Association between acid-suppressant use and self-reported stroke

The association between proton pump inhibitors (PPIs) and stroke remains controversial, while H₂ receptor antagonists (H₂RAs) have been less studied in relation to stroke. This study aimed to explore the relationship between PPIs, H₂RAs, and overall acid-suppressant medications with stroke using data from the National Health and Nutrition Examination Survey (NHANES) 1999-2018.

Data on acid-suppressant use were obtained from the NHANES prescription medication survey for the past month, classifying individuals using PPIs or H2RAs as acid-suppressant users. Stroke data were derived from the disease questionnaire. Propensity score matching (PSM) was applied to reduce intergroup differences, followed by weighted multivariable logistic regression and subgroup analyses to examine the association between acid-suppressant use (including the use of PPIs and H2RAs) and stroke.

A total of 54,918 participants were included before matching, among whom 5,776 (10.5%) were acid-suppressant users. Stroke was reported in 551 (9.5%) acid-suppressant users and in 1,643 (3.3%) of the 49,142 non-users. After 1:3 propensity score matching, 14,390 participants were retained, including 3,740 acid-suppressant users and 10,650 non-users. Among them, 316 users (8.4%) and 614 non-users (5.8%) reported having had a stroke. Weighted logistic regression analysis indicated a significant association between acid-suppressant use and stroke. The association remained statistically significant both before matching (OR: 1.37, 95% CI: 1.15-1.63, P < 0.001) and after matching (OR: 1.33, 95% CI: 1.10-1.60, P = 0.003). In addition, the association between H2RA use and stroke remained significant before and after matching, whereas the association between PPI use and stroke appeared less stable, showing borderline significance.

Our findings suggest a significant association between the use of acid-suppressive medications and stroke; however, due to the limitations of cross-sectional study design and self-reported data, this result requires further verification in future research.