Hypoglycemia and Cardiovascular Risk: Is There a Major Link?

Hypoglycemia is a frequent adverse complication of glucose-lowering treatment of diabetes, particularly with insulin and sulfonylureas. Severe hypoglycemia has been identified to be one of the strongest predictors of macrovascular events, adverse clinical outcomes, and mortality in people with type 2 diabetes (1–4). The adjusted hazard ratios for total mortality of patients experiencing at least one episode of severe hypoglycemia (defined as requiring assistance for recovery) compared with those with no hypoglycemia in large prospective randomized trials have been found to be between 1.67 and 4.28; by contrast, nonsevere (self-treated) hypoglycemia was not associated with a significant increase in either total mortality or cardiovascular death (3–5). A meta-analysis of prospective and retrospective cohort studies in people with type 2 diabetes has supported this interpretation and reported a hazard ratio of 2.05 (95% CI 1.74, 2.42) for cardiovascular outcomes (i.e., death from cardiovascular causes, myocardial infarction, stroke, or surgical intervention for vascular disease) (6). In type 1 diabetes, large prospective studies such as the Diabetes Control and Complications Trial (DCCT) and the EURODIAB Prospective Complications Study have not demonstrated an increased risk of mortality or fatal cardiovascular disease related to hypoglycemia (7,8), but a recent retrospective cohort study using the Clinical Practice Research Datalink database reported a hazard ratio for all-cause mortality of 1.98 (1.25, 3.17) in patients with type 1 diabetes who had experienced at least one episode of severe hypoglycemia (9).

In large prospective studies in type 2 diabetes, the frequency of severe hypoglycemic events is much higher with intensification of glucose-lowering therapy compared with standard treatment by a factor of at least two to three, but the reported event rate differs considerably between the trials and is not exclusively associated with the HbA_1c value but also depends on age, comorbidities, diabetes duration, and treatment regimens (3,8,10–13). A high degree of underreporting of hypoglycemia is likely in these long-term trials, since many events occur during sleep and are not perceived by the patients, and continuous glucose monitoring was not undertaken (14). Despite the strong evidence for a relationship between hypoglycemia and both cardiovascular and all-cause mortality in large prospective outcome studies of patients with type 2 diabetes, it is unclear whether there is a direct pathophysiological link with hypoglycemia or whether hypoglycemia is primarily a marker of vulnerability to these events.

During hypoglycemia, the brain becomes neuroglycopenic and the central autonomic nervous system is stimulated as part of the counterregulatory response to hypoglycemia. With respect to the cardiovascular system, the most important response is sympathoadrenal activation with the release of catecholamines, but the parasympathetic division is also activated and will also influence cardiac responses. The autonomic neural stimulation, accompanied by the surge in plasma catecholamines and the diminished energy supply of glucose per se, has profound cardiovascular effects. In one hypoglycemia clamp study (plasma glucose 3.0 mmol/L [54 mg/dL]) in participants without diabetes, a 12-fold increase in plasma epinephrine concentration was observed that was associated with increased heart rate and cardiac stroke volume and decreased peripheral resistance (15). Acute hypoglycemia provokes a substantial rise in myocardial contractility (16) and cardiac output (17). These hemodynamic changes are accompanied by an increase in elasticity of large blood vessels and a fall in central arterial pressure, with both measures being diminished in young men with type 1 diabetes of long duration (>15 years) in whom arterial stiffness has developed (18). A recent analysis of coronary artery calcification scores in patients with type 1 diabetes suggested an association of higher scores—indicating a higher degree of calcification and hence arterial stiffness—with a higher rate of severe hypoglycemia (19). Progressive arterial stiffness in people with diabetes allows the reflected arterial pressure wave to return to the central aorta and heart more quickly, arriving during diastole instead of systole, which may compromise coronary filling. These hemodynamic responses to hypoglycemia may be detrimental for people with diabetes of longer duration, particularly patients with type 2 diabetes who have underlying cardiac disease including coronary heart disease, autonomic abnormalities, and dysfunction of cardiac muscle (20,21).

Evidence that hypoglycemia can cause myocardial ischemia is accumulating. In an American study of 19 patients with type 2 diabetes who had documented coronary heart disease, simultaneous continuous glucose monitoring and Holter ECG recording were applied; 10 episodes of hypoglycemia were associated with chest pain (out of 54 in total), and acute ischemic ECG abnormalities were present during 6 of these episodes (22). A surrogate measure of myocardial perfusion during hypoglycemia in young adults with and without type 1 diabetes has recorded lower values in the group with diabetes (23). How often severe hypoglycemia may cause myocardial infarction is unknown, as only a few anecdotal cases have been reported, while a large retrospective Japanese study of 414 subjects presenting with severe hypoglycemia revealed newly diagnosed vascular disease in five patients (with only 1 case of myocardial infarction and 4 cases of cerebrovascular disease) (24).

Even more important than the association between hypoglycemia and cardiac ischemia is the possible induction of serious cardiac arrhythmias, which can lead to sudden death, as has been documented in several case reports (25–27). Epinephrine induces hypokalemia, which affects electrophysiological function and may predispose to the development of an arrhythmia (28). A very frequent finding during hypoglycemia is QT prolongation of the surface ECG. This has been demonstrated in both type 1 and type 2 diabetes (24,29–31). QT prolongation is a strong risk factor for severe ventricular arrhythmia and sudden death (32) and has been suggested as an explanation for the “dead in bed” syndrome of patients with type 1 diabetes (29). Hypoglycemia can be proarrhythmogenic via different mechanisms: a low blood glucose has a direct effect on the ether-á-go-go–related gene (HERG) ion channel, and the catecholamine release causes hypokalemia and action potential prolongation and increases the risk of a dangerous ventricular arrhythmia (33,34).

