Dispositional Mindfulness Predicts Enhanced Smoking Cessation and Smoking Lapse Recovery

Abstract

Background

Although mindfulness has been hypothesized to promote health behaviors, no research has examined how dispositional mindfulness might influence the process of smoking cessation.

Purpose

The current study investigated dispositional mindfulness, smoking abstinence, and recovery from a lapse among African American smokers.

Methods

Participants were 399 African Americans seeking smoking cessation treatment (treatments did not include any components related to mindfulness). Dispositional mindfulness and other psychosocial measures were obtained pre-quit; smoking abstinence was assessed 3 days, 31 days, and 26 weeks post-quit.

Results

Individuals higher in dispositional mindfulness were more likely to quit smoking both initially and over time. Moreover, among individuals who had lapsed at day 3, those higher in mindfulness were more likely to recover abstinence by the later time points. The mindfulness-early abstinence association was mediated by lower negative affect, lower expectancies to regulate affect via smoking, and higher perceived social support.

Conclusions

Results suggest that mindfulness might enhance smoking cessation among African American smokers by operating on mechanisms posited by prominent models of addiction.

Mindfulness has been defined as purposeful attention to present-moment experiences (1). Dispositional mindfulness (the tendency for mindful attention in daily life) appears to be multifaceted (2), and one of its most critical components is present-focused awareness (3). Although researchers have hypothesized that mindfulness might promote positive health behavior changes such as smoking cessation and have developed interventions to increase dispositional mindfulness for this purpose (e.g., 4), very little research has examined associations between dispositional mindfulness and smoking cessation. The current study examined whether dispositional mindfulness (i.e., present-focused awareness in daily life) predicts smoking cessation and to elucidate potential mechanisms underlying any association.

Mindfulness and Smoking Cessation: Rationale and Potential Mechanisms

Although the majority of smokers express interest in quitting, only a small portion quit successfully (5). Most smokers attempting to quit lapse within the first few days (6), and as many as 95% of lapses lead to full-blown relapse (7). Thus, researchers have emphasized the study of “milestones” within the cessation process (i.e., initial abstinence, lapses to smoking, and the transition from lapse to relapse; 8). A better understanding of protective factors that promote initial abstinence, prevent lapses, and inhibit lapses from progressing into full-blown relapse could help inform intervention strategies at various stages of the cessation process (8). Unfortunately, little is known about protective factors or personal characteristics that promote lapse recovery.

Mindfulness

Mindfulness might promote cessation, lapse prevention, and lapse recovery. Smokers tend to have lower levels of dispositional mindfulness than non-smokers (as measured by the Mindful Attention Awareness Scale [MAAS], a commonly used measure of mindfulness that is also used in the present study; 3, 9). Furthermore, among smokers, greater dispositional mindfulness (as assessed with the MAAS) is associated with lower tobacco dependence (10), lower negative affect and perceived stress (11), greater positive affect (12), and higher expectancies regarding abilities to regulate emotions without smoking (10). In the only known study of the association between dispositional mindfulness and smoking cessation, Spears et al. (13) found that nonjudgmental acceptance (a subscale of the Five Facet Mindfulness Questionnaire; 14) predicted higher odds of abstinence up to 26 weeks post-quit among Spanish-speaking smokers of Mexican heritage. In addition, mindfulness-based smoking cessation interventions (aimed at increasing mindful awareness) show promise (4, 15). As such, mindfulness is hypothesized to enhance cessation through its effects on emotions, nicotine dependence and withdrawal, agency, and social support.

Emotions

Negative affect and depressive symptoms predict increased smoking urges, difficulties with cessation, and greater risk for relapse (16, 17). Given that mindfulness is linked to lower negative affect (18), mindfulness might improve cessation by reducing negative affect, and may also reduce the tendency to react to negative affect by smoking (19, 20). Along these same lines, rather than becoming overwhelmed by negative affect associated with an early lapse (21), a highly mindful person might move on from this setback and resume their quit attempt. On the other hand, positive affect may protect against relapse (22). Thus, mindfulness might increase cessation through reduced negative affect and enhanced positive affect. Mindful attention to positive life experiences (e.g., each day of abstinence) may generate further increases in positive emotions (23), which could counter negative affective processes during cessation (e.g., withdrawal-induced negative affect), and thus lessen urges to smoke in attempt to reduce distress.

