Pain Intensity, E-Cigarette Dependence, and Cessation-Related Outcomes: The Moderating Role of Pain-Related Anxiety

Abstract

Pain and nicotine dependence are prevalent, co-occurring conditions posited to interact in the manner of a positive feedback loop; however, most research to date has been conducted among tobacco cigarette smokers. Initial evidence suggests that pain is a risk factor for greater e-cigarette dependence, and additional research is needed to examine covariation between pain and e-cigarette use. There is reason to suspect that pain-related anxiety (i.e., the tendency to respond to pain with anxiety or fear) may be associated with greater e-cigarette dependence and difficulty quitting, and that pain intensity and pain-related anxiety may interact to confer greater risk for e-cigarette use. The current study represents the first examination of cross-sectional associations between pain intensity, pain-related anxiety, and e-cigarette dependence, motivation to quit, history of lifetime e-cigarette quit attempts, perceived barriers to cessation, and negative expectancies during abstinence from e-cigarettes. Participants (N = 520 e-cigarette users, 52.1% female, M_age = 34.85) completed an online survey assessing health behaviors. Results indicated that pain-related anxiety was positively associated with e-cigarette dependence and perceived barriers to cessation (ps < .05). Pain-related anxiety was found to moderate relations between pain intensity and primary outcomes, such that pain intensity was positively associated with motivation to quit, likelihood of past failed quit attempt, and negative abstinence expectancies among participants who endorsed high (but not moderate or low) levels of pain-related anxiety. Future research would benefit from examining prospective associations between pain-related anxiety, pain intensity, and e-cigarette use/cessation trajectories among individuals with chronic pain.

1. Introduction

Pain and nicotine/tobacco use are highly prevalent and frequently co-occurring conditions that engender a substantial economic toll in the United States each year (National Institute on Drug Abuse, 2020; Ma, Chan, & Carruthers, 2014; Orhurhu, Pittelkow, & Hooten, 2015; Phillips, 2009; Zvolensky, McMillan, Gonzalez, & Asmundson, 2009). Rates of tobacco cigarette smoking are considerably higher among persons with pain (~24–68%; e.g., Michna et al., 2004; Orhurhu et al., 2015; Zvolensky et al., 2009), relative to the general population (~14%; Clarke, Norris, & Schiller, 2019). Smokers with pain (vs. without pain) are more likely to use electronic cigarettes (e-cigarettes; Powers, Heckman, LaRowe, & Ditre, 2019), and greater pain-related interference has been associated with a 45% increased likelihood of meeting criteria for nicotine use disorder (McDermott, Joyner, Hakes, Okey, & Cougle, 2018). An established reciprocal model posits that pain and nicotine use interact in the manner of a positive feedback loop, leading to progression and maintenance of both conditions over time (Ditre, Brandon, Zale, & Meagher, 2011; Ditre, Zale, & LaRowe, 2019).

Consistent with this framework, greater lifetime nicotine/tobacco exposure has been associated with more severe pain, heightened sensitivity to experimental pain, and an increased risk for developing persistent pain (De Vita, Maisto, Ansell, Zale, & Ditre, 2019; Pisinger et al., 2011; Sugiyama et al., 2010). Indeed, nicotine use may lead to dysregulated pain processing and poorer outcomes via both nicotine specific (e.g., increased serum proteolytic enzyme activity; Aamodt, Stovner, Hagen, Bråthen, & Zwart, 2006; Fogelholm & Alho, 2001) and general neurobiological effects (e.g., allostatic load on overlapping pain, stress, and reward neurocircuitry; Elman & Borsook, 2016). Nicotine has also been shown to confer acute analgesia (Ditre, Heckman, Zale, Kosiba, & Maisto, 2016), and converging evidence indicates that pain can be a powerful motivator of nicotine use (Dhingra et al., 2014; Ditre & Brandon, 2008; Ditre, Heckman, Butts, & Brandon, 2010; Kosiba, Zale, & Ditre, 2018). Smokers often report using cigarettes to cope with pain (e.g., Jamison, Stetson, & Parris, 1991; Patterson et al., 2012), and expectancies for pain-coping via smoking have been shown to predict lower odds of achieving bio-verified smoking abstinence at six month follow-up (Endrighi, Rueras, Dunsiger, & Borrelli, 2019).

