Opponent-Process Theory: Welcome to the dark side

May 25 2009 Published by under Addiction, Neuroscience

You people. You people and your REQUESTS. Requests to do things like blog more about opponent-process theory. Well. Sci hears you. She obeys. At least this time. And for all your drug addiction experts out there asking me to read Koob, I can assure you that I have read a LOT of Koob in my time. For those of you not necessarily familiar with the drug abuse lit, George Koob is considered one of the greatest minds in current drug abuse research, and has done a lot to conform the motivationally-focused opponent-process theory to the model of drug addiction that exists today. Guy even has a wikipedia entry! That's how you know you've hit the big time.

ResearchBlogging.org And so, Sci continues her discussion of opponent-process theory in this second installment, with many thanks to Koob and his co-author, Le Moal.
Remember this?
OP Theory1.png
You'll need it.

When we talk about addiction, we need to be very careful about WHAT, exactly, we are talking about. Addiction in the colloquial sense actually isn't addiction in the scientific sense in all cases.
So: drug addiction (excluding other addictions right now like food and sex and the internet) is characterized by a compulsion to seek and take the drug of choice, lack of control in limiting how much drug is taken, and a negative state when the drug is taken away. Keep in mind that addiction is also characterized by chronic relapse. The concept of chronic relapse is more than someone getting out of rehab and going back to the bottle, it's every time someone drinks again after saying they're not going to or drinks more than they said they were intending to.
So why is this important? It's important because it distinguishes between drug ABUSE and drug ADDICTION. After all, about 10% of the population is going to try cocaine in their lifetime. But only some of those 10% will become addicted to the drug. What is the difference between those people, and the people who don't become addicted? And what triggers the descent from controlled, occasional drug use to selling yourself for your next hit?
Obviously, there is more than one factor at work here. According to Koob and others, the main factors at work as someone become addicted to a drug include changes in your natural reward mechanisms, changes in your natural negative state mechanisms, and changes in your response to stress.

Natural Reward Mechanisms
(I am SO HIGH right now)
As you probably know by now from my many posts on dopamine, it's a very important neurotransmitter in the mechanisms of natural reward systems. It's hypothesized right now that every drug of abuse activates the dopamine system at some level, whether they be via changing the dopamine transporter at the terminals like cocaine and amphetamine, or by activating dopamine neurons in the ventral tegmental area like alcohol and opioids. Keep in mind, though, that dopamine isn't all there is. It's also thought that serotonin, the neurotransmitters GABA, opioid peptides, and endogenous (meaning homegrown cannabinoids that naturally occur in your brain) cannabinoids also play a role in reward mechanisms.
As we all know by now, drugs of abuse hijack these natural reward systems, and feel a lot better than anything that you could get naturally. And this means that chronic exposure to drugs of abuse can alter your natural reward systems, meaning that the same circuits (like the dopamine system, for example) that are involved in the positive effects of drug taking, can become disrupted, a change called within-system neuroadaptation. The main neuroadaptation that takes place occurs with something that scientists like to call the "brain reward threshold", which basically means how much stimulation is needed to cause your brain reward mechanisms to notice. When Sci thinks of 'thresholds', she thinks of stairs.

Think of your normal brain reward threshold as shallow. It doesn't take a lot of drug to make your brain go "dude, this feels GREAT!" However, after prolonged exposure, within-system neuroadaptations take place, and your reward threshold becomes steeper:

