Comparison of three measurement models of discounting among individuals with methamphetamine use disorder.

<bold>Background and Objectives: </bold>Delay discounting is associated with numerous clinically significant aspects of substance use disorders (SUDs). Recent studies have demonstrated that different models for assessing discounting may result in disparate conclusions. The current study compared two discounting tasks: money now versus money later (M-M) and methamphetamine now versus money later (MA-M) among non-treatment seeking individuals (N = 59) with methamphetamine use disorder (MAUD). Results from each task were assessed using three different models for assessing delay discounting.<bold>Methods: </bold>Discounting data were fit to three models of discounting, log k using Mazur's hyperbolic formula, area under the curve (AUC), and an alternative AUC model in which the delay values have been log transformed (AUClog).<bold>Results: </bold>For both discounting tasks, the distribution of model-related outcomes were normally distributed when using log k and AUClog, but skewed for AUC. Discounting in the MA-M task was significantly greater compared to the M-M task when using log k and AUClog but not AUC.<bold>Conclusion: </bold>To our knowledge, the current study is the first to report significantly greater discounting in a MA-M relative to M-M discounting task among individuals with MAUD, an outcome consistent with other psychomotor stimulants and drugs of abuse.<bold>Scientific Significance: </bold>The differential results observed across the three discounting models reaffirm potential issues with AUC noted in recent studies and highlight that researchers must be cautious when deciding on their final model of discounting. (Am J Addict 2018;XX:1-8). [ABSTRACT FROM AUTHOR]