Glycemia and Corrected QT Interval Prolongation in Young Type 1 Diabetic Patients: What is the relation?

Nine children and adolescents with stable type 1 diabetes (five males and four females aged 9–19 years) were prospectively recruited from the patient population that regularly attends the Diabetes Clinic for Children and Adolescents of the Antwerp University Hospital. Diabetes duration varied between 1 and 9 years with a median of 6 years. All of the patients were treated with a basal-bolus insulin regimen with at least four subcutaneous injections daily. Exclusion criteria were the presence of other disorders, medication known to modify QT, obvious clinical complications or signs of persistent microalbuminuria (albumine excretion >15 μg/min in an overnight timed urine collection), and neuropathy or proliferative retinopathy.

A 24-h electrocardiogram (Dynacord 423; Delmar-Reynolds, Hertford, U.K.) was recorded in each patient, with automatic QT, RR interval, and QTc measurements (QTc = QT/√ RR). The automatic electrocardiogram analysis provided hourly mean values for heart rate, QT, and QTc together with minimum and maximum values. Simultaneously, interstitial glucose concentrations (IGC) were measured via a subcutaneous abdominal catheter using a continuous glucose monitoring system (CGMS) (Minimed; Medtronic, Minneapolis). This technique provides 12 glucose measurements hourly. The system was calibrated first against standard glucometer values.

Calculations were performed using SPSS v.11 statistical software. The Holter datasets are presented as means with minimum and maximum values. Spearman’s correlation analysis was used to evaluate the relation between QTc values and glucose concentrations for the patients individually. P ≤ 0.05 was considered statistically significant.

Holter data were available in all nine patients. CGMS data were lost in one patient due to a technical problem. High values for QT and QTc were found in these patients, with a percent QTc >440 ms up to 56% per hour in one patient and a percent QTc >460 ms up to 31% per hour in another patient. QTc values >550 ms were measured in four of nine patients. The lowest IGC values were found during the night and early morning and the highest concentrations at daytime. The opposite curve was observed for the QTc pattern over 24 h, with the higher QTc values at night and in the early morning (e.g., patient 3 in Fig. 1A).

When the correlations were calculated for the patients individually, a strongly significant correlation between QT and glycemia was found in five of eight patients and between QTc and QTc_max and glycemia in three of eight patients (Fig. 1B).

In this study, we confirmed that young type 1 diabetic patients have important QTc prolongation, with QTc values up to 578 ms. QTc values above the normal limit of 440 ms were frequent (up to 56% per hour in one patient and >460 ms up to 31% per hour in another patient). The critical QTc value that confers particular vulnerability to ventricular arrhythmias probably varies individually but is considered to be ∼550 ms (15). Four of nine patients in this study exceeded this threshold.

We also evaluated the relation between IGC and QT or QTc combining Holter recordings and continuous glucose monitoring with an abdominal subcutaneous catheter. A significant correlation was found between QT, QTc, and QTc_max and IGC in some patients, with the higher QT and QTc values coinciding with lower glucose readings.

Although CGMS readings are imperfect reflections of glycemia (due to time lag, intersensor variability, and reliability in identifying true hypoglycemia), they make it possible to measure the relationship of glucose to QT over longer periods without disturbing the sleep pattern (16). Our results are in agreement with other reports that describe an abnormal repolarization during both experimental and spontaneous clinical nocturnal hypoglycemia (7–9,17). Those data indicated that hypoglycemia can cause an acquired long QT syndrome. Some studies associated unexpected sudden deaths in young diabetic patients (the “dead in bed” syndrome) with cardiac dysrhythmias following abnormal cardiac repolarization during hypoglycemia (10–13). It is not clear whether hypoglycemia and the concomitant reactive adrenergic stimulation per se or rather insulin itself creates this prolonged repolarization. Recently, Robinson et al. (8) induced an acquired long QT syndrome during experimental hypoglycemia in normal men that could be prevented by β-blockade rather independently of extracellular potassium. Murphy et al. (9) showed in a young type 1 diabetic population that a prolonged QTc occurred more frequently during spontaneous overnight hypoglycemia, possibly related to insulin-induced hypokalemia. In our study with simultaneous IGC and QTc measurements, we confirm that the highest QTc values are seen at night and in the early morning, when glucose values are low, due to the treatment scheme. The “dead in bed” syndrome could be explained by a physiologically longer cardiac repolarization at night and thus an electrically instable myocardium, worsened by the hypoglycemia-induced adrenergic stimulation (18), maybe in combination with insulin-induced hypokalemia. When extra medication that prolongs the QT interval or alcohol is taken, the risk for a potentially lethal degree of QT prolongation and fatal arrhythmias will increase. Further work in a larger population is absolutely necessary to explain a mechanism that possibly causes sudden death in young people.

This study was supported by the National Foundation for Research in Pediatric Cardiology.