Despite well-documented interrelations between pain and nicotine use, most research to date has focused on tobacco cigarette smoking. The uptake of e-cigarettes nationwide is unprecedented (McMillen, Gottlieb, Shaefer, Winickoff, & Klein, 2014), with sales of e-cigarettes increasing over 640% from 2016–2017 alone (King, Gammon, Marynak, & Rogers, 2018), and a critical next step is to examine the role of pain in e-cigarette use. Initial work has indicated that pain may be associated with poorer e-cigarette-related outcomes. For example, a recent study of 322 e-cigarette users/dual users of cigarettes and e-cigarettes found that those who reported more severe pain were also more dependent on e-cigarettes and perceived a greater number of barriers to quitting (Zvolensky et al., 2018). Thus, individuals with pain may be at-risk for poorer e-cigarette use outcomes.

In addition to further elucidating relations between pain and use of e-cigarettes, it is also important to identify malleable cognitive-affective factors that offer potential to inform the development of tailored interventions. A recent review posited that pain-related anxiety (i.e., the tendency to respond to pain with anxiety or fear) is an important transdiagnostic factor underlying co-occurring pain and cigarette smoking (LaRowe, Zvolensky, & Ditre, 2018). Pain-related anxiety has been identified as a risk factor in the transition from acute to chronic pain (Costal, Maherl, McAuleyl, Hancockl, & Smeetsl, 2011), and has been positively associated with pain intensity and the use of maladaptive pain-coping strategies (McCracken, Gross, Sorg, & Edmands, 1993). Higher levels of pain-related anxiety have also been associated with smoking dependence motives (Ditre, Langdon, Kosiba, Zale, & Zvolensky, 2015; Ditre, Zale, Kosiba, & Zvolensky, 2013), the use of tobacco to cope with pain (Patterson et al., 2012), expectations that smoking can alleviate negative mood (Gonzalez, Hogan, McLeish, & Zvolensky, 2010), and perceived barriers to cigarette cessation (Ditre, Langdon, Kosiba, Zale, & Zvolensky, 2015). Most recently, pain-related anxiety was shown to predict early lapse and relapse among smokers participating in a self-guided quit attempt (LaRowe, Langdon, Zvolensky, Zale, & Ditre, 2017). Despite a growing empirical literature linking pain-related anxiety to cigarette smoking, we are not aware of any work that has examined this factor in the context of e-cigarette use.

Drawing on the theoretical and empirical literature linking pain, pain-related anxiety, and tobacco cigarette smoking (Ditre et al., 2011; Ditre et al., 2019; LaRowe et al., 2018), we hypothesize that pain intensity and pain-related anxiety interact to confer greater risk for e-cigarette use. Pain is a potent motivator of nicotine use (Dhingra et al., 2014; Ditre & Brandon, 2008; Ditre et al., 2010; Kosiba et al., 2018), and persons who experience more intense pain may be more likely to use e-cigarettes to cope with pain symptoms (Ditre & Brandon, 2008; Ditre et al., 2016; Khantzian, 1997). It is also possible that associations between pain intensity and e-cigarette use are stronger among individuals who are more fearful/anxious about pain. Previous work has shown that pain-related anxiety can increase propensity to engage in escape/avoidance behaviors in response to pain (Crombez, Vlaeyen, Heuts, & Lysens, 1999), including substance use (e.g.,McCracken, Zayfert, & Gross, 1992). Pain-related anxiety may also amplify the actual and anticipated analgesic effects of nicotine (LaRowe et al., 2018). Thus, persons with elevated pain-related anxiety may derive greater negative reinforcement from e-cigarette use (LaRowe et al., 2018), contributing to increased levels of dependence and greater difficulties quitting.

The current study tested whether pain intensity and pain-related anxiety conferred greater e-cigarette dependence and poorer cessation-related outcomes among a sample of adult e-cigarette users. Specifically, we hypothesized that pain intensity would be positively associated with e-cigarette dependence, perceived barriers to quitting e-cigarettes, and negative abstinence expectancies among participants with higher (vs. lower) levels of pain-related anxiety. We also hypothesized that participants with elevated pain intensity and pain-related anxiety would report less motivation to quit, lower likelihood of attempting to quit e-cigarettes in the past, and fewer number of lifetime quit attempts.

