Anhedonia in Tobacco Withdrawal among African American Smokers

Abstract

Background:

Anhedonia—diminished interest and pleasure in response to rewards—may be a symptom of tobacco withdrawal that is understudied in priority populations. This experiment investigated the magnitude and correlates of various dimensions of anhedonia during tobacco withdrawal among African American (AA) smokers—a population subject to health disparities.

Method:

AA smokers (N=607; ≥10 cigarettes/day, 37.8% female, M[SD] age=50.0[10.6] years) completed self-report measures assessing expected pleasure from (i.e., consummatory anhedonia) and desire to engage in (i.e., anticipatory anhedonia) various types of hypothetically experienced rewards at counterbalanced 16-hr tobacco deprived and non-deprived lab visits. Other tobacco withdrawal symptom measures (e.g., craving, negative affect, hunger) were also assessed.

Results:

Tobacco deprivation most robustly increased scores on a composite measure of consummatory anhedonia directed towards various reward domains (i.e., hobbies, sensory experiences, social activities; d=.32, p<.001). Deprivation modestly increased consummatory and anticipatory anhedonia directed toward sexual rewards (ds=.09 to .12, ps<.02), did not significantly change anhedonia toward food rewards, and decreased anhedonia directed toward psychoactive drug rewards (i.e., increased desire for and pleasure from drugs; ds=−.21 to −.19, ps<.001). Deprivation-induced changes in anhedonia were modestly correlated with other withdrawal symptoms (average ∣r∣s=.04 to .23) and were amplified among participants with higher nicotine dependence and lower positive affect-related traits (∣β∣s=.10 to .12, ps<.01).

Conclusion:

Some dimensions of anhedonia may be genuine expressions of acute tobacco withdrawal in AA smokers. Applying multi-dimensional anhedonia conceptualizations might advance basic knowledge and treatment of tobacco use disorder, improve smoking cessation outcomes, and address tobacco-related health disparities facing AA smokers.

African American (AA) smokers bear a disproportionate brunt of the public health burden of smoking (Henley et al., 2016; Simmons et al., 2016). Relative to white smokers, AA smokers are considerably more likely to relapse following quit attempts despite being more likely to initiate them (Kulak et al., 2016; Pagano et al., 2018; Weinberger et al., 2019). A major barrier to tobacco cessation both within the general smoking population and among AA smokers, the tobacco withdrawal syndrome is a set of unpleasant symptoms that emerge shortly after the discontinuation of sustained tobacco use (American Psychiatric Association, 2013; Hooper et al., 2017; Weinberger et al., 2017). Considerable evidence indicates that craving, negative affect, disturbances in appetite and sleep, and impaired concentration are valid expressions of tobacco withdrawal and are consequently included in the Diagnostic and Statistical Manual of Mental Disorders Fifth Edition (DSM-5) definition of tobacco use disorder (American Psychiatric Association, 2013; Hughes, 2007). Recently, it has been proposed that additional symptoms outside of those currently identified within the DSM-5 may also be valid manifestations of tobacco withdrawal (e.g., anhedonia, reduced positive affect, impulsivity; Hughes et al., 2014; Hughes et al., 2020; Klemperer et al., 2020; Piper, 2015). A comprehensive understanding of potentially novel (i.e., currently excluded from the DSM-5) symptoms of tobacco withdrawal is not only critical to advancing the science and treatment of tobacco use disorder but may also generate insight into risk factors that contribute to poor cessation outcomes and tobacco-related health disparities among AA smokers.

Emergent data suggests that anhedonia—diminished pleasure and interest in response to (non-drug) rewarding stimuli—may be one such novel symptom of tobacco withdrawal (Cook et al., 2017; Cook et al., 2015; Hughes et al., 2017; Piper et al., 2017; for a review, see Hughes et al., 2020). Defined by diminished appetitive functioning (i.e., behaviors, thoughts, and feelings evoked in response to commonly rewarding stimuli to the end of continued reward attainment; Hatzigiakoumis et al., 2011; Leventhal & Zvolensky, 2015), anhedonia is one of two core symptoms that make up major depressive disorder (MDD; American Psychiatric Association, 2013), alongside poor mood. Anhedonia is distinct from poor mood and depression in that it is possible to have poor mood and meet criteria for MDD without experiencing anhedonia (i.e., anhedonic individuals have a reduced capacity to experience pleasure irrespective of their level of poor mood, whereas those with poor mood often report still being able to experience pleasure; Ho & Sommers, 2013). Anhedonia is also distinct from negative or positive affect in that it does not reflect a current internal mood state as much as a cognitive appraisal of an external stimulus (i.e., reward), hence its alternative designation as decreased reward sensitivity/responsivity (Der-Avakian & Markou, 2012). Moreover, anhedonia is multifactorial in that it is driven by impairment(s) in at least one of three reward processing mechanisms (i.e., wanting, liking, learning) that each involve varying degrees of affective processing (Rømer Thomsen et al., 2015). Thus, although experiences of pleasure and positive affect are functionally and neuroanatomically connected (Kringelbach & Berridge, 2009), the anhedonia construct represents more than just the affective component of the subjective pleasure response.

To be considered a valid manifestation of withdrawal, any new symptom should conform to the following criteria: (a) biological plausibility; (b) reliable emergence after drug use discontinuation following a period of chronic use; (c) suppression by reinstatement of drug use or an agonist-like compound; and (d) association with other key manifestations of drug withdrawal and drug use disorder (e.g., loss of control, relapse; Siegel, 1983). Biological plausibility is supported by extensive cross-species evidence that chronic nicotine exposure alters the brain’s reward system homeostasis such that (1) normal reward processing becomes contingent on the presence of nicotine in the bloodstream and (2) nicotine deprivation diminishes behavioral and biological responsiveness to non-drug rewards (Epping-Jordan et al., 1998; Oliver et al., 2017; Pergadia et al., 2014; Sweitzer et al., 2014). Evidence for anhedonia’s conformity to the remaining three withdrawal criteria has been established in a series of pioneering studies evaluating whether anhedonia displays the requisite features of a tobacco withdrawal symptom among human tobacco users (Cook et al., 2017; Cook et al., 2015; Hughes et al., 2017; Piper et al., 2017).

For example, anhedonia has been shown to reliably emerge upon acute tobacco deprivation experienced in relation to both naturalistic tobacco cessation (Cook et al., 2017; Cook et al., 2015; Piper et al., 2017) and experimentally-manipulated abstinence (Hughes et al., 2017). In support of the third withdrawal criteria, another set of studies demonstrated that deprivation-induced increases in anhedonia are subsequently attenuated by nicotine administration (Cook et al., 2015; Guillot et al., 2015; Piper et al., 2017). Finally, prior work on the topic has also provided preliminary evidence of associations between deprivation-induced increases in anhedonia and manifestations of tobacco use disorder (Cook et al., 2017; Cook et al., 2015; Piper et al., 2017), tobacco withdrawal (Cook et al., 2017; Piper et al., 2017), as well as poor cessation outcomes (i.e., lower post-quit date abstinence rates, shorter latency to relapse; Cook et al., 2015; Piper et al., 2017). Hence, anhedonia appears to meet the four aforementioned criteria for consideration as a valid tobacco withdrawal symptom.

