Serious bias in 20 year follow-up study of statin trialBMJ 2017; 359 doi: https://doi.org/10.1136/bmj.j4906 (Published 31 October 2017) Cite this as: BMJ 2017;359:j4906
- 1Magle Stora Kyrkogata 9, 22350 Lund, Sweden
- 2Nagoya City University and Institute for Consumer Science and Human Life, Kinjo Gakuin University, Nagoya, Japan
- 3Western Vascular Institute, University Hospital Galway and Galway Clinic, National University of Ireland and Royal College of Surgeons of Ireland affiliated Hospital, Galway, Republic of Ireland
The BMJ’s news story on the follow-up study of the WOSCOPS trial says that total and cardiovascular mortality were considerably reduced in men allocated to the treatment group at 20 year follow-up; in particular among those with the highest concentrations of low density lipoprotein (LDL) cholesterol.12
But the study has serious bias. As mentioned in the supplement: “At five years after the completion of the randomised trial 38.7% and 35.2% of patients originally allocated to pravastatin and placebo arms, respectively, were taking statins (P<0.001). No later data on the proportion of individuals on statins were available.”
Many patients stop taking statins. In a study of over 140 000 elderly people, two thirds of those with cardiovascular disease (and even more of those without) had stopped treatment after two years.3 The question is, therefore, whether the mortality benefit among those with the highest LDL cholesterol is due to statin treatment or to their high LDL cholesterol.
The question is relevant because a recent systematic review of 19 cohort studies including 68 094 elderly people (≥60 years) followed for several years found an inverse association between all cause mortality and LDL cholesterol in 92% of participants.4 In the largest study those with the highest LDL cholesterol lived longer than those taking statins.5
One possible explanation is the little known but solidly documented fact that LDL cholesterol participates in immune system responses by inactivating most types of micro-organisms, and evidence shows that infections participate in the pathogenesis of cardiovascular disease.67
Competing interests: UR and HO have written books with criticism of the cholesterol hypothesis.
Full response with all authors at: http://www.bmj.com/content/358/bmj.j4171/rr-0.