PROJECT SUMMARY Despite the common belief that prescription drug use is a leading cause of motor vehicle crashes, data are scarce and controversy remains about the effects of medications on crashes in older adults aged ≥65 years. Unlike other determinants of crashes (e.g., medical conditions), medications are one of few modifiable potential determinants of the 6,800 crash-related deaths and 191,000 crash-related non-fatal injuries that occur annually among older adults. In particular, psychoactive drugs are commonly used among older drivers, but may interfere with safe driving. As more adults continue to drive into older age, there is an urgent need to understand the effects of medications and distinguish them from the effects of contemporaneous age-related medical conditions, impairments, and physiological changes. The overall objective of this proposal is to examine the causal effects of medications on crashes in older drivers, and the extent to which these medications disproportionately affect crash risk across subgroups [e.g., Alzheimer's disease and related dementias (ADRD), polypharmacy] and by medication adherence status. The central hypothesis is that sedating psychoactive medications (opioids, nonbenzodiazepine hypnotics, antidepressants, and antipsychotics) will increase crash risk while central nervous system (CNS)-activating drugs (cholinesterase inhibitors, CNS simulants) and others (non-steroidal anti- inflammatory drugs) will decrease the risk, and that these effects will be greatest among individuals with ADRD and those who are adherent to their medications. This hypothesis will be tested by pursuing three specific aims: 1) Estimate the effect of initiating sedating psychoactive, CNS-activating, and other medications, including dose, on crashes in older adults; 2) Quantify the effect of initiating sedating psychoactive, CNS-activating, and other medications on crashes across important subgroups of older adults, including those with ADRD; polypharmacy; multimorbidity; and sleep, psychiatric, neurological, and musculoskeletal disorders; and 3) Evaluate the effect of non-adherence to sedating psychoactive, CNS-activating, and other medications, each separately compared to adherence, on crash risk. To accomplish the three aims, our team will develop a unique database that combines data on older drivers' licensing and crash histories; Medicare health insurance and drug claims; and data on important determinants of medication use and crashes (e.g., access to transportation alternatives). This approach is innovative because it is the first to compile high-quality U.S. data on all three domains necessary to study the effect of medications on crashes—1) medical conditions (covariates); 2) medication use (exposure); and 3) crashes (outcome)—in a dataset that is large enough to precisely estimate effects using causal inference methods while accounting for differential driving frequency between drivers. The proposed research is significant be...