Inspiratory and leg muscle blood flows during inspiratory muscle metaboreflex activation in heart failure with preserved ejection fraction

Abstract

The purpose of this study was to determine the cardiovascular consequences elicited by activation of the inspiratory muscle metaboreflex in patients with heart failure with preserved ejection fraction (HFpEF) and controls. Patients with HFpEF (n = 15; 69 ± 10 yr; 33 ± 4 kg/m²) and controls (n = 14; 70 ± 8 yr; 28 ± 4 kg/m²) performed an inspiratory loading trial at 60% maximal inspiratory pressure (P_IMAX) until task failure. Mean arterial pressure (MAP) was measured continuously. Near-infrared spectroscopy and bolus injections of indocyanine green dye were used to determine the percent change in blood flow index (%ΔBFI) from baseline to the final minute of inspiratory loading in the vastus lateralis and sternocleidomastoid muscles. Vascular resistance index (VRI) was calculated. Time to task failure was shorter in HFpEF than in controls (339 ± 197 s vs. 626 ± 403 s; P = 0.02). Compared with controls, patients with HFpEF had a greater increase from baseline in MAP (16 ± 7 vs. 10 ± 6 mmHg) and vastus lateralis VRI (76 ± 45 vs. 32 ± 19%) as well as a greater decrease in vastus lateralis %ΔBFI (−32 ± 14 vs. −17 ± 9%) (all, P < 0.05). Sternocleidomastoid %ΔBFI normalized to absolute inspiratory pressure was higher in HFpEF compared with controls (8.0 ± 5.0 vs. 4.0 ± 1.9% per cmH₂O·s; P = 0.03). These data indicate that patients with HFpEF exhibit exaggerated cardiovascular responses with inspiratory muscle metaboreflex activation compared with controls.

NEW & NOTEWORTHY Respiratory muscle dysfunction is thought to contribute to exercise intolerance in heart failure with preserved ejection fraction (HFpEF); however, the underlying mechanisms are unknown. In the present study, patients with HFpEF had greater increases in leg muscle vascular resistance index and greater decreases in leg muscle blood flow index compared with controls during inspiratory resistive breathing (to activate the metaboreflex). Furthermore, respiratory muscle blood flow index responses normalized to pressure generation during inspiratory resistive breathing were exaggerated in HFpEF compared with controls.

INTRODUCTION

The pathophysiology underlying heart failure with preserved ejection fraction (HFpEF) is complex and involves multiple physiological systems (1–3). Impaired exercise capacity (e.g., reduced peak oxygen uptake) and exercise intolerance (e.g., limited submaximal exercise endurance) are defining characteristics of HFpEF symptomology and are strongly associated with reduced quality of life and mortality (4–6). In addition to well-known central limitations [e.g., reduced cardiac reserve (7, 8)], peripheral abnormalities also contribute to exercise intolerance in patients with HFpEF (9–11). Respiratory muscle dysfunction including inspiratory muscle weakness and reduced fatigue resistance represents one potential mechanism that may contribute to exertional symptoms (i.e., dyspnea and fatigue) and limit exercise capacity in HFpEF (9). Previous studies have demonstrated that respiratory muscle dysfunction in HFpEF is associated with reduced functional capacity (12), poor patient prognosis (13), and greater mortality (13). However, the underlying mechanisms by which respiratory muscle dysfunction contributes to exercise intolerance in these patients are unclear.

In healthy adults, stimulation of neural afferents in response to metabolite accumulation during high inspiratory muscle work (i.e., inspiratory metaboreflex activation) elicits increases in mean arterial pressure (MAP) and limb vascular resistance through greater sympathetic outflow resulting in decreased limb blood flow (14–18). It is unclear whether the inspiratory muscle metaboreflex is abnormal in HFpEF. Impaired inspiratory muscle blood flow could accentuate metabolite accumulation and thereby amplify activation of the inspiratory muscle metaboreflex (19). To this point, recent studies have suggested that skeletal muscle blood flow responses may be reduced during submaximal exercise in patients with HFpEF (20). Alternatively, if the inspiratory muscle metaboreflex is exaggerated in HFpEF despite greater augmentation of inspiratory muscle blood flow, this may reflect intrinsic metabolic abnormalities of the respiratory muscles in these patients such as a shift in fiber type composition that promotes a greater oxygen cost of respiration as well as higher levels of metabolite accumulation per amount of respiratory work (21). Until now, however, inspiratory muscle blood flow responses have not been measured in human HF.

