Impaired central hemodynamics in chronic obstructive pulmonary disease during submaximal exercise

Abstract

It is unknown whether central hemodynamics are impaired during exercise in chronic obstructive pulmonary disease (COPD) patients. We hypothesized that, at a similar absolute V̇o₂ during exercise, COPD patients would have a lower stroke volume and cardiac output compared with healthy controls. Furthermore, we hypothesized that greater static hyperinflation [ratio of inspiratory capacity to total lung capacity (IC/TLC)] and expiratory intrathoracic pressure would be significantly related to the lower cardiac output and stroke volume responses in COPD patients. Clinically stable COPD (n = 13; FEV₁/FVC: 52 ± 13%) and controls (n = 10) performed constant workload submaximal exercise at an absolute V̇o₂ of ~1.3 L/min. During exercise, inspiratory capacity maneuvers were performed to determine operating lung volumes and cardiac output (via open-circuit acetylene rebreathe technique) and esophageal pressure were measured. At similar absolute V̇o₂ during exercise (P = 0.81), COPD had lower cardiac output than controls (COPD: 11.0 ± 1.6 vs. control: 12.2 ± 1.2 L/min, P = 0.03) due to a lower stroke volume (COPD: 107 ± 13 vs. control: 119 ± 19 mL, P = 0.04). The heart rate response during exercise was not different between groups (P = 0.66). FEV₁ (%predicted) and IC/TLC were positively related to stroke volume (r = 0.68, P = 0.01 and r = 0.77, P < 0.01). Last, esophageal pressure-time integral during inspiration was positively related to cardiac output (r = 0.56, P = 0.047). These data demonstrate that COPD patients have attenuated cardiac output and stroke volume responses during exercise compared with control. Furthermore, these data suggest that the COPD patients with the most severe hyperinflation and more negative inspiratory intrathoracic pressures have the most impaired central hemodynamic responses.

NEW & NOTEWORTHY Chronic obstructive pulmonary disease leads to cardiac structural changes and pulmonary derangements that impact the integrative response to exercise. However, it is unknown whether these pathophysiological alterations influence the cardiac response during exercise. Herein, we demonstrate that COPD patients exhibit impaired central hemodynamics during exercise that are worsened with greater hyperinflation.

INTRODUCTION

Patients with chronic obstructive pulmonary disease (COPD) exhibit reduced exercise capacity (31). The putative underlying mechanisms responsible are multifactorial and include ventilatory, central hemodynamic, and peripheral limitations (1, 10, 25, 30). One of the strongest predictors of exercise capacity in COPD is peak oxygen pulse, a surrogate for stroke volume (28). Furthermore, previous studies have suggested that COPD patients demonstrate a reduced or similar central hemodynamic response during submaximal exercise (4, 16, 31, 35, 43, 46). However, these studies have either not included a healthy control group or compared COPD with controls at relative intensities (i.e., 40% peak workload). Because of the compromised exercise capacity in COPD, the lower cardiac output and stroke volume in COPD compared with controls reported when relative intensity is matched (16, 35) is consistent with the lower absolute oxygen uptake (V̇o₂) and workload. Bogaard et al. (4) compared patients with COPD and controls at matched absolute workloads during exercise and found an attenuated cardiac output (measured via impedance cardiography) and stroke volume in COPD patients; however, the central hemodynamic responses were variable, and the underlying mechanisms contributing to these responses were not investigated.

One of the primary mechanisms suggested to be responsible for the impaired central hemodynamic response to exercise in COPD patients is airway obstruction resulting in hyperinflation and consequently dysfunctional negative intrathoracic pressure swings as well as an augmented expiratory load (23, 24, 29, 42, 43, 45). For example, COPD patients with greater static hyperinflation [i.e., lower ratio of inspiratory capacity to total lung capacity (IC/TLC)] often exhibit a lower cardiac output and stroke volume at peak exercise compared with COPD patients with less static hyperinflation (43), potentially arising from the aberrant negative intrathoracic pressures and/or elevated expiratory load (29, 42). However, it is important to note that these patients exhibiting greater hyperinflation also have lower exercise capacity (43, 45), which may account for the lower cardiac output. Thus, the relationships between static hyperinflation and intrathoracic pressures with cardiac output and stroke volume during exercise remain unclear.

