Twenty-eight states and the District of Columbia have legalized marijuana for a variety of medical uses, and eight of those states plus the district have also legalized it for recreational use.

Now, a new report from the National Academies of Sciences, Engineering, and Medicine offers a rigorous review of scientific research published since 1999 about what is known about the health impacts of cannabis and cannabis-derived products – such as marijuana and active chemical compounds known as cannabinoids – ranging from their therapeutic effects to their risks for causing certain cancers, diseases, mental health disorders, and injuries.

The committee that carried out the study and wrote the report considered more than 10,000 scientific abstracts to reach its nearly 100 conclusions. The committee also proposed ways to expand and improve the quality of cannabis research efforts, enhance data collection efforts to support the advancement of research, and address the current barriers to cannabis research.

The committee found evidence to support that patients who were treated with cannabis or cannabinoids were more likely to experience a significant reduction in pain symptoms. For adults with multiple sclerosis-related muscle spasms, there was substantial evidence that short-term use of certain “oral cannabinoids” – man-made, cannabinoid-based medications that are orally ingested – improved their reported symptoms. Furthermore, in adults with chemotherapy-induced nausea and vomiting, there was conclusive evidence that certain oral cannabinoids were effective in preventing and treating those ailments.

But the report dismisses most of the drug’s other supposedly ‘medical benefits’ as unproven. Crucially, the researchers concluded there is not enough research to say whether marijuana effectively treats epilepsy – one of the most widely-recognized reasons for cannabis prescriptions.

The report also casts doubt on using cannabis to treat cancers, irritable bowel syndrome, or certain symptoms of Parkinson’s disease, or helping people beat addictions.

Turning to potential harms, the committee concluded:

1) Strong evidence links marijuana use to the risk of developing schizophrenia and other causes of psychosis, with the highest risk among the most frequent users.
2) Some evidence suggests a small increased risk for developing depressive disorders, but there’s no evidence either way on whether it affects the course or symptoms of such disorders, or the risk of developing post-traumatic stress disorder.
3) There’s strong evidence that using marijuana increases the risk of a traffic accident, but no clear indication that it promotes workplace accidents or injuries, or death from a marijuana overdose.
4) There’s only weak evidence for the idea that it hurts school achievement, raises unemployment rates or harms social functioning.
5) For pregnant women who smoke pot, there’s strong evidence of reduced birth weight but only weak evidence of any effect on pregnancy complications for the mother, or an infant’s need for admission to intensive care. There’s not enough evidence to show whether it affects the child later, like sudden infant death syndrome or substance use.
6) Some evidence suggests there’s no link to lung cancer in marijuana smokers. But there’s no evidence, or insufficient evidence, to support or rebut any link to developing cancers of the prostate, cervix, bladder, or esophagus.
7) Substantial evidence links pot smoking to worse respiratory symptoms and more frequent episodes of chronic bronchitis.
8) There’s weak evidence that suggests smoking marijuana can trigger a heart attack, especially for people at high risk of heart disease. But there’s no evidence either way on whether chronic use affects a person’s risk of a heart attack.
8) Some evidence suggests a link between using marijuana and developing a dependence on or abuse of other substances, including alcohol, tobacco and illicit drugs.

Currently, cannabis is the most popular illicit drug in the United States, in terms of past-month users. Based on a recent nationwide survey, 22.2 million Americans ages 12 and older reported using cannabis in the past 30 days. This survey also reports that 90 percent of adult cannabis users in the United States said their primary use was recreational, with about 10 percent reporting use solely for medical purposes. Around 36 percent reported mixed medical and recreational use. In addition, between 2002 and 2015, the percentage of past-month cannabis users in the U.S. population ages 12 and older has increased steadily from 6.2 percent to 8.3 percent.