A new randomized trial has the goods.

I visited a legal cannabis dispensary in Massachusetts a few years ago, mostly to see what the hype was about. There I am, knowing basically nothing about pot, as the gentlestoner behind the counter explained to me the differences between the various strains. Acapulco gold is buoyant and energizing, purple kush is sleepy, relaxed, dissociative. Here’s a strain that makes you feel nostalgic, here is one that helps you focus. It was as complicated and as oddly specific as a fancy wine tasting and, I had a feeling, about as reliable.

And while a strain that evokes memories of your first kiss is beyond the reach of modern cultivation practices, it is true that not all marijuana is created equal. It’s a plant, after all, and though delta 9 tetra-hydro-cannabinol (THC) is the chemical responsible for its euphoric effects, it is far from the only substance in there.

The second most important compound in cannabis is cannabidiol (CBD), and most people will tell you that CBD is the gentle Yin to THCs paranoiac yang. Hence your local ganja barrista reminding you that, if you don’t want all those anxiety-inducing side-effects of THC, grab a strain with a nice CBD balance.

But… is it true? A new study appearing in JAMA Network Open suggests, in fact, that its quite the opposite.

We are talking about this study from Dr. Austin Zamarripa and colleagues who clearly sit at the researcher cool kid table.

18 adults who had abstained from marijuana use for at least a month participated in this trial which is way more fun than anything we do at my lab at Yale. In random order, separated by at least a week, they ate some special brownies. 

Condition one was a control brownie, condition two a brownie with 20mg of THC, and condition three a brownie with 20mg of THC and 640mg of CBD. Participants were assigned each condition in random order, separated by at least a week.

A side-note on doses for those of you who, like me, are not totally weed-literate. 20mg of THC is about a third of what you might find in a typical joint these days (though it’s about double the THC content of a joint in the 70s – I believe the technical term is doobie).  640mg of CBD is a decent dose – as 5mg per kilogram is what some folks start with to achieve therapeutic effects.

Both THC and CBD interact with the cytochrome p450 system in the liver. This matters when you’re ingesting them instead of smoking them because you have first-pass metabolism to contend with. And, because of that p450 inhibition, it’s possible that CBD might actually increase the amount of THC that gets into your blood stream from the brownie, or gummy, or pizza sauce or whatever.

Let’s get to the results – starting with blood THC concentration. It’s not subtle. With CBD on board the THC concentration gets higher faster, with roughly double the area under the curve.

And, unsurprisingly, the subjective experience correlated with those higher levels. Individuals rated the “drug effect” higher with the combo. But, interestingly, the “pleasant” drug effect didn’t change much, while the unpleasant effects were substantially higher. No mitigation of THC anxiety here – quite the opposite. CBD made the anxiety worse.

Cognitive effects were equally profound. Scores on a digit symbol substitution test and a paced serial addition task were all substantially worse when CBD was mixed with THC.

And for those of you who want some more objective measures – check out heart rate – despite the purported “calming” nature of CBD, heart rates were way higher when individuals were exposed to both chemicals.

OK the picture here is quite clear, though the mechanism is not. At least when talking edibles, CBD enhances the effects of THC and not necessarily for the better. It may be that CBD is competing with some of the proteins that metabolize THC, prolonging its effects. CBD may also directly inhibit those enzymes. But whatever the case, I think we can safely say the myth that CBD makes the effects of THC more mild or more tolerable, is busted.

A version of this commentary first appeared on Medscape.com.