A measure of expectancies for alcohol analgesia: Preliminary factor analysis, reliability, and validity

Highlights

• Expectancies for alcohol analgesia may influence pain-alcohol interrelations.

• The EAA is a novel 5-item measure of expectancies that alcohol will reduce pain.

• The EAA evinced a single-factor structure and evidence of reliability/validity.

• Future work should include further validation and tests of predictive utility.

Abstract

Rates of alcohol consumption are substantially higher among persons with pain, and recent research has focused on elucidating bidirectional pain-alcohol effects. Expectancies for alcohol analgesia could influence the degree to which alcohol confers acute pain-relieving effects, and may amplify the propensity to respond to pain with drinking behavior. However, no validated measures of expectancies for alcohol analgesia are available. Therefore, we developed a five-item measure of Expectancies for Alcohol Analgesia (EAA), which assesses the perceived likelihood that alcohol will reduce pain. The goal of this project was to examine psychometric properties of the EAA among a sample of 273 current alcohol users with chronic pain (M_age = 32.9; 34% female) who completed an online survey of pain and substance use. Confirmatory factor analysis (CFA) results indicated that the hypothesized single-factor structure of the EAA provided good model fit (Bollen-Stine bootstrap p = .13). The EAA also showed excellent internal consistency (α = 0.97), and scores were positively associated with quantity/frequency of alcohol use, alcohol outcome expectancies, coping-related drinking motives, and pain severity (ps < 0.01). These findings provide initial support regarding the single-factor structure, reliability, and validity of the EAA. Examination of predictive utility and further validation are important next steps.

Introduction

Epidemiological estimates indicate that individuals who suffer from chronic musculoskeletal pain are twice as likely to meet diagnostic criteria for alcohol dependence, when compared to their pain-free counterparts (Von Korff et al., 2005). Similarly, the prevalence of pain appears to be substantially higher among problem drinkers (vs. non-problem drinkers; Brennan, Schutte, & Moos, 2005), and up to three-quarters of substance use treatment patients who identify alcohol as their drug of choice also report moderate-to-severe past-month pain (Larson et al., 2007). Given the high co-occurrence of pain and alcohol use, recent work has begun to elucidate bidirectional effects in pain-alcohol relations (e.g., Ditre et al., 2019, Edwards et al., 2020, Zale et al., 2015).

An established reciprocal model posits that pain and alcohol use interact in the manner of a positive feedback loop, resulting in the exacerbation and maintenance of both conditions over time (Ditre et al., 2019, Zale et al., 2015). Although alcohol can confer acute analgesia (Thompson, Oram, Correll, Tsermentseli, & Stubbs, 2017), excessive alcohol consumption is associated with the onset and severity of numerous painful conditions. For example, heavy alcohol use is a causal factor in the development of alcohol-induced pancreatitis (Lerch et al., 2003) and alcohol-related neuropathy (Chopra & Tiwari, 2012), and may increase the risk of developing osteoarthritis (Cheng et al., 2000) and pain following musculoskeletal injury (Sá, Baptista, Matos, & Lessa, 2008). There is also converging research indicating that pain can increase motivation to drink alcohol. For example, laboratory pain induction increases self-reported urge to consume alcohol (Moskal, Maisto, De Vita, & Ditre, 2018), greater levels of pain unpleasantness have been associated with increased motivation to drink (Lawton & Simpson, 2009), nearly one-quarter of patients enrolled in both pain treatment and inpatient substance abuse programs have endorsed using alcohol to cope with pain (Goebel et al., 2011, Sheu et al., 2008), and pain intensity has been positively associated with alcohol coping motives (Rogers, Zegel, Tran, Zvolensky, & Vujanovic, 2020). Indeed, acute alcohol analgesia may negatively reinforce alcohol use and strengthen beliefs about the pain-relieving effects of alcohol (Ditre et al., 2019). Importantly, using alcohol to reduce pain can lead to increased consumption over time (Brennan et al., 2005).

