I'm saying that probably a very tiny percentage of Americans get 0 IU. Which makes 0 IU a very poor choice of a baseline. Since you need some concrete amount to

I'm saying that probably a very tiny percentage of Americans get 0 IU. Which makes 0 IU a very poor choice of a baseline. Since you need some concrete amount to use for your control group, what would you propose using?