Loss, but not absence, of control – How choice and addiction are related

In a recent post the notion that “loss of control” is an addiction myth was raised by our contributing author, Christopher Russell, a thoughtful graduate student studying substance abuse in the U.K. Though I obviously personally believe in control- and choice-relevant neurological mechanisms playing a part in addiction, this conversation is a common one both within and outside of the drug abuse field. Therefore, I welcome the discussion onto our pages. I’d like to start out by reviewing some of the more abstract differences between my view and the one expressed by Christopher and follow those with some evidence to support my view and refute the evidence brought forth by him.

Addiction conceptualization – Philosophical and logical differences and misinterpretations

One of the first issues I take with the argument against control as a major factor in drug addiction is the interpretation of the phrase “loss of control” as meaning absence, rather than a reduction, in control over addiction and addictive behavior. Clearly though, one of the definitions of loss is a “decrease in amount, magnitude, or degree” (from Merriam-Webster.com) and not the destruction of something. Science is an exercise in probabilities so when scientists say “loss”, they mean a decrease and not a complete absence in the same way that findings showing that smoking cigarettes causes cancer do not mean that if an individual smokes cigarettes they will inevitably develop cancerous tumors. Similarly, the word “can’t” colloquially means having a low probability of success and not the complete inability to succeed. Intervention that improve the probability of quitting smoking (like bupropion or quitlines for smoking) success are therefore said to cause improvements in the capacity for quitting.

Next, Christopher wants scientists to identify the source of “will” in the brain but I suggest that “will” itself is simply a term he has given a behavioral outcome – the ability to make a choice that falls in line with expectations. In actuality, “will” is more commonly used as a reference to motivation, which while measurable, isn’t really the aspect of addiction involved in cognitive control. Instead, what we’re talking about is “capacity” to make a choice. The issue is a significant, not semantic one, since the argument most neuroscientists make about drug abuse is that addicts suffer a reduced capacity to make appropriate behavioral choices, especially as they pertain to engaging in the addictive behavior of interest. If someone is attempting to get into a car but repeatedly fails, we say they can’t get in the car (capacity), not that they don’t want to (will). Saying that they simply “don’t” get in the car doesn’t get at either capacity or will but instead is simply descriptive. I don’t believe that science is, or should be, merely descriptive but instead that it allows us to form conclusions based on available information.

That there is a segment of individuals who develop compulsive behavioral patterns tied to alcohol and drug use and who attempt to stop but fail is, to my mind, evidence that those individuals have a difficulty (capacity) in stopping their drug use. Their motivation (will) to quit is an aspect that has been shown to be associated with their probability of success but the two are by no means synonymous. It is important to note, and understand, that the attribution for the performance should not fall squarely on the shoulders of the individuals. We humans are so prone to making that mistake that it has a name, “The fundamental attribution error,” and indeed, individuals who show compulsive, addictive, behavior do so because of neuropharmacological, environmental, and social reasons in addition to the complex interactions between them all. But no one is disputing that and in fact, the article used by Christopher to point out the notion of a “tipping point” in addiction directly points out that fact in the next paragraph (Page 4), which he chose not to reference or acknowledge.

“Of course, addiction is not that simple. Addiction is not just a brain disease. It is a brain disease for which the social contexts in which it has both developed and is expressed are critically important… The implications are obvious. If we understand addiction as a prototypical psychobiological illness, with critical biological, behavioral, and social-context components, our treatment strategies must include biological, behavioral, and social-context elements.” (Lashner, 1997)

Lastly, Christopher’s philosophical musings are interesting, but they seem to stray away from trying to find an explanation for behavior and instead simply deconstruct evidence. In a personal communication I explained that while most addiction researchers understand that addiction, like most other mental health disorders is composed of a continuum of control ranging from absolute control over behavior to no control whatsoever (with most people fitting somewhere in the middle and few if any at the extreme ends), categorization is a necessary evil of clinical treatment. The same is true for every quantitative measure from height (Dwarfism is sometimes defined as adults who are shorter than 4’10”) to weight (BMI greater than 30 kg/m²). I think it’s equally as tough to argue that someone with a BMI of 29.5 is distinctly different from an individual with a BMI of 30 as it is to argue that there is no utility in the classification. Well, the same applies for drug addiction, although some people categorically object to classification and believe it has no utility or justification.

