Lipid abnormalities, hypertension, diabetes, smoking, and obesity are the traditional atherosclerotic risk factors. These risk factors cause endothelial dysfunction, inflammation, and altered biological pathways [¹–³] which are well-recognized final common pathways that lead to the development of systemic artery atherosclerosis. These risk factors are also associated with impaired systemic circulation, which can also cause peripheral vascular remodeling. Peripheral vascular remodeling significantly influences systemic peripheral vascular resistance (SVR). SVR is thought to be one of the major determinants of cardiac function during exercise [¹]. Thus, in patients with heart disease, peripheral vascular remodeling plays a key role in the pathophysiology of impaired exercise tolerance [⁴].

Flow-mediated vasodilation (FMD) in the brachial artery is one of the most widely used noninvasive diagnostic modalities [⁴, ⁵] for cardiovascular morbidity and mortality. However, though there are many studies dealing with endothelial function and SVR at rest, there is scant information about the relationship between peripheral vascular endothelial function and SVR during exercise. As well, the relationship between peripheral endothelial function and cardiac function reserve in response to exercise has not been fully investigated in patients with heart disease. This lack of information is related to the methodological difficulties that preclude precise evaluation of dynamic SVR and cardiac function during exercise [⁶].

However, recent advances in technology have enabled us to measure dynamic SVR and cardiac function during exercise at the same time. Recently, a new type of ambulatory radionuclide monitoring (VEST) system has been developed, which can be connected to computerized ambulatory-electrocardiogram monitoring using an IC card [⁷]. This VEST system is a useful and reliable piece of equipment for monitoring both SVR and left ventricular function during exercise [⁸, ⁹].

The purposes of the present study were (1) to investigate the role that peripheral endothelial function and dynamic changes in SVR during exercise have on left ventricular performance during exercise and (2) to clarify the relationship between peripheral endothelial function and SVR during exercise in patients with heart disease by using the new VEST system combined with cardiopulmonary function testing.

All patients exercised beyond AT (16.3 ± 3.3 mL/min/kg) until exhaustion without any ischemic ST-segment changes on electrocardiography. Heart rate and systolic blood pressure significantly increased from rest to AT or peak exercise. The double product and the body weight-corrected VO₂ showed similar changes to the heart rate (Table 1).

The sequential data obtained using the VEST system are shown in Table 2. The left ventricular ejection fraction and stroke volume significantly increased from rest to AT or peak exercise. The left ventricular ejection fraction and left ventricular peak ejection ratio significantly increased from rest to AT or peak exercise (Table 2).

The FMD measurement results showed a significant increase of brachial artery diameter and calculated brachial artery blood flow (diameter, from 5.1 ± 0.7 mm to 5.2 ± 7.3 mm, P < .05; blood flow, from 15 ± 13 mL/sec to 65 ± 21 mL/sec, P < .05). From these data, the calculated FMD was 2.5% ± 2.0%, which was below the normal range obtained in our laboratory [¹³]. In contrast, NTG-D was 13.5 ± 4.3, which was within the normal limits obtained in our laboratory.

There was a significant inverse correlation between the changes in the SVR from rest to AT and the changes in left ventricular ejection fraction or left ventricular peak ejection ratio from rest to AT (Figure 1). FMD was significantly correlated with the change in left ventricular ejection fraction from rest to AT, and it tended to correlate with the changes seen in left ventricular peak ejection ratio from rest to AT (Figure 2). However, these relationships were not observed for NTG-D.

ACE-inhibitors have a potent effect on endothelial function (FMD); their use may confound the relationship between these various parameters [¹⁴, ¹⁵]. When FMD values were compared in patients with and without ACE-inhibitors, the patients with ACE-inhibitors tended to have better values of FMD than those without ACE-inhibitors (2.5 ± 1.9% versus 1.8 ± 2.2, P < .1). Thus, when patients on ACE-inhibitor treatment were excluded from the analysis, the FMD significantly correlated with both the left ventricular ejection fraction and the left ventricular peak ejection ratio change that occurred from rest to AT (Figure 3). In addition, if the correlation between the FMD and SVR during exercise was analyzed in this group of patients, there were a tendency to an inverse correlation (Figure 4).

The present study has demonstrated that: (1) peripheral endothelial function possibly contributes to left ventricular performance during exercise; (2) SVR during exercise can determine left ventricular performance in response to aerobic exercise; and (3) peripheral endothelial function may partly contribute to SVR during aerobic exercise in patients with heart disease.

The correlation coefficient obtained in this study supports the possible presence of a significant relationship between peripheral endothelial function and left ventricular performance during aerobic exercise when patients on ACE-inhibitors were excluded. As well, a significant relationship between SVR and left ventricular performance was confirmed. In addition, peripheral endothelial function tended to correlate with the SVR obtained during aerobic exercise. Before reaching the anaerobic threshold, exercise is considered to be aerobic if no ischemia exists. In this study, no ischemic ECG changes were observed at the AT points in any of the patients studied. Thus, all of the patients were exercising aerobically when the contributions of peripheral endothelial function and SVR on left ventricular performance were observed. However, when the relatively low correlation coefficient r value and the influence of ACE-inhibitors were considered, the definite role of peripheral endothelial function on left ventricular performance during exercise should be confirmed by duplicated larger cohort study of our observation.

