One would think that you'd get a much more accurate results (e.g. control more variables) from looking at a cohort that lives meters away from nuclear power pla

One would think that you'd get a much more accurate results (e.g. control more variables) from looking at a cohort that lives meters away from nuclear power plants for months each year to study, e.g. personnel on nuclear subs, ships and carriers.