With limited exception (Guillot et al., 2015), AA and other racial/ethnic minority smokers are vastly underrepresented across the extant literature on anhedonia as a tobacco withdrawal symptom. Nonetheless, anhedonia-related tobacco withdrawal may be particularly germane to understanding processes that underlie tobacco-related health disparities among AA smokers. Conclusions drawn from the nascent literature base focused on racial differences in tobacco withdrawal have been equivocal, with some studies observing less severe tobacco deprivation effects among AA (versus non-AA) smokers (Riedel et al., 2003; Robinson et al., 2014; Weinberger et al., 2017) and others finding the reverse pattern (Bello et al., 2015; Soyster et al., 2016). For example, AA smokers have been found to report larger reductions in positive affect following tobacco deprivation compared to smokers of other racial/ethnic backgrounds (Bello et al., 2015). In addition, AA smokers also report greater motivation to smoke to obtain pleasure and enhance positive affect compared to White smokers (Goldenson et al., 2015; Lam et al., 2008). Collectively, this research suggests that AA smokers may be especially vulnerable to expressions of tobacco withdrawal that tap affective and hedonic processing, such as anhedonia. Identifying whether deprivation-induced anhedonia constitutes a significant component of the tobacco withdrawal syndrome among AA smokers could shed light on smoking cessation approaches that may help reduce tobacco-related disparities faced by this population.

One barrier to the scientific translation of anhedonia-withdrawal findings is the insufficient phenotypic characterization (i.e., observed variation in type, form, severity) of the multidimensional phenomenology of anhedonia during tobacco withdrawal. Anhedonia can be parsed into two key domains: (1) anticipatory anhedonia (i.e., diminished interest or desire to engage in rewarding experiences), and (2) consummatory anhedonia (i.e., diminished pleasure or enjoyment from rewarding experiences; Ho & Sommers, 2013; Treadway & Zald, 2013). These two dimensions of anhedonia may be underpinned by neurobiologically distinct mechanisms (Leventhal & Zvolensky, 2015; Treadway & Zald, 2011), which suggests that they may be differentially impacted by nicotine deprivation and warrant different smoking cessation strategies. Despite these differences, anticipatory and consummatory facets of anhedonia symptoms have not typically been separated in prior tobacco research (Cook et al., 2017; Cook et al., 2015; Hughes et al., 2017; Piper et al., 2017). Nor has prior work examined whether nuanced differences in withdrawal-related anhedonia exist as a function of the specific type of reward in question, but instead tends to collapse deprivation-induced anhedonic responses to a series of typically rewarding activities into a single composite measure to capture general reward responsivity (Cook et al., 2017; Cook et al., 2015; Hughes et al., 2017; Piper et al., 2017).

Another remaining gap in the evidence base relates to the lack of empirical data on individual difference characteristics that may predict greater propensity toward experiencing anhedonia during tobacco withdrawal. Beyond tobacco dependence, which would theoretically demarcate vulnerability to experiencing withdrawal-related anhedonia were it to constitute a valid tobacco withdrawal symptom, a host of other behavioral traits may differentially predict withdrawal-related anhedonia. For example, the Situation × Trait Adaptive Response (STAR) model of nicotine dependence purports that between-person differences in behavioral traits among smokers moderate the qualitative profile of nicotine administration and deprivation effects in a trait-consistent manner (Gilbert, 1997; Gilbert & Gilbert, 1998). Hence, traits that are pathognomonically-concordant (i.e., overtly characteristic) of a given tobacco withdrawal symptom should theoretically be robust predictors of that symptom’s expression during tobacco deprivation. Knowledge of risk-factors for experiencing acute deprivation-induced anhedonia is critical to characterizing anhedonia-related withdrawal phenotypes and providing more precise intervention targets that can be used in personalized medicine strategies tailored to patients’ clinical profiles.

The Current Study

This study characterized the multi-faceted phenotypic expressions of self-reported anhedonia induced by overnight tobacco deprivation in AA daily cigarette smokers. Study objectives were: (1) to establish the effect sizes of tobacco deprivation on anhedonia indicators and compare those to established symptoms of tobacco withdrawal (i.e., urges, affect, concentration problems, hunger, motivation to reinstate smoking); (2) to investigate the extent to which different manifestations of anhedonia during tobacco withdrawal correlate with one another and correlate with established tobacco withdrawal symptoms; and (3) to determine whether individual difference characteristics (i.e., affective/psychopathological symptoms, tobacco dependence) predict the magnitude of increase in anhedonia provoked by acute tobacco withdrawal.

Related to the first objective, we hypothesized that both self-reported anticipatory anhedonia and consummatory anhedonia would increase with tobacco deprivation. Second, we hypothesized that anticipatory and consummatory anhedonia would be correlated with one another and with established tobacco withdrawal symptoms. True to the conceptualization of the anhedonia phenotype, it is possible that diminished reward responsiveness following nicotine deprivation might be preferentially expressed toward natural non-pharmacologic rewards and not observed in response to non-nicotine drug rewards. Thus, we also hypothesized that deprivation-induced changes in anhedonia directed toward primary (non-drug/alcohol) rewards (i.e., food, sex) would be stronger than deprivation-induced changes in anhedonia directed toward drug-related rewards. Finally, we hypothesized a strong, positive association between tobacco dependence severity and the magnitude of deprivation-induced increases in anhedonia. Based on the STAR model of nicotine dependence (Gilbert, 1997; Gilbert & Gilbert, 1998), we also hypothesized that affective/psychopathological constructs that are pathognomonically-concordant with anhedonia (i.e., related to diminished positive affect and reward processing) would predict greater deprivation-induced increases in anhedonia, whereas psychopathological symptoms related to negative affect or other pathognomonically-discordant problems (e.g., somatic features, trauma) would not predict anhedonia during tobacco deprivation.