The purpose of this study was to determine the cardiovascular consequences elicited by inspiratory muscle metaboreflex activation and characterize the inspiratory muscle blood flow response to an exhaustive inspiratory resistive breathing task (IRBT) in controls and patients with HFpEF. We hypothesized that patients with HFpEF, compared to controls, would have reduced tolerance to inspiratory resistive breathing, exaggerated cardiovascular responses with inspiratory muscle metaboreflex activation, and attenuated inspiratory muscle blood flow responses to inspiratory resistive breathing.

METHODS

Participants

Fifteen patients with HFpEF and 14 age- and sex-matched control participants were recruited for this study. HFpEF was defined by a clinical diagnosis including left ventricular ejection fraction of ≥50%, clinical symptoms (e.g., exertional dyspnea), and elevated left heart filling pressures at rest and/or with exercise in accordance with established guidelines (22). Exclusion criteria for the patients with HFpEF included history of arrhythmias, chronic kidney disease, overt pulmonary disease, body mass index >40 kg/m², smoking history of >15 pack yr, and limited ambulatory capabilities. Patients with HFpEF performed all testing while remaining on standard pharmacological therapy. The control group included a subset of participants with hypertension as well as overweight or obesity, but were free from cardiovascular (besides hypertension), pulmonary, and muscular diseases. All aspects of the study were approved by the Mayo Clinic Institutional Review Board and conformed to the Declaration of Helsinki, except for registration in a database. All participants were informed about the experimental procedures and potential risks involved before providing written and verbal informed consent.

Experimental Design

Participants completed two separate study visits. During the first visit, participants were familiarized with all experimental procedures and measurements. During the second visit, participants performed IRBTs at 2% and 60% of their maximal inspiratory mouth pressure (P_IMAX), which were separated by at least 20 min. The 2% IRBT was performed for 10 min and served as the control condition. The 60% IRBT was performed until task failure. Blood pressure, heart rate, and mouth pressure were continuously measured. At baseline and at the end of the IRBTs (i.e.,10 min for 2% IRBT and task failure for 60% IRBT), venous blood was sampled to measure plasma norepinephrine concentrations via high-performance liquid chromatography and a bolus of indocyanine green (ICG) dye was injected to determine the blood flow index (BFI) for the left sternocleidomastoid and left vastus lateralis muscles.

Inspiratory Muscle Strength

P_IMAX was assessed according to the American Thoracic Society and European Respiratory Society guidelines (23). P_IMAX was used to determine inspiratory muscle strength and was measured from the residual volume, as previously reported (24–26). Equations previously established by Neder et al. were used to predict P_IMAX for each participant according to age and sex (27). Inspiratory muscle weakness was defined as a P_IMAX value that was <70% of the predicted value (13, 23, 28).

Leg Tissue Composition

Dual-energy X-ray absorptiometry scans (Lunar iDXA Series X, GE Healthcare) were used to quantify leg tissue composition in patients with HFpEF and controls. Percent fat mass of the legs is reported. Before each scan, the scanner was calibrated against a standardized calibration phantom to control for possible baseline drift according to manufacturer recommendations.

Inspiratory Resistive Breathing Tasks

At baseline and during the IRBTs, participants were seated and breathed through a nonrebreathing three-way pneumatic switching valve connected to a pneumotachometer (Hans Rudolph, Kansas City, MO) and gas mass spectrometer (MGA 1100; Perkin Elmer, Wesley, MA). Participants rested for at least 20 min before baseline measurements were made. During the IRBTs, a resistor was added to the inspiratory conduit to elicit the desired pressure (2% or 60% P_IMAX) while expiration remained unimpeded. The resistor diameter necessary to elicit 2% and 60% P_IMAX were determined during the familiarization trial for each participant. Mouth pressure was continuously measured from a port in the mouthpiece via a calibrated differential pressure transducer (MP45; Validyne Engineering Corporation, Northridge, CA). The target inspiratory pressure (i.e., 2% and 60% PI_MAX) was displayed on a computer screen for the participant and breathing frequency (20 breaths per min) at a prolonged duty cycle (1.5 s inspiration and 1.5 s expiration) was maintained with a metronome. The time-integral of inspiratory pressure was calculated for each breath throughout the IRBTs. Sternocleidomastoid BFI was normalized to the time integral of mouth pressure during inspiration during the 60% IRBT (i.e., %ΔBFI per cmH₂O·s) and reported as a positive value. Vastus lateralis BFI was also normalized to the time integral of mouth pressure during inspiration during the 60% IRBT (i.e., %ΔBFI per cmH₂O·s) and reported as a negative value.

