ABC | Volume 110, Nº4, April 2018

Original Article Silva et al Postural hypotension: spectral analysis Arq Bras Cardiol. 2018; 110(4):303-311 Methods This was an observational, prospective, cross-sectional study. Our sample was composed of 105 outpatients aged 60 years or older, included during the period from February 2013 to August 2014. Patients with dementia, autonomic dysfunction-related neurologic diseases, persistent or permanent atrial fibrillation, pacemakers, institutionalized patients, and those using antiarrhythmic agents (Class I, III or IV agents according to Vaughan Williams classification) or digoxin were excluded. The study was approved by the local Research Ethics Committee and all participants signed the informed consent form. For sample estimation, a one-tailed test was used, with significance level at 5%, power of 90%, two controls per case, frequency of OH of 30%. Two groups, age matched, were studied – a case group (n = 39) with OH, and a control group (n = 66) without OH. Participants underwent clinical assessment, clinical pathology tests, 12-lead electrocardiography, measurements of blood pressure (BP) at supine position at the 5 th minute of rest and at the 3 rd minute of orthostatism or before, in case they had OH symptoms according to well-established conditions, 2 monitored by the Holter system. Measurements were performed at a temperature-controlled room, in the afternoon, at least two hours after lunch to exclude the possibility of post-prandial hypotension. Holter monitoring was performed using a three‑channel digital recorder (Cardioflash) (modified V1 and V5 and DIII) version 1.0, at supine and orthostatic positions for 15 and 10 minutes, respectively, for analysis of HRV in the frequency domain by the fast Fourier transform method. Measurements of high-frequency (HF) and low-frequency (LF) components that indicate parasympathetic and sympathetic activities, respectively, aswell as the LF/HF ratio 10 were calculated. This analysis was performed after manual edition of recordings to remove artifacts and correct arrhythmias. Measures were obtained during 5 minutes at 10 th minute of supine position and 5 th minute of orthostatic position. Results of the spectral analysis were expressed in ms 2 . The Framingham 14 and the PROCAM 15 risk scores were also calculated using clinical and laboratory data, which included plasma levels of cholesterol and its fractions, triglycerides, and fasting glucose levels. Statistical analysis For data analysis, we used the International Business Machines (IBM) Statistical Package for Social Sciences (SPSS) Statistics 19. Results were expressed as numbers and proportions for categorical variables and as central tendency (mean and median) and dispersion measures for continuous variables. Associations between categorical variables were assessed by the chi-square test or the Fisher’s exact test, as appropriate. Data normality was not tested. The Mann‑Whitney test was used for comparisons between continuous variables, and correlations between categorical variables were assessed by the Spearman's rank correlation test. The Wilcoxon test was used to compare the two periods HRV components in the spectral analysis (supine and orthostatic positions). Stepwise multivariate analysis was performed to evaluate predicting values of OH, considering the variables with a p ≤ 0.10 in the univariate analysis. Receiver operating characteristic curve was analyzed for the stable variable postural response. The level of significance was set at 5%. Results General characteristics of the population Mean and median age were 71.9 and 73.0 years, respectively; 64 (61%) were women. Clinical variables of the study population are described in Table 1. With respect to cardiovascular risk factors, systemic arterial hypertension (SAH) and dyslipidemia were the most frequent, found in 80 (76.2%) and 42 (40%) patients, respectively. Diabetes was found in 17.1% of patients. Thiazide diuretics were the most used antihypertensive drugs; 42 patients (40%) used them isolated or in combination with other antihypertensive agents. Following thiazide diuretics, angiotensin II receptor blockers (29.8%), angiotensin-converting enzyme (ACE inhibitor) (28.6%) and beta-blockers (27.6%) were the most common, with similar frequencies of use. Also, 14.3% of patients were using calcium antagonists (amlodipine or nifedipine). Symptoms characterized by previous history of dizziness, falls, and presyncope and/or syncope were reported by 64 patients (61%). Impaired conduction in the left bundle branch was detected at electrocardiography in 9.5% of patients, with mean PR and QT intervals of 166.9 ms (120-280) and 403.0 ms (320-520), respectively. Comparison between case and control groups No difference was found in age (mean of 73.5 ± 8.0 years; median of 74.0 in the case group and 71.0 ± 6.8 years and 72.0 years in the control group, p = 0.119), but a significant difference in sex was observed between the groups (56.4% of men in the case group and 27.8% of men in the control group, p = 0.005). No correlation was found between these two variables (Spearman’s coefficient correlation of 0.274). Results of other comparisons between the two groups are described in Table 2. No patient had dizziness, presyncope or syncope in orthostatismwhen BP was measured. No difference considered abnormal in BP between the upper limbs was detected in seated position. Significant differences were found in the frequency of previous symptoms (dizziness, prepsyncope and syncope) – 77% in the case group versus 51.5% in the control group (p < 0.001). However, no difference between patients with and without previous symptoms were found in age (mean or median) – 71.4 ± 7.4 years; 72.0 years versus 72.7 ± 7.8 years, 74.0 years; respectively (p = 0.38) – nor in BP measured in the supine position. With respect to hypertension, no difference was found between the case and control groups (p = 0.54). Forty‑nine patients (74.2%) in the control group and 31 in the case group were hypertensive (79.4%). There was no difference 304