Persistently elevated heart rate accelerates the progression of arterial stiffness

A large number of epidemiologic studies have shown that heart rate (HR) is an important risk factor for cardiovascular disease, independent of currently accepted risk factors [1,2]. This association has been demonstrated by various research groups in general populations, in hypertensive or diabetic patients, and in patients with cardiac diseases [1,2]. It was estimated that an increase in office HR of 10 bpm (beats per minute) was associated with a 20% increased risk of cardiac death [3]. According to results from the Ohasama study [4], a 5 bpm increase in home-measured resting HR was associated with a 17% increase in cardiovascular mortality. These data leave little doubt about the clinical importance of an increase in HR and indicate that this clinical variable should be taken into account when evaluating the risk profile of a hypertensive, diabetic, or cardiac patient. Although the association between HR and cardiovascular disease is well documented, it is difficult to establish the actual role of HR in determining cardiovascular events because of the complex interaction between the various risk factors. Elevated HR is directly related to an autonomic imbalance characterized by sympathetic predominance, and it is known that sympathetic hyperactivity may favor the development of atherosclerotic lesions and trigger cardiovascular events [5]. However, a crucial and still controversial issue, in this context, is whether fast HR is only a marker of autonomic imbalance or can also be considered as a causative factor in the induction of cardiovascular risk. Several lines of research have demonstrated that the relation between HR and atherosclerotic lesions can also be explained by a number of direct mechanisms [2,5]. A relationship between HR and coronary artery lesions has been demonstrated in experimental studies on monkeys, whereas an association between HR and carotid stenoses has been found in human beings, suggesting that tachycardia might produce a chronic stress on the arterial wall favoring the progress of atherosclerosis [2,5]. It has been postulated that a fast HR increases the magnitude and frequency of the tensile stress on the arterial wall, and that, by shortening the diastolic phase of the cardiac cycle, it facilitates the exposure of the endothelium to the deleterious effect of an increased pressure load and oscillatory shear stress [6]. Moreover, the extent of HR variations was found responsible also for changes in blood pressure (BP) variability [7]. These mechanisms may facilitate the development of atherosclerosis and vascular stiffening. Indeed, a cross-sectional association between HR and large artery stiffness has been documented in epidemiologic studies [8,9]. Furthermore, Benetos et al.[10] found that increased HR was one of the most powerful predictors of accelerated progression in pulse wave velocity (PWV) in a 6-year longitudinal study. Participants with a HR more than 80 bpm at the time of the first visit had a more pronounced progression in arterial stiffness compared to participants with a HR less than 60 bpm. An association between high HR and increased arterial stiffness has been found also in experimental studies on animals. An increase in HR obtained with atrial pacing produced a progressive reduction in carotid distensibility in rats [11]. The same phenomenon was observed in sympathectomized animals, suggesting that an increase in HR itself, and not an underlying hyperactivity of the sympathetic system, was the responsible factor [11]. Taken together, these data indicate that hemodynamic features are an important factor in determining arterial stiffness, and emphasize the importance of HR in the progression of arterial rigidity. However, in the above-mentioned epidemiologic studies, the occurrence of a causal link between high HR and arterial stiffness was established on the basis of HR measured at baseline, whereas no information was available about changes in HR during follow-up. Changes in HR over time, due to lifestyle modifications or induced by cardioactive drugs, may weaken the association between baseline HR measurement and future progression of large artery stiffness. In this respect, the article by Tomiyama et al.[12] published in the present issue of the Journal sheds new light on such a possibility. The authors specifically studied the temporal association between rates of increase in PWV and HR measured not only at baseline but also during a 5–6-year follow-up in 1795 apparently healthy Japanese participants (mean age 39 years). Brachial-ankle PWV, measured through a volume-plethysmographic apparatus, and serum C-reactive protein (CRP) levels were assessed at baseline and end of the follow-up period. BP measured via an oscillometric device and HR derived from an electrocardiogram were simultaneously recorded during the measurement of the brachial-ankle PWV. Indeed, a reason of strength of the present study is that the HR values used for the data analysis were determined based on electrocardiographic recordings. The authors found that both HR at the baseline examination and changes in HR during the follow-up period were significantly associated with the corresponding changes in PWV during the study period. These relationships held true also after adjustment for a variety of atherogenic risk factors. The predicting value of HR was evaluated also by dividing the participants into four categories