CARDIFF, Wales, July 27 -- Marijuana users face a risk of psychotic illness that increases with the frequency of exposure, results of a systematic review suggest.
Individuals who reported any use of cannabis had a 41% greater risk of psychotic illness than people who had never used it, investigators here reported in the July 28 issue of The Lancet.
And risk increased with use -- the most frequent users had a psychosis risk double that of never-users.
"We conclude that there is now sufficient evidence to warn young people that using cannabis could increase their risk of developing a psychotic illness later in life," concluded Stanley Zammit, Ph.D., of Cardiff University, and colleagues.
Evidence of an adverse effect of cannabis on affect is less compelling, they added.
Marijuana, or cannabis -- the parent plant -- is the most commonly used illegal substance in most countries, the authors noted. About 20% of young people report using the hallucinogen at least once a week or heavy use, defined as using on more than 100 occasions.
Experimental evidence and surveys of users have indicated that cannabis intoxication can lead to transient, usually mild, psychotic and affective experiences. The development of chronic symptoms that persist beyond or independently of intoxication provides reason for greater concern, the authors wrote.
Whether cannabis use increases the incidence of established syndromes, such as schizophrenia or depression, is unclear, they added.
In an attempt to clarify the relationship between cannabis use and psychotic or affective disorders, Dr. Zammit and his associates systematically reviewed longitudinal studies of cannabis use and subsequent psychotic or affective symptoms.
The review identified 35 reports from 22 cohort studies: 11 reports involving cannabis use and psychotic outcomes and 24 related to affective outcomes.
The unadjusted results of all the psychosis studies provided evidence of an increased risk of psychotic outcomes among cannabis users versus non-users. Pooling of adjusted estimates resulted in an odds ratio of 1.41 for cannabis users.
Pooled estimates from studies that examined frequency of cannabis use resulted in an adjusted odds ratio of 2.09 for frequent or heavy cannabis users versus non-users. Separate analysis of disabling psychotic outcomes, such as schizophrenia, yielded similar results.
"A dose-response effect was observed in all studies that examined the relation to increasing cannabis exposure," Dr. Zammit and his co-authors wrote.
Cannabis use and depression was evaluated in all 15 of the cohort studies. Analysis of pooled risk estimates suggested a possible relationship between frequency of cannabis use and depressive symptoms, but overall, the results failed to substantiate an association.
Studies that evaluated suicidal ideation produced little evidence of an association with cannabis use, the authors stated. The same was true of studies that examined anxiety and cannabis use.
Although the review was unable to prove causality in the association between cannabis use and psychosis, the results do provide policymakers with information to advise the public about the risks of a widely used drug, the authors concluded.
In an accompanying commentary, Merete Nordentoft and Carsten Hjorthoj, of Copenhagen University, noted that Dr. Zammit and colleagues were able to adjust for the effect of psychotic or imminent psychotic symptoms and thus were able "to ensure that psychotic outcomes were not related to the transitory effect of intoxication."
Although the ultimate proof of a causal relationship between cannabis and psychosis would be a large-scale placebo-controlled randomized trial, they said, such a trial cannot be done "because of practical and ethical reasons."