In research published in the PLoS Medicine journal, the scientists said that while their finding of a link does not prove that such drugs cause people to be more violent, further studies should be conducted and extra warnings may be needed in future when they are prescribed to people aged 15 to 24.
Selective Serotonin Re-uptake Inhibitors (SSRIs) are a class of widely prescribed drugs, including fluoxetine, branded by Eli Lilly as Prozac, and GlaxoSmithKline's paroxetine, branded as Paxil or Seroxat, designed to ease symptoms of anxiety and depression.
For this work, researchers led by Seena Fazel of Britain's Oxford University used a unique study design which aimed to avoid confounding factors by comparing the same individuals' behaviour while they were on and while they were off medication.
"The point of the design is that we're comparing people with themselves," Fazel told reporters at a briefing, adding that this helped minimise the impact of genetics or lifestyle factors.
Using matched data from Sweden's prescribed drug register and its national crime register over a three-year period, they found about 850,000 people were prescribed SSRIs, and 1.0 percent of these were convicted of a violent crime.
While in most age groups the likelihood of criminal violence was not significantly different when people were taking SSRIs and when they were not, in 15-24 year-olds there was a distinct increase - of 43 percent - in their risk of committing violent crime while on the medication.
The results also found a higher risk of young people being involved in violent arrests, non-violent convictions and arrests, non-fatal injuries and having alcohol problems when they were taking the antidepressants - but also that those who took lower doses had a higher risk of being violent.
Fazel stressed that the findings raised several questions and should be investigated further before any changes were recommended on prescribing SSRIs. He said it was possible that young people taking lower doses of antidepressants were not being "fully treated" for their mental disorder, leaving them more likely to engage in impulsive behaviour.
He added, however, that if the results are confirmed in further studies, "warnings about the increased risk of violent behaviour among young people taking SSRIs might be needed."