Few topics in drug research are as tangled as the relationship between cannabis and violence. Does one lead to the other? Do both come from something else entirely? A new systematic review and meta-analysis from King’s College London does not settle the question of cause, but it is the largest attempt so far to measure how strongly the two are linked.

The big picture

Researchers at the Institute of Psychiatry, Psychology & Neuroscience pooled data from 63 earlier studies covering roughly 265,000 people. They looked at the general population and at people receiving psychiatric care as two separate groups, and they asked two questions: are people who use cannabis more likely to commit violence, and are they more likely to experience it?

“Violence” was defined broadly, following the World Health Organization: physical, sexual and psychological harm directed at another person, whether inside or outside intimate relationships. The team also kept self-reported violence apart from violence that ended in a criminal conviction, and when a study reported several types, they used the most serious one.

The numbers

  • General population: people who used cannabis were about twice as likely to have committed a violent act as those who did not.
  • Psychiatric patients: among people in mental health care, cannabis users were about two and a half times as likely to be violent as patients who did not use it.
  • Criminal convictions: for violence serious enough to lead to a conviction, the association was stronger, at roughly three to four times the likelihood.
  • Being a victim: cannabis users were about one and a half times as likely to experience violence, and this link was stronger in women than in men.

The association also appeared in studies that followed people over time, which matters because it suggests the pattern is not purely a snapshot artifact. It was, however, smaller in those longitudinal studies than in studies that measured everyone at a single point in time.

Association is not causation

The authors are explicit that the review shows a link, not a proven cause. Cannabis use often travels with other risk factors for violence, including alcohol and other drug use and difficult social circumstances such as instability or deprivation. Someone who is at higher risk of violence for those reasons may also be more likely to use cannabis, which would produce a statistical link without cannabis driving the behavior.

That is the classic problem of confounding, and a meta-analysis can only be as clean as the studies it combines. It is also worth noting that the same pattern showing up for victimization hints that lifestyle and environment, not only individual behavior, are part of the story.

Limitations worth knowing

  • Only a small number of the included studies recorded how much, or how often, people used cannabis, so a dose-response picture is largely missing.
  • The conviction findings rest on a relatively limited set of studies.
  • Most of the underlying research predates the spread of high-potency cannabis. The authors suggest daily use of today’s stronger products could show a different, possibly stronger, pattern, but this review cannot test that.

Why it matters for health services

Based on the results, the researchers argue that mental health services should routinely ask about cannabis use when assessing the risk of violence, and that public health messaging about cannabis should mention violence alongside the more familiar concerns about psychosis, memory and dependence. Co-author Marta Di Forti stressed that the goal is better risk understanding without stigmatizing people who use cannabis more broadly, and senior author Robin Murray pointed out that as cannabis becomes more commercialized and legal, the risk conversation has focused mostly on the user, while violence, in both directions, has been largely absent.

The takeaway

The statistics are robust enough that the link deserves to be taken seriously by clinicians and policymakers, but they stop short of saying that cannabis makes people violent. The most useful next steps are studies that measure dose and potency, track people over longer periods, and carefully account for alcohol, other substances and social conditions. Until then, the honest summary is: a real association, an unresolved cause.

Sources