RT Journal
A1 Redberg RF, Katz MH
T1 USe of statins in healthy men—reply
JF JAMA
JO JAMA
YR 2012
FD August 15
VO 308
IS 7
SP 666
OP 667
DO 10.1001/jama.2012.8732
UL http://dx.doi.org/10.1001/jama.2012.8732
AB 
Dr Beckman suggests that Ray et al1 should have used a fixed-effects model instead of a random-effects model. However, given the different types of studies involved, it is more appropriate to assume that the true effect varies between studies and perform a random-effects model. Furthermore, even with the fixed-effect model, the result is not significant (ie, the upper confidence limit includes 1.00) or no benefit from statins.