By using a combination of the VEST system and blood pressure measurements [¹⁶], we could calculate dynamic SVR. The changes in the SVR that occur during exercise reflect peripheral vasomotion during exercise. The VEST system is useful for evaluating not only serial cardiac function but also peripheral vasomotion during exercise in the upright position. By using this device, one can simultaneously evaluate left ventricular ejection fraction, cardiac output, stroke volume, left ventricular end diastolic volume, left ventricular end systolic volume, and the left ventricular peak ejection ratio [¹⁷, ¹⁸].

The findings in this study give a new insight into the exercise physiology of patients with heart disease. Cardiac performance is closely associated with dynamic SVR during exercise obtained using the new VEST system. Cardiopulmonary stress testing in combination with the VEST system can be more widely applied in clinical practice to help manage heart disease. Using this modality, we can evaluate the efficacy of a new medication or a new therapeutic option, such as biventricular pacing, in patients with heart failure. Furthermore, the mechanisms of action of new treatments can also be elucidated.

Peripheral endothelial function contributes not only to resting cardiac function but also to the dynamic cardiac function change that occurs during exercise. One could speculate that direct and indirect mechanisms are responsible for this phenomenon. However, the present study's findings cannot help clarify these mechanisms. While in one study nitric oxide aggravated cardiac function in patients with heart disease [¹⁹], other reports have shown that nitric oxide improved cardiac function in the failing heart [²⁰, ²¹]. There are three distinct nitric oxide synthases (NOS), neural (nNOS), inducible (iNOS), and endothelial (eNOS), that probably behave differently in the failing heart at rest and during exercise. Generally, nitric oxide is reported to modulate β adrenergic responsiveness [²²], and nitric oxide can regulate myocardial oxygen consumption (MVO₂) in the failing heart [²³, ²⁴]. Therefore, nitric oxide could possibly affect cardiac performance even during exercise [²⁵]. With respect to indirect mechanisms, exercise-induced hyperemia in either skeletal or cardiac muscle is at least partly nitric oxide-dependent [²⁶]. A previous report found that nitric oxide was related to microvascular function in humans [²⁷]. Exercise training and exercise therapy in patients with heart disease improved both endothelial function and peripheral vascular circulation [²⁸–³²]. Brachial artery endothelial function may be indirectly associated with cardiac function during exercise. Since endothelial function could possibly be related with cardiac function during exercise, medical therapy, including prescribing ACE-inhibitors, statins and/or antioxidant therapy, as well as exercise training, might improve cardiac function during exercise by improving endothelial function in patients with heart disease.

In our study, brachial (peripheral) endothelial function tended to correlate with the systemic vascular response that occurred during exercise. However, this tendency was only observed when patients on ACE-inhibitors were excluded from the analysis, as ACE-inhibitors have potent effects on endothelial function even after having been discontinued for 24 hours [¹⁴, ¹⁵]. There are many neurohormonal factors that play a role in exercise physiology and pathophysiology [³³]. Among them, nitric oxide has been reported to be involved in 20%–40% of the entire control of the vasomotion during exercise. Thus, the moderate correlation obtained in this study agrees with previously reported findings.

Study limitations: this study has several limitations. First, the study population was not homogeneous; thus, different baseline characteristics might have confounded the main findings of the present study. This study should be duplicated with more homogeneous populations. Second, endothelial function during exercise could not be measured due to technical problems. If a new technology that measures endothelial function during exercise could be developed, the role of endothelial function in systemic vascular resistance and left ventricular function during exercise would be further elucidated. Third, since recently statisticians criticize investigators for utilizing the correlation analysis between the changes in two measured values [³⁴] and recommend to use Blomquist's formula, our results might have less impact on the relation between peripheral endothelial function and left ventricular performance during exercise. However, the criticizing report is relatively new and, in addition, commercially available personal computer package of this statistics does not exist so that we could not analyze our data by using Blomquist's formula. Our findings should be further confirmed by new analysis in the near future. Fourth, by the recent advance of technology, FMD and NTG-D now can be measured by on-line system [³⁵] so that our method of measuring FMD and NTG-D is somewhat out of date. We did not have duplex measurements to obtain flow and shear rates, and the time to peak for FMD could not be determined by our method. In contrast, recent report [³⁵] especially showed that the time to peak for FMD exceeded more than 90 seconds. In addition, about NTG-D, on-line continuous measurement for NTG-D, is supposed to be important to obtain peak NTG-D value. These somewhat out-of-date methods likely confound the interpretation of FMD, NTG-D and the reactive hyperemic response of brachial artery flow. Our data again should be duplicated in recent sophisticated equipment. However, since the huge amount of articles on FMD and NTG-D have been published by using same measurement method as ours, our findings have significant message on the relation between peripheral endothelial function and left ventricular performance during exercise. Lastly, our Vest system can only measure relative change of left ventricular dimension so that absolute value of the left ventricular volume parameters such as stroke volume that directly influence exercise tolerance could not be evaluated. This is another limitation of this study.

Despite these limitations, we can conclude that peripheral endothelial function and vasomotion, including SVR, might play a significant role in determining cardiac performance during exercise in patients with heart disease. This information provides a new insight into the understanding of the exercise physiology of patients with heart disease.