Using marijuana seems to increase the chance of becoming psychotic, researchers report in an analysis of past research that reignites the issue of whether pot is dangerous.
The new review suggests that even infrequent use could raise the small but real risk of this serious mental illness by 40 percent.
Doctors have long suspected a connection and say the latest findings underline the need to highlight marijuana's long-term risks. The research, paid for by the British Health Department, is being published Friday in medical journal The Lancet.
"The available evidence now suggests that cannabis is not as harmless as many people think," said Dr. Stanley Zammit, one of the study's authors and a lecturer in the department of psychological medicine at Cardiff University.
The researchers said they couldn't prove that marijuana use itself increases the risk of psychosis, a category of several disorders with schizophrenia being the most commonly known.
There could be something else about marijuana users, "like their tendency to use other drugs or certain personality traits, that could be causing the psychoses," Zammit said.
Marijuana is the most frequently used illegal substance in many countries, including the United Kingdom and the United States. About 20 percent of young adults report using it at least once a week, according to government statistics.
Zammit and colleagues from the University of Bristol, Imperial College and Cambridge University examined 35 studies that tracked tens of thousands of people for periods ranging from one year to 27 years to examine the effect of marijuana on mental health.
They looked for psychotic illnesses as well as cognitive disorders including delusions and hallucinations, bipolar disorder, depression, anxiety, neuroses and suicidal tendencies.
They found that people who used marijuana had roughly a 40 percent higher chance of developing a psychotic disorder later in life. The overall risk remains very low.
For example, Zammit said the risk of developing schizophrenia for most people is less than 1 percent. The prevalence of schizophrenia is believed to be about five in 1,000 people. But because of the drug's wide popularity, the researchers estimate that about 800 new cases of psychosis could be prevented by reducing marijuana use.
The scientists found a more disturbing outlook for "heavy users" of pot, those who used it daily or weekly: Their risk for psychosis jumped to a range of 50 percent to 200 percent.
One doctor noted that people with a history of mental illness in their families could be at higher risk. For them, marijuana use "could unmask the underlying schizophrenia," said Dr. Deepak Cyril D'Souza, an associate professor of psychiatry at Yale University, who was not involved in the study.
Dr. Wilson Compton, a senior scientist at the National Institute on Drug Abuse in Washington, called the study persuasive.
"The strongest case is that there are consistencies across all of the studies," and that the link was seen only with psychoses — not anxiety, depression or other mental health problems, he said.
Scientists cannot rule out that pre-existing conditions could have led to both marijuana use and later psychoses, he added.
Scientists think it is biologically possible that marijuana could cause psychoses because it interrupts important neurotransmitters such as dopamine. That can interfere with the brain's communication systems.
Some experts say governments should now work to dispel the misconception that marijuana is a benign drug.
"We've reached the end of the road with these kinds of studies," said Dr. Robin Murray of King's College, who had no role in the Lancet study. "Experts are now agreed on the connection between cannabis and psychoses. What we need now is for 14-year-olds to know it."
In the U.K., the government will soon reconsider how marijuana should be classified in its hierarchy of drugs. In 2004, it was downgraded and penalties for possession were reduced. Many expect marijuana will be bumped up to a class "B" category, with offenses likely to lead to arrests or longer jail sentences.
Two of the authors of the study were invited experts on the Advisory Council on the Misuse of Drugs Cannabis Review in 2005. Several authors reported being paid to attend drug company-sponsored meetings related to marijuana, and one received consulting fees from companies that make antipsychotic medications.