Started Smoking Again After 6 Months

• Journal List
• HHS Author Manuscripts
• PMC4517970

Addiction. Writer manuscript; available in PMC 2015 Jul 28.

Published in final edited form equally:

PMCID: PMC4517970

NIHMSID: NIHMS708771

Predictors of smoking relapse by duration of abstinence: findings from the International Tobacco Command (ITC) Four Country Survey

N Herd

^a Department of Psychology, The University of Melbourne, Australia

R Borland

^b VicHealth Middle for Tobacco Command, The Cancer Quango Victoria, Australia

A Hyland

^c Department of Health Behavior, Roswell Park Cancer Plant, Buffalo, New York, The states

Abstract

Aim

To explore predictors of smoking relapse and how predictors vary co-ordinate to duration of forbearance.

Design, setting and participants

A longitudinal survey of 1296 ex-smokers recruited every bit part of the International Tobacco Control (ITC) Four Country Survey (Australia, Canada, UK, and USA).

Measurements

Quitters were phone interviewed at varying durations of abstinence (from one twenty-four hours to approximately iii years) so followed-up approximately i year later. Theorised predictors of relapse (i.east., urges to fume; effect expectancies of smoking and quitting; and abstinence self-efficacy) and nicotine dependence were measured in the survey.

Findings

Relapse was associated with lower forbearance self-efficacy and a college frequency of urges to smoke, but only after the first calendar month or and so of quitting. Both of these measures mediated relationships between perceived benefits of smoking and relapse. Perceived costs of smoking and benefits of quitting were unrelated to relapse.

Conclusions

Challenging perceived benefits of smoking may exist an effective way to increase abstinence self-efficacy and reduce frequency of urges to smoke (particularly after the initial weeks of quitting), in club to subsequently reduce relapse run a risk.

Keywords: smoking, tobacco, abeyance, relapse

Introduction

Nigh smokers would like to quit (1) and the majority have tried to exercise and so (2, 3), only virtually of those who try end upwards relapsing. (4-7). Relapse occurs most often during the initial days of quitting (6); still, longitudinal studies have shown that a substantial proportion of quitters who remain abstinent early in the quit attempt, actually keep to relapse later being quit for months or even years (5, eight-10). Despite the prevalence of tardily relapse in that location is limited agreement of its precipitates and whether information technology differs from early on relapse.

In a companion paper (thirteen), we showed that smoking related behavior and experiences change systematically after quitting. Measures of most variables studied (due east.chiliad., frequency of urges to fume, abstinence self-efficacy) inverse relatively speedily during the first few weeks of quitting before beginning to asymptote; however, the rate of asymptoting varied from a rapid, logarithmic, function to a slower, foursquare-root function of fourth dimension. Differing rates of change means that the influences of these beliefs are likely to change and it is possible that this is related to the power of these factors to influence relapse at different points in fourth dimension.

Real time data and retrospective accounts of relapse have plant that cravings and urges to smoke are often cited equally precipitates of early on relapse (12, xiv, 15). Similarly, perceived benefits of smoking predict relapse early in quit attempts (16, 17); even so, it is unclear what role they play in long-term relapse. Depression self-efficacy has proven to exist a reliable predictor of early relapse (5, 11, 18-20). One written report found an interactive relationship between self-efficacy and time (11): high self-efficacy predicted success among those quit for less than a calendar week or those experiencing at least daily strong urges to fume, merely predicted relapse among participants who had been quit for more than a week and who reported less than daily strong urges to smoke. The authors suggested that overconfidence might play a pregnant role in late relapse. Other inquiry has plant that higher levels of behavioural change process predicted relapse, but only during the showtime month of quitting (v). Enquiry into postal service-quitting weight gain has shown that during the initial weeks of quitting gains predict abstinence, whereas gains afterwards in the quitting process predict relapse (12).

Using longitudinal data from the International Tobacco Control (ITC) Four Country Survey, our aim was to explore the rate at which quitters relapsed over fourth dimension, and how smoking related beliefs (i.east., forbearance self-efficacy, benefits of smoking/costs of quitting, benefits of quitting/costs of smoking) and experiences (i.east., frequency of urges to smoke) precipitate relapse over time.

Given that these factors are unlikely to trigger relapse independently of one another, we likewise explored possible interactive processes by which predictors precipitate relapse. In line with Bandura's social cognitive theory (21) and previous research findings (17) we hypothesised:

• 1)

  that perceived benefits of smoking would but threaten sustained abstinence when self-efficacy was low; and

• 2)

  that self-efficacy would only protect confronting relapse when the task of quitting was deemed to exist more difficult (i.e., in this study, having high perceived benefits of smoking).

Next we explored the mediating mechanisms though which predictors precipitated relapse. In line with Marlatt and Gordon's model of relapse prevention (22), Dijkstra and Borland (17) establish that the relationship between perceived benefits of smoking and relapse was mediated by cravings, but not self-efficacy. Given that greater perceived benefits of smoking are probable to make the job of quitting seem more difficult, it is surprising that the authors did not notice a causal pathway in which greater perceived benefits of smoking as well exert their influence on relapse past decreasing self-efficacy. We sought to examination these mediating models of relapse individually and as part of a multiple mediation model.

