Review: Does hypoglycaemia cause cardiovascular events?

The British Journal of Diabetes & Vascular Disease http://dvd.sagepub.com/

Review: Does hypoglycaemia cause cardiovascular events? Alex J Graveling and Brian M Frier British Journal of Diabetes & Vascular Disease 2010 10: 5 DOI: 10.1177/1474651409355113 The online version of this article can be found at: http://dvd.sagepub.com/content/10/1/5

Published by: http://www.sagepublications.com

Additional services and information for The British Journal of Diabetes & Vascular Disease can be found at: Email Alerts: http://dvd.sagepub.com/cgi/alerts Subscriptions: http://dvd.sagepub.com/subscriptions Reprints: http://www.sagepub.com/journalsReprints.nav Permissions: http://www.sagepub.com/journalsPermissions.nav Citations: http://dvd.sagepub.com/content/10/1/5.refs.html

>> Version of Record - Feb 17, 2010 What is This?

Downloaded from dvd.sagepub.com by guest on October 11, 2013

review

Does hypoglycaemia cause cardiovascular events? Alex J Graveling, Brian M Frier Abstract

S

trict glycaemic control is strongly advocated in people with type 2 diabetes to prevent vascular disease. However, the outcomes of two large clinical trials have indicated the potential dangers of pursuing this policy in those at high risk of cardiovascular disease, with an excess of fatal vascular events being associated with a higher frequency of severe hypoglycaemia. Hypoglycaemia secondary to insulin and sulphonylurea therapy is often associated with serious morbidity; anecdotal evidence has long implicated hypoglycaemia as a potential cause of myocardial ischaemia or a cardiac arrhythmia. Hypoglycaemia provokes sympatho-adrenal activation and counterregulatory hormone secretion, which exert pronounced cardiovascular effects. Although well tolerated in healthy people, the superimposition of these profound physiological effects on a diseased coronary vasculature and a dysfunctional cardiac conductive system may induce serious or even fatal cardiovascular events. These risks should influence therapeutic targets and the approach to diabetes management in people with diabetes with established vascular disease in whom exposure to severe hypoglycaemia could be dangerous. Br J Diabetes Vasc Dis 2010;10:5–13 Key words: cardiac arrhythmia; cardiovascular event; diabetes mellitus; hypoglycaemia; myocardial ischaemia

Introduction Diabetes and cardiovascular mortality Diabetes is a major risk factor for cardiovascular disease and is the principal cause of death in people with type 2 diabetes mellitus.1 Up to 80% of people with diabetes die from a major cardiovascular event, principally myocardial infarction but also stroke.2 Diabetes per se doubles the risk of cardiovascular events, which is equivalent to that of people without diabetes Department of Diabetes, Royal Infirmary of Edinburgh, Edinburgh, Scotland, UK. Correspondence to: Professor Brian M Frier Department of Diabetes, Royal Infirmary of Edinburgh, Edinburgh, EH16 4SA, UK. Tel: +44 (0)131 242 1475; Fax: +44 (0)131 242 1485 E-mail: [email protected]

Abbreviations and acronyms ACCORD ACS ADA ADVANCE DIGAMI ECG HbA1C ICU VADT

Action to Control Cardiovascular Risk in Diabetes acute coronary syndrome American Diabetes Association Action in Diabetes and Vascular Disease: Preterax and Diamicron Modified Release Controlled Evaluation Diabetes Mellitus Insulin-Glucose Infusion in Acute Myocardial Infarction electrocardiogram glycated haemoglobin A1C intensive care unit Veterans Affairs Diabetes Trial

who have already suffered a myocardial infarction.3,4 Microvascular disease is considered to be principally a risk of people with type 1 diabetes mellitus while those with type 2 diabetes are thought to be mostly at risk of developing premature macrovascular disease.5 However, although patients with type 2 diabetes undoubtedly have an increased risk of developing premature cardiovascular disease, patients with type 1 diabetes have a similar level of risk for cardiovascular disease when matched for age.6,7 Conversely, around 65% of patients admitted with a myocardial infarction, who have no history of diabetes, are found subsequently to have either impaired glucose tolerance or frank type 2 diabetes.8

