• CGIC score at last observation [ Time Frame: 8 weeks after baseline ] [ Designated as safety issue: No ]
  
• Change from baseline to last observation in NPI total score [ Time Frame: 8 weeks after baseline ] [ Designated as safety issue: No ]
  

• Number of study days completed [ Time Frame: 8 weeks after baseline ] [ Designated as safety issue: No ]
  
• Change from baseline to last observation in BPRS total score [ Time Frame: 8 weeks after baseline ] [ Designated as safety issue: No ]
  

Although the occurrence of disruptive agitation behaviors likely are influenced by environmental/ interpersonal factors, it is also likely that behaviorally relevant neurobiologic abnormalities lower the threshold for the expression of such behavior in AD. Because of the success prazosin has had in the treatment of PTSD, it is thought that it could be used similarly with disruptive agitation. Originally designed to evaluate AD patients in nursing homes, the study now includes outpatients. It is a 9-week placebo-controlled trial.

Inclusion Criteria:

• No age limit
• probable/possible AD diagnosis
• disruptive agitated behaviors (e.g., irritability, aggression, uncooperativeness, pacing)
• no hypotension
• no concurrent use of alpha-1-blockers
• no delirium, schizophrenia, mania, psychotic symptoms.

Exclusion Criteria:

• Cardiovascular: unstable angina, recent MI, second or third degree AV block, preexisting hypotension (systolic BP less than 110) or orthostatic hypotension
• Other medical exclusions: chronic renal or hepatic failure, or any unstable medical condition
• Exclusionary medications: current treatment with prazosin, other alpha-1-blockers
• Current enrollment in a separate investigational drug trial
• Psychoactive medications: subjects may be psychoactive medication-free or be partial responders (by subjective assessment of referring health care professional) to one psychoactive medication from any of the following classes: antipsychotics, anticonvulsants, mood stabilizers, antidepressants, benzodiazepines, or buspirone. Partial response is defined as some improvement in agitated behavior but persistence of agitated behaviors severe enough to cause patient distress and/or difficulty with caregiving. Although not formally rated, this improvement is equivalent to a CGIC rating of no more than minimal improvement (improvement is noticed by not enough to improve patient function or caregiver's practical management of the patient).
• Psychiatric/behavioral: lifetime schizophrenia; current delirium, mania, depression, or uncontrolled persistent distressing psychotic symptoms (hallucinations, delusions), substance abuse, panic disorder, or any behavior which poses an immediate danger to patient or others or which results in the patient being too uncooperative to meet the requirements of study participation.