Interaction of processes

If these processes all actuate the clinical manifestation of addiction, it would seem that at least some of these processes must interact and mutually drive the process to stability in the addictive state. We consider just two of the processes and illustrate a plausible means by which this interaction may occur. One of the clinically relevant aspects of addiction is the relative insensitivity to price for the addictive commodity. This might result from an interaction between euphoriogenic rewards or reinforcement (Vulnerability 3) and temporal discounting (Vulnerability 9). Consider that sensitivity to price is affected by the availability of substitutes. For example, a consumer would not buy a commodity for a high price if a nearly identical commodity (a substitute) were available for a substantially lower price. One source of substitution is inter-temporal. For example, a consumer would not buy a commodity for a high price now if a nearly identical commodity (a substitute) were available later at a substantially lower price. However, if the future value of a commodity (say a drug of dependence) is radically discounted, as is often the case (e.g., Bickel & Marsch Reference Bickel and Marsch2001; Yi et al., in press), then the opportunity for inter-temporal substitution is limited, which, in turn, would result in less sensitivity to price. We think that there is opportunity for some of the other processes illuminated by Redish and colleagues to interact in ways that may plausibly result in known clinical features of addiction.

Co-identity of processes

Some of the processes identified by Redish et al. have had a very limited history of investigation. As such there may be processes that are treated separately for reason of historical contingency of its exploration and naming but are very closely related, if not referring to the same phenomena. Let us again consider two: discounting (Vulnerability 9) and balance between the planning and habit system (Vulnerability 8). Redish et al. note that drugs may inhibit structures involved with planning over habit systems. These two systems have been variously named (e.g., Bechara Reference Bechara2005), and we have referred to them as “impulsive” and “executive” (Bickel et al. Reference Bickel, Miller, Yi, Kowal, Lindquist and Pitcock2007). Importantly, evidence suggests that temporal discounting serves as a summary measure of these two systems. Specifically, McClure et al. (Reference McClure, Laibson, Loewenstein and Cohen2004) used functional magnetic resonance imaging to scan the brains of normal adults engaged in discounting procedures. Choices favoring the immediately available option were associated with greater relative activation in structures associated with the habit, or “impulsive,” system (e.g., limbic region). In contrast, choices favoring the temporally remote option were associated with greater relative activation in structures associated with the planning, or “executive,” system (e.g., prefrontal cortex). As such, temporal discounting provides a specific measure of the balance between these competing systems, suggesting that these processes are in fact the same.

Centrality of some processes

To the extent we know how these processes affect and interact with each other, we may come to understand the organization or topology of these processes in addiction. The centrality of these processes may not all be equivalent, and knowing their organization may suggest the most effective means of “treating” the system; that is, some processes may have a greater effect on the system than others. As an example, consider protein synthesis in the yeast. The yeast cell produces approximately 1,870 different proteins (Jeong et al. Reference Jeong, Mason, Barabasi and Oltvai2001). However, these proteins are not all equally essential to the yeast cell. Understanding the link among these proteins has shown that highly connected proteins are three times more essential (i.e., deletion results in cell death) than proteins with a small number of links. Suggestions of how to adapt these typological organizations for the study of addiction have been offered (Chambers et al. Reference Chambers, Bickel and Potenza2007). Application of such approaches may indicate which component processes are best targeted to ameliorate addiction.

In conclusion, we applaud the work done by Redish et al. It is an important first step, and it puts us on the right footing in our task to understand, prevent, and treat addiction. We may need to know many things to reach our goal, and we hope we will come to know a small number of things that have very big effects.