Government Study Casts Doubt on Legal Definitions of Stoned Driving

Simulator tests also confirm that marijuana impairs drivers less than alcohol.


National Advanced Driving Simulator

A new study by researchers at the National Institute on Drug Abuse (NIDA) confirms that marijuana has a less dramatic effect on driving than alcohol does and casts further doubt on the standard that Colorado and Washington use for determining when someone is too stoned to drive. In the double-blind study, 18 occasional cannabis consumers—defined by marijuana use at least once in the previous month, but no more often than three days a week—took tests at the National Advanced Driving Simulator in Iowa City after vaping marijuana, drinking alcohol, or taking a placebo. The researchers, led by NIDA senior investigator Marilyn Huestis, found that a THC level of 13.1 nanograms per milliliter of blood had an impact on weaving within the lane (a.k.a. standard deviations of lateral position, or SDLP) similar to a blood alcohol concentration (BAC) of 0.08 percent, the cutoff used for most drivers in every state.

Notably, that THC level is more than twice as high as the five-nanogram standard established by Colorado and Washington when they legalized marijuana for recreational use. This result is especially striking because the subjects were occasional users. People who use cannabis every day, including patients who use it for medical reasons, may develop enough tolerance that they can drive safely at THC levels even higher than 13 nanograms. Yet according to Time, "Huestis believes that the 5 ug/L limit is not strict enough, particularly when you take into account those with low tolerance." That's another way of looking at it, I suppose, but the five-nanogram cutoff in practice means that many regular cannabis consumers can never legally drive, even when they're not impaired, which hardly seems sensible or fair.

Huestis and her colleagues found that alcohol and marijuana together had an additive effect on SDLP, with five nanograms of THC per milliliter plus a 0.05 percent BAC having an impact similar to a 0.08 percent BAC. In addition to SDLP, Huestis and her colleagues measured lateral acceleration (weaving speed) and lane departures per minute. Both were affected by alcohol but not by marijuana. This particular study did not consider any of the many other indicators used to measure driving impairment, although Huestis et al. plan to publish more on this subject to help illuminate the question of how best to define stoned driving.

Another result of this study that is relevant to enforcing such limits: Saliva tests accurately confirmed exposure to cannabis, but "with greater THC concentration variability than paired blood samples," which "poses challenges in concentration-based effects interpretation." In other words, saliva tests may be easier and less invasive, but they are less reliable in determining whether someone has exceeded whatever the legal threshold might be.

Also interesting, especially for people who have heard the frequent warnings from anti-pot prohibitionists about the dangers posed by increased potency: Most of the subjects, who were allowed to vape at will for 10 minutes, achieved about the same blood THC levels whether they were given weak cannabis (with a THC content of about 3 percent) or stronger cannabis (with a THC content of about 7 percent). As the researchers put it, "smokers can self-titrate cannabis dose to achieve desired pharmacological response."

Although this study provides additional evidence that marijuana poses less of a road hazard than drug warriors often claimTime health reporter Eliza Gray manages to obscure that point. "While drunk driving is on the decline in the U.S.," she writes, "driving after having smoked or otherwise consumer marijuana has become more common. According to the most recent national roadside survey from the National Highway Traffic Safety Administration of weekend nighttime drivers, 8.3 percent had some alcohol in their system and 12.6 percent tested positive for THC—up from 8.6 percent in 2007." Gray neglects to mention that a landmark NHTSA crash risk study released the same day found that, once the data were adjusted for confounding variables, cannabis consumption was not associated with an increased probability of getting into an accident.