Article Text

Download PDFPDF

Differential effects of heart rate reduction by atenolol or ivabradine on peripheral endothelial function in type 2 diabetic patients
  1. Roberto Nerla1,
  2. Antonino Di Franco1,
  3. Maria Milo1,
  4. Dario Pitocco1,
  5. Francesco Zaccardi1,2,
  6. Pierpaolo Tarzia2,
  7. Filippo Maria Sarullo2,
  8. Angelo Villano1,
  9. Giulio Russo1,
  10. Alessandra Stazi1,
  11. Giovanni Ghirlanda1,
  12. Gaetano Antonio Lanza1,
  13. Filippo Crea1
  1. * 1Institute of Cardiology and Diabetes Unit, Institute of Internal Medicine, Università Cattolica del Sacro Cuore, Roma, Italy
  2. 2Cardiology Division, Buccheri la Ferla Fatebenefratelli Hospital, Palermo, Italy
  1. Correspondence to Dr Gaetano Antonio Lanza, Istituto di Cardiologia, Università Cattolica del Sacro Cuore, Largo A. Gemelli, 8, Roma 00168, Italy; g.a.lanza{at}


Objective To assess whether reduction of heart rate (HR) has beneficial effects on endothelial function in patients with type 2 diabetes mellitus (T2DM).

Design Randomised, double-blind, placebo-controlled study.

Setting University hospital.

Patients 66 T2DM patients without overt cardiovascular disease.

Interventions Patients were randomised to receive for 4 weeks, in addition to their standard therapy, one of the following treatments: atenolol (25 mg twice daily), ivabradine (5 mg twice daily) or placebo (1 tablet twice daily).

Main outcome measures Systemic endothelial function, assessed by flow-mediated dilation (FMD); endothelium-independent vasodilation, assessed by nitrate-mediated dilation (NMD); cardiac autonomic function, assessed by HR variability (HRV).

Results 61 patients completed the study (19, 22 and 20 patients in atenolol, ivabradine and placebo groups, respectively). Compared with baseline, HR was similarly reduced by atenolol (87±13 vs 69±9 bpm) and ivabradine (86±12 to 71±9 bpm), but not by placebo (82±10 vs 81±9 bpm) (p<0.001). FMD improved at follow-up in the atenolol group (4.8±1.7 vs 6.4±1.9%), but not in the ivabradine group (5.2±2.5 vs 4.9±2.2%) and in the placebo group (4.8±1.5 vs 4.7±1.7%) (p<0.01). NMD did not change significantly in any group. HRV parameters did not change in the placebo group; they, instead, consistently increased in the atenolol, whereas a mild increase in SDNNi was only observed in the ivabradine group. A significant correlation was found in the atenolol group between HR and FMD changes (r=−0.48; p=0.04).

Conclusions Despite a comparable reduction in HR, atenolol, but not ivabradine, improved FMD in T2DM patients suggesting that changes in HR are by themselves unlikely to significantly improve endothelial function.

Statistics from

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.


Heart rate (HR) has emerged as a simple, but relevant risk factor for cardiovascular events in several clinical settings, including coronary artery disease and heart failure, and also in apparently healthy subjects.1 ,2 Importantly, the association of HR with cardiovascular mortality was found to be independent of traditional cardiovascular risk factors.3–5 The increase in HR has several negative effects on cardiac function, including enhanced myocardial work and oxygen consumption, as well as reduced diastolic time for blood perfusion.

Of note, data from experimental studies also suggest that HR elevation may favour the development of atherosclerotic lesions in arterial vessels due to the activation of cellular signalling events which lead to vascular oxidative stress, endothelial dysfunction and acceleration of atherogenesis.6

Whether reducing HR may have beneficial effects on vascular endothelial function remains controversial. HR reduction by β-blockers has given controversial results, with favourable effects being found in some studies,7 but not in others.8–11 β-Blockers, however, are not the ideal drugs to assess the effects of HR lowering on endothelial function due to their more complex effects on the cardiovascular system.

Ivabradine is a novel specific HR lowering antianginal agent, which acts through inhibition of the If current that modulates the spontaneous diastolic depolarisation in the sino-atrial node.12 In experimental studies, ivabradine was found to improve endothelium-dependent vasorelaxation in peripheral arteries,13 but its effects in patients with endothelial dysfunction are unknown.