Dependence and Withdrawal

Greater tobacco dependence consistently predicts worse cessation outcomes (24). However, smokers with greater mindfulness have lower levels of dependence (10). Through conscious attention to thoughts, feelings, and external stimuli, mindfulness may “de-automate” the habitual processes and reactions that underlie addiction (25, 26). Withdrawal symptoms are also strong predictors of relapse (27). Negative affect is a core aspect of withdrawal (28), thus mindfulness might lessen withdrawal symptoms related to negative affect. Mindfulness might also reduce craving (29), and/or individuals' responses to the experience of craving (20).

Agency

A sense of agency (30) about one's ability to quit smoking can include both smokers' beliefs that they can regulate affect without smoking and self-efficacy for avoiding smoking in high-risk situations (10). Lower agency predicts greater difficulties quitting smoking (31, 32). Mindfulness may enhance agency by broadening perceived available thoughts and actions, thus strengthening individuals' personal coping resources, and loving kindness meditation (which increases mindful awareness) has been associated with increases in agentic thinking (that one has been/will be able to achieve one's goals; 33). Similarly, a broader perception of coping resources might enhance expectations to regulate emotions by means other than smoking. Indeed, more mindful smokers report both higher self-efficacy for not smoking and greater expectancies that they can regulate affect without smoking (10).

Social Support

Mindfulness may lead to improved relationships, possibly through attunement to others' thoughts and perspectives or enhanced communication/interaction styles (34). Mindfulness could foster more supportive relationships, which could promote abstinence and decrease likelihood of lapses (35, 36). No research has examined links between mindfulness and social support for smoking cessation. However, mindfulness is linked to improved social outcomes including reduced aggression (37), increased relationship satisfaction and constructive responses to conflict (38), and increased empathy (39).

Mindfulness and Smoking Cessation in African Americans

African Americans have disproportionately high rates of cancer, other chronic diseases, and mortality associated with smoking compared to the general population (40, 41). Although African Americans are more likely to express interest in quitting and to have attempted to quit in the past year compared to Whites, they are less likely to have successfully quit smoking (42). Higher levels of stress and depressive symptoms are associated with worse cessation outcomes among African American smokers (43, 44) and many stressors that are particularly prominent in minority populations (e.g., financial strain, discrimination) predict a lower likelihood of cessation (45, 46). Given that mindfulness has been linked to lower levels of stress and depressive symptoms (3), mindfulness may reflect an important individual characteristic that could reduce negative affect and improve cessation outcomes among African American smokers attempting to quit. Unfortunately, few studies of mindfulness have included significant proportions of African Americans (47).

Current study

No known research has examined whether dispositional mindfulness predicts smoking cessation or recovery from a lapse in African Americans. The current study investigated three questions: 1) Does mindfulness predict smoking abstinence early in a quit attempt and over time?, 2) Does mindfulness predict recovery of abstinence among smokers who lapse early in a quit attempt?, and 3) What are the mechanisms underlying relationships between mindfulness and cessation? The current study investigated whether naturally-occurring individual differences in dispositional mindfulness predict abstinence. This study utilized a sample of African American smokers enrolled in smoking cessation treatment. We predicted that associations between mindfulness (assessed at baseline) and cessation outcomes over time (up to 26 weeks post-quit) would be mediated by emotions, dependence and withdrawal, agency, and social support.

Method

Data were collected as part of a randomized clinical trial examining the efficacy of a culturally tailored, palmtop computer-delivered smoking cessation treatment for African Americans compared to treatment-as-usual. Neither treatment specifically taught mindfulness. Rather, the current study examined associations between dispositional mindfulness and smoking cessation. Although analyses revealed no effect of treatment on smoking abstinence (49), treatment group was included as a covariate in all analyses.

