Aerobic training conducted at different times of day in elderly patients with hypertension: a controlled trial

Abstract

Previous data showed greater blood pressure (BP) reduction after aerobic training performed in the evening than in the morning in middle-aged hypertensive men. This study investigated whether the time of day in which aerobic training is performed also influences BP, hemodynamics, and vascular function in elderly hypertensives. In this controlled trial, 54 elderly patients receiving antihypertensive medication (≥ 60 years, 26 women) were randomized to one of the following groups: morning training (MT, n = 19), evening training (ET, n = 18), or control (CG, n = 17). For 10 weeks, they underwent the proposed interventions, consisting of progressive moderate-intensity aerobic training in MT (7–10 a.m.) and ET (5–8 p.m.), and passive stretching in CG (half of the patients in each time of the day). Assessments included BP, systemic hemodynamics (cardiac output and peripheral vascular resistance), and vascular parameters (carotid intima–media thickness, arterial stiffness, brachial vascular conductance, and endothelial function) measured before and after the interventions. Aerobic capacity improved only and similarly in both training groups (P = 0.035), while no significant changes were observed in BP, systemic hemodynamics, or vascular parameters in either group (all, P > 0.05). The effects of the interventions adjusted for the pre-intervention values also showed no differences among the groups (all, P > 0.05). Thus, in medicated elderly hypertensives, 10 weeks of aerobic training, whether performed in the morning or evening, improved cardiorespiratory fitness but did not reduce BP nor modify systemic hemodynamics and vascular function. These findings suggest population-specific responses to aerobic training in hypertension.

Graphical Abstract

Introduction

Systemic arterial hypertension is a chronic asymptomatic disease characterized by elevated systolic (SBP) and/or diastolic (DBP) blood pressure (BP) [1, 2]. Its global prevalence is substantial, affecting approximately 1.3 billion adults worldwide in 2019 [3]. Additionally, hypertension prevalence increases considerably with aging, reaching 59% in people over 65 years [2].

The pathophysiology of hypertension involves structural and functional alterations in the vasculature, such as an increased arterial wall thickness [4, 5] and stiffness [6], as well as impaired vascular conductance [4] and endothelial function [7, 8]. Together, these alterations contribute to increase peripheral vascular resistance (PVR) and BP [9, 10]. Therefore, interventions specifically targeting these mechanisms are important for more effective hypertension management, particularly for elderly patients in whom vascular alterations are often more pronounced and associated with greater risks [4, 5, 11].

Current hypertension management encompasses both pharmacological and non-pharmacological approaches [2, 12]. Among the non-pharmacological strategies, aerobic training has significant effects on BP control, with a meta-analysis reporting reduction between 8–12 mmHg in SBP and 4–6 mmHg in DBP [13]. Additionally, the BP reductions induced by aerobic training have been primarily attributed to decreased PVR [14, 15], resulting from vascular adaptations such as reduction of arterial wall thickness [16] and arterial stiffness [17], as well as improvement in vascular conductance [18] and endothelial function [19].

Despite the well-documented BP-lowering effect of aerobic training, there is a huge variation among the studies, suggesting that some factors may affect this response. Along this line, previous studies from our group showed that aerobic training performed in the evening elicits a greater BP-lowering effect due to a greater reduction in PVR and muscular sympathetic nerve activity than in the morning [20, 21]. However, the impact of training time on vascular function has not been investigated yet, leaving an important knowledge gap, particularly for elderly hypertensives.

In parallel to the 24-h BP fluctuations, vascular function also shows circadian variations. Thus, the vascular components associated with BP elevation, such as arterial stiffness, typically peak in the morning [22], while components associated with BP decrease, including vascular conductance and endothelial function, reach their peak activity in the evening [23]. Based on these circadian dynamics and on our previous results, it was possible to hypothesize that evening aerobic training may potentiate the depressor components of the vascular function, leading to a greater reduction in PVR and BP. This hypothesis needed to be checked in elderly patients with hypertension who may specially benefit from the vascular improvements induced by aerobic training.

Therefore, this study was designed to evaluate and compare the effects of aerobic training performed at different times of the day (morning and evening) on BP, systemic hemodynamics, and vascular function in elderly patients with hypertension.

Methods

Patients were recruited through digital and print media or through direct contact at public health campaigns, other social venues, and the University hospital. This study was conducted in accordance with the Declaration of Helsinki. All patients signed a consent form after being informed about the procedures, risks, and benefits of the study’s participation. This study is part of a bigger trial that was approved by the local Ethics Committee (CAAE: 02518918.7.0000.5391) and registered on the Brazilian Clinical Trials platform (U1111-1242–4972). Data regarding the autonomic adaptations obtained with training in part of the present sample has been published elsewhere [21]. The present paper includes data from the whole sample regarding the effects of training on vascular function. All experimental procedures were conducted following the approved ethics submission document.