Experimental animal studies of lethal hypoglycemia in rats have provided insight into disturbances of cardiac electrical stability, although these findings may not be directly applicable to humans. Profound biochemical hypoglycemia (blood glucose 0.6 mmol/L [12 mg/dL]) in rats initially induced sinus tachycardia and QT prolongation, followed by ventricular premature beat ventricular tachycardia and atrio-ventricular blockade. At the time of cardiorespiratory arrest, a markedly reduced heart rate (bradycardia) was the prominent finding (35). Interestingly, the reported ECG changes were significantly reduced after potassium supplementation or the administration of intracerebral glucose to prevent neuroglycopenia—both of these findings indicating a strong role for sympathoadrenal activation in the initiation of arrhythmias. Simultaneous recording of continuous interstitial glucose and ECG monitoring in humans with insulin-treated type 2 diabetes and cardiovascular disease demonstrated that a much higher incidence of bradycardia and atrial and ventricular premature beats was associated with spontaneous hypoglycemia, particularly during the night when the patients were asleep (36). In a comparable study population, our group in Germany observed a high number of complex ventricular arrhythmias but with no obvious synchronized occurrence of the cardiac arrhythmias with documented hypoglycemia (14,37). In this case-control study, the duration of severe hypoglycemia was the most important independent predictor of developing a serious ventricular arrhythmia. An example of a parallel recording is shown in Fig. 1. Finally, there are several case reports of hypoglycemia-induced cardiac arrhythmias, of which atrial fibrillation is one of the most commonly reported (29,38), though this has not been demonstrated by recent studies with parallel recording of ECG and continuous glucose monitoring.

Another mechanism that may link hypoglycemia with cardiovascular events is the prothrombotic effect of low blood glucose levels. Catecholamines and other hormones released during hypoglycemia increase blood viscosity and promote platelet aggregation and activation (21,39–41). In addition, insulin-induced hypoglycemia in people with type 1 diabetes increases coagulation factors such as fibrinogen and factor VIII (42) and promotes the release of inflammatory cytokines to the circulation (41,43,44), which may remain elevated for several days after the hypoglycemia has resolved (40). Taken together, these proinflammatory and procoagulant effects could directly affect vascular flow, leave the affected individual vulnerable for some time after hypoglycemia, and potentially contribute to the occurrence of major vascular events. The major cardiovascular effects of hypoglycemia are shown in Table 1.

The counterregulatory responses to hypoglycemia serve to restore euglycemia and protect the brain from protracted neuroglycopenia. However, some aspects of the counterregulatory mechanism, particularly the sympatho-adrenal response, can potentially promote a serious cardiovascular event, and the question remains whether there is a major link between low blood glucose level per se (or its related effects) and major cardiovascular and cerebrovascular events—myocardial infarction and sudden death from an arrhythmia or stroke. Despite the evident possibility of a causal relationship that has been clearly demonstrated by mechanistic studies, the evidence from large prospective studies that hypoglycemia is a major causal contributor to cardiovascular events is limited. The major limitation is the inability to assign the cause of death to hypoglycemia with absolute certainty, as many episodes are asymptomatic and are difficult to identify in clinical trials and the temporal relationship between hypoglycemia and a subsequent vascular outcome is often difficult to determine (14,20). Therefore, anecdotal case reports in which cardiac arrhythmia or sudden death is associated with documented hypoglycemia remain the principal source of evidence. Furthermore, most associations between hypoglycemia and cardiovascular events have been detected in patients with a high cardiovascular risk—defined by age, comorbidities, and concomitant medication. It is probable that preexisting cardiovascular risk factors other than the hypoglycemia per se were a major link to the cardiovascular events that were observed in these patients. This hypothesis is supported by additional analyses of the Action in Diabetes and Vascular Disease: Preterax and Diamicron MR Controlled Evaluation (ADVANCE), the Action to Control Cardiovascular Risk in Diabetes (ACCORD) study, and the Outcome Reduction With Initial Glargine Intervention (ORIGIN) study, which demonstrated indeed a higher rate of severe hypoglycemia but a lower rate of hypoglycemia-related mortality in the intensified treatment arm compared with the standard treatment (3–5).

In some mechanistic studies that focused on a high-risk population, the association between hypoglycemia and arrhythmias was influenced by covariates; for example, our group described a higher rate of ventricular tachycardia (VT) in patients with a hypoglycemic event, but VT occurred with the same frequency during daytime and sleep, whereas hypoglycemia was recorded twice as often at night than during daytime (14). This differs from the findings of Chow et al. (36), where bradycardia and ventricular ectopics were strongly associated with nocturnal rather than daytime hypoglycemia and may reflect the obtunded sympatho-adrenal response during sleep (43) allowing compensatory parasympathetic activity. Furthermore, the multivariate regression analyses of our data revealed—besides the duration of severe hypoglycemia—a low normal TSH level as an independent predictor for the occurrence of a VT (14). This would suggest that an interaction of risk factors, among them hypoglycemia, but not one factor alone may be a prerequisite for the final event.

So is hypoglycemia a major causal factor for cardiovascular events or is it a marker of vulnerability to these events? This question is not yet resolved. In conclusion, hypoglycemia can precipitate cardiac arrhythmias and may diminish myocardial perfusion, so leading to a major cardiovascular event (38) but probably in interaction with other risk factors. Through extrapolation of the evidence that is currently available, it would appear that severe hypoglycemia can trigger cardiovascular events in vulnerable patients who are at high cardiovascular risk and should therefore be avoided at all costs in people with known cardiac disease who require an individualized approach to diabetes care.

Duality of Interest. No potential conflicts of interest relevant to this article were reported.