In the current paper nosotros besides explore two models of relapse: a fixed threshold model and a relative threshold model. If relapse occurs when smoking related behavior and experiences are above given fixed thresholds so the probability of relapse should refuse over time as these measures fall further below the threshold. As the predictors asymptote, all other things existence equal, rates of relapse should stabilise at very low levels. By contrast, the relative threshold model of relapse predicts that the probability of relapse will continue to vary over fourth dimension because the thresholds at which smoking related conventionalities and experiences precipitate relapse are relative and, therefore, also vary over time (i.e., the threshold would get progressively lower, and thus rates of relapse would be college than predicted past a fixed threshold model).

Method

The ITC-four Survey is an annual longitudinal survey of smokers that was established to evaluate the psychosocial and behavioural impact of tobacco control policies (23). Participants who choose to quit smoking are retained in the cohort, thus providing the opportunity to report predictors of relapse over time.

Participants

Participants in the current study were 1296 adults from the first five waves of the ITC-iv Survey, recruited as smokers who were then subsequently quit on at least one wave and who reported smoking status at the subsequent wave. Participants' beliefs and reported experiences taken from the first iv waves of data were used to predict relapse in waves ii through five (2003-2006). All predictors, except sociodemographics, time to starting time cigarette, and cigarette consumption, were assessed while participants were quit. Fifty-7 percent of participants were female and the hateful age was 43.64 years (SD=14.16). Overall, 30% were from Australia, 27% from the Great britain, 26% from Canada, and 18% from the USA. Before quitting, 39% smoked i-10 cigarettes per day, 42% 11-twenty, and fifteen% more than than 20. Mean heaviness of smoking alphabetize (HSI), the combination of cigarettes smoked per day and time to get-go cigarette (range 0-vi), was 2.18 (SD=ane.58). At baseline (while nonetheless smoking), most participants had smokers among their five closest friends (24% none; 19% one; 22% two; 16% 3; 20% iv or five).

Participants who were quit at more than than one moving ridge (and reported more than 1 wave of follow-upward information) contributed multiple response sets: 896 contributed one fix; 296 two sets; and 104 three sets (no participants completed all four sets of information). At that place were 1800 sets of responses beyond the five waves from the 1296 respondents. Number of days quit across all five waves ranged from one to 1121, with a median of 151 and an interquartile range of 341.five (146 were surveyed during the first week mail service quitting, 239 were surveyed 1 week – 1 calendar month, 655 were surveyed one-vi months, 291 were surveyed six months - one year, 326 were surveyed one-two years, and 143 were surveyed >ii years). Nicotine replacement therapy was used by viii% of participants at the fourth dimension of being interviewed.

Measures

The demographic and baseline (while smoking) measures used are described above. Smoking status outcome at each moving ridge was adamant by asking participants if they were still quit, and if and then, whether they had stayed quit since the terminal survey engagement. Participants who were back smoking and those who were currently quit merely reported that they had not stayed quit since the final survey were considered to have relapsed. Participants who reported smoking at to the lowest degree one time a month were considered to be still smokers.

Proposed predictors of relapse (presented in Error! Reference source non establish.) were fatigued from established psychosocial models of wellness behaviour (refer to Fong et al (23)) and have been used in past research exploring predictors of quitting (24). Responses to each item were recorded on four- or v-point Likert scales (for more details of each measure see Herd et al. (thirteen)).

Statistical analysis

Results in the respective trends newspaper (13) establish that the majority of proposed predictors of relapse inverse according to a logarithmic or square-root function. Therefore, for duration of abstinence in the current study, we used the transformation that best fitted each of the dependent variables in the previous paper; a log transformation of days quit was used for proposed predictors that changed over time co-ordinate to a log function and a square root transformation was used for those that changed according to a square root office. All analyses were conducted using STATA 10 (and significant findings are indicated past p-values of 0.05, 0.01, and 0.001).

Logistic regression assay was used to explore relationships betwixt demographic variables (i.e., sex activity, age, state, HSI, and nicotine replacement therapy) and relapse at the following wave approximately one year later, after adjusting for log transformed elapsing of forbearance. Wald tests were used to determine the overall significance of each chiselled demographic variable. Hierarchical logistic regression analysis was used to examine the relationships betwixt proposed predictors of relapse and relapse at the subsequent wave, after adjusting for demographic variables and appropriately transformed duration of abstinence. Interactions between proposed predictors and duration of abstinence were entered in the final stride to determine if the relationships between these variables and relapse varied according to duration of abstinence. Due to missing data for HSI at recruitment and nicotine replacement therapy, nosotros conducted analyses with and without these variables. Results including these variables are only reported if they were substantially different from the pattern of results obtained earlier they were added. The postgr3 control for STATA (25, 26), was used to graph the adjusted predicted probability of relapse co-ordinate to the logistic regression interpretation models. The postgr3 command holds covariates entered in the logistic regression models constant at their mean.