The 60% IRBT has previously been shown to elicit reductions in inspiratory muscle blood flow in canines (29) as well as inspiratory muscle metaboreflex activation (15–17, 30) and inspiratory muscle fatigue development in humans (14, 18). Therefore, the participants performed the 2% IRBT first and for 10 min, which served as the control condition. Following >20 min of rest, participants performed the 60% IRBT until task failure. Task failure was defined as when participants were unable to generate the target pressure for three consecutive breaths despite strong verbal encouragement from the study team. Once task failure was reached, participants were provided with continued strong verbal encouragement and performed the IRBT for an additional 90 s while final measurements were made (i.e., injection of ICG and venous blood sampling) (14). Participants were closely monitored by the research team to ensure proper timing, breathing technique, and effort. Supplementary CO₂ was added to the inspired gas to maintain end-tidal partial pressure of carbon dioxide (P_ETCO₂) near baseline levels thereby preventing hypocapnia-induced alterations in cardiovascular responses when compared with baseline.

Blood Pressure

Beat-by-beat heart rate, systolic blood pressure, diastolic blood pressure, and MAP were continuously measured via finger photoplethysmography (Nexfin, BMEYE, Amsterdam, the Netherlands). The Nexfin system was calibrated to brachial artery blood pressure. Real-time arterial pressure waveforms were sampled at 200 Hz and transmitted from the Nexfin system and converted from analog to digital signal using a PowerLab data acquisition module and digital oscilloscope (LabChart Pro, ADInstruments, v. 7, Colorado Springs, CO). Continuous beat-by-beat measurement of arterial pressure from the Nexfin system has been suggested to reliably estimate arterial-derived blood pressure (31). The final 60 s of baseline and final 30 s of each min during the IRBTs were averaged to calculate changes in MAP.

Near-Infrared Spectroscopy with ICG

Relative changes in leg and inspiratory muscle blood flow were measured by continuous-wave near-infrared spectroscopy (NIRS; NIRO-200NX, Hamamatsu Photonics KK, Hamamatsu, Japan) combined with bolus injection of ICG. This methodology has been described previously in detail (32) and has been validated against direct Fick principles for determining muscle blood flows (33). Each bolus of ICG (5 mg) was rapidly injected into the antecubital vein and immediately followed by 20 mL of 0.9% normal saline. Light-emitting diodes operating at wavelengths of 735, 810, and 850 nm and a photodiode detector were placed over the belly of the vastus lateralis on the left leg (i.e., a locomotor muscle) and the left sternocleidomastoid (i.e., an inspiratory muscle). The emitter-to-detector separation distance for each NIRS sensor was 4.0 cm. Due to the nearly complete absorption of infrared light in larger blood vessels, NIRS measurements are almost exclusively associated with the microcirculation (i.e., vessels <1 mm in diameter) (34). Changes in light attenuation at each of the NIRS wavelengths were used to isolate and quantify ICG concentration using the modified Beer–Lambert Law (35) and a matrix inversion operation, as previously described (36). BFI was calculated based on the detailed methodology previously described by Vogiatzis et al. in healthy adults and patients with chronic obstructive pulmonary disease (33, 37). Specifically, BFI was calculated as the rate of increase in ICG concentration following the bolus injection. To determine this rate, the peak change in ICG concentration was divided by the time interval between 10% and 90% of the overall ICG response. Figure 1 shows ICG concentration tracings for the vastus lateralis and sternocleidomastoid at baseline and task failure during the 60% IRBT for a single representative female patient with HFpEF.

Figure 1.

ICG concentration tracings for a representative female patient with HFpEF. ICG concentration tracings from the vastus lateralis (A) and sternocleidomastoid (B) muscles at baseline (dotted line) and task failure (solid line) during the 60% IRBT in a single female patient with HFpEF. For the vastus lateralis ICG tracing, there is a decrease in slope and peak ICG concentration from baseline to task failure during the 60% IRBT indicating a slower rate of ICG dye accumulation within the muscle and therefore, decreased vastus lateralis blood flow index. For the sternocleidomastoid ICG tracing, there is an increase in slope and peak ICG concentration from baseline to task failure during the 60% IRBT indicating a faster rate of ICG accumulation within the muscle and subsequently increased sternocleidomastoid blood flow index. HFpEF, heart failure with preserved ejection fraction; ICG, indocyanine green; IRBT, inspiratory resistive breathing task.

