Nassim Nicholas Taleb @nntaleb
@bigfatsurprise @StevenMBelknap @LuisCLCorreia @DrDamluji @cochranecollab @nationallipid You are posting "no effect on mortality events", which is a misrepresentation, instead of "there is too much noise around a signal of a risk factor". It is because there are ~1 O(.) magnitude more CVD events for every death, hence smaller sample. I will summarize this in a post. — PolitiTweet.org