Some responses in the current analysis were repeated measures and, therefore, cannot be considered contained from ane another. Therefore, generalised estimating equation models (GEE) were also fitted to the data (27). An exchangeable within-subject correlation construction was used, equally this immune for unequal spacing in duration of abstinence between observations. An unstructured correlation structure was initially tried, only did non e'er allow the information to converge. All of the results from GEE modelling supported the results obtained from logistic regression assay and, therefore, are non reported in the results.

Hypothesised moderating effects were explored past adding an interaction term between contained variables to the model after both variables had already been entered. Mediation analysis with a dichotomous dependent variable was carried out according to the methods described by MacKinnon and Dwyer (28) and multiple mediators were tested simultaneously co-ordinate to the methods described by Kenny at al. (29). Once again, all analyses adjusted for demographic measures and duration of abstinence.

Results

Overall, 37% (n=668) of participants quit at ane wave relapsed before the subsequent wave, with the remaining 63% (north=1132) remaining abstinent. Not surprisingly, relapse by the subsequent wave was more than prevalent early in the quit effort (Table 1). Sexual activity, age, country, and use of nicotine replacement therapy did not predict relapse later on adjusting for duration of abstinence. The HSI also did not predict relapse. However, amidst the minor sample surveyed during the first month of quitting (n=289), during which nicotine dependence would be expected to exist most probable to influence quitting success, relapse was marginally (albeit not-significant) college among heavier smokers (11-20 cigarettes, OR=ane.02, 95% CI=0.59-1.78; 21-30 cigarettes, OR=1.58, 95% CI=0.65-3.83; 31+ cigarettes, OR=i.41, 95% CI=0.27-7.46).

Table i

Smoking status at the subsequent wave by elapsing of abstinence at the preceding wave.

Smoking status at follow-up moving ridge	Duration of forbearance at preceding wave
	1-seven days	viii-30	31-182	183-365	366-730	>730
Relapsed	114 (78%)	154 (64%)	274 (42%)	65 (22%)	54 (17%)	7 (v%)
Continued abstinence	32 (22%)	85(36%)	381 (58%)	226 (78%)	272 (83%)	136 (95%)

Results in Table 2 show that for every one signal increase in log days quit, the odds of relapse decreased by a factor of 0.17.

Tabular array ii

Results of logistic regression modelling identifying significant predictors of relapse by the subsequent wave: demographics and elapsing of abstinence (log transformed) (n=1678).

	Variables	Wald test	Odds ratio	95% CI	p
Days quit (log transformed)			0.17	0.thirteen-0.21	<0.001
NRT	Not using	2.56	1
	Using		1.44	0.92-two.26	0.11
HSI			one.00	0.92-1.07	0.xc
Sex	Female	0.52	1
	Male person		0.92	0.73-1.16	0.47
Age	18-24 years	6.71	i
	25-39 years		0.73	0.48-ane.11	0.fourteen
	xl-54 years		0.69	0.45-1.05	0.08
	55+ years		0.56	0.36-0.88	<0.05
State	Australia	two.62	1
	Canada		0.87	0.65-1.xviii	0.38
	United Kingdom		0.89	0.65-one.21	0.45
	United States		0.75	0.53-ane.07	0.eleven

After controlling for demographics and elapsing of abstinence, the number of smokers among participants' five closest friends significantly predicted relapse; for each friend who smoked, the odds of relapse increased by 1.12 (95% CI=1.04-1.xx, p<0.01). In that location was a significant interaction between number of friends who smoked and duration of abstinence (OR=1.xvi, 95% CI=1.04-1.29, p<0.01) suggesting that a higher proportion of friends who smoke was only associated with an increased risk of relapse afterwards approximately a calendar month post quitting. Error! Reference source not constitute. shows the probability of relapse at each measured fourth dimension point as a role of reported number of smoker friends at baseline.

Post-quitting belief and experiences equally predictors of relapse

Tabular array 3 presents results from logistic regression analysis identifying predictors of relapse. A college frequency of urges to smoke measured at waves two to four was significantly related to an increased likelihood of relapse. A significant interaction between urges and elapsing of abstinence indicated that urges predicted relapse differently according to elapsing of abstinence (encounter Figure 2). This, and subsequent figures, show the probability of relapse (12 months later) for the measure taken at the fourth dimension indicated. Given that the interaction appeared to cantankerous over at approximately one month postal service quitting, we conducted separate logistic regression analyses for 1 month or less mail service quitting and more than one month post quitting. Results showed that urges during the first month of quitting were unrelated to relapse (OR=1.09, 95% CI=0.83-one.43, p>0.05); yet, frequent urges reported after a month or more were associated with a greater likelihood of relapse at follow-upward (OR=1.42, 95% CI=1.25-ane.60, p<0.001). We tested to meet if this may have been due to the use of nicotine replacement therapy early in the try, only found no effect.

The interaction betwixt duration of abstinence and frequency of urges to fume as a predictor of relapse.

Table 3

Results of logistic regression modelling identifying predictors of relapse.

