Data Used for Tinnitus Stats

I was wondering about the nature of the T stats we all like to use. Tte "5-20" percent of people suffer from tinnitus ones. Does this mean chronic tinnitus? Because I heard recently that this statistic did not include war veterans (maybe because it would artificially ramp up the number idk) but also that the definition of tinnitus in this case was rather wishy-washy. As in, tinnitus does not have to present at all times nor even be true cycling t but as little as 5 min twice a week for something? If anyone could shed light on the nature of this statistic I'd be really interested to know. Maybe that's not true I just wanted to confirm.