a calculation mistake ?
I dont know if this would change the author's conclusions on the non clear
benefit of omega 3 but I draw your attention on a simple calculation (or
mistake that might be relevant : in fig 2, "Effect of omega 3 fatty acids
mortality", in the subcategory "RCT data, a linolenic acid only", the
for low omega 3 /control should be 68, not 58, which makes the 72 events
the high omega 3 group look not as bad.
Competing interests: No competing interests