Prediabetic states and diabetes are important risk factors for cardiovascular morbidity and mortality (1–3). Cardiovascular death or death of unknown origin was in the 0.4–0.5% range in the subgroups of a 3-year follow-up study on patients with impaired glucose tolerance (IGT) and/or impaired fasting glucose (IFG) level (4). In a 23-year follow-up study on Japanese American men, the relative risks for sudden cardiac death were 2.22 in subjects with asymptomatic hyperglycemia, and 2.76 in diabetic patients (5). Diabetes status was a strong risk factor for sudden death, but not for fatal myocardial infarction in men during the population-based Paris Prospective Study I (6). Higher risk of sudden cardiac death was associated with borderline diabetes, diabetes with or without microvascular disease, compared to subjects without diabetes in a population-based case–control study of patients experienced out-of-hospital cardiac arrest due to heart disease (1). Out-of-hospital sudden cardiac deaths were 1.79-fold for non-diabetic men with impaired fasting plasma glucose and 2.26-fold for men with type 2 diabetes, and a 1 mmol/l increment in fasting plasma glucose was related to an increase of 10% in the risk of sudden cardiac death in Finnish men (7).

The prevalence of confirmed cardiovascular autonomic neuropathy (AN), an impairment of autonomic control of the cardiovascular system during diabetes, was between 16 and 20% in unselected type 1 and type 2 diabetic patients (8). Cardiovascular AN is a risk marker of cardiovascular morbidity, and it causes a 3.65-fold increase in the relative risk of mortality (8). Cardiac AN promotes ventricular repolarization disturbances [QTc prolongation, increased QT dispersion (QTd)] and may increase the risk of sudden cardiac death. Prolongation of QT interval could lead to increased myocardial electrical instability, predisposing diabetic subjects with AN to potentially fatal ventricular arrhythmias (9). Cardiac AN with QT interval prolongation proved to be a poor prognostic factor for sudden cardiac death in diabetic patients in a 5-year follow-up study (10). Prolonged QTc is more frequent in patients with IFG (30%) and with diabetes (42%) than in subjects with normal glucose tolerance (22%), and both IFG and diabetes increased the risk of prolonged QTc (11). QTc interval duration was found to be significantly higher both during the day and night using ECG Holter recordings in patients with IGT compared to subjects with normal glucose tolerance (12). IGT was confirmed in 15% of men and 23% of women with QTc prolongation (>440 ms) in the population-based Hisayama study in Japan (13).

In the clinical setting, the risk assessment of serious ventricular arrhythmias in individual patients is challenging since the prolongation of repolarization that manifests as QT interval prolongation on the ECG does not always correlate with subsequent development of ventricular arrhythmias (14–16). Cardiac repolarization reserve may be reduced even without significant changes in the duration of cardiac repolarization; therefore, QT interval prolongation cannot reliably predict the development of ventricular arrhythmias (17). The short-term variability of the QT interval (STV_QT) was introduced as an early and sensitive indicator of repolarization instability (18) that more reliably predicted ventricular arrhythmias and sudden cardiac death than prolongation of repolarization in previous experimental (16, 19–22) and clinical studies (23–27). Type 1 diabetes mellitus moderately lengthened ventricular repolarization, attenuated repolarization reserve, and enhanced the risk of sudden cardiac death in dogs (27, 28), and similar mechanisms might also occur in patients suffering from prediabetic states and diabetes.

The aim of the present study was to determine beat-to-beat STV_QT for assessment of repolarization instability and possible proarrhythmic risk, together with cardiovascular autonomic function in patients with IGT.

Cardiac autonomic dysfunction present in prediabetes may lead to repolarization disturbances and may increase the risk of ventricular arrhythmias and sudden cardiac death. In this study, we show for the first time that beat-to-beat STV_QT, an early and sensitive parameter of repolarization instability, is increased even before QTc prolongation or enhanced QTd could be detected in patients with IGT.

