Year of Publication



Public Health

Date Available


Degree Name

Master of Public Health (M.P.H.)

Committee Chair

Steve Browning, Ph.D.

Committee Member

Emily Brouwer, Pharm.D., Ph.D.

Committee Member

Daniela Moga, M.D., Ph.D.



Proton pump inhibitors (PPIs) are widely used drugs in the United States. Atrial fibrillation (AF) affects 2.7 million people in this country and often leads to other serious medical conditions. There is a growing body of evidence suggesting that PPIs may be proarrhythmic. This study seeks to better appreciate the relationship between these popular drugs and the most common arrhythmia.


This is a nested case-control study using a large, nationally representative insurance database. Members with an incident AF diagnosis between 2007 and 2009 were identified. Each case was matched with one control member, using incidence density sampling, and both were assigned an index date of the case’s date of AF diagnosis. PPI exposure was assessed through pharmacy claim records for both cases and controls prior to their assigned index date. Multivariable analysis was accomplished through conditional logistic regression as a means of comparing PPI exposure among cases and controls.


The study included 40,484 eligible members (20,265 cases and 20,219 controls). Bivariate analyses revealed that cases were more likely to have each of the potentially confounding comorbidities of interest than controls. An unadjusted multivariable analysis suggested that cases were more likely to have redeemed a PPI prescription OR= 3.53 PPIs and AF 4 [3.12 – 3.99]. Following adjustment, the association diminished but remained significant OR= 1.98 [1.63 – 2.41]. A sensitivity analysis among cases and controls with redeemed prescriptions for PPIs and H2RAs failed to suggest that cases were significantly more likely to have been exposed to PPIs OR= 1.68 [0.88 – 3.19].


Exposure to PPI therapy was more likely among those with an incident AF diagnosis. This study is in agreement with similar findings that PPIs may cause arrhythmias. However, the insignificance of the sensitivity analysis weaken the study’s ability to declare with more certainty that the observed association between PPIs and AF are independent of GERD, which has also been hypothesized as a causative condition of AF.

Included in

Public Health Commons