Ditre and colleagues (2019) proposed that bidirectional pain-alcohol effects are likely influenced by outcome expectancies (i.e., estimates that a given behavior will lead to specific outcomes), which are considered to be important determinants of motivation and behavior (e.g., Bandura, 1989, Rotter, 1954). There is a vast literature documenting the role of outcome expectancies in the initiation, progression, and maintenance of alcohol use, and several measures have been developed to assess alcohol outcome expectancies (e.g., Brown et al., 1987, Fromme et al., 1993, Leigh and Stacy, 1993, Solomon and Annis, 1989). These measures assess both general and specific alcohol outcome expectancies, across a variety of domains (e.g., social facilitation, tension reduction, cognitive impairments). Higher scores on measures of positive alcohol outcome expectancies (i.e., estimates that alcohol use will result in desired consequences) have consistently been correlated with greater drinking motives and quantity/frequency of alcohol consumption (e.g., Jones et al., 2001, Madden and Clapp, 2019, Monk and Heim, 2013). In contrast, negative alcohol outcome expectancies (i.e., estimates that alcohol use will result in undesired consequences) have been associated with reduced consumption and a greater desire to restrain from drinking (e.g., Jones et al., 2001, Monk and Heim, 2013). Given that positive (vs. negative) outcomes of alcohol use are often more immediate, it has been suggested that positive outcome expectancies are often more influential on drinking behavior (e.g., Stacy, Widaman, & Marlatt, 1990). Moreover, it has also been noted that the predictive utility of expectancy measures further improves with greater specificity of measurement (e.g., Fromme et al., 1993).

Previous work has demonstrated the importance of assessing specific expectancies that substance use will reduce pain. For example, a measure of pain and smoking expectancies (PSE; Ditre, 2006) was developed to assess beliefs that cigarette smoking will have acute analgesic effects. The PSE has demonstrated excellent internal consistency (Ditre, 2006, Ditre et al., 2010), and has been shown to account for nearly one-third of the variance in pain-induced urge to smoke cigarettes (Parkerson & Asmundson, 2016). Alcohol users likely hold similar expectations for pain relief (Ditre et al., 2019, Zale et al., 2015), which may increase propensity to respond to actual or anticipated pain with drinking behavior. Consistent with evidence that positive alcohol outcome expectancies are correlated with greater quantity/ and frequency of alcohol consumption (Jones et al., 2001), it is also possible that expectancies for alcohol analgesia may lead to increased alcohol consumption and the development/maintenance of hazardous drinking patterns. Moreover, an accumulating literature demonstrates that analgesic outcome expectancies can influence the experience of pain (e.g., Atlas and Wager, 2014, Bingel et al., 2011, Butcher and Carmody, 2012, Ossipov et al., 2010, Peerdeman et al., 2016), and expectancies for alcohol analgesia may influence the degree to which alcohol use confers acute pain-relieving effects. However, we are not aware of any validated measure of alcohol outcome expectancies for pain relief.

We developed a measure of Expectancies for Alcohol Analgesia (EAA) to assess the perceived likelihood that alcohol consumption will reduce pain. The EAA consists of 5 items that were adapted from the PSE and are hypothesized to reflect a single-factor. All authors, who are experts in the domain of pain and substance use, reviewed and approved the adapted items for content validity. The primary goal of this study was to conduct an initial evaluation of the EAA factor structure, reliability, and validity among a sample of current alcohol users with chronic musculoskeletal pain. We hypothesized that the EAA would demonstrate (1) a single-factor structure, (2) acceptable internal consistency (α > 0.7), (3) initial evidence of concurrent validity via medium-to-large sized correlations with outcomes related to both alcohol consumption and clinical pain experience, and (4) initial evidence of discriminant validity via the absence of associations with cannabis use (a theoretically distinct construct). An exploratory aim of this study was to assess whether EAA scores differed as a function of sociodemographic characteristics (e.g., gender, race) and/or the presence of a high level of alcohol problems.