Effect of HbA1C on cardiovascular mortality

An elevated HbA1C is associated with an increased risk of sustaining a cardiovascular event.9 While good glycaemic control limits the risk; these benefits appear to persist for several years after glycaemic control has been relaxed both in type 1 and type 2 diabetes,10,11 suggesting the existence of a metabolic (glycaemic) memory. However, three large clinical trials, ACCORD, ADVANCE and VADT, have failed to show that intensive treatment to obtain strict glycaemic control (HbA1C < 7.0%) will lower the frequency of major cardiovascular events.12-14 While a subsequent metaanalysis indicated a benefit in selected outcomes, including death from cardiovascular events, no beneficial effect was observed on all cause mortality (see figure 1).15 The DIGAMI study showed that the use of insulin to treat and prevent hyperglycaemia in people

© The Author(s), 2010. Reprints and permissions: http://www.sagepub.co.uk/journalsPermissions.nav

10.1177/1474651409355113

5

REVIEW

Figure 1.   Composite forest plot of clinical outcomes of intensive glucose control on cardiovascular outcomes in a meta-analysis of randomised controlled trials

Reprinted with permission.15 ©2009 Elsevier.

with diabetes who presented with an ACS, resulted in fewer deaths.16 A subsequent study of patients who had better baseline glycaemic control did not show the same mortality benefit.17

Putative mechanisms of myocardial dysfunction in diabetes Diabetes, and hyperglycaemia in particular, can affect the coronary vasculature in several ways.18 Although the risk of coronary heart disease is higher in people with type 2 diabetes, the magnitude of risk varies widely between individuals and is not fully explained by traditional risk factors.2,19 In addition, the risk for women with diabetes is disproportionately raised, to the extent that they lose the protective effect of their sex and exhibit a risk equivalent to that of men with diabetes.20 The incidence of heart failure in people with diabetes is also substantially increased compared with matched non-diabetic controls (11.8% versus 3.2%).21 The putative disease process responsible is erroneously referred to as diabetic cardiomyopathy, although a more accurate term would be ‘specific heart disease of diabetes’.22 The aetiology of this complication is likely to be complex and multifactorial but a microvascular basis has been suggested as it is associated with a higher incidence of retinopathy23 and evidence of various pathological changes in the heart.24

6

Endothelial dysfunction is an early pathological feature in the development of atherosclerosis25 but the degree of impairment does not always correlate with the extent of coronary arterial disease.26 Severe endothelial dysfunction, even in the absence of obstructive coronary artery disease, is associated with an increase in cardiac events.27 People with diabetes develop arterial stiffness prematurely, with a significant increase being observed in young adults with type 1 diabetes of long duration (> 15 years), as compared with age-matched non-diabetic volunteers and those with a shorter duration of type 1 diabetes (< 5 years).28 This arterial stiffness has important implications for coronary vascular flow. Coronary artery filling mostly occurs during diastole. Normal elasticity of the arterial wall ensures that the reflected pressure wave from the high-pressure arterioles, generated during each myocardial contraction, returns to the heart during early diastole so increasing diastolic pressure and thereby enhancing coronary arterial perfusion.29 Progressive stiffening of the arterial wall results in the earlier arrival of the reflected wave during late systole, which may interfere with coronary arterial perfusion.30 The increased vascular stiffness that occurs in people with type 2 diabetes has been shown to predict cardiovascular mortality.31 VOLUME 10 ISSUE 1  .  January/February 2010

REVIEW

Table 1.  Haemodynamic changes during hypoglycaemia36-38

• Increased heart rate • Increased peripheral systolic blood pressure • Decreased central blood pressure • Increased myocardial contractility with an increased ejection fraction • Decreased peripheral resistance (resulting in a widened pulse pressure)

75

240

70

220

65

200

60

180

55

160

50

140

45

120

40

100 0

Cardiac output (% of basal)

Leftventricular ejection fraction (%)

Figure 2.   Cardiac function during hypoglycaemia. R corresponds with the acute autonomic reaction observed at the glucose nadir (mean 1.0 ± 0.2 mmol/L)

R + 90 R + 60 R + 30 R Time from autonomic reaction (R) in minutes Leftventricular ejection fraction

Cardiac output

Composite of figures 2 and 3 from Fisher et al.39

Hypoglycaemia

not important for counterregulation, such as the spleen and the skin. By contrast it is increased to the myocardium, splanchnic circulation (carrying 3-carbon precursors to the liver) and the brain. Autonomic activation, principally of the sympathoadrenal system, results in end-organ stimulation and a profuse release of epinephrine (adrenaline), which provokes haemodynamic changes (see table 1 and figure 2).35-39 The increased activity of the sympathetic nervous system and secretion of other hormones and peptides such as the potent vasoconstrictor, endothelin, have pronounced effects on intravascular haemorheology, coagulability and viscosity. 40 Increased plasma viscosity occurs during hypoglycaemia because of an increase in erythrocyte concentration, while coagulation is promoted by platelet activation and an increment in factor VIII and von Willebrand factor. Endothelial function may be compromised during hypoglycaemia because of an increase in C-reactive protein, mobilisation and activation of neutrophils and platelet activation. Figure 3 indicates how these changes could promote vascular occlusion and localised tissue ischaemia.