Patients and eligibility

The participants were elderly (≥ 60 years) patients of both sexes who were regularly taking antihypertensive medication, with the drug class and dose maintained for at least the last 4 months. The patients were excluded if they (1) had secondary hypertension, target organ damage, and/or other cardiovascular diseases besides hypertension; (2) were taking β-blockers, non-dihydropyridine calcium channel inhibitors, and/or insulin; (3) were obese grade II or higher; (4) had resting SBP/DBP > 160/105 mmHg; (5) had an extremely morning or evening chronotype; (6) were physically active or performed structured physical training more than once a week; (7) had musculoskeletal limitations that precludes physical exercise; (8) showed ischemia or complex arrhythmias during exercise; and (9) did not complete at least 75% of the programed interventions’ sessions. In addition, all the women had to be post-menopausal and not taking hormone therapy. Adherence to the study criteria was verified by a set of preliminary assessments.

Preliminary assessments

Clinical assessment

Patients were interviewed by a physician regarding their personal health characteristics and medication use and were excluded when presenting any health problem or medication precluded by the study criteria.

Anthropometric assessment

Body mass and height were measured using a scale with a stadiometer (Filizola S.A, Personal, Campo Grande, Brazil), and body mass index (BMI) was calculated. The patients with BMI ≥ 35 kg/m² were excluded [24].

BP assessment

In two visits to the laboratory, BP was measured in triplicate in both arms after 5 min of seated rest using the auscultatory method, a mercury column sphygmomanometer (Unitec, São Paulo, Brazil), and employing phases I and V of the Korotkoff sounds to identify SBP and DBP, respectively. The mean value of the six measures was calculated for each arm and the greater value between the arms was accepted as the patient’s BP. Patients were excluded if SBP was greater than 160 mmHg and/or DBP was greater than 105 mmHg [25].

Chronotype identification

Chronotype was assessed using the Horne and Ostberg questionnaire [26] that categorizes chronotype as morning, evening, or intermediate. Only patients with scores > 30 and < 70 were studied, excluding those with extremely morning or evening chronotypes to avoid any possible influence of preference for one time of the day.

Maximal cardiopulmonary exercise test

All patients underwent a maximal cardiopulmonary exercise test on a cycle ergometer (Corival Cycle; Lode, Groningen, Netherlands) carried out between 1 and 5 p.m. in a thermoneutral laboratory (20–22 °C) [27]. The test was preceded by a resting ECG (EMG System do Brazil, EMG 030110/00B, São Paulo, Brazil) and followed a protocol of 15-W increment per minute until exhaustion. During the test, respiratory gases were analyzed by a metabolic cart in averages of 30 s (Medical Graphics Corporation CPX/Ultima, Minnesota, USA). The test followed the recommendations of the Brazilian Society of Cardiology [27]. The patients who presented ECG alterations suggesting ischemia and/or complex arrhythmia were excluded. Additionally, the ventilatory thresholds (i.e., anaerobic threshold and respiratory compensation point) were visually identified by two experienced researchers using the Skinner and McLellan criteria [28], and peak oxygen consumption (VO₂peak) was determined by the highest value reached during the exercise. This test was repeated at the end of the study, keeping an interval of at least 48 h from the last intervention session.

Experimental protocol

The patients who met all the study criteria underwent an initial evaluation in which BP, systemic hemodynamics, and vascular function were assessed. Then, they were allocated to one of three groups: morning training (MT—between 7 and 10 a.m.), evening training (ET—5 to 8 p.m.), or control (CG—half in the morning and half in the evening). Initially, the patients were randomly allocated for the three groups in blocks of three. Within each block, the first patient selected a paper from a brown envelope that indicated the allocation to MT, ET, or CG; the second patient selected another paper from the same envelope; and the third patient was allocated to the remaining group. Due to the difficulty observed after the COVID-19 for recruiting elderly volunteers, the last eight patients included in the study were allocated according to their availability for coming in the morning or the evening, and were sorted for the control or training intervention at the available time of day. After the initial evaluation, the patients participated 3 times/week in the intervention sessions according to their group allocation. Then, after 10 weeks of intervention, they underwent the final evaluation with the same experimental protocol of the initial evaluation.

The initial and final evaluations were carried out between 2 and 4 p.m. in a temperature-controlled laboratory (20–22 °C). For these evaluations, after arriving at the laboratory, patients rested in the seated position for 20 min, and then seated BP and systemic hemodynamics were measured in triplicate. Afterwards, they lay down and rested in the supine position for 70 min. During this period, from 15 to 25 min, the carotid artery structure and function were assessed, and from 25 to 40 min, BP and HR were measured, and the brachial artery function was assessed. Finally, from 50 to 70 min, supine BP and systemic hemodynamics were measured in triplicate.

