Introduction: Distress tolerance (DT), or the ability to withstand negative affect, has been associated with a variety of psychological dysfunctions including alcohol use disorders (AUDs). Specifically, DT of negative emotions may be an important mechanism in the initiation and maintenance of binge drinking (BD). The Distress Tolerance Scale (DTS) is a self-report measure that is widely used within the substance use literature, yet its psychometric properties have not been consistently replicated. It remains unclear whether the underlying structure of the DTS reflects a single or hierarchical model, which may explain why studies on alcohol misuse have reported mixed findings with the DTS and have inconsistently used the DTS. Furthermore, despite the use of the DTS to understand the role of DT in problematic drinking among young adults in college, its psychometric properties among this high-risk population have not been examined. The aim of the current study was to examine the validity and reliability of the DTS in college students who binge drink.
Methods: The study included a sample of 147 undergraduate students (Mage = 19.92, SDage = 1.31; 55% female) who reported binge drinking (4/5 drinks for females/for males in two hours). Confirmatory factor analyses (CFAs) using AMOS version 20 (Arbuckle, 2006) were conducted comparing the single-factor and hierarchical multidimensional model described in previous research. Convergent validity of the DTS was examined by conducting bivariate Pearson correlations between DTS total score and its subscales and drinking motives (DMQ-R), depressive symptoms (BDI), alcohol-related consequences (BYACCQ), and binge drinking frequency (TLFB).
Results: The CFA testing the single factor model yielded a χ2 (90, n = 120) = 271.86, p< .001, indicating a poor fit. Further evaluation of the goodness of fit showed the fit was poor across various indices using both conservative and liberal cut-offs: CMIN/DF = 3.02; CFI = .84; and RMSEA = .12. The CFA examining a hierarchical model also indicated a poor fit: χ2 (89, n = 120) = 180.46, with p < .001. Additional indices suggested the fit was mixed, CMIN/DF = 2.03, CFI = .92, and RMSEA = .09. However, the overall DTS (α = .93) and its four subscales (.74 ≤ α ≥ .87) demonstrated good internal consistency. Convergent validity was established with self-reported depressive symptoms (-.35 ≤ r ≥ -.69), drinking motives (-.29 ≤ r ≥ -.47), and alcohol-related problems (-.21 ≤ r ≥ -.29), but not with drinking quantity or frequency.
Discussion: This is the first study to examine the psychometric properties of the DTS, a commonly used self-report measure of DT, in college students who binge drink. The DTS failed to demonstrate adequate structural validity in both the hierarchical and single-factor models despite good internal consistency and convergent validity. Thus, although the DTS consists of related items that appear to be indicators of DT, the scale does not adequately measure the construct of perceived DT of negative emotions in college students who binge drink. Recommendations for strengthening the validity of the DTS and an alternative for assessing distress tolerance in substance-use are offered.
Amanda Khan– Clinical Psychology Doctoral Candidate, Psychology Department, Suffolk University, Cambridge, Massachusetts
Benjamin Shapero– Depression Clinical & Research Program, Massachusetts General Hospital
Lauren Fisher– Depression Clinical & Research Program, Massachusetts General Hospital
Maren Nyer– Depression Clinical & Research Program, Massachusetts General Hospital
Amy Farabaugh– Depression Clinical & Research Program, Massachusetts General Hospital
Laura MacPherson– Greenebaum Comprehensive Cancer Center, University of Maryland
Paola Pedrelli– Depression Clinical & Research Program, Massachusetts General Hospital
Clinical Psychology Doctoral Candidate
Psychology Department, Suffolk University
Depression Clinical & Research Program, Massachusetts General Hospital