The article seems to imply that lots of p-hacking took place before they found a "positive" result worth reporting. Not sure how to feel about that. Also, why

The article seems to imply that lots of p-hacking took place before they found a "positive" result worth reporting. Not sure how to feel about that. Also, why on earth would one talk about mortality "from cancer" and not mention overall mortality from all causes? This is classic "smoking cures most diseases of old age" statistical thinking...