Type 2 diabetes mellitus (T2DM) is known to be associated with an increased risk of cardiovascular events,14 and recent data suggest that, in this population, a higher resting HR is also associated with an increased risk of death and cardiovascular events.15 Of note, an impairment of endothelial function is also a frequent finding in T2DM patients.16 In this study, we aimed at investigating whether pure HR lowering by ivabradine was able to improve systemic vascular endothelial function in patients with T2DM.


Study population

We enrolled 66 consecutive patients with a diagnosis of T2DM, according to the American Diabetes Association guidelines, who presented at the Diabetic Care Unit of our University hospital for a follow-up clinical visit and fulfilled the following inclusion criteria: (1) no evidence of cardiovascular disease, according to clinical history, physical examination and standard ECG; (2) no evidence of any acute or chronic inflammatory disease; (3) no liver or renal dysfunction; (4) HR at rest >60 bpm; (5) systolic blood pressure >110 mm Hg; and (6) no previous or present consumption of β-blockers or ivabradine. The protocol was approved by the Ethics Committee of our Institution and all subjects gave written informed consent to participate in the study.

Study protocol

In this randomised double-blind, placebo-controlled trial, patients after basal assessment were randomised to receive one of three treatments for a period of 1 month: (1) atenolol, 25 mg twice daily (n=22); (2) ivabradine, 5 mg twice daily (n=22); or (3) placebo, 1 tablet twice daily (n=22).

The randomisation process was done following a computer-generated table of random numbers, with the last patients randomised in order to have the same number of patients in the three groups. Coauthors involved in the preparation of anonymous drug packages and in the randomisation process did not participate in the clinical assessment of patients. Cardiologists involved in the clinical and laboratory assessment of patients and/or analyses of data were blinded to treatment allocation.

After 1 week of treatment, patients underwent a clinical visit and a standard 12-lead ECG to assess compliance with treatment and potential side effects.

HR and blood pressure were measured at baseline and at 1-month follow-up. All other medications were left unchanged throughout the whole period of the study.

Systemic vasodilator function

Peripheral vasodilator function was assessed by the same expert operator using methods previously described in detail.17 ,18 Systemic endothelium-dependent vasodilation was assessed by measuring flow-mediated dilation (FMD). Briefly, subjects rested in the supine position for at least 10 min in a warm, quiet room (22–24°C) before testing. A 10-MHz multifrequency linear array probe attached to a high-resolution ultrasound machine was used to acquire images of the right brachial artery. Brachial artery diameter was measured throughout the whole test using a totally automated system19 that automatically identifies the internal edges of the vessel and tracks the walls of the artery via the brightness intensity of the walls versus the lumen of the vessel. The software provides a diameter measurement every second throughout the test. A mechanical support keeps the probe in a fixed position throughout the whole examination.

After baseline images of the brachial artery were obtained for 1 min, a forearm cuff, positioned 1 cm under the antecubital fossa, was inflated to 250 mm Hg and released after 5 min, with a consequent forearm reactive hyperaemia. The basal diameter was defined as the average of all measures collected during the first minute, and FMD was calculated as the maximum per cent change of the brachial artery diameter during hyperaemia compared with the basal diameter.

After recovery of brachial artery diameter to basal values, endothelium-independent vasodilation was assessed by measuring nitrate-mediated dilation (NMD). To this aim, 25 μg of sublingual glyceryl trinitrate was given and NMD was measured as the maximum per cent change of the brachial artery diameter compared with the basal diameter.

HR variability

Cardiac autonomic function was assessed by measuring HR variability (HRV) on 24-h 3-channel Holter ECG recordings.20 Time-domain HRV parameters included SDNN (SD of all RR intervals) and SDNNi (mean of the SD of RR intervals of all 5 min segments in the recording). Frequency-domain HRV parameters, derived from power spectrum analysis of RR intervals by fast Fourier transform, included amplitudes of RR oscillations in the range of low frequency (0.04–0.15 Hz) and high frequency (0.15–0.40 Hz).20

Statistical analyses

Preliminary data in T2DM patients indicated that we could expect an average basal FMD value of 5.0% with an SD of 1.5%. Assuming no significant effect on FMD of placebo, we calculated that we needed to enrol 17 patients per group to have a statistical power of 80% to show as significant (at p<0.05) an improvement of FMD of 20% with the active drugs. By considering possible drop-outs and adjustments for multiple comparisons (see below), we planned to enrol in the study 22 patients per group.

