Analysis of 24-h noninvasive ambulatory blood pressure profiles by a third-degree polynomial approach

Abstract

Extensive experience of the circadian variation of blood pressure (BP) has been reported. But some problems exist concerning the correct and satisfactory evaluation of 24-h BP profiles. A well-known approach is the Halberg's 'cosinor' method. But it could be considered inadequate for blood pressure recordings because it is based on a rigid model that does not practically fit empirical data. We suggest an approach based on a cubic function as a suitable method for such an analysis. Our investigation studied the behavior of 24-h blood pressure by a model where the function f(t) is a third-degree polynomial. We recruited 52 untreated patients affected by essential hypertension. Ambulatory BP recordings were made using a portable noninvasive system. Data were analyzed by a computerized procedure for both estimating parameters of the cubic-polynomial model and obtaining conventional and function-derived parameters. Eleven of 52 hypertensives showed no common rhythm: 8 patients had no statistically significant rhythm and 3 had a reversed rhythm. The 8 of 11 hypertensives who had no statistically significant rhythms were also 'nondippers' according to the criteria of Verdecchia et al.. The remaining 41 patients were matched and compared to 41 normotensive subjects. In general, function-derived parameters stressed differences between the two groups more than the conventional ones. We suggested an analysis of 24-h BP data based on a cubic function model. We verified that such a methodology is feasible. At the moment, it seems a useful approach for clinical investigation.