BFI measurements were not collected on some participants due to either methodological issues with the NIRS sensors (e.g., signal disruption via ambient light) or previous history of allergies to iodine-containing agents (n = 1 patient with HFpEF). Therefore, vastus lateralis BFI was measured in 11 patients with HFpEF and 11 controls and sternocleidomastoid BFI was measured in 12 patients with HFpEF and 14 controls. Vascular resistance index (VRI) was calculated for the sternocleidomastoid and vastus lateralis muscles as MAP divided by BFI. An important characteristic of NIRS measurements is the potentially confounding influence of overlying adipose tissue and skin pigmentation (34, 38). As adipose tissue thickness and skin pigmentation are assumed to remain unchanged within an individual during each trial, we report NIRS-derived values as a % change from baseline to minimize the intersubject variability in ICG concentration measured via NIRS due only to differences in adipose tissue thickness across individuals. It should be noted that absolute blood flows (reported in mL/100 mL/min) can be measured with NIRS and ICG coupled with arterial catheterization and photodensitometry (39). However, arterial catheterization was not performed in the present study and therefore, vastus lateralis and sternocleidomastoid BFIs were calculated. As absolute blood flows were not measured herein, an additional methodological consideration of the present study is that vascular resistance was calculated using BFI and reported as an index.

Statistical Analyses

In the tables, data are reported as means ± SD. In the figures, data are reported as individual values and means ± SE for clarity. Statistical analyses were performed using SigmaStat 2.0 (Jandel Scientific, San Rafael, CA). Normality and equal variance were assessed using the Shapiro–Wilk and Levene tests, respectively, and nonparametric tests were used when appropriate. Participant characteristics, as well as % changes in BFI and VRI were compared using unpaired t tests. For the 2% IRBT, MAP and heart rate were compared within (baseline, min 2, and min 10) and between groups (HFpEF vs. control) via two-way mixed factorial measures analysis of variance (ANOVA). For the 60% IRBT, MAP and heart rate were compared within (baseline, min 2, and at task failure) and between groups (HFpEF vs. control) via two-way mixed factorial measures ANOVA. A Tukey’s post hoc analysis was performed to determine where significance existed. Statistical significance was set at P < 0.05 for all analyses.

RESULTS

Participant Characteristics

Table 1 shows participant characteristics. Patients with HFpEF had higher weight and body mass index and lower P_IMAX and P_IMAX (% predicted) compared with controls. Inspiratory muscle weakness was present in 47% of the patients with HFpEF (n = 7).

Table 1.

Participant characteristics

	Control	HFpEF	P Value
n	14	15
Men/women	10/4	10/5	0.90
Age, yr	70 ± 8	69 ± 10	0.87
Height, cm	171 ± 8	171 ± 7	0.78
Weight, kg	83 ± 12	96 ± 10	<0.01
Body mass index, kg/m2	28.4 ± 3.6	33.2 ± 4.3	<0.01
PIMAX, cmH2O	108 ± 26	72 ± 25	<0.01
PIMAX, % predicted	121 ± 32	80 ± 26	<0.01
Leg fat mass, %	32 ± 11	40 ± 10	0.08
Hypertension, n, %	5 (36)	8 (53)	0.56
NYHA Class II/III		7/8
LV ejection fraction, %		60 ± 5
LA volume index, mL/m2		42 ± 11
Mitral E/e' ratio		13 ± 4
ACE inhibitor or ARBs, n, %	2 (14)	8 (53)	0.03
Diuretics, n, %	1 (7)	12 (80)	<0.01
β-blocker, n, %	0 (0)	9 (60)	<0.01
Ca2+ channel blocker, n, %	2 (14)	3 (20)	0.72
Aspirin, n, %	3 (21)	8 (53)	0.08

Data are presented as means ± SD. ACE, angiotensin-converting enzyme; ARB, angiotensin-receptor blockers; HFpEF, heart failure with preserved ejection fraction; LA, left atrial; LV, left ventricle; NYHA, New York Heart Association; P_IMAX, maximal inspiratory mouth pressure.

Ventilatory Responses during IRBTs

The ventilatory responses during the 2% and 60% IRBTs are shown in Table 2. For both 2% and 60% IRBTs, inspiratory mouth pressure (time-integral) and breathing frequency increased during the IRBTs compared with baseline for controls and patients with HFpEF. Compared with baseline, P_ETCO₂ was higher at min 2 during both the 2% and 60% IRBTs for patients with HFpEF and at min 2 during the 60% IRBT for controls (all, P < 0.04).

Table 2.

