CHICAGO — Medical marijuana has not been proven to work for many illnesses that state laws have approved it for, according to the first comprehensive analysis of research on its potential benefits.

The strongest evidence is for chronic pain and for muscle stiffness in multiple sclerosis, according to the review, which evaluated 79 studies involving more than 6,000 patients. Evidence was weak for many other conditions, including anxiety, sleep disorders, and Tourette’s syndrome and the authors recommend more research.

The analysis is among several medical marijuana articles published Tuesday in the Journal of the American Medical Association. They include a small study suggesting that many brand labels for edible marijuana products list inaccurate amounts of active ingredients. More than half of brands tested had much lower amounts than labeled, meaning users might get no effect.

Highlights from the journal:

The analysis

The researchers pooled results from studies that tested marijuana against placebos, usual care or no treatment. That’s the most rigorous kind of research but many studies found no conclusive evidence of any benefit. Side effects were common and included dizziness, dry mouth and sleepiness. A less extensive research review in the journal found similar results.

It’s possible medical marijuana could have widespread benefits, but strong evidence from high-quality studies is lacking, authors of both articles say.

Edible products

Researchers evaluated 47 brands of medical marijuana products, including candy, baked goods and drinks, bought at dispensaries in Los Angeles, San Francisco and Seattle.

Independent laboratory testing for THC, marijuana’s leading active ingredient, found accurate amounts listed on labels for just 13 of 75 products. Almost 1 in 4 had higher amounts than labeled, which could cause ill effects. Most had lower-than-listed amounts. There were similar findings for another active ingredient. Products were not identified by name.

Johns Hopkins University researcher Ryan Vandrey, the lead author, said he was surprised so many labels were inaccurate. The researchers note, however, that the results may not be the same in other locations.

Marijuana laws

Twenty-three states and Washington, D.C., have laws permitting medical marijuana use. Approved conditions vary but include Alzheimer’s disease, epilepsy, glaucoma, kidney disease, lupus and Parkinson’s disease.

An editorial in the journal says approval in many states has been based on poor-quality studies, patients’ testimonials or other nonscientific evidence.

Marijuana is illegal under federal law and some scientists say research has been stymied by government hurdles, including a declaration that marijuana is a controlled substance with no accepted medical use.

But in a notice published Tuesday in the Federal Register, the Department of Health and Human Services made it a little easier for privately funded medical marijuana research to get approved. The department said that a federal Public Health Service review of research proposals is no longer necessary because it duplicates a required review by the Food and Drug Administration.

The future

Colorado, one of a few states where recreational marijuana use is legal, has pledged more than $8 million in state funds for several studies on the drug’s potential medical benefits, including whether it can reduce veterans’ symptoms of post-traumatic stress disorder. That study may begin recruiting participants later this year, said Vandrey, one of that study’s leaders.

The recommendations

The editorial by two Yale University psychiatrists suggests enthusiasm for medical marijuana has outpaced rigorous research and says widespread use should wait for better evidence. Federal and state governments should support and encourage such research, the editorial says.