Unpopular opinion: whenever I read a paper based on #Mendelian #Randomization techniques, I sense vibes of quackery.
(No offense intended: this surely comes right from my ignorance. Can't help bring sorely sceptical, though)