2. Methods

2.1. Participants

Data from the present study are from an online survey examining e-cigarette use behaviors among adult e-cigarette users. Participants were recruited nationally via Qualtrics Inc, an online survey management system that has been found to yield valid and reliable representative data among e-cigarette and substance use populations (Manning, Garey, Mayorga, Nizio, & Zvolensky, 2020; Zvolensky, Shepherd, Garey, Case, & Gallagher, 2020). Qualtrics aims to generate a representative sample based upon the most recent United States Census (Schulte & Gearhardt, 2018). Individuals with a Qualtrics Panels account were invited to complete the survey that was advertised using only the survey name listed in English. Interested individuals (N = 1007) were screened for eligibility; 224 were found ineligible and 147 provided substantial missing data that invalidated their data. Eligible participants were redirected to the survey. Prior to starting the survey, participants provided informed consent. Eligibility criteria included being 18–65 years old, reporting use of an e-cigarette in the last 30 days, and ability to speak English and provide informed consent. The survey took approximately 30 minutes to complete, and participants were compensated with credit through their Qualtrics account commensurate to their participation. Specifically, each participant was given the option to choose their preferred form of compensation based off their credit, however, the total amount for completing the survey remained the same ($8.50). To ensure the validity of the data, IP addresses were obtained to prohibit duplicate responses, 116 responses were omitted based on constant responding, short completion time, nonsensical responses (e.g., responded with gibberish to open text responses), and responding to the quality question (i.e., “We care about the quality of our data. In order for us to get the most accurate measures of your opinions, it is important that you thoughtfully provide your best answers to each question in this survey. Do you commit to thoughtfully providing your best?”) with something other than “I commit to giving my best answers.” The study protocol was approved by the Institutional Review Board at the University of Houston.

2.2. Measures

2.2.1. Pain Intensity.

Past three-month pain was assessed using the Graded Chronic Pain Scale (Von Korff, Ormel, Keefe, & Dworkin, 1992), which is commonly used to assess pain severity among both clinical and nonclinical samples (Turk & Melzack, 2011). A Characteristic Pain Intensity (CPI) score can be computed by summing responses to three questions asking participants to rate current pain and average/worst pain over the past three months on an 11-point scale (0 = no pain to 10 = pain as bad as it could be). Total scores range from 0–30 (α = .82).

2.2.2. Pain-Related Anxiety.

Pain-related anxiety was assessed using the Pain Anxiety Symptoms Scale −20 (PASS-20, McCracken & Dhingra, 2002). The PASS-20 includes 20 items (range: 0–100) that assess how often participants engage in various thoughts (e.g., “When I feel pain, I am afraid that something terrible will happen”) and behaviors (e.g., “I avoid important activities when I hurt”) using a 6-point Likert scale (0 = never to 5 = always). The PASS-20 demonstrated excellent internal consistency in the current sample (α = .96).

2.2.3. Motivation to Quit E-Cigarettes and Lifetime Quit Attempts.

A single item was used to assess current motivation to quit e-cigarettes. On a 11-point Likert scale ranging from 0 (Not at all motivated) to 10 (Extremely motivated), participants were asked to “Please rate how motivated you are to quit e-cigarettes.” The use of a single motivation to quit item has consistently been utilized in the combustible cigarette literature (Kotz, Brown, & West, 2013; Sciamanna, Hoch, Duke, Fogle, & Ford, 2000). For lifetime quit attempts, participants first indicated (yes/no) whether they had ever tried to quit e-cigarettes. Those who endorsed a lifetime quit attempt were then asked to indicate the total number of times they have tried to quit using e-cigarettes in their lifetime.