Measurements

Blood pressure

BP was measured in the non-dominant arm using the auscultatory method, a mercury column sphygmomanometer (Uniteq, São Paulo, Brazil), and considering phases I and V of the Korotkoff sounds to identify SBP and DBP, respectively. Mean blood pressure (MBP) was calculated by MBP = (SBP − DBP)/3 + DBP. These measurements were performed by the same researcher in the same patient before and after the intervention.

Systemic hemodynamics

Cardiac output (CO) was estimated using the adapted Fick equation [29] and the CO₂ rebreathing technique following the protocol suggested by Collier [30]. For that, the NICO software from a gas analyzer (Medical Graphics Corporation CPX/Ultima, Minnesota, USA) and a gas mixture with 8% CO₂ and 35% O₂ were employed. Thus, PVR was calculated by MBP/CO. Additionally, heart rate (HR) was assessed by radial pulse count for 15 s, and stroke volume (SV) was calculated by CO/HR.

Carotid artery structure and function

The intima–media thickness (IMT) and the carotid artery stiffness were determined by duplex ultrasound using a linear matrix (L9-3 RS probe; Logiq E (R7), California, USA) with an insonation angle of 60° [31]. To assess IMT, longitudinal images of the carotid artery were recorded in the Brightness “B” mode for 30 s. Then, the images were analyzed for an automatic detection of the artery walls (Cardiovascular Suite, Quipu srl, Pisa, Italy), and the IMT of the common carotid artery was defined as the mean distance between the leading edge of the first echogenic line (lumen–intima) and the leading edge of the second echogenic line (media–adventitia) [31]. The carotid artery stiffness was calculated using the beta index (β), a gold-standard parameter due to its high correlation with pulse wave velocity, as proposed by Spronck et al. [32]. These evaluations were carried out by an experienced evaluator who was blinded to the interventions.

Brachial artery function

The brachial artery measurements were taken on the patient’s right arm (~ 10 cm above the cubital fossa) using a duplex ultrasound and a linear matrix (L9-3 RS probe; Logiq E (R7)) with an insonation angle of 60° [31]. Initially, the artery images were recorded for 1 min to obtain blood flow velocity and the artery diameter in ultrasound Doppler mode using an automatic software (Cardiovascular Suite, Quipu Srl, Pisa, Italy). Then, a cuff positioned on the forearm of the same arm (just after the cubital fossa) was inflated to 250 mmHg and promoted arterial occlusion for 5 min. After this period, the cuff was deflated and the images to obtain blood flow velocity and artery diameter were recorded for 3 min.

Baseline blood flow was calculated for the minute before occlusion by multiplying the diameter of the artery by the mean blood flow velocity and was expressed in milliliters per minute. Baseline vascular conductance (VC) was calculated by dividing baseline blood flow by MBP taken immediately before the blood flow measurement. Peak blood flow was assessed as the blood flow at the peak diameter after occlusion release. Endothelial function was assessed by flow-mediated dilation (FMD) that was calculated as the maximum percentage increase in artery diameter after the occlusion release by the following formula: FMD (%) = [(peak diameter − baseline diameter)/baseline diameter] × 100. These evaluations were carried out by an experienced evaluator who was blinded to the interventions. In addition, all vascular images were coded and so the evaluator who analyzed the videos was also blinded for the allocation group.

Interventions

MT occurred between 7 and 10 a.m., ET occurred between 5 and 8 p.m., and in the CG group, half of the patients participated between 7 and 10 a.m. and the other half between 5 and 8 p.m. Both training groups (MT and ET) used the same training protocol that was already employed in our previous studies [20, 21]. Briefly, aerobic training was carried out on a cycle ergometer and consisted of 30 sessions. The training duration increased from 30 to 45 min in the first 2 weeks and then was kept constant. Training intensity began at the HR of the anaerobic threshold and after the 4 weeks, it increased progressively to achieve the HR corresponding to 10% below the respiratory compensation point at the fifth week. The CG group performed 30 sessions of passive stretching following a protocol that has been applied before and did not result in any change in BP in patients with hypertension [20].

Statistical analysis

The minimum sample size required for this study was calculated as 39 patients (13 in each group). This calculation considered SBP as the primary outcome, a statistical power of 0.90, an α value of 0.05, and an effect size of 0.61 based on our previous study [20]. Considering the greater variability of the other variables and the occurrence of COVID-19 pandemic during the study execution, a greater number of patients was recruited.

The normality of the distribution and the homogeneity of the data were checked using the Shapiro–Wilk and the Levene tests, respectively. To compare the characteristics of the groups, one-way ANOVAs were used for the continuous variables and chi-square or Fisher’s exact tests for the categorical variables.