Kolmogorov–Smirnov test showed variable distributions not different from normal. Comparisons among groups of baseline continuous variables were done by one-way analysis of varianca (ANOVA). ANOVA for repeated measures was applied to compare the effects of treatments on the outcome variables. Multiple between-groups and within-group analyses were done by unpaired and paired t test, respectively, with Bonferroni correction. Categorical variables were compared by χ2 test or Fisher exact test, as indicated. Correlations were assessed by Pearson test. Data are reported as mean±SD or proportions. A two-tailed p value <0.05 was considered as statistically significant. Statistical analyses were performed by the SPSS V.17.0 statistical software (SPSS Italia, Inc., Florence, Italy).


General data

The main clinical characteristics of the three groups of patients are summarised in table 1. Five patients dropped out from the study (three in the atenolol group, two in the placebo group). Thus, 61 patients (20 in the placebo group, 19 in the atenolol group and 22 in the ivabradine group) constituted the final population of the study. There were no differences with regard to age, gender, cardiovascular risk factors and drug therapy. No significant changes in clinical features, as well as in main laboratory variables, were shown during the short term follow-up of the study (data not shown).

Table 1

Main clinical characteristics of patients enrolled in the study

Vascular function

At enrolment, basal brachial artery diameter and blood flow velocity, peak blood flow velocity, FMD and NMD were comparable among groups (table 2). No significant correlation was found between HR and FMD at baseline in the whole population (r=−0.14; p=0.27).

Table 2

Indexes of peripheral vasodilator function at baseline and at 1-month follow-up.

A comparable HR reduction was observed after 1 month of treatment in the atenolol group (from 87±13 to 69±9 bpm, p<0.01) and in the ivabradine group (from 86±12 to 71±9 bpm, p<0.01), whereas no significant changes in HR were observed in the placebo group (82±10 vs 81±9 bpm, p=0.26; p<0.001 for HR group interaction).

Basal brachial artery parameters did not change significantly at follow-up in any group. A significant difference in FMD changes was instead observed at 1 month among groups, which was related to a significant improvement in FMD in the atenolol group, but not in the other two groups (table 2 and figure 1). The significant difference in FMD changes among groups persisted even after adjustment for gender and changes in diastolic blood pressure (p<0.001).

Figure 1

Flow-mediated dilation (FMD) in the three groups of patients at baseline and after 4 weeks of treatment. p Values in the figure refer to among groups comparisons at the two assessments. *p<0.05 compared with baseline and compared with ivabradine and placebo at follow-up. This figure is only reproduced in colour in the online version.

No differences among groups or significant changes in any groups were found in NMD values (table 2 and figure 2).

Figure 2.

Nitrate-mediated dilation (NMD) in the three groups of patients at baseline and after 4 weeks of treatment. p Values in the figure refer to among groups comparisons at the two assessments. This figure is only reproduced in colour in the online version.

HR variability

HRV results are summarised in table 3. There were no differences at baseline among groups in any variables. Mean RR interval showed a similar increase in the atenolol and ivabradine groups, whereas remained unchanged in the placebo group. After treatment, no changes in HRV were observed in the placebo group, whereas a consistently significant improvement of HRV parameters was observed in the atenolol group. A mild increase in SDNNi, but not in other HRV parameters, was observed in ivabradine treated patients.

Table 3

HR variability (HRV) results at baseline and at 1-month follow-up in the three groups of patients.

In the atenolol group, a significant correlation was found between the per cent increase in RR interval during 24-h Holter monitoring at follow-up, compared with baseline, and the per cent change in FMD (r=−0.48; p=0.04; figure 3). However, no correlation could be demonstrated between the per cent changes in HRV parameters and the per cent changes in FMD (best r correlation coefficient −0.36 for SDDNi, p=0.13).

Figure 3

Correlation between the per cent increase in flow-mediated dilation (FMD) and in RR interval during 24-h Holter monitoring after 1 month of therapy, compared with baseline, in patients treated with atenolol. This figure is only reproduced in colour in the online version.


Our study shows that a β-blocker (atenolol) was able to improve systemic endothelial function after 1 month of therapy in patients with T2DM. In contrast, ivabradine, a pure HR lowering agent, exclusively acting on the sino-atrial node, did not exert any significant effect, in spite of HR reduction. Accordingly, our data indicate that HR reduction per se is unable to achieve a significant improvement of endothelium-mediated vasodilator function, at least in T2DM and after a short period of treatment.

