Time-Series Data Analysis of Long-Term Home Blood Pressure Measurements in Relation to Lifestyle

We conducted a long-term time-series analysis of an individual's home blood pressure measurements, stored on a personal healthcare system in cloud, relative to the individual's life-style. In addition to daily scattering, apparent seasonal variations were observed in both systolic and diastolic blood pressure measurements. We examined the effect of seasonal variations on the outcome of a healthcare data mining process that extracts rules between blood pressure measurements and life-style components such as exercise and diet, and found that the daily blood pressure data approached a normal distribution when adjusted for the seasonal variations. This implies that an adjustment is desirable in order to produce appropriate rules in the healthcare data mining process.