You are here

Article Text

Article menu

Download PDFPDF

Systematic review
‘Warm-up Angina’: harnessing the benefits of exercise and myocardial ischaemia
  1. Rupert P Williams1,2,
  2. Vasiliki Manou-Stathopoulou2,
  3. Simon R Redwood1,2,
  4. Michael S Marber1,2
  1. 1Department of Cardiology, King's College London British Heart Foundation Centre of Excellence, The Rayne Institute, St. Thomas’ Hospital Campus, London, UK
  2. 2Department of Cardiology, National Institute for Health Research Biomedical Research Centre at Guy's and St. Thomas’ NHS Foundation Trust, London, UK
  1. Correspondence to Dr Michael S Marber, Department of Cardiology, Cardiovascular Division, King's College London British Heart Foundation Centre of Excellence, The Rayne Institute, St. Thomas’ Hospital Campus, London SE1 7EH, UK; mike.marber{at}kcl.ac.uk

Abstract

The phenomenon of warm-up angina was first noted over 200 years ago. It describes the curious observation whereby exercise-induced ischaemia on second effort is significantly reduced or even abolished if separated from first effort by a brief rest period. However, the precise mechanism via which this cardio-protection occurs remains uncertain. Three possible explanations for reduced myocardial ischaemia on second effort include: first, an improvement in myocardial perfusion; second, increased myocardial resistance to ischaemia similar to ischaemic preconditioning; and third, reduced cardiac work through better ventricular–vascular coupling. Obtaining accurate coronary physiological measurements in the catheter laboratory throughout exercise demands a complex research protocol. In the 1980s, studies into warm-up angina relied on great cardiac vein thermo-dilution to estimate coronary blood flow. This technique has subsequently been shown to be inaccurate. However exercise physiology in the catheter laboratory has recently been resurrected with the advent of coronary artery wires that allow continuous measurement of distal coronary artery pressure and blood flow velocity. This review summarises the intriguing historical background to warm-up angina, and provides a concise critique of the important studies investigating mechanisms behind this captivating cardio-protective phenomenon.

  • Coronary Physiology
  • Coronary Artery Disease

https://doi.org/10.1136/heartjnl-2013-304187

Statistics from Altmetric.com

    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.

    Warm-up angina; a historical perspective

    Warm-up angina describes attenuation or abolition of angina on a second period of exertion when separated from the first period of exertion by a brief rest.1 ,2 There are interchangeable terms for warm-up angina including ‘first effort,’ or amusingly, ‘first hole’ angina, referring to the first hole on a golf course.

    Heberden was the first to describe a patient with potential warm-up angina in 1772: Opium taken at bed-time will prevent the attacks at night. I know one who set himself a task of sawing wood for half an hour every day, and was nearly cured.3

    Another compelling account was in a letter from an unknown gentleman to Heberden in 1785. He wrote: As well as I can recollect, it is about 5 or 6 years since, that I first felt the disorder which you treat of; it always attacked me when walking, and always after dinner. The first symptom is a pretty full pain in my left arm a little above the elbow; and in perhaps half a minute it spreads across the left side of my breast, and produces either a little faintness, or a thickness in my breathing; at least I imagined to, but the pain generally obliges me to stop. I have frequently, when in company, borne the pain, and continued my pace without indulging it; at which times it has lasted for five to perhaps ten minutes, and then gone off, as well as I can recollect, rather suddenly as it came.4

    Although the passage above provides a vivid description of angina and its adaptation, it better refers to the related phenomenon of walk-through angina.5

    Following these observations there was a period of confusion regarding the organ responsible for angina and even the symptoms it encompassed. Europeans included sudden pain in the cardiac area and anxiety associated with paroxysmal tachycardia within their definition. However, Wenkebach preferred the unambiguous English definition that described characteristic heavy sternal pressure on exertion, as opposed to “the pain of the neurotic patient; who with dramatic emphasis and sweeping gestures wants to make an impression upon his doctor, and points with the finger to the exact point of discomfort”.6

    Warm-up angina was first reviewed in the British Heart Journal in 1951,5 and again by us in 1994.7 The purpose of this review is to provide an update on this fascinating and ancient observation.

    Definition

    Warm-up angina is objectively defined by reduced ischaemia (ST-segment depression) or a raised ‘ischaemic threshold’ on second, compared to first, exertion (figure 1). A raised ‘ischaemic threshold’ is defined by an increase in the rate pressure product (RPP) at which a prestated amount of ST segment depression occurs.8 Rather than cardiac work (RPP), an ‘ischaemic threshold’ can also be defined by external work achieved (watts on cycle ergometer or METs/mins on treadmill).

    Figure 1
    Figure 1

    Warm-up angina. Demonstration of reduced ST segment depression and symptoms on second exercise despite increased treadmill speed (10 percent incline constant throughout study, transthoracic central back lead CB5 or CB6 used). Reproduced with kind permission from MacAlpin et al.1

    Clinical importance

    Warm-up angina is not simply a mere intellectual curiosity since it is also observed in patients taking standard anti-anginal medication.7 Moreover, Edwards et al demonstrated a reduction in ischaemic ventricular arrhythmias, chest discomfort and ST segment depression on second exertion (figure 2).9 ,10 This occurred independent of recruitment of collateral coronary vessels. In addition, Banning's group demonstrated protection against ischaemic left ventricular (LV) dysfunction on second exercise.11 Regional ejection fractions were calculated during and after supine bicycle ergometry, using equilibrium radionuclide imaging.

    Figure 2
    Figure 2

    The appearance of the surface ECG during repetitive exercise. Patients were subjected to three consecutive treadmill exercise tests. Test 2 started 15 min after test 1 had been completed, and test 3 was begun 90 min after the end of test 2. Panels A and B consist of serial ECG recordings taken at identical time points into each of the three exercise tests in two different patients. (A) ST segment depression occurs in tests 1 and 3, but is absent at the identical time point in test 2 despite a similar rate–pressure product. (B) The record of a patient who developed ventricular bigeminy during tests 1 and 3. During these tests, the complexes with a narrow QRS morphology represent normally conducted beats with 0.4 mV of ST segment depression. At the corresponding time point during test 2, ST segment depression is less pronounced (0.2 mV) and is no longer accompanied by ventricular bigeminy. Reproduced with kind permission from Edwards et al.10

    Currently there are no data to suggest warm-up angina confers a prognostic benefit, although in 1951 it was suggested that patients with warm-up angina may not need to be referred to a cardiologist!5 However, its potential to protect against malignant ventricular arrhythmias and ischaemic LV dysfunction is significant and may contribute to the lower mortality associated with myocardial infarction that is preceded by angina.12 Consequently, if the mechanism underlying warm-up angina could be understood and mimicked, an additional therapy that reduces morbidity and mortality in ischaemic heart disease (IHD) could result.

