It's not similar diet to the outside world. But the two Amish groups with and without the mutation have similar diets so the Amish without the mutation act as t

It's not similar diet to the outside world. But the two Amish groups with and without the mutation have similar diets so the Amish without the mutation act as the control group and have 7% incidence of T2D while the Amish group with the mutation have 0%. Presumable the p-value for normal Amish 7% vs. rest-of-US 8% is not significant in the numbers they studied, but the 7% vs. 0% was.