We’ve come a long way in understanding addiction and the underlying neurobiology; however, there are a lot of theories attempting to explain how, exactly, addiction arises. Partly, this is due to the complex nature of addiction, as it arises from a combination of factors. How someone was brought up, who his or her parents are/were, what genes were passed on, personality traits, how much money someone makes, whether someone has a co-occurring mental health disorder, all play a role in driving addictive behaviors.
I don’t claim to have a definitive answer for addiction, and I am more inclined to agree with many of the theories, and that perhaps one or more explain addiction in a person’s life. In other words, I generally think everyone is right to a certain degree, but that strict adherence to a single theory means that others are being overlooked or even undermined.
Recently, I saw an article by Dr. Nora Volkow, director of NIDA (National Institute of Drug Abuse), in which she said that addiction is a disease of free will. In other words, people engage in addictive behaviors because they have lost the ability to choose. On the other side are scientists, such as Marc Lewis and Gene Heyman, arguing that addiction is not a disease; rather it’s a disorder of choice. The brain can change and addictive behaviors can be undone. It’s not a life sentence. To make matters more complex, Dr. Carl Hart argues that addiction is rare, and that many scientists fail to consider the environment in which a person was raised or lives now. By improving the environment, fewer people will be inclined to use drugs. By offering substance-dependent individuals an attractive choice, many people will choose not to use. This forms the basis of Contingency Management treatment, something I have talked about before and have published a peer-reviewed article on.
Would it be copping out to say that I agree with all of these scientists? Maybe. My grad school adviser published something along these lines, claiming that several credible theories fit into a decision-making framework, and addiction influences different decision-making systems for different people.
In the previous post, part 1 in my series on theories of addiction, I discussed compulsivity and opponent process theories, two sides of the same addiction coin. Compulsivity is characterized by more automatic processes and only occurs in a subset of people (and animals), whereas opponent process may imply more purposeful actions in order to avoid negative situations (e.g., withdrawal). I made the argument that individuals could abuse drugs or be dependent on drugs because of one of these theories, suggesting that approaches to treating individuals should differ. I will explore this topic in full in the next post.
In this article, I will talk about the change from voluntary to habit behaviors and the underlying neural processes that drive them.
You know that person (or maybe you are that person) who used to drink or do drugs or smoke cigarettes because it was fun? Now, he or she doesn’t seem to enjoy it anymore but still does it anyway. With anyone and with any behavior, and except in extreme examples, the initial action is voluntary. Taking that action repeatedly creates something like a well-worn path, one that’s easy to slip into, and the behavior eventually becomes a habit: more automatic; sometimes subconscious; and difficult to change.
In animal studies, this has been studied by training an animal, like a rat, to associate two levers with two different food rewards. Then, one of the rewards is paired with a substance that makes the rat sick. Think about a time when you or someone you know was sick, ate, and then threw up after eating. It’s a common phenomena, and it typically results in an aversion to that food. I have a family member who was sick, ate rice, and threw up, and this family member hated rice for years. Anyway, rats quickly associate the sickness with the specific food reward and don’t press that lever anymore. That is, unless the rats have been extensively trained on that particular lever. Then, the rats don’t care, and they keep pressing that lever. It becomes habitual.
In several remarkable studies, scientists found that voluntary behavior is associated with certain brain areas, such as the hippocampus, dorsal medial striatum (caudate in people), and pre-frontal cortical areas. On the other hand, habit-based behavior is mediated by dorsal lateral striatum (putamen in people) and sensory/motor cortical areas. The remarkable thing is that when they temporarily inactivated the dorsal lateral striatum in a rat that was in full habit mode, the rat went back to behavior that looked like voluntary behavior. It was like the functionality of these brain areas that mediated these differing types of behavior were layered. By removing the habit layer, the voluntary layer was uncovered.
Areas of the brain form systems that work in conjunction. When behavior is voluntary, areas such as the hippocampus, ventral striatum, and pre-frontal cortex work together as an individual makes decisions about what to do, where to go, what action to take, and so on. Voluntary actions take into consideration an individual’s current state and potential future outcomes of actions. Let’s call this the choice system.
As same or very similar actions are performed over and over, eventually the neural system shifts, and areas such as the dorsal lateral striatum and motor areas of the cortex work together to make actions happen quickly and with little thought. Let’s call this the habit system. In a sense, after the habit system takes over, the choice system is bypassed. You can read Barry Everitt’s article to learn more about this process. Also, here’s a brief interview with him discussing some of what I’ve covered, so far.
This type of learning is normal and it happens with any action over time. It frees up an individual to learn new associations while still executing well-learned actions. But, this article is about addiction, so the question is what is unique about addiction? Do drugs of abuse exploit this system, perhaps causing actions to become habit more quickly than normal habits? Or, perhaps the drug habit is formed more slowly, so that the value of drugs increase and become more valuable than natural rewards?
