A large new evidence review has found that cannabis use was associated with a higher likelihood of both carrying out violence and experiencing it. The findings bring together research involving more than 265,000 people, but the authors stress that they do not prove cannabis causes violence.
The systematic review and meta-analysis, made public on 1 October by a south London mental-health research team, pooled 63 studies. It considered people in the general population separately from people receiving psychiatric care, rather than treating either group as a single, uniform population.
Its definition of violence was broad, covering physical, sexual and psychological violence towards another person, in intimate relationships and elsewhere. The researchers also distinguished between violence reported by participants and violence that resulted in a criminal conviction. Where an individual study measured several forms of violence, the review used its most serious reported outcome.
What the pooled evidence found
Among general-population studies, people who used cannabis were about twice as likely to have been reported as committing a violent act as people who did not use it. Within psychiatric-care samples, the association was around two and a half times. For violence that led to a criminal conviction, the reported association was larger, at roughly three to four times, although that result drew on a small number of studies.
The review also examined victimisation. Cannabis users were about one and a half times as likely to have experienced violence, with a stronger association reported for women than for men. That is an association in the collected research, not evidence that a person’s cannabis use causes another person’s violence or that any individual user will experience or commit harm.
Studies that followed people over time still found an association, but it was smaller than in studies which took a snapshot at one point. This matters because long-term studies can help researchers assess whether patterns persist beyond a single moment, while still not resolving every reason why the groups may differ.
Why the findings cannot settle cause and effect
A meta-analysis can strengthen the picture of a recurring relationship, but it cannot by itself establish a cause. Cannabis use may occur alongside alcohol or other drug use, difficult social circumstances and other factors that can affect the likelihood of violence or victimisation. The studies did not consistently measure all of those influences in the same way.
There were other limits. Relatively few studies recorded the amount or frequency of cannabis use, so the review cannot give a reliable threshold at which risk changes. Most of the underlying research was carried out before high-potency cannabis became widespread. The evidence therefore should not be read as a precise estimate for every product, pattern of use, place or person today.
The authors also identified signs that publication bias or small-study effects may have influenced parts of the evidence base. Their additional analyses continued to show associations, but in some groups the estimates were reduced. Such checks are one reason the overall results need to be understood as evidence of a relationship requiring careful interpretation, rather than a simple prediction of behaviour.
Implications for care and public information
The researchers said mental-health services should routinely ask about cannabis use when considering violence risk, alongside the many other clinical and social factors they already need to assess. They also argued that public-health information about cannabis should include the possible links with violence, including victimisation, rather than concentrating only on effects experienced by the user.
That recommendation is not a basis for treating cannabis users as inherently dangerous or for assuming that they will be harmed. Violence has many causes and contexts, and people who use cannabis are not a homogeneous group. For clinicians and public-health teams, the practical point is to discuss use in a non-judgemental way, assess wider circumstances and offer appropriate support where concerns arise.
The review gives a substantial new synthesis of a difficult area of research. Its strongest message is twofold: the observed associations are serious enough to consider in care and prevention, but they must not be turned into a claim of proven causation or a stigma attached to everyone who uses cannabis.
