ABC | Volume 114, Nº6, June 2020

Original Article Velten Orthostatic hypotension and pressure variation: ELSA Arq Bras Cardiol. 2020; 114(6):1040-1048 Methods Study design and population This is a descriptive study carried out with data collected at the baseline (2008 – 2010) of the Longitudinal Study of Adult Health (ELSA-Brazil), with a cohort of 15,105 civil servants of both sexes, between 35 and 74 years of age, whose main objective was to determine the incidence of chronic diseases and their determinants in theBrazilianpopulation. The study is being carried out in six centers of investigation located inpublic higher education and research institutes, the participants being active or retired civil servants from these institutes. Details on sampling, recruiting, and data collected at the baseline have been previously published. 12,13 This study included all the participants of the ELSA-Brazil, with the exception of those who did not have complete data on the postural change maneuver. The final sample was composed of 14,833 individuals (Figure 1). Postural change maneuver and orthostatic hypotension To perform the postural change maneuver, participants remained lying down for approximately 20 minutes while they were submitted to the protocol for measuring ankle-brachial index (ABI). Three BP measurements were obtained in the right arm in the supine position, with two-minute intervals between them. The average of the last two measurements was used as the supine BP value. Subsequently, the assessor instructed the participant the standup (withhelp, if necessary) and tomaintain an upright posture, standing only on his or her feet. BPwasmeasured again at 2, 3, and 5 minutes after standing, without supporting the participant’s arm. 14 Assessors were instructed to take note of spontaneously reported symptoms (dizziness, visual alterations, nausea, etc.) on a specific form. Depending on the intensity of symptoms, it was possible to alter the protocol and measure BP in the seated position. Assessors received routine training, certification, and periodic recertification. Supervisors who were trained and certified on the central level trained local teams. 14 All BP measurements were obtained using a validated oscillometric device (Omron HEM 705CPINT, Japan), 15 and cuff sizewas chosen according to armcircumference. It was necessary to use a mercury sphygmomanometer (Unitec, Brazil) for 27 participants, owing to failure to read the oscillometric device. Another 14 participants were unable to maintain orthostasis for all BPmeasurements, and their BP increasedwhen they returned to the supine position. For these individuals, a correction is made based on the average BP variations or the individuals who remained standing with the same values of reduced pressure. OH was defined as the presence of a reduction in SBP of ≥ 20 mmHg and/or DBP of ≥ 10 mmHg in the measurement at 3 minutes after standing. 6,7 Subsequently, prevalencewas evaluated considering a BP reduction in any measurement or applying a reduction of ≥ 30 mmHg in SBP for patients with hypertension. Statistical analysis The prevalence of OH was determined by sex, age range, race/color, and level of schooling. Data on prevalence were shown as frequency and 95% confidence intervals (CI). With the aim of avoiding the influence of the cardiovascular diseases or diabetes, the prevalence of OH was recalculated for a subsample generated by the removal of patients with hypertension (whether or not they were using anti- hypertensive medication), diabetes, self-reported heart failure, prior coronary disease (infarction or stent placement), and stroke. The average and standard deviation (SD) for age were also described for each subsample. Furthermore, the prevalence of symptoms related to postural change in individuals with or without OH was verified. Average and SD were also described for variations in BP (pressure variation: orthostatic BP minus supine BP) by age range and overall, for variations in both SBP and DBP. Finally, the prevalence of OH was calculated considering pressure reductions at 2, 3, and 5 minutes and applying the criterion of a reduction of ≥ 30 mmHg in SBP at 3 minutes for patients with hypertension. A Venn diagramwas also developed for the three different measurements. Analyses were carried out using Microsoft Office Excel and IBM SPSS Statistics 21. Figura 1 - Fluxograma do estudo. 1041