Methods

Participants

Data were drawn from the Southern California Tobacco Addiction Phenotype Project (SC-TAPP), a laboratory study of individual differences in the expression of tobacco withdrawal among AA smokers in the Los Angeles, CA, USA Metropolitan Region conducted from August 2013 to May 2017 (N = 607; Bello et al., 2017; Kelly et al., 2020; Liautaud et al., 2018). Non-treatment-seeking daily cigarette smokers were recruited via advertisements announcing opportunities to participate in research on the causes of smoking. Eligible participants were: (1) ≥ 18 years of age; (2) self-reported non-Hispanic AA ethnicity/race; (3) smoked ≥ 10 cigarettes per day for ≥ 2 years; and (4) fluent in English. Participants were excluded if they: (1) met diagnostic criteria for current DSM-IV non-nicotine substance dependence, including cannabis dependence (to minimize alcohol and/or other drug withdrawal symptoms during experimental sessions); (2) exhaled carbon monoxide (CO) levels < 10 parts per million (ppm) at intake; (3) expressed a strong desire to cut down or quit smoking in the next 30 days; (4) reported current use of nicotine replacement therapy or psychiatric medications; (5) reported current daily use of other tobacco products or cannabis; or (6) reported currently breast feeding or being pregnant. Participants were compensated approximately $230 USD for completing the entire study. The University of Southern California Institutional Review Board approved study procedures. All participants provided written informed consent prior to enrolling in the study.

Procedures

After a telephone eligibility screen, participants completed an in-person baseline visit involving additional screening and assessments, which included completing self-report questionnaires, CO and breath alcohol concentration (BrAC) analyses, and structured clinical interviews (Structured Clinical Interview for DSM-IV Research Edition; First et al., 2002). Participants then attended two counterbalanced experimental visits starting at approximately 12pm in which they were instructed to either abstain from smoking after 8pm the night before their visit (16-hour tobacco-deprived condition) or smoke as they normally would prior to their visit (non-deprived condition). Participants were also instructed to avoid the use of any alcohol, other tobacco products, or cannabis for 24 hours prior to their visits. For the non-deprived visit, participants smoked a single cigarette of their preferred brand in the laboratory at the beginning of the visit in order to standardize abstinence levels across the sample and were then tested for breath alcohol (BrAC = .000 required to continue participation) and CO. For deprived sessions, the visit began with breath alcohol and CO assessment. Participants with CO levels > 9 ppm during their abstinent session (n = 42) were considered non-abstinent and rescheduled to complete their session on another day. Participants who did not meet CO criteria for abstinence (≤ 9 ppm) at their second attempt were discontinued (n = 8). At both experimental visits following breath CO assessment, participants were given an identical battery of self-report measures that assessed anhedonia and other manifestations of tobacco withdrawal at a single time point, and they then completed a behavioral smoking task that began approximately 75 minutes after the start of the visit.

Baseline Visit Questionnaires

Demographic and Smoking History Questionnaire

Participants were asked to self-identify on demographic and smoking characteristics of the sample (i.e., age, gender, educational attainment, employment status, annual income, cigarettes/day, menthol preference, and age of onset of regular smoking) using an investigator-designed survey.

Fagerström Test for Cigarette Dependence (FTCD; Fagerström, 2011; Heatherton et al., 1991)

The FTCD is a 6-item self-report measure that assesses tobacco dependence severity. A total score is created by summing all 6 items, with scores ranging from 0 to 10 (higher scores indicate greater tobacco dependence). The first item on the FTCD instructs participants to indicate how soon they typically have their first cigarette of the day upon waking in the morning (i.e.; time to first cigarette [TTFC]; 1=Within 5 minutes, 2=6 to 30 minutes, 3=31 to 60 minutes, and 4=After 60 minutes); since TTFC is a strong predictor of smoking relapse (Baker et al., 2007), this item was analyzed as an additional standalone indicator of dependence.

Inventory of Depressive and Anxiety Symptomology (IDAS; Watson et al., 2007)

The IDAS is a 58-item measure consisting of 10 domain-specific psychopathologic symptom subscales: Dysphoria (e.g., sadness, guilt, concentration problems, worry), Lassitude (i.e., anergia and hypersomnia), Insomnia, Appetite Loss, Appetite Gain, Ill Temperament (i.e., irritability), Well-Being (i.e., positive affect), Panic, Social Anxiety, and Traumatic Intrusions. Participants are instructed to rate the extent to which they experienced symptoms (e.g., “I felt exhausted.”) during the past two weeks on 5-point scales (1=Not at all to 5=Extremely). Mean rating per item is calculated for each subscale. Previous studies have found adequate internal consistency and construct and factorial validity of this measure (Watson et al., 2008; Watson et al., 2007).

Center for Epidemiologic Studies Depression Scale (CES-D; Radloff, 1977)

The CES-D is a well-validated 20-item scale that instructs participants to rate how often they experienced depressive symptoms (e.g., “I felt that I could not shake off the blues even with the help from my family”) during the past week on 4-point scales (0=Rarely or None of the time [0-1 days] to 3=Most or all of the time [5-7 days]). Based on prior work showing a four-factor structure of the CES-D in smokers (Leventhal et al., 2008; Shafer, 2006), we computed subscale scores based on each scales respective items’ average score (negative affect [7 items], positive affect [4 items], somatic features [7 items], and interpersonal problems [2 items]), as in prior work. The CES-D was only included in the baseline questionnaire battery for only the first 463 participants (see Study Sample below).

Subjective Happiness Scale (SHS; Lyubomirsky & Lepper, 1999)

The SHS is a 4-item scale that assesses dispositional happiness on a general level, as well in relation to others. Participants are instructed to rate their happiness on 7-point Likert scales, with higher scores representing higher levels of happiness (e.g., “Compared to my peers, I consider myself” from 1=Less Happy to 7=More Happy). The SHS has shown adequate psychometric properties in previous research (Ameringer & Leventhal, 2015; Lyubomirsky & Lepper, 1999; Mattei & Schaefer, 2004). Cronbach’ α for all baseline multi-item measures are reported in Supplemental Table 1.

Experimental Visit Questionnaires

Anhedonia

Snaith Hamilton Pleasure Capacity Scale (SHAPS; Snaith et al., 1995).

The 14-item SHAPS is a unidimensional measure of generalized consummatory anhedonia across a broad domain of rewards. The SHAPS state-version instructs participants to rate the degree of pleasure (0= Strongly Agree to 3=Strongly Disagree) they would hypothetically feel “right now, in the current moment” in response to various interests/hobbies, social activities, and sensory experiences that are typically pleasant (“e.g., I would enjoy being with family or close friends”). A mean score is generated across the 14 items, and responses were reflected, such that higher scores indicate greater anhedonia (i.e., diminished pleasure capacity; Cronbach’s α for deprivation-induced change score in this sample = .91). The SHAPS was selected as the primary outcome variable because it is well-established clinical tool in anhedonia assessment with high translation potential and has demonstrated robust sensitivity to tobacco deprivation effects as well as excellent construct validity, internal consistency, and test-retest reliability in prior work (Dawkins et al., 2009; Franken et al., 2007; Leventhal et al., 2006; Powell et al., 2004).

Sensitivity to Reinforcement of Addictive and other Primary Rewards (STRAP-R; Goldstein et al., 2010).