2.2.4. E-Cigarette Barriers to Cessation.

Perceived barriers to quitting e-cigarettes were assessed via The E-Cigarette Barriers for Cessation Scale (E-BCS). The E-BCS is a 19-item questionnaire that was modified from the original Barriers of Cessation Scale (BCS; Garey et al., 2017; Macnee & Talsma, 1995). The E-BCS assesses perceived barriers to or stressors resulting from using e-cigarettes (e.g., “No encouragement or help from friends”). Responses are scored on a 4-point Likert scale with items ranging from 0 (Not a barrier) to 3 (A large barrier). The E-BCS total score was used for the present study and demonstrated excellent internal consistency (α = .96)

2.2.5. E-Cigarette Abstinence Expectancies.

A modified version of the Smoking Abstinence Expectancies Questionnaire (Abrams, Zvolensky, Dorman, Gonzalez, & Mayer, 2011) was administered to assess abstinence expectancies for e-cigarettes. The E-Cigarette Abstinence Expectancies Questionnaire (EAEQ) included 22 items, and responses are scored on a 11-point Likert scale with items ranging from 0 (Not likely) to 10 (Likely). Higher total scores on the EAEQ indicate greater negative expectancies from abstaining from e-cigarettes. The EAEQ also demonstrated excellent internal consistency (α = .97).

2.2.6. Anxiety and Depression.

The Overall Anxiety Severity and Impairment Scale (OASIS) was used to assess severity of anxiety symptoms (Norman, Hami Cissell, Means-Christensen, & Stein, 2006). The OASIS is brief, five-item continuous measure that demonstrated excellent internal consistency (α = .90). Depression was assessed via the Overall Depression Severity and Impairment Scale (ODSIS; Bentley, Gallagher, Carl, & Barlow, 2014). The ODSIS includes five items and generates a continuous total score of severity and impairment of depressive symptomatology. Examination of Cronbach’s alpha for the ODSIS revealed excellent internal consistency (α = .93).

2.2.7. Sociodemographic and E-Cigarette Use Characteristics.

A range of sociodemographic characteristics were assessed via self-report (e.g., age, gender, race/ethnicity, education, marital status, and income). Participants also reported daily use of e-cigarettes and cigarettes. E-cigarette dependence was assessed using the Penn State Electronic Cigarette Dependence Index (Foulds et al., 2015). This measure includes 10 items that were modified from a measure of cigarette dependence (Penn State Cigarette Dependence Index; Foulds, Veldheer, & Berg, 2011). Total scores range from 0–20, with higher scores indicating greater levels of dependence.

2.3. Data Analysis

All analyses were conducted using SPSS Version 24 (IBM Corp., 2016). First, distributions for all variables were assessed for normality, and skewness and kurtosis fell within acceptable ranges (George & Mallery, 2003). Examination of Variance Inflation Factor (VIF) indicated no issues with multicollinearity between pain intensity and pain-related anxiety (VIF = 1; Craney & Surles, 2002). Second, we conducted separate hierarchical linear regression models to test the interaction between pain intensity and pain-related anxiety on: (1) e-cigarette dependence, (2) motivation to quit e-cigarettes, (3) number of lifetime quit attempts, (4) perceived barriers to e-cigarette cessation, and (5) e-cigarette abstinence expectancies. All models met assumptions for normality of residuals and homoscedasticity (Jarque & Bera, 1987; Jarque & Bera, 1980). We also conducted a hierarchical logistic regression model to test the interaction between pain intensity and pain-related anxiety on the likelihood of reporting a lifetime e-cigarette quit attempt. Given previously observed associations with e-cigarette use (Kasza et al., 2017; Stanton & Halenar, 2018), gender, race, age, income, and daily cigarette use were included as covariates in each model. E-cigarette dependence was also included as a covariate for models examining motivation to quit, perceived barriers to quitting, and abstinence expectancies. Adjusted models also covaried for depression and anxiety, as these variables have been associated with both pain and pain-related anxiety (Carleton, Abrams, Asmundson, Antony, & McCabe, 2009; Woo, 2010). For each model, predictors were entered in the following order: Step 1 (covariates); Step 2 (pain intensity, pain-related anxiety); Step 3 (pain intensity x pain-related anxiety interaction). We interpreted the pain intensity x pain-related anxiety interaction by examining results at Step 3 of each model. Significant interactions were probed by testing the conditional effects of pain intensity at each level of pain-related anxiety using the PROCESS Macro (Hayes, 2013). Consistent with recommendations, associations were probed at three percentiles of pain-related anxiety (16^th, 50^th, 84^th; Hayes, 2013). If a non-significant interaction was observed, Step 2 of the model was examined. Because running multiple models increases the likelihood of a Type I error (Chen, Feng, & Yi, 2017), a Bonferroni correction was employed (Benjamini & Hochberg, 1995; Bland & Altman, 1995) and all interaction models were tested at more stringent significance levels (ps < .008).