Participants

Participants were recruited via local print advertisements. Individuals were eligible if they were African American, between 21-65 years old, had smoked ≥ 5 cigarettes per day for ≥12 months, had an expired carbon monoxide (CO) level of ≥8 ppm, planned to quit smoking within the next 2 weeks, possessed a functioning home telephone number and permanent home address, and were able to understand English at a sixth grade literacy level or higher. Exclusion criteria were: regular use of tobacco products other than cigarettes, use of pharmacological smoking cessation treatments other than nicotine patches, medical contraindication for the nicotine patch, and current pregnancy/lactation. (See 48 for further details).

Procedure

Data were collected from participants at seven times: 3 pre-quit (days -19, -12, and -5 before quit date) and 4 post-quit (days 3, 10, and 31, and 26 weeks following quit date). The present study utilized data from day -19 pre-quit (baseline), day -5 pre-quit, days 3 and 31 post-quit, and week 26 post-quit. Following baseline, participants were randomly assigned to either standard treatment (ST) that included the nicotine patch, culturally sensitive self-help materials, and individual counseling, or to the palmtop computer-delivered treatment (CDT, which also included all components of ST). Participants received five brief in-person counseling sessions (from pre-quit day -12 through post-quit day 31). Counselors were blind to the baseline data.

Measures

Smoking abstinence

Three abstinence assessments were utilized for the current study, which were chosen to reflect three critical smoking cessation milestones: early abstinence, lapse recovery, and long-term abstinence. Specifically, day 3 abstinence reflected early abstinence as the majority (74%) of participants had lapsed by day 3, and week 26 reflects a common long-term abstinence timepoint in smoking cessation studies. Among participants who had lapsed by day 3, day 31 was selected to represent a reasonable time frame in which to regain abstinence and to permit assessment of 7-day point prevalence abstinence. Abstinence was biochemically verified through expired CO <10 ppm (61) and/or cotinine <20 ng/ml (62). Self-reports of abstinence that were not consistent with biochemical verification were coded as not abstinent.

Mindfulness

The Mindful Attention Awareness Scale (MAAS; 18) was administered at baseline. Participants responded to 15 statements (e.g., “I find myself doing things without paying attention”) on a 6-point Likert scale with reversed endpoints (1=Almost Always, 6 =Almost Never). Higher scores reflect greater mindfulness (i.e., present focused attention). The MAAS has been associated with greater acting with awareness (51) and indicators of better psychological functioning (e.g., lower depression, anxiety, and rumination; higher positive affect; 3), as well as neural correlates of attention and emotion regulation (52). Furthermore, meditation practitioners score higher on the MAAS than non-meditators (3), and MAAS scores increase after mindfulness-based treatment (53, 54). The MAAS showed good internal consistency in the current sample (α = 0.92).

Mediators

Emotions

The Positive and Negative Affect Schedule (55) consists of 20 adjectives that form subscales for positive affect and negative affect. Higher scores indicate higher positive and negative affect. Both positive and negative affect were assessed at day -5, day 3 and day 31, with Cronbach's alpha ranging 0.92 - 0.93. The 20-item Center for Epidemiological Studies— Depression (56) scale, designed to assess symptoms of depression in the general population, was used to assess depressive symptoms at day -5, day 3 and day 31 (α = 0.86 - 0.88 in current sample).

Dependence and Withdrawal

The 68-item Wisconsin Inventory of Smoking Dependence Motives (57) provides scores indicating “primary dependence” (automaticity, craving, loss of control, and tolerance) and “secondary dependence” (e.g., cognitive enhancement, positive and negative reinforcement), as well as a total score. This measure was administered at baseline and at 31 days post-quit (α at both time-points = 0.98). The 28-item Wisconsin Smoking Withdrawal Scale (28) yields a total score and 7 subscale scores (i.e., anger, anxiety, concentration difficulty, craving, hunger, sadness, and sleep problems) and was administered at day -5, day 3 and day 31 (α = 0.78 – 0.90).