			Chief effect model		Interaction model
	Contained Variables	North	OR	95% CI	OR	95% CI
Urges to smoke	Frequency of stiff urges to smoke	1727	0.65*	0.45-0.93	1.45***	1.22-one.72
Perceived benefits of smoking	Perceived weight control benefits of smoking	1785	1.06	0.97-1.16
	Enjoy smoking too much to give information technology up for good	1764	one.26***	1.11-i.42
	Smoking is an important part of your life†	1792	one.14*	1.01-ane.28
	Smoking calms you down when y'all are stressed	1782	1.10*	i.004-1.212
	Thoughts about the enjoyment of smoking	1796	0.72*	0.54-0.96	ane.29***	1.12-1.48
Perceived costs of smoking	Thoughts virtually the harms of smoking to you and others†	1788	1.06	0.96-i.16
	Thoughts about the coin spent on smoking†	1797	i.00	0.93-1.09
Perceived benefits of quitting	Perceived health and other benefits of not smoking	1723	0.98	0.88-1.10
	Perceived hazard of centre affliction in time to come vs. not-smoker	1502	0.93	0.84-1.04
	Perceived quality of life since quitting	1679	1.08	0.95-1.23
Forbearance self-efficacy	How sure are you that y'all tin can stay quit?	1786	0.62***	0.56-0.70

Iv of the five perceived benefits of smoking we measured predicted relapse, with only perceived weight command benefits of smoking beingness unrelated. Figure 3 shows that higher understanding with three perceived benefits was related to increased relapse contained of time. Higher frequency of thoughts nearly the enjoyment of smoking was besides significantly related to an increased likelihood of relapse, only the outcome varied by fourth dimension quit (Figure four). There was no outcome during the commencement calendar month of quitting (OR=0.96, 95% CI=0.79-i.xvi, p>0.05); however, after a calendar month at that place was an increased likelihood of relapse with college frequency of enjoyment thoughts (OR=1.23, 95% CI=ane.12-ane.36, p<0.001). For those quit for less than one month in that location were limited numbers of cases (as low as 256), so the results need to be treated with caution. The only perceived benefit of smoking to be independently predictive for this sub-sample was the belief that smoking is too enjoyable to give up for good.

Elapsing of abstinence and perceived benefits of smoking as predictors of relapse.

The interaction between duration of abstinence and frequency of thoughts almost the enjoyment of smoking as a predictor of relapse.

We found that perceived costs of smoking and perceived benefits of quitting did not predict relapse (Error! Reference source non found.). There were also no significant interactions betwixt each of these measures and duration of abstinence. Figure v shows that higher self-efficacy was associated with a lower probability of relapse.

Elapsing of forbearance and abstinence cocky-efficacy as predictors of relapse.

Moderation of relapse

Moderating models of relapse were used to determine if perceived benefits of smoking merely threatened sustained forbearance when self-efficacy was low. After adjusting for master effects, demographics, and duration of abstinence, the interactions between cocky-efficacy and each perceived do good of smoking detail was non meaning, indicating no moderation effect. We besides tested to see if self-efficacy was only important when urges were high by looking at its effects amidst those quit for more i month and reporting less than daily strong urges. Withal, self-efficacy was still a strong predictor (OR=0.69, 95% CI=0.59-0.81).

Arbitration of relapse

Mediating models of relapse explored whether frequency of urges to fume and self-efficacy mediated the relationships between perceived benefits of smoking and relapse. Duration of abstinence, sexual activity, historic period, and country were entered every bit covariates at each step. Given that frequency of urges to smoke just predicted relapse afterwards the beginning calendar month of quitting, nosotros just explored whether information technology acted as a mediator after this point. Also, the mediation of frequency of thoughts about the enjoyment of smoking was only explored later on one month postal service quitting.

The relationships betwixt relapse and the beliefs that smoking calms you downwardly, that it is an important role of life, and that information technology is too enjoyable to surrender for good, were all mediated by self-efficacy (Mistake! Reference source non found.). Frequency of thoughts about the enjoyment of smoking was simply partially mediated by self-efficacy. For those quit for more one month, the relationships between relapse and frequency of thoughts about the enjoyment of smoking, and the belief that smoking calms yous down, were both mediated by frequency of urges to smoke. The conventionalities that smoking is too enjoyable to requite upwardly for skillful was just partially mediated by urges. The relationship betwixt relapse and the belief that smoking is an of import part of life was non pregnant when data from the first month of quitting was excluded.

Given the overlapping mediation, we side by side explored whether the above effects were simultaneously mediated by urges and self-efficacy. Not surprisingly, frequency of urges and self-efficacy were negatively correlated (r=−0.32, p<0.001). Results confirmed that when both proposed mediators were added to the models predicting relapse, each perceived benefit of smoking no longer predicted relapse. Sobel tests plant that the indirect effects of perceived benefits of smoking on relapse were carried past both urges and cocky-efficacy (see Figure 6). This figure does not show the relationship for the belief that smoking is an of import function of life as information technology was not a significant predictor of relapse post 1 month quit.

The indirect effect of perceived benefits of smoking on relapse through frequency of urges to smoke and abstinence self-efficacy.