Now for the evidence – “Choice” and “control” are not the same as “will”

Some people quit, even without help – Christopher and a number of the people he cites in support (Peele, Alexander), suggest that because some people do stop using that it can’t be said that there is a problem with any individuals’ capacity to stop. The problem with that argument is that it supposes that everyone is the same, a fact that is simply false. As an example I would like to suggest that we compare cognitive control with physical control and use Huntington’s Disease (HD or Huntington’s Chorea) as an example.

HD patients suffer mental dementia but the physical symptoms of the disease, an inability to control their physical movement resulting in flailing limbs often referred to as the Huntington Dance, are almost always the first noticeable symptoms. Nevertheless, HD sufferers experience a number of debilitating symptoms that originate in brain dysfunction (specifically destruction of striatum neurons, the substantia nigra, and hippocampus) and that alter their ability (capacity) to control their movements and affect their memory and executive function leading to problems in planning and higher order thought processes. So, while it is true that most people can control their arm movements, here is an example of individuals who progressively become worse and worse at doing so due to a neurophramacological disorder. There is currently no cure for HD but some medications that help treat it no doubt restore some of the capacity of these patients to control their movements. If a cure is found it would be difficult to say, as Christopher suggests of addiction, that the cure does not affect the capacity of HD patients to control what they once could not. I chose HD for its physiological set of symptoms but a similar example could easily be constructed for schizophrenia and a number of other mental health disorders (including ADHD and drug addiction). Importantly, cognitive control is a function of brain activity, activity that can become compromised as the set of experiment I will discuss next show.

An experiment conducted at UCLA (1) has shown that cocaine administrations reduced animals’ ability to change their behavior when environmental conditions called for it. Even more meaningful was the finding that once animals are exposed to daily doses of drugs, the way their learning systems function is altered even when the drugs themselves are no longer on board and even when the learning has nothing to do with drugs per se.

In the experiment, conducted by Dr. David Jentsch and colleagues, monkeys were given either a single dose (less than the equivalent of a tenth of a gram for a 150lb human) or repeated doses (1/8 to 1/4 of a gram equivalent once daily for 14 days) of cocaine. The task involved learning an initial association between the location of food in one of three boxes and then learning that the location of the food has changed. We call this task reversal learning since animals have to unlearn an established relationship to learn a new one.

Obviously, the animals want the food, and so the appropriate response once the location is changed is to stop picking the old location and move on to the new one that now holds the coveted food. This sort of thing happens all the time in life and indeed, during addiction it seems that people have trouble adjusting their behavior when taking drugs is no longer rewarding and is, in fact, even troublesome (as in leading to jail, family breakups, etc.).

In the experiment, animals exposed to cocaine had trouble (when compared to control animals that got an injection of saline water) learning to reverse their selection when tested 20 minutes after getting the drug, which is not surprising but still an example of how drug administration can causally affect an individual’s ability to make appropriate choices. As pointed above, the most interesting finding had to do with the animals that got a dose of cocaine every day for 14 days. Even after a full week of being off the drug, these animals showed an interesting effect that persisted for a month – while their ability to learn that initial food-box association, they had significant trouble changing their selection once the conditions changed. Remember, this effect was present with no cocaine in their system and with learning conditions that had nothing whatsoever to do with cocaine.

If that’s not direct evidence that having drugs in your system can alter the way your brain makes choices, I don’t know what is.

Another study conducted by Calu and colleagues with rats found similar (or even more pronounced) reversal learning problems after training the animals to take cocaine for themselves, clarifying that it is the taking of cocaine and not the method that causes the impairments.

