Spectral analysis of the systolic blood pressure signal in secondary hypertension

Abstract

To seek in hypertensive patients rhythmic variations of the systolic blood pressure signal obtained by ambulatory blood pressure monitoring of any inherent cycle of intermediate value between 1 and 24 h. Subjects (62 hypertensive, 39 normotensive) were evaluated by 24-h ambulatory blood pressure monitoring. The hypertensive group consisted of 48 patients with essential hypertension, nine with renovascular hypertension and five with phaeochromocytoma. The groups were matched for age and weight. The ambulatory systolic blood pressure recording served as the input for a filtering procedure that rejected unacceptable values according to predetermined criteria. The whole-day systolic blood pressure series thus obtained were subjected to Fourier analysis to obtain a spectral analysis of daily systolic blood pressure fluctuations. Daily (first 12 h), nightly (second 12 h) and whole-day average systolic blood pressure values were calculated and compared for the various groups. The average nocturnal systolic blood pressure was found to be lower than its daily counterpart in the normal subjects and in the patients with essential hypertension, whereas in the patients with renovascular hypertension or phaeochromocytoma no such nocturnal decrease was found. The power spectrum of patients with phaeochromocytoma was statistically different from that of other aetiologies of hypertension. This was achieved due mainly to a statistically significant difference in the power spectrum integral over the low-frequency band (0-0.2 cycles/h) of the power spectrum of the 24-h systolic blood pressure signal. Resection of the phaeochromocytoma normalized the power spectrum as found by analysis of the postoperative ambulatory blood pressure monitoring data in two patients who underwent a repeat recording. The technique described enables the discrimination of patients with phaeochromocytoma as a cause of hypertension from other aetiologies of hypertension. Patients with renovascular hypertension could not be distinguished from those with essential hypertension on the basis of their power spectrum. However, this technique may prove to be a valuable modality for characterizing hypertensive patients of different aetiologies.