Detection of normotension, borderline and hypertension cutoffs in a population evaluated by non invasive blood pressure monitoring.
The 24-hour blood pressure (BP) average value in normotensive and hypertensive subjects is not defined. In 329 subjects three occasional blood pressure data were measured. The same day a non invasive 24-hour blood pressure monitoring was performed. The average of the occasional blood pressure and the average of the 24-hour blood pressure data were compared. In many cases the occasional blood pressure average was not strictly linked with the 24-hour blood pressure average. A statistical procedure (K-Means Cluster Analysis) was performed separately on occasional blood pressure values and on the hourly blood pressure average of the whole day. By this method three clusters of subjects were selected to evaluate the occasional and the 24-hour blood pressure cluster's cutoffs. The cluster's blood pressure cutoffs evaluated on occasional blood pressure values were not dissimilar from the blood pressure values suggested by literature for normotensive, borderline and hypertensive groups. The K-Means Clusters Analysis seems an interesting statistical method to research the subsets in a population evaluated by 24-hour blood pressure monitoring. We suggest that the 24-hour BP cutoffs evaluated in the three cluster groups may be considered as indicative for a clinical diagnosis of normotensive, borderline and hypertensive state.