To assess and compare the effects of aerobic training executed at different times of the day, two-way mixed ANOVAs were used, employing group (MT, ET, and CG) as the between main factor and evaluation (initial and final) as the within factor. Additionally, the changes observed in each group (i.e., Δ = final − initial values) were adjusted for the initial values of each variable and compared between the groups (MT, EDT, and CG) by one-way ANCOVAs. Bonferroni post hoc tests were used when appropriate. Finally, the partial eta-squared effect size (η_p²) was calculated, being considered non-existent when < 0.01; small when between 0.01 and 0.06; moderate between 0.06 and 0.14; and large when greater than 0.14 [33].

All analyses were conducted using JASP (version 0.19.3, Amsterdam, The Netherlands) and the significance level adopted was P ≤ 0.05. Data are expressed as absolute and relative frequencies for categorical variables and as mean ± standard deviation for the continuous variables.

Results

This study was conducted from August 2018 to December 2024 with interruptions or reductions in data collection for some periods during the COVID-19 pandemic.

Sample characteristics

After contacting almost 380 volunteers, 123 patients accepted to participate and signed the informed consent (Fig. 1). Of them, 25 were excluded during the preliminary assessments for not fulfilling the study criteria, and 30 dropped out due to unavailability to undergo the study procedures. Thus, 68 patients underwent the initial evaluation and were randomized for the 3 groups (CG = 23, MT = 23, and ET = 22). However, 14 patients dropped out during the intervention period (most of them because of COVID pandemic or personal reasons). Therefore, 54 patients (CG = 17, MT = 19, and ET = 18) completed the experimental protocol.

Fig. 1.

Participants’ flowchart

There were no differences in the anthropometric, hemodynamic, and clinical characteristics among the three groups at the beginning of the study (Table 1, all P > 0.05). However, chronotype score and the use of combined therapy were lower in the ET compared to the other groups.

Table 1.

Characteristics of the study sample

	CG	MT	ET	P
n	17	19	18	–
Males (n/%)	8 (28%)	10 (36%)	16 (36%)	0.878
Age (years)	69 ± 4	67 ± 7	65 ± 5	0.665
Chronotype (score)	63 ± 5	65 ± 5	57 ± 10 #†	0.003
Anthropometrics
Body mass (kg)	79.8 ± 13.5	76.6 ± 12.0	80.4 ± 18.5	0.705
Height (m)	1.7 ± 0.1	1.6 ± 0.1	1.6 ± 0.1	0.921
BMI (kg/m2)	29.1 ± 3.8	28.4 ± 3.2	29.1 ± 3.9	0.788
Cardiovascular risk factors
Smoking (n/%)	2 (67%)	1 (33%)	0 (0%)	0.315
Diabetes (n/%)	7 (47%)	6 (40%)	2 (13%)	0.126
Hemodynamics
Resting SBP (mmHg)	131 ± 15	127 ± 15	134 ± 12	0.310
Resting DBP (mmHg)	82 ± 9	79 ± 9	85 ± 8	0.109
Resting HR (bpm)	76 ± 9	73 ± 9	72 ± 8	0.346
Number of antihypertensive medications
One (n/%)	6 (29%)	4 (19%)	11 (52%) #†	0.041
Two or more (n/%)	11 (33%)	15 (46%)	7 (21%) #†	0.041
Antihypertensive drugs
ARB (n/%)	12 (30%)	15 (37%)	13 (33%)	0.829
ACEi (n/%)	3 (30%)	4 (40%)	3 (30%)	0.937
DHP-CCB (n/%)	5 (28%)	6 (33%)	7 (39%)	0.821
Diuretics (n/%)	10 (38%)	11 (42%)	5 (20%)	0.106

Data: mean ± SD or n (%). Analysis: one-way ANOVA or chi-square or Fisher’s exact tests

BMI body mass index, SBP systolic blood pressure, DBP diastolic blood pressure, HR heart rate, ARB angiotensin receptor blocker, ACEI angiotensin-converting enzyme inhibitors, DHP-CCB dihydropyridine calcium channel blockers

^#Different from CG (P < 0.05)

^†Different from MT (P < 0.05)

Adherence to the training sessions was similar and greater than 90% in both training groups. Training intensity increased similarly and as planned in both groups (from 97 ± 7 and 101 ± 6% of the HR of the anaerobic threshold to 85 ± 8 and 86 ± 4% of the HR of the respiratory compensation point for MT and ET, respectively, all P > 0.05).