classified on the basis of baseline HR (higher or lower than the median in the group) and increase/decrease in HR during the follow-up. Serum CRP levels and atherogenic risk factors were significantly worse in the participants with high baseline HR accompanied by increased HR during the study period. Also changes in PWV during the study period were significantly higher in this group compared to the other three groups, suggesting a synergistic role of high baseline HR and increase in HR during the study period in accelerating age-associated increases in arterial stiffness. The importance of follow-up HR in predicting outcome over and above baseline HR has been documented in several prospective studies. In the Hypertension and Ambulatory Recording Venetia Study (HARVEST), both baseline and follow-up HRs were potent predictors of subsequent development of hypertension needing drug therapy [13]. Participants with permanent but not transient tachycardia had an increased risk of hypertension during 6 years of observation. A more recent analysis showed that fast HR at both baseline and follow-up assessment was a significant predictor of weight gain (HARVEST study, unpublished observations). In addition, in a subgroup of 124 participants in whom also arterial elasticity was studied, high HR and high systolic BP were found to be the main determinants of impaired arterial elasticity over time (HARVEST study, data on file). In the HARVEST study, large artery (C1) and small artery (C2) compliance were assessed from radial arterial pulse waves recorded with a tonometer sensor array at baseline and study end. PWV and augmentation index were also obtained with applanation tonometry using a different instrumentation. In agreement with the results by Tomiyama et al.[12], an association of arterial stiffness with high HR was observed when considering both baseline and follow-up measurements. In multivariate regression analyses including sex, age, smoking, alcohol and coffee use, physical activity, parental hypertension, antihypertensive drug treatment (yes/no), height, baseline and final body mass, and baseline distensibility parameters as independent variables, baseline BP and spontaneous or drug-induced changes in BP were the main determinants of changes in C1 (both P < 0.001) and augmentation index (P = 0.03 and 0.04). The other main determinant of changes in the distensibility parameters was HR measured at baseline (P < 0.001 for C1) and at final assessment (P < 0.001 for C1 and P = 0.002 for PWV). Spontaneous [14] or drug-induced [15] changes in HR were found to be also predictive of cardiovascular mortality. Recently, Jouven et al.[14] in 5139 asymptomatic men free of clinically detectable cardiovascular disease assessed whether changes in HR during the first 5 years of follow-up were associated with changes in later mortality risk over a 23-year follow-up. After adjusting for confounding factors and baseline HR, participants with decreased HR during the 5-year observation period had a 14% decreased mortality risk, whereas participants with increased HR during the 5 years had a 19% increased mortality risk. This relationship was observed for both cardiovascular and noncardiovascular mortality. Participants with both a low baseline HR at rest and a decreased HR during the 5 years had the lowest mortality risk, whereas participants with a high HR at baseline and increased HR during the 5 years had the highest risk. The decrease in HR observed over 5 years and the associated reduction in the risk of mortality might be due to treatment with β-blocking agents or other HR-lowering drugs. However, according to the authors of this study, such a possibility was unlikely because use of these compounds was uncommon. In addition, results were unchanged after adjustment for modifications in other risk factors during the follow-up, including BP, making it unlikely that the association between changes in HR and mortality was due to a modified influence by other risk factors. Results obtained within the frame of the International Verapamil–Trandolapril Study (INVEST) are in keeping with the above data [15]. The purpose of the INVEST study was to investigate the BP control and adverse outcomes using either a verapamil-slow release-based or an atenolol-based treatment strategy in hypertensive patients with coronary artery disease. In this study, mean HR during an average follow-up of 2.7 years was calculated using all available follow-up visits. Elevated baseline resting HR was associated with increased incidence of adverse outcomes with an approximately two-fold increase among patients with HR more than 100 bpm. Mean follow-up HR was also strongly associated with the risk of adverse outcomes and a J-shaped relationship was observed with a nadir at 59 bpm. However, when tested in the same Cox model with follow-up HR, baseline HR was no longer a significant predictor of outcome, suggesting that follow-up HR contributes more information to the prediction of adverse outcomes than baseline pretreatment HR. Recently, Paul et al.