Even assuming that by theoat you actually meant to write “the most,” the assertion seems unlikely at best. By what metric are you measuring violence?

EDITED to add I found this in Wikipedia, which I think accounts for the bizarre twist:

There are several problems with data collection about domestic violence in same-sex relationships, including lack of reporting, biased sampling, and lack of interest in studying same-sex domestic violence. For example, if samples are taken at a center that caters to LGBT individuals, the results may be artificially high, as those who use these services may also be those in need of help, or artificially low, as abusers often limit their partner’s exposure to others, especially those who may recognize their relationship as abusive.