Abstract | OBJECTIVES: METHODS: This was a retrospective nationwide study using database from Taiwan National Health Insurance and involved subjects aged ≥20 years. In propensity score-matched analysis, patients with current PPI use were compared with propensity score-matched PPI non-use controls at a 1:1 ratio. Patients with prior stroke or hospitalization before the index date were excluded. The primary outcome measure was hospitalization with a primary diagnosis of ischemic stroke during 120-day follow-up. A parallel analysis adopting a nested case-control design was carried out. Patients hospitalized for a first-time ischemic stroke were identified and were compared with matched controls using conditional logistic regression analyses focusing on PPI use before the index date. RESULTS: The propensity score-matched analysis included 198,148 PPI treatment courses and control periods without PPI use. PPI use was associated with a higher risk of hospitalization due to ischemic stroke with a hazard ratio of 1.36 (95% confidence interval (CI) 1.14-1.620, P=0.001). Based on subgroup analysis, patients aged <60 years were more susceptible (P=0.043 for interaction), whereas gender, history myocardial infarction, diabetes mellitus, hypertension, use of antiplatelet agents of non-steroidal anti-inflammatory drugs, or type of PPIs had no effect on the risk. In the nested case-control analysis, 15,378 patients hospitalized owing to ischemic stroke were identified and were compared with 15,378 matched controls. An association between PPI use and increased cerebrovascular risks was identified, and the adjusted odds ratios for PPI use were 1.77 (95% CI 1.45-2.18, P<0.001) within 30 days, 1.65 (95% CI 1.31-2.08, P<0.001) between 31 and 90 days, and 1.28 (95% CI 1.03-1.59, P=0.025) between 91 and 180 days before the onset of first-time ischemic stroke. CONCLUSIONS: PPI use is associated with an increased risk of first-time ischemic stroke in the general population, and the risk is independent of antiplatelet agents. However, caution should be exercised when considering its clinical relevance as the magnitude of association was modest and a cause-and-effect relationship remained to be established.
|
Authors | Yen-Feng Wang, Yung-Tai Chen, Jiing-Chyuan Luo, Tzeng-Ji Chen, Jaw-Ching Wu, Shuu-Jiun Wang |
Journal | The American journal of gastroenterology
(Am J Gastroenterol)
Vol. 112
Issue 7
Pg. 1084-1093
(Jul 2017)
ISSN: 1572-0241 [Electronic] United States |
PMID | 28397874
(Publication Type: Journal Article)
|
Chemical References |
|
Topics |
- Case-Control Studies
- Female
- Humans
- Longitudinal Studies
- Male
- Middle Aged
- Propensity Score
- Proton Pump Inhibitors
(adverse effects)
- Retrospective Studies
- Stroke
(chemically induced, epidemiology)
- Taiwan
(epidemiology)
|