Another entire set of studies has shown that stimuli (also known as cues or triggers) that have become associated with drugs can bring back long-forgotten drug-seeking behavior once they are reintroduced. This was shown in that Calu paper I mentioned above and in so many other articles that it would be wasteful to go through all the evidence here. Importantly, this evidence shows that drug associated cues direct behavior towards drug seeking in a way that biases behavior regardless of any underlying will. My own research has shown that animals who respond greatly to drugs (nicotine in our case) likely learn to integrate more of these triggers than animals who show a reduced response, indicating once again that these animals bias  their behavioral selection towards drug-seeking more than usual. While we have more studies to conduct, we believe that genetic differences relevant to dopamine and possibly other neurotransmitters important for learning (like Glutamate) are responsible for this effect.

While we can’t do these kinds of experiments with people (research approval committee’s just won’t let you give drugs to people who haven’t used them before), there is quite a bit of evidence showing an association between trouble in reversal learning and chronic drug use in humans (see citation 3 for example) as well as research showing very different brain activity among addicted individuals to drug-associated versus non-drug cues (like seeing a crack pipe versus a building). All this evidence suggests that drug users are different in the way they learn generally, and more specifically about drugs, than individuals not addicted to drugs. When it comes to genetics, we know quite a bit about the  association between substance abuse and specific genes, especially when it comes to dopamine function. As expected, genetic variation in dopamine receptor subtypes important in learning about rewards (D4 and D2) has been revealed to exist between addicts and non addicts. Without getting into the techniques and analysis methods involved in these genetic studies, their sheer number and the relationship between substance abuse and other impulse disorders points to a direct relationship between drug use disorders (and possibly other addictive disorders) and a reduced capacity to exert behavioral control. Less capacity for control is what researchers have found sets addict apart from non-addicts.

Summary, conclusions, and final thoughts

The toyota Prius is slow but efficientIn closing, there are undoubtedly imperfections about the ways we diagnose addiction (drug addiction and others). It would probably be nice if we could figure out a way to incorporate what we know about the continuous nature of the disorder with the need for clinical delineation of who requires addiction treatment and who doesn’t. Addiction researchers are far from the only ones who wonder about this question though (the same issues are relevant for schizophrenia, depression, and nearly every mental health disorder) and I am certain that better and better solutions will emerge.

However, the discussion of stigma in this context needs to allow us to discuss the reality of addiction without having to resort to blaming and counter-blaming. If I describe the Toyota Prius as being slow but incredibly efficient I am no more stigmatizing than if I describe a Ferrari as being incredibly fact but wasteful in terms of fuel. The same applies, or should apply, to health and mental health diagnoses – Just because an individual is less able to exert cognitive control over impulses should not by definition call into question their standing as a human being. We are complex machines and by improving our understanding of the nuts and bolts that make us function we can only, in my opinion, improve our ability to make the best use of our capabilities while understanding our relative strengths and weaknesses. Any other way of looking at it seems to me to be either wishful (I can do anything if I want it badly enough) or defeatist (I will never be anything because I’m not good at X) and neither seem like good options to me.

Citations:

1) Jentsch, Olausson, De La Garza, and Tylor (2002): Impairments of Reversal Learning and Response Perseveration after Repeated, Intermittent Cocaine Administrations to Monkeys. Neuropsychopharmacology, Volume 26, Issue 2, Pages 183-190

2) Calu et al (2007) Withdrawal from cocaine self-administration produces long-lasting deficits in orbitofrontal-dependent reversal learning in rats. Learning & Memory, 14, 325-328.

3) Some evidence in humans from Trevor Robbins’ group: Reversal deficits in current chronic cocaine users.

People, places, and things – How important are drug-related triggers for addiction relapse?

In cognitive behavioral therapy they’re a big part of the “Five W’s” = When, Where, Why, With, and What. In the various 12-step programs they’re simply referred to as “People, places, and things.” But no matter how you refer to them, drug-associated cues, or “triggers” as they are more commonly known, obviously play a big role in reminding addicted individuals about their drug-seeking behavior, and they are often enough to restart old behavior, even among those who have been abstinent for a while and especially when unprepared for their effect.

