While media attention has swirled around a purported link between antidepressants and violence, a new European study suggests the medications have only a weak association with homicide.
But, the researchers found, there is somewhat stronger evidence that some people on prescription painkillers -- like OxyContin, Vicodin and even migraine drugs -- could have an increased likelihood of homicide.
The same was true of prescription sedatives called benzodiazepines, which include drugs such as Valium, Xanax and Ativan.
The findings, published June 1 in World Psychiatry, do not prove that painkillers or sedatives drive certain people to murder, experts stressed.
But they should offer some reassurance on the safety of antidepressants in that regard, according to lead researcher Dr. Jari Tiihonen, of the Karolinska Institute in Stockholm, Sweden.
Media outlets have focused on the possible role of antidepressants in several high-profile mass killings, including school shootings in the United States and elsewhere, Tiihonen pointed out.
But as far as the science goes, there's been little evidence that the medications carry such a risk. And the new findings are in line with that, according to Tiihonen.
The new study was based on records from nearly 960 Finnish adults and teenagers who had been convicted of homicide between 2003 and 2011. Tiihonen's team found that the odds of committing homicide were 31 percent higher during time periods when offenders were on antidepressants, versus when they were not.
But while that relative increase might sound large, it actually represents a small shift in a risk that is already quite low, the researchers said.
And when they focused on offenders younger than 26 -- the group for which concerns about antidepressants are greatest -- there was no connection between the medications and homicide risk.
"I think we can conclude that antidepressants are not hazardous treatments in this respect," Tiihonen said.
Two U.S. psychiatrists who reviewed the study agreed.
It's "very difficult" to weed out the potential effects of a medication from the slew of other factors that drive violent behavior, said Dr. Michael First, a clinical psychiatry professor at Columbia University, in New York City.
"But this study suggests that if there is a risk [of homicide] associated with antidepressants, it's a very small one," First said.
No medication is going to turn a normally non-aggressive person into a killer, said Dr. Alan Manevitz, a clinical psychiatrist at Lenox Hill Hospital, also in New York City.
Instead, he explained, it's possible that a drug with particular brain effects could be the tipping point for certain individuals who are prone to violence.
Both Manevitz and First said it makes sense that benzodiazepines, used for anxiety and insomnia, were associated with an increased risk of homicide. The study showed an elevated risk, at 45 percent. The risk was doubled for people under age 26. That's because the drugs can have "disinhibiting" effects on the brain, which means a person's usual impulse control may be lessened.
For someone who already struggles with controlling aggression, that could be dangerous, Manevitz said.
First noted that the sedatives are already known to carry serious risks, including dependence if they're used long-term. They've also been linked to dementia in older adults.
"I think this study gives us yet another cautionary message about these drugs," First said.
He and Manevitz both said doctors should be careful about prescribing sedatives to people with anger issues or problems with impulse control.
In contrast, the connection between painkillers and homicide is not readily explained, and Tiihonen said it requires more study.
His team found that when offenders were on valid prescriptions for opiate painkillers, like OxyContin and Vicodin, their odds of committing homicide were roughly doubled, versus periods when they were off the drugs.
But while the study looked at prescriptions, opiates are commonly abused, Manevitz said. And it's hard to know, he added, whether people in this study were drug abusers, or had other risk factors for violent behavior that explain the link to opiates.
The researchers accounted for the risk factors that they could, Manevitz said, but it's impossible to account for all of them.
In addition, there was another puzzling painkiller finding: Non-opiate painkillers -- like migraine drugs and prescription acetaminophen -- were linked to a three-fold increase in the risk of homicide.
"It's just not clear why that is," First said.
As for antidepressants, they said, the small risk seen in this study should be balanced against the potential benefits for people suffering from serious depression.
"There's a popular perception that antidepressants are associated with violence, and some people might think it's a big risk that affects the whole population," First said. "This study offers some reassurance that this isn't the case."