THURSDAY, Sept. 15, 2016 (HealthDay News) -- A new study of drivers who died in auto accidents suggests people in states with medical marijuana laws may be using fewer opioid painkillers, the study authors contend.

"After the implementation of a medical marijuana law, there appears to be less opioid use, at least among young and middle-aged adults," study lead author June Kim said. He's a graduate student in epidemiology at Columbia University Mailman School of Public Health in New York City.

However, two addiction experts not involved with the research were critical of the methodology used, saying the study authors did not prove the point they were trying to make.

The study sought to understand how laws allowing the medical use of marijuana -- now legal in 25 states and Washington, D.C. -- might affect the use of opioid painkillers such as oxycodone (OxyContin) and hydrocodone (used in Vicodin and Vicoprofen).

Medical officials have linked the abuse of these painkillers to widespread addiction and overdose deaths.

"A study that came out a few years ago suggested that states with medical marijuana laws have a reduced rate of opioid overdoses," Kim said. "I thought that if these laws were actually reducing overdoses, we should expect to see a similar reduction in opioid use."

The researchers looked for signs of trends in an unusual place: traffic fatalities. The researchers looked at the records of people who died in car crashes to see if they had tested positive for opioid use. The accidents took place in 18 states from 1999 to 2013.

There were more than 68,000 traffic fatalities included in the study. Forty-two percent of accidents occurred in states with medical marijuana laws that were up and running. About one-quarter happened in states that had passed medical marijuana laws, but hadn't yet implemented them. And 33 percent of the accidents occurred in states without medical marijuana laws.

About 1 percent to 8 percent of drivers tested positive for opioid painkillers, the study reported.

The study didn't look at whether drivers had marijuana in their systems, since not all the states tested for it, Kim said.

The researchers found that far fewer drivers in states with active medical marijuana laws died with opioids in their system.

"If you are a driver aged 21 to 40, you were about half as likely to test positive for opioids if you crashed in a state with a medical marijuana law versus if you had crashed in a state before a law was implemented," Kim said.

The study authors stressed that it's not clear if the opioid painkillers -- or, for that matter, marijuana -- contributed to any of the car accidents.

Kim said the study's findings suggest people are turning to legal pot for pain relief instead of opioid painkillers. However, the study did not prove medical marijuana was being used in place of opioids.

Jason Hockenberry is an associate professor and director of graduate studies with the Department of Health Policy and Management at Emory University in Atlanta. He was critical of the study, calling it "a bit of a mess."

A variety of explanations for the findings are possible, Hockenberry said. State policies regarding opioids could also be at play, he pointed out.

He also noted that there's no information about whether the drivers were using marijuana.

Hockenberry added that "any benefits of medical marijuana need to be balanced against the negative effects of marijuana, which are not trivial. Our own work finds that abuse of marijuana and dependency are increasing in states with medical marijuana laws."

Brendan Saloner is an assistant professor of Health Policy and Management at Johns Hopkins Bloomberg School of Public Health in Baltimore. He also studies drug addiction.

He said the sample in the study -- drivers who died in car accidents -- "is not necessarily generalizable to the population as a whole."

There's also a question of the wider effect of medical marijuana laws, Saloner said.

"On the one hand, they could very well reduce harmful opioid use. But on the other hand, they could have offsetting effects on other risky behaviors including impaired driving," he noted.

Still, Saloner said, his team's own research has "documented that states passing medical cannabis laws experienced a 25 percent reduction in fatal opioid overdoses relative to states not implementing these laws. Other studies since ours have reached similar conclusions."

The study appears Sept. 15 in the American Journal of Public Health.

More information

For more about opioid painkillers, try the U.S. National Institute on Drug Abuse.