Obviously, it's going to require more drug on board to get your brain reward mechanisms in on the action. We like to call this tolerance. In terms of our old opponent-process graph, it's going to go from this:
OP Theory3.png
To this:
OP Theory5.png
It's the same thing, you require more drug to get the same positive feelings you had before. Your brain reward system has become tolerant. This has been found to occur in humans. Cocaine addicts have low-functioning dopamine systems, and we think that this not only contributes to tolerance to the rewarding effects, but also contributes to the negative feelings associated with withdrawal.
Antireward Systems (what a downer)
Some people seem surprised that we have antireward systems, but really, it's a no-brainer. You need to be aware when you don't feel good, otherwise you're not going to do anything to fix the situation. Unfortunately, this is another set of systems (and it's many systems, involving regulation of emotional pain, irritability, pain perception, malaise, and motivation) that changes during a drug addicted state. When you first take a drug, the overactivation of the reward system induces the recruitment of the antireward system, tying to bring you back down from your high. These antireward systems often persist longer than the drug is onboard, resulting in the acute withdrawal seen with things like alcohol hangovers and the depression experienced after cocaine use. To get a little funky with the graph, think of an alcohol hangover:
opponent process theory alcohol.png
The yellow is the positive effects, the red is the aversive effects. You can see that the positive effects of the drug peak before the antireward systems (in this case the effects of dehydration and nausea and other things induced by alcohol being on board) of your brain can get recruited, and that the antireward systems remain "on" after the drug effects are gone. The result? A hangover. In a longer term sense, there are other antireward mechanisms that get recruited and contribute to a negative state, things like increased anxiety and craving.
And these antireward systems become more sensitive over time, producing an aversive state that is triggered a lot faster this normal. This state of feeling crappy can actually drive drug abuse even more than the memory of how great the high was. Now you need the high produced by the drug, just to overcome your negative state and feel normal again. This contributes to the compulsive nature of drug addiction.
However, addicts don't always crave drug. There are situations which set off the drug craving mechanisms, both those of positive reinforcement (remembering how good it was to get high), and those of the antireward mechanisms. The first of these situations what we like to call cue-related. And cue is a pretty broad term. Basically, a cue is anything that can remind you of past drug use. And if, say, you're using the drug in your room, around your neighborhood, in your car, or with your friends, those cues can be really hard to avoid. Pretty soon, you're remembering how good it was to get high, and you've relapsed. But there's a second situation which gets recruited in drug addiction: stress.
The State of Stress (I'm so stressed OUT, I just need to RELAX)
Everyone's heard about it before. An addict comes out of rehab, swearing they've seen the light, they hate their drug, and they're never using again. And they mean it. They believe this very firmly. And then something happens. A family member gets sick, there's a divorce, they lose their job, something. And then...relapse. This is another system that gets out of whack following chronic exposure to addictive drugs, the stress system.
This isn't surprising. After all, negative states like withdrawal are...stressful. Negative states take a toll on the body and brain. Acute withdrawal is known to release chemicals in the brain associated with stress, such as corticotropic-releasing factor and norepinephrine. Not only that, these stress signals feed-forward to the antireward system, producing a negative state because you're stressed, which in turn increases feelings of stress, and so on. What this means in addicts is that stress can trigger craving by producing feelings associated with drug withdrawal. And the next thing you know, you NEED that drug. Just once.
And this is currently how we think opponent process theory works with regard to addiction. It is called the allostatic model, referring to the changes that take place in the brain over the time in a drug addict, as the brain takes a continuous reevaluation of the situation (how stressed are you? how do you feel right now?), and acts accordingly, increasing or decreasing craving for a drug and changing the emotional setpoint. And there are some pictures that show this really well.
First, we have picture 1, the drug naive individual:

You can see this looks a bit better that my opponent-process diagrams, but the principle is the same. Following the exposure to the drug in the naive person, reward systems are activated, things like dopamine and opioids, and net affect is positive. Then, the drug effect fades, and negative feelings set in, activating stress systems. You get a hangover, or feel slightly bad, but it's nothing extreme.
But now, see the drug-dependent individual:

This person is taking drug pretty continuously in order to avoid withdrawal. You can see at the beginning, on the left of the graph, their set point has been reset (0' instead of 0), they feel crappy, and they take drug. But the positive effects of the drug are only just enough to overcome the negative set point, and then, when withdrawal kicks in, it's more severe, and recovery is only partial, to the now, low set point where they always feel crappy.
But of course, some people aren't high ALL the time, and alcoholics aren't drunk ALL the time. Instead, they do something called drug-taking binges or episodes, and those end up looking like this:

