The thing is, they kind cheat with those statistics. They say smoing RELATED deaths. For example, if you die of a heart attack, they ask whether you smoked or have ever smoked. They then put you in the category of 'smoking related death', even if the real cause was the fact you were 400 pounds...