The design and analyses of Alzheimer Disease (AD) studies are often complicated by factors such as the progressive and degenerative nature of the disease and the multiple dimensions of the symptoms. In the past five drugs have been developed and approved by FDA utilizing the traditional single time point analysis and parallel design. Since such design and analysis could only demonstrate certain symptomatic effects from the treatments, newer compounds are being developed and intended to demonstrate an effect altering the course of the underlying dementia process. In this presentation, we will review the study design and statistical methods utilized from the past trials to the newer trials; discuss various disease modification strategies and viewpoints; and finally focus on Randomized Start Design and present approaches on the trial implementation and analysis.