Endothelial dysfunction is the earliest abnormality in the chain of events leading to atherosclerosis.21 Moreover, endothelial dysfunction has been shown to be an independent predictor of cardiovascular events in several clinical settings.22–24 Increased HR also is associated with a worse prognosis, both in patients with known coronary artery disease and in subjects without overt cardiovascular disease,1–5 including diabetic patients.13 It remains controversial, however, whether increased HR per se is responsible for the enhanced risk or cardiovascular events or it is rather a marker of prognosis, likely reflecting sympatho-vagal imbalance.2

Interestingly, a possible link between increased HR and a worse cardiovascular outcome might consist in a deleterious effect on vascular endothelial function. Indeed, previous experimental studies suggested that increased HR can by itself induce endothelial dysfunction and favour development and progression of vascular atherosclerotic lesions,6 perhaps by a chronic increase of shear stress on the arterial wall or even by modifying the quality of shear stress acting on endothelial cells. Persistently elevated shear stress can indeed result in uncoupling in endothelial cells of NO synthase activity, with increased production of oxygen reactive species,25 whereas modification of shear waveform and frequency can influence gene expression in endothelial cells,26 possibly leading to an increased expression of proinflammatory, proapoptotic and procoagulant transcripts,27 as well as a reduced eNOS expression.28

Whether HR reduction may improve endothelial function in the clinical setting has remained poorly known, due to the difficulty to achieve a lower HR without affecting other significant haemodynamic parameters, able to influence endothelial function. The recent availability of ivabradine, a pure bradycardic agent,10 can allow to selectively investigate the effects of a reduction of HR on vascular function. Ivabradine, indeed, acts by inhibiting the If current, which modulates spontaneous diastolic depolarisation in the sino-atrial node;10 thus, its effects on HR are not mediated by effects on autonomic nervous function, as instead it is with β-blockers. Of note, ivabradine has been reported to improve clinical outcome in patients with coronary artery disease and heart failure with higher HR,29 ,30 thus suggesting that a selective HR reduction has by itself favourable effects. In this study, we failed to find any significant effects of ivabradine on endothelial function, thus suggesting that the favourable clinical benefits of ivabradine are unlikely to include a significant improvement of vascular function.

In contrast, we found that atenolol improved impaired endothelial function in T2DM patients. Previous studies have given conflicting results about the effects of β-blockers on endothelial function.7–11 Atenolol, in particular, failed to improve FMD in T2DM patients in a previous study.8 The reasons for the different results between our study and others are not clear. However, there are several ways through which β-blockers may improve endothelial function. Modification of autonomic balance towards vagal predominance might favour vagal (acetylcholine) stimulation of endothelial cells, resulting in increased endothelial NO synthesis.21 In our study, a significant inverse correlation was indeed found between the changes in HR and FMD. The lack of significant correlation between HRV variables and FMD in our atenolol group may in fact question the role of sympatho-vagal changes in the effects of atenolol on endothelial function. The changes in HR, however, might correlate better with the systemic autonomic effects of β-blockade than the changes in HRV, which might more selectively reflect the β-blocking effects on sinus node activity. Furthermore, in previous studies atenolol was found to reduce subclinical inflammation31 and endothelial permeability to low density lipoproteins,32 suggesting that these mechanisms might contribute to improve endothelial function.

Interestingly, in our study we found a mild but significant improvement of SDNNi in patients treated with ivabradine, suggesting that this pure HR lowering agent can in some way also modulate the effects of nervous autonomic activity on sinus node, although this finding needs to be further assessed in other studies.

Limitations of the study

Some limitations of our study should be acknowledged. First, we used low drug daily doses (50 mg for atenolol and 10 mg for ivabradine). Thus, we cannot exclude that different results might be obtained with higher doses. Second, our results in T2DM patients cannot necessarily be extrapolated to other patient populations. Finally, our study was limited to 4 weeks of follow-up. Thus, we cannot exclude that different results might be obtained with longer periods of treatment.


Our study shows that atenolol, but not ivabradine, was able to improve FMD in T2DM patients. This finding points out that HR lowering is by itself unable to achieve significant beneficial effects on endothelium-mediated vascular function.



  • Contributors Study planning and responsible for the study: GAL; Selection of patients: DP, FZ; Conduction of the study: NR, RN AdF (FMD/NMD) and MM, AS (HRV); Randomisation and database management: PT, GR, FMS; Data analysis: GAL, AV; Manuscript writing: GAL; Manuscript review: GG, FC.

  • Funding This study was conducted with Research funds from the ‘Ministero Italiano dell'Università e della Ricerca’.

  • Competing interests None.

  • Patient consent Obtained.

  • Ethics approval Ethics Committee of Catholic University of Rome.

  • Provenance and peer review Not commissioned; externally peer reviewed.

  • Data sharing statement No additional data.