A Finnish-Swedish joint study of over 900 convicted homicides has found a link between use of drugs and violent crimes like homicide. The study notes that recent massacres around the world, such as neo-Nazi Anders Brevik, and sinister Ronald McDonald cosplayer James Holmes has drawn new attention on a link between mental illness, psychotropic drugs, and violent tendencies. “It has been repeatedly claimed that it was the anti-depressants used by the persons who committed these massacres that triggered their violent behaviour. It is possible that the massive publicity around the subject has already affected drug prescription practices,” study co-author, Professor Jari Tiihonen says.

It’s here we must be cautious however, and remember, correlation – a link or association – is not causation. Case-controlled studies like this can only show an association. It takes a randomized controlled trial to establish causation. Certain pop science outlets might put out flashy headlines like “Painkiller users 3 times more likely to kill” or “Could antidepressants push you to murder?” to pull you in.

Very famously, ice cream sales and homicides are correlated. It doesn’t mean ice cream makes you want to kill, or that killing makes you want to get an ice cream. Lots of things can correlate, and there are plenty of spurious correlations. Something else may be happening. We need to take a closer look at the stats.

Movie madmen, like Travis Bickle in Taxi Driver, or Tyler Durden’s insomniac alter-ego in Fight Club, has schizophrenia, bipolar disorder, or multiple personality disorder. In this study, antipsychotic and neuroleptic drugs such as Zyprexa (olanzapine) and Seroquel (quetiapine), used to treat those serious diseases, were NOT associated with an increased risk of violence. In reality, those who suffer from these disorders are more than twice as likely TO BE killed, than to kill.

Antidepressant drugs like Prozac (fluoxetine) or Effexor (venlafaxine) were associated with a slightly elevated risk (+31%) of killing someone. Benzodiazepines, prescribed for anxiety and insomnia, which include drugs like Xanax (alprazolam) and phenazepam were associated with a “significantly elevated risk (+45%)” for homicide.

The press release mentions one of the effects of benzodiazepines, anxiety reduction, which can reduce inhibitions, this is also similar to the effects of alcohol. Benzodiazepines are known to be linked with violent behavior. A growing body of literature suggests this inhibition-reducing effect can have serious consequences.

The strongest link was between opiate painkillers (92%) and anti inflammatory painkillers (206%). Let’s take a look at the painkiller stat, a 92% increase with opiate painkillers like heroin or codeine. Opiates are known for something called “opiate rage.” While high on opiates, users report that they are quick to snap at people. Along with an increased propensity to talk, reduced sensation of pain, and an exaggerated sense of well-being, it is a bad combination. Run your mouth, take some punches, lose control.

However, the strongest association, a 206% increase, was with anti-inflammatory painkillers, NSAIDs, like Advil (ibuprofen) or aspirin. These are drugs not known for psychiatric side effects at all. I don’t believe there’s been an ignored side effect of aspirin, in fact, studies point to certain forms of aggression are actually reduced by aspirin. This association could just be that NSAIDs are extremely common drugs.

The press release concludes with a caution against prescribing highly addicting benzodiazepines and addictive painkillers to people with drug abuse problems. Which seems like obvious advice — addicts are willing to do anything for a fix, which is why they need a medical intervention. Not mentioned, however, is that people taking psychotropic drugs aren’t Hollywood-style killers; often, they are the victimized. The mentally ill and addicts are marginalized enough, they need compassion — not suspicion that they’re killers.