Assessing blood pressure response to exercise: methodological issues and clinical relevance.

Abstract

A number of studies have focused on individual differences in blood pressure (BP) response to exercise, either to investigate cardiovascular control mechanisms or to explore the clinical relevance of this BP pattern, in particular when facing an exaggerated BP elevation during a stress test. Indeed, an increased exercise systolic (S) BP has been reported to be associated with a higher incidence of myocardial infarction, stroke and all-cause or cardiovascular mortality [1–8]. Moreover, increased exercise SBP has also been associated with incident hypertension independently of baseline resting BP or other cardiovascular risk factors [9–13]. On the contrary, due to methodological difficulties, only a few studies have addressed the clinical relevance of increased exercise diastolic (D) BP, also in this case providing evidence of its association with cardiovascular mortality [5], cardiovascular diseases [4] and incident hypertension [12,13]. Given the well known difficulties in measuring BP during exercise and given the great variety of stress test protocols applied in different laboratories, available data on this issue have been critically addressed in the 2013 European Society of Hypertension/European Society of Cardiology Guidelines for hypertension management [14]. It is well known that during dynamic exercise, either performed on a bicycle ergometer or on a treadmill, BP displays a variable increase, different in different individuals, with SBP increasing more than DBP [15]. This represents a physiologic phenomenon, due to both central and reflex neural mechanisms modulating the cardiovascular system during physical exercise. However, when the BP response to exercise is exaggerated, it may have a pathophysiologic relevance, as it may reflect or predict pathologic cardiovascular alterations. Theoretically, an exaggerated BP reactivity to exercise should also have clinical relevance, indicating, among other mechanisms, a more pronounced reactivity of the sympathetic nervous system to physical activity and/or the presence of structural changes in peripheral arteries that prevent their physiologic dilatation during exercise. Indeed, the degree of SBP increase observed at submaximal exercise has been found to be related to preexercise BP, age, arterial stiffness and abdominal obesity and is somewhat greater in women than in men and in unfit than in fit individuals [15–19]. In spite of its pathophysiological interest, however, the clinical relevance of the quantification of BP response to exercise in a stress test laboratory has been questioned, as it may not be sufficiently accurate to represent a clinically useful parameter. In fact, only SBP can be measured reliably enough with noninvasive methods in these conditions, while DBP assessment during exercise is largely inaccurate [14]. On the background of these methodological problems, it is therefore not surprising that the definition of ‘normal’ BP response to dynamic exercise is still a matter of debate, with no consensus among experts also because, at variance from the definition of arterial hypertension, there is no evidence from randomized clinical trials to support either a definition of normal exercise BP nor the definition of thresholds indicating ‘exercise hypertension’. Notwithstanding these difficulties, levels of SBP more than 210 mmHg in men and more than 190 mmHg in women achieved during exercise have nevertheless been proposed as indicating the occurrence of ‘exercise hypertension’ in a number of studies, although also other different definitions of an exaggerated BP response to exercise have been suggested [16,17]. The most relevant clinical information reported in most although not in all studies addressing BP reactivity to exercise in normotensive individuals is the possibility that an excessive rise of BP in this condition predicts the future development of arterial hypertension, independently of BP levels measured at rest [9,15,16]. In spite of these interesting observations, however, exercise testing is not currently recommended to predict future hypertension because of the above-mentioned methodological limitations of this approach, such as the acknowledged inaccuracy of BP measurements and lack of standardization of the exercise methodology and of the related BP assessment and thresholds. Furthermore, there is conflicting evidence on the association of exercise BP with organ damage, such as left ventricular hypertrophy, after adjustment for resting BP levels and other covariates, either in normotensives or in hypertensive patients [15,16]. An additional area of controversy is the prognostic significance of exercise BP in either normotensive or hypertensive individuals in relation to cardiovascular events incidence, for which available results are not consistent [17]. Such a controversy may be due to the fact that the two haemodynamic components determining BP levels, namely cardiac output and peripheral resistance, variably change in opposite directions during dynamic exercise. In fact both the degree of reduction in systemic vascular resistance and the degree of increase in cardiac output may be largely different in different individuals. We may hypothesize that the most important prognostic factor in this regard could be a blunted reduction of systemic vascular resistance during exercise, compatible with structural pathophysiological changes in peripheral arteries [15,19,20]. However, the possibility that an impaired arterial dilatation translates into an excessive rise of BP during exercise may at