Skip navigation
Link to Back issues listing | Back Issue Listing with content Index | Subject Index

Statins and the law of unintended consequences

Changing statins in primary care
Withdrawing statins in patients with acute coronary disease

A reader asked Bandolier whether it was true that it was dangerous to stop taking statins. Given that statins are among the safest of drugs, this seemed a rather curious question to which the answer was blisteringly obvious. If statins were given to prevent something bad happening, then stopping them might make the bad thing more likely to happen. A quick search turned up two pieces of evidence which imply that stopping or changing statins could increase vascular risk by about three times. There are some complexities in the biology that might make this an interesting area to keep an eye on.

Changing statins in primary care [1]

This report was an audit of all patients who had changed their statin in the Otago region of New Zealand in a response to a reference pricing scheme. There were 126 such patients, and their hospital records were examined for fasting lipids and hospital admission for unstable angina, myocardial infarction, thrombotic stroke or peripheral artery occlusion for the six months before and after substitution of fluvastatin for simvastatin.

The mean dose of simvastatin of 22 mg was changed to a mean dose of 37 mg of fluvastatin. The change was accompanied by a significant rise in total cholesterol of 18%, LDL cholesterol by 34% and triglyceride by 13%. Significant increases occurred in 94% of the patients.

There was also a three-fold increase in thrombotic events (Table 1), from nine in the last six months on simvastatin to 27 in the first six months on fluvastatin.

Table 1: Thrombotic events in six months before and after change of statin

Mean dose (mg)
Events over six months
Myocardial infarction
Unstable angina
Non-haemorrhagic stroke
Acute limb ischaemia
Total vascular events
126 patients over six months before and after substitution of fluvastatin for simvastatin

Withdrawing statins in patients with acute coronary disease [2]

A randomised trial set out to test the efficacy of platelet receptor inhibition in the first 40 hours after the onset of chest pain in 3232 patients. Those results are of no interest here, but a retrospective analysis examined the outcomes of death, myocardial infarction, ischaemia and revascularisation in the 30 days after onset of chest pain according to statin therapy. Where there were full records:

Patients in these three groups were similar (age about 67 years, mostly men). Those on statin treatment had more hypercholesterolaemia diagnoses, but a 10% lower median total cholesterol than those who did not receive statins. Total cholesterol did not change in the 72 hours after withdrawal.

The main difference was in the rates of death and myocardial infarction in the 30 days after onset of chest pain. Patients on statins before and after admission had lower event rates than those not on statins (Figure 1). Those who had statin withdrawn had higher rates, not just higher than those continuing on statins (relative risk 2.9; 1.6 to 6.3), but higher than those never treated with a statin, though not significantly so (1.7; 0.9 to 3.6).

Figure 1: Outcome of death or MI for patients never treated with a statin, those in whom statin continues after onset of chest pain, and those in whom statin was discontinued after onset of chest pain


It may be pure coincidence that in both these reports there was a three-fold increased risk of an event after stopping a statin or changing to an ineffective dose, but it was the expected result. Neither study is definitive and merely serve to generate hypotheses, though ethics for the trial needed to prove the hypothesis may be difficult to come by, especially if the idea was to prove a larger than expected effect.

Statins are likely to do things other than just affect cholesterol levels. Endothelial nitric oxide production, leukocyte adhesion, platelet activation and LDL oxidation are all postulated mechanisms for which there is some experimental evidence. That is all very academic, but a better handle on the risk with stopping statins would be welcome.

There is an enormous experiment going on, in which millions of people are being prescribed statins and begin to take them. But most patients stop taking their statins after some time. If there is an increased risk with stopping, might an unintended consequence be that we actually cause more thrombotic events?

Unintended consequences also come from tinkering with therapy. The Otago “experiment” came from changes in reimbursement policy, so patients received insufficient doses of a less potent drug that probably altered their lipid control for the worse. When statin was changed from simvastatin to atorvastatin and lipid control changed for the better, the number of thrombotic events after the change was low, and no different from before the change [3].

Perhaps the take home message is not to mess about with statin therapy without a very good reason, and to make sure that lipid control is not impaired. Sure we need more data, but we know how to avoid the law of unintended consequences.


  1. M Thomas, J Mann. Increased thrombotic vascular events after change of statin. Lancet 1998 352: 1830-1831.
  2. C Heeschen et al. Withdrawal of statins increases event rates in patients with acute coronary syndromes. Circulation 2002 105: 1446-1452.
  3. AS Wierzbicki et al. Increased thrombotic vascular events after change of statin. Lancet 1999 353: 845.

previous or next story