Well, not to be argumentative, but association is not causation. Let's assume that they were also caucasian, you could also conclude that being caucasian kills.

Maybe a more accurate picture would be to say that their obesity was a sympton of something else, maybe their lack of activity, their diet (undoubtedly high in processed carbs), their lifestyle, or all of the above with a series of other yet-to-be identified factors.

The link Walt posted reminded me that science doesn't know everything and tends to make way too many quick assumptions because that's how the system works. Money and prestige goes to scientists with politically correct views.