I assume you already know that correlation does not equal causation.

In this study, the higher smoking, higher diabetic, lowest exercising, and higher male cohort were more likely to die. To ascribe causality on the diet, one that is not even a low carb diet is absolute bullshit.