Here you see that the person is set at a lower set point (0'), similar to figure B. But they aren't doing drug continuously. Instead, they are set off by a stressor, which drives them even LOWER, to set point 2 (0"). Driven by stress and negative feelings, they take drug. But the normal dose of drug doesn't even serve to get them high anymore. Rather, it just brings them up to a normal motivational state, which is quickly overcome by withdrawal, and they return to a motivational state no better than where they started.
But of course, many people taking drug have gone into long periods of abstinence, where they no longer crave the drug on a daily basis. How does relapse then occur?

Here's a person in protracted abstinence. They don't feel great, but they are still abstinent, and have been for a while. When they do take the drug (triggered either by drug cues or stress), they get high, and in fact get almost as high as they used to. But the antireward system states are still sensitized, and withdrawal is worse than you would see in a naive individual. Thus, they only return to a lower set point (0'). And then they either again become dependent to avoid withdrawal, or remain at the lower set point until a stressor comes along.
These pictures allow you to see how the set points of a persons emotions and motivations change over the course of drug addiction, and how these changes can contribute to continued use, or to relapse. And they highlight the fact that drug addiction is a chronic physiological disorder resulting in long term changes, and that these changed systems can be triggered by things in the environment or within the brain itself.
Koob, G., & Le Moal, M. (2008). Addiction and the Brain Antireward System Annual Review of Psychology, 59 (1), 29-53 DOI: 10.1146/annurev.psych.59.103006.093548

11 responses so far

  • Terry says:

    Hi Sci! you must get HAMMERED on this topic. Th basic flaw here is OP theory relies on behaviorism. Behaviorism relies on only that which we can observe OUTSIDE the black box. It therefore starts postulating what is happening inside the black box, based upon only outward behaviorally observable process, and based upon a tool design in that processes that happen internally are unknowable. Hopefully the problem there is obvious to most, even if not to die hard behaviorists.
    So a simple question, based on my theories: opponent processes or competing processes, with some substances and emotional states altering the bias in the competition though degeneration of areas, or altering of synaptic plasticity in areas above what can be attained in a non-drug state (biologically represented state dependent learning)?
    I think the latter fits far more accurately.
    Far too many attempt to find things things/concepts in the brain based upon concepts created prior to our current insights into the brain, and based in theories that said we can't know the internal workings of the brain. To me it seems at best modern phrenology, and more usually just folly. And your closing paragraph highlights the problems - it neglects "learning" and how learning about environmental interactions is altered, not just that "some things change, but how much could THAT matter over all?" It is the classic over simplification of simple neural triggers as opposed to altered group/population dynamic that therefore starts a different group/population dynamic that results in behavior emitted. There is no "take drug" cell any more than there is "grandma" cell in vision. but looking at the problems that simplistically has gotten non-vision science to where vision neuroscience was 30 years ago. Seems as scientists we should have learned from past mistakes, instead of repeating them. Just too bad repeating past mistakes from other areas has better odds at getting a grant than exploring new areas.

  • Pinus says:

    There is no implicit 'drug-taking cell' in the opponent process. It speaks to brain-wide systems that are engaged by extended exposure. And I think it is pretty clear that the idea is there is pathological learning in the circuts...salience is giving to normally non-salient cues, which leads to inappropriate emotional expression.
    Further, these systems do not stand alone...one of the more interesting ideas is that beyond this 'affective' endpoint, there is engagement of dorsal striatal circuitry which accounts for some of the compulsive aspects of drug-taking.