Therefore, the purpose of this study was to 1) compare central hemodynamics between COPD patients and healthy controls and 2) examine the impact of static hyperinflation and elevated intrathoracic pressure on cardiac output and stoke volume in COPD patients. We hypothesized that COPD patients would have lower stroke volume and cardiac output compared with controls at matched V̇o₂ during exercise. Furthermore, we hypothesized that lower IC/TLC and increased expiratory intrathoracic pressure would result in lowered cardiac output and stroke volume in COPD patients relative to age-matched healthy controls.

METHODS

Participants.

Thirteen clinically stable COPD patients with moderate to severe COPD [forced expiratory volume in 1 s/forced vital capacity (FEV₁/FVC): 52 ± 13% predicted] and 10 healthy control participants matched for age and sex volunteered for this study and provided written informed consent. Exclusion criteria for this study included supplemental oxygen dependence as well as cardiovascular, metabolic, and/or musculoskeletal diseases. All aspects of this study were approved by the Mayo Clinic Institutional Review Board and conformed to the Declaration of Helsinki.

Experimental design.

For this cross-sectional study, participants performed all protocols and measurements during two study visits. On the first study visit, participants were first familiarized with all experimental measurements and protocols, performed resting pulmonary function testing, and then completed an incremental exercise test to volitional fatigue to determine peak oxygen uptake (V̇o_2peak). On the second study visit, participants performed constant workload submaximal exercise at an absolute V̇o₂ of ~1.3 L/min. During constant workload exercise, inspiratory capacity (IC) maneuvers were performed to determine operating lung volumes, and cardiac output and esophageal pressure were measured.

V̇o_2peak.

Participants performed an incremental cycling test to volitional fatigue to determine V̇o_2peak, using an electronically braked cycle ergometer (Lode Corival, Groningen, The Netherlands). The incremental test consisted of increasing workloads of 10- to 20- and 20- to 50-W increments for COPD and controls, respectively based on the participants' perceived peak exercise workload. During the incremental test, participants maintained a pedal frequency of 60 rpm and remained seated. Ventilatory and gas exchange variables were collected during the incremental cycling test (MedGraphics CPX/D, St. Paul, MN) and averaged over 30 s. Data acquired included V̇o₂, V̇co₂, respiratory exchange ratio (RER), ventilation (V_E), breathing frequency (f_B), and tidal volume (V_T). In addition, the ventilatory equivalents for V̇o₂ and V̇co₂ (V_E/V̇o₂ and V_E/V̇co₂, respectively) were calculated at peak exercise. Percent (%)predicted V̇o_2peak was calculated from Kaminsky et al. (20). Heart rate was measured continuously via electrocardiogram and recorded at the end of each stage.

Submaximal exercise.

On the second visit, all participants completed a submaximal constant-load exercise test. First, resting data were collected for 3 min, and then participants exercised at a workload resulting in an absolute V̇o₂ of ~1.3 L/min. This absolute V̇o₂ was chosen as being similar to moderate physical activity in COPD patients (48) and was submaximal for all participants. The initial resistance on the cycle ergometer was determined using data from the incremental test, and then the workload was titrated during minutes 3 to 5 to achieve a V̇o₂ of 1.3 L/min. Minutes 5 to 8 were used to achieve steady state (as evident by the heart rate, RER, and V̇o₂ responses), at which time an IC maneuver was performed, and cardiac output and blood pressure were measured. Ventilatory and gas exchange variables were collected at rest and during exercise, using the same methodology used during the incremental test, and the 30-s average corresponding to the IC maneuver and cardiac output is reported.

Cardiac output.