VO₂peak increased significantly only in the trained groups (MT = 19.0 ± 4.8 vs. 20.8 ± 5.8, ET = 18.2 ± 4.3 vs. 20.0 ± 5.3 and CG = 17.4 ± 3.9 vs. 17.5 ± 3.6 ml.kg⁻¹.min⁻¹, P_interaction = 0.035). Body mass index (MT = 28.4 ± 3.4 vs. 28.2 ± 3.6, ET = 29.1 ± 4.0 vs. 29.1 ± 3.9, and CG = 28.9 ± 3.7 vs. 28.8 ± 3.7 kg/m², P_interaction = 0.575) did not change from the initial to final evaluation in either group.

When assessed in the seated position, SBP, DBP, MBP, CO, PVR, SV, and HR did not change significantly from the initial to the final evaluation in either group (Table 2). Additionally, the changes in these variables adjusted for their initial values also showed no significant differences among the groups (Fig. 2). Nevertheless, regardless of the evaluation (initial or final), DBP and MBP were significantly higher in the ET than MT (Table 2).

Table 2.

Hemodynamic variables measured in the seated position at the initial and final evaluations in the control (CG), morning training (MT), and evening training (ET) groups

		Initial	Final
SBP (mmHg)
	CG	127 ± 13	123 ± 15	Pgroup = 0.291
	MT	125 ± 12	119 ± 11	Pphase = 0.057
	ET	127 ± 10	126 ± 11	Pint = 0.572
DBP (mmHg)
	CG	81 ± 8	81 ± 10	Pgroup = 0.023
	MT	79 ± 5	75 ± 8	Pphase = 0.241
	ET	84 ± 9 (†)	83 ± 7 (†)	Pint = 0.478
MBP (mmHg)
	CG	96 ± 8	95 ± 11	Pgroup = 0.020
	MT	94 ± 5	90 ± 8	Pphase = 0.094
	ET	98 ± 8 (†)	98 ± 6 (†)	Pint = 0.500
CO (l/min)
	CG	4.9 ± 1.6	4.7 ± 1.6	Pgroup = 0.955
	MT	4.7 ± 1.4	4.6 ± 1.3	Pphase = 0.610
	ET	4.7 ± 0.9	4.7 ± 0.9	Pint = 0.989
PVR (U)
	CG	22 ± 9	22 ± 6	Pgroup = 0.941
	MT	22 ± 9	21 ± 6	Pphase = 0.630
	ET	22 ± 5	22 ± 5	Pint = 0.837
SV (ml)
	CG	68 ± 21	67 ± 20	Pgroup = 0.995
	MT	66 ± 19	69 ± 25	Pphase = 0.899
	ET	69 ± 18	67 ± 16	Pint = 0.645
HR (bpm)
	CG	72 ± 11	71 ± 8	Pgroup = 0.850
	MT	72 ± 9	68 ± 8	Pphase = 0.244
	ET	70 ± 9	70 ± 10	Pint = 0.339

Data: mean ± SD. Analysis: two-way mixed ANOVAs. † different from MT (P < 0.05), () main factor effect

int interaction, SBP systolic blood pressure, DBP diastolic blood pressure, MBP mean blood pressure, CO cardiac output, PVR peripheral vascular resistance, SV stroke volume, HR heart rate

Fig. 2.

Changes observed from the initial to the final evaluations and adjusted for the initial values (Δa) for the hemodynamic variables assessed in the seated position in the control (CG), morning training (MT), and evening training (ET) groups

When assessed in the supine position, SBP, DBP, MBP, CO, PVR, SV, and HR did not change significantly from the initial to the final evaluation in any group (Table 3). In addition, the changes in these variables adjusted for their initial values also revealed no significant differences among the groups (Fig. 3).

Table 3.

Hemodynamic variables measured in the supine position at the initial and final evaluations in the control (CG), morning training (MT), and evening training (ET) groups

		Initial	Final
SBP (mmHg)
	CG	127 ± 17	126 ± 20	Pgroup = 0.528
	MT	123 ± 14	124 ± 14	Pphase = 0.671
	ET	130 ± 12	127 ± 13	Pint = 0.646
DBP (mmHg)
	CG	82 ± 8	81 ± 9	Pgroup = 0.154
	MT	81 ± 7	79 ± 7	Pphase = 0.364
	ET	85 ± 9	85 ± 8	Pint = 0.812
MBP (mmHg)
	CG	97 ± 9	96 ± 11	Pgroup = 0.223
	MT	95 ± 7	94 ± 8	Pphase = 0.473
	ET	100 ± 9	99 ± 8	Pint = 0.977
CO (l/min)
	CG	4.8 ± 1.1	4.8 ± 1.4	Pgroup = 0.457
	MT	5.1 ± 1.1	4.9 ± 1.5	Pphase = 0.982
	ET	4.4 ± 0.8	4.7 ± 1.0	Pint = 0.700
PVR (U)
	CG	21 ± 7	21 ± 6	Pgroup = 0.452
	MT	20 ± 6	21 ± 6	Pphase = 0.935
	ET	23 ± 4	22 ± 6	Pint = 0.212
SV (ml)
	CG	71 ± 19	73 ± 21	Pgroup = 0.534
	MT	78 ± 25	77 ± 25	Pphase = 0.674
	ET	70 ± 13	72 ± 17	Pint = 0.848
HR (bpm)
	CG	69 ± 10	66 ± 8	Pgroup = 0.331
	MT	66 ± 8	65 ± 8	Pphase = 0.411
	ET	63 ± 6	65 ± 8	Pint = 0.206