To provide a supplementary multidimensional anhedonia assessment, we included the STRAP-R, which measures both consummatory and anticipatory anhedonia directed toward three different reward domains (i.e., sex, food, drugs/alcohol). The STRAP-R prompts participants to think about experiencing rewards from each of the three domains (i.e., “your favorite food,” “your favorite sexual activity,” and “your favorite alcohol or drug [including caffeine or any other substance that changes the way that you feel]”). Participants provide two ratings for each reward domain: 1) “How pleasant would it be to do it right now?” (consummatory) and 2) “Do you want to do it right now?” (anticipatory), utilizing 5-point Likert scales (5=Somewhat, 4=Slightly, 3=Moderately, 2=Very, 1=Extremely). Each rating is analyzed as a separate outcome, with higher scores reflecting greater anhedonia (i.e. diminished pleasure). Previous research suggests preliminary divergent validity and sensitivity of the STRAP-R to pharmacological manipulations (e.g., methylphenidate; Goldstein et al., 2010). The STRAP-R was only included in the experimental questionnaire battery for the first 463 participants (see Study Sample below).

Other Manifestations of Tobacco Withdrawal

Each of the following measures has shown sufficient sensitivity to tobacco deprivation effects and adequate internal consistency across tobacco deprived and non-deprived states as well as deprivation-induced changes (Leventhal et al., 2010; Liautaud et al., 2018). See Supplemental Table 2 for Cronbach’s α for each measure.

Profile of Mood States (POMS; McNair et al., 1971).

The POMS is an affect check list that instructs participants to rate how much they are feeling 72 different affect adjectives “right now” on 5-point Likert scales (0=Not at all to 4=Extremely). As in prior work (Leventhal, Ameringer, Osborn, Zvolensky, & Langdon, 2013), we computed 4 negative mood subscales (Anger [12 items], Anxiety [8 items], Confusion [7 items], and Depression [15 items]) and 3 positive mood subscales (Elation [6 items], Friendliness [8 items], and Vigor [8 items]) based on mean rating per item within respective subscale items and overall scores for negative mood valence (NM; mean of 4 negative mood subscale scores) and positive mood valence (PM; mean of 3 positive mood subscales).

Brief Questionnaire of Smoking Urges (QSU; Cox et al., 2001).

The QSU consists of 10 items that evaluate two dimensions of smoking urges: appetitive urge (Factor 1: desire to smoke for pleasure and intention to smoke; e.g., “A cigarette would taste good.”) and aversive urge (Factor 2: desire to smoke to relieve negative affect; e.g., “Smoking would make me less depressed.”). Participants rate the extent to which they agreed with experiences based on how they felt “right now” on 6-point Likert scales (0=Strongly Disagree to 5=Strongly Agree). Each subscale is computed based on mean rating per item within each respective subscale (Factor 1 and Factor 2).

Wisconsin Smoking Withdrawal Scale (WSWS; Welsch et al., 1999).

The 28-item WSWS is a multidimensional measure that assesses the major symptom elements of the tobacco withdrawal syndrome. Participants are instructed to rate the extent to which they experienced tobacco withdrawal symptoms (e.g., “I have been tense or anxious.”) on 5-point Likert scales (0=Strongly Disagree to 4 = Strongly Agree) based on how they have felt “so far today,” which yields a composite index based on mean response across all items. The WSWS also contains 6 subscales representing severity of specific tobacco withdrawal symptoms (i.e., anxiety, anger, hunger, concentration problems, craving, and sadness) based on mean score per item response.

Smoking Reinstatement Task (McKee, 2009; McKee et al., 2006).

This behavioral-economics-based task assesses the relative reward value of initiating smoking and of continuing smoking once given the opportunity to smoke (McKee, 2009). At the onset of the task, participants are provided with 8 cigarettes of their preferred brand and an ashtray. In the “delay period,” participants are told they will be permitted to smoke any time during the next 50 min but for each 5 min that they delay smoking, they can earn $0.20, up to a maximum of $2.00. The delay period ends when participants decide to smoke or at the end of the 50 min. Immediately afterwards, participants enter a 60-minute “self-administration period,” where they are told they have a $1.60 credit from which they can pay $0.20 for each cigarette they wish to smoke. Thus, a total of $3.60 can be earned throughout the smoking lapse task. The selected $0.20 incentive has been sensitive to experimental manipulations in socioeconomically similar study populations to our own (Aguirre et al., 2015; Ameringer & Leventhal, 2015; Liautaud et al., 2018; Pang et al., 2018; Pang & Leventhal, 2013) and is based off piloting conducted in prior samples from our study population that showed adequate variance in lapse behaviors at the $0.20 (n=300) level, but not at the $1.00 (n=15; i.e., incentive validated by [McKee et al., 2006]) or $0.50 (n=15) levels. Following the self-administration period, participants enter a rest period for the remainder of the visit during which they are not allowed to smoke (rest time range: 60-110 min) to minimize the influence of participants’ impending ability to smoke following session completion on lapse behaviors during the task. The primary outcomes for this measure are latency to smoking initiation during the delay period in minutes (range: 0-50) and number of cigarettes purchased during self-administration period (range: 0-8).

Statistical Analysis

For each anhedonia measure, we computed descriptive statistics and intercorrelations among deprivation-induced change scores (i.e., score in deprived – nondeprived conditions; the primary deprivation effect outcome used in analyses). Given the overrepresentation of males in the sample (n=378; 62.2%), we also conducted paired sample t-tests for each anhedonia measure to determine whether deprivation-induced changes in anhedonia differed by sex.

To evaluate the effects of tobacco deprivation on anhedonia relative to that of other expressions of withdrawal (i.e., Study Objective 1), paired sample t-tests were run for each experimental measure to evaluate the difference in scores between deprived and non-deprived conditions. To investigate the extent to which different manifestations of anhedonia during tobacco withdrawal correlate with one another and with established tobacco withdrawal symptoms (i.e., Study Objective 2), Pearson correlations were calculated among deprivation-induced changes in each anhedonia measure and among deprivation-induced changes in other established tobacco measures. To identify potential risk factors for deprivation-induced changes in anhedonia (i.e., Study Objective 3), linear regression models were run to test whether nicotine dependence, affective psychopathology, and affective traits predicted the magnitude of deprivation-induced changes in generalized consummatory anhedonia (SHAPS). Separate models were tested for each dependence and affective measure; each model included the non-deprived SHAPS score as a covariate to adjust for baseline anhedonia.

With the exception of planned missingness for the STRAP-R and CES-D (see Study Sample below), missing data was fairly rare (see number of observations available in Supplemental Table 3) and was handled by list-wise deletion complete case analysis. Deprivation effects are reported as Cohen’s d effect size estimates and regression results are reported as standardized regression coefficients (β). Significance was set at p < .05 two-tailed. To correct for multiple tests, the Benjamini-Hochberg procedure was used to maintain a study-wise false discovery rate of .05 (Benjamini & Hochberg, 1995). Data were analyzed in January 2020 using IBM SPSS Version 24.