Cardiac output was measured using an open-circuit acetylene technique that has previously been validated against direct Fick measurements (17). Previous studies have validated rebreathe techniques (using carbon dioxide and nitrous oxide) against direct Fick, thermodilution, and cardiac magnetic resonance imaging in patients with cardiopulmonary diseases (e.g., COPD, pulmonary hypertension, etc.) (13, 15, 26, 37). Importantly, the Bunsen solubility coefficient of acetylene in blood is approximately two times greater than that of carbon dioxide and nitrous oxide (12). Briefly, participants breathed into a nonrebreathing three-way pneumatic switching valve connected to a pneumotachometer (Hans Rudolph, Kansas City, MO) and gas mass spectrometry (MGA-1100; PerkinElmer Wellesley, MA) integrated with custom analysis software for measurement of cardiac output (17, 32, 44a). The pneumatic switching valve allowed for rapid switching from room air to the gas mixture (0.65% C₂H₂, 21%O₂, 9% He, and balanced N₂). Stroke volume was calculated by dividing cardiac output by heart rate. Systemic vascular resistance (SVR) was calculated as the quotient of mean arterial pressure and cardiac output.

Pulmonary function tests.

Pulmonary function tests were performed according to established American Thoracic Society and European Respiratory Society guidelines (27). Total lung capacity (TLC), residual volume (RV), FVC, FEV₁, FEV₁/FVC ratio, and lung diffusion capacity for carbon monoxide (DL_CO) were measured and reported as %predicted values (21, 22, 44).

Respiratory mechanics.

Esophageal pressure was measured using a latex balloon-tip catheter (CooperSurgical, Trumbull, CT) inserted via the nares ~45 cm following local anesthetic. Correct placement of the balloon (i.e., lower one-third of the esophagus) was confirmed using the “occlusion” method (2) following inflation of the balloon with 1 mL of air. The catheter was connected to a differential pressure transducer (MP45; Validyne Engineering, Northridge, CA) and calibrated using a water manometer before each test. Prior to exercise, participants performed multiple IC maneuvers until the maneuvers were valid and consistent. The standard deviation of end-expiratory lung volume (EELV) at rest for the controls and COPD patients were 0.3 and 0.2 L, respectively. At rest and during submaximal exercise, participants performed IC maneuvers from EELV. A “typical” breath was determined if it had similar volume, flow, and pressure characteristics to the previous breaths before the IC maneuver (6, 39–41). A computer program was used to correct for physiological drift during exercise (i.e., when there were unequal inspiratory and expiration volumes) using the IC ratio method (11, 38). TLC was assumed not to significantly change during exercise (18, 19). End-inspiratory lung volume (EILV) was calculated as the sum of V_T and EELV and expressed as %TLC. Peak esophageal pressure during inspiration and expiration (P_es,ins and P_es,exp, respectively) and esophageal pressure-time integral during inspiration and expiration (P_es,insTI and P_es,expTI) for each breath were calculated from 8–10 breaths during exercise.

Statistical analyses.

Values are reported as means ± SD. 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 peak exercise, submaximal exercise, and resting data were compared between groups (COPD vs. controls) using unpaired t-tests. Relationships were determined via linear regression. Statistical significance was set at P < 0.05.

RESULTS

Participant characteristics.

COPD patients had a lower %predicted FEV₁, FVC, DL_CO, and FEV₁/FVC than controls but greater body mass index, %predicted RV, and RV/TLC (all, P ≤ 0.03; Table 1). Peak exercise workload, V̇o₂, V̇co₂, RER, heart rate, V_E, and V_T were lower in COPD than in controls (Table 2). The %predicted V̇o_2peak for the COPD patients was lower than for controls (COPD: 53 ± 15 vs. controls: 83 ± 17%, P < 0.01).

Table 1.

Participant characteristics

	Control	COPD
n	10	13
M/W	8/2	11/2
Age, yr	59 ± 11	59 ± 9
Height, cm	174 ± 8	172 ± 5
Weight, kg	84 ± 10	99 ± 25
BMI, kg/m2	28 ± 3	31 ± 4*
FEV1, %pred	96 ± 13	51 ± 10*
FVC, %pred	97 ± 15	79 ± 13*
FEV1/FVC, %	79 ± 5	52 ± 13*
TLC, %pred	105 ± 16	115 ± 23
RV, %pred	108 ± 17	177 ± 58*
RV/TLC, %	35 ± 6	52 ± 9*
DLCO, %pred	96 ± 14	71 ± 14*

Values are means ± SD. COPD, chronic obstructive pulmonary disease; M/W, men/women; BMI, body mass index; FEV₁, forced expiratory volume in 1 s; FVC, forced vital capacity; TLC, total lung capacity; RV, residual volume; DL_CO, difusion capacity for carbon monoxide; pred, predicted.