Patients with prediabetic conditions or diabetes have higher risk for sudden cardiovascular death (1, 5–7). Cardiac AN and instability of cardiac repolarization, detected by QTc prolongation or increased QTd, contribute to the increased risk for sudden cardiac death (9, 10). Prolonged QTc was related to a progressive worsening of glucose tolerance after adjustment for possible confounding factors in elderly women with IGT or diabetes (34). Impairment of cardiac parasympathetic and sympathetic innervation as well as QT interval prolongation may play a partial role in the pathogenic mechanism of sudden unexpected death in diabetic patients. Cardiovascular adaptation mechanisms, including baroreflex sensitivity and HR variability, are also impaired in diabetic AN that may further increase the risk for arrhythmia development (35).

However, decreased repolarization capacity and increased arrhythmia susceptibility is not necessarily preceded by significant changes in the duration of cardiac repolarization, and in these cases, cardiac repolarization reserve may be reduced without manifest QT interval prolongation (17). Importantly, a wide range of non-cardiovascular drugs or even dietary constituents with only mild repolarization blocking effects can increase the risk for serious ventricular arrhythmias and sudden cardiac death in patients with impaired repolarization reserve (17). Therefore, in this clinical setting, the prediction of lethal ventricular arrhythmias is especially challenging. STV_QT has been suggested as an early and sensitive indicator of temporal repolarization instability based on previous experimental and clinical studies (16, 18, 20, 24–26).

Our present study is the first to indicate that patients with IGT, a prediabetic condition, have repolarization instability indicated by elevated beat-to-beat STV_QT. This study was not designed to assess the exact mechanisms responsible for repolarization disturbances in patients with IGT; however, several possible mechanisms may be considered. Compelling recent evidence suggests a direct link between type 2 ryanodine receptor (RyR2) dysfunction in the endo/sarcoplasmic reticulum leading to altered intracellular calcium homeostasis, glucose intolerance, and impaired insulin secretion in patients with catecholaminergic polymorphic ventricular tachycardia (CPVT) (36, 37). The known RYR2 mutations identified in these CPVT patients were previously linked to reduced binding affinity of calstabin2 to the RyR2 channel resulting in intracellular Ca²⁺ leak (37–39). In knock-in mouse models where these CPVT-linked mutations leading to RyR2-mediated Ca²⁺ leak were reconstituted, mitochondrial dysfunction and blunted ATP production with concomitantly increased sarcolemmal K_ATP channel function (reversible by the K_ATP blocker glibenclamide) were found in pancreatic β-cells to cause reduced insulin secretion and consequently, IGT (36). In addition to causing altered glucose metabolism and providing triggers for cardiac arrhythmias (CPVT), the RyR2-mediated Ca²⁺ leak—by depleting Ca²⁺ stores—may also contribute to arrhythmia substrate creation via reduced I_Ks current, i.e., decreased Ca²⁺-dependent I_Ks activation (40) and consequently, impaired repolarization reserve (17). Interestingly, and in accordance with this mechanism, reduced I_Ks density, impaired repolarization reserve, and increased risk for sudden cardiac death were described in diabetic dogs (28). Although there is no doubt that RyR2 channel dysfunction is directly linked to heart failure (41), cardiac arrhythmia development (42, 43), IGT, and reduced insulin release (36, 44), however, further clinical studies are needed to determine whether RYR2 mutations leading to leaky RyR2 channels are frequently present in patients diagnosed with IGT in general.

Repolarization instability can be a long-standing risk factor for cardiovascular morbidity and mortality in prediabetic states and during development of diabetes. However, the role of additional cardiovascular risk factors cannot be excluded in early prediabetic conditions. Early sympathetic nerve dysfunction and insulin resistance may also play a role in the development of decreased coronary flow reserve in patients with normoglycemia (45). In this regard, increased QT interval variability associated with sympathetic dysinnervation was observed in patients with type 2 diabetes in the supine position and the QT variability was further elevated in the context of sympathetic activation upon standing (46).

Relative sympathetic predominance was observed in cardiovascular reflex tests during IGT, as sympathetic parameters (systolic BP fall after standing up and diastolic BP increase after sustained handgrip) were unchanged, whereas two of three parasympathetic parameters measured (HR variation and Valsalva ratio) were significantly decreased. In addition, a significant negative correlation was seen between the values measured in the third parasympathetic test (30/15 ratio) and STV_QT in our study. The significantly lower STV_RR values observed also represent this parasympathetic dysinnervation and subsequent relative sympathetic predominance in patients with IGT. Sympathetic predominance acutely evoked by graded head-up tilt test resulted in similar changes, such as decreased variance of HR and increased variance of repolarization duration (47, 48).