Agency

Agency was assessed in two ways: self-efficacy for avoiding smoking in high-risk situations (the Self-Efficacy Scale; 58) and beliefs about ability to regulate affect both by smoking and by other means (the Affective Information Processing Questionnaire; 59). The Self-Efficacy Scale (specific to smoking) yields a total and 3 subscale scores: positive affect/social situations, negative affect situations, and habitual/craving situations. Higher scores reflect greater confidence in one's ability to avoid smoking. The 9-item version of the Self-efficacy Scale was administered at day -5, day 3 and day 31 (α = 0.87 – 0.94). The Affective Information Processing Questionnaire assessed individuals' expectations that they could regulate their mood either (a) by smoking or (b) by means other than smoking in negative affect situations described by 10 vignettes and was administered at baseline and at day 31 (α = 0.90 – 0.96).

Social Support

The Interpersonal Support Evaluation List (60) assessed perceived social support. Responses to this 12-item measure were scored such that higher scores indicate higher social support. In addition to a total score, the items form three subscales: appraisal, belonging, and tangible support. This measure was administered at baseline, day 3 and day 31 post-quit (α: 0.86 – 0.87 across scales and times).

Covariates

Demographics

Demographics included age, gender, education (≤ high school vs. > high school), and partner status (single/separated/divorced/widowed vs. married/living with partner).

Pre-Quit Tobacco Use

Two items were used to assess pre-quit tobacco use: “How many cigarettes a day do you smoke on average?” and “How soon after you wake up do you smoke your first cigarette?” (50).

Data Analysis Overview

To examine the relationship between mindfulness and smoking abstinence over time, logistic random coefficients regression models were utilized. An intention-to-treat procedure was followed, such that those with missing data were considered not abstinent. Models specified an unstructured covariance matrix for the vector of random intercept and slope of time for each subject. To examine associations between mindfulness and the specific processes of early lapse and lapse recovery, logistic regression analyses were conducted with the full sample and with the subsample of early lapsers at all applicable time points. Day 3 abstinence was used to separate early lapsers from early abstainers.

To examine mediators of associations between mindfulness and smoking abstinence, a series of analyses identified mediators of the relationship between: 1) baseline mindfulness and early abstinence, and (2) baseline mindfulness and lapse recovery among early lapsers. Potential mediators were tested in separate models. Simple mediation effects were obtained from bootstrapping analyses utilizing 5,000 re-sampling iterations and 95% confidence intervals (63, 64). Following typical mediation nomenclature, two paths were estimated for each simple mediation model: path a denotes the relationship between mindfulness and the candidate mediator, and path b denotes the relationship between the candidate mediator and abstinence (controlling for mindfulness).

For early smoking abstinence models, the total sample was utilized and each simple mediation model included baseline mindfulness as the predictor, day 3 abstinence as the outcome, and each mediator of interest (individually). Potential mediators were measured prior to the quit date. For lapse recovery models, potential mediators for day 31 and week 26 abstinence (analyzed separately) were examined among the subset of early lapsers. Potential mediators were drawn from the previous abstinence assessment (day 3 mediators for day 31 abstinence, day 31 mediators for week 26 abstinence). Final models controlled for socio-demographics (gender, education, age, partner status), treatment group (ST versus CDT), and pre-quit tobacco use (cigarettes per day, time to first cigarette). In analyses predicting week 26 abstinence, day 31 smoking status was entered as an additional covariate.

Results

Participant Characteristics

Participants were 399 African American smokers. Table 1 displays participant characteristics for the full sample and for the subsample of early lapsers. Participant attrition was 11.8% on day -5, 18.5 % on day 3, 27.6% on day 31, and 35.6% at week 26. Compared to participants who did not attend the final week 26 visit, those who did attend tended to be older (p = 0.02). Further descriptive data and relations among demographic variables, psychosocial variables, and smoking variables in this sample are presented by Businelle, Kendzor et al. (65), Kendzor, Businelle et al. (66), and Kendzor, Cofta-Woerpel et al. (67). See Table 2 for means and standard deviations of potential mediators at each time point.

Table 1. Participant characteristics.

