Author’s reply to RogersBMJ 2012; 344 doi: http://dx.doi.org/10.1136/bmj.e1243 (Published 21 February 2012) Cite this as: BMJ 2012;344:e1243
- Morris J Brown, professor of clinical pharmacology1
- 1Clinical Pharmacology Unit, University of Cambridge, Addenbrooke’s Hospital, Cambridge CB2 2QQ, UK
Unfortunately, there are two answers to the question,1 2 depending on whether absolute or relative risk is plotted against blood pressure.
The first is the “raw data,” but the second is, in the Oxford view of risk, the more valid way of disentangling the influence of blood pressure from confounders such as age.3 The log-linear relation between …