It’s in the study under the controls section. “The Population Information System provided four male and four female controls, matched for birth year and munici

It’s in the study under the controls section. “The Population Information System provided four male and four female controls, matched for birth year and municipality of birth for each gender-referred individual. The index date of the gender-referred person was assigned to all controls.” Obviously the control population can’t also be GR so they’re going to have different experiences, but that’s presumably offset by matching eight to one. After all at the population level you’ll have people with plenty of trauma and illness over 25 years, even if the causes will differ from GR individuals.