	Full Sample (N=399)	Early Lapsers (N=295)
Female (%)	51	50
Married or living with partner (%)	21	18
≤ High school education (%)	52	56
Age (mean, SD)	42.4 (9.74)	41.5 (9.75)
Mindfulness (mean, SD)	4.29 (1.03)	4.20 (1.04)
Cigarettes per day (mean, SD)	20.56 (12.16)	21.68 (12.90)
Smoke within 5 min of waking (%)	59	63.4

Table 2. Descriptive statistics of potential mediating variables across all time points.

Potential	Mediator	Baseline	Day 3	Day 31	Week 26
Emotion
	Positive Affect		35.87 (8.77)	36.60 (9.01)	34.98 (9.75)
	Negative Affect		19.41 (8.88)	19.04 (8.63)	20.11 (8.90)
	Depressive Symptoms		14.80 (11.12)	14.11 (11.29)	14.74 (11.37)
Dependence
	Primary	4.87 (1.42)		2.85 (1.48)	3.09 (1.62)
	Secondary	4.18 (1.37)		2.70 (1.27)	2.84 (1.34)
Withdrawal
	Anger		1.69 (1.13)	1.50 (1.12)	1.55 (1.02)
	Anxiety		1.82 (0.89)	1.69 (1.05)	1.71 (0.94)
	Sadness		1.41 (0.83)	1.40 (0.91)	1.47 (0.84)
	Concentration		1.43 (0.87)	1.38 (1.01)	1.42 (0.85)
	Craving		2.06 (1.01)	1.82 (1.03)	1.87 (1.01)
	Hunger		2.31 (0.84)	2.19 (0.92)	2.02 (0.73)
	Sleep		1.94 (0.98)	1.83 (1.07)	1.88 (0.91)
Self-efficacy
	Positive Affect		3.50 (0.96)	3.50 (1.07)	3.19 (1.11)
	Negative Affect		3.10 (1.00)	3.20 (1.13)	2.95 (1.19)
	Habitual/Craving		3.62 (0.93)	3.60 (1.04)	3.29 (1.11)
	Total		3.41 (0.88)	3.43 (1.00)	3.14 (1.06)
Affect Regulation Expectancies
	By Not Smoking	4.89 (1.29)		5.46 (1.31)	5.16 (1.40)
	By Smoking	4.20 (1.58)		3.62 (1.94)	3.99 (1.78)
Social Support
	Appraisal	12.05 (3.04)	12.96 (2.58)	13.06 (2.66)	12.84 (2.72)
	Belonging	12.08 (2.65)	12.88 (2.43)	13.07 (2.41)	12.90 (2.59)
	Tangible	12.00 (2.89)	12.70 (2.55)	12.80 (2.66)	12.63 (2.78)
	Total	36.12 (7.56)	38.54 (6.50)	38.86 (6.74)	38.37 (7.17)

Mindfulness and Smoking Abstinence over Time

Both MAAS and time predicted abstinence longitudinally, such that increasing time from quit date reduced the odds of abstinence—from day 3 to day 31 (OR = 0.12, p < 0.001) and from day 3 to week 26 (OR = 0.007, p < 0.001)—and higher mindfulness scores increased the odds of abstinence such that an increase in MAAS score of 1 point approximately doubled the odds of abstinence (OR = 2.22, p < 0.01). MAAS scores significantly predicted abstinence in the unadjusted model (OR = 2.22, p = 0.004) as well as in the model adjusted for demographics, treatment, and pre-quit tobacco use (OR = 1.75, p = 0.04). This analysis was also conducted with the subset of participants who completed the study (N = 259). The pattern of results remained the same, although the finding that higher MAAS scores were associated with an increased odds of abstinence only approached significance (OR = 1.42, p = .088). The interaction between mindfulness and time was not significant for any time contrasts (ps > .16).¹ Figure 1 displays abstinence rates over time for individuals low and high in mindfulness (using values +/- 1 SD from the mean).

Figure 1.

Abstinence rates for individuals low and high in dispositional mindfulness.

Low MAAS and High MAAS rates generated from values +/- 1 SD from the mean on the total sample MAAS scores. MAAS = Mindful Attention Awareness Scale.