Results

Study Sample

The SC-TAPP sample was recruited over the course of four years from August 2013 to May 2017. Of the 776 participants who attended an in-person screen, 684 were deemed eligible for the SC-TAPP study, enrolled, and completed the baseline visit. Of baseline completers, 16 and 61 participants did not complete one or both experimental visits, respectively. Thus, the resulting analytic sample consisted of 607 individuals.

In July 2016 (i.e., 3 years into accrual for the SC-TAPP and 4 years prior to the conduct of data analyses for the current manuscript; n=463 participants completed), the STRAP-R and CES-D were among a broader selection of measures that were removed from the assessment battery at the study investigators’ discretion to accommodate inclusion of other measures while mitigating burden for the remaining 144 participants; these additional measures are not the focus of the secondary analysis reported here and are not described further. Thus, analyses that included the STRAP-R and CES-D measures utilized the subsample of 463, whereas analyses that include the SHAPS (and every other measure described in the current paper) included all 607 participants. Abstinence-induced changes in SHAPS scores did not differ between participants with versus without STRAP-R/CES-D data, t(605)=0.89, p=0.38.

The analytic sample (N=607) was 37.8% female, middle-aged (M[SD] = 50.0 [10.6] years old), predominantly low-income (i.e., total pre-tax household annual income < $15,000; 64.7%), mostly unemployed (50.8%) or retired or on disability (30.1%), and heterogenous in educational attainment (15.5% did not complete high school, 38.3% with high school degree, 33.3% some college, and 12.1% with college degree). Participants were, on average, moderate-to-heavy smokers (M[SD] cigs/day = 15.2 [7.4]) with medium levels of nicotine dependence (M[SD] FTCD-Total = 5.5 [2.0]) and chronic smoking histories (years of daily smoking, M[SD]=30.3[12.1]), with a majority endorsing regular use of menthol-flavored cigarettes (61.8%).

Effects of Tobacco Deprivation on Anhedonia and Other Expressions of Withdrawal

Tobacco deprivation increased generalized consummatory anhedonia on the SHAPS, generating a modest effect size estimate (Cohen’s d=.32; Table 1). Deprivation produced small, statistically significant increases in anticipatory and consummatory anhedonia in relation to sexual activity (ds=.09 to .12). Deprivation significantly reduced anhedonia, or increased reward responsiveness, directed toward drug rewards (ds=−.21 to −.19). No significant deprivation effects emerged for consummatory or anticipatory anhedonia in relation to food. Across all anhedonia outcomes evaluated, no significant differences emerged between males and females in the magnitude of deprivation-induced changes in anhedonia (Supplementary Table 4). In comparison to tobacco deprivation effects on anhedonia outcomes, deprivation effects on other tobacco withdrawal measures were generally more robust, producing a mean Cohen’s d deprivation effect of .54 (SD = .35; range = .11 to 1.63) across the 19 measures (Table 2).

Table 1.

Intercorrelations and Effect of Smoking Deprivation on Generalized and Domain-Specific Anhedonia Outcomes

	Deprived		Non-deprived		Deprivation-induced change effect		Intercorrelations of deprivation-induced change scoresa (r)
Anhedonia Outcomes	M	(SD)	M	(SD)	d	p	1.	2.	3.	4.	5.	6.
1. Generalized Consummatoryb	1.89	(0.53)	1.72	(0.47)	0.32	<.001
2. Food Consummatoryc	2.54	(1.28)	2.43	(1.28)	0.08	.07	.26
3. Food Anticipatoryc	3.06	(1.44)	3.02	(1.42)	0.03	.67	.13	.59
4. Sex Consummatoryc	2.67	(1.51)	2.49	(1.41)	0.12	.004	.27	.20	.11
5. Sex Anticipatoryc	3.22	(1.55)	3.08	(1.53)	0.09	.02	.15	.10	.19	.55
6. Drug Consummatoryc	3.76	(1.41)	4.02	(1.29)	−0.19	<.001	.06	.18	.16	.13	.02
7. Drug Anticipatoryc	4.05	(1.28)	4.31	(1.16)	−0.21	<.001	−.001	.17	.21	.08	.07	.71

^aDeprivation-Induced Change Score = Score in Deprived Condition – Score in Non-Deprived Condition.

^bSnaith Hamilton Pleasure Capacity Scale (higher score indicates higher anhedonia, range: 1 - 4), n=607.

^cSensitivity to Reinforcement of Addictive & Other Primary Rewards (higher score indicates higher anhedonia, range: 1-5), n=463.

Shaded values indicate anticipatory-consummatory associations within common reward domains. Bolded values indicate statistically significant correlations after applying the Benjamini-Hochberg correction for multiple tests to maintain a study-wise false discovery rate of .05.

Table 2.

Deprivation-Induced Changes in Other Tobacco Withdrawal Measures and Correlations with Deprivation-Induced Changes in Anhedonia

	Deprivation Effect	Correlations between deprivation-induced changea in anhedonia with other tobacco withdrawal measures, Pearson’s r
Other Tobacco Withdrawal Measures	Cohen’s d	Generalized (Cons)b	Food (Cons)c	Food (Ant)c	Sex (Cons)c	Sex (Ant)c	Drug (Cons)c	Drug (Ant)b
Behavioral Smoking Task
Latency to smoke	−0.83	−.10	−.08	.001	−.03	.03	−.02	.01
Cigarettes smoked	0.22	.05	−.02	−.01	.09	.01	.02	.01
Smoking Urges
Craving (W)	0.94	.12	.05	.01	.08	.03	.01	−.003
Appetitive Urge (Q)	1.63	.02	−.02	.003	.04	.01	−.04	−.04
Aversive Urge (Q)	0.97	.17	.04	.02	.18	.12	−.03	−.06
Affect States
Negative Mood (P)	0.25	.32	.06	−.04	.20	.15	−.05	−.06
Anger (W)	0.60	.21	.08	−.06	.11	.05	.03	.02
Anger (P)	0.28	.34	.05	−.05	.21	.18	−.07	−.07
Anxiety (W)	0.53	.26	.04	−.04	.14	.05	−.05	−.05
Anxiety (P)	0.41	.31	.07	−.08	.20	.15	−.05	−.07
Sadness (W)	0.44	.27	.09	−.03	.14	.08	−.13	−.13
Depression (P)	0.11	.23	.05	−.06	.15	.11	−.03	−.03
Positive Mood (P)	−0.49	−.37	−.20	−.03	−.22	−.08	−.02	.04
Vigor (P)	−0.40	−.30	−.17	−.04	−.19	−.05	.003	.02
Friendliness (P)	−0.53	−.36	−.19	−.05	−.21	−.08	−.05	.01
Elation (P)	−0.46	−.35	−.20	.003	−.21	−.08	−.01	.07
Confusion (P)	0.23	.19	.03	−.02	.13	.14	−.02	−.05
Other Symptoms
Concentration Problems (W)	0.54	.32	.08	−.02	.23	.14	−.02	−.06
Hunger (W)	0.41	.02	−.25	−.26	.06	.04	.06	.03
Mean (SD) across 19 measures,d ∣d∣ or ∣r∣	0.54 (0.35)	.23 (.12)	.09 (.07)	.04 (.06)	.15 (.06)	.08 (.05)	.04 (.04)	.04 (.03)

Note. Sample size varies across analyses due to missing data (see Supplemental Table 1). Cons = Consummatory; Ant = Anticipatory; Q = subscale on the Questionnaire of Smoking Urges; W = subscale on the Wisconsin Smoking Withdrawal Scale; P = subscale on the Profile of Mood States.

