As described in our previous work,11 we derived covariate matched cohorts for AUD, epilepsy, HS and TBI (i.e., neuroinflammation-related conditions) (Figure 1A). We included individuals with ≥60 years of age and ≥2 years of enrollment. First, we identified patients with neuroinflammation-related conditions. Specifically, we identified patients that had their earliest known neuroinflammation-related diagnosis after 365 days of the enrollment start date. We defined the index date as the first diagnosis date for patients with neuroinflammation-related conditions. We excluded patients with AD, AIDS, cancer, dementia, and/or hemiplegia/paraplegia prior to the index date. Second, we matched patients with neuroinflammation-related conditions to individuals without the corresponding conditions on index date, age, gender, race, and comorbidities (e.g., cerebrovascular disease, congestive heart failure, chronic pulmonary disease, diabetes, depression, liver disease, myocardial infarction, peptic ulcer disease, peripheral vascular disease, other neurological disorder, psychoses, renal disease, and rheumatoid arthritis). Specifically, for each and every patient, we randomly selected a matched individual from all individuals met the following criteria: (i) without the corresponding neuroinflammation-related condition, (ii) without AD, AIDS, cancer, dementia, and/or hemiplegia/paraplegia prior to the corresponding patient’s