Ventilatory and norepinephrine data during IRBTs

	Baseline		Min 2		Min 10		ANOVA
	Control	HFpEF	Control	HFpEF	Control	HFpEF	Group	Time	Interaction
2% IRBT
Pm,insp TI, cmH2O·s	0 ± 0	0 ± 0	−3 ± 1†	−2 ± 1*†	−3 ± 2†	−2 ± 1*†	0.04	<0.01	0.03
Breathing frequency, breaths/min	15 ± 3	19 ± 4*	20 ± 1†	21 ± 2†	20 ± 1†	21 ± 2†	0.01	<0.01	<0.01
PETCO2, mmHg	34 ± 3	34 ± 5	36 ± 3	37 ± 7†	37 ± 3	36 ± 4	0.86	<0.01	0.64
MAP, mmHg	91 ± 6	88 ± 6	93 ± 9	90 ± 10	93 ± 7	89 ± 7	0.96	0.07	0.99
Heart rate, beats/min	64 ± 11	68 ± 9	68 ± 11†	70 ± 9†	66 ± 10	69 ± 9	0.56	<0.01	0.60
Norepinephrine concentration, pg/mL	598 ± 183	587 ± 280			588 ± 176	571 ± 279	0.88	0.33	0.82

	Baseline		Min 2		Task failure
	Control	HFpEF	Control	HFpEF	Control	HFpEF
60% IRBT
Pm,insp TI, cmH2O·s	0 ± 0	0 ± 0	−57 ± 19†	−33 ± 13*†	−52 ± 16†	−31 ± 12*†	<0.01	<0.01	<0.01
Breathing frequency, breaths/min	15 ± 3	18 ± 5	21 ± 2†	21 ± 3†	21 ± 4†	21 ± 2†	0.28	<0.01	0.06
PETCO2, mmHg	34 ± 4	34 ± 5	38 ± 3†	37 ± 6†	36 ± 3	36 ± 6	0.85	<0.01	0.83
MAP, mmHg	92 ± 8	88 ± 7	101 ± 14†	100 ± 11†	102 ± 12†	104 ± 9†	0.75	<0.01	0.07
Heart rate, beats/min	62 ± 10	65 ± 9	76 ± 12†	74 ± 11†	78 ± 11†	76 ± 11†	0.87	<0.01	0.08
Norepinephrine concentration, pg/mL	593 ± 171	551 ± 241			673 ± 177†	658 ± 263†	0.76	<0.01	0.42

Data are presented as means ± SD. HFpEF, heart failure with preserved ejection fraction; IRBT, inspiratory resistive breathing task; P_m,insp, inspiratory mouth pressure; P_ETCO₂, end-tidal partial pressure of carbon dioxide; TI, time-integral;. *Significantly different than control. †Significantly different than baseline.

Cardiovascular Responses during 2% IRBT

MAP was not different during the 2% IRBT compared with baseline for either controls or patients with HFpEF (time main effect: P = 0.07, time × group: p = 0.96). The heart rate response during the 2% IRBT was different across time (time main effect: P < 0.01, time × group: P > 0.60), with significant increases in heart rate present only at min 2 for both groups (both, P < 0.04). There were no differences in the % change in vastus lateralis BFI from baseline to min 10 of the 2% IRBT between controls and patients with HFpEF (−0.9 ± 11.0 vs. 0.8 ± 14.2%, P = 0.77). Venous norepinephrine concentration was not different between baseline and min 10 for controls or patients with HFpEF (Table 2).

Cardiovascular Responses during 60% IRBT

Figure 2 shows the time to task failure during the 60% IRBT trial for both groups. The time to task failure was less for HFpEF compared with controls (P = 0.02). Figure 3 shows the MAP and heart rate responses during the 60% IRBT. MAP and HR during the 60% IRBT were different across time and between groups (time main effect: P < 0.01, group main effect: P < 0.04, time × group: P ≥ 0.05). Compared with baseline, controls and HFpEF patients had increases in MAP and heart rate at min 2 and task failure (all, P < 0.01). The increase in MAP at task failure was greater for HFpEF compared with controls during the 60% IRBT (P < 0.01). The increases in heart rate at min 2 and at task failure during the 60% IRBT were greater in controls compared with patients with HFpEF (both, P < 0.02). There was no difference in the heart rate response between the patients with HFpEF prescribed β-blockers (n = 9) compared with those who were not (n = 6) (P = 0.18). However, there was a relationship between the increase in heart rate during the 60% IRBT and the inspiratory load set for the 60% IRBT (r = 0.60, P = 0.02). Figure 4 shows the % change in vastus lateralis BFI and VRI from baseline to task failure during the 60% IRBT in controls and patients with HFpEF. Patients with HFpEF, compared with controls, had greater increases in vastus lateralis VRI and decreases in vastus lateralis BFI from baseline to task failure during the 60% IRBT (both, P < 0.01). These differences between patients with HFpEF and controls remained when vastus lateralis BFI was normalized to inspiratory mouth pressure (−1.31 ± 0.87 vs. −0.33 ± 0.24%/cmH₂O·s, P < 0.01). From baseline to the task failure, venous catecholamines increased in controls and patients with HFpEF, while no differences were present between groups (Table 2).

