Mixing cannabis and alcohol simultaneously doubles drunk driving risk and triples social consequences
Simultaneous versus concurrent use of alcohol and cannabis in the National Alcohol Survey.
Bottom Line
People who use cannabis and alcohol at the same time have 2-3x higher odds of drunk driving and social consequences, while those who use both separately show much lower risk.
Why It Matters
First study to distinguish simultaneous from concurrent alcohol-cannabis co-use in the general population, revealing that mixing is dramatically riskier than using both but separately.
The Backstory
Cannabis is the most commonly used drug among people who drink. Everyone knows this. What nobody had measured — until Meenakshi Subbaraman asked the right question — was the difference between people who use cannabis and alcohol at the same time versus people who use both but keep them separate.
The distinction turned out to be enormous. And it reframed the entire debate about whether legal cannabis helps or hurts public health by changing how we think about alcohol.
The Question Nobody Had Asked
Researchers studying cannabis and alcohol had been lumping all co-users together. If you drank and you also used cannabis, you were a "co-user" — regardless of whether you lit a joint while drinking a beer or used cannabis on weeknights and drank on weekends.
Subbaraman, a researcher at the Alcohol Research Group (Public Health Institute) in Emeryville, California, recognized that this was like studying food by combining everyone who eats breakfast and dinner into one category. The pattern matters as much as the behavior.
Three Populations, Three Risk Profiles
65.7 vs. 13.9
days per year of heavy drinking (5+ drinks) among simultaneous cannabis-alcohol users versus alcohol-only users. Simultaneous users aren't just using cannabis — they're drinking dramatically more than everyone else. Nearly five times as many heavy drinking days per year.
This number challenges a common assumption: that people who combine cannabis with alcohol are simply doing two things at once. In reality, simultaneous users are a distinct, higher-risk population characterized by heavier overall substance use patterns.
Subbaraman & Kerr (2015)
The Risk That Doubles
After controlling for demographics and drinking patterns, the findings were unambiguous:
The Substitution Question
The individual-level data tells one story: mixing is dangerous. But Subbaraman's broader body of work — including a comprehensive review of 39 studies on alcohol-cannabis substitution and complementarity — reveals a more nuanced population-level picture.
Why This Matters for Policy
The substitution question has enormous implications for legalization debates. If legal cannabis replaces even a fraction of alcohol consumption, the net public health impact could be positive — because alcohol causes far more harm than cannabis on every systematic assessment.
But Subbaraman's data adds a critical caveat: the benefit depends on whether people substitute or stack. The policy goal should be to encourage substitution (cannabis instead of alcohol) while discouraging simultaneous use (cannabis plus alcohol). This argues for public health messaging that doesn't just warn about cannabis risks in isolation, but explicitly addresses the higher risks of combining the two.
For individuals, the message is clear: if you're going to use cannabis and you also drink, keeping them separate is dramatically safer than mixing. The pharmacological interaction — alcohol increases THC absorption, producing the "spins" and greater impairment — makes simultaneous use one of the riskiest patterns in recreational substance use.
Key Takeaways
Simultaneous versus concurrent use of alcohol and cannabis in the National Alcohol Survey
Subbaraman MS, Kerr WC () · Alcoholism: Clinical and Experimental Research
Frequently Asked Questions
Cite this study
Subbaraman, Meenakshi S; Kerr, William C. (2015). Simultaneous versus concurrent use of alcohol and cannabis in the National Alcohol Survey.. Alcoholism: Clinical and Experimental Research, 39(5), 872-879. https://doi.org/10.1111/acer.12698