3. Results

3.1. Participant Characteristics

Participants included N = 520 e-cigarette users (52.1% female; M_age = 34.85, SD = 11.17). The mean Penn State Electronic Cigarette Dependence Scale score was 8.91 (SD = 4.47) indicating low-to-moderate levels of e-cigarette dependence (Foulds et al., 2015), and participants reported using e-cigarettes on an average of 14.91 days over of the past month (SD = 10.17). Approximately 60% of the sample reported using e-cigarettes every day, and 35.6% of participants endorsed daily cigarette use. When asked to describe current use of e-cigarettes and cigarettes, 40.4% of participants reported exclusive e-cigarette use, whereas 28.5% reported that they considered themselves to be a dual user of both products. Average pain intensity was 14.45 (SD = 8.28) indicating moderate intensity of pain. Most of the sample was white (81.2%), married/living with someone (56.2%), earned more than $35,000 per year (61.0%) and had earned at least a four-year college degree (42.5%). Additional sociodemographic data and e-cigarette use characteristics are presented in Table 1.

Table 1.

Sociodemographic, E-Cigarette, and Pain Characteristics

	Total N = 520
	n (%)
Gender
Female	271 (52.1%)
Race
White	422 (81.2%)
Black or African American	61 (11.7%)
Other	37 (7.1%)
Ethnicity
Hispanic	103 (19.8%)
Marital Status
Married/Living with Someone	292 (56.2%)
Widowed	11 (2.1%)
Divorced/Separated	67 (12.9%)
Never Married	150 (28.8%)
Income
< $10,000	59 (11.3%)
$10,000 - $34,999	143 (27.5%)
More than $35,000	317 (61.0%)
Education
Did not graduate high school	22 (4.2%)
High school graduate	105 (20.2%)
Some college/Technical school/Associate’s degree	172 (33.1%)
College Degree	77 (14.8%)
School beyond college	144 (27.7%)
Daily Cigarette Smoking	185 (35.6%)
	M (SD)
Age	34.85 (11.17)
Current Pain Intensity a	14.45 (8.28)
E-Cigarette Dependence b	8.91 (4.47)
Motivation to Quit	5.99 (3.22)
# of Lifetime Quit Attempts	6.71 (12.73)
Pain-Related Anxiety c	51.34 (26.56)

Note.

^aGraded Chronic Pain Scale – Characteristic Pain Intensity,

^bPenn State Electronic Cigarette Dependence Index,

^cPain Anxiety Symptoms Scale − 20.

3.2. E-Cigarette Dependence and Motivation to Quit E-Cigarettes

Both pain intensity (Step 2: β = .11, p = .041) and pain-related anxiety (Step 2: β = .28, p < .001) were positively associated with e-cigarette dependence at Step 2 of the model. However, only pain-related anxiety remained significant after correcting for multiple models. Pain-related anxiety did not moderate the association between pain intensity and e-cigarette dependence (Step 3: β = .11, p = .387; Table 2). Pain-related anxiety was found to moderate associations between pain intensity and motivation to quit e-cigarettes (Step 3: β = .36; ΔR² = .01, p = .008; Table 2). Conditional analyses revealed that pain intensity was positively associated with motivation to quit e-cigarettes among participants with high pain-related anxiety (b = .07, SE = .03, p = .008), but not among those with moderate (b = .03, SE = .02, p = .179) or low levels of pain-related anxiety (b = −.02, SE = .03, p = .476; Graph 1).

Table 2.