Mindfulness, Early Smoking Abstinence, and Lapse Recovery

Given the significant association between mindfulness and abstinence over time, simple logistic regression analyses were conducted examining associations of mindfulness with early abstinence (day 3) and recovery of abstinence among individuals who had lapsed by day 3. Among the total sample, greater mindfulness predicted greater odds of abstinence at day 3 (OR = 1.48, p = 0.001). Among early lapsers (n = 295), those with higher mindfulness were more likely to be abstinent at day 31 (OR = 1.77, p = 0.05) and week 26 (OR = 2.17, p = 0.009). See Table 3.

Table 3. Dispositional mindfulness predicts smoking abstinence across time and cessation milestones.

Logistic Random Coefficients Regression Models			Abstinence Across Time (full sample)
			OR		p
Mindfulness (unadjusted)			2.22		0.004
Mindfulness (adjusted for Demographics and Treatment)			1.69		0.005
Mindfulness (adjusted for Demographics, Treatment, and Pre-quit tobacco use)			1.75		0.04
Simple Logistic Regression Models	Early Abstinence (Day 3) among full sample		Lapse Recovery (Day 31) among early lapsers		Lapse Recovery (Week 26) among early lapsers
	OR	p	OR	p	OR	p
Mindfulness (unadjusted)	1.48	0.001	1.77	0.050	2.17	0.009
Mindfulness (adjusted for Demographics and Treatment)	1.35	0.019	1.94	0.032	2.41	0.005
Mindfulness (adjusted for Demographics, Treatment, and Pre-quit tobacco use)	1.27	0.069	2.04	0.024	2.27	0.011

Note. Odds ratios reported are predicting smoking abstinence. Full sample N = 399. Early lapsers (those who smoked by day 3) N = 295. Demographic covariates were gender, education, age, and partner status. Pre-quit tobacco use covariates were cigarettes per day and time to first cigarette.

Mediators of the Association between Mindfulness and Early Smoking Abstinence

Of the individual models examined, three general types of variables emerged as significant mediators of the association between mindfulness and early abstinence: 1) negative emotional experiences, 2) affect regulation expectancies, and 3) social support. First, mindfulness predicted lower levels of anger, sadness, and depressive symptoms (each associated with smoking lapses), which then predicted higher likelihood of abstinence. Second, more mindful individuals indicated lower expectations that they would regulate their affect by smoking, which predicted higher likelihood of abstinence. Third, greater mindfulness was associated with higher levels of social support (total score, appraisal, and belonging), which was associated with greater likelihood of abstinence. See Table 4.

Table 4. Potential mediators of baseline mindfulness to day 3 abstinence.

Potential Mediator	Coefficients		Indirect effect (SE)	BC 95% CI		PME
	a	b		Lower	Upper
Emotion
Positive Affect	2.41	.02	.05 (.04)	-.024	.140	.20
Negative Affect	-3.00	-.03	.09 (.06)	-.023	.223	.37
Depressive Symptoms	-4.71	-.03	.16 (.08)	.029	.329	.64
Dependence
Primary	-.46	.03	-.016 (.05)	-.116	.088	.05
Secondary	-.45	-.04	-.018 (.05)	-.132	.087	.07
Total	-5.84	.004	-.020 (.06)	-.135	.088	.08
Withdrawal
Anger	-.43	-.37	.16 (.07)	.032	.303	.66
Anxiety	- .37	-.29	.11 (.07)	-.024	.252	.45
Sadness	-.28	-.38	.10 (.06)	.001	.233	.44
Concentration	-.41	-.11	.04 (.08)	-.101	.207	.19
Craving	-.18	-.09	-.02 (.04)	-.100	.050	.06
Hunger	-.13	-.10	.01 (.03)	-.028	.080	.05
Sleep	-.32	-.21	.07 (.06)	-.029	.188	.29
Self-efficacy
Positive Affect	.07	.52	.04 (.03)	-.009	.119	.15
Negative Affect	.07	.38	.03 (.03)	-.012	.091	.11
Habitual/Craving	.05	.49	.02 (.03)	-.016	.092	.10
Total	.06	.62	.04 (.03)	-.009	.118	.15
Affect Regulation Expectancies
By Not Smoking	.34	.10	.04 (.04)	-.040	.119	.14
By Smoking	-.21	-.15	.03 (.02)	.001	.091	.13
Social support
Appraisal	1.10	.11	.12 (.07)	.006	.271	.49
Belonging	.93	.13	.12 (.06)	.018	.251	.50
Tangible	1.11	.09	.100 (.06)	-.006	.236	.41
Total Score	3.14	.05	.151 (.08)	.025	.315	.63

