Psychometric properties of the German Penn Alcohol Craving Scale

Abstract

Abstract Craving for alcohol is an important diagnostic criterion in alcohol use disorder (AUD) and an established predictor of future relapse. The 5-item Penn Alcohol Craving Scale (PACS) is one of the most widely used questionnaires to quantify craving and has been translated into different languages. It is assumed that the PACS constitutes one factor, although theoretical considerations suggest an additional second factor. We conducted stability and factor analyses (principal component and confirmatory factor analyses) of the German PACS (PACS-G) in samples of patients with AUD from the following three German study sites: Erlangen, N = 188 (mean age: 47.1 years, 43.5% female); Mannheim, N = 440 (45.5 years, 28.6% female); Hannover, N = 107 (48.1 years, 48.6% female). In our samples, the 2-factor solution of the PACS-G version is more stable than the internationally assumed 1-factor solution. The resulting two PACS-G subscores ‘difficulty to resist’ (items 4 and 5) and ‘thoughts about alcohol’ (items 1, 2, and 3) have an internal consistency (Cronbach’s alpha) of 0.80 ≤ α ≤ 0.90, m = 0.86 and 0.86 ≤ α ≤ 0.91, m = 0.89 with an overlap of R2 = 62%. We found good convergent validity assessed via the Craving Automatized Scale-Alcohol and the Obsessive-Compulsive Drinking Scale, but also correlations with depression and anxiety assessed via the Beck’s Depression and Anxiety Inventories. This study is the first to provide evidence for a 2-factor solution (‘difficulty to resist’ and ‘thoughts about alcohol’) underlying the PACS-G version.