^aDeprivation-Induced Change Score = Score in Deprived Condition – Score in Non-Deprived Condition.

^bSnaith Hamilton Pleasure Capacity Scale (higher score indicates higher anhedonia, range: 1 - 4), n=607.

^cSensitivity to Reinforcement of Addictive & Other Primary Rewards (higher score indicates greater anhedonia, range: 1-5), n=463.

^dBased on distribution of absolute value of Cohen’s d or Pearson’s r across all 19 measures.

Bolded values indicate statistical significance after applying the Benjamini-Hochberg correction for multiple tests to maintain a study-wise false discovery rate of .05.

Intercorrelations Among Deprivation-Induced Changes in Consummatory and Anticipatory Anhedonia and Other Tobacco Withdrawal Outcomes

Deprivation-induced changes in anticipatory and consummatory anhedonia toward common reward domains (e.g., anticipatory toward food with consummatory toward food) were strongly correlated with one another (rs = .55 to .71; Table 1). Conversely, correlations among deprivation-induced changes in anhedonia that cut across different reward domains (e.g., anticipatory toward sex with anticipatory toward food) were relatively small in magnitude and in some cases not statistically significant (rs <.21). Deprivation-induced changes in generalized consummatory anhedonia on the SHAPS were modestly correlated with anticipatory and consummatory anhedonia directed toward food and sex (rs = .15 to .27) but not significantly correlated with pleasure response to drug/alcohol reward (rs ≤ .06; Table 1).

Deprivation-induced changes in generalized consummatory anhedonia were significantly correlated in the expected direction with deprivation-induced increases in most other withdrawal outcomes (i.e., positive associations with craving and unpleasant symptoms and inverse associations with positive affect and latency to smoking reinstatement on the behavioral smoking task), producing a mean ∣r∣ of .23 (SD = .12; range = .02 to .37) across the 19 measures (Table 2). Correlations of anhedonia toward specific reward domains and other tobacco withdrawal measures were generally of small magnitude (rs = −.26 to .23) and often non-significant. The sole exception was deprivation-induced changes in consummatory anhedonia toward sex, which were consistently correlated with other tobacco withdrawal expressions in the expected direction, producing significant correlations with 14 of the measures and a mean ∣r∣ of .15 (SD = .06; range = .03 to .22) across all 19 measures (Table 2). Other than those cases, deprivation-induced changes in consummatory anhedonia toward food were associated with reductions in all 4 positive affect measures, and deprivation-induced increases in anticipatory anhedonia for sex were associated with increases in 3 of the 6 negative affect measures and with urge to smoke to alleviate negative affect. No other statistically significant associations were observed between STRAP-R items and other tobacco withdrawal symptoms (Table 2).

Associations Between Affective and Smoking Characteristics with Deprivation-Induced Changes in Generalized Consummatory Anhedonia

As shown in Table 3, higher baseline levels of subjective happiness, CES-D positive affect, and IDAS Well-Being were all inversely associated with deprivation-induced changes in generalized consummatory anhedonia (i.e., higher happiness and well-being were protective against anhedonia during tobacco deprivation; βs = −.12 to −.09; ps <.04). Additionally, higher levels of tobacco dependence severity (FTCD-Total) and shorter latency to first cigarette upon waking were positively associated with deprivation-induced changes in generalized consummatory anhedonia (βs =.10 to .12; ps < .01; Table 3). No significant associations were found between negative affect and deprivation-induced changes in anhedonia, nor between other psychopathologic expressions of distress (i.e., IDAS subscales with deprivation-induced changes in anhedonia [Table 3]).

Table 3.

Associations of Psychological and Smoking Characteristics with Deprivation-Induced Changes in Generalized Consummatory Anhedonia

	Descriptive Statistics	Association
Predictor Variables	M (SD) or N (%)	β	p
FTCD-Total, M(SD)	5.54 (1.97)	.10	.01
FTCD-Time-to-First-Cigarette, M(SD)	2.30 (0.79)	.12	<.001
Within 5 minutes, N(%)	282 (46.5%)
6 to 30 minutes, N(%)	251 (41.4%)
31 to 60 minutes, N(%)	48 (7.9%)
After 60 minutes, N(%)	25 (4.1%)
Subjective Happiness Scale, M(SD)	5.36 (1.24)	−.11	.005
CES-D, M(SD)	0.71 (0.49)
Positive Affect	2.06 (0.80)	−.09	.04
Negative Affect	0.56 (0.63)	−.02	.67
Somatic Features	0.78 (0.60)	−.04	.35
Interpersonal Problems	0.54 (0.74)	−.06	.16
IDAS, M(SD)	1.91 (0.51)
Dysphoria	1.79 (0.77)	−.01	.79
Lassitude	1.75 (0.67)	.001	.99
Insomnia	2.07 (0.95)	−.03	.46
Appetite Loss	1.73 (0.84)	−.02	.65
Appetite Gain	2.18 (0.95)	−.07	.08
Ill Temperament	1.54 (0.75)	.02	.65
Well Being	3.26 (0.87)	−.12	.002
Social Anxiety	1.56 (0.74)	.01	.86
Traumatic Intrusions	1.49 (0.73)	−.02	.68
Panic	1.34 (0.55)	.001	.97

Note. Sample size varies across analyses due to missing data (Ns=460-607; See Supplementary Table 1). All regressor variables were tested independently in separate models while controlling for respective Non-Deprived Scores. Outcome: Snaith Hamilton Pleasure Capacity Scale (i.e., generalized consummatory anhedonia; range 4 = Strongly Disagree to 1 = Strongly Agree). Deprivation-Induced Change Score = Score in Deprived Condition – Score in Non-Deprived Condition; FTCD = Fagerström Test for Cigarette Dependence (range 0 to 10); Time to first cigarette coded as Within 5 minutes (=3), 6 to 30 minutes (=2), 31to 60 minutes (=1), after 60 minutes (=0);CES-D = Center for Epidemiologic Studies-Depression Scale (each subscale range 0 to 3); IDAS = Inventory of Depressive and Anxiety Symptomatology (each subscale range 1 to 5); Subjective Happiness Scale (range 1 to 7).

