Objectives: Dependence on addictive substances is indicated by impaired control, strong desires/craving, withdrawal symptoms with abstinence, challenges with quitting, etc. Most measures of dependence are substance-specific, which inhibits comparisons across substances. In light of increasing prevalence of cannabis and tobacco co-use, we aimed to adapt indicators of tobacco dependence to assess cannabis dependence (CD).
Methods: In 2024, past-year cannabis consumers (n = 320; Mage = 35) from an ongoing 24-year longitudinal study completed self-report questionnaires. We examined the factor structure and measurement invariance across sex of the adapted 15-item scale, assessed the likelihood of endorsing and discriminative power of each item in participants with different degrees of CD, and explored the internal consistency, and construct and criterion-related validity of the scale.
Results: Items loaded strongly on a single factor, and strict measurement invariance across sex was confirmed. Items demonstrated strong discriminative power among individuals with varying degrees of CD. The scale demonstrated high internal consistency (ωcat=0.94), excellent convergent reliability (r = 0.76 with the Cannabis Abuse Screening Test), and a small though statistically significant inverse correlation with age at first use (ρ=-0.20). There was a dose-response relationship with frequency of cannabis use (ρ=0.81).
Conclusions: The adapted scale has excellent psychometric properties and is invariant across sex. Additional research is needed to determine whether these properties hold with larger and more diverse samples and across types of cannabis products (e.g., leaf, edibles, hashish). The scale may permit direct comparison of dependence across tobacco vs. cannabis consumers.
Keywords: Cannabis dependence; Confirmatory factor analysis; Item response theory; Measurement invariance; Reliability and validity.
Copyright © 2025 Elsevier B.V. All rights reserved.