Discussion

This study is 1 of the few to assess relapse in the general population, rather than equally a function of a clinical abeyance intervention. Our results showed that the charge per unit of smoking relapse decreased over time consistent with previous studies (5, 6), dropping to around 5% later on more than than two years. The two main dynamic predictors of relapse announced to be self-efficacy, which protects against relapse, and frequency of urges to smoke, which promotes relapse, but simply after existence quit for around i month. Both these variables appear to mediate the predictive relationships between relapse and perceived benefits of smoking/barriers to quitting. The other important predictor we identified was number of friends who smoke, only once more merely from a month or and so after quitting. Curiously, cigarette consumption prior to quitting, an indicator of dependence, was not a predictor of relapse. Nosotros as well found no effect for perceived costs of smoking/benefits of quitting and 1 potential barrier to quitting, perceived weight control benefits of smoking. In the following paragraphs we attempt to integrate these findings with each other and with previous enquiry.

We were surprised to notice no result for dependence given that this is known to exist a strong predictor of smoking cessation (24), although the literature regarding relapse is scarce (12). This may suggest that high levels of addiction (at least as measured by the behavioural indices of the HSI) generally only predicts very early relapse, something we were underpowered to study. If these findings were to be replicated, and then information technology would be skillful news for dependent smokers who might anticipate greater difficulty in staying quit for sustained periods, suggesting that if they survive the early days they are equally likely to succeed as would anyone else. That said, frequency of potent urges to smoke predicted relapse afterward one month, and this is conspicuously related to dependence (30). It may be that our behavioural measures of dependence are missing a vital element of dependence that becomes important post-cessation.

Number of friends who smoked was predictive of relapse, but simply after a month or and so. Early those with more smoking friends seemed to do somewhat amend. We suspect this is partly a function of those with many smoking friends taking this into account early on; nevertheless, the bear upon of friends being continual, ways the ease of staying quit does not improve as it might for those who alive in a more non-smoking environment. Whatsoever such upshot could be magnified if quitters tended to avoid socialising with smoking friends in the early days of their endeavor, just, understandably, did non sustain this over time.

Lower self-efficacy was a significant predictor of relapse independent of duration of forbearance and frequency of urges, consistent with much previous research (5, 19, twenty). The only study to detect time-dependent effects (11) looked at predictors curt term (effectually 3 weeks), while our report considers them over a longer time period, suggesting that in the long term, at to the lowest degree, loftier self-efficacy for the maintenance of abstinence is critical.

Frequency of strong urges to smoke too predicted relapse, but merely after the beginning month or so of quitting. Retrospective accounts of relapse suggest that urges to fume precipitate relapse (15), as also, does real fourth dimension information of initial smoking lapses amongst recent quitters (14). Our findings are consequent with this. However, they do propose that the frequency of such urges may non be a problem early on in a quit try. Possibly early in the quit attempt, quitters are prepared to bargain with urges and this helps proceed them focussed on the task; even so, if urges persist and then they may experience cocky-regulatory fatigue and become more susceptible to relapse (31). Although urges may not precipitate early on relapse, nosotros exercise non question the utility of learning how to cope with such urges.

Perceived benefits of smoking/barriers to quitting appear to only exist associated with relapse to the extent that they pb to more urges to fume and/or reduce cocky-efficacy for staying quit. The barriers we identified as being important were frequency of thoughts nearly the enjoyment of smoking, and agreement with the beliefs that smoking calms you down when stressed, that it is an important office of life, and that it is too enjoyable to give up for good. 2 of these may exist particularly of import early on in quit attempts; smoking being an important office of life (see Effigy 3B), as it was non significant when analyses were restricted to those quit for more than 1 month, and that smoking is as well enjoyable to give upwards for skilful, the only contained predictor during the first month. These beliefs might be expected to persist, and then the fact that they reject and may lose influence is reassuring, every bit it suggests that they may lose potency with fourth dimension, also as becoming less prevalent (thirteen). The belief that smoking calms you down when stressed or upset was one of the few predictors of relapse that was still persistent among many participants even years later on quitting (13). Given that this belief was significantly associated with relapse, it may contribute to a substantial proportion of late relapse, particularly when stressful or upsetting experiences occur.

The conventionalities that smoking helps control weight was the only perceived do good of smoking that was unrelated to relapse, and interestingly, also the merely perceived benefit for which agreement increased over time (13). Given that agreement with this belief was probable to have increased in response to actual weight gain, the results suggest that weight gain alone was unlikely to have been a precipitate of relapse. By research has found that baseline concerns about potential weight proceeds are unrelated to subsequent relapse (32-35). Still, given that quitters increasingly gain weight the longer they are quit it is likely that concern with weight proceeds besides increases over time. Time to come inquiry would benefit from exploring the relationship between relapse, actual weight proceeds, and concerns about weight that has been gained.

