Do Drugs Cause Addiction?

Talk to anyone about drug policy, and the dangerousness of drugs comes to the fore. One of the major dangers mentioned is that drugs cause addiction. There are two lines of evidence for this: human epidemiology and animal experimentation. It seems on the surface to be an established fact based on science, but science is the process of finding the fewest assumptions that explain the observed phenomena. So let's take a look at the assumptions behind this "fact."

Animal Experimentation

We don't think it is ethical to experiment on humans, so we use animals that share some trait with humans to examine that trait. With addiction we're in luck because every mammal shares the same reward biology of the midbrain; it's a structure and system that's been giving mammals their edge for over 100 million years. Every mammal that's been looked at shares the same brain structures that calculate salience, reward, and hedonic tone. So we have good models for addiction to study.

In spite of all of that we don't actually have studies of animals with the human illness of addiction. That's because humans haven't been interested in studying addiction. Instead, what they study are substance use disorders. So all the studies in the literature, both human and animal, have not been studies of a primary brain illness called addiction, but rather studies of a small piece of addiction. There in lies the first assumption. We are ignoring phenomena other than drug use.

So, we'll study drug use in animals. That will still tell us a lot. But there's a problem. It's hard to get animals to use drugs. When it comes right down to it, they don't want to. But why the hell not? Drugs make you feel good, right? Why won't animals use them? Well, there's another assumption.

Actually it turns out that drugs don't make most normal mammals feel good. That feel good signal has to do with the dopamine signal to a part of the brain called the Nucleus Accumbens, commonly referred to as the reward center (I'll just call it the accumbens). Too little signal at the accumbens and we don't feel good; too much signal at the accumbens and we don't feel good in a different way. So if you give a mammal a drug that raises dopamine and they already have enough, they now have too much and don't feel good. So what's a scientist who wants to study drug use to do?

Well, I said normal mammals won't take drugs, so we need to make them not normal. Scientists have three ways. First, you can give them the drug against their will until the drug changes the brain enough to make them take the drug on their own. This sounds like a good idea because that's how most of us think addiction starts in humans anyway, but really, that's only how a very small percent of humans with addiction become ill. The second way is putting the animal under some stress that lowers the dopamine signal and makes them feel bad enough to like the drug. We can do that by confining them, restricting their movement, causing food insecurity, or putting them in subservient social roles. Sadly, we do those things to humans too. The final way for us to get lab animals to take drugs is to breed them to "like" drugs. The way you do this is you take 100 rats (rats are cheaper than monkeys but you could do this with any mammal) and offer them sugared water with various amounts of alcohol in it. The rats all vary, so some will have slightly lower dopamine signals than the others and those will be willing to drink the water with the more alcohol in it. Take the male and female that put up with the most alcohol and breed them. Take 100 of their offspring and do the alcohol and water thing again. Take the male and female of that group who tolerate the most alcohol and breed them. Repeat this process 14 generations. You now have a new strain of rats that will drink alcohol straight out of the bottle, but I wouldn't call them normal. Nature has done this experiment in us too, and about 10-20% of us are walking around genetically programmed to "like" drugs.

As an aside on public policy, wouldn't it be better to identify and educate the 20% of us at risk than spend billions trying to prevent the other 80% to not do something they don't want to do much of in the first place? But that's the government for you.

Human Epidemiology

I said there was another line of reasoning that leads people to believe that drugs cause addiction. It turns out that where a drug is more available or cheaper, it's used more, and so, one would think, that one thing leads to another. There's a lot of research on this. If you make cigarettes more expensive people smoke less. If you make alcohol more available, people drink more. If you make cannabis legal people use more pot. It seems obvious.

But we mistake correlation for causation in these studies, and we forget that they are all based on the assumption that people (a type of mammal) will freely use drugs when available under all circumstances. That's an assumption we know now that we can't make. So what are these studies measuring? 

Well, if you assumed instead that 10-20% of the population had addiction and the need for increased dopamine signal, and you knew that some behaviors and chemicals increased that signal, then you'd expect that most of those chemicals would be consumed by that 10-20%. And if you (or nature) restricted those chemicals, you'd expect that any relaxation of those restrictions would increase use until that 10-20% got everything they needed. You could also expect that if you restricted a commonly used chemical more, then that 10-20% would go find the next best cheapest alternative to get the dopamine. 

Well that's exactly what happens. When we smoke and drink less, we use more high fructose corn syrup. When we stop using as much of that we'll take more opioids. This goes on and on and on and has for more than 100 years of us picking an "evil" drug and knocking it down. Our congressmen get involved and pass laws to interdict, prohibit, etc this drug to solve the problem. After years of work and billions of dollars the drug isn't as prevalent as it once was. The Congress tells itself that it solved a problem, the epidemiologists confirm their success, and off they go to the congratulatory cocktail party. Then someone from the New York Times comes in a few months later and tells them we have a new "drug problem" and we start the whole thing all over again.

Epidemiology only shows a correlation between drugs and cost when we limit the discussion to "a drug." No one has done any studies on the epidemiology of addiction as a whole, and until they do, it's not a line of inquiry that should inform our policy.

The Way Forward

So how can we proceed? We should examine the assumptions behind the studies we do have so as not to over generalize them. Facts that are true in rats confined to cages may not be true in freely moving humans in a free society. Once we've done that, we should take a look at the problem we want to actually solve and start asking the questions we need to solve it. 

I truly believe we can end addiction as a problem in American life in my working lifetime, but we won't get there believing drugs cause addiction.