Data: mean ± SD. Analysis: two-way mixed ANOVAs

int interaction, SBP systolic blood pressure, DBP diastolic blood pressure, MBP mean blood pressure, CO cardiac output, PVR peripheral vascular resistance, SV stroke volume, HR heart rate

Fig. 3.

Changes observed from the initial to the final evaluations and adjusted for the initial values (Δa) for the hemodynamic variables assessed in the supine position in the control (CG), morning training (MT), and evening training (ET) groups

Considering the variables related to the vascular structure and function (i.e., carotid IMT and stiffness as well as brachial vascular conductance, baseline blood flow, peak blood flow, and FMD), they did not change significantly from the initial to the final evaluation in either group (Table 4), and the changes in these variables adjusted for their initial values also revealed no significant differences among the groups (Fig. 4).

Table 4.

Vascular variables measured at the initial and final evaluations in the control (CG), morning training (MT), and evening training (ET) groups

		Initial	Final
Carotid IMT (mm)
	CG	0.69 ± 0.14	0.69 ± 0.13	Pgroup = 0.519
	MT	0.71 ± 0.13	0.71 ± 0.12	Pphase = 0.880
	ET	0.68 ± 0.10	0.66 ± 0.14	Pint = 0.864
Carotid stiffness (m/s)
	CG	6.9 ± 2.0	6.1 ± 1.7	Pgroup = 0.684
	MT	6.4 ± 1.5	7.1 ± 1.7	Pphase = 0.858
	ET	6.8 ± 2.1	7.0 ± 1.9	Pint = 0.117
Baseline brachial blood flow (ml/min)
	CG	125.4 ± 79.9	111.1 ± 58.2	Pgroup = 0.843
	MT	150.7 ± 134.1	112.6 ± 53.5	Pphase = 0.328
	ET	115.2 ± 42.4	136.4 ± 62.4	Pint = 0.072
Baseline brachial vascular conductance (ml.min-1.mmHg-1)	CG	1.30 ± 0.81	1.18 ± 0.60	Pgroup = 0.716
	MT	1.59 ± 1.36	1.27 ± 0.64	Pphase = 0.599
	ET	1.16 ± 0.45	1.42 ± 0.70	Pint = 0.083
Peak brachial blood flow (ml/min)
	CG	828.4 ± 364.4	749.3 ± 327.7	Pgroup = 0.874
	MT	770.6 ± 323.9	768.1 ± 332.8	Pphase = 0.522
	ET	733.1 ± 246.6	898.8 ± 296.6	Pint = 0.076
Braquial FMD (%)
	CG	8.9 ± 5.8	8.4 ± 6.1	Pgroup = 0.049
	MT	5.1 ± 3.1	5.9 ± 3.6	Pphase = 0.284
	ET	5.3 ± 4.0	7.0 ± 4.5	Pint = 0.430

Data: mean ± SD. Analysis: two-way mixed ANOVAs

IMT intima–media thickness, FMD flow-mediated dilation

Fig. 4.

Changes observed from the initial to the final evaluations and adjusted for the initial values (Δa) for the vascular variables in the control (CG), morning training (MT), and evening training (ET) groups

For most of the variables, the effect sizes of the interventions adjusted for their initial values were considered inexistent or small. For seated DBP (η²_p = 0.088), carotid stiffness (η²_p = 0.082), baseline vascular conductance (η²_p = 0.074), and peak blood flow (η²_p = 0.086), the effect sizes were moderate but near the limit for small (η²_p = 0.06). Only for baseline blood flow, the effect size was moderate and near large (η²_p = 0.105).

Discussion

The main findings of this study were that in medicated elderly hypertensive patients, aerobic training, whether performed in the morning or the evening, did not reduce BP as assessed both in the seated and supine positions. Additionally, the proposed training protocol did not modify PVR, carotid structure and function, as well as brachial vascular function. These findings contrast with previous research employing the same training protocol in middle-aged hypertensive men, highlighting subpopulation-specific responses to aerobic training in hypertension.

