The mean serum cholesterol in American males over 60 years of age is 220 with a standard deviation of 21. An investigator wants to test if the mean serum cholesterol is different in males over 60 years of age living in Mediterranean regions.

The investigator plans to perform a two-sided hypothesis test and wants to be 95% confident that she can detect a diference of 20 points in mean total cholesterol level with 80% statistical power. She assumes that 10% of the participants will be lost to follow-up during the study. How many males over 60 years of age living in Mediterranean regions must be enrolled in the study?