Participants

The sample consisted of 20 women with RA [16] from the Rheumatoid Arthritis clinic of the Clinics Hospital of the School of Medicine of the University of Sao Paulo. Potential participants were identified from the clinic data records and all of those who complied to the study eligibility criteria (detailed below) were invited to participate in the study via a phone call. To participate in the study, the participants should have a diagnosis of RA according to the American College of Rheumatology criteria [16], and be on a stable drug therapy in the last 3 months and throughout the study. All participants should also be hypertensive, which was defined according to established criteria [17]. Exclusion criteria were as follows: a body mass index (BMI) ≥ 35 kg/m2; presence of heart, pulmonary or renal diseases, stroke, diabetes, and severe musculoskeletal and joint disorders that could preclude the execution of the tests; use of medications that could affect cardiovascular responses to exercise; abnormal resting or exercise electrocardiograms.

Prior to participation in the study, participants received a detailed explanation about the experimental procedures and provided their written informed consent. The study followed the principles of the Declaration of Helsinki and was approved by the local Institutional Ethics Committee (Clinics Hospital, University of Sao Paulo. CAAE: 90250718.0.0000.0068).

Study design

This was a randomized crossover trial (trial registration: https://ensaiosclinicos.gov.br/rg/RBR-867k9g). Participants attended the laboratory on four separate occasions. In the first two visits, participants were screened to determine eligibility, signed informed consent form, and performed the baseline tests. In the third and fourth visits, participants performed two experimental sessions (exercise and control), which were conducted in a random manner (simple randomization using https://www.randomizer.org/), with an interval of at least 48 h between them.

Baseline tests (visits one and two)Clinical evaluation

Demographic and clinical data (disease duration, disease activity, drug therapy, co-morbidities) were obtained through personal interview and medical records. Disease activity was determined by means of the Disease Activity Score-28 with C-reactive protein (DAS28-CRP) [18], whereas functional capacity was assessed through the Health Assessment Questionnaire (HAQ) [19]. Participant’s current pain level was assessed by a 0–10 visual analogue scale (Pain-VAS). Participants were also questioned about their physical activity level and the amount of weekly leisure time physical activity and were defined as active if they performed at least 150 min per week of moderate-to-vigorous physical activity [18] or insufficiently active if they did not reach that threshold.

Blood pressure (BP) was measured in two different days (visits 1 and 2), by an automated oscillometric device (Mindray MEC-1000, Shenzhen Mindray Bio-Medical Electronics Co, China) after 10-min seated rest, following the recommendations of the Brazilian Guidelines of Hypertension [19]. Briefly, BP was measured three times on both arms on visit 1 and then BP was measured three additional times on the arm with highest BP on visit 2. Baseline BP was defined as the mean of six measures on the arm with the highest BP across visits 1 and 2.

Blood samples (30 ml) were collected after a 12-h overnight fast for characterization of the cardiometabolic profile of the participants. We assessed glucose, lipid profile (i.e., high-density lipoprotein (HDL) cholesterol, low-density lipoprotein (LDL) cholesterol, very low-density lipoprotein (VLDL) cholesterol, total cholesterol, and triglycerides), C-reactive protein (CRP) and erythrocyte sedimentation rate (ESR).

Cardiopulmonary exercise test

Participants performed a maximal cardiopulmonary exercise test (CPET) on a treadmill, following a stepwise protocol with increments of speed and incline every 3 min. Initial workload and workload increments were set according to participants’ perceived fitness estimated based on an initial familiarization with the treadmill ~1 h before the CPET. During the CPET, oxygen consumption (VO2) was collected breath-by-breath using a metabolic cart (Metalyzer III B/ breath-by-breath, Leipzig, Germany) and heart rate (HR) was measured using a 12-lead ECG. Peak VO2 (VO2peak) and HR (HRpeak) were defined as their maximal values attained at the end of the exercise test (average of 30 s).

Experimental sessions (visits 3 and 4)

Participants attended the laboratory in two separate occasions to perform the experimental sessions (exercise and control). These sessions were performed in a temperature-controlled laboratory (20–22 °C) during the afternoon. Participants were instructed to avoid the intake of caffeinated and alcoholic beverages and not to perform strenuous exercise in the 24 h preceding the sessions. Additionally, they were instructed to have a light meal 2 h before the sessions. To avoid the confounding effects of different eating habits before the experimental sessions, participants completed a 24-hour food recall prior to the first experimental session and were asked to repeat a similar diet prior to the second session. Women of reproductive age performed the sessions on the first 7 days of their menstrual cycle, or during the inactive/placebo phase of the oral contraceptive pack for oral contraceptive pill users.

The experimental sessions were composed by (a) pre-intervention assessments of BP and HR at rest and in response to different stress; (b) intervention (exercise or control); (c) post-intervention assessments of BP and HR at rest and in response to different stress and; (d) 24-h (out of the lab) ambulatory BP assessment. Importantly, participants remained seated at rest for 20 min between the intervention and the post-intervention assessments. Figure 1 presents a summary with the timeline of the assessments and interventions performed in both sessions.

Fig. 1: Summary of the experimental sessions.

The exercise (upper panel) and control sessions (lower panel) were performed in different days and the order of the experiments was randomised. Both sessions were composed by pre- and post-interventions of blood pressure (BP) and heart rate (HR) at rest and in response to the Stroop Color and Word Test (SCWT), cold pressor test (CPT) and handgrip test. After the post-intervention assessments, the participants were fitted with an ambulatory BP monitor for 24-h BP assessment. The black and white icons were created by Freepik/Flaticon.

