A Brownie, a Drink, and a Driving Test That Doesn't Work: The Week's Most Important Cannabis Science

A Brownie, a Drink, and a Driving Test That Doesn't Work: The Week's Most Important Cannabis Science

Science can sometimes feel like it moves slower than the culture around it. For cannabis, that gap has been a defining feature of the last decade — millions of Americans using a substance while the research needed to understand its risks and benefits lags years, sometimes decades, behind the conversation. This past week produced a cluster of studies that narrow that gap in meaningful ways, with findings that will matter for consumers, patients, policymakers, and anyone who has ever wondered whether the roadside sobriety test actually catches impaired cannabis users.

The most urgent finding came from Johns Hopkins Medicine, which amplified a study published May 1 in JAMA Network Open with an updated public press release this week. The study's central finding is both simple and alarming: combining cannabis edibles with alcohol — even at relatively low doses of each — produces driving impairment that exceeds what either substance causes on its own, and that impairment is largely invisible to the standard field sobriety tests currently used by law enforcement across the country.

The study's design was careful. Participants were given either a cannabis-infused brownie containing 10 or 25 milligrams of THC, or a placebo, then given an alcoholic or non-alcoholic beverage calibrated to bring their blood alcohol content to either 0.05% or 0.08% — the lower threshold used in some states and the standard federal legal limit, respectively. They were then evaluated using standard field sobriety tests and driving simulation tasks. The results showed clear synergistic impairment: the combination of THC and alcohol degraded driving performance far more than a simple additive model would predict, particularly at the 0.05% alcohol level — meaning people who would test legally sober in most states were exhibiting meaningful impairment when cannabis was also in their system.

The detection gap is what makes this finding most practically significant. Standard field sobriety tests failed to flag most cannabis-impaired drivers, regardless of whether they had also consumed alcohol. The legal framework built around drunk driving — the breathalyzer, the 0.08% threshold, the standardized tests — was designed for a world where alcohol was the primary impairment concern. It is increasingly not calibrated for the world we actually live in, where roughly one in five American adults report using cannabis, and where edibles, which have delayed onset and can be consumed hours before driving, have become one of the fastest-growing product categories. Policymakers at the state and federal level have known this gap exists for years; the Johns Hopkins data makes the cost of that gap harder to ignore.

The week's science conversation also circled back to a landmark study from earlier in 2026 that continues to generate discussion. A systematic review published in The Lancet Psychiatry, conducted by researchers in Australia who analyzed more than 50 clinical trials, found no evidence that medical cannabinoids effectively treat the conditions their users most commonly cite as reasons for using them: anxiety, depression, and PTSD. The review found no significant effect for those conditions, nor for psychotic disorders, OCD, opioid use disorder, or anorexia nervosa.

The findings don't mean cannabis has no medical value. The review did find some evidence for cannabinoid efficacy in managing cannabis use disorder itself (using THC and CBD combinations, counterintuitively), in autism spectrum disorder, insomnia, and Tourette's syndrome. But the overall quality of the evidence was rated low, and the gap between what patients report and what clinical trials have been able to demonstrate remains substantial. That gap has multiple explanations — decades of federal restrictions on cannabis research have made high-quality randomized controlled trials scarce — but it's a gap the industry and advocacy community must grapple with honestly if the push for medical legitimacy is to succeed on scientific rather than purely political grounds. The Center for Medicinal Cannabis Research held an Investigators' Meeting focused on cannabis and mental health on May 28, suggesting the research community is trying to address exactly this deficit.

A third study, published earlier in May by researchers affiliated with University of Utah Health and the University of Colorado Boulder, offers a window into the real-world decision-making of cannabis consumers. The study found that older adults — a fast-growing demographic in cannabis retail — are predominantly turning to edibles seeking alternatives to pharmaceutical drugs for sleep problems, chronic pain, and anxiety. Most of them are making these decisions based on word of mouth from friends or family, not guidance from a health care provider. The finding is a pointed reminder that consumers are making meaningful health decisions in an information environment that is not yet equal to the task of supporting them.

Taken together, this week's research underscores a persistent tension at the heart of cannabis in 2026: the substance is already woven into the fabric of American life, shaping how millions of people sleep, manage pain, unwind, and, apparently, get behind the wheel. The science is catching up, but unevenly. Some of what it finds confirms what users already suspected. Some of it challenges assumptions that have become culturally settled. All of it deserves to be part of a more honest public conversation than the one we've been having.