Figure 2.

Time to task failure during the 60% IRBT. Time to task failure during the 60% IRBT in controls (n = 14, open circles) and patients with HFpEF (n = 15, closed circles). Vertical bars indicate the means ± SE for each group. Time to task failure was shorter in patients with HFpEF compared with controls (P = 0.02). *Significantly different between groups. Data were analyzed using unpaired t tests. HFpEF, heart failure with preserved ejection fraction; IRBT, inspiratory resistive breathing task.

Figure 3.

MAP and heart rate responses to the 60% IRBT. Change in mean arterial pressure (MAP) (A) and heart rate (HR) (C) from baseline during the 60% IRBT performed until task failure in controls (n = 14, open circles) and patients with HFpEF (n = 15, closed circles). Group data are presented as means ± SE. Individual MAP and heart rate responses (absolute change from baseline to task failure) are shown in B and D, respectively. Patients with HFpEF had a greater increase in MAP from baseline to task failure compared with controls (P < 0.01). Compared with patients with HFpEF, controls had a greater increase in heart rate from baseline at min 2 and at task failure during the 60% IRBT (both, P < 0.02). *Significantly different between groups. Data were analyzed using two-way mixed factorial ANOVA with Tukey’s post hoc tests. HFpEF, heart failure with preserved ejection fraction; IRBT, inspiratory resistive breathing task.

Figure 4.

Vastus lateralis BFI and VRI during the 60% IRBT. Change in vastus lateralis blood flow index (BFI) (A) and vascular resistance index (VRI) (B) from baseline to task failure during the 60% IRBT in controls (n = 11, open circles) and patients with HFpEF (n = 11, closed circles). Vertical bars indicate the means ± SE for each group. The % reduction in vastus lateralis BFI was greater in patients with HFpEF compared with controls (P < 0.01). The % increase in vastus lateralis VRI was greater in patients with HFpEF compared with controls (P < 0.01). *Significantly different between groups. Data were analyzed using unpaired t tests. HFpEF, heart failure with preserved ejection fraction; IRBT, inspiratory resistive breathing task.

Inspiratory Muscle Blood Flow Responses

The % increase in sternocleidomastoid BFI during the 60% IRBT was not different between patients with HFpEF and controls (257 ± 202 vs. 192 ± 82%, P = 0.98). There were positive relationships between the % change in sternocleidomastoid BFI and P_IMAX as well as % predicted P_IMAX in patients with HFpEF (r = 0.63–0.64, P = 0.03 for both) and controls (r = 0.54–0.60, P = 0.02–0.045), whereas no relationships were present in the % change in sternocleidomastoid BFI and age or body mass index (both, P > 0.10). Considering the relationships with P_IMAX, sternocleidomastoid BFI responses were normalized to inspiratory pressure generation (time-integral) for each participant (Fig. 5). The % increase in sternocleidomastoid BFI normalized to inspiratory pressure generation was greater in patients with HFpEF compared with controls (P = 0.03). No relationships were present between the % increase in sternocleidomastoid BFI normalized to inspiratory pressure generation and age, body mass index, or P_IMAX (all, P > 0.20). The % decrease in sternocleidomastoid VRI during the 60% IRBT was not different between patients with HFpEF and controls (−56 ± 23 vs. −59 ± 11, P = 0.90).

Figure 5.

Sternocleidomastoid BFI responses normalized to inspiratory pressure generation. Change in sternocleidomastoid blood flow index (BFI) from baseline to task failure during the 60% IRBT normalized to the time-integral of inspiratory pressure generation in controls (n = 14, open circles) and patients with HFpEF (n = 12, closed circles). Vertical bars indicate the means ± SE for each group. The % increase in sternocleidomastoid BFI per absolute inspiratory pressure generation was greater in patients with HFpEF compared with controls (P = 0.03). *Significantly different between groups. Data were analyzed using unpaired t tests. HFpEF, heart failure with preserved ejection fraction; IRBT, inspiratory resistive breathing task.

Impact of BMI on Responses in HFpEF

The primary outcomes were compared in the patients with HFpEF categorized by BMI [i.e., higher BMI (n = 8): 36 ± 3 vs. lower BMI (n = 7): 30 ± 1 kg/m², P < 0.01)]. There were no differences between the higher and lower BMI groups in time to task failure (357 ± 181 vs. 318 ± 226 s), % change in vastus lateralis BFI (−35 ± 10 vs. −34 ± 14%), or % change in sternocleidomastoid BFI normalized to inspiratory pressure generation (7.6 ± 6.3 vs. 8.3 ± 4.0% per cmH₂O·s) during the 60% IRBT (all, P > 0.72).