Different triggers to reactivate old behavior

Research on relapse (what researchers call reinstatement) has long shown that there are a number of things that can return a person, or an animal, to drug seeking after they have been abstinent for a while. Stress, small drug doses, and the presentation of triggers are all very capable of doing this, even after months of abstinence and likely even years. It’s probably not surprising that giving drugs to an abstinent person can make them want the drug again. In fact, I would venture to guess that most readers believe that this is the most powerful way to induce a relapse (assuming the initial exposure was out of a person’s control and doesn’t count).

Well, recent research suggests that in actuality, triggers, or those people, places, and things, might be more powerful or at least longer lasting relapse risks than even taking drugs!

Triggers, not drugs, are shown to be longest lasting relapse risk

Researchers in Japan trained mice to press a lever for meth, getting them to poke their nose into a hole 60 times for a total of 30 meth administration per three hour session. Every time they poked their nose in the right hole they got a shot of meth and a little light above their nose-poke hole went on (this will become the trigger in the end). Once they were doing this reliably the researchers took away the meth and the animals learned, within 10-20 days, that pressing the lever no longer got them a drug and reduced their number of presses to less than 15 presses per session.

After all this the researchers gave the mice an injection of meth 30 minutes before putting them back in the box – leading the mice to start pressing again for the drug even though in the previous session they has pretty much stopped pressing knowing that no drug was coming. Obviously, the drug injection caused the mice to relapse back into their drug seeking. But, as you can see from the figure below (on the left side, the right side shows that the mice didn’t poke their nose into a hole that did nothing as a control), this little trick only worked once, and the next time the mice were given a shot of meth before being put in the box (after once again being taken through extinction training teaching them that pressing the lever did nothing), they didn’t press the lever any more and just around not doing much.

For the following part of the study the researchers once again took the animals through extinction training (and once again the mice stopped pressing the lever for meth) and then in a following session reintroduced the little light that used to go on every time the mice originally got meth. Just like they did with the meth the animals immediately went back to pressing the lever like crazy, hoping that now that the light was back, so was their meth. Just like with the drug relapse experiment above, the researchers repeated this whole process over two months later, only this time, the little light managed to re-trigger the lever pressing again, unlike the one-trick-pony meth. Seeing this, the researchers went for broke and tried another run of this with the same animals, now following up five months after the last time the animals received meth when they pressed the lever. Again the little light got the animals to increase their pressing, only this time it was a little less impressive than the first two tries (but still significantly higher). All in all, the little light managed to restart the lever pressing by the mice three times and a full five month after the meth-relapse experiment had failed!!!

Conclusion, thoughts, and implications about triggers, relapse, and addiction

In a completely different article I’d written that researchers found a number of different patterns of relapse among alcoholics who went to rehab and that in fact, the vast majority of those who did relapse never went back to the kind of heavy drinking that characterized their earlier problem (see here for One is too many, a thousand not enough). While this research touches on a different aspect of relapse, it once again challenges our thinking about the crucial factors in relapse prevention among addicts. Everyone knows that triggers are important, but the fact that they are at least as powerful and apparently longer lasting dangers than even being re-exposed to the addictive drug is a novel one. Still, this isn’t very surprising given the very long-lasting impact of drugs of abuse (especially stimulants like crystal meth) on learning mechanisms. In my opinion, and based on my own experience, those changes are essentially permanent and the only thing that makes an ex-user less likely to run back to pressing that drug lever when being re-triggered 10 years later is the life they’ve built, the experience they have, and the training they’ve undergone in reacting to those triggers. As you can see from the graph above, if a person runs back to the drugs and actually starts using again on that first, second, or third exposure to a trigger they are likely to start the whole cycle again, possibly making it ever more difficult to escape the next time.

Obviously preventing trigger-induced relapse should be a major strategy of addiction treatment and indeed, from CBT relapse prevention strategies to groundbreaking medications that have been shown to be effective for relapse rate reduction (like Vivitrol, Buprenorphine, Bupropion, and more), there is quite a bit of effort going exactly that way.

Citation:

Yijin Yan, Kiyofumi Yamada, Atsumi Nitta  and Toshitaka Nabeshima (2007). Transient drug-primed but persistent cue-induced reinstatement of extinguished methamphetamine-seeking behavior in mice. Behavioral Brain Research, 177, 261-268.