    More directly, warm-up angina studies have also shown that it is, after all, safe to exercise patients to the point of angina. A warm-up period of exercise may enable higher intensity exercise training to be carried out safely, which in turn may offer prognostic benefit.13 Although only 20% of patients1 describe warm-up symptoms, approximately 80% of coronary artery disease (CAD) patients2 ,14 ,15 have been shown to objectively demonstrate warm-up angina. Therefore there are a substantial proportion of CAD patients who may derive benefit from ‘warming up’.

    Aerobic exercise training in CAD patients has been shown to reduce cardiovascular mortality and improve quality of life in a Cochrane meta-analysis of 11 000 patients.16 Previous American guidelines recommended CAD patients to exercise at ‘low aerobic intensities’, namely, below their ischaemic threshold.17 This reflected concern from triggering of acute myocardial infarction with sudden physical exertion,18 and the suggestion of ischaemic exercise conferring an increased arrhythmia risk.19 Bogaty's group demonstrated that with a preceding warm-up period, exercise training beyond the ischaemic threshold can be safely performed without increased arrhythmic risk or myocardial injury.13 These safety findings have been recently confirmed in a larger trial.20

    The importance of precise exercise protocol design

    Exercise protocols investigating warm-up angina usually compare RPPs at equivalent and increasingly severe amounts of myocardial ischaemia (0.1, 0.15 or 0.2 mV ST segment depression). The RPP is used as an index of cardiac work, which in turn, is an index of myocardial oxygen consumption. These protocols therefore aim to examine myocardial oxygen consumption at the point of myocardial ischaemia. Counter-intuitively, warm-up angina demonstrates an objective reduction in myocardial ischaemia despite equivalent or greater myocardial oxygen consumption.

    The RPP was validated as an index of myocardial oxygen consumption in adventurous physiological studies carried out in the 1960s,21 which originally included left ventricular ejection time. Various workloads of differing exercise stresses (eg, upright cycle-ergometry, emotional stress, running up and down a flight of stairs) were performed while simultaneous central arterial pressures were measured with a 0.75 mm polyethylene catheter introduced through the brachial artery.

    In 1971, Redwood et al extended the validation work described above. They attempted to design a standardised and incremental upright cycle-ergometry protocol that could demonstrate reproducible and unequivocal RPPs at sequential anginal thresholds.22 Using their protocol with average workloads of 50–60 watts, they did just this. They concluded that their exercise protocol could be used to describe patients with ‘40 watt’ or ‘80 watt’ angina and therefore use this to grade severity, and monitor the response to therapeutic interventions. Of note, with this protocol there was no significant evidence of warm-up angina.

    However, these findings differed dramatically to a study by Quinn's group.2 They again used a standardised exercise protocol with two consecutive periods of exercise. Peak RPPs on second exercise were significantly higher than first exercise. Furthermore ECG tracings from both periods were examined in order to grade which ECG had greater maximal ST-segment depression. This was done in a blinded fashion with two independent observers. Twenty one of 22 patients had less ST segment depression on their second exercise period. This was the first study to objectively demonstrate warm-up angina, with the counter-intuitive reduction in ST segment depression despite the augmented RPP on second exercise.

    Why should there be such a discrepancy above? Quinn's group highlighted the significantly higher workloads used in their study; 166 watts and 112 watts for men and women respectively, versus an average of 50–60 watts in Redwood's study. The higher intensity of first exercise is the principle explanation and highlights the differential effects of small alterations in exercise protocol design. This is discussed in further detail below.

    Prerequisites for cardio-protection with warm-up angina

    Therefore what intensity of the first exercise is required to produce the warm-up effect? Kay et al demonstrated no warm-up on second exercise if this was preceded by a low-intensity subanginal or subischaemic first exercise period.23 However, exercising to ischaemia on first exercise, either rapidly or slowly; afforded equal cardio-protection. This was evidenced by a reduction in ST segment depression, increase in exercise duration and increase in RPP.

    Interestingly, Bogaty et al24 subsequently demonstrated that exercise of moderate intensity during the first effort still prolongs time to ischaemia and time to angina on second effort. However, in order to reduce maximal ST segment depression on second effort, the subject must exercise to ischaemia during the first effort. Thus it appears that myocardial ischaemia rather than exercise is the predominant stimulus required to initiate the maximal protective effect. The main difficulty with interpreting studies of this design is that it is not possible to separate exercise intensity from myocardial ischaemia, since ischaemia is only associated with the higher intensity exercise.

    The severity of CAD may also influence the degree of cardio-protection from warm-up angina. Again comparing Redwood et al and Jaffe et al, it was also noted that Redwood's group had more advanced CAD.2 This is significant as Tuomainen et al have since shown a more pronounced warm-up effect in patients with less severe CAD on coronary angiography.25 A caveat to this observation is that patients with cardiac syndrome X have been shown not to exhibit warm-up despite greater ST segment depression than the CAD patient cohort on first exercise.26 This may reflect differences in microvascular dysfunction.

    The interval between the first and second exercise periods is also very important. Tuomainen et al compared rest periods of 15, 30, 60 and 120 min in between first and second exercise periods.25 A significant increment in the ischaemic threshold on second exercise was demonstrated in 85%, 31%, 31% and 46% of each group respectively. These data reinforce those of Stewart et al who demonstrated greater increments in the ischaemic threshold and reductions in myocardial ischaemia, with a 10 min versus a 30 min intervening period of rest between exertions.27

    There also appears to be a spectrum of cardio-protection; which can be defined by the degree of reduction in ST segment depression or percent increment in the ischaemic threshold. Approximately 80% of CAD patients have a significant increment in their ischaemic threshold on second effort, which usually varies from 10% to 20%.24 Intriguingly however, this cardio-protective effect does not extend to a third exercise period.10 In addition, there seems to be significantly less protection against ischaemia in older (∼75 years) patients.28

    Mechanism

    Despite recognition of this cardio-protective phenomenon for over 200 years, the mechanisms that underlie warm-up angina remain uncertain. However three principal mechanisms have been proposed:

    1. An improvement in myocardial perfusion

    2. Increased myocardial resistance to ischaemia, similar to ischaemic preconditioning (IPC)

    3. A reduction in myocardial work

    An improvement in myocardial perfusion?