DISCUSSION

Major Findings

The major findings of this study were threefold. First, time to task failure during the 60% IRBT was shorter for patients with HFpEF compared with controls (Fig. 2). Second, inspiratory muscle metaboreflex activation (via the 60% IRBT) elicited greater increases in MAP (Fig. 3A) and vastus lateralis VRI (Fig. 4B) as well as decreases in vastus lateralis BFI (Fig. 4A) in patients with HFpEF compared with controls. Finally, patients with HFpEF had an exaggerated sternocleidomastoid BFI response compared with controls when normalized to inspiratory pressure generation (Fig. 5). Taken together, these findings demonstrate that the cardiovascular responses to inspiratory muscle metaboreflex are exaggerated in HFpEF despite augmented inspiratory muscle blood flow responses.

Time to Task Failure Is Reduced in HFpEF

During the 60% IRBT, time to task failure was ∼45% shorter for patients with HFpEF compared with controls. This reduction in inspiratory muscle endurance in HFpEF was observed despite an exaggerated sternocleidomastoid BFI response (and presumably greater oxygen availability) for a given inspiratory pressure generation (discussed in Inspiratory Muscle Blood Flow Is Augmented in HFpEF). These data are important as increased inspiratory muscle fatigue resistance is associated with greater exercise capacity (40). Alterations in respiratory muscle morphology (i.e., fiber-type shift and muscle atrophy) and impaired mitochondrial function have been previously observed in rats fed a western diet (to induce HFpEF-like symptomology) (21). Furthermore, these morphological and mitochondrial changes were concomitant with greater inspiratory muscle fatiguability. Combined with these previous findings, our data support the notion that HFpEF is associated with inspiratory muscle abnormalities that result in impaired inspiratory muscle fatigue resistance. Although inspiratory muscle weakness (and likely fatiguability) appears to be independent of systemic muscle weakness (41, 42), it is important to note that exercise training and physical activity status may improve inspiratory muscle fatigue resistance and endurance in patients with HFpEF, as reported previously in healthy adults (13, 26, 28, 43).

Inspiratory Muscle Metaboreflex Is Exaggerated in HFpEF

Previous studies have demonstrated that fatiguing inspiratory muscle work in otherwise resting healthy participants results in increases in muscle sympathetic nerve activity, MAP, and leg vascular resistance as well as decreases in leg muscle blood flow (14–18). The changes in MAP, leg VRI, and leg BFI in the control participants reported herein are consistent with these previous studies.

The primary novel finding of the present study was that patients with HFpEF had greater increases in MAP (Fig. 3A) and vastus lateralis VRI (Fig. 4B) during the 60% IRBT compared with controls. Consequently, patients with HFpEF had greater decreases in vastus lateralis BFI (Fig. 4A). These findings indicate that the cardiovascular responses to inspiratory muscle metaboreflex activation are exaggerated in patients with HFpEF. In addition to other pulmonary system and respiratory muscle abnormalities [e.g., attenuated increases in lung diffusion capacity with exercise (44)], an exaggerated inspiratory muscle metaboreflex may have significant implications for eliciting abnormal neural and cardiovascular responses during exercise (45–47) thereby contributing to exertional symptomology and exercise intolerance in these patients. In the present study, venous norepinephrine concentration increased during the 60% IRBT in both controls and patients with HFpEF, whereas no differences were present between groups. Although surprising, these data are consistent with a previous study reporting similar norepinephrine levels at peak exercise between controls and patients with HFpEF (48). Future studies with direct measurement of muscle sympathetic nerve activity during the 60% IRBT may provide additional insight as to the underlying mechanisms responsible for exaggerated inspiratory muscle metaboreflex-mediated vasoconstriction in HFpEF.

It should be noted that other sources may have contributed to the neural and cardiovascular responses observed in the present study. The 2% IRBT did not result in increases in MAP (with minimal changes in the % change in vastus lateralis BFI) in the present study, suggesting the mechanoreflex was not significantly contributing to the cardiovascular responses observed during the 60% IRBT. Regarding central command, previous studies have performed the IRBTs at near maximal inspiratory pressures without inducing inspiratory muscle fatigue and reported no changes in muscle sympathetic nerve activity or MAP (14, 18). Furthermore, a recent study has provided evidence that central command is not altered during exercise in patients with HFpEF compared with controls (49). The arterial baroreceptors have been shown to interact with the skeletal muscle metaboreflex for the control of neural and cardiovascular responses (50). Spontaneous baroreflex modulation of muscle sympathetic nerve activity was recently found to be similar in HFpEF compared with controls (51); however, it is unclear whether the HFpEF syndrome impacts baroreflex sensitivity during inspiratory muscle metaboreflex activation.