Frequency of thoughts about the enjoyment of smoking was the one perceived benefit of smoking that became a stronger predictor with time. This highlights the importance of helping the quitter develop strategies for extinguishing craving-evoking cues in as many contexts as possible. Nostalgic beliefs about the value of smoking may be one specially important set of cues for sustaining cravings and threatening self-efficacy. Our findings suggest a more dynamic model of the interrelationships between these factors and self-efficacy than that found past Dijkstra and Borland (17), in that it is not high self-efficacy itself that is disquisitional in preventing barriers from precipitating relapse, but rather the capacity to maintain high self-efficacy in the context of strongly felt barriers to quitting that is disquisitional.

Although perceived costs of smoking and benefits of quitting often precipitate a quit endeavour (3, 24) and are oftentimes used in the media to encourage quitting, results in the current study found that agreeing with, or thinking more near, these issues did non increase the likelihood of successful abstinence after quitting. Similar enquiry past Hyland and colleagues (24) besides establish that perceived costs of smoking (measured before quitting) were unrelated to relapse; however, larger expected benefits of quitting at baseline did predict relapse (unexpectedly). The wellness benefits of quitting are typically hard to observe and were probably still largely in the future for well-nigh participants in this report, thus they may be largely insignificant in helping maintain abstinence. An alternative explanation is that some levels of such beliefs are universal and the variability in our measures was non meaningful. Nosotros practise not question the utility of quitters understanding the wellness benefits of staying quit, indeed why would they carp if they thought there was no benefit, particularly those that run across benefits from smoking? Instead, it might be that knowledge of the harms helps to maintain forbearance only if it is specifically accessed during periods of relapse vulnerability (something that was not measured in this report).

The finding that the probability of relapse reduces over time for all variables studied is notable. I t supports a relative threshold model of relapse, in which the threshold at which determinants precipitate relapse varies over time. We might have expected that persistence or strong pro-smoking attitudes, frequent urges, and low self-efficacy might have been even more predictive of relapse over time. It is possible the effect is because the ratings are made relative to their electric current state of affairs, rather than to an absolute, only even if this is then, it is reassuring. These findings complement those plant in our companion paper (13) that showed changes in levels of behavior over fourth dimension. These changes should add to the reduced predictive value for relapse to further reduce overall relapse rates. Information technology suggests that failure to successfully arbitrate to reduce the threat from these factors might not exist a complete recipe for relapse. However, it remains important to challenge these beliefs and experiences, because they remain predictors of relapse, which as nosotros take seen, occurs at unacceptably high rates.

Caution needs to exist exercised still in generalising too strongly from our results. To further explore the mode in which these potential determinants of relapse interact, enquiry is required in which these variables are measured more frequently and closer to the smoking status outcome. Our mediation analysis was besides limited by the predictor variables and mediators being measured at the aforementioned time. It would be especially instructive to experimentally induce changes in perceived benefits of smoking and appraise their impact on urges and self-efficacy, and then on relapse in lodge to confirm this mediational pathway.

The current study was express by the varying intervals between our measures and when relapse occurred. It could have been merely days subsequently the survey, in which case the predictors were measured proximally, or it could accept been up to a year. Given the potential gap betwixt the survey and consequence measures, it is notable that we still found strong predictors of relapse. We acknowledge that we lacked sensitivity to observe the furnishings of variables that modify considerably day to day. However, the variables that are most likely to modify in this way, urges and frequency of thoughts, were identified equally predictors, and then nosotros think it unlikely that nosotros have missed other major predictors for this reason. However, we acknowledge that the strength of the association between the predictors we found and relapse is likely to be stronger than we estimate here. We as well acknowledge that predictors of relapse may vary according to factors not measured hither, and might vary for some population sub-groups (e.g., those with psychiatric or melancholia disorders, those from unlike subcultures), but it is equally possible that the determinants of relapse are relatively abiding and all that would vary is the frequency of predictors of relapse and maybe the rates at which they alter with time.

Overall, the results confirm a considerable level of relapse even amid those who take been abstinent for a year or longer. The model of relapse that emerges from this is that perceived benefits of smoking play a key office in effecting the frequency of urges to smoke and lowering cocky- efficacy, which then subsequently co-determine relapse. Rather than reminding ex-smokers about the costs of smoking or benefits of quitting to encourage sustained abstinence, it may be more beneficial to provide persuasive information or experiences that claiming perceived benefits of smoking, to the extent that this is possible. However, there is only express show that such a strategy works to reduce alcohol consumption (36), so circumspection is required. Our findings also suggest that at that place may be a need to adopt somewhat different strategies for preventing relapse early in the quit try to subsequently on. Early on on, coping with challenges would appear to be important, while by a month or and so, it is important to have fewer smoking urges and bolstered self-efficacy.

​

The interaction betwixt duration of abstinence and number of smokers among five closest friends as a predictor of relapse.

Table 4

The mediation of perceived benefits of smoking on relapse by frequency of urges to smoke and abstinence self-efficacy.