The prevalence of distal symmetric polyneuropathy that may result in weakness, sensory loss, pain, autonomic dysfunction, gait impairment, falls, and disability has been reported to be 11% in patients with IGT (49). It is known that IGT is present in about 40% of patients with idiopathic peripheral neuropathy and abnormal microvascular endothelial dysfunction is common in both patient groups (50). It has long been known that IGT is associated with AN and a shift is observed in sympathovagal balance to sympathetic overactivity (51–54). Prevalence of parasympathetic dysfunction was 25%, whereas the prevalence of sympathetic dysfunction was 6% in 268 patients with IGT in the Finnish Diabetes Prevention Study (55). Abnormal sinus arrhythmia test (55 vs 33%; P = 0.004) and abnormal Valsalva maneuver (34 vs 7%; P = 0.004) were significantly more frequent in patients with IGT than in control subjects; however, the frequency of abnormal postural test was not different in these two groups (P = 0.334) (51). Insulin resistance was associated with global autonomic dysfunction and an increased LF/HF (low frequency/high frequency) ratio indicating sympathetic overactivity (52). However, the autonomic dysfunction was less significant in IGT patients than in diabetic subjects (52). IGT induced decrease in parasympathetic modulation (decreased HF power and 30/15 ratio) and a shift toward augmented sympathetic tone (increased LF/HF ratio) were also confirmed in an epidemiological study (54).

Putz et al. (53) described a mainly subclinical, asymptomatic small-fiber neuropathy, and mild impairment of cardiovascular autonomic function in IGT subjects. Similar to our present findings, HR variation and Valsalva ratio were decreased, whereas 30:15 ratio was unchanged among the tests evaluating parasympathetic activity; however, sympathetic function was also mildly impaired in patients with IGT (53). Moreover, these IGT patients also have abnormal circadian blood pressure regulation and increased diastolic blood pressure (56). Abnormal HR recovery was more common in patients with IFG (42%) and diabetes (50%) than in participants with normal glucose tolerance (31%) in a population-based Italian study; the relative risks were 1.34 (95% confidence intervals = 1.2–1.5) and 1.61 (95% CI = 1.35–1.92), respectively (57).

Fasting plasma glucose found to be an independent predictor of abnormal HR recovery (P < 0.0003) even after adjustments for other confounders (57). Moreover, impaired glucose regulation significantly (P < 0.006) correlated with adrenergic autonomic dysfunction when age, an important confounder, was removed from the model (58). The self-assessment of autonomic symptoms by patients with IGT and early diabetes correlated to the degree of autonomic dysfunction defined by abnormal 30:15 ratio and reduced quantitative sudomotor axon reflex test sweat volume (59).

The present study is the first to show that short-term QT interval variability is higher in patients with IGT. The elevated temporal STV_QT and concomitant cardiac AN may serve as early indicators of the increased instability of cardiac repolarization and elevated risk for sudden cardiac death in patients with prediabetic states.

The studies described here were carried out in accordance with the Declaration of Helsinki (2000) of the World Medical Association and were approved by the Scientific and Research Ethical Committee of the Medical Research Council at the Hungarian Ministry of Health (ETT-TUKEB), under ethical approval No. 4987-0/2010-1018EKU (338/PI/010). All subjects have given written informed consent of the study.

The authors listed below gave the following contributions: IB, AN, TV, LB, GÁ, PK, JP, AV, and CL had substantial contributions to the conception of the work and design of the paper; AO, SN, AK, ZP, RT, LB, and CL contributed to the measurements and analyses of data; AO, IB, PK, JP, AV, and CL drafted the paper or revised it critically for important intellectual content. All authors (AO, IB, SN, AK, ZP, RT, AN, TV, LB, GÁ, PK, JP, AV, and CL) have read and approved the final manuscript.

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.