    In the 1930s, Gallavardin and Wayne et al believed that the principal mechanism behind warm-up angina was due to either coronary spasm on first exercise or enhanced coronary dilatation on second exercise.29

    Is increased coronary flow on second exercise plausible? Well, Bogaty et al did demonstrated a greater reduction in maximal ST segment depression with aminophylline,30 an arteriolar vasodilator. However, in this study, aminophylline was expected to block warm-up angina through its action as an adenosine antagonist!

    The resistance of the myocardial microvasculature is expected to be minimal at the point of ischaemia. Therefore it is hard to envisage a further reduction on second exercise. Nonetheless, there seems little doubt that vasodilator reserve still exists during ischaemia.31 Hence, the focus has shifted towards the potential mechanisms for increased myocardial blood flow on second effort, including:

    1. Collateral flow recruitment.1 ,5

    2. Augmentation of coronary blood flow (CBF).32

    3. Relative redistribution of myocardial perfusion towards the vulnerable subendocardial layer.30

    Each mechanism will be discussed in turn.

    Collateral flow recruitment1,5

    There is abundant evidence that well-developed collaterals can limit ischaemia during percutaneous coronary intervention (PCI)33 and acute myocardial infarction.34 The first work was by Rentrop's group who demonstrated significantly less angina and ST segment elevation in patients with a higher collateral grade, following vessel occlusion with an angioplasty balloon.33

    However, initial studies investigating the role of collateral recruitment as a mechanism of warm-up angina have shown contradictory results. Cribier et al demonstrated enhanced collateral angiographic grades in patients exhibiting warm-up angina.35 Kay et al observed a greater reduction in ischaemia on second exercise in patients with chronic total occlusions subtending collateral-dependent myocardium, compared to patients with isolated 70% coronary stenoses.23 In contrast, Ylitalo et al observed warm-up effects with no correlation to the degree of collateralisation at rest.36 However, as all of these studies employed observational methods rather than direct measurement of collateral flow, their conclusions can only be speculative.

    We subsequently provided strong evidence against a role for enhanced collateral recruitment on second exercise. 10 We found no increase in quantitatively assessed collateral flow between two consecutive, prolonged coronary artery balloon occlusions, in 33 patients with single vessel coronary disease exhibiting warm-up angina. Collateral flow was calculated as [Poccl−CSP]/[Pao−CSP37], where Poccl=distal coronary artery occlusion pressure, Pao=aortic pressure and CSP=coronary sinus pressure. While this study directly measured collateral flow, it did not do this simultaneously with exercise. Nonetheless warm-up angina on exertion was observed in patients with no recruitable collaterals despite a total of 6 min of balloon occlusion (2×3 min) during their subsequent PCI.

    Interestingly, Togni et al have recently demonstrated doubling of collateral flow during supine bicycle exercise, in 30 patients with non-occlusive coronary artery disease.38 They randomised patients to a protocol with either exercise or rest first to avoid the confounding effect of collateral recruitment by coronary artery balloon occlusions. Collateral flow was calculated as [Poccl−CVP]/[Pao−CVP], where CVP=central venous pressure.

    There is agreement that serial coronary artery balloon occlusions do not differentially affect collateral recruitment despite attenuation of ST deviation on second occlusion.10 ,38 Furthermore warm-up angina still occurs in patients without measurable collaterals despite prolonged epicardial coronary artery occlusion. Therefore while there is no doubt that collateral flow can be augmented on exercise, its contribution to warm-up angina is questionable and it is not a necessary prerequisite.

    Augmentation of CBF32

    Resistance beyond a culprit coronary artery stenosis should be near minimal at the time of angina. Hence it seems intuitively unlikely that CBF can be augmented on second exercise. In order to assess this mechanism, CBF needs to be directly measured during repeated exercise stress. It does not take a significant amount of imagination to envisage the complexity of such a catheter laboratory set-up!

    Hence initially, direct measurements of CBF were taken in response to pacing induced tachycardia. Williams et al recruited 11 patients with significant left anterior descending (LAD) stenoses.15 They used right atrial pacing to induce angina and measured great cardiac vein flow by thermo-dilution to estimate distal LAD flow during sequential peak heart rates. This study demonstrated no difference in CBF between first and second peak heart rate measurements, despite a reduction in angina score and ST-segment depression during the second bout of pacing.

    Okazaki et al subsequently measured CBF in 13 patients during a serial exercise protocol on a supine ergometer.14 Subjects had severe LAD stenoses and minimal evidence of collaterals. The investigators again determined CBF using great cardiac vein thermo-dilution. CBF was measured at rest, after 3 min of exertion, and at the anginal threshold. Similar to Williams et al, they demonstrated no increase in CBF between the first and second exercise periods, despite a significantly prolonged time to angina and lesser ST segment depression on second exercise.14

    However, these studies are not conclusive. Great cardiac vein thermo-dilution has been shown to be inaccurate and highly dependent on the position of the catheter,39 and also only approximates myocardial blood flow beyond the LAD stenosis. Alternative routes of venous drainage through the thebesian circulation also confound this technique. Moreover, both studies were very small and had to employ additional exclusions for technical reasons.15 In addition, prior to inclusion into the study, neither study screened patients with treadmill tests to see if they exhibited a reproducible warm-up effect; therefore the investigators decreased their chances of discerning differences in CBF between exercise periods.

    Importantly, the above studies did quantify myocardial oxygen extraction during pacing and exercise stress, using the Fick principle, and neither demonstrated a significant difference in oxygen extraction between first and second exercise periods. This is in agreement with previous studies showing a minimal change in oxygen extraction with varying intensity of exercise,40 and reflects the high level of basal oxygen extraction at rest within heart muscle. Therefore it is very unlikely that oxygen extraction can be preferentially increased on second exercise to explain the warm-up effect.

    We32 have recently published data demonstrating a significant increase in CBF on second exercise in 16 CAD patients. We used a ‘combowire’ through radial arterial access. This enabled simultaneous measurement of pressure and flow beyond a coronary artery stenosis during supine exercise. The demands of this protocol are highlighted by the high proportion of excluded patients (11/27). Distal flow and pressure were measured by a Doppler probe near the tip of the coronary artery wire.

    Contrary to expectation, we32 demonstrated a significant increase in CBF velocity (27 cm/s vs 22 cm/s) on second exercise (figure 3). This was associated with significantly lower microvascular resistance on second exercise. Interestingly, microvascular resistance fell throughout first exercise and remained low during second exercise. These data do suggest progressive microvascular dilatation that is sustained during second exercise.