least partly depend on individual cardiac function, in particular on how cardiac output changes in this condition. Indeed, in normotensive individuals and in mild hypertensive patients with adequate increase of cardiac output during exercise, an exaggerated BP response in this condition does predict a poorer long-term outcome [17,21], although this might not be the case in individuals with impaired cardiac output modulation during physical activity. Indeed, when hypertension is associated with cardiac dysfunction and blunted exercise-induced increase of cardiac output, the prognostic significance of exercise BP is likely to be lost [20]. Somehow paradoxically, a higher BP during exercise may even carry a better prognosis in specific clinical settings, such as in elderly individuals [22], in patients with suspected cardiac disease [23] or with heart failure [24], because in all these conditions, a higher exercise BP implies relatively preserved systolic cardiac function [16]. Although all the above controversies in relation to the clinical relevance of an excessive BP response to exercise still prevent a dynamic stress test to be recommended as a routine diagnostic or prognostic evaluation in hypertensive patients, it has nevertheless been proposed that exercise-induced hypertension, in the presence of normal BP levels at rest, should at least represent an indication to perform out-of-office BP monitoring, as it may indicate the occurrence of a masked hypertension condition [25]. Because of all these still ongoing discussions, the recent European hypertension management guidelines [14] have raised doubts on the clinical utility of BP measurements during exercise testing for diagnostic and prognostic purposes in patients with high blood pressure, although acknowledging that exercise testing can be useful as a general prognostic indicator when focusing on exercise capacity and on electrocardiographic parameters (exercise test is indeed commonly performed for these reasons in cardiac patients). At the same time, these guidelines support the view that an abnormal BP response to exercise, although not accurate enough to represent a prognostic indicator and a clinical target in itself, warrants performance of a 24-h ambulatory BP monitoring, because, as mentioned above, such an exaggerated pressor response may indicate the occurrence of a more generalized BP reactivity to daily life challenges, with the possible occurrence of masked hypertension, which is best assessed through 24-h ambulatory BP recordings. An additional interesting contribution to the discussion on the clinical importance of BP response to exercise is provided by the study by Lorbeer et al.[26] published in this issue of the Journal of Hypertension. The aim of this study was to identify relevant cut-off values for increased exercise BP able to predict incident hypertension. This was done by considering different cut-off values to define increased exercise SBP, DBP and pulse pressure measured at a submaximal work load of 100 Watts (W) and at the maximal achieved work load during a cycle ergometer test in women and men selected from a population sample. The authors analysed data of 661 normotensive individuals (386 women) aged 25–77 years from the Study of Health in Pomerania, Germany, including observations obtained at baseline and after a 5-year follow-up. Exercise BP was measured at a submaximal level of 100 W and at maximum work load level of a symptom-limited cycle ergometry test. Cut-off values for increased exercise BP were defined by considering the maximum sum of sensitivity and specificity for the prediction of incident hypertension. To this aim, the area under the curve (AUC) and net reclassification index (NRI) were calculated to investigate whether increased exercise BP adds predictive value for incident hypertension beyond established cardiovascular risk factors. In the study by Lorbeer et al.[26], only in men was exercise SBP at baseline positively associated with incident hypertension both at the 100 W level and at the maximum level of work load independently of cardiovascular risk factors. Exercise DBP was not associated with incident hypertension either in women or in men. Exercise pulse pressure was only associated with incident hypertension in men at the 100 W level. More specifically, analysis of this study data identified in men values of 160 mmHg for a 100 W exercise level, and of 210 mmHg for a maximum exercise level as optimal cut-off values for the definition of increased exercise SBP able to predict incident hypertension during the follow-up period. In women, despite the lack of a statistically significant association, the authors suggest a cut-off SBP value of 190 mmHg at maximum exercise level as possibly predicting incident hypertension. The study by Lorbeer et al.[26] thus offers interesting indications for a better standardization of exercise BP measurement in healthy and normotensive individuals, as well as for the definition of cut-off values to predict incident hypertension based upon a prospective population study with a 5-year follow-up. The results of this study on one side confirm the threshold SBP values for exercise hypertension of more than 210 mmHg for men and more than 190 mmHg for women at maximum level of exercise, which had been already proposed by a number of previous studies. On top of this, however, the findings from the study by Lorbeer et al.