How does hypoglycaemia affect myocardial function? Hypoglycaemia has profound effects on cardiac function but initially these can be difficult to distinguish from the effects of insulin per se. Insulin is a coronary vasodilator and has proinflammatory actions.41,42 The administration of intravenous insulin has a small, immediate effect to promote sympathetic neural activation and to increase left ventricular ejection fraction, before any fall in blood glucose occurs.22 These changes become more pronounced with a progressive decline in blood glucose; the maximal responses coincide with the glucose nadir. Significant increments in stroke volume and cardiac output also occur during hypoglycaemia.36 The haemodynamic changes during hypoglycaemia that are observed in non-diabetic adults are attenuated in some people with type 1 diabetes who have strict glycaemic control, which has been attributed to attenuated sympathetic stimulation.43

Hypoglycaemia is a very common side effect of insulin therapy and to a lesser extent of treatment with sulphonylureas. In type 1 diabetes the requisite for strict glycaemic control using intensive insulin therapy to minimise long-term microvascular complications, is associated with a three-fold increase in the risk of severe hypoglycaemia.32 Risk factors for severe hypoglycaemia include age, duration of diabetes, strict glycaemic control, sleep, impaired awareness of hypoglycaemia, renal impairment, C-peptide negativity and a previous history of severe hypoglycaemia.33,34 Hypoglycaemia results in significant morbidity and in younger patients with type 2 diabetes (aged from 20–49 years), between 6 and 18% of deaths have been attributed to hypoglycaemia.35

Antecedent hypoglycaemia induces a reduction in baroreflex sensitivity and sympathetic response to hypotensive stress, which may predispose susceptible individuals to the development of a cardiac arrhythmia.44,45 Diminished sympatho-adrenal responses also occur in type 2 diabetes following exposure to antecedent hypoglycaemia46 but the possible effects on the autonomic innervation to the heart have not been studied in type 2 diabetes. Impairment of the autonomic neural control of heart rate is associated with an increased risk of mortality.47

Pathophysiology of hypoglycaemia

Hypoglycaemia and mortality rate

Acute hypoglycaemia provokes pronounced physiological responses, the important consequences of which are to maintain the supply of glucose to the brain and promote the hepatic production of glucose. Blood flow is reduced to organs that are

A direct relationship between hypoglycaemia and a fatal cardiovascular event is difficult to demonstrate as blood glucose and cardiac monitoring are seldom performed simultaneously. In 2008 an excess of deaths in the intensive treatment arm of the

THE BRITISH JOURNAL OF DIABETES AND VASCULAR DISEASE

Does previous exposure to hypoglycaemia modify the risk?

7

REVIEW

Figure 3.    Haematological and inflammatory processes that may lead to vascular complications

Haemodynamic changes

↑Endothelin

↑Plasma viscosity

↑Factor VIII

↓Blood flow

Vasoconstriction

Capillary closure

↑vWF

Platelet activation

Neutrophil activation

↑Coagulation

Endothelial damage

Thrombosis

Atherogenesis

↑CRP

VASCULAR COMPLICATIONS

Adapted with permission.40

ACCORD study led to early discontinuation of the study.12 This has prompted much speculation about the likely reasons and potential underlying mechanisms that could be responsible for the greater number of deaths in people with strict glycaemic control. In ACCORD hypoglycaemia was three times more common in the intensively treated group, with an annual prevalence of severe hypoglycaemia of 3.3% compared to 1.1% in the group receiving standard treatment. Mortality in the 93% of patients who had no episodes of severe hypoglycaemia during the study was 1.2% compared with 3.1% in those with a history of one or more episodes of severe hypoglycaemia.48 Although the authors of the ACCORD study have persistently maintained that the cardiovascular deaths were not caused by hypoglycaemia (ADA meeting, New Orleans, 2009), this assertion cannot be proven as neither continuous blood glucose nor Holter monitoring was used in the study. A panel of cardiologists, that adjudicated the cause of death in the fatal cases, decided that none of these had been caused by hypoglycaemia. As concomitant blood glucose data were not available in these fatal cases it is unclear how the potential predisposing effects of hypoglycaemia could be discounted in this situation. In the smaller study of veterans with type 2 diabetes, VADT,14 severe hypoglycaemia was found to increase the risks of adverse events and death. In both treatment groups combined, those who had experienced hypoglycaemic coma had an 88% increase in primary cardiovascular events and a threefold higher rate of cardiovascular death (ADA meeting, New Orleans, 2009). Information from other clinical studies support the premise that exposure to hypoglycaemia has inherent cardiovascular risks, particularly in people who are seriously ill. Patients with

