To the Editor I read with interest the article by Gomm and colleagues,1 which examined the association between the use of proton pump inhibitors (PPIs) and incident dementia in elderly individuals in a prospective study. The authors adopted time-dependent Cox regression analysis, and the time-dependent covariates were polypharmacy and the comorbidities of depression, diabetes, ischemic heart disease, and stroke. Age and sex were also used as confounding factors. The hazard ratio of PPIs for incident dementia was 1.44 (95% CI, 1.36-1.52), and the authors recommended randomized clinical trials to confirm the causal association.1 I appreciate their conservative opinion and editorial comments seem appropriate in combination with information from past research.2 I have some concerns on their study.
Kawada T. Proton Pump Inhibitors and Dementia Incidence. JAMA Neurol. 2016;73(8):1025-1026. doi:10.1001/jamaneurol.2016.1491