Bolded values indicate statistical significance after applying Benjamini-Hochberg correction for multiple tests to maintain study-wise false discovery rate of .05.

Discussion

The current study provides a comprehensive, multidimensional characterization of the subjective manifestations of anhedonia during tobacco deprivation. Moreover, it extends this literature to AA smokers—a population facing extensive tobacco-related health disparities (Henley et al., 2016; Simmons et al., 2016). The main finding was that tobacco deprivation produced generalized consummatory anhedonia toward various interests/hobbies, social activities, and sensory experiences, supporting anhedonia as a valid element of the tobacco withdrawal syndrome in AA smokers. By contrast, measures of perceived reactions toward specific primary reward domains of food, sex, and alcohol/drugs mostly did not produce results concordant with tobacco withdrawal (though sex-related anhedonia was an exception to some extent).

Several findings here suggest that generalized consummatory anhedonia may be a valid and unique phenotypic expression of the acute tobacco withdrawal syndrome. First, paralleling previous work (Cook et al., 2015; Hughes et al., 2017), the magnitude of deprivation-induced increases in generalized consummatory anhedonia observed in the current study were larger than or similar to that of several other moderate-severity tobacco withdrawal symptoms (i.e., number of cigarettes smoked, confusion, hunger, certain mood states). However, deprivation-induced changes in generalized consummatory anhedonia were considerably smaller compared to relatively higher severity tobacco withdrawal symptoms, such as cravings, latency to smoke, appetitive and aversive smoking urge. This may partially reflect inherent challenges and shortcomings that accompany the set of self-report measures of anhedonia used in the current study (see Limitations for additional details). Notably, deprivation-induced increases in generalized consummatory anhedonia were modestly associated with a shorter latency to smoking reinstatement on the behavioral smoking task. This coheres with prior work demonstrating that deprivation-induced anhedonia experienced during and after a quit attempt is associated with shorter latency to smoking relapse and lower 8-week abstinence rates (Cook et al., 2015; Piper et al., 2017).

Additional support for anhedonia as a valid tobacco withdrawal symptom can be inferred from the nature of the correlations between deprivation-induced changes in generalized consummatory anhedonia and deprivation-induced changes in other withdrawal symptoms. Correlations were generally significant and in the expected direction, which aligns with results from a previous study of predominately White smokers making a quit attempt (Cook et al., 2017). Most of these correlations were significant yet moderate in magnitude (∣r∣s=.10-.37), rather than null or large. This suggest that generalized consummatory anhedonia represents a distinct symptom domain of a common tobacco withdrawal syndrome that is non-redundant with other tobacco withdrawal symptoms, including deprivation-induced changes in constructs that are functionally related to anhedonia (i.e., depression, positive and negative mood states [∣r∣s=.11-.60]). In other words, anhedonia does not appear to merely be an alternate indicator of an existing symptom construct that is already tapped in withdrawal assessment.

Finally, results from the predictor analysis of risk factors for withdrawal-related anhedonia are also consistent with study hypotheses. In concordance with results from a prior clinical study (Cook et al., 2015) and the fundamental conceptualization that drug withdrawal is a core component of the drug dependence syndrome (Wikler, 1971), tobacco dependence severity predicted greater deprivation-induced increases in generalized consummatory anhedonia. Behavioral traits and emotional states that were pathognomonically-concordant with diminished pleasure and reward (i.e., low happiness, well-being, and positive affect) were also associated with greater severity of generalized consummatory anhedonia during acute tobacco deprivation, while emotional symptoms and traits related to negative affect, cognitive features, and somatic signs were not. This is consistent with both the STAR model of nicotine dependence (Gilbert, 1997) and other established withdrawal symptoms that show pathognomonic concordance with specific psychopathologic states (e.g., tobacco-deprived smokers with depressed mood, eating disorder symptoms, and ADHD tend to experience relatively greater deprivation-induced increases in sadness, hunger, and attention problems, respectively; Leventhal et al., 2013; McClernon et al., 2008; Pomerleau et al., 2000). These results shed mechanistic insight into the role of deprivation-induced anhedonia in the smoking maintenance of AA smokers with severe tobacco dependence and/or deficits in positive emotional functioning. Our findings also point toward the possibility of personalized treatment approaches for cessation, whereby patients with a psychopathological profile indicative of diminished positive affect might benefit from clinical attention to anhedonia during the cessation process. Even so, translational work of this nature is contingent on the empirical extension of the current findings to clinical settings and other racial/ethnic groups.

Although similar findings have been observed in mixed race samples (Cook et al., 2015), it is unknown whether the current findings generalize to racial/ethnic groups beyond AA smokers. Incidentally, the effect sizes that we observed for withdrawal-related changes in positive and negative affect (d=−.26-.49) were much closer in magnitude than in prior research with predominately non-AA samples (e.g., ds=.34-.97; Leventhal et al., 2010). This may be a consequence of racial differences in emotion socialization, whereby AAs are socialized to downregulate expression of negative emotions and attend to positive emotions relative to Whites (Labella, 2018). In part, this may reflect the relatively graver sociocultural consequences that negative emotional expression has for AAs in the context of chronic oppression and discrimination (Nelson et al., 2012; Odom et al., 2016). Prior laboratory work demonstrating relatively larger reductions in positive affect following tobacco deprivation among AA (vs. White) smokers (Bello et al., 2015) suggests that this socialized tendency may generalize to the expression of withdrawal-related affective disturbance. Given that anhedonia and positive affect are both expressions of affective disturbance that are not distress-related, it is possible that sociocultural differences in emotion socialization (Labella, 2018) or genetic variation in reward sensitivity (Zhao et al., 2012) may render AA smokers particularly vulnerable to these types of reward-related withdrawal expressions.

Domain-specific anhedonia measures largely did not conform to criteria for tobacco withdrawal phenomena. This may indicate that anhedonic responses during acute tobacco withdrawal do not generalize across all types of rewards, particularly to food- and drug-related rewards. Although prior human research studies have shown that tobacco deprivation increases the desire to consume sweet and carbohydrate-rich foods (Ogden, 1994; Perkins et al., 1995; Rodin, 1987; Spring et al., 2003), we did not observe any significant differences in reward responsiveness toward food between deprived vs. non-deprived conditions. This parallels previous human research demonstrating significant tobacco deprivation effects on the relative reward value of cigarettes but not food (Betts & Tiffany, 2019; Lawn et al., 2015). Given that acute nicotine administration increases the rate of metabolic processing (i.e., energy expenditure) throughout the body (Perkins, 1992), hunger is a common tobacco withdrawal symptom (Hughes, 1992, 2007; Hughes & Hatsukami, 1986). Hence, it is possible that withdrawal-related increases in hunger during tobacco deprivation may produce a corresponding motivation to eat that counterpoises any competing deprivation-induced reward deficits directed toward food. This explanation is consistent with the current finding that deprivation-induced changes in hunger were inversely correlated with deprivation-induced changes in food-related anhedonia.

