U.S. drivers who tested positive for cannabis over a 10-year period had a 29 per cent higher risk of causing a fatal crash than motorists not taking the drug, a new Canadian-led study suggests.

Michel Bédard, director of public health at Lakehead University in Thunder Bay, Ont., and his team studied the effects of cannabis on driving using test results from 32,543 drivers in the U.S.

Between 1993 and 2003, 1,567 of those tested, or five per cent, were positive for cannibis but not alcohol, the researchers report in Wednesday's edition of the Canadian Journal of Public Health. The journal's publisher has launched a campaign against drug-impaired driving.

"Those who tested positive for cannabis had 29 per cent more risk of having committed a driving action that led to the crash than those who did not," Bédard told CBC News.

The higher risk existed after controlling for age, sex and previous driving record.

"It tells us that cannabis is not a safe substitute for alcohol, and I especially mean that for young people," said Bédard.

Perceptions about marijuana use differ

Dozens of studies have been done on whether being stoned affects driving ability, but none has the final word. Last year, the Canadian government introduced a bill to crack down on drug-impaired driving.

Surveys suggest that there is a perception among some young people that it is OK to smoke marijuana and drive, the researchers wrote. 

About seven per cent of Canadians reported using marijuana at least once in the previous year, and among Canadian undergraduate students, the figure jumped to 18 per cent during the academic year.

David Malmo-Levine, a long-time marijuana user in Vancouver, believes steady pot smokers are actually more cautious on the road.

"You're going to be relaxed, you're going to be focused, it might help fight fatigue and might help you deal with road rage and be able to mellow out," said Malmo-Levine.

Cannabis combines many of the properties of alcohol, tranquillizers, opiates and hallucinogens, and it is a sedative, analgesic and psychedlic that stimulates appetite and other systemic effects.

There are problems with Bedard's study, said pharmacology professor David Cook of the University of Alberta.  The study only found the presence of cannabis, not evidence of intoxication — a missing link in the research. 

"First of all, at what level of intoxication do we have to become concerned?" asked Cook.

Part of the problem is marijuana affects different people in different ways.  Some can tolerate a lot with little effect, while others become severely intoxicated on only a little.  Traces of cannabis can stay in the body for more than a week, well after the high has worn off.

The question of what dose of cannabis starts to pose a crash risk, such as  the 0.08 blood alcohol limit, remains to be answered, the team acknowledged.

Safe driving under the influence of cannabis relates not only to recreational users, but also for those taking the drug for medicinal purposes such as for glaucoma or cancer, they said.

Other studies suggest that roadside screening for cannabis using saliva are predictive of actual use.