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

Statins in clinical practice


Do statins reduce cholesterol and mortality? Of course they do, although not everyone is happy to agree about the exact mechanism providing the benefit, with argument still going on about cholesterol lowering and anti-inflammatory effects. We have a large number of large randomised trials that tell us so, with detailed pooled analysis of those trials to back it up. Trials run independently of pharmaceutical companies give the same answers as those run by the companies.

Even so, there are always those who are concerned that clinical trials are not the same as clinical practice. Clinical practice is populated by hosts of patients who wouldn't get into any clinical trial, and the real world is complicated by such issues of compliance and comorbidity that would make trialists shudder. So an observational study of statins in clinical practice [1] is worth a few moments of our time.


The study was a data mining exercise using a US veterans database with demographic, laboratory test, vital sign, drugs prescribed, outpatient visits, hospital admissions, and disease classifications on about 1.5 million patients, 95% of whom were men. Those with a prescription for any statin were identified, and relevant information collated on the 229,000 statin users and 1,262,000 nonusers.

The study looked at the effect of statin use on overall death. To examine the influence of statin use according to different levels of risk, a seven-point (0-6) risk score was used, scoring one point each for:


Statin users were rarely aged less than 50 years (6%), but almost half (46%) were aged 70 years or more. Almost 90% had used a statin for at least one year, and about half for three years or more. Simvastatin was most commonly used, in almost 80%.

Statin users were very different from nonusers. They were much more likely to have hypertension, diabetes, and coronary artery disease, be depressed, or be current smokers. They were much more likely to be treated with drugs like beta-blockers, ACE inhibitors, calcium channel blockers, or aspirin.

Statin users had initially higher levels of total and LDL cholesterol and triglycerides than nonusers. After treatment, they had lower levels of total and LDL cholesterol, and triglyceride concentrations were the same.

Use of statin had a highly significant negative association with death in a statistical model, with a mean age at death two years older than for those not using statins. The distribution of age at death was somewhat different in statin users, with fewer dying very young (<45 years) or very old (>90 years). Statin use was associated with a statistically significant reduced risk of death when risk scores were 1 or above (Figure 1).

Figure 1: Reduction in odds ratio for death by statins according to initial risk score


By any standard, the patients prescribed a statin in this observational study were not a good bet. Nearly 80% had hypertension, and 40% diabetes, with a mean initial total cholesterol of about 5.3 mmol/L. Half were taking antihypertensives or aspirin, and almost all were men. Yet they lived two years longer than their lower risk contemporaries.

This large and important study shows that the expected benefits from clinical trials are found in clinical practice. It also demonstrates that finding and treating higher risk patients in primary care is a sensible and useful process. Whatever the mechanism, this is useful background evidence away from clinical trials, and provides encouragement in these cold and dark Northern mornings.


  1. JL Mehta et al. Comparison of mortality rates in statin users versus nonstatin users in a United States veteran population. American Journal of Cardiology 2006 98: 923-928.

previous or next story