Intervention

In the exercise session, participants performed a 30-min exercise on a treadmill at a heart rate (HR) corresponding to 50% VO2max. In the control session, participants remained stood in treadmill during 30 min, but without performing any exercise.

Pre- and post-intervention assessments

Assessments were taken during the pre- and post-intervention periods with the subjects in the sitting position. BP was assessed in the left arm using an automated oscillometric device, and HR by a 3-lead ECG connected to a multiparameter monitor (Mindray MEC-1000, Shenzhen Mindray Bio-Medical Electronics Co, China). These assessments were performed at rest, and during the Stroop Color and Word Test (SCWT), the Cold Pressor Test (CPT) and the handgrip test (detailed below). A 3–5 min recovery period was provided between each test for the return of HR and BP to baseline.

Rest

Resting BP and HR were assessed after 10-min resting in the sitting position. BP and HR were measured three times, and resting BP and HR were considered the mean of all three measurements. PEH was calculated as the net effect of exercise on BP (i.e., the difference between responses observed in the exercise and control sessions) by the formula: PEH net effect = (post-exercise BP – pre-exercise BP) – (post-control BP – pre-control BP) [20, 21].

Strop color and word test

A modified version of the SCWT was used to promote mental stress in the study participants. The test started with 3-min baseline, and then the participants were presented with a sequence of words of different colors (red/green/blue) printed in either same (congruent condition) or different colors (incongruent condition). Participants were requested to respond the color of the ink rather than the words meaning [22]. The participants were asked to name each word color in the fastest possible way. At the end of the test, participants were asked to rate their perceived stress during the test on a 5 point scale (0–4, where 0 was ‘non-stressful’ and 4 was ‘extremely stressful’) [23]. Participants could only choose a whole number value (i.e., decimals were not allowed).

HR and BP were measured in 1-min intervals during the test, and SBP, DBP and HR reactivity to the SCWT were calculated as the absolute changes (delta changes – Δ) from mean baseline to their highest values achieved during the stress phase (usually within the 2nd of 3rd minute of stress).

Cold pressor test

The CPT is a reactivity test that evokes significant increases in BP and HR as a consequence of pain-induced sympathoexcitation [24]. The test started with 3-min baseline, and then the participants were instructed to immerse their right hand until the wrist in water at 4 °C for 3 min. At the end of the test, participants were asked to rate their pain on a 0-10 visual analogue pain scale (0 = not painful, 10 = maximal tolerable pain).

HR and BP were measured in 1-min intervals during the test, and SBP, DBP and HR reactivity to the CPT were calculated as the absolute changes (delta changes – Δ) from mean baseline to their highest values achieved during the hand immersion phase (usually within the 2nd of 3rd minute of hand immersion).

Handgrip test

The handgrip test was performed with a hand dynamometer (Crown Manual-50KGF, INSTRUTEMP, Sao Paulo, Brazil). The test consisted of 3 min baseline and then participants were asked to perform 3-min of sustained isometric contraction in the hand dynamometer at 30% of the maximal isometric voluntary contraction (MVC) using their right hand. During the test, participants received feedback from the investigators in order to sustain the targeted force. At the end of the test, participants were asked to rate their pain on a 0–10 visual analogue pain scale (0 = not painful, 10 = maximal tolerable pain).

HR and BP were measured in 1-min intervals during the test, and SBP, DBP and HR reactivity to the handgrip test were calculated as the absolute changes (delta changes – Δ) from baseline to their highest values achieved during the handgrip (usually within the 2nd of 3rd minute of the handgrip test).

24-hours blood pressure

After the post-intervention testing (~17:00), participants underwent an ambulatory BP monitoring for 24-h. Measurements were taken every 15 min for 24 h (this same frequency of measurements were kept during sleep) using an oscillometric device (CardioMapa, CARDIOS, São Paulo, Brazil). Participants were instructed to complete a diary with time of sleeping, waking, medication use and daytime activities. On the completion of the test, data was downloaded to an offline PC using the software Dyna-MAPA (CARDIOS, Sao Paulo, Brazil), that calculated the mean of SBP, DBP and HR for the entire 24 h, and for the asleep and awake periods as reported by the patients in the diary. Additionally, SBP and DBP nocturnal fall (%) was calculated by the difference between asleep SBP/DBP and awake SBP/DBP, divided by awake SBP/DBP and multiplied by 100. Only exams with more than 80% of measurements were considered valid for analysis [25].

Statistical analysis

Normality of data and homogeneity of variance in each group were checked with the Shapiro-Wilk and the Levene tests, respectively. A two-way (session vs. time) repeated measures analysis of variance was used to compare resting BP and BP reactivity to the SCWT, CPT and handgrip test between sessions (exercise and control) and times (pre and post). In case of significant interaction or main effects, the Tukey’s a post-hoc test was performed to perform the multiple comparisons. Paired t-tests were used to compare ambulatory BP parameters between session. For all tests, significance level was set at 5%. Continuous data are presented as mean ± standard deviation (SD) and categorical data are presented as relative frequencies. The minimum sample size required for the current study was calculated for changes in resting SBP (i.e. primary outcome), considering an alpha of 0.80 and effect size of 1.0 (SBP PEH net effect = −11 ± 11 mmHg) [21], which resulted in 13 participants.