The hypothesis of this study was that aerobic training would decrease BP, especially when conducted in the evening. Nevertheless, neither seated nor supine SBP or DBP changed significantly in either MT or ET groups. These findings were confirmed when the effects of the interventions were adjusted for the initial values as well as by the non-existent to small effect sizes observed for BPs. Thus, the aerobic training protocol employed in this study was unable to reduce BP in elderly hypertensives. The absence of BP reduction contrasts with the accepted concept that aerobic training lowers BP in hypertensives [1, 2, 34] and with our previous study in which ET induced greater BP reduction than MT [20]. Some factors may be discussed to explain this null result.

First, a possible concern may be whether the training protocol was properly executed since another study using a similar protocol reported BP decrease in hypertensives [35]. Nevertheless, the training intensity progressed as planned, and the adherence was high in both training groups. Additionally, and most importantly, VO₂peak, a marker of aerobic training efficacy [25], increased 9–10% in both training groups, which is similar to our previous study that reported BP decrease after aerobic training [20]. Therefore, the training protocol was properly employed and effective in inducing aerobic improvements.

Another explanation may be the characteristics of the subpopulation of hypertensives studied. The current study enrolled elderly patients of both sexes, receiving anti-hypertensive medication, and with BP levels lower than 160/105 mmHg. Age-related differences in exercise responsiveness have been reported. Multicenter research [36] comparing the effect of aerobic training in young (30–49 years) and middle-aged/elderly (50–69 years) patients reported greater BP reduction in the younger participants. This blunted responsiveness in the elderly has been attributed to an age-related difficulty for vascular adaptations [37, 38]. Regarding sexes, a meta-analysis reported similar BP responses to aerobic training in men and women [39]. Additionally, a complementary analysis of the present data, including sex as a between main factor in ANOVAs, revealed no change in the results (data not shown, all P > 0.05). Considering the medication status, systematic reviews suggest that the BP-lowering effect of aerobic training may be lower in medicated patients [39, 40], which may reflect the difficulty of aerobic training to induce further adaptations on BP regulatory mechanisms beyond medication. In addition, the absence of response may reflect the lower initial BP level induced by the medication. It is well established that training induces smaller BP reduction when the initial BP levels are lower [41], and well-controlled hypertensives (i.e., SBP and DBP lower than 140 and 90 mmHg, respectively) present attenuated responses to training [40, 42]. In the present study, 59% of the patients in the training groups had well-controlled BP and 22% had normal BP (i.e., ≤ 120/80 mmHg) at the initial assessments. Together, all these aspects (i.e., elderly patients, medication use, and lower initial BP) may be responsible for the absence of BP reduction in the present study.

The lack of BP reduction can also be explained by the fact that neither morning nor evening training modified the BP mechanisms assessed in the present study. Aerobic training did not change CO, PVR, VS, and HR assessed either at the seated or the supine positions. Although aerobic training is expected to reduce HR, this adaptation is often associated with higher training intensities [43]. Additionally, several patients (32%) were taking amlodipine that can increase sympathetic nerve activity [44] and potentially counteract the training-induced HR decrease. The absence of HR decrease explains the failure to decrease CO, which is sometimes reported after aerobic training [45]. Concerning the carotid artery, the lack of change in its structure and function is in line with a previous meta-analysis conducted with pre-hypertensive and hypertensive patients [46], although studies with healthy adults have reported reduction in IMT and increase in carotid distensibility with aerobic training [16, 17]. Thus, the presence of hypertension together with the age of the participants may have affected the arterial adaptations to training since these two aspects (age and hypertension) promote structural changes in the vessels (i.e., increased collagen deposition and elastin fragmentation) that may require longer or more intensive training to reverse [38]. Concerning the lack of effect of training on brachial artery function, although a previous meta-analysis reported an average increase of + 1.45% in FMD in hypertensive patients [19], a more comprehensive meta-analysis [46] concluded that there is not enough evidence that exercise can improve endothelial function in sedentary healthy elderly adults. In addition, antihypertensive medications, particularly those targeting the renin–angiotensin–aldosterone system, used by 95% of the present sample, can improve endothelial function [47], limiting any additional benefit from aerobic training. Finally, other physiological mechanisms associated with hypertension, such as autonomic dysfunction, oxidative stress, and inflammation, that can be modulated by aerobic training [9, 18] were not assessed in the present study. However, it is probable that they had also not changed with training, which can be attributed to the same explanations applied for the vascular mechanisms, i.e., the patient’s age combined with the use of medication blunting or mitigating the responsiveness to training.