    Figure 3
    Figure 3

    The alterations in pressure and flow velocity occurring during serial exercise. The left panels were recorded at baseline immediately before first (Ex1, upper traces) and second (Ex2, lower traces) exertion. The right panels were recorded at peak equivalent workload during Ex1 and Ex2. Pa indicates proximal, or aortic pressure; Pd, distal coronary pressure; and U, coronary flow velocity recorded from the distal coronary artery. The mean flow velocity is higher during Ex2 (27 vs 22 cm/s, p<0.03), causing a greater mean pressure gradient across the coronary stenosis, δP (Pa–Pd; 18 vs 11, p<0.02 mm Hg), with a resultant reduction in micro-vascular resistance (3.6 vs 4.6 mm  Hg×cm−1× s−1, p=0.01) that occurs on second effort. Reproduced with kind permission from Lockie et al.32

    Relative redistribution of myocardial perfusion towards the vulnerable subendocardial layer30

    In 12 subjects with previous positive exercise tolerance tests and prior demonstration of an exercise-induced single photon emission computerized tomography (SPECT) perfusion deficit, Bogaty et al observed no improvement in SPECT perfusion on second versus first exercise effort, despite an observed improvement in the ischaemic threshold on second exercise.30 Again patients were not screened prior to the study to ensure substantive warm-up effects. Moreover, this study only assessed pan-mural myocardial perfusion. This is a key criticism, as theoretically the warm-up effect could be achieved through a relative redistribution of blood from the epicardium to the endocardium in the second period of exercise. This hypothesis is derived from basic principles of coronary physiology; in the absence of a coronary artery stenosis, flow distribution across the myocardium is relatively uniform at rest through local autoregulation. Once the vasodilatory reserve of these resistance vessels is exhausted, ischaemia ensues.

    During exercise, the contractile forces within the heart have a disproportionate effect on the subendocardial layer, reducing coronary perfusion pressure (mean coronary artery pressure—minimal LV pressure) and vasodilator reserve. Consequently this layer becomes ischaemic first.41 Subepicardial perfusion in contrast is generally unaffected by such contractile forces.41 Cardiac Magnetic Resonance imaging perfusion scans now provide sufficient spatial resolution to distinguish between subepicardium and subendocardium, and may therefore provide a possible means to study this hypothesis.

    Increased myocardial resistance to ischaemia, similar to IPC?

    When warm-up angina was last reviewed in Heart,7 nearly 20 years ago, the authors’ conclusions were that the mechanism underlying warm-up angina likely reflected an enhanced endogenous resistance to ischaemia on second exercise. IPC describes the phenomenon whereby repetitive brief periods of ischaemia protect against a subsequent period of lethal ischaemia. IPC was first demonstrated in dogs with an LAD occlusion model, whereby IPC afforded a 40% reduction in infarct size.42

    IPC, like warm-up angina, is not explained by a downregulation of contractile function induced by initial exercise,43 and the time course for protection with warm-up is similar to that of IPC.7 Bradykinin and adenosine (acting on selective A1 receptors) are able to trigger preconditioning of the myocardium in the absence of an ischaemic insult.44 Subsequently, IPC exerts end-effects through opening of cardiomyocyte mitochondrial KATP channels,45 preventing opening of the mitochondrial permeability transition pore.46 The role for these endogenous substances in warm-up angina is not as clear.

    Adenosine: Although traditional views supported the hypothesis of a build-up of adenosine within myocardium as a trigger for warm-up angina,7 the four studies investigating its role are universally negative.30 ,47–49

    Initial studies were confounded by the unavailability of a selective A1 adenosine antagonist.30 ,47 ,48 A2 adenosine antagonists cause coronary artery vasodilatation and therefore partial blockade of this receptor could clearly confound results. Of note, Bogaty et al actually noted a greater reduction in maximal ST segment depression that may reflect inadvertent A2 adenosine antagonist effects.30 However, a specific A1 adenosine antagonist has also failed to block warm-up angina.49

    Bradykinin: Angiotensin-converting enzyme (ACE)-inhibitors exert part of their action through inhibition of bradykinin breakdown. Hypothetically, buildup of bradykinin may trigger warm-up angina. The addition of an ACE-inhibitor to long-term isosorbide dinitrate on exercise-induced ischaemia has been assessed in CAD patients. Maximal ST-segment depression was reduced and time to ST segment depression was increased with an ACE-inhibitor compared to placebo.50 Unfortunately a subsequent larger study showed no differences in the same measures.51 We also studied the effect of ACE-inhibitors in the context of a warm-up angina protocol.52 ACE-inhibition did not significantly potentiate the reduction in ischaemia or increment in ischaemic threshold seen on second exercise.

    KATP Channels: From a mechanistic perspective, the role of KATP channels has attracted the most interest. Table 1 below summarises the relevant studies. All studies have used the ischaemic threshold as their primary endpoint apart from Correa et al.53 From an efficacy perspective it is easier to assess the effect of KATP channel blockade; patients with warm-up angina increase their ischaemic threshold on second exercise. Thus successful blockade will prevent such an increment. Studies investigating the effect of a KATP channel opener are more challenging to interpret; should we see a greater increment in the ischaemic threshold on second exercise, or perhaps a significantly higher ischaemic threshold on first exercise?

    View this table:
    Table 1

    Protocols for studies investigating the role of KATP channels in warm-up angina

    With the exception of the study by Correa et al, the effects of KATP channel blockade are fairly straightforward to digest in all other studies.53 Glibenclamide, given either long-term or just before exercise, appears to completely abolish the increment in the ischaemic threshold on second exercise.54–58 This applies in diabetics and non-diabetics.54–56 A significant reduction in maximal ST-segment depression on second exercise was demonstrated in four control groups.43 ,55 ,57 ,58 This protection was abolished with glibenclamide.43 ,55 ,57 ,58 It is difficult to assess the effects of glibenclamide in Ferreira et al, as they failed to show a significant reduction in maximal ST-segment depression on second exercise in their control group.56

    Why are the results of Bogaty et al43 and Correa et al53 in disagreement with the above? Well, neither study directly compared ischaemic thresholds in control and diabetic groups. Therefore we cannot assess whether glibenclamide may have abolished increments in this parameter on second exercise. Moreover, the intervals between first and second exercise periods were not constant in either study, and as a result, the RPP did not return to baseline in the glibenclamide arm of Bogaty et al.43 In addition, Correa et al started the exercise protocols 180 min after glibenclamide administration.53 Glibenclamide serum levels peak 60–120 min after administration.54 Hence it is possible that Correa et al did not adequately block KATPchannels, given their patients were glibenclamide naïve.53