^*Significantly different from control.

Table 2.

Peak and submaximal exercise data

	Control	COPD	P Value
Peak exercise
Workload, W	173 ± 50	96 ± 21*	<0.01
V̇o2, mL·kg−1·min−1	27.1 ± 5.5	16.9 ± 4.1*	<0.01
V̇o2, L/min	2.3 ± 0.5	1.6 ± 0.3*	<0.01
V̇co2, L/min	2.6 ± 0.7	1.7 ± 0.5*	<0.01
RER	1.15 ± 0.08	1.04 ± 0.09*	0.01
Heart rate, beats/min	157 ± 17	125 ± 24*	<0.01
VE, L/min	93 ± 24	54 ± 20*	<0.01
fB, breaths/min	37 ± 4	33 ± 8	0.12
VT, L	2.5 ± 0.8	1.6 ± 0.3*	<0.01
VE/V̇co2	35 ± 4	33 ± 6	0.62
VE/V̇o2	40 ± 5	35 ± 7	0.06
Submaximal exercise
Workload, W	62 ± 14	45 ± 10*	<0.01
V̇o2, mL·kg−1·min−1	15.4 ± 2.3	13.4 ± 2.8	0.08
V̇o2, L/min	1.3 ± 0.2	1.3 ± 0.1	0.81
VCO2, L/min	1.2 ± 0.2	1.2 ± 0.1	0.63
RER	0.90 ± 0.04	0.93 ± 0.07	0.25
Heart rate, beats/min	102 ± 8	103 ± 10	0.66
VE, L/min	33 ± 4	37 ± 6	0.11
fB, breaths/min	23 ± 4	27 ± 5*	0.04
VT, L	1.5 ± 0.3	1.4 ± 0.2	0.37
IC, L	3.7 ± 0.9	2.5 ± 0.5*	<0.01
IC/TLC, %	55 ± 7	36 ± 10*	<0.01
EILV/TLC, %	67 ± 8	85 ± 7*	<0.01
Peak Pes,ins, cmH2O	−10 ± 3	−19 ± 6*	<0.01
Peak Pes,exp, cmH2O	8 ± 7	19 ± 7*	<0.01
Pes,inspTI, cmH2O·s	−7 ± 2	−11 ± 2*	<0.01
Pes,expTI, cmH2O·s	4 ± 2	14 ± 7*	<0.01
SaO2, %	97 ± 1	96 ± 1*	<0.01
RPE	11 ± 2	14 ± 3*	0.03
Dyspnea	2 ± 1	4 ± 2*	<0.01
MAP, mmHg	101 ± 8	107 ± 15	0.40
SVR, mmHg·L−1·min−1	8.3 ± 1.3	9.9 ± 1.9*	0.04

Mean ± SD. Data from 23 participants (CTL = 10; COPD = 13). Submaximal exercise data were averaged during min 5–8 of the submaximal exercise test. COPD, chronic obstructive pulmonary disease; V̇o₂, oxygen uptake; V̇co₂, carbon dioxide production; RER, respiratory exchange ratio; V_E, ventilation; f_B, breathing frequency; V_T, tidal volume; IC, inspiratory capacity; TLC, total lung capacity; EILV, end inspiratory lung volume; P_es,ins, inspiratory esophageal pressure; P_es,exp, expiratory esophageal pressure; TI, time integral; RPE, rating of percieved exertion; MAP, mean arterial pressure; SVR, systemic vascular resistance.

^*Significantly different from control.

Resting cardiopulmonary data.