Note. All models controlled for age, gender, education, partner status, pre-quit tobacco use, and treatment group. When tobacco dependence variables were tested as mediators without controlling for baseline cigarettes per day and time to first cigarette, the results did not change. Significant indirect effects are shown in bold. Unstandardized coefficients are reported for substantive interpretation (85). a = unstandardized coefficient for relationship between mindfulness and potential mediator. b = unstandardized coefficient for relationship between potential mediator and abstinence. SE = standard error of indirect effect. BC 95% CI = bias-corrected 95% confidence intervals for total indirect effect. PME = Proportion of Mediated Effect. PME was calculated with the equation ab/(c′+ab) (86) utilizing the absolute values of the direct and indirect effects (87). PMEs have been shown to be unstable with sample sizes < 500 (88) but are reported here for illustrative purposes.

Mediators of the Association between Mindfulness and Lapse Recovery

In predicting day 31 abstinence among early lapsers, none of the potential mediators (measured at day 3) produced significant indirect effects. In predicting week 26 abstinence among early lapsers, only negative affect emerged as a having a significant indirect effect on week 26 abstinence. Mindfulness predicted lower negative affect at day 31 (b_{path a} = -2.81, p < .001), and lower negative affect predicted higher rates of abstinence (b_{path b} = -0.12, p < .05). The indirect effect of mindfulness on abstinence through lower negative affect was statistically significant, b_indirect = .349 (95% CIs [0.096, 0.755], SE = 0.23). Thus, whereas negative affect, social support, and affect regulation expectancies each mediated the association between mindfulness and early abstinence, none of the hypothesized variables was a mediator with regard to day 31 abstinence, and only negative affect mediated the relationship between mindfulness and long-term (week 26) abstinence.

Discussion

Dispositional mindfulness predicted smoking abstinence over time, early smoking abstinence, and lapse recovery among African American smokers seeking treatment. The link between greater dispositional mindfulness and early abstinence was mediated by lower negative affect (i.e., depressive symptoms, anger, and sadness), lower expectancies regarding the ability of smoking to regulate negative affect, and higher perceived social support. The association between dispositional mindfulness and lapse recovery by week 26 was only mediated by lower negative affect. Overall, the mediators that most consistently linked mindfulness to abstinence and lapse recovery were indicators of negative affect. The importance of negative emotions as mediators is consistent with theories of addiction (68) that assert that avoidance of or escape from negative affect is a key determinant of relapse. In addition, theories on mechanisms underlying mindfulness (69, 70) place central importance on the role of mindfulness in promoting more effective regulation of unpleasant thoughts and emotions. The current findings extend previous work by showing that lower negative affect is a primary mechanism through which mindfulness predicts smoking cessation.

Social support, which also plays an important role in theories of cessation (35), has not yet been a prominent factor in models of mindfulness. Although mindfulness has been linked to positive interpersonal outcomes (37-39), the current study is the first known to link mindfulness to perceptions of social support specifically. The findings underscore the importance of social support as a key factor influencing early abstinence and extend the potential mechanisms of mindfulness from intrapersonal to interpersonal factors. Research suggests that having a strong social support network could be a protective factor with regard to smoking for African Americans (71-73). Perhaps mindfulness-based interventions for smoking cessation, especially among African Americans, could be enhanced either by encouraging the use of mindfulness in social interactions, or by utilizing social media to capitalize on social support.