    Glibenclamide has also been shown to abolish protection afforded from IPC.46 Glibenclamide and tolbutamide have the highest selectivity for KATP channels within cardiac myocytes and therefore may cause adverse cardiovascular effects in patients with type 2 diabetes and IHD. A Danish study with 100 000 diabetic patients highlighted the increased cardiovascular risk with these medications.59 However, these medications continue to be prescribed in diabetic patients, which is a serious concern. In contrast, Gliclazide did not significantly affect warm-up angina in Bilinska et al,58 which is reassuring and is consistent with its higher selectivity for pancreatic KATP channels.59

    With regard to the effects of the KATP channel openers pinacidil and nicorandil, the results were slightly disappointing. Nevertheless, we demonstrated a tendency to a higher ischaemic threshold on second effort with nicorandil versus placebo, 12% versus 7% respectively, although this did not reach statistical significance.52 Furthermore, interpretation is complicated by the antianginal effect of nicorandil on first exercise. Intriguingly, Lindhardt et al observed a reduction in the ischaemic threshold on second exercise with pinacidil.57 They concluded that this may be a direct pharmacological effect of pinacidil on the ST segment.57 Pinacidil and nicorandil block both sarcolemmal and mitochondrial KATP channels,52 ,57 but it is possible that nicorandil is more specific for the latter.

    In conclusion, while there are notable discrepancies between warm-up angina and IPC, there is a suggestion of activation of KATP channels following a period of exercise-induced ischaemia. Although this may not be the sole mechanism underlying warm-up angina, blockade of KATP channels appears to significantly affect the ability to afford cardio-protection.

    A reduction in myocardial work?

    Studies demonstrating an enhanced degree of external work,7 measured in metabolic time equivalents, on second compared with first exercise exclude less external work as an explanation for warm-up. Similarly increments in RPPs on second exercise suggest higher myocardial oxygen consumption and enhanced global myocardial work on second exercise.

    However, the RPP does not incorporate the relative contributions of contractility and timing of systole in determining MVO2.31 Thus, it remains plausible that changes in unmeasured determinants of cardiac workload on second exercise may explain the warm-up effect. These could occur through a reduction in the central arterial pressure waveform, hence diminishing afterload, an effect well documented in response to glyceryl trinitrate (GTN).

    GTN has previously been demonstrated to reduce augmentation in the central and peripheral pulse waveforms, and also the ‘peripheral augmentation index (AI)’; which approximates the ratio of central to peripheral pulse pressure (PP).60 A reduction in afterload by GTN may in turn enhance ventricular–vascular coupling to produce a more efficient systole; prolonging the time fraction of diastole and reducing ventricular work.

    Interestingly, Munir et al demonstrated similar reductions in systolic augmentation indices with exercise to that which occur with GTN (in healthy volunteers). These effects persisted after the heart rate returned to baseline.60 Therefore exercise may also provide a sufficient stimulus to invoke a similar beneficial reduction in ventricular work. However, this mechanism also seems counter-intuitive, as both GTN and exercise cause a reduction in diastolic pressure and therefore intuitively should decrease myocardial perfusion.

    Lockie et al have recently clarified these diametrically opposed effects on myocardial work and supply.32 We used a high fidelity pressure wire positioned in the aortic root, which enabled accurate aortic pressure waveforms to be analysed during exercise, alongside the ‘combowire’ inserted into the distal coronary artery. The AI, a measure of central systolic blood pressure augmentation thought to arise from pressure-wave reflection, was calculated as the difference between the second (P2) and first (P1) peaks, expressed as a percent of the PP (figure 4). We also calculated the tension time index (TTI) and diastolic time index (DTI) in all subjects (see figure 4). These indices relate to myocardial oxygen demand and coronary perfusion, respectively.

    Figure 4
    Figure 4

    (A) Typical pressure waveform at rest recorded from the ascending aorta in a healthy middle-aged man. Two systolic peaks are labelled P1 and P2. The area under the curve (AUC) during systole is the tension time index , and AUC during diastole is diastolic time index. TR is defined as the time between the foot of the wave (TF) and the inflection point (Pi). (B) Example of an aortic pressure trace from one of the subjects taken at peak equivalent workload during each exercise period, demonstrating the striking change in wave morphology between first (Ex1) and second (Ex2) exercise, with a reduction in the overall amplitude of the wave and specifically a marked reduction in pressure augmentation. Reproduced with kind permission from Lockie et al.

    We demonstrated a significant 33% reduction in AI in all subjects on second exercise (see figure 4). Interestingly on an individual basis, a greater increment in the ischaemic threshold was associated with a larger percent reduction in AI. Both TTI and DTI were also significantly lower during second exercise. However, despite the suggestion of a lower coronary perfusion pressure on second exercise, we actually demonstrated augmented CBF velocity secondary to reduced microvascular resistance (defined as mean distal coronary artery pressure/mean CBF velocity).32 We also demonstrated a reduction in left ventricular ejection time on second exercise, after accounting for heart rate.32

    The reduction in AI can be attributed to vasodilatation of systemic muscular arteries, rather than changes in arterial stiffness.60 These more favourable haemodynamic conditions during second exercise result in reduced afterload and shorten systole; through more efficient ventricular–vascular coupling. In turn, this results in enhanced diastolic relaxation.61 Our study was the first demonstration of exercise-induced peripheral vasodilatation as an important mechanism underlying warm-up angina.

    Furthermore to provide mechanistic insight, we applied the technique of ‘wave-intensity analysis’ (WIA) to tie in the observed changes in CBF velocities and central aortic pressure waveforms.62 Central to WIA is the understanding that phasic changes in coronary artery pressures and flows can be explained by a series of wave fronts. Quantification of these wave fronts can assess the relative importance of aortic and microcirculatory contributions to CBF.62

    The degree of diastolic left ventricular relaxation greatly influences the backward expansion wave in the coronary artery.62 We demonstrated a significant 21% increase in the energy of the backward-travelling expansion wave throughout second exercise.32 Therefore there is further evidence of enhanced diastolic relaxation, which actively increased CBF.

    It is difficult to determine which of the haemodynamic mechanisms above is the initial cause and which the consequence, but these data suggest evidence of positive cardiac–coronary interaction following a period of exercise-induced ischaemia. We concluded that the initial stimulus is likely to be progressive coronary and systemic vasodilatation, which through enhanced ventricular–vascular coupling and cardiac–coronary interaction, enhance diastolic relaxation, reduce microvascular resistance, and hence, augment CBF.