Resting V̇o₂, V̇co₂, and V_E were greater in the COPD patients compared with the controls (all, P < 0.03), while relative V̇o₂, RER, heart rate, V_T, f_B, and MAP were not different between groups (all P < 0.18). Resting cardiac output (COPD: 5.3 ± 1.1 vs. controls: 5.6 ± 0.6 L/min, P = 0.37), stroke volume (COPD: 77 ± 18 vs. controls: 87 ± 16 mL, P = 0.23), and systemic vascular resistance (COPD: 18.2 ± 4.8 vs. controls: 16.7 ± 3.0 mmHg·L⁻¹·min⁻¹, P = 0.41) were not different between groups. Moreover, resting cardiac output and stroke volume indexed to body surface area were not different for COPD patients compared with controls (COPD: 2.6 ± 0.6 vs. controls: 2.8 ± 0.3 L·min⁻¹·m⁻² and COPD: 38 ± 11 vs. controls: 43 ± 9 mL/m², respectively; both P > 0.20). Last, COPD patients, compared with controls, had more negative peak P_es,insp (COPD: −11 ± 4 vs. controls: −6 ± 2 cmH₂O) and P_es,inspTI (COPD: −9 ± 4 vs. controls: −5 ± 1 cmH₂O·s) as well as more positive peak P_es,exp (COPD: 5 ± 4 vs. controls: 1 ± 3 cmH₂O) and P_es,expTI (COPD: 7 ± 7 vs. controls: 1 ± 1 cmH₂O·s) (all P < 0.02).

Submaximal exercise responses.

At similar absolute V̇o₂, COPD patients had lower workload, IC, and IC/TLC compared with controls (Table 2). Furthermore, COPD patients had greater SVR, EILV/TLC, RPE, and dyspnea than controls. Last, compared with controls, COPD patients had more negative peak P_es,insp and P_es,inspTI as well as more positive peak P_es,exp and P_es,expTI. Cardiac output was lower in COPD than in controls (COPD: 11.0 ± 1.6 vs. controls: 12.2 ± 1.2 L/min, P = 0.03) due to a lower stroke volume (COPD: 107 ± 13 vs. controls: 119 ± 19 mL, P = 0.04), as heart rate was not different (P = 0.66) (Fig. 1). Furthermore, cardiac output and stroke volume indexed to body surface area were significantly lower for COPD patients than for controls (COPD: 5.3 ± 0.6 vs. controls: 6.1 ± 0.6 L·min⁻¹·m⁻² and COPD: 51 ± 5 vs. controls: 59 ± 10 mL/m², respectively; both P ≤ 0.01). In COPD patients, FEV₁ (%predicted) was positively related with stroke volume and cardiac output during exercise (r = 0.68, P = 0.01 and r = 0.78, P < 0.01, respectively; Fig. 2). IC/TLC was positively related to stroke volume and negatively related to systemic vascular resistance in COPD patients (r = 0.77, P < 0.01 and r = −0.75, P < 0.01, respectively). Furthermore, P_es,inspTI was positively related to cardiac output (r = 0.56, P = 0.047), whereas P_es,expTI was not related to cardiac output (r = 0.06, P = 0.42). These relationships were not present in controls (P ≥ 0.43). Last, dyspnea was negatively related to cardiac output and stroke volume during exercise in COPD patients (r = −0.61 and r = −0.74, both P < 0.03).

Fig. 1.

Central hemodynamics in controls and chronic obstructive pulmonary disease (COPD). Cardiac output (A), stroke volume (B), and heart rate (C) in controls (n = 10) and COPD patients (n = 13) during exercise. COPD had lower cardiac output and stroke volume than controls (both P < 0.04), whereas heart rate was not different between groups (P = 0.66). Data are reported as median and 25–75 percentile interquartile range. Significant differences were determined via unpaired t-tests. *Significantly different from control. Men and women are presented in closed and open circles, respectively.

Fig. 2.

Relationships between central hemodynamics and hyperinflation and intrathoracic pressure. Relationships between central hemodynamics and hyperinflation, inspiratory intrathoracic pressure, as well as systemic vascular resistance in controls (○) and chronic obstructive pulmonary disease (COPD) patients (●). There was a positive relationship between FEV₁ and stroke volume during exercise in COPD patients (A: r = 0.68, P = 0.01). Furthermore, there were positive and negative relationships between ratio of inspiratory capacity to total lung capacity (IC/TLC) and stroke volume (B) and systemic vascular resistance (D) (r = 0.77, P < 0.01 and r = −0.75, P < 0.01) in COPD patients during exercise. Last, inspiratory esophageal pressure time integral (P_es,inspTI) was positively related to cardiac output (C) (r = 0.56, P = 0.047). No significant relationships were present in the controls (all P ≥ 0.43).