Theories of addiction also highlight the role of agency (e.g., self-efficacy to avoid drug use) as key in predicting abstinence (21, 35). Fredrickson's broaden-and-build framework explains how mindfulness might enhance positive emotions, which could strengthen personal resources including agency (33). However, the current study was the first known to explore agency as a potential mediator in the mindfulness-smoking abstinence link. The current study provides only partial support for the role of agency, specifically through expectancies to regulate negative affect by smoking. Given that mindfulness has been related to greater self-efficacy and expectations to regulate affect without smoking (10), future research on the role of agency in mediating the mindfulness-abstinence association may be warranted.

The current findings shed light on the process of abstinence recovery following a lapse, a critical “milestone” in the cessation process (8). Given that most lapses lead to relapse (7), understanding the processes and personal characteristics that contribute to lapse recovery is critical. This study suggests that mindfulness could help prevent lapses from interfering with longer-term cessation. Furthermore, lower negative affect at day 31 mediated the effect of mindfulness on week 26 abstinence among early lapsers. Lower distress associated with lapses may be due to a “decentered” perspective (74). The tendency to observe thoughts and feelings without reacting to them might enable more mindful individuals to view a lapse as a single slipup rather than as a devastating failure. This mindful responding might lessen the likelihood of smoking as a way of alleviating lapse-associated distress, and thus promote long-term abstinence.

Overall, our findings most consistently highlight lower negative affect as a primary mediator of associations between mindfulness and abstinence in African American smokers. Although greater negative affect is a consistent predictor of lapse and relapse across various populations, the role of negative affect as a trigger for smoking may be even more prominent among African Americans. African Americans often face unique stressors, including chronic racial discrimination, which increase negative affect and contribute to health disparities (75-77). Given that chronic stress and discrimination negatively impact smoking cessation in minority populations (44, 45), future research should also examine whether mindfulness-based interventions improve coping and reduce negative affect related to specific culturally-relevant stressors.

Limitations and Future Directions

Given that the sample was entirely African American, it is unclear whether results will generalize to other racial/ethnic groups. This study is also limited by exclusive use of the MAAS as a unidimenional measure of mindfulness. Other research has highlighted that mindfulness may be a multidimensional construct (2). Thus, the MAAS may not capture the full complexity of mindfulness (78), but it does appear to capture a core element of mindfulness (i.e., present-focused attention and awareness). Future studies should examine whether particular facets of mindfulness predict cessation. For example, recent research (13) suggests that non-judgment is one aspect of mindfulness that predicts cessation. Given that mindfulness is associated with lower neuroticism (3, 79, 80), which has been linked to cessation (81), research might also examine whether personality variables might account for relationships between mindfulness and smoking.

This study found no significant mediators of abstinence recovery by day 31, and only negative affect as a mediator of abstinence recovery by week 26. This could be due to an insufficient sample size or high relapse rates among the sample. In addition, mediation analyses were likely hampered by the significant time lags between measurement of the mediators and abstinence. Shiffman and Waters (82) demonstrated that rapid shifts in negative affect predicted relapse. Future research should measure moment-to-moment changes in potential mediators and examine how trajectories of these constructs over time might predict relapse (83, 84).

Finally, although this study examined potential mediators in separate models, these mediators likely affect one another. For example, the availability of a strong social support system may increase positive and reduce negative emotions during the quit process. In addition, mindfulness reduces negative emotions, which may then enhance agency. Future research might examine more integrative models and study designs that enable a more fine-grained analysis of the reciprocal relations among mindfulness and potential mechanisms.

Conclusions

This is the first known prospective study of associations between dispositional mindfulness and smoking cessation at both early and later stages of a quit attempt in African American smokers. Results suggest that dispositional mindfulness promotes smoking cessation and ability to recover from an early lapse. This improved lapse recovery appears to emerge primarily through lower negative affect. This research adds to the small but growing literature on the benefits of mindfulness in racial/ethnic minority populations. Findings support the development of mindfulness-based interventions (which aim to enhance dispositional mindfulness) for promoting smoking cessation at various stages within a quit attempt.