THE ASSOCIATION OF PSYCHOTROPIC MEDICATIONS WITH THE COGNITIVE AND FUNCTIONAL
TRAJECTORY OF MCI AND ALZHEIMERS DISEASE 
   Lead Investigator:    Paul Rosenberg
   Institution      :    John Hopkins
   E-Mail           :    prosenb9@jhmi.edu
   Proposal ID      :    143


   Publications:


1  Rosenberg PB, Han D, Leoutsakos JM, Rattinger GB, Stuart EA, Oh ES, Lyketsos CG
   Psychotropic medication exposure and cognitive, functional, and neuropsychiatric outcomes in Alzheimer's Disease (AD)
   2017   Journal Article  J Am Geriatr Soc , 

   REVISION





   Proposal Description:
   Psychotropic medications are widely used in Alzheimer's disease (AD) and mild cognitive impairment (MCI) for treatment of psychiatric symptoms, but their effect on cognitive and functional outcomes is not well characterized.  There are preclinical models of AD suggesting that enhancement of monoaminergic neurotransmission might be beneficial, but there are observational studies of antipsychotics and benzodiazepines suggesting deleterious effects.  We hypothesize that use of antidepressants (particularly selective-serotonin reuptake inhibitors [SSRIs]) will be associated with beneficial or null effects  on cognitive and functional outcomes, while use of antipsychotics and benzodiazepines will be associated with poorer outcomes. SPECIFIC AIMS: 1. To examine the association of psychotropic medication use (antidepressants, antipsychotics, and benzodiazepines) with cognitive outcomes (evidenced by neuropsychological performance) and functional outcomes (evidenced by Clinical Dementia Rating Scale-Sum of Boxes) in NACC participants with MCI and AD, using mixed-effects models controlling for propensity scores as well as other relevant covariates (age, education, ethnicity among others)  2. To examine the association of psychotropic medication use (antidepressants, antipsychotics, and benzodiazepines) with prognosis of NACC participants with MCI as evidenced by transition from baseline diagnosis of MCI to diagnosis of dementia and/or AD at at least one followup visit, using Cox proportional-hazards models controlling for propensity scores as well as other relevant covariates (age, education, ethnicity among others).