The ex-wife of Orlando shooter Omar Mateen described him as a volatile and violent spouse who abused steroids and beat her during their brief marriage. “He started abusing me physically, very often, and not allowing me to speak to my family, keeping me hostage from them,” Yusufiy told reporters gathered in front of her home yesterday. After four months of marriage, Yusufiy was physically rescued from Mateen by her parents and, she said, filed a police report about his abusive treatment.

Since the shooting, former coworkers of Mateen have come forward to speak out about his “unhinged and unstable” behavior. But his violent behavior toward his then-wife was perhaps a leading indicator of the future course he might take. A 2015 study by Everytown for Gun Safety, an organization that studies mass shootings in America, found that more than a quarter of all mass shootings between January 2009 and July 2015 were perpetrated by individuals with prior domestic violence charges. And many of those killings have also been targeted at families themselves.

“At this point, it’s clear that there is some correlation between domestic violence and mass shooters and we have statistics that show it,” says Samar Kaukab, the former director of the Ohio Alliance to End Sexual Violence. “People who experience a loss of control in the home often try and express their anger about that loss of control against society on a larger scale.”

FBI Director James Comey said today that previous investigations into Mateen found that he had made conflicting claims of being connected to the Islamic State, al Qaeda, and even the Lebanese militant group Hezbollah. He also claimed to have been acquainted with the Boston Marathon bombers Dzhokhar and Tamerlan Tsarnaev — which FBI officials later determined to be false. During his rampage at the Pulse nightclub, Mateen called 911 operators and pledged allegiance to ISIS leader Abu Bakr al-Baghdadi, Comey said.

Amarnath Amarasingam, a radicalization researcher with George Washington University’s Program on Extremism, says that domestic violence does not factor into models for charting individuals’ involvement with terrorist groups.

“Law enforcement, and most terrorism scholars for that matter, don’t use domestic violence as a risk factor for radicalization,” he told The Intercept. “The main problem is that it is not entirely clear what it would be an indicator for — many individuals who are violent domestically never get radicalized, and many who are radicalized are either single or come from normal families.”

In recent years, the U.S. government has earmarked tens of millions of dollars for programs to counter violent extremism, intended to identify and stop individuals before they go on to commit violent acts. These programs have been criticized for focusing primarily on ideology and beliefs, and have tended not to incorporate more prosaic indicators of future mass violence, such as behavior toward intimate partners and family members. Kaukab says that the blind spot toward domestic violence misses an opportunity to more accurately identify people who have the potential to commit greater violence in the future.

We don’t know the precise reasons why people become mass shooters, whether they have mental illness, control issues or if they’re radicalized,” says Kaukab. “But what we do know is that willingness to do violence in the home is a good indicator that they are willing to do violence outside the home.”

“Domestic violence advocates have been screaming about related issues for years, but policymakers don’t listen,” she adds. “Issues of violence in the home don’t stop in the home, they snowball and affect society at large.”