Inspiratory Muscle Blood Flow Is Augmented in HFpEF

Inspiratory muscle blood flow has been shown to increase in proportion to absolute work in healthy individuals (39), which is important as P_IMAX was markedly reduced in HFpEF (Table 1). Furthermore, we found a positive relationship between P_IMAX and the % change in sternocleidomastoid BFI in the patients with HFpEF and controls. To elucidate whether the patients with HFpEF had an abnormal sternocleidomastoid BFI during inspiratory loading, we compared the sternocleidomastoid BFI responses between groups after normalizing for absolute inspiratory pressure during the 60% IRBT. This revealed exaggerated sternocleidomastoid BFI responses in HFpEF (Fig. 5) and suggests a greater stimulus for blood flow at a given level of inspiratory pressure generation. Notably, there were no differences in sternocleidomastoid VRI between patients with HFpEF and controls. This is in contrast to a previous study reporting exaggerated vasoconstriction of the contracting skeletal muscles under ischemic conditions during exercise in a pacing-induced canine model of HF (52). An important consideration of these observations is the likely contributions from multiple inspiratory pressure-generating muscles during the 60% IRBT. Although muscle activation levels were not assessed in the present study, the heterogeneity of inspiratory muscle recruitment strategies among patients with HFpEF may have contributed to the variability in BFI responses (Fig. 5).

Methodological Considerations

There are three primary methodological considerations for the present study. First, patients with HFpEF had a higher body mass index compared with controls (Table 1). The impact of obesity on metaboreflex control of the circulation is unclear as obese adults have been found to have reduced (53), similar (54), and increased (55) autonomic and cardiovascular responses to metaboreflex activation compared with healthy adults. Although we found no significant associations with body mass index, future studies are necessary to investigate whether obesity impacts inspiratory muscle metaboreflex-mediated cardiovascular responses. Second, inspiratory muscle fatigue was not directly assessed in the present study; however, this protocol has previously been shown to elicit inspiratory muscle fatigue in healthy adults (14, 17, 18, 30, 56). Future studies are necessary to determine the impact of HFpEF on inspiratory muscle fatigue development. Third, lung volumes impact P_IMAX. In the present study, P_IMAX was measured from residual volume; however, lung volumes were not measured during the IRBTs.

Conclusions

We show that patients with HFpEF exhibit an exaggerated inspiratory muscle metaboreflex associated with exaggerated peripheral vasoconstriction and augmented inspiratory muscle blood flow response to high inspiratory loads. Our data suggest that intrinsic respiratory metabolic dysfunction may exist in HFpEF and contribute to well-described cardiovascular hemodynamic perturbations that develop during exercise. Future studies are necessary to identify effective strategies that improve the overall function of the respiratory muscles to minimize neural and cardiovascular consequences of inspiratory metaboreflex activation in these patients.

GRANTS

This work was supported by the National Institutes of Health (T32HL007111 to S.M.H. and J.R.S., K12 HD065987 to J.R.S, HL126638 to T.P.O., HL128526 to B.A.B.) and American Heart Association (18POST3990251 to J.R.S). This publication was also made possible through the support of the Mary Kathryn and Michael B. Panitch Career Development Award in Hypertension Research Honoring Gary Schwartz, M.D. (to J.R.S).

DISCLOSURES

No conflicts of interest, financial or otherwise, are declared by the author(s).

AUTHOR CONTRIBUTIONS

S.M.H., E.J.B., T.G.B., J.D.M., N.V., B.A.B., T.P.O., and J.R.S. conceived and designed research; S.M.H., E.J.B., T.G.B., J.D.M., N.V., B.A.B., T.P.O., and J.R.S. performed experiments; S.M.H., E.J.B., T.G.B., J.D.M., N.V., B.A.B., T.P.O., and J.R.S. analyzed data; S.M.H., E.J.B., T.G.B., J.D.M., N.V., B.A.B., T.P.O., and J.R.S. interpreted results of experiments; S.M.H., E.J.B., T.G.B., J.D.M., N.V., B.A.B., T.P.O., and J.R.S. prepared figures; S.M.H., E.J.B., T.G.B., J.D.M., N.V., B.A.B., T.P.O., and J.R.S. drafted manuscript; S.M.H., E.J.B., T.G.B., J.D.M., N.V., B.A.B., T.P.O., and J.R.S. edited and revised manuscript; S.M.H., E.J.B., T.G.B., J.D.M., N.V., B.A.B., T.P.O., and J.R.S. approved final version of manuscript.