V_E-matched submaximal exercise responses.

In a subset of COPD (n = 10) and controls (n = 8), the submaximal exercise responses were compared at similar V_E (COPD: 35 ± 4 vs. controls: 34 ± 3 L/min, P = 0.71) and V̇o₂ (COPD: 1.3 ± 0.1 vs. controls: 1.3 ± 0.2 L/min, P = 0.27). COPD patients, compared with controls, had more negative peak P_es,insp (COPD: −19 ± 6 vs. controls: −11 ± 3 cmH₂O) and P_es,inspTI (COPD: −12 ± 2 vs. controls: −7 ± 3 cmH₂O·s) as well as lower IC (COPD: 2.4 ± 0.5 vs. controls: 3.8 ± 0.9 L) and IC/TLC (COPD: 33 ± 10 vs. controls: 55 ± 7%) (all, P < 0.01) during exercise. COPD patients had lower cardiac output (COPD: 10.5 ± 1.5 vs. controls: 12.4 ± 1.0 L/min, P < 0.01) and stroke volume compared with controls during exercise (COPD: 105 ± 15 vs. controls: 122 ± 18 mL, P = 0.03), whereas heart rate was not different between groups (COPD: 101 ± 10 vs. controls: 101 ± 8 beats/min, P = 0.82).

DISCUSSION

Major findings.

The major findings of the present study were threefold. During exercise, COPD patients exhibited lower cardiac output and stroke volume than age-matched healthy adults working at the same absolute V̇o₂. Second, stroke volume was positively related to IC/TLC such that the COPD patients with greater hyperinflation (lower IC/TLC) had a lower stroke volume during exercise. Third, and in contrast to our hypothesis, more negative intrathoracic pressure during inspiration was associated with a lower cardiac output during exercise in COPD patients. These findings suggest that hyperinflation and aberrant intrathoracic pressures are likely significant contributors to COPD-related impairments in central hemodynamics.

COPD and cardiac output.

COPD is associated with cardiac structural and functional abnormalities resulting in lower resting stroke volume (16, 35, 47). During exercise, however, the effect of COPD has been highly controversial with studies suggesting cardiac output is similar or reduced compared with controls during exercise (4, 9, 16, 25, 31). Likely explanations for these discrepant findings include exclusion of a control group and comparisons at relative exercise intensities. This latter experimental strategy is problematic, as cardiac output is compared at different absolute V̇o₂ and workloads. To circumvent this issue in the present study, we compared COPD and controls at an absolute V̇o₂ during exercise. We found that cardiac output was lower in COPD owing to a lower stroke volume during exercise. Importantly, the cardiac output reported herein (i.e., 11.0 L/min) during exercise at a V̇o₂ of 1.3 L/min is in line with cardiac output measured via direct Fick (i.e., ~11.7 L/min at a V̇o₂ of ~1.1 L/min) (43). Our findings are consistent with those of Bogaard et al. (4) who found an attenuated increase in stroke volume and cardiac output in COPD compared with controls at matched workloads during exercise. It is also important to note that the COPD patients had a greater cost of exercise than the healthy controls, likely resulting from a greater cost of breathing and reduced locomotor mechanical efficiency (33, 34). These findings have important implications during submaximal exercise, as they indicate that COPD patients have a greater reliance on peripheral oxygen utilization than controls.

Previous studies have highlighted the variable cardiac output and stroke volume responses during exercise in COPD (4, 43, 45). For example, cardiac output in COPD has been reported to range between 50 and 190% predicted at a V̇o₂ of 800 mL/min during submaximal exercise (43). In line with this, Vassaux et al. (45) found a positive relationship between IC/TLC and oxygen pulse at peak exercise in COPD. However, the interpretation of this latter finding is limited, as oxygen pulse was used as a surrogate for stroke volume, and the COPD patients had different peak exercise capacities (and presumably central hemodynamics). In the present study, we found a positive relationship between stroke volume and IC/TLC at matched V̇o₂ during exercise in COPD patients. These findings indicate that stroke volume is reduced in COPD patients during exercise, and this impairment is further exacerbated by greater hyperinflation. Future studies are necessary to determine whether partial alleviation of the hyperinflation during exercise improves cardiac hemodynamics during exercise in COPD patients. Taken together, these data suggest that significant hyperinflation incurs critical cardiovascular consequences in COPD patients during exercise.

