The threshold for publishing health studies used to be a 3-4 times greater risk, but now with so much muddiness in the data, it's difficult to make conclusive statements about research studies that link smoking, lack of sleep, and heart attacks.