In addition to the previous discussion, it is also possible that specific antihypertensive drugs differently influence the circadian BP pattern, potentially affecting the time-of-day effects of aerobic training [48–50]. Along this line, calcium channel blockers may affect the diurnal variation of BP, potentially reducing the differential impact of morning and evening exercise [51, 52]. In the present study, approximately 40% of the patients in the ET group were receiving calcium channel blockers, which can have impacted the results. However, many patients were receiving more than one class of medication and the administration schedule of the doses (only morning or morning and evening) differed according to the medication class. Thus, the present data does not allow for clarifying this possible influence that should be investigated in the future.

By revealing no effect of aerobic training in elderly hypertensives, the results of the present study have important implications. Publishing studies with null findings helps to avoid publication bias and may guide future research and clinical applications for specific populations. Along this line, by showing that an aerobic training protocol that reduces BP in middle-aged hypertensive men could not decrease BP in elderly hypertensives, the present study highlights the importance of replicating studies in different subpopulations of hypertensives. This finding also challenges the “one-size-fits-all” approach for exercise prescription in hypertension management, suggesting that exercise recommendations should be tailored to specific subpopulations, particularly considering age and medication status. Therefore, it is reasonable to recommend training protocols with longer duration, higher intensity, or different modalities to reduce BP for elderly hypertensive patients. Additionally, the discussion revealed a knowledge gap regarding the interplay between antihypertensive medication and aerobic training, emphasizing the need for future research to consider the type and time of medication administration. Finally, despite the null effect on BP, the present results support the recommendation of aerobic training to elderly hypertensive patients since it improved cardiorespiratory fitness that is independently and negatively associated with cardiovascular risk and mortality [53] with an increase of 1-MET representing an improvement of 12% in survival [54].

This study’s limitations include the following: First, the findings cannot be generalized beyond the specific subpopulation included in the present study. Second, although powered for the primary outcome (i.e., SBP), the number of subjects could seem small for the secondary outcomes, which is unlikely since the effect sizes were non-existent or small for almost all variables. Third, biological preference for a phase of day may differently influence the effects of morning and evening training. However, the present study only included patients with intermediate chronotype, and even within this subtype, complementary analyses incorporating chronotype value as a covariate revealed no change in the results (data not shown). Fourth, the assessments were conducted at a standardized time of the day (i.e., afternoon), and the results may be different if evaluated at other times since training adaptations can be potentiated for the specific time of training [55]. Finally, the patients were taking different medications that can differently influence responses to aerobic training at different times of the day. Therefore, further multicenter research is needed for enrolling a substantial sample of hypertensive patients taking different classes of medications and undergoing evaluations at various times of the day.

Among the strengths noted in the study, the inclusion of both sexes can be highlighted as well as the excellent adherence to the training protocol, the similar results obtained with measures taken in the supine and seated positions, and the similar characteristics of the experimental groups. Together, these aspects strengthen that no differences between the groups or the training protocols had affected the study findings, and that the absence of effect was not limited to a specific body position.

Conclusion

Ten weeks of an aerobic training protocol that lowered BP in middle-aged hypertensive men, whether performed in the morning or evening, did not reduce BP nor modify systemic hemodynamics or vascular function in medicated elderly hypertensive patients. These findings highlight the need for guidelines to include subpopulation-specific exercise prescriptions for controlling BP in hypertension. Additionally, given that increment in cardiorespiratory fitness is associated with reduced cardiovascular morbidity and mortality, clinicians should recommend aerobic training for medicated elderly hypertensive patients even in the absence of additional BP reduction.

Author contribution

L.A., L.B., and C.F.: conception. L.A., L.B., G.O., T.M., and S.A.: data collection. N.SJ. and L.C.: blinded data analysis. L.A.: analysis and draft of manuscript. C.F.: supervision of all phases. L.B. and C.F.: critique, revision, and approval of final manuscript. All people designated as authors qualify for authorship and all those who qualify for authorship are listed and have read and approved the final version of this manuscript and agree to be accountable for all aspects of this work.

Funding

This study was supported by the São Paulo Research Foundation (FAPESP—2022/12605–3; 2018/05226–0), the Brazilian National Council for Scientific and Technological Development (CNPQ—302309/2022–5), the Coordination for the Improvement of Higher Education Personnel (CAPES—0001), and the American Heart Association (24CDA1267757). Basically, the funding consisted of scholarships and financial support for purchasing the equipment and materials necessary to carry out the research.

Data availability

Data are available from the corresponding author upon reasonable request.

Declarations

Compliance with ethical standards

This study was conducted in accordance with the Declaration of Helsinki. All patients signed a consent form after being informed about the procedures, risks, and benefits of the study’s participation. The study protocol was approved by the local Ethics Committee and registered on the Brazilian Clinical Trials platform (U1111-1242–4972). All experimental procedures were conducted following the approved ethics submission document.

Conflict of interest

The authors declare no competing interests.