The study they cite has real problems. FAERS data is pretty limited - can't actually determine real risk numbers, can't separate what's actual medical signal versus media hype, can't tell if it hits one group harder than another, can't tell severity. Hard to draw firm conclusions when the data...