I went into McD's a couple of days ago - yeah, yeah, but I happen to think their coffee is passable (and cheap, relatively).
30 patrons, not counting 4 children.
ALL at least 50 lbs overweight. And I mean ALL. And a distinct emphasis on "at least". I  would qualify them all as "obese".
And the children were also greatly overweight.
In addition, of the 6 servers, 3 were also overwieght.
That left me and 3 people behind the counter as moderately close to a reasonable BFI.
So: Excluding the observer, Correlation of obesity to customer sample: r+1.00.
Correlation or causation?
Just askin'.