Polypharmacy, Hyperpolypharmacy, and Associated Dementia: A Southeast Texas Nested Case-Control Study

Polypharmacy has been hypothesized to contribute to cognitive decline, but population-based evidence remains limited. This nested case-control study examined the associations between polypharmacy and the diagnoses of all-cause dementia (ACD), Alzheimer's disease (AD), and mild cognitive impairment (MCI) in a large sample of patients from Southeast Texas. EHR data was collected from a regional neurology practice, comprising 508 ACD patients, 149 CE patients, 579 MCI patients, and 730 cognitively healthy controls. Odds ratios (ORs) and 95 % confidence intervals (CIs) were calculated for associations between cognitive outcomes and demographic, clinical, and medication-related variables. Associations were further analyzed based on the number of concurrently consumed medication groups. Several clinical characteristics and high-risk medication groups were significantly associated with increased odds of cognitive impairment. Antipsychotics, antidepressants, benzodiazepines, antihypertensives, and dopaminergics showed increased odds for all adverse cognitive outcomes; corticosteroids for ACD and AD; and non-benzodiazepine hypnotics and anticonvulsants for MCI. Polypharmacy itself demonstrated a strong dose-dependent relationship with cognitive impairment. Compared to unmedicated individuals, those taking increasing numbers of medication groups showed increasing odds of cognitive impairment, with those taking ≥ 10 medication groups showing markedly elevated odds of ACD (OR: 20.21, 95 % CI: 14.51-28.14), AD (OR: 12.64, 95 % CI: 7.57-21.08), and MCI (OR: 47.94, 95 % CI: 34.97-65.73). Receiver operating characteristic analysis supported medication group number as an indicator of cognitive impairment. This study identified significant associations between high-risk medication groups, overall polypharmacy, and hyperpolypharmacy with increased odds of ACD, AD, and MCI, underscoring the importance of medication overview in cognitively vulnerable populations.