  • Scicurious says:

    Hi Terry (and Pinus)! Excellent comments! Let's play. 🙂
    I actually agree with you that alteration of synaptic plasticity has a great deal to do with what's going on in the brain in an addicted state. I think Koob would agree with you, too. Your brain is indeed learning in response to a drug state, and will change accordingly. This is not just due to the presence of drug, but to the cues and emotional states that precipitate drug use. I would strongly disagree, though, that the allostatic model of addiction precludes learning. Of course changes in reward circuitry involve changes in cell populations and their relationships to each other, and these are going to include changes in basal and stimulated neuronal firing rates. These change drastically not just due to the presence of drug, but also change the way the brain will react to subsequent environmental stimuli, connected with drug or otherwise.
    I would also disagree that the opponent-process theory is like modern phrenology. After all, the brain creates behavior, and when we analyze things like electrophysiological traces and pathways in the brain, we will learn very little unless we look for their outputs. Often, those outputs are behavioral. And of course we know there's not "take drug" cell. That was indeed simplistic thinking, but I think there are very few drug addiction specialists who would ascribe to that theory now.
    I think the allostatic model has been very important in terms of an angle from which to approach research in the field, and sure, the angle might end up being wrong, or overly simplistic, but it doesn't stop you from accruing valuable knowledge. In the absence of this model, based on behavior, what model should we have used? Another model may be more valid now, but we had to start somewhere.
    And Pinus, absolutely there are massive cortico-striatal circuits involved. In fact, if you grab the particular Koob review I'm referencing (the 2008 one with Le Moal) he spends a great deal of time talk about the switch over from "impulsive" drug taking to "compulsive", and I know many scientists who are now working on how this might take place.

  • Pinus says:

    I had figured you were aware of that. I mainly was trying to point out to Terry that there are layers of complexity to the theory.

  • Terry says:

    Pinus - thoroughly aware of the layers - no worries there. I have taught learning theory before...
    I still think one of the biggest issues in neuroscience is with taking these externally identified dichotomies (or whatever arbitrary level observers split behavior into) back into the brain, and saying "okay - I know what the output is, now lets find what part is responsible for each of the parts of that output", instead of looking at the machinery and resolving how they could be CREATING the behavior that we then artificially have defined.
    So instead of looking for discrete internal mechanism, it is more the case of needing to look at what simultaneous equation are being computed that then create the appearance of discrete competing equations.
    to me it is this taking preconceived categorized aspects of behavior and then looking for which 'bump' is responsible for that aspect that is modern phrenology.
    on a realated aside - have you ever read (or its related paper):
    McKearney JW (1969) Fixed-interval schedules of electric shock presentation: Extinction and recovery of performance under different shock intensities and fixed-interval durations. Journal of the Experimental Analysis of Behavior 1969; 12(2): 301-313.

  • leigh says:

    well i knew YOU had read Koob's papers, i was merely pointing out further reading for the benefit of your readers 🙂

  • Scicurious says:

    Terry: Hm. I think what we're really after here are differences in approach rather than actual doctrine. I see the idea of starting with a behavior (either defined as normal, or pathological) and attempting to trace it back to its mechanism in the brain as a more clinically-oriented outcome, and one more likely to lend itself to translational research. While I think that looking at the equations are very interesting and could be a very effective approach, I think that while you may find something, you won't necessarily be looking for something with a specific clinical syndrome in mind.
    While I think your approach has a great deal of value, I think it's a difference in perspective. I tend to approach my research from the angle of "there is a problem in the population, I want to know how this behavior comes about in an attempt to provide treatments for those who suffer from it", while yours might be more something like "I see these interesting patterns in neuronal firing rates, I wonder if these patterns are a sign of something that has a function". Am I right about this? I think that both are very valid methods of experimentation, but they lend themselves to very different ways of coming up with hypotheses and mechanistic theories.
    Pinus: w00t, I say. w00t.
    Leigh: You weren't the only one asking if I'd read Koob. 🙂 He's truly a giant in the field.

  • Scicurious says:

    Oh, right, sorry. Yes, I've read that paper, though it's been a while. Your thoughts?