    Conclusion

    Warm-up angina describes a fascinating cardio-protective phenomenon first described over 200 years ago.3 Understanding the mechanism behind this phenomenon has presented a huge challenge, given the complexity of the systemic and coronary haemodynamic changes that occur on exercise, and the difficulty of simultaneous recordings of flow and pressure.

    Recent evidence suggests that periods of exercise are capable of triggering systemic arterial vasodilatation to the point at which left ventricular afterload is significantly reduced, both in healthy volunteers60 and in CAD patients.32

    During second compared to first effort, we have also noted a reduction in coronary microvascular resistance and an increase in CBF, despite a reduction in diastolic perfusion pressure. Is exercise itself a sufficient stimulus to cause this? Bogaty et al and Kay et al demonstrate that this is not the case, and suggest warm-up angina requires a preceding period of ischaemia.23 ,24 Ischaemia may activate contributory myocardial mechanisms possibly through the opening of KATP channels. We therefore postulate that warm-up angina harnesses the beneficial cardio-protective effects of both exercise and ischaemia, and in both the systemic and coronary circulations.

    How far have we come in 200 years? We now know warm-up angina is due to the attenuation of myocardial ischaemia on second effort. However, we are no closer to identifying a single underlying mechanism and harnessing it for patients’ benefit. Just carry on exercising!

    Acknowledgments

    This study was funded by a British Heart Foundation Clinical Training Fellowship awarded to Dr Williams (FS/11/90/29087) and by the National Institute for Health Research (NIHR) Biomedical Research Centre based at Guy's and St Thomas’ National Health Service Foundation Trust and King's College London.

