Birdy,
I've read that paper that motivated McCarty:
http://ije.oxfordjournals.org/content/32/4/536.full.pdfThey didn't look at vegans per se.
The examined diets for content of beef + pork, eggs, milk, chicken, and fish. They sorted by each category of animal product, but there was no category or comparison for "none of the above." See table 4.
There was one study (Heidelberg) that did not find an increased risk of cerebrovascular disease for vegans, but when Fraser et al did their meta-analysis "magic" (obtained by combining the results of all studies with use of a random-effects model) this somehow turned into an increased risk. I think somebody might have made a typo or something.
raw data for heidelberg, table 2 here
http://cebp.aacrjournals.org/content/14 ... l.pdf+htmlmeta-analysis, table 3 here
http://www.direct-ms.org/sites/default/ ... tality.pdfIn other analyses, the data for cerebrovascular disease comparing vegans to non-vegans rarely achieve statistical significance (P<0.05 generally for that, an arbitrary cutoff but somewhat useful).
See table 3. Most of the stroke column has very high P-values
http://ajcn.nutrition.org/content/89/5/ ... nsion.htmlI question the Key/Fraser manipulations on the Heidelberg study for cerebrovascular disease is all. I don't know what their machinations were. They didn't spell it out and it's possible that they messed it up.
In short, I would not worry about it.