Maybe limited funding? If there were variables between those groups to find concrete patterns within each group you would need a larger pool to study. You could say add white, but why just to Black and White and not Asian, Native Americans? As the relationship might not hold true in each group (we can be sure it would, but you cant presume that, scientifically) then you need a large enough pool in each category to get significant results, so that's a much larger study.

The funding probably also stipulated to look at populations deemed to have particular under-researched issues or needs.