A Fourier series approach for comparing groups of subjects on ambulatory blood pressure patterns

We develop an approach to the statistical analysis of 24-hour ambulatory blood pressure monitoring where we represent each subject's profile by a different mathematical model. We first smooth the data and then use a Fourier series approach to determine the best model for each subject. We then estimate summary variables based on each subject's model to compare distinct groups of subjects. In comparing 15 adult black male hypertensives to eight adult black male normotensives we found that the two groups differ on the shift away from the mesor for both the systolic and the diastolic blood pressure profile.