The objectives of exercise testing in congestive heart failure (CHF) may be summarized as follows: (a) detect impaired cardiac performance, (b) grade severity of cardiac failure and classify functional capability, and (c) assess effects of interventions. Several different methods are available to make these assessments, and we have to ask ourselves how well exercise testing achieves these objectives. It has to be kept in mind that the power generated by the exercising muscles is dependent on the oxygen delivery to the skeletal muscles. Oxygen uptake is the result of an integrated performance of the lungs, heart, and peripheral circulation. In patients, as well as in normal subjects, oxygen uptake is related to hemodynamic indices such as cardiac output, stroke volume, or exercise duration when a stepwise regulated maximal exercise protocol is used. However, there are major differences in the concept of a true maximum in normal subjects versus heart failure patients. Fit-normal subjects will achieve a real maximal oxygen uptake, whereas patients may stop testing before a maximum is reached because of symptoms such as dyspnea or leg fatigue. Therefore, it is better if the actual oxygen uptake can be measured. ''Peak'' rather than true maximal oxygen uptake has been suggested for the classification of the severity of heart failure. Peripheral factors modify the cardiac output through such factors as vascular resistance, organ function, and hormonal release. Maximal exercise will stress the cardiovascular system to a point where the weakest chain will impose a limiting effect. In CHF this is supposed to be the heart. However, in a condition in which cardiac performance is already limited, evaluation at submaximal exercise may be more physiologically relevant to evaluate the whole system. Several submaximal protocols are available such as the 6-min walk and the self-powered treadmill. It has been assumed that improvement in exercise performance is related to improved survival. This concept has been questioned, and recent studies have not been able to confirm this assumption. In view of these problems of exercise testing, it is easy to appreciate that pharmacological effects on exercise performance are difficult to interpret. In individual patients, the effects of drugs on myocardial contraction and structure, vascular resistance, fluid retention, and so on, may be of differing importance. For example, angiotensin-converting enzyme (ACE) inhibitors have been evaluated in many studies and, overall, the evidence available supports the view that maximal exercise capacity indeed can be improved by treatment. However, from critical examination of the results it is evident that the generally held impression of a strong effect by such drugs as ACE inhibitors seems to be an overexaggeration of reported data. Maximal exercise testing with an increasing work load is valuable in evaluating patients' symptomatology, severity of heart failure, and prognosis. It is of more limited value for the assessment of pharmacological therapy. The use of other end points such as mortality, symptomatology, and/or neuroendocrine activation is advocated when the effect of drug treatmentis assessed in CHF.