Given modern understanding of the nature of anhedonia as being expressed toward non-pharmacological rewards only (Leventhal et al., 2014), we hypothesized that deprivation would not have an effect on responsiveness to drug-related rewards. Thus, the observed increase in the desire and perceived hedonic response to drug-related rewards as a function of tobacco deprivation in this study was not expected. Even so, our finding that tobacco deprivation may actually decrease anhedonic responding to drug-related reward potentially parallels some evidence that drugs can act as substitute reinforcers for one another, whereby reductions in the opportunity to use one drug are sometimes compensated by increased reinforcing value of other drugs (Sussman & Black, 2008). Understanding the influence that increased reward responsiveness to psychoactive drugs during smoking cessation may have on the use of other substances is an avenue for future research that merits additional attention. With respect to anhedonic responses toward sex, only consummatory, and to a lesser extent anticipatory, anhedonic responses, were modestly increased by tobacco deprivation and correlated with other withdrawal symptoms. Even though the results for sex-related anhedonia were not robust or entirely consistent, this topic merits further research given the biological plausibility that sex-related anhedonia might be a manifestation of nicotine withdrawal and the lack of research in this area.

Limitations

The findings of the current study should be considered in the context of its limitations. First, the reverse-scored SHAPS was the only measure of generalized anhedonia used, and the SHAPS has traditionally been conceptualized as a self-report measure of consummatory pleasure (Rizvi et al., 2016; Snaith et al., 1995). However, some studies have shown that the SHAPS tends to correlate about as well with self-reported anticipatory pleasure as it does with another measure of self-reported consummatory pleasure (i.e., Temporal Experience of Pleasure Scale; Geaney et al., 2015; Liu et al., 2012; Loas & Godefroy, 2014). Second, the anhedonia scales used in the current study (i.e., SHAPS, STRAP-R) may be subject to psychometric limitations. Because consummatory items assess hypothetical responses to a reward that is not immediately available, responses may be confounded if anticipatory reward processing is inadvertently activated. Because anticipatory items assess desire experienced “right now” to engage in sex-, drug/alcohol-, and food-related activities, responses may be confounded by a host of contextual considerations (e.g., social acceptability in current setting, time of day). This may explain why the tobacco deprivation effects we observed for anticipatory and consummatory responses to specific reward domains were so small. Alternatively, single- vs. multi-item measures (i.e., STRAP-R vs. SHAPS) may simply be less effective at tapping the underlying anhedonic construct due to their relatively poorer psychometric properties.

Another set of limitations relate to the nature of the study sample, which only included non-treatment seeking AA smokers who underwent a brief period of externally imposed deprivation that was not self-motivated. Thus, our findings may not generalize to smokers whose smoking deprivation is part of an intrinsically initiated quit attempt, though evidence suggests that tobacco withdrawal during experimentally imposed deprivation is predictive of withdrawal following a naturalistic self-initiated quit attempt (Strong et al., 2011). In addition, although withdrawal symptom severity and withdrawal-induced relapse tends to occur most often during the first 48 hours of a cessation attempt (Shiffman, 1982), our results may not generalize to withdrawal-related anhedonia experienced during periods of deprivation that exceed 16 hours. Finally, it remains unclear whether the clinical manifestations of anhedonia during withdrawal observed in the current study would be more or less robust in other racial or ethnic groups. Future work examining anhedonia as a component of the tobacco withdrawal syndrome could build upon the present findings by using samples that counter these limitations, incorporating multiple, separate self-report measures of generalized anhedonia to better capture its multi-faceted nature (i.e., state vs. trait; anticipatory vs. consummatory), and by employing multi-method protocols that integrate subjective assessments with neurobiological, cognitive, and behavioral indices of anhedonia.

A final limitation relates to our inability to demonstrate that anhedonia-related withdrawal offers incremental predictive value over existing withdrawal symptoms that are similar in nature (i.e., depression, low positive affect). Although the design of this study did not permit an evaluation of this kind, future work could advance an understanding of the clinical utility of assessing withdrawal-related anhedonia by evaluating the relative explanatory power that deprivation-induced changes in anhedonia, positive and negative affect, and poor mood have in predicting clinically meaningful endpoints (e.g., odds of smoking relapse during cessation).

Implications

The principle finding of the current study—that generalized consummatory anhedonia increased as a function of acute tobacco deprivation among African American smokers—has the potential to advance tobacco-related health disparities research and clinical practice in several meaningful ways. Several clinical interventions (e.g., behavioral activation therapy, positive mental imagery training) have demonstrated promise in reducing symptoms of anhedonia (Blackwell et al., 2015; Carl et al., 2016; Pictet et al., 2016; Williams et al., 2015) and potentially increasing smoking cessation success (MacPherson et al., 2010). There is also some evidence that pharmacologic interventions, such as bupropion and nicotine replacement, improve anhedonia during tobacco deprivation (Cook et al., 2017; Guillot et al., 2015; Paterson et al., 2007), although whether anhedonia alleviation is a mechanism of efficacy for these medications merit further inquiry.

In furtherance of the burgeoning base of literature in support of anhedonia as a novel component of the tobacco withdrawal syndrome (Cook et al., 2017; Cook et al., 2015; Hughes et al., 2017; Piper et al., 2017), our findings suggest that deprivation-induced anhedonia may be implicated in the smoking patterns of AA smokers, who are considerably more likely to relapse following a quit attempt (Kulak et al., 2016; Pagano et al., 2018; Weinberger et al., 2019). Given the motivational salience that deprivation-induced generalized consummatory may have on smoking lapse and relapse among AA smokers, the current findings provide empirical justification for incorporating anhedonia into standard tobacco withdrawal assessment and targeting anhedonia in smoking cessation behavioral interventions used with AA smokers. Future research on the intersection of anhedonia and tobacco withdrawal may be fruitful for advancing the science and treatment of tobacco addiction and efforts to reduce tobacco-related health disparities facing AA smokers.

Public Significance Statement:

This study found that generalized consummatory anhedonia (i.e., diminished pleasure response toward rewarding experiences) increased as a function of acute tobacco deprivation among African American (AA) smokers. These findings support the conceptualization of anhedonia as a genuine component of the tobacco withdrawal syndrome that warrants being incorporated into standard tobacco withdrawal assessment and targeted in smoking cessation behavioral interventions in AA smokers.