8

diabetes, who were admitted to hospital with ACS and who experienced severe hypoglycaemia at some point during their hospital stay, exhibited double the mortality rate compared with those who had no hypoglycaemia. This effect persisted despite adjustment for potential confounders.49,50 The cardiovascular causes of death in patients who had a history of recent hypoglycaemia suggested that the susceptibility of those patients to cardiac arrhythmia may have been increased by preceding exposure to low blood glucose. A possible mechanism may involve the demonstrated effect of antecedent hypoglycaemia to diminish the cardiac vagal baroreflex sensitivity and the sympathetic response to drug-induced hypotension, and thereby attenuating the autonomic responses to cardiovascular stress for up to 16 hours.44 Extremes of blood glucose on admission to hospital in patients with ST-elevation ACS, were associated with a greater 30-day mortality (figure 4).51 In a different study, patients who did not have diabetes but who developed spontaneous hypoglycaemia during their hospital stay had a significantly poorer outcome than those with insulintreated diabetes who developed iatrogenic hypoglycaemia.50,52 In this situation the development of hypoglycaemia may be a surrogate marker for severity of illness, but it may also contribute directly to a fatal outcome.

Hypoglycaemia in the ICU In hospitalised patients with critical illness, hyperglycaemia is associated with poorer clinical outcomes but the degree to which glucose should be lowered during the management of these patients has provoked debate.53,54 Despite various observational studies that have shown a direct linear relationship between

VOLUME 10 ISSUE 1  .  January/February 2010

REVIEW

Figure 4.   Association of admission blood glucose level with mortality at 30 days

12.0% 10.4%

Mortally (%)

10.0% 8.0% 6.3% 5.2%

6.0% 4.0%

3.1%

2.6%

2.3%

4.5–5.5

5.6–6.9

2.0% 0.0% 11.1

Blood glucose level (mmol/L)

Reprinted with permission51©2008 Elsevier.

Table 2.  Summary data from randomised clinical trials of intensive insulin therapy in critically ill patients55

Trial name

No. of Type of Blood glucose level patients ICU targeted



Blood glucose level achieved

Primary outcome

Intensive Conventional Intensive Conventional glucose glucose glucose glucose control control control control



mmol/L

Rate of outcome

Odds ratio

Intensive Conventional glucose glucose control control

mmol/L

%

Leuven 1 1548

Surgical

4.4–6.1

10.0–11.1

5.7

8.5

Death in ICU

4.6

8.0

0.58 (0.38–0.78)

Leuven 2 1200

Medical

4.4–6.1

10.0–11.1

6.2

8.5

Death in hospital

37.3

40.0

0.94 (0.84–1.06) 1.10 (0.84–1.44)

Glucontrol 1101

General

4.4–6.1

7.8–10.0

6.6

8.0

Death in ICU

16.7

15.2

VISEP

537

General

4.4–6.1

10.0–11.1

6.2

8.4

Death at 28 days

24.7

26.0

Not reported

NICE-

6104

General

4.4–6.1

8.0–10.0

6.6

8.1

Death at 90 days

27.5

24.9

1.14 (1.02–1.28)

SUGAR

decrements in blood glucose and an increase in mortality rates, it is unclear whether hypoglycaemia per se is responsible for these outcomes or whether it is simply a surrogate marker indicating the sickest patients who are at the greatest risk of dying.55 A seminal study from Belgium showed that mortality was lower in patients who were treated with an intensive insulin regimen in the setting of an ICU;56 only 13% of these patients were known to have diabetes. Subsequent studies have failed to replicate these findings in other inpatient populations (table 2), which has challenged how much these results can be generalised.

THE BRITISH JOURNAL OF DIABETES AND VASCULAR DISEASE

The relationship of glycaemic control to outcome was examined in a large, multi-centre, randomised controlled trial of patients treated in ICUs in Australia.57 Strict control of blood glucose (4.5–6.0 mmol/L) was compared with standard control (