Pain Intensity, Pain-Related Anxiety, and E-Cigarette Dependence and Motivation to Quit E-Cigarettes

E-Cigarette Dependence
	β	t	p	ΔR2	p for ΔR2
Step 1				.15	<001**
Gender	−.17	−3.89	<.001
Age	−.05	−1.12	.264
Race	−.05	1.16	.248
Anxietya	.11	1.67	.096
Depressionb	.20	3.01	.003*
Income	.10	2.31	.022*
Daily Smoker	.12	2.70	.007*
Step 2				.08	<001**
Pain Intensityc	.11	2.05	.041*
Pain-Related Anxietyd	.28	5.34	<001**
Step 3				.00	.387
Pain Intensity * Pain-Related Anxiety	.11	.87	.387
Motivation to Quit
	β	t	p	ΔR2	p for ΔR2
Step 1				.05	.002*
Gender	−.03	−.62	.536
Age	.04	.86	.404
Race	.11	2.36	.019*
Anxietya	.15	2.06	.040*
Depressionb	−.11	−1.51	.131
Income	.12	2.47	.014*
Daily Smoker	.04	.90	.367
E-Cigarette Dependencee	.11	2.34	.020*
Step 2				.04	<001**
Pain Intensityd	.08	1.37	.171
Pain-Related Anxietyd	.22	3.64	<001**
Step 3				.01	.008*
Pain Intensity * Pain-Related Anxiety	.36	2.66	.008*

Note: β = standardized beta weights;

^aOverall Anxiety Severity and Impairment Scale;

^bOverall Depression Severity and Impairment Scale;

^cGraded Chronic Pain Scale – Characteristic Pain Intensity;

^dPain Anxiety Symptoms Scale − 20;

^ePenn State Electronic Cigarette Dependence Index;

^*p < .05;

^**p < .001.

Graph 1. Conditional Effects of Pain-Related Anxiety on Associations Between Pain Intensity and Motivation to Quit E-Cigarettes.

Note: Pain Intensity – Graded Chronic Pain Scale, Characteristic Pain Intensity; Pain-Related Anxiety - Pain Anxiety Symptoms Scale − 20; Levels of pain-related anxiety include 16^th, 50^th, and 84^th percentiles; Conditional Effects: Low (b = −.02, SE = .03, p = .476; Graph 1), Moderate (b = .03, SE = .02, p = .179), High (b = .07, SE = .03, p = .008),

3.3. Lifetime Quit Attempts

The positive association between pain intensity and likelihood of endorsing a lifetime quit attempt was significantly moderated by pain-related anxiety (Step 3: OR = 1.00, p = .007; Table 3). Conditional analysis revealed that greater pain intensity was associated with a greater likelihood of endorsing a previous quit attempt among participants with high pain-related anxiety (b = .06, SE = .02, p = .003), whereas this association was nonsignificant among participants with moderate (b = .03, SE = .02, p = .085) or low levels of pain-related anxiety (b = −.01, SE = .02, p = .566). Among individuals who reported a lifetime quit attempt (n = 276), there was no significant pain intensity x pain-related anxiety interaction (β = .05; ΔR² = .00, p = .806), nor was there a main effect of pain intensity or pain-related anxiety on number of reported quit attempts (ps > .05).

Table 3.

Logistic Regression Examining Pain Intensity, Pain-Related Anxiety, and History of a Lifetime Quit Attempt

	Lifetime Quit Attempt (Yes/No)
Variable	OR	95% CI	P
Gender	1.37	.93–2.03	.111
Race	1.61	1.13–2.29	.009*
Age	.99	.98–1.01	.493
Anxietya	.98	.92–1.05	.596
Depressionb	1.01	.96–1.07	.694
Income	1.13	.85–1.50	.412
Daily Smoker	1.20	.80–1.82	.401
E-Cigarette Dependencec	1.06	1.01–1.12	.011*
Pain Intensityd	.97	.92–1.02	.192
Pain-Related Anxietye	.99	.98–1.01	.367
Pain Intensity x Pain-Related Anxiety	1.00	1.00–1.00	.007*

Note: OR = Odds Ratio;

^aOverall Anxiety Severity and Impairment Scale;

^bOverall Depression Severity and Impairment Scale;

^cPenn State Electronic Cigarette Dependence Index;

^dGraded Chronic Pain Scale – Characteristic Pain Intensity;

^ePain Anxiety Symptoms Scale − 20;

^*p < .05;

^**p < .001.

