[krackers]

Isn't there an obvious confounding variable here where probably the hearing-aid using population has bad hearing, and if their hearing aid was not updated then it's more likely they would develop dementia? So the results from the paper could technically be true for the subset of hearing aid users that don't go for checkups.