Hypoglycaemia, thrombosis and vascular events in diabetes

KEYWORDS:

• Hypoglcyaemia
• thrombosis
• cardiovascular disease
• endothelial function
• inflammation

1. Introduction

Hypoglycemia is commonly encountered in the treatment of diabetes and is increasingly seen following intensification of treatment in order to achieve glycated hemoglobin (HbA1c) targets. Estimates of severe hypoglycemia (defined as the need for third-party assistance) in type 1 diabetes (T1DM) vary from 0.98 to 3.2 episodes per patient per year, affecting 22–46% of patients over a 12-month period [1]. Landmark trials such as the DCCT in T1DM [2] and UKPDS in type 2 diabetes (T2DM) demonstrated that improved glucose control leads to a clear reduction in microvascular complications in the medium/long term and a probable reduction in macrovascular disease in the long term but at the expense of an increase in hypoglycemia rates [3,4]. Although HbA1c has been a useful marker of overall glycemic control, it fails to address other important glycemic markers, including hypoglycemia, and therefore, clinical guidelines advocate personalization of HbA1c targets as part of a patient-centered approach to diabetes management [5].

The increased mortality observed in the intensive glycemic arm of the ACCORD study created an ongoing debate on the potential link between hypoglycemia and adverse clinical outcome [6]. A number of studies have since demonstrated close associations between hypoglycemia, increased rate of cardiovascular disease, and mortality in T2DM patients [7–13] and possibly T1DM [10,14], further emphasizing the limitations of HbA1c as the sole measure of glycemic control.

2. Mechanisms linking hypoglycemia and cardiovascular disease

Hypoglycemia is associated with numerous physiological and cellular changes that can combine to devastating effect. These include (1) an increase in cardiac ischemia and dysrhythmia, (2) an increased inflammatory response leading to endothelial cell dysfunction and atherosclerosis, and (3) a disordered coagulation system. Low blood glucose levels trigger an acute, compensatory catecholamine response which along with a concomitant drop in potassium levels affects cardiac depolarization prolonging the QT interval generating a proarrhythmogenic environment. During sleep, it is likely that exaggerated sympathoadrenal activation, with or without electrolyte disturbances, leads to more lethal ventricular arrhythmias and is thought to be the cause of the so-called ‘dead in bed’ scenario whereby young T1DM patients are found dead in an undisturbed bed [15,16]. Acute, insulin-induced hypoglycemia has been shown in numerous studies to induce a proinflammatory state. Increased levels of interleukin-6 (IL-6), tumor necrosis factor-alpha (TNF-α), reactive oxygen species, and adhesion molecules have detrimental effects on the vascular endothelium leading to increased atherosclerotic plaque formation [17–19]. Numerous changes are also observed in hemostatic proteins leading to a prothrombotic environment, which are further detailed later.

3. Effect of hypoglycemia on the coagulation system

It is well recognized that chronic hyperglycemia promotes the development of atherothrombosis in diabetes by a concatenation of events that leads to increased atherosclerotic plaque formation and rupture, enhanced platelet activation, and increased thrombus formation [20,21]. Mechanisms for increased thrombosis risk include increased platelet activation as well as quantitative and qualitative changes in coagulation proteins such as plasminogen activator inhibitor-1 (PAI-1), fibrinogen, and plasminogen generating a denser fibrin clots that are resistant to fibrinolysis [22–26]. Similar adverse changes have been noted during experimental hypoglycemia, and thus, acute hypoglycemia may be an equivalent atherothrombotic risk factor to chronic hyperglycemia as discussed later.

3.1. Hypoglycemia and the cellular phase of coagulation

The cellular phase of coagulation relates predominantly to platelet activation which in turn leads to thrombin generation. This, together with thrombin generated by activation of the coagulation proteins, results in the conversion of soluble fibrinogen into insoluble fibrin networks, which form the skeleton of the blood clot. Acute hypoglycemia triggers the release of counter-regulatory catecholamines, and these have been implicated in increased platelet activation [27–29]. Recent studies of acute, insulin-induced hypoglycemia have yielded variable results in relation to platelet activation with no difference observed in markers of platelet aggregation between hypoglycemia and euglycemia in control and T1DM patients in some studies [17,19] and increased aggregation in healthy controls after repeated hypoglycemia in one study [18]. This discrepancy in the data may be related to the methodologies used and the difficulties encountered in accurately measuring platelet activation or simply due to chronic activation of platelets in hyperglycemic states making the ‘additional’ effects of hypoglycemia difficult to measure.