    References

    1. ↵
      1. MacAlpin RN,
      2. Kattus AA
      . Adaptation to exercise in angina pectoris: the electrocardiogram during treadmill walking and coronary angiographic findings. Circulation 1966;33:183–201.
      OpenUrlAbstract/FREE Full Text
    2. ↵
      1. Jaffe MD,
      2. Quinn NK
      . Warm-up phenomenon in angina pectoris. 1980;2:934–6.
    3. ↵
      1. Heberden W
      . Some account of a disorder of the breast. Med Trans, R Coll Physicians London 1772;2:59–67.
      OpenUrl
    4. ↵
      1. Heberden W
      . Letter to Heberden. Med Trans, R Coll Physicians London 1785;1:3.
      OpenUrl
    5. ↵
      1. Price RJLR
      . First effort angina. Second wind in angina pectoris. Brit Heart J 1951;13:197.
      OpenUrlFREE Full Text
    6. ↵
      1. Wenkebach KF
      . Angina pectoris and the possiblilities of its surgical relief. Brit Med J 1924;1:809.
      OpenUrlFREE Full Text
    7. ↵
      1. Marber MS,
      2. Joy MD,
      3. Yellon DM
      . Is warm-up in angina ischaemic preconditioning? Br Heart J 1994;72:213–15.
      OpenUrlFREE Full Text
    8. ↵
      1. Maybaum S,
      2. Ilan M,
      3. Mogilevsky J,
      4. et al
      . Improvement in ischemic parameters during repeated exercise testing: a possible model for myocardial preconditioning. Am J Cardiol 1996;78:1087–91.
      OpenUrlPubMedWeb of Science
    9. ↵
      1. Tuomainen P,
      2. Hartikainen J,
      3. Vanninen E,
      4. et al
      . Warm-up phenomenon and cardiac autonomic control in patients with coronary artery disease. Life Sci 2005;76:2147–58.
      OpenUrlCrossRefPubMedWeb of Science
    10. ↵
      1. Edwards RJ,
      2. Redwood SR,
      3. Lambiase PD,
      4. et al
      . Antiarrhythmic and anti-ischaemic effects of angina in patients with and without coronary collaterals. Heart 2002;88:604–10.
      OpenUrlAbstract/FREE Full Text
    11. ↵
      1. Kelion AD,
      2. Webb TP,
      3. Gardner MA,
      4. et al
      . The warm-up effect protects against ischemic left ventricular dysfunction in patients with angina. J Am Coll Cardiol 2001;37:705–10.
      OpenUrlCrossRefPubMedWeb of Science
    12. ↵
      1. Kloner RA,
      2. Jennings RB
      . Consequences of brief ischemia: stunning, preconditioning, and their clinical implications: part 2. Circulation 2001;104:3158–67.
      OpenUrlAbstract/FREE Full Text
    13. ↵
      1. Noel M,
      2. Jobin J,
      3. Marcoux A,
      4. et al
      . Can prolonged exercise-induced myocardial ischaemia be innocuous? Eur Heart J 2007;28:1559–65.
      OpenUrlAbstract/FREE Full Text
    14. ↵
      1. Okazaki Y,
      2. Kodama K,
      3. Sato H,
      4. et al
      . Attenuation of increased regional myocardial oxygen consumption during exercise as a major cause of warm-up phenomenon. J Am Coll Cardiol 1993;21:1597–604.
      OpenUrlCrossRefPubMedWeb of Science
    15. ↵
      1. Williams DO,
      2. Bass TA,
      3. Gewirtz H,
      4. et al
      . Adaptation to the stress of tachycardia in patients with coronary artery disease: insight into the mechanism of the warm-up phenomenon. Circulation 1985;71:687–92.
      OpenUrlAbstract/FREE Full Text
    16. ↵
      1. Heran BS,
      2. Chen JM,
      3. Ebrahim S,
      4. et al
      . Exercise-based cardiac rehabilitation for coronary heart disease. Cochrane Database Syst Rev 2011;(7):CD001800.
    17. ↵
      1. Fletcher GF,
      2. Balady GJ,
      3. Amsterdam EA,
      4. et al
      . Exercise standards for testing and training: a statement for healthcare professionals from the American Heart Association. Circulation 2001;104:1694–740.
      OpenUrlFREE Full Text
    18. ↵
      1. Mittleman MA,
      2. Maclure M,
      3. Tofler GH,
      4. et al
      . Triggering of acute myocardial infarction by heavy physical exertion. Protection against triggering by regular exertion. Determinants of Myocardial Infarction Onset Study Investigators. N Engl J Med 1993;329:1677–83.
      OpenUrlCrossRefPubMedWeb of Science
    19. ↵
      1. Hoberg E,
      2. Schuler G,
      3. Kunze B,
      4. et al
      . Silent myocardial ischemia as a potential link between lack of premonitoring symptoms and increased risk of cardiac arrest during physical stress. Am J Cardiol 1990;65:583–9.
      OpenUrlCrossRefPubMedWeb of Science
    20. ↵
      1. Rognmo O,
      2. Moholdt T,
      3. Bakken H,
      4. et al
      . Cardiovascular risk of high- versus moderate-intensity aerobic exercise in coronary heart disease patients. Circulation 2012;126:1436–40.
      OpenUrlAbstract/FREE Full Text
    21. ↵
      1. Robinson BF
      . Relation of heart rate and systolic blood pressure to the onset of pain in angina pectoris. Circulation 1967;35:1073–83.
      OpenUrlAbstract/FREE Full Text
    22. ↵
      1. Redwood DR,
      2. Rosing DR,
      3. Goldstein RE,
      4. et al
      . Importance of the design of an exercise protocol in the evaluation of patients with angina pectoris. Circulation 1971;43:618–28.
      OpenUrlAbstract/FREE Full Text
    23. ↵
      1. Kay P,
      2. Kittelson J,
      3. Stewart RA
      . Relation between duration and intensity of first exercise and “warm up” in ischaemic heart disease. Heart 2000;83:17–21.
      OpenUrlAbstract/FREE Full Text
    24. ↵
      1. Bogaty P,
      2. Poirier P,
      3. Boyer L,
      4. et al
      . What induces the warm-up ischemia/angina phenomenon: exercise or myocardial ischemia? Circulation 2003;107:1858–63.
      OpenUrlAbstract/FREE Full Text
    25. ↵
      1. Tuomainen P,
      2. Vanninen E,
      3. Halonen P,
      4. et al
      . Characterization of the warm-up phenomenon in patients with coronary artery disease. Am Heart J 2002;144:870–6.
      OpenUrlCrossRefPubMedWeb of Science
    26. ↵
      1. Lupi A,
      2. Lanza GA,
      3. Lucente M,
      4. et al
      . The “warm-up” phenomenon occurs in patients with chronic stable angina but not in patients with syndrome X. Am J Cardiol 1998;81:123–127.
      OpenUrlCrossRefPubMed
    27. ↵
      1. Stewart RA,
      2. Simmonds MB,
      3. Williams MJ
      . Time course of “warm-up” in stable angina. Am J Cardiol 1995;76:70–3.
      OpenUrlCrossRefPubMedWeb of Science
    28. ↵
      1. Longobardi G,
      2. Abete P,
      3. Ferrara N,
      4. et al
      . “Warm-up” phenomenon in adult and elderly patients with coronary artery disease: further evidence of the loss of “ischemic preconditioning” in the aging heart. J Gerontol A Biol Sci Med Sci 2000;55:M124–9.
      OpenUrlAbstract/FREE Full Text
    29. ↵
      1. Wayne EJ,
      2. Graybiel A
      . Observations on the effect of food, gastric distension, external temperature, and repeated exercise on angina of effort, with a note on angina sine dolore. Clin Sci 1934;1:287–304.
      OpenUrl
    30. ↵
      1. Bogaty P,
      2. Kingma JG,
      3. Guimond J,
      4. et al
      . Myocardial perfusion imaging findings and the role of adenosine in the warm-up angina phenomenon. J Am Coll Cardiol 2001;37:463–9.
      OpenUrlCrossRefPubMedWeb of Science
    31. ↵
      1. Duncker DJ,
      2. Bache RJ
      . Regulation of coronary blood flow during exercise. Physiol Rev 2008;88:1009–86.
      OpenUrlAbstract/FREE Full Text
    32. ↵
      1. Lockie TP,
      2. Rolandi MC,
      3. Guilcher A,
      4. et al
      . Synergistic adaptations to exercise in the systemic and coronary circulations that underlie the warm-up angina phenomenon. Circulation 2012;126:2565–74.
      OpenUrlAbstract/FREE Full Text
    33. ↵
      1. Cohen M,
      2. Rentrop KP
      . Limitation of myocardial ischemia by collateral circulation during sudden controlled coronary artery occlusion in human subjects: a prospective study. Circulation 1986;74:469–76.
      OpenUrlAbstract/FREE Full Text
    34. ↵
      1. Charney R,
      2. Cohen M
