first 8 days after quitting (Hughes et al., 2004), and second, not smoking in the first week after quitting is highly predictive of long-term abstinence (Ashare et al., 2013).
Conclusion
In summary, we found several associations between contextual factors and temptations and lapses. Most results corroborated prior research, but previously unrecognized factors (e.g., sexual intercourse) were also found. Digital just-in-time interventions may be well suited to address the identified contextual factors by providing support to participants in contexts associated with increased risk (Naughton, 2016). Further research in the field of smoking relapse could usefully explore the relationships between motivational and contextual factors by developing and testing a conceptual model. This model could incorporate motivational factors such as self-efficacy as moderators of relapse outcomes (e.g., a person high in self-efficacy to decline offers to smoke may be less influenced by the presence of other smokers). The model could also incorporate emotional factors such as stress or boredom (e.g., stress may trigger temptations, but whether temptations develop into lapses might vary based on the person’s location). Furthermore, the model could consider more general factors such as cigarette availability (e.g., the impact of activities such as sexual intercourse may depend on the sheer availability of cigarettes). To ensure focused and rigorous investigations, it is crucial to define (parsimonious) models a priori, facilitating hypothesis-driven research. Moreover, this study demonstrates the applicability of ecological momentary assessments to examine contextual factors, which may also provide new insights into related health behaviors, such as alcohol abuse or obesity.
References
(2013). The first 7 days of a quit attempt predicts relapse: Validation of a measure for screening medications for nicotine dependence. Journal of Addiction Medicine, 7(4), 249–254. 10.1097/ADM.0b013e31829363e1
(2014). lme4: Linear mixed-effects models using Eigen and S4. R package version 1.1.27. Retrieved April 28, 2022, from https://cran.r-project.org/web/packages/lme4/lme4.pdf
(2010). Nicotine addiction. New England Journal of Medicine, 362(24), 2295–2303. 10.1056/NEJMra0809890
(2018). Predicting smoking lapses in the first week of quitting: An ecological momentary assessment study. Journal of Addiction Medicine, 12(1), 65–71. 10.1097/ADM.0000000000000365
(2012). Cessation assistance reported by smokers in 15 countries participating in the International Tobacco Control (ITC) policy evaluation surveys. Addiction, 107(1), 197–205. 10.1111/j.1360-0443.2011.03636.x
(2016). How can relapse be prevented? Cancer Council Victoria. Retrieved April 6, 2022, from https://www.tobaccoinaustralia.org.au/chapter-7-cessation/7-8-how-can-relapse-be-prevented
(1991). The Fagerstrom test for nicotine dependence: A revision of the Fagerstrom Tolerance Questionnaire. British Journal of Addiction, 86(9), 1119–1127. 10.1111/j.1360-0443.1991.tb01879.x
(2004). Shape of the relapse curve and long-term abstinence among untreated smokers. Addiction, 99(1), 29–38. 10.1111/j.1360-0443.2004.00540.x
(1980). Coffee drinking and cigarette smoking: I. Coffee, caffeine and cigarette smoking behavior. Addictive Behaviors, 5(4), 389–394. 10.1016/0306-4603(80)90012-X
(2022). movisensXS, Version 1.5.23 (movisens GmbH, Karlsruhe, Germany). Retrieved April 25, 2022 from https://www.movisens.com/en/products/movisensXS/
(2016). Delivering “just-in-time” smoking cessation support via mobile phones: Current knowledge and future directions. Nicotine & Tobacco Research, 19(3), 379–383. 10.1093/ntr/ntw143
(2023). Within-person associations between psychological and contextual factors and lapse incidence in smokers attempting to quit: A systematic review and meta-analysis of ecological momentary assessment studies. Addiction, 118(7), 1216–1231. 10.1111/add.16173
(2021). Spatial, temporal, and demographic patterns in prevalence of smoking tobacco use and attributable disease burden in 204 countries and territories, 1990-2019: A systematic analysis from the Global Burden of disease Study 2019. The Lancet, 397(10292), 2337–2360. 10.1016/S0140-6736(21)01169-7
(2020). Blokhuis: Alle rookruimtes op het werk en in openbare gebouwen vanaf 2022 definitief gesloten. Retrieved July 22, 2022, from https://www.rijksoverheid.nl/actueel/nieuws/2020/10/06/blokhuis-alle-rookruimtes-op-het-werk-en-in-openbare-gebouwen-vanaf-2022-definitief-gesloten
(2022). Coronavirus tijdlijn. Retrieved May 29, 2023, from https://www.rijksoverheid.nl/onderwerpen/coronavirus-tijdlijn
(2019). Evaluating the temporal relationships between withdrawal symptoms and smoking relapse. Psychology of Addictive Behaviors, 33(2), 105–116. 10.1037/adb0000434
(1982). Relapse following smoking cessation: A situational analysis. Journal of Consulting and Clinical Psychology, 50(1), 71–86. 10.1037/0022-006X.50.1.71
(2005). Dynamic influences on smoking relapse process. Journal of Personality, 73(6), 1715–1748. 10.1111/j.0022-3506.2005.00364.x
(2006). Reflections on smoking relapse research. Drug and Alcohol Review, 25(1), 15–20. 10.1080/09595230500459479
(2014). Smoking patterns and stimulus control in intermittent and daily Smokers. PLoS ONE, 9(3),
Article e89911 . 10.1371/journal.pone.0089911(1996). Temptations to smoke after quitting: A comparison of lapsers and maintainers. Health Psychology, 15(6), 455–461. 10.1037/0278-6133.15.6.455
(2009). Immediate antecedents of cigarette smoking: An analysis from ecological momentary assessment. American Psychological Association. 10.1037/11855-015
(2006). Smoking patterns and dependence: Contrasting chippers and heavy smokers. Journal of Abnormal Psychology, 115(3), 509–523. 10.1037/0021-843X.115.3.509
(1996). First lapses to smoking: Within-subjects analysis of real-time reports. Journal of Consulting and Clinical Psychology, 64(2), 366–379. 10.1037/0022-006X.64.2.366
(2020). Ecological momentary assessment of temptations and lapses in non-daily smokers. Psychopharmacology, 237(8), 2353–2365. 10.1007/s00213-020-05539-3
(2008). Ecological momentary assessment. Annual Review of Clinical Psychology, 4(1), 1–32. 10.1146/annurev.clinpsy.3.022806.091415
(2018). Ecological momentary assessment of smoking behaviors in native and converted intermittent smokers. The American Journal on Addictions, 27(2), 131–138. 10.1111/ajad.12690
(1994). Caffeine and nicotine: A review of their joint use and possible interactive effects in tobacco withdrawal. Addictive Behaviors, 19(3), 229–256. 10.1016/0306-4603(94)90027-2
(2022). Cijfers roken. Retrieved May 16, 2022, from https://www.trimbos.nl/kennis/cijfers/roken/
(2014). The Health Consequences of Smoking—50 Years of Progress: A Report of the Surgeon General. https://www.cdc.gov/tobacco/sgr/50th-anniversary/index.htm
(2011). Predictors of attempts to stop smoking and their success in adult general population samples: A systematic review. Addiction, 106(12), 2110–2121. 10.1111/j.1360-0443.2011.03565.x
(2011). Conditioned cues for smoking elicit preparatory responses in healthy smokers. Psychopharmacology, 213(4), 781–789. 10.1007/s00213-010-2033-2
Appendix A: Constructs and Items of the Questionnaire Used in the EMA Protocol
Appendix B: R Script