[26] suggest that a SBP cut-off point of 160 mmHg can be used to identify exercise hypertension at a submaximum level of workload in men, and that this value can also be of clinical relevance. This is an important finding for clinical practice, because it may be more practical for clinicians and researchers to use a BP cut-off value at 100 W exercise level, given that such a level is more feasible and easy to be standardized in a clinical setting, thus avoiding possible test failures during exercise at maximum level. Unfortunately, a reliable exercise SBP cut-off point criterion during submaximum level of exercise could not be found in women, probably because of the relatively small number of study participants. Among the positive features of the study by Lorbeer et al.[26], we have to include the longitudinal population-based cohort design, the accurate assessment of a large variety of established cardiovascular risk factors and the systematic search for optimal cut-off values for increased exercise BP, based on incident hypertension during follow-up. Some limitations also need to be acknowledged, however. Due to the existence of different exercise testing methods and protocols, the results of this study have to be limited to bicycle ergometry at the 100 W level and at the maximum exercise capacity level, and to middle-aged participants characterized by an average fitness level and with a significant experience of bicycling. Indeed, there are mainly two methods of performing a graded exercise test in clinical practice, namely by means of a bicycle ergometer or a treadmill. Kim et al.[27] compared both methods in terms of exercise BP and found that the SBP response is more pronounced when using a bicycle ergometer than a treadmill. Therefore, different ergometric tests performed with these two different methodologies may exert a different burden on the cardiovascular system in humans mainly because the exercise muscles involved are different between the two test methods. The study by Lorbeer et al.[26] did not compare the BP response between these two test methods. Therefore, given that potential differences caused by related methodological issues cannot be excluded, future studies comparing these two frequently used exercise test approaches are needed, in order to clarify whether the proposed BP thresholds during exercise apply only to bicycle ergometry or also to a stress test carried out by using a treadmill. Another limitation of the study by Lorbeer et al.[26] is the relatively small sample size, due to the fact that a number of individuals within the explored population did not meet the inclusion criteria and to the fact that others were lost at follow-up. This may have been one of the reasons for the limited evidence gathered in women. Similarly, the statistical power was also not sufficient to analyse age-specific cut-off values. Additionally, duration of follow-up was relatively short in this study. Currently, setting up diagnostic criteria for exercise hypertension could best be done by correlating baseline exercise BP with the development of future hypertension and with the occurrence of cardiovascular events during several years of follow-up. In this cohort, 661 normal adult individuals were followed up for 5 years by focusing on incident hypertension, but during such a relatively short follow-up, the incidence of cardiovascular endpoints was low. This prevented authors to explore the association between exercise BP and incidence of hard endpoints, also because their study was not powered enough for such an analysis. A longer follow-up of this cohort would thus be required for a reliable future analysis of the possible relation between exercise BP and hard outcomes in this study. By pooling all the available data on BP response to exercise together and by considering, as mentioned above [14], that exercise hypertension is not only related to the exercise itself but also related to preexercise resting BP levels, age and arterial stiffness, a careful clinical evaluation of cardiovascular risk profile can be recommended in an individual whenever a pronounced BP reactivity to physical exercise is found. This recommendation can be issued even if controversial evidence is at present available on the correlation between exercise BP and target organ damage, such as left ventricular hypertrophy, after adjustment for baseline resting BP and other confounders in normotensive or in mild hypertensive individuals. A practical question at this point would be whether an exaggerated BP response to exercise warrants being buffered by pharmacological treatment. Overall, while waiting for additional evidence to be provided, on the basis of the data so far available it seems premature to prescribe antihypertensive treatment in this condition. On the contrary, as higher preexercise BP is more frequently found in those individuals with an exaggerated elevation of exercise BP, which may indicate a masked hypertension condition, out-of-office BP measurements, and in particular ambulatory BP monitoring should be implemented under these circumstances. In conclusion, the results of the interesting study by Lorbeer et al.[26] provide some practical indication on how to interpret the degree of SBP response to exercise in a clinical setting at least in men and further underline that an excessive elevation of BP during bicycle exercise may not be a benign phenomenon in normal individuals and thus may warrant further investigation, given its relation with incident hypertension during follow-up. At the same time, these data emphasize the need for further evidence to be collected in a larger number of individuals and by considering a longer follow-up, in order to more precisely characterize the clinical value of exercise BP, in particular in women. ACKNOWLEDGEMENTS Conflicts of interest There are no conflicts of interest.