ABC | Volume 110, Nº4, April 2018

Original Article Silva et al Postural hypotension: spectral analysis Arq Bras Cardiol. 2018; 110(4):303-311 Figures 1 and 2 depict the analysis of HRV components of a patient with OH in the supine and orthostatic positions, respectively. Figures 3 and 4 show the analysis of HRV components of a patient without OH in the supine and orthostatic positions, respectively. Stepwise multivariate analysis Variables with a p ≤ 0.10 in the univariate analysis – sex, use of ACE inhibitors, presence of previous symptoms, HR, LF and LF/HF in supine and orthostatic positions, and Framingham score – were considered for the multivariate analysis. The independent variable with statistical significance was HR in the supine position, with p = 0.001, 95% confidence interval of -0.0022 – -0.006. Analysis of the receiver operating characteristic curve Using the receiver operating characteristic curve for the stable variable postural response without OH, and considering the variable HR in the supine position, an area under the curve of 0.70 was obtained (Figure 5), with a p = 0.001 (95% confidence interval of 0.595-0.796). The best cutoff point was 61 bpm, with a sensitivity of 77.3% and specificity of 51.3%. Positive predictive value was 61.3%, and negative predictive value 69.3%. Odds ratio was 3.23 for OH in patients with a HR lower than 61 bpm. Discussion The main finding of this study was that, in contrast to HRV components, HR in the supine position was an independent predictor for the occurrence of OH in the study population. Median HR in the supine position was significantly lower in the case group than in the control group in the same position. Although this variable was a predictor of OH, with an odds ratio of 3.23 for patients with HR < 61 bpm, it was not considered a good diagnostic test, as confirmed by the analysis of the receiver operating characteristic curve. Aging is one of the main predicting factors for OH, which may be explained by changes in the autonomic regulation Table 3 – Comparison of heart rate spectral analysis between case and control groups Variables Case group Median (Q1 – Q3) Control group Median (Q1 – Q3) p-value LF supine position (ms 2 ) 157.0 (83.6 – 323.3) 275.0 (164.0 – 439.5) 0.014 HF supine position (ms 2 ) 111.0 (50.5 – 368.5) 141.0 (65.0 – 342.5) 0.873 LF/HF supine position (ms 2 ) 1.5 (0.7 – 2.4) 1.8 (0.9 – 4.1) 0.054 LF orthostatism (ms 2 ) 161.5 (71.5 – 333.6) 242.0 (128.5 – 375.0) 0.075 HF orthostatism (ms 2 ) 66.0 (29.0 – 229.5) 91.0 (33.5 – 247.1) 0.898 LF/HF orthostatism (ms 2 ) 1.8 (1.0 – 3.3) 2.4 (1.2 – 6.1) 0.096 SD: standard deviation; LH: low frequency; HF: high frequency; LH/HF: low frequency/high frequency ratio; ms: milliseconds. Mann-Whitney test; Q1: 25 th percentile; Q3: 75 th percentile Figure 1 – Spectral analysis of a male patient (67 years of age) with orthostatic hypotension in supine position. RR: number of QRS in sinus rhythm; VLF: very low frequency; LF: low frequency; HF: high frequency; HFnu: HF normalized unit. 18000 13500 4500 9000 0 0.0 0.1 0.2 0.3 0.4 0.5 Frequency (c/b) Duration: 5 min Total RR: 188 corrected: 0 RR average: 1135 ms 53 bpm Total potential: 1579 VLF Power: 906 (0.00–0.04)Hz LF Power: 434 (0.04–0.15)Hz LFnu: 64.6 HF Power: 237 (0.15–0.40)Hz HFnu: 35.2 LF/HF: 1.83 306