Date of Award

Fall 2010

Degree Type


Degree Name

Doctor of Philosophy (PhD)

Graduate Group

Epidemiology & Biostatistics

First Advisor

Brian L. Strom, MD, MPH

Second Advisor

Warren Bilker, Ph.D.

Third Advisor

Sean Hennessy, PharmD, Ph.D.


Statins reduce the risk of major coronary outcomes and all cause mortality. They are generally well tolerated, but are associated with uncommon but serious adverse events. Pharmacokinetic studies show statins metabolized by the CYP3A4 isoenzyme (statin 3A4 substrates) are susceptible to drug interactions when concomitantly administered with drugs that inhibit the CYP3A4 isoenzyme (CYP3A4 inhibitors) - potentially increasing the risk for adverse events. Studies to evaluate the clinical importance of the statin-CYP3A4 inhibitor interaction are limited to anecdotal findings. This research endeavored to evaluate the clinical importance of the statin-CYP3A4 inhibitor drug interaction in two empiric investigations and a methodologic study.

The preliminary empiric study was an analysis of spontaneous rhabdomyolysis reports. It showed an increased rhabdomyolysis reporting rate for simvastatin (a statin 3A4 substrate) but not for pravastatin (a statin non-3A4 substrate) with a concomitant CYP3A4 inhibitor. Substantial internal validity limitations, inherent in spontaneous reporting analyses, warranted additional research.

To further assess the clinical importance of this drug interaction, we evaluated the validity of the multinomial propensity score as a confounding adjustment method in a simulated drug interaction study. The results from the simulation study provided support for using the multinomial propensity score in the second empiric study. The results showed the multinomial propensity score reduced bias, had greater coverage probability, and increased precision compared to binary propensity score methods. Investigators studying multinomial exposures, such as drug interactions, should consider the multinomial propensity score for confounding adjustment.

The second empiric study was a large retrospective cohort study. The objective was to evaluate the hazard of muscle toxicity, renal dysfunction, and hepatic dysfunction among patients exposed to statin 3A4 substrates (atorvastatin and simvastatin) compared to statin non-3A4 substrates (fluvastatin, pravastatin, and rosuvastatin) with and without CYP3A4 inhibitor concomitancy. We found no overall increased hazard of muscle toxicity, renal dysfunction, or hepatic dysfunction associated with statin 3A4 substrates compared to statin non-3A4 substrates with versus without a concomitant CYP3A4 inhibitor. Given the magnitude and validity of this investigation, the drug interaction between statins and CYP3A4 inhibitors therefore does not represent a substantial public health concern.

Files over 3MB may be slow to open. For best results, right-click and select "save as..."