Possible mechanisms.

Potential mechanisms affecting stroke volume during exercise in COPD are breathing mechanics and pulmonary vascular alterations (7–9, 29, 43). The impact of respiratory mechanics and, specifically, intrathoracic pressures on central hemodynamics has been controversial. For example, Stark-Leyva et al. (42) found that expiratory loading resulted in reductions in cardiac output and stroke volume during exercise in healthy humans. Furthermore, voluntary hyperinflation (without expiratory loading) increased stroke volume to control levels during exercise due to the more negative intrathoracic pressures (42). In contrast, Montes de Oca et al. (29) found that greater inspiratory intrathoracic pressure was significantly predictive of lower oxygen pulse during peak exercise in COPD patients. Additionally, Cheyne et al. (7, 8) found that more negative intrathoracic pressure during inspiration as well as dynamic hyperinflation were associated with decreased left ventricular stroke volume, likely mediated by direct ventricular interaction in healthy humans at rest. In the present study, we found that more negative intrathoracic pressure (per breath) during inspiration was positively related to cardiac output, and IC/TLC was negatively related to systemic vascular resistance in COPD patients during exercise. These data suggest that the more negative intrathoracic pressure during inspiration in COPD patients impairs the central hemodynamic response, likely due to augmented afterload and/or direct ventricular interaction. Since the COPD patients also exhibited substantially greater expiratory pressures than controls during exercise, the augmented expiratory load likely also contributed to the impaired cardiac response (42). To further elucidate these underlying mechanisms, future studies are required to determine the impact of the intrathoracic pressure swings during a cardiac cycle in COPD patients during exercise. It is possible that pulmonary vascular alterations may have also impacted the stroke volume response in COPD (3, 16, 35). For example, pulmonary vascular resistance has been reported to not change from rest to exercise in COPD patients (with some presenting with pulmonary hypertension), which can significantly influence stroke volume via elevated right ventricular afterload (16, 35). Future studies are needed to investigate the contribution of the pulmonary vasculature in limiting stroke volume in COPD patients with moderate disease severity without pulmonary hypertension.

Methodological considerations.

First, the open-circuit acetylene technique was used to noninvasively measure cardiac output during exercise. COPD patients exhibit inherent ventilation-perfusion mismatch (36), which can influence the mixing of acetylene within the lung. As a result, the cardiac output measurements during exercise may have been underestimated (5). Second, we acknowledge the relatively small sample size in the present investigation. Studies with larger sample sizes may be necessary to confirm our findings. Third, invasive pulmonary hemodynamics were not measured to determine the incidence of pulmonary hypertension or measure pulmonary vascular resistance. However, a recent study found the prevalence of pulmonary hypertension to be low in GOLD (Global Initiative for Chronic. Obstructive Lung Disease) stages II and III (i.e., 5 and 27%) (14). Finally, gastric pressure and thus expiratory muscle pressures were not measured in the present study. Future studies are warranted to determine the contribution of gastric pressure to stroke volume in COPD during exercise.

Conclusions.

During exercise, COPD patients exhibited an attenuated central hemodynamic response during exercise compared with controls. On the basis of these data, we conclude that, in adults with COPD, those with the most severe hyperinflation and more negative inspiratory intrathoracic pressures will exhibit the most significant impairment in exercise-related central hemodynamics. Future studies using chest wall kinematics are necessary to determine whether mitigation of hyperinflation and aberrant intrathoracic pressures will improve central hemodynamics during exercise in COPD patients.

GRANTS

This work was supported by the National Heart, Lung, and Blood Institute (HL-126638 to T. P. Olson and HL-071478 to B. D. Johnson) and the American Heart Association (18POST3990251 to J. R. Smith].

DISCLOSURES

No conflicts of interest, financial or otherwise, are declared by the authors.

AUTHOR CONTRIBUTIONS

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