Lifestyle factors have been shown to increase the risk of developing Alzheimer's disease (AD) later in life. Specifically, an unfavorable cholesterol profile, and insulin resistance are associated with increased risk of developing AD. One way to non-pharmacologically affect the levels of plasma lipids is by exercise, which has been shown to be beneficial in cognitively healthy individuals. In this randomized controlled trial y, we therefore aimed to clarify the effect of physical exercise on the lipid profile, insulin and glucose in patients with AD. In addition, we investigated the effect of apolipoproteinE genotype on total cholesterol, high-density lipoprotein-cholesterol (HDL-C), low-density lipoprotein-cholesterol (LDL-C), and triglycerides (TG) in plasma from patients with AD. Plasma samples from 172 patients who underwent 16 weeks of moderate-to-high intensity exercise (n = 90) or treatment as usual (n = 82) were analyzed change from baseline for the levels of total cholesterol, LDL-C, HDL-C, TG, glucose, and insulin. In addition, we analyzed those from the exercise group who adhered to the protocol with an attendance of 2/3 or more of the exercise session and who followed the protocol of an intensity of 70% of the maximum heart rate. We found a significant increase in plasma HDL-C levels between the "high exercise sub-group" compared to control group. After intervention HDL-C was increased by 4.3% in the high-exercise group, and decreased by 0.7% in the control group, after adjustment for statin use. In conclusion, short term physical activity may be beneficial on the cholesterol profile in patients with AD.