      . The role of the coronary collateral circulation in limiting myocardial ischemia and infarct size. Am Heart J 1993;126:937–45.
      OpenUrlCrossRefPubMedWeb of Science
    35. ↵
      1. Cribier A,
      2. Korsatz L,
      3. Koning R,
      4. et al
      . Improved myocardial ischemic response and enhanced collateral circulation with long repetitive coronary occlusion during angioplasty: a prospective study. J Am Coll Cardiol 1992;20:578–86.
      OpenUrlCrossRefPubMedWeb of Science
    36. ↵
      1. Ylitalo K,
      2. Jama L,
      3. Raatikainen P,
      4. et al
      . Adaptation to myocardial ischemia during repeated dynamic exercise in relation to findings at cardiac catheterization. Am Heart J 1996;131:689–97.
      OpenUrlCrossRefPubMedWeb of Science
    37. ↵
      1. Pijls NH,
      2. van Son JA,
      3. Kirkeeide RL,
      4. et al
      . Experimental basis of determining maximum coronary, myocardial, and collateral blood flow by pressure measurements for assessing functional stenosis severity before and after percutaneous transluminal coronary angioplasty. Circulation 1993;87:1354–67.
      OpenUrlAbstract/FREE Full Text
    38. ↵
      1. Togni M,
      2. Gloekler S,
      3. Meier P,
      4. et al
      . Instantaneous coronary collateral function during supine bicycle exercise. Eur Heart J 2010;31:2148–55.
      OpenUrlAbstract/FREE Full Text
    39. ↵
      1. Bagger JP
      . Coronary sinus blood flow determination by the thermodilution technique: influence of catheter position and respiration. Cardiovasc Res 1985;19:27–31.
      OpenUrlAbstract/FREE Full Text
    40. ↵
      1. Grubbstrom J,
      2. Berglund B,
      3. Kaijser L
      . Myocardial blood flow and lactate metabolism at rest and during exercise with reduced arterial oxygen content. Acta Physiol Scand 1991;142:467–74.
      OpenUrlCrossRefPubMedWeb of Science
    41. ↵
      1. Fokkema DS,
      2. VanTeeffelen JW,
      3. Dekker S,
      4. et al
      . Diastolic time fraction as a determinant of subendocardial perfusion. Am J Physiol Heart Circ Physiol 2005;288:H2450–6.
      OpenUrlAbstract/FREE Full Text
    42. ↵
      1. Murry CE,
      2. Jennings RB,
      3. Reimer KA
      . Preconditioning with ischemia: a delay of lethal cell injury in ischemic myocardium. Circulation 1986;74:1124–36.
      OpenUrlAbstract/FREE Full Text
    43. ↵
      1. Bogaty P,
      2. Kingma JGJ,
      3. Robitaille NM,
      4. et al
      . Attenuation of myocardial ischemia with repeated exercise in subjects with chronic stable angina: relation to myocardial contractility, intensity of exercise and the adenosine triphosphate-sensitive potassium channel. J Am Coll Cardiol 1998;32:1665–71.
      OpenUrlCrossRefPubMedWeb of Science
    44. ↵
      1. Downey JM,
      2. Cohen MV
      . Signal transduction in ischemic preconditioning. Adv Exp Med Biol 1997;430:39–55.
      OpenUrlCrossRefPubMedWeb of Science
    45. ↵
      1. Sato T,
      2. O'Rourke B,
      3. Marban E
      . Modulation of mitochondrial ATP-dependent K+ channels by protein kinase C. Circ Res 1998;83:110–4.
      OpenUrlAbstract/FREE Full Text
    46. ↵
      1. Yellon DM,
      2. Hausenloy DJ
      . Myocardial reperfusion injury. N Engl J Med 2007;357:1121–35.
      OpenUrlCrossRefPubMedWeb of Science
    47. ↵
      1. Tomai F,
      2. Crea F,
      3. Danesi A,
      4. et al
      . Effects of A1 adenosine receptor blockade on the warm-up phenomenon. Cardiologia 1997;42:385–92.
      OpenUrlPubMed
    48. ↵
      1. Kerensky RA,
      2. Franco E,
      3. Schlaifer JD,
      4. et al
      . Effect of theophylline on the warm-up phenomenon. Am J Cardiol 1999;84:1077–80, A9.
      OpenUrlCrossRefPubMedWeb of Science
    49. ↵
      1. Kelion AD,
      2. Webb TP,
      3. Gardner MA,
      4. et al
      . Does a selective adenosine A(1) receptor agonist protect against exercise induced ischaemia in patients with coronary artery disease? Heart 2002;87:115–20.
      OpenUrlAbstract/FREE Full Text
    50. ↵
      1. Murohara T,
      2. Tayama S,
      3. Tabuchi T,
      4. et al
      . Effects of angiotensin-converting enzyme inhibitor alacepril in patients with stable effort angina during chronic isosorbide dinitrate treatment. Am J Cardiol 1996;77:1159–63.
      OpenUrlCrossRefPubMedWeb of Science
    51. ↵
      1. Pepine CJ,
      2. Rouleau JL,
      3. Annis K,
      4. et al
      . Effects of angiotensin-converting enzyme inhibition on transient ischemia: the Quinapril Anti-Ischemia and Symptoms of Angina Reduction (QUASAR) trial. J Am Coll Cardiol 2003; 42:2049–59.
      OpenUrlCrossRefPubMed
    52. ↵
      1. Edwards RJ,
      2. Redwood SR,
      3. Lambiase PD,
      4. et al
      . The effect of an angiotensin-converting enzyme inhibitor and a K+(ATP) channel opener on warm up angina. Eur Heart J 2005;26:598–606.
      OpenUrlAbstract/FREE Full Text
    53. ↵
      1. Correa SD,
      2. Schaefer S
      . Blockade of K(ATP) channels with glibenclamide does not abolish preconditioning during demand ischemia. Am J Cardiol 1997; 79:75–8.
      OpenUrlCrossRefPubMedWeb of Science
    54. ↵
      1. Tomai F,
      2. Danesi A,
      3. Ghini AS,
      4. et al
      . Effects of K(ATP) channel blockade by glibenclamide on the warm-up phenomenon. Eur Heart J 1999;20:196–202.
      OpenUrlAbstract/FREE Full Text
    55. ↵
      1. Ovunc K
      . Effects of glibenclamide, a K(ATP) channel blocker, on warm-up phenomenon in type II diabetic patients with chronic stable angina pectoris. Clin Cardiol 2000;23:535–9.
      OpenUrlCrossRefPubMedWeb of Science
    56. ↵
      1. Ferreira BM,
      2. Moffa PJ,
      3. Falcao A,
      4. et al
      . The effects of glibenclamide, a K(ATP) channel blocker, on the warm-up phenomenon. Ann Noninvasive Electrocardiol 2005;10:356–62.
      OpenUrlCrossRefPubMedWeb of Science
    57. ↵
      1. Lindhardt TB,
      2. Abedini S,
      3. Olesen RM,
      4. et al
      . Effects of pharmacological modulation of the ATP-sensitive potassium channels on the development of warm-up angina pectoris. Cardiology 2006;105:17–21.
      OpenUrlCrossRefPubMed
    58. ↵
      1. Bilinska M,
      2. Potocka J,
      3. Korzeniowska-Kubacka I,
      4. et al
      . ‘Warm-up’ phenomenon in diabetic patients with stable angina treated with diet or sulfonylureas. Coron Artery Dis 2007;18:455–62.
      OpenUrlCrossRefPubMedWeb of Science
    59. ↵
      1. Schramm TK,
      2. Gislason GH,
      3. Vaag A,
      4. et al
      . Mortality and cardiovascular risk associated with different insulin secretagogues compared with metformin in type 2 diabetes, with or without a previous myocardial infarction: a nationwide study. Eur Heart J 2011;32:1900–8.
      OpenUrlAbstract/FREE Full Text
    60. ↵
      1. Munir S,
      2. Jiang B,
      3. Guilcher A,
      4. et al
      . Exercise reduces arterial pressure augmentation through vasodilation of muscular arteries in humans. Am J Physiol Heart Circ Physiol 2008;294:H1645–50.
      OpenUrlAbstract/FREE Full Text
    61. ↵
      1. Brutsaert DL,
      2. Sys SU,
      3. Gillebert TC
      . Diastolic failure: pathophysiology and therapeutic implications. J Am Coll Cardiol 1993;22:318–25.
      OpenUrlCrossRefPubMedWeb of Science
    62. ↵
      1. Davies JE,
      2. Whinnett ZI,
      3. Francis DP,
      4. et al
      . Evidence of a dominant backward-propagating “suction” wave responsible for diastolic coronary filling in humans, attenuated in left ventricular hypertrophy. Circulation 2006; 113:1768–78.
      OpenUrlAbstract/FREE Full Text
    View Abstract

    Footnotes

    • Contributors RPW and VM-S wrote the initial drafts of the manuscript. All authors were involved in critical review of the manuscript. All authors reviewed and accepted final version of the manuscript in its entirety.

    • Competing interests None.

    • Provenance and peer review Commissioned; externally peer reviewed.