(Reuters Health) - Americans seem to have developed a very rosy view of marijuana that isn’t backed up by science, a new study suggests.

A nationally representative online survey of 16,280 U.S. adults found that many ascribe health benefits to marijuana that haven’t been proven, researchers report in Annals of Internal Medicine.

“The American public has a much more favorable point of view than is warranted by the evidence,” said the study’s lead author, Dr. Salomeh Keyhani of the University of California, San Francisco. “Perhaps most concerning is that they think that it prevents health problems.”

While studies have shown that cannabis can help quiet seizures in children with hard to treat epilepsy, quell the nausea and vomiting that can accompany chemotherapy and sooth nerve pain, there’s no evidence that it can help with the vast majority of other medical conditions, Keyhani said. And yet, people think of it as a cure all, she added.

“Google marijuana benefits and the top hit is a Business Insider story titled ‘23 Health Benefits of Marijuana,’” Keyhani said.

The new survey looked at public perceptions of the benefits and risks of cannabis. Keyhani and her colleagues found that more than one third (36.9 percent) of respondents thought edible marijuana could prevent health problems, and more than a quarter (29.2 percent) believed smoking or vaping marijuana was protective.

While 76 percent believed marijuana could be addictive, 22.4 percent thought it had no addictive potential. Scientists have found that nine percent of people become addicted to cannabis.

Heartening news - that 92.1 percent thought marijuana was unsafe during pregnancy - was undercut by the fact that 7.3 percent thought it was somewhat or completely safe. “That 7 percent is a lot of people who think that it’s OK during pregnancy,” Keyhani said.

Also worrisome is the fact that 27.6 percent thought driving while under the influence of marijuana was safer than driving while drunk.

Keyhani suspects that a big part of the problem is that there is little, if any, regulation of cannabis advertising. “It’s a multibillion dollar industry,” she said. “That’s big business.”

Another part of the problem is that Americans seem to conflate legality with safety, said Dr. Ajay Wasan, a professor of anesthesiology and psychiatry at the University of Pittsburgh School of Medicine and vice chair for pain medicine at the University of Pittsburgh Medical Center.

“People perceive it as more like a dietary supplement than as a Schedule I drug with addiction potential,” said Wasan, who is unaffiliated with the new study.

Classification of marijuana as a Schedule I drug, which puts it in the same category as heroin, LSD and ecstasy, has gotten in the way of a lot of research, said Dr. Timothy Fong, a professor of addiction psychiatry at the University of California, Los Angeles, and a faculty member of the UCLA Cannabis Research Initiative.

When a substance is Schedule I, scientists need to jump through a lot of regulatory hoops to do research, Fong said. “We want to do more studies, but we can’t do a darn thing if the federal government handcuffs us,” he added. “This is the kind of study (Keyhani’s) that I think elevates the discussion. And it shows we have a long way to go.”

Fong suspects that with little actual data on cannabis, the public gets their information from “pop culture, television shows, celebrities, social media and cannabis conventions.”

One big reason people think cannabis is safe is there aren’t overdose deaths, said Fong. But that doesn’t mean there aren’t any deaths. “People get intoxicated and they fall off roofs or they get into automobile accidents,” Fong said. “Sometimes they show up in the hospital with a psychotic condition.”

Without research, laws are driven by politics, Fong said, adding “I like science based legislation.”