Predictors	Outcomes (Standardized regression coefficients and SE)			Sobel examination (z)
	ASE	Relapse	Mediated relapse
Thoughts about the enjoyment of smoking†	−0.x (0.01)***	0.14 (0.03)***	0.10 (0.03)**	4.72***
ASE			−0.21 (0.03)***
Smoking calms me downward when stressed	−0.09 (0.01)***	0.07 (0.03)*	0.02 (0.03)	five.40***
ASE			−0.26 (0.03)***
Smoking is an important function of life	−0.10 (0.01)***	0.06 (0.03)*	0.02 (0.03)	five.46***
ASE			−0.26 (0.03)***
I enjoy smoking also much to quit for good	−0.15 (0.01)***	0.11 (0.03)***	0.04 (0.03)	6.48***
ASE			−0.26 (0.03)***

	Urges	Relapse	Mediated relapse
Thoughts about the enjoyment of smoking†	0.23 (0.01)***	0.xiv (0.04)***	0.06 (0.04)	4.06***
Urges			0.16 (0.04)***
Smoking calms me down when stressed†	0.x (0.01)***	0.08 (0.04)*	0.05 (0.04)	4.25***
Urges			0.18 (0.04)***
I enjoy smoking also much to quit for good†	0.07 (0.01)***	0.xi (0.03)**	0.08 (0.03)*	3.45***
Urges			0.17 (0.03)***

Acknowledgements

The first writer was supported by an Australian Postgraduate Award. This enquiry was funded by grants from the National Cancer Institute of the United States (R01 CA 100362), the Roswell Park Transdisciplinary Tobacco Utilize Research Center (P50 CA111236), Robert Forest Johnson Foundation (045734), Canadian Institutes of Health Inquiry (57897 and 79551), National Health and Medical Enquiry Council of Australia (265903 and 450110), Cancer Research United kingdom (C312/A3726), and Canadian Tobacco Command Research Initiative (014578), with additional support from the Middle for Behavioural Inquiry and Program Evaluation, National Cancer Institute of Canada/ Canadian Cancer Society.

Footnotes

Address where work was carried out: Department of Psychology, School of Behavioural Scientific discipline, 12th Floor, Redmond Barry Building, The University of Melbourne, Victoria 3010 Australia

Conflict of interest declaration: None

References

ane. Mullins R, Borland R. Practice smokers desire to quit? Australian and New Zealand Journal of Public Wellness. 1996;20:426–427. [PubMed] [Google Scholar]

2. Borland R, Yong H, King B, Cummings KM, Fong GT, Marshall TE, et al. Use of and beliefs well-nigh light cigarettes in four countries: Findings from the International Tobacco Control Evaluation Survey. Nicotine and Tobacco Research. 2004;6:S311–S321. [PubMed] [Google Scholar]

3. Hyland A, Li Q, Bauer J, Giovino G, Steger C, Cummings KM. Predictors of cessation in a cohort of current and former smokers over 13 years. Nicotine and Tobacco Research. 2004;6:S363–S369. [PubMed] [Google Scholar]

4. Piasecki TM, Fiore MC, McCarthy DE, Baker TB. Have we lost our fashion? The need for dynamic formulations of smoking relapse proneness. Habit. 2002;97:1093–1108. [PubMed] [Google Scholar]

five. Segan CJ, Borland R, Greenwood KM. Tin can transtheoretical model measures predict relapse from the activeness phase of modify among ex-smokers who quit subsequently calling a quitline? Addictive Behaviors. 2006;31:414–428. [PubMed] [Google Scholar]

6. Hughes JR, Keely J, Naud Due south. Shape of the relapse bend and long-term abstinence amongst untreated smokers. Habit. 2004;99:29–38. [PubMed] [Google Scholar]

7. Marlatt GA, Curry South, Gordon JR. A longitudinal analysis of unaided smoking cessation. Journal of Consulting and Clinical Psychology. 1988;56:715–720. [PubMed] [Google Scholar]

8. Wetter DW, Cofta-Gunn L, Fouladi RT, Cinciripini PM, Sui D, Gritz ER. Belatedly relapse/sustained forbearance amid former smokers: a longitudinal study. Preventive Medicine. 2004;39:1156–1163. [PubMed] [Google Scholar]

9. Brandon Thursday, Lazev AB, Juliano LM. Very delayed smoking relapse warrants research attention. Psychological Reports. 1998;83:72–74. [PubMed] [Google Scholar]

x. Blondal T, Gudmundsson LJ, Olafsdottir I, Gustavsson M, Westin A. Nicotine nasal spray with nicotine patch for smoking cessation: randomised trial with six year follow up. British Medical Journal. 1999;318:285–289. [PMC free article] [PubMed] [Google Scholar]

eleven. Borland R, Balmford J. Perspectives on relapse prevention: An exploratory study. Psychology and Health. 2005;20:661–671. [Google Scholar]

12. Norregaard J, Tonnesen P, Petersen L. Predictors and reasons for relapse in smoking cessation with nicotine and placebo patches. Preventive Medicine. 1993;22:261–271. [PubMed] [Google Scholar]

xiii. Herd N, Borland R. The natural history of quitting smoking: findings from the International Tobacco Control (ITC) Four State Survey. Addiction. Submitted. [PMC free article] [PubMed] [Google Scholar]

