ABC | Volume 110, Nº2, February 2018

Original Article Casali et al Short & very short-term blood pressure variability Arq Bras Cardiol. 2018; 110(2):157-165 however, studies on the association between BP variability evaluated by 24h-ABPM indices and target organ damage have shown contradictory results. 4,14 Our report was aimed to compare three different parameters of short-term BP variability in 24h-ABPM, with very short-term BP variability measured by indices obtained from the FBPM in healthy subjects and in a population at high cardiovascular risk comprised of diabetic hypertensive subjects. Methods Study design and population This cross-sectional study was conducted at the outpatient clinic of the Hospital de Clínicas de Porto Alegre (Porto Alegre, RS, Brazil), a tertiary teaching hospital, and Instituto de Cardiologia do Rio Grande do Sul, Fundação Universitária de Cardiologia, from January 2009 to December 2012. The study was approved by the Ethics Committee of both Institutions (nº 0469.0.001.000-08 and 4313/09, respectively), which is accredited by the Office of Human Research Protections as an Institutional Review Board, in agreement with the principles outlined in the Declaration of Helsinki. After protocol approval, all subjects signed a written informed consent for participation. Adult patients of both genders, aged 18-65 years, and with hypertension and type 2 diabetes mellitus were invited to participate (DH group). Control group (C) consisted of healthy subjects, that is, without diagnosis or medication for hypertension and diabetes. Clinical evaluation Patients underwent demographic and clinical baseline data collection. Diabetes mellitus was defined by two fasting plasma glucose ≥ 126 mg/dl or use of antidiabetic agents or personal history of diabetes. Blood pressure was measured with an office aneroid sphygmomanometer and the mean values were estimated after an two measures on average. The cuff size was selected according to arm circumference. Hypertension was defined by mean blood pressure ≥ 140/90 mmHg or use of antihypertensive medication. After baseline data collection, subjects were randomly assigned to evaluations, being first submitted to 24h-ABPM or to FBPM. The interval between the two examinations was of no more than 15 days. Short-term blood pressure variability evaluation (24h-ABPM) All individuals were submitted to a 24h-ABPM in a usual working day, using a monitor (Spacelabs 90207, Spacelabs, Redmond,WA). Measurements were obtained every 15minutes from 7 a.m. to 11 p.m., and every 20 minutes from 11 p.m. to 7 a.m. to complete 24 hours of the period studied. Cuff size was selected according to subjects’ arm circumference. 14 Based on the results of 24h-ABPM, the mean 24-hour systolic (SBP) and diastolic (DBP) blood pressures were calculated for each patient. Three different parameters of SBP variability were calculated: 1) time-rate index (rate of change in SBP over time in mmHg/min, defined as the first derivative values of SBP by time); 2) coefficient of variation of systolic BP within 24 hours (SD/mean pressure x 100%); and 3) mean of standard deviation of 24-hour systolic BP. The time-rate index allows the calculation of angular coefficients’ sum and aims to measure how fast or slow and in which direction SBP values change. The measure was calculated using the following formula, where r is the rate of BP variability over time (considering the differences between BP measurements at each time interval) and N is the number of recordings: 15 R = | r | = N – 1 N – 1 i = 1 | ri | ∑ In addition, considering circadian variations of BP and possible differences between daytime and nighttime 24h-ABPM parameters, data were divided into daytime and nighttime according to patients’ reports and were also analyzed separately, considering both periods. Circadian behavior differences were calculated by subtracting nighttime from daytime values for each parameter. Very short-term blood pressure variability evaluation (FBPM) Blood pressure was recorded continuously, on a beat‑to‑beat basis, using the FINAPRES system (Ohmeda 2300, Monitoring Systems, Englewood, CO, USA). 16 In this method, the pressure wave can be continuously monitored by a sensor placed on the patient’s non-dominant middle finger, detecting small oscillations only. The experimental protocol had measurements at two different stages: ten minutes at rest in a sitting position and ten minutes after standing-up maneuver (sympathetic activation). The BP signal was digitized by the CODAS system (Computer Operated Data Acquisition Software; DATAQ, Instruments, AKRON, OH, USA), sampling at 1 kHz and analyzed for each condition. Pulse interval (PI) tachogram and systolic arterial (SA) systogram series were constructed through the algorithm of Windaq/DATAQ, which identifies systolic peaks from BP waves. Systogram and tachogram series were analyzed by spectral analysis (frequency domain analysis) using an autoregressive model, applied to stationary intervals, which were selected in each segment condition. The stationarity of each time series was tested as previously reported. 17 Short-term BP and HR variabilities were evaluated based on systogram and tachogram analyses, respectively. In humans, the frequency domain analysis considers three distinct bands: high frequency (HF), which includes the interval between 0.15 and 0.4 Hz; low frequency (LF) between 0.04 and 0.15 Hz; and very low frequency (VLF), lower than 0.04 Hz. 18,19 The same analysis was applied to tachogram series. Among parameters obtained by frequency domain analysis, LF and HF components are distinguished by physiological significance. They are mainly related to sympathetic and parasympathetic cardiac modulations, respectively; the relation between them—LF/HF index—is related to sympathetic-vagal balance; 20 and the absolute powers of LF and VLF components are predominantly related to vascular sympathetic modulation and to renin-angiotensin system modulation on SBP, respectively. 1 The alpha index was obtained from the square root of the ratio between the LF component of tachogram and systogram when coherence, assessed by cross-spectral analysis, exceeded 0.5 in these 158