Commentary on Martínez‐Loredo et al.: Where do we go from here? Increasing the clinical utility of alcohol purchase tasks by expanding our definition of constraint

Abstract

Alcohol purchase tasks provide clinically meaningful indices of motivation to consume alcohol by measuring decisions to consume alcohol across price constraints. The convergent validity of these tasks could be enhanced by extending the definition of constraints to include clinically relevant contextual factors that could help to inform ideographic treatment approaches. Imagine a Friday night pre-pandemic. You are attending a concert with your closest companions. Before the show starts, you meander to the bar to buy a drink. While you review the menu, you overhear the bartender say that drinks will cost $10 each. How many would you purchase? What if drinks only cost $5? $2? Many of us have experienced these moments. We might be willing to buy three drinks when they are $5, but only one at the $10 price. Of course, we all know of someone who refuses to pay more for a single beer than what it would cost in a six-pack (those hyper-rational agents). What about those infrequent, but remarkable circumstances where drinks are free, such as at a wedding, or a university party? Now, imagine you are home alone and have an important event the next morning—how many drinks would you have under these circumstances? These real-life scenarios demonstrate a fundamental concept in behavioral theory: as the costs of alcohol increase, choices to purchase and consume alcohol decrease, and both the amount of consumption in the absence of constraints and the degree to which constraints influence choice can serve as markers of alcohol value. Borrowing from economic theory, behavioral scientists have developed alcohol purchase tasks (APTs) modeled after laboratory progressive ratio tasks (and real-life choices) to measure this alcohol value, or alcohol demand. Demand indices incorporate data across constraints and should theoretically serve as robust predictors of extended periods of observable real-life behavior. Martínez-Loredo and colleagues [1] update previous meta-analytical effect sizes [2], demonstrating a robust association between alcohol demand indices and alcohol consumption, problems, and heavy drinking. Intensity (consumption under no constraints) demonstrated uniquely robust utility in predicating alcohol problems. Opportunities for unconstrained drinking may be relatively infrequent (e.g. parties, weddings) and thus may be difficult to predict based on standard assessments of drinking. However, these ‘high-intensity’ drinking episodes can be extremely risky and are thus important to account for in models of alcohol problem severity used to inform alcohol interventions. Intensity may be a useful index in predicting risk for these high-intensity drinking events over time, and should be examined in future research. An additional unique utility of APTs may lie in their dynamic measurement of fluctuations in alcohol valuation due to intentional modeling of behavior across constraints. Although there are numerous valid indices of alcohol use severity (e.g. quantity/frequency, heavy drinking, alcohol use disorder symptoms), these are often ‘slices’ of past behavior which took place within a set of environmental constraints, and do not account for possible fluctuations in these constraints under future conditions. In other words, an individual who drank four drinks last week may have drunk more if drinks were more available or if they had seen their friends. Although standard APTs define constraints as increasing price, a substantial literature has demonstrated that alcohol demand changes under a variety of other constraints [3], such as when one has to drive after drinking [4], when there is a next-day responsibility (opportunity cost [5]) and when friends are no longer present [6]. In other words, demand elasticity might be better conceptualized more broadly as the change in consumption as a function of constraint, with monetary price being just one important and easily quantified type of constraint [3]. Thus, one future direction to increase the convergent validity of APTs might be to quantify the effect of other clinically relevant constraints. The present review provides an excellent starting-point for what we hope is the next stage of APT research, which carefully models relative utility as a dynamic indicator of motivation to consume alcohol in any given moment. Purchase tasks exploring novel constraints could identify idiographic high-risk situations and protective factors for informing person-centered treatment and intervention. This research may be facilitated specifically with brief APTs administered through ecological momentary assessment to determine dynamic contextual influences on demand. This could greatly enhance the clinical utility of the APT. For example, an individual may report low drinking while alone but heavy drinking at low drink prices when with friends. These tasks would provide nuanced contextual data that could help clinicians to identify vulnerable moments to target through intervention. None. Samuel F. Acuff: Conceptualization, Writing, Editing. James G. Murphy: Conceptualization, Editing.