14. Shiffman Southward, Paty JA, Gnys M, Kassel JA, Hickcox 1000. Start lapses to smoking: within-subjects analysis of real-fourth dimension reports. Journal of Consulting and Clinical Psychology. 1996;64:366–379. [PubMed] [Google Scholar]

15. Cummings KM, Jaen CR, Giovino G. Circumstances Surrounding Relapse in a Group of Recent Ex-smokers. Preventive Medicine. 1985;fourteen:195–202. [PubMed] [Google Scholar]

sixteen. Gwaltney CJ, Shiffman Southward, Balabanis MH, Paty JA. Dynamic self-efficacy and outcome expectancies: prediction of smoking lapse and relapse. Journal of Aberrant Psychology. 2005;114:661–675. [PubMed] [Google Scholar]

17. Dijkstra A, Borland R. Residuum consequence expectations and relapse in ex-smokers. Health Psychology. 2003;22:340–346. [PubMed] [Google Scholar]

eighteen. Condiotte MM, Lichtenstein E. Self-efficacy and relapse in smoking abeyance programs. Journal of Consulting and Clinical Psychology. 1981;49:648–658. [PubMed] [Google Scholar]

19. Stuart K, Borland R, Murray Due north. Self-efficacy, health locus of control, and smoking cessation. Addictive Behaviours. 1994;19:ane–12. [PubMed] [Google Scholar]

xx. Shiffman S, Balabanis MH, Paty JA, Engberg J, Gwaltney CJ, Liu KS, et al. Dynamic effects of self-efficacy on smoking lapse and relapse. Health Psychology. 2000;xix:315–323. [PubMed] [Google Scholar]

21. Bandura A. Social foundations of thought and action: a social-cognitive theory. Prentice-Hill; Englewood Cliffs, New Jersey: 1986. [Google Scholar]

22. Marlatt GA, Gordon JR, editors. Relapse Prevention: Maintenance Strategies in the Treatment of Addictive Behaviours. Guilford Press; New York: 1985. [Google Scholar]

23. Fong GT, Cummings KM, Borland R, Hastings G, Hyland A, Giovino GA, et al. The conceptual framework of the International Tobacco Control (ITC) Policy Evaluation Project. Tobacco Control. 2006;15:iii3–iii11. [PMC free article] [PubMed] [Google Scholar]

24. Hyland A, Borland R, Li Q, Yong H-H, McNeill A, Fong GT, et al. Individual-level predictors of cessation behaviours among participants in the International Tobacco Control (ITC) Four State Survey. Tobacco Control. 2006;15(Suppl III):iii83–iii94. [PMC free article] [PubMed] [Google Scholar]

25. Mitchell MN, Chen X. Visualizing main effects and interaction for binary logit models. The Stata Journal. 2005;five:64–82. [Google Scholar]

26. UCLA Academic Applied science Services Stata Programs for Data Assay. How can I use postgr3 for making graphs afterward estimation commands. 2007 [Google Scholar]

27. Liang Thou, Zeger SL. Longitudinal data analysis using generalized linear models. Biometrika. 1986;73:13–22. [Google Scholar]

28. MacKinnon DP, Dwyer JH. Estimating mediated furnishings in prevention studies Evaluation review. 1993;17:144–158. [Google Scholar]

29. Kenny DA, Kashy D, Bolger N. Data analysis in social psychology. In: Gilbert D, Fiske S, Lindzey Grand, editors. Handbook of social psychology. McGraw-Hill; New York: pp. 233–265. [Google Scholar]

30. Etter J, Houezec JL, Perneger TV. A self-administered questionnaire to measure out dependence on cigarettes: the cigarette dependence calibration. Neuropsychopharmacology. 2003;28:359–370. [PubMed] [Google Scholar]

31. Baumeister RF, Heatherton TF, Tice DM. Losing control: How and why people fail at self-regulation. Academic Press; San Diego, CA: 1994. [Google Scholar]

32. Pisinger C, Jorgensen T. Weight concerns and smoking in a general population: The Inter99 study. Preventive Medicine. 2007;44:283–289. [PubMed] [Google Scholar]

33. Borrelli B, Mermelstein R. The role of weight business organization and self-efficacy in smoking cessation and weight gain amid smokers in a dispensary-based cessation program. Addictive Behaviors. 1998;23:609–622. [PubMed] [Google Scholar]

34. French SA, Jeffery RW, Pirie PL, McBride CM. Practise weight concerns hinder smoking cessation efforts? Addictive Behaviors. 1992;17:219–226. [PubMed] [Google Scholar]

35. Sepinwall D, Borrelli B. Older, medically ill smokers are concerned about weight gain after quitting smoking. Addictive Behaviors. 2004;29:1809–1819. [PubMed] [Google Scholar]

36. Jones BT, Corbin W, Fromme K. A review of expectancy theory and alcohol consumption. Addiction. 2001;96:57–72. [PubMed] [Google Scholar]

elderquirld.blogspot.com

Source: https://www.ncbi.nlm.nih.gov/pmc/articles/PMC4517970/

Share :