[16] analyzed the predictive value of baseline HR, final HR, and HR change during follow-up in the patients attending the Glasgow Blood Pressure Clinic. Analysis of final HR data showed that the highest risk of all-cause mortality was associated with a HR of 81–90 bpm, and the lowest risk was associated with a HR of 61–70 bpm. Using a threshold of 80 bpm, patients were classified into those who had a consistently high (high->high) or low (low->low) HR or patients whose HR increased (low->high) or decreased over time (high->low). Compared to low->low patients, high->high patients had an increased risk of all-cause and cardiovascular mortality. The next highest risk group was low->high followed by high->low. Patients whose HR increased or remained high throughout follow-up had significantly higher BPs. This indicates that HR in treated hypertensive patients might be a surrogate for BP control. However, in a multivariate model, which included the achieved systolic BP, HR remained a significant predictor of outcome indicating an additional specific effect of HR itself. Overall, these data from the Glasgow Blood Pressure Clinic and the data provided by previous studies strongly indicate that changes in HR achieved during follow-up represent a better predictor of risk than baseline HR. This may be due to the fact that final HR, or the HR achieved during follow-up, is a more stable trait than baseline HR and is related with changes in HR from baseline. Thus, it may be more representative of the overall hemodynamic bulk on the arterial wall over time than baseline HR is. This may account for the close relationship between final HR and follow-up changes in PWV found by Tomiyama et al. in their study [12]. But why should an increased resting HR cause a faster progression of arterial stiffness? A moderate increase in resting HR (<100 bpm) results in a greater cardiac output and, accordingly, in an increased BP and BP-related tensile stress [6]. The increased stress on the endothelium exerted by a high frequency of blood pressure fluctuations may ultimately lead to endothelial damage. In addition, increased magnitude of the tensile stress imposed on the wall upregulates the expression of proatherosclerotic genes by the endothelial cells (Fig. 1) [17]. Indeed, a correlation between high stress index (rate-pressure product) on the arterial wall and atherosclerotic lesions in the infrarenal aorta and iliac arteries was found in monkeys, emphasizing the importance of the synergistic effect of fast HR and hypertension on the formation of atherosclerotic plaques [18]. Thus, both the amount of mechanical load and the frequency of the mechanical swings imposed on the arterial wall seem to play a key role in the development of atherosclerosis.Fig. 1As mentioned above, one of the mechanisms that implicate high HR in the induction of atherosclerosis is an accelerated progression of arterial stiffness. As shown with direct intra-arterial recordings in humans, when HR increases, the duration of cardiac cycle diastolic phase gradually decreases, resulting in an increase in the total time spent on systole. As a consequence, the endothelium is exposed over time to a higher mean BP and to the atherogenic effect of the systolic load and oscillatory shear stress for longer periods. The increased load and oscillatory shear stress associated with high HR may impair the physiological balance between the synthesis and breakdown of collagen and elastin, thus favoring collagen deposition and promoting vascular stiffening [6,17]. The above considerations may account for the association between HR and arterial stiffness found by Tomiyama et al. in their study [12]. However, as the authors themselves admit, this might be a spurious relationship due to the algorithm employed by the device to measure arterial stiffness. If HR was used by the algorithm to identify the start of the rising limb of the pulse wave, it is possible that high HR mathematically leads to an increased PWV. However, as mentioned above, in the HARVEST study, a longitudinal association of final HR was found not only with PWV but also with large artery compliance assessed with a different device, making it unlikely that the arterial distensibility–HR relationship is a mere mathematical artefact. If a permanently elevated HR is associated with progressive worsening of arterial elasticity, it follows that HR reduction with nonpharmacological measures and/or drug treatment would retard the progression of arterial stiffening. Results from the HARVEST study suggest that whenever HR remained elevated throughout a 6-year period of observation, a worsening of arterial distensibility could be observed in spite of a good BP control with treatment (data on file). The results obtained in the recent analysis of the Glasgow Blood Pressure Clinic study [16] are in keeping with these findings because also in that study, the risk was high in the patients whose BP was well controlled but HR remained elevated. The interesting perspectives offered by the study by Tomiyama et al.[12], however, need to be considered in the light of some study limitations, acknowledged by the authors themselves. Besides the possible mathematical artefacts in the assessment of the association between high HR and increased PWV mentioned above, it has to be emphasized that most of the study participants were healthy middle-aged men, with only a small number of female participants included in the population being examined. Because of previous reports on sex-related differences in the relationship between HR and PWV [19], the findings by Tomiyama et al. would need to be confirmed by additional studies also including elderly and female participants. Additional data on such relationship obtained in participants with high cardiovascular risk and/or advanced atherosclerosis would also be required to fully exploit the clinical relevance of these intriguing observations [1].