3.4. Perceived Barriers to Cessation

Pain-related anxiety was positively associated with greater perceived barriers to cessation from e-cigarettes (Step 2: β = .37; p < .001; Table 4). There was no association between pain intensity and this outcome (Step 2: β = −.01; p = .923). Additionally, pain-related anxiety was not found to moderate these associations (Step 3: β = .14; p = .176).

Table 4.

Pain Intensity, Pain-Related Anxiety, and Cessation-Related Outcomes

Number of Lifetime Quit Attempts
	β	t	p	ΔR2	p for ΔR2
Step 1				.04	.184
Gender	−.07	−1.09	.278
Age	.04	.60	.548
Race	−.06	−.87	.385
Anxietya	.16	1.55	.123
Depressionb	−.04	−.40	.692
Income	−.01	−.06	.955
Daily Smoking	.03	.46	.649
E-Cigarette Dependencec	.05	.77	.444
Step 2				.01	.522
Pain Intensityd	.04	.42	.676
Pain-Related Anxietye	−.10	−1.14	.256
Step 3				.00	.806
Pain Intensity * Pain-Related Anxiety	.50	.25	.806
Barriers to Cessation
	β	t	p	ΔR2	p for ΔR2
Step 1				.39	<001**
Gender	−.14	−3.78	<.001**
Age	−.04	−.08	.279
Race	.01	.24	.808
Anxietya	.20	3.36	.001*
Depressionb	.10	1.75	.081
Income	.06	1.56	.120
Daily Smoking	.05	1.43	.154
E-Cigarette Dependencec	.43	10.99	<001**
Step 2				.09	<001**
Pain Intensityd	−.01	−.10	.923
Pain-Related Anxietye	.37	8.08	<001**
Step 3
Pain Intensity * Pain-Related Anxiety	.14	1.36	.176	.00	.176
Negative Abstinence Expectancies
	β	t	p	ΔR2	p for ΔR2
Step 1				.47	<001**
Gender	−.12	−3.39	.001*
Age	.01	.03	.978
Race	.02	.54	.592
Anxietya	.15	2.69	.007*
Depressionb	.12	2.20	.029*
Income	.10	2.52	.012*
Daily Smoking	.15	4.11	<001**
E-Cigarette Dependencec	.50	13.33	<001**
Step 2				.12	<001**
Pain Intensityd	−.02	−.41	.681
Pain-Related Anxietye	.45	11.06	<001**
Step 3				.01	.001*
Pain Intensity * Pain-Related Anxiety	.30	3.27	.001*

Notes: β = standardized beta weights;

^aOverall Anxiety Severity and Impairment Scale;

^bOverall Depression Severity and Impairment Scale;

^cPenn State Electronic Cigarette Dependence Index;

^dGraded Chronic Pain Scale – Characteristic Pain Intensity;

^ePain Anxiety Symptoms Scale − 20;

^*p < .05;

^**p < .001.

3.5. Negative Abstinence Expectancies

Pain-related anxiety moderated associations between pain intensity and abstinence expectancies (Step 3: β = .30; ΔR² = .01, p = .001; Table 4). Examination of conditional effects revealed that among participants with high pain-related anxiety, pain intensity was associated with greater negative abstinence expectancies (b = .56, SE = .22, p = .010). There was no significant association of pain intensity with negative abstinence expectancies among participants with low (b = −.39, SE = .23, p = .090) or moderate levels of pain-related anxiety (b = .12, SE = .17, p = .489).

4. Discussion

This is the first study to examine the role of pain-related anxiety in relation to pain intensity and e-cigarette dependence/cessation-related outcomes. Results indicated that pain intensity and pain-related anxiety were each associated with greater e-cigarette dependence, and that pain-related anxiety was also associated with more perceived barriers to quitting e-cigarettes. Moderation analyses further showed that pain intensity was positively associated with motivation to quit e-cigarettes, likelihood of previously attempting to quit, and negative expectancies for abstaining from e-cigarettes among participants who reported high (but not moderate or low) pain-related anxiety.

