N-of-1 approach to determine when adverse effects are caused by statins

Ebrahim and Smith suggest in their letter that myalgia symptoms in patients taking statins or placebo were "indistinguishable" , concluding that statins did not cause myalgia. Their summary appesrs to suppose that statin myalgia is of sudden onset and equally sudden disappearance, as they used a three week swap period.

From my own experience of statin side-effects (N=1) with sequential trials of simvastatin, atorvastatin, rosuvastatin and prvastatin I can assure your correspondents that their conclusion is worthless. While atorvastatin produced side-effects after two days with a polymyalgia pattern of symptoms, all of the others produced effects that came on between 4 and 12 weeks - but more importantly the effects took between four weeks and four years to subside. I have no idea why different statins should behave differently (and I might add that ezetimibe produced similar symptoms); nor have I any other explanation for my symptoms or the fact that with three of the offending drugs my creatine kinase became significantly abnormal. But if my experience is anything to go by the trial quoted is flawed because its timescale is too short.

Of course, the whole debate is informed by the re-analysis of old studies now shown to be flawed, and the growing realisation that sugar, and not cholesterol, is the villain. It is also true that abandoning statin therapy in the millions who are being treated for something they don't have would go a long way to helping the NHS through its current financial meltdown.