  • Terry says:

    Sci - not quite - I actually see what you describe from my perspective as the current approach of the microbiology based research, so completely agree that that is not a very good way to approach *behavioral* neuroscience!
    I think the key word you use is mechanism - trace something back to the 'mechanism'. Depending on the approach and how you define mechanism is where I see far too many trying to do modern phrenology, as far too often mechanism is meant as which nucleus or which subtype of which receptor.
    to me, the current state of research mirrors ecology and biology, of 100 years or so ago. The is a problem in this ecosystem, so lets go in and introduce this species to 'help' or lets just eradicate species X to help, instead of looking at it as a system and saying how do we get things into a better balance across the ecosystem?
    I wanted to introduce the term neurecology, but got beat to that one, so then micro neuroecology, in that we need to look at a complex system (which is represented by simultaneous equations, not are equations...)
    Like in the good v bad centers in teh brain, and not a valence system that can signal in both directions. You make a good point in your "firing rate CHANGE" (emphasis added. it is also usually forgotten that the brain is a noisy state mechanism. So a neutral stimulus can be encoded by no change in firing, while aversive/rewarding aspects can be encoding in teh same units and system by increases or decreases in that base. It isn't a digital on/off signal (so can only represent one bit of information), but a variable firing analog signal that can code a great deal of information. How do what we do to the brain change its overall activity? In the case of learning - there are many disruptions that can cause memory failure/attenuation. loss of dopamine signal. loss of 5HT signal, loss of septal (et al.) input to the hippocampus, loss of hippocampus, loss of Ach effectivity , loss of GABA activity, disregulation of GABA/ACh balance, etc. etc. Obviously in certain cases there is a discreet target, but in the case of addiction and other perseverative and acquired behaviors, it is syndrome of related effects, so trying to find THE mechanism if you restrict mechanism to single thing, is proving intractable. But science and science training is currently in that mode. Works many times, but sometimes you have to try for a bigger picture and answer to discover the real issue - just like in ecology - or say micro-neuroecology. Obvioulsy there is much more to it, but I am trying to convey the essence here (ironic, isn't it?).
    similar to the lever press=shock=reward in the paper, and similarity to many phenomena (say like kids that commit bad acts to get bad attention). If there is a reward system, and a separate aversive system, how then does the aversive system take over the reward system? Occam's razor - a single bidirectional system, but a system that can process in parallel (simultaneous equations...).
    and I think you will see how much of this fits with Koob's work, but expands it.

  • Scicurious says:

    I like the idea of neuro-ecology (or whatever, micro-neuroecology...), and I think it might provide a better picture of the brain than we currently have. However, I don't think this makes the approach that many people are currently using in behavioral neuroscience invalid. The process of scientific inquiry is one that lends itself to constant reevaluation of hypotheses, so that, over time, what starts out as seemingly simple can be shown to be complex.
    For example, in the beginning of an inquiry into something, we may start with something like "x appears to be responsible for y behavior". Obviously, this is an overly simplistic explanation. Over time, you will start to see accruing evidence that "only increases in x cause y, and not in a dose-dependent fashion", "increases in x cause y under z conditions", "x is modulated by a and b, and a and b in turn can indirectly modulate y on their own". Although the method we have for approaching neuroscience right now may not be the most efficient, it is self-correcting, and will show the systems in increasing complexity over time. I think many people simply package their findings as "THE" mechanism because it looks better. Most of us know inside that the answer is probably more complex.
    So I think that a "neuroecology" kind of idea is a good one, but what way should one go about investigating it in all its complexity? In rainforest ecology, it's very hard to walk into a rainforest and see all the incredibly complex species interactions which are occurring all the time. Rather, scientists break down the system to its component parts, studying species by species, and how those particular species interact with others. Over time, the complexity of the system becomes apparent. I cannot think off the top of my head of a better way in which to approach a complex system, especially as, with the brain, we can't even walk in and look around.
    I agree that his does not have negative consequences for Koob's hypothesis, his idea of an allostatic model certainly allows for extremely complex changes to take place in both 'positive' and 'negative' systems.

  • […] usually think of dopamine linked more with things like reward or drug-addiction, but what dopamine actually does is more complex than that. Dopamine is involved in movement, for […]

Leave a Reply