Just like normal decision making, drug use is voluntary at first and becomes habit-based after extensive use and is mediated by brain areas that govern voluntary and habit-based behavior, respectively. It is also known that inactivating brain areas that govern normal habits also reduces drug use after extensive drug use.
People and animals with addiction have abnormal dopamagergic signaling, and it was proposed that this abnormality causes habits to develop faster. And, in one study, there was a facilitated shift from the choice to the habit system. In another, investigating subjects that were more impulsive , meaning they were generally more vulnerable to addiction, the transition from voluntary to habit brain areas was slower than normal.
So then, studies so far are mixed and have not answered whether addictive drugs cause faster or slower developing habits, yet. Regardless, my guess is that only a subset of people develop habit-based addiction. If so, how might this happen?
A compelling theory involves something called noncompensable dopamine. Basically, the theory proposes that drugs of abuse cause the brain to over-value drugs by increasing the selection of actions that lead to drug use. My adviser in grad school (who also came up with this theory) had a great analogy for this.
Think about getting a candy bar from a vending machine and imagine you have a choice from three vending machines side by side. Usually there’s a predictable relationship between you, your money, and the candy bar/s. But, what if one day the vending machine in the middle gives you two candy bars? Wouldn’t you be more likely to select that machine and that particular candy bar on subsequent visits? This bonus is understood in the context of temporal difference reinforcement learning and is computed as a reward-prediction error. You expect one candy bar, but you get two, so the reward-prediction error is +1, and dopamanergic mechanisms in the brain account for this error by increasing the value of that machine and that particular selection.
The theory goes that drugs produce this bonus (reward-prediction error +1) every single time you use them, creating a perpetual increase in their value; a value that is stored in the brain. If this theory is true, then no wonder drugs are so hard to kick and undermines other (regular) rewards. When we scientists talk about drugs hijacking the brain, noncompensable dopamine provides a highly probable mechanism for what we mean.
Is that it? Has this explained addiction? Unfortunately, no. Well maybe, but we don’t know yet, since there are only a few studies testing the theory, and in one study the value of cocaine did not continue to increase. But preliminary data (shown at neuroscience conferences) indicated that subjects more vulnerable to drug addiction may be susceptible to some of the things proposed by this theory.
However, noncompensable dopamine fits into the framework of the actor/critic model (for more on this, read Yael Niv and colleagues), and together with temporal difference learning (remember the candy bar example?) and a brain circuit called the basal ganglia, we may be getting closer to the crux of the problem.
The “critic” is the evaluator of current and potential future states and is updated by the reward-prediction error. The “actor” performs reward-related actions with guidance from the “critic”. Drugs of abuse may affect both the “critic” and the “actor”. In other words, using drugs might affect reward valuation, in that the choice system would perpetually associate increased value with drug cues. This might have the side-effect of causing multiple cues in the environment to be associated with the drug, so that multiple drug cues would elicit drug seeking by increasing craving for the drug and making relapse more likely. In this same actor/critic framework, others have proposed that drugs of abuse cause a lack or deficit of the reward-prediction error signal, or that drugs cause the “critic” to only update value for the present action but doesn’t integrate past and potential future rewards. Either way, ideas about this are consistent in that drugs affect the critic’s ability to correctly evaluate and update reward values.
Until more studies are done to test the noncompensable dopamine theory, and just like other posited theories, individual variability in addiction may provide a temporary explanation. As noted, preliminary evidence suggests that noncompensable dopamine may only occur with a subset of drug users, acting as yet another vulnerability to increased drug use. Additionally, actor/critic evaluations manipulated by addicting drugs may help explain the appeal of drug use more broadly, helping us to understand how addictive behaviors arise and continue despite negative consequences.
Applying what we know
Is it useful to consider all these different theories? Will this help guide treatment for addiction? I think it is, and I think it will. If we can understand how a person develops addiction, we’ll be in a better position to treat that person.
That’s the ultimate goal of this series. To be practical and effective. During my time studying addiction, it has become more and more clear that addiction is a complex issue, an issue that is divided with many voices on each side. That makes sense, since it’s also very personal. Even if you have not experienced the incessant tug of craving or the physical and emotional lows of withdrawal, I bet you know someone who has. In the next post I’m going to cover the more philosophical theories of addiction before laying the groundwork for what I would consider to be a more effective approach to dealing with addiction. I believe it is these philosophies that cause so much conflict. It may appear to many people – scientists, policy makers, AA goers – that proponents of the choice model are going to undo years of work and advocacy that helped the public to finally see that addiction is not a moral failing. True, it’s not a moral failing, but it’s also irresponsible to view addiction as only driven by neurobiology. We need to be aware of these other factors, along with the neurobiological changes, because it will help to guide people in recovery as well as people making decisions about what is considered valuable treatment.
About the author
Paul Regier has a PhD in neuroscience and studies addiction at UPenn. You can follow him here on The Real Edition, on Medium, and on Twitter @Form_Tell. Please, correct him when you think he’s wrong and provide feedback and comments.
Header picture via io9