Matt Vogel, an assistant professor of criminology and criminal justice, gave a presentation titled “Empirical Reality of Methodological Uncertainty: The Promises and Perils of the Gateway Hypothesis in Today’s Drug Landscapes” at last month’s Addiction Science Conference hosted by the Missouri Institute of Mental Health. (Photo by August Jennewein)

The Missouri Institute of Mental Health at the University of Missouri–St. Louis last month hosted the Addiction Science Conference centered on “The Science and Policy of Cannabis” at the Clayton Plaza Hotel.

MIMH helped organize a similar conference last year examining opioid overdose through Missouri Opioid-Heroin Overdose Prevention and Education Project.

This year’s event brought together researchers and advocates in such fields as psychiatry, pharmacology and public health from around the region and as far away as California to share their insights with Assistant Professor Rachel Winograd serving as the moderator.

Matt Vogel, an assistant professor of criminology and criminal justice at UMSL, also served as one of the presenters. His talk, titled “Empirical Reality or Methodological Uncertainty: The Promises and Perils of the Gateway Hypothesis in Today’s Drug Landscapes,” examined the idea that marijuana use serves as a gateway to other, more illicit drugs.

For the latest installment of the “Ask an Expert” series, Vogel sat down with UMSL Daily to discuss that presentation and research at the heart of the broader debate about marijuana in the United States.

Why did you make the Gateway Hypothesis a central theme of your presentation?

Rachel asked me to speak, and when I agreed, she said, “Great, that’s awesome. And by the way, here’s the title of your talk.” So my back story is not that interesting, but the talk itself was. When she asked me, I said, “I don’t know, are we even talking about the Gateway Hypothesis?” D.A.R.E. was proven ineffective 15 years ago. We sort of realized the War on Drugs wasn’t working. But then I started doing a little bit of research and realized that it’s been brought up again in the past year, including by Attorney General Jeff Sessions.

How did your presentation fit into the rest of the conference?

It fit in, sort of. I was the only criminologist there, and I’m really not a drug scholar in the way these folks were. I have published a couple of papers that look at marijuana use among high school kids. They’ve spent their entire career studying this. I kind of came as an outsider. But I think it went well.

The conference is all about the science of cannabis in today’s drug landscape. So we have two things going on in the country. We have states moving toward recreational marijuana or legalized medicinal marijuana, and then we have an opiate crisis. It’s really easy to see those things could be related – cause and effect.

The Gateway Hypothesis is essentially one of causality. Kids start smoking marijuana. We know not every kid who smokes pot is going to go on and become a heroin addict, but it hypothesizes that there’s something causal in nature that without smoking marijuana, these kids never would have experimented with heroin.

I said, “Well, let me look at the evidence.” And in doing some background research, what I did find was a lot of caveats. The Gateway Hypothesis is true but only under certain conditions – only when kids start using young, only when they use a lot and so on. So that’s where I started.

The Gateway Hypothesis is an idea that predates empirical research, right?

You know Harry Anslinger? He was the one that first proposed the Gateway Hypothesis when the United States was passing the first marijuana legislation in 1937. Anslinger was the director of the Federal Bureau of Narcotics, and he was testifying before the Ways and Means Committee, and he actually first proposes this notion of a Gateway Hypothesis, that marijuana leads to hardcore drug addiction. That would have been in 1936. The scientific evidence for the Gateway Hypothesis didn’t come until 1975. That was the first piece of information. Denise Kandel published it in the journal “Science.”

Did you find evidence to support the Gateway Hypothesis?

Absolutely. That’s what I took as my point of departure for the talk, and I said, what if we find longitudinal data, so we follow kids over time, we get them when they’re young enough that they haven’t started smoking pot yet, we identify when some of them have started smoking pot and then follow them out through their 30s and see what happens. If you look at those kids – the early onset, heavy use kids – they’re a lot more likely to start using other drugs, meth, cocaine, to keep smoking marijuana at age 18, at age 28, and they have a lot more problems.

But it’s not all of the evidence?

It’s not all of the evidence, and it’s somewhat superficial. The approach I took is at the end of the day, we need to consider not only whether or not these kids start using marijuana at a young age but, more importantly, who are the kids that start using marijuana at a young age? Do they differ from the rest of the general population from the starting gate? As it turns out, the answer’s yes and in meaningful ways.

How do people go about determining those differences?

The way I went about it, I used an empirical technique called propensity score analysis. I took marijuana use as a treatment, and because we can’t use a randomized control trial – I can’t randomly assign a 12-year-old to smoke marijuana and another 12-year-old to not smoke and see what happens – so I take as my point of departure the fact that some 12-year-olds start smoking marijuana when they’re 12. Others don’t, and that’s my comparison group.

I then assess based on background characteristics the probability that a kid started smoking marijuana, so I look at things like: Do they have friends who use drugs? Are they from an intact family? Do they live in a nice neighborhood? Have they been victimized? Do they have adverse childhood experiences? And so on.

I use that to predict whether or not the kid starts smoking pot, and then that generates a propensity score – a predictive probability. And then I just look for kids who have similar scores but just by coincidence haven’t started smoking pot yet for whatever reason. When I make that my comparison group, then we can start looking at these differences and we can start isolating a causal effect.

When you look at the kids who are more likely to start using or who do start using when they’re 12, what are some of the other traits that are often present?

Kids who start smoking marijuana at a young age are more likely to have tried cigarettes, to have consumed alcohol, to report having drugs available in their home – which makes sense, because of opportunities – to have adverse childhood experiences. They’re more likely to report being victimized as a young kid, to being bullied by their friends and so on and so forth. There’s neighborhood-level predictors, so kids from poorer areas tend to be higher risk.

Why didn’t people initially see the holes in the Gateway Hypothesis?

I don’t have a good answer to that. I would say this idea of selection bias and treatment effects in the social sciences – so the realm in which I work in – is something we’ve really started appreciating in the last 15, 20 years, and we have had models that allow us to do these sorts of things maybe in the last 10. So it’s relatively new.

Since the push in recent years to decriminalize marijuana in many states, has there been research on if it’s led to an increase in the use of other drugs in those states?

There’s very, very recent research, but there have been a couple of articles that have come out just in the last couple of weeks now. There was one study in Washington state, and I think in Vancouver as well, that suggested street-level heroin addicts were actually much more likely to try to wean themselves off using marijuana. When they started decriminalizing and legalizing marijuana, they saw rates of opiate abuse go down.

The other argument is in states like Colorado and states that have legalized medicinal marijuana, you see lower rates of opiate use. The idea there is that people, when given the choice, would actually prefer to medicate with something like marijuana that’s not addicting than with something harder. In states like Colorado – the first one to legalize recreational marijuana – the prevalence of high school marijuana use went down, not up, because they’re investing the tax dollars they generate from it in prevention. You see the same thing in the Netherlands too. The Dutch have the lowest rate of cannabis consumption in the western world among countries for which they collect data.