These findings contribute to a growing empirical literature implicating pain-related factors in the maintenance of e-cigarette dependence. A previous survey conducted among 322 current e-cigarette users/dual users of cigarettes and e-cigarettes showed that pain was positively associated with e-cigarette dependence, perceived risks of e-cigarette use, and self-reported barriers to quitting (Zvolensky et al., 2018). In addition, pain-related anxiety has been identified as an important transdiagnostic factor in pain-tobacco cigarette dependence comorbidity (e.g., LaRowe et al., 2018). The current results integrate and extend this work by indicating that e-cigarette users who endorse more intense pain and high levels of pain-related anxiety may be especially susceptible to e-cigarette dependence and poor cessation outcomes. Similar to traditional tobacco cigarettes, e-cigarette use may be negatively reinforced via acute analgesic effects of nicotine (Ditre et al., 2016; LaRowe & Ditre, in press), and e-cigarette users with high pain-related anxiety may be more likely to employ maladaptive escape/avoidance behaviors, especially when experiencing pain (Crombez et al., 1999; McCracken et al., 1992).

Clinically, it is important to note that e-cigarette users in this sample who endorsed more intense pain and higher pain-anxiety reported greater motivation to quit than those who endorsed lower levels of pain intensity/anxiety. This finding is consistent with observations in the pain-tobacco cigarette literature (e.g., Zale, Ditre, Dorfman, Heckman, & Brandon, 2014), and suggests that e-cigarette users with pain may be amenable to participating in more intensive cessation interventions (Mohiuddin et al., 2007). For example, such individuals may benefit from integrated treatments that teach more adaptive skills for coping with pain and pain-related anxiety in the absence of e-cigarette use, decrease perceived barriers to quitting, challenge negative cessation-related outcome expectancies (Eccleston & Crombez, 1999; Schnoll et al., 2011; Vlaeyen & Linton, 2000).

Several limitations and directions for future research warrant discussion. First, the current study was cross-sectional in nature, and prospective research is needed to test for causal effects. For example, future work should examine pain-related anxiety as a prospective predictor of e-cigarette use and cessation-related variables, and test whether pain-related anxiety interacts with changes in pain intensity to predict e-cigarette use trajectories (e.g., using ecological momentary assessment; Dhingra et al., 2014). Second, approximately 80% of the current sample identified as White, and the mean level of e-cigarette dependence can be characterized as low-to-moderate (Foulds et al., 2015). Thus, the extent to which these findings may translate to more racially diverse samples or individuals with high levels of e-cigarette dependence remains unclear. Third, participants were not recruited based on the presence of chronic pain, and future research is needed to examine covariation between pain intensity, pain-related anxiety, and e-cigarette use among individuals with chronic pain. Fourth, participants were not excluded for use of combustible cigarettes. Although models covaried for daily cigarette smoking, future work may benefit from recruiting larger samples of exclusive e-cigarette users.

In summary, results of the current study represent an initial yet important step towards better understanding the role of pain and pain-related anxiety in e-cigarette dependence and quit-relevant constructs. This and future research have the potential to identify e-cigarette users who may be at-risk for greater dependence and difficulty quitting. Clinical implications include the possibility that e-cigarette users with pain may benefit from tailored cessation interventions that target pain-related anxiety, and that pain-related anxiety may be an important transdiagnostic factor underlying co-occurring pain and e-cigarette use (LaRowe et al., 2018).

Graph 2. Conditional Effects of Pain-Related Anxiety on Associations Between Pain Intensity and E-Cigarette Negative Abstinence Expectancies.

Note: Pain Intensity – Graded Chronic Pain Scale, Characteristic Pain Intensity; Pain-Related Anxiety - Pain Anxiety Symptoms Scale – 20; Levels of pain-related anxiety include 16^th, 50^th, and 84^th percentiles; Conditional Effects: Low (b = --.39, SE = .23, p = .090), Moderate (b = .12, SE = .17, p = .489), High (b = .55, SE = .21, p = .010).

Highlights.

• Pain intensity was positively related to motivation/attempt to quit e-cigarettes.

• Pain intensity was associated with greater negative abstinence expectancies.

• These associations were stronger among individuals with high pain-related anxiety.

• Pain-related anxiety was related to greater dependence and barriers to quit.

• Future work should test if pain-related anxiety predicts cessation trajectories.