3.2. Hypoglycemia and the fluid phase of coagulation

Acute, insulin-induced hypoglycemia appears to have a deleterious effect on hemostatic proteins tipping the balance toward a prothrombotic, hypofibrinolytic state as is observed in hyperglycemia. Increased levels of the antifibrinolytic agent PAI-1 have been demonstrated following acute [17] and repeated [18] hypoglycemia. Moreover, hypoglycemia is associated with increased thrombin generation [18] and plasma levels of clotting factor VIII [30], thereby enhancing thrombus formation. Contrary to hyperglycemia which can persist beyond the acute episode, hypoglycemic attacks are generally terminated quickly by patients in everyday life, except in those with hypoglycemia unawareness where hypoglycemic can persist for longer periods. Hyperinsulinemic hypoglycemia clamp studies indicate that deranged coagulation proteins return to normal once euglycemia is achieved, and thus, how transferrable these data are to clinical practice may be debated. However, we have shown that ex vivo fibrinolysis is impaired for up to 1 week following hypoglycemia, mediated at least partly by altered levels of coagulation factors; therefore, the thrombotic effects of a single hypoglycemic episode are not necessarily short-lasting [31]. Moreover, lower fasting blood glucose in T2DM patients at high cardiovascular risk has also been linked to elevated thrombin generation, and formation of less porous clots that are more resistant to fibrinolysis, further providing evidence for a link between hypoglycemia and increased thrombosis risk [32].

3.3. Hypoglycemia and endothelial function

In the absence of conditions such as atrial septal defects or infective endocarditis, the clinical entities of myocardial infarction and cerebrovascular disease mainly occur following the formation of an occlusive arterial thrombus secondary to atheromatous plaque rupture.

The formation of these lipid laden plaques is a chronic inflammatory process that takes many years and is exacerbated by factors such as hyperglycemia leading to endothelial dysfunction. There is evidence to support a proinflammatory state associated with hypoglycemia with increased inflammatory cytokines such as IL-6 and TNF-α [17,18,33], along with markers of endothelial dysfunction such as vascular adhesion molecules and reduced nitric oxide-induced vasodilation [18,19] and increased oxidative stress [33]. For reasons cited earlier, it is unclear to what extent brief episodes of hypoglycemia have on the development of atherosclerotic plaque formation as opposed to the more prolonged hyperglycemia, and studies in this area are urgently needed.

4. Future directions and management strategies

Until recently, an ongoing challenge in glycemic management has been the inability to capture all hypoglycemic episodes due to the sporadic nature of capillary glucose testing and the difficulties encountered in long-term continuous glucose monitoring. This has compromised our understanding of the role of hypoglycemia in the pathogenesis of atherothrombosis in patients with diabetes. However, the recent development of continuous glucose monitoring devices that are affordable and easy to use will undoubtedly have a major impact on our understanding of the relationship between hypoglycemia, thrombosis, and adverse vascular outcome. Moreover, the new devices will help to disentangle the importance of various components of hypoglycemia in relation to atherothrombosis risk including frequency, severity, and duration of hypoglycemia.

Long-term continuous glucose monitoring will have the opportunity to: (1) clarify the true extent of hypoglycemia in diabetes together with the role of various therapies, (2) accurately identify the adverse hypoglycemic profile by addressing the effects of frequency, severity, and time spent in hypoglycemia, and (3) understand various mechanistic pathways linking low glucose levels to increased atherothrombotic events with the potential to individualize risk and draw up appropriate management strategies according to the needs of each patient.

5. General conclusions

It is well accepted that hyperglycemia in diabetes is prothrombotic and is associated with increased risk of vascular occlusive events. However, several pieces of emerging evidence indicate that hypoglycemia also increases thrombosis potential and is associated with increased cardiovascular events, leading to high vascular mortality. Our efforts in understanding the relationship between hypoglycemia and vascular thrombosis have been hampered by the inability to capture all episodes of low glucose levels in patients with diabetes. Moreover, mechanistic work has so far mainly concentrated on hypoglycemic clamp studies, a highly artificial environment which only partly reflects real life situations.

With the advent of a new generation of glucose testing devices, future prospective and observational studies should enable better understanding of the hypoglycemia profile that enhances thrombosis risk and which contributes to increased vascular events in diabetes. These mechanistic and clinical studies will further assist in characterizing the adverse hypoglycemia-induced thrombotic profile, and this in turn will help to identify the favorable glycemic targets that are associated with the lowest risk of vascular thrombosis.

While HbA1c has so far served us well as a glycemic measure, there is more to glycemia than a single glycated protein. A more modern approach is now required that takes into account other glycemic markers, most notably hypoglycemia, in order to better understand the complex relationship between glucose levels, thrombosis risk, and vascular events in patients with diabetes.

Declaration of interest

R Ajjan has received research funding and/or honoraria from Astra Zeneca, Bayer, Boehringer Ingelheim, Bristol-Myers Squibb, Eli Lilly, Johnson & Johnson (Lifescan), GlaxoSmithKline, Merck Sharpe & Dohme, NovoNordisk, Roche, Takeda and Abbott. The authors have no other relevant affiliations or financial involvement with any organization or entity with a financial interest in or financial conflict with the subject matter or materials discussed in the manuscript apart from those disclosed.

Additional information

Funding

This paper was not funded.