Arrhythmias and diabetes: heart block, ventricular arrhythmias, and death

Abstract

In 1942, a 58-year-old dentist with insulin-treated diabetes was admitted to a Jerusalem hospital with a history of episodic fainting and convulsions which were attributed to epilepsy. ‘[P]rior to his admission to the Medical Division, insulin was decreased from 40 to 30 units a day because of a continuous fall in blood sugar. The evening injection…was given from an ampoule which contained 60 units per c.c. instead of the usual 20 units; as this difference in content was accidentally overlooked the patient got 45 instead of 15 units. Sometime later, when feeling that he was “suffering from insulin shock,” he took some sweets and glucose. Shortly afterwards a continuous series of convulsive seizures set in accompanied by complete loss of consciousness and of control of the sphincters. During these attacks a very slow pulse was observed by his house physician. Atropine, camphor, glucose intravenously, and ephedrine were subsequently administered but proved to be of little avail. On the following morning when the attacks finally stopped, his blood sugar was 597 mg. per 100 c.c. [33.2mmol/l].’ Despite resolution of the hypoglycaemia, his heart rate remained 26–28 bpm. His ECG showed complete heart block which eventually persisted. He died 10 months later, after many collapses and convulsions, with increasing heart block and episodes of asystole.1 Errors in diabetes drugs in hospital are common. Although there have been improvements, insulin and other errors still occur.2 Having diabetes increases the likelihood of heart block. A US study compared 293,124 people with diabetes, among whom 1.1% had a record of third degree heart block, with 552,623 people with hypertension but not diabetes, among whom 0.6% had heart block; odds ratio 3.1 (3.0 to 3.3).3 A Swedish study compared the need for a cardiac pacemaker in 416,247 people with type 2 diabetes with 2,081,235 age- and sex-matched people without diabetes. Having diabetes increased pacemaker need; HR 1.56 (1.51–1.60) after correcting for confounders. Risk increased by increasing age, HbA1c, BMI and diabetes duration, and by requiring lipid- and blood pressure- lowering medication.4 Obstructive sleep apnoea is common among people with type 2 diabetes. About 5–10% of people with sleep apnoea have heart block, and sometimes asystole. In sleep studies of 16 people with sleep apnoea and nocturnal heart block without electrophysiological conduction abnormalities, there were 651 episodes of heart block. Over 90% of episodes were during oxygen desaturation of 4% or more. There were only 79 episodes during continuous positive airway pressure (CPAP) treatment.5 Sleep apnoea is also associated with frequent premature ventricular beats, and non-sustained ventricular tachycardia.6 In India, a 62-year-old man with type 2 diabetes had dizziness due to heart block. Both his brothers had type 2 diabetes and heart block, as did their mother who died of a stroke aged 75. The family could not afford detailed tests. The authors postulate that they had a variant of maternally- inherited mitochondrial disorder, for example, maternally- inherited diabetes and deafness (MIDD) or mitochondrial encephalopathy with lactic acidosis and stroke-like episodes (MELAS).7 Variable clinical features include diabetes and cardiac conduction disorders.8 The Jerusalem dentist's underlying heart block was revealed in an episode of hypoglycaemia which probably exacerbated the conduction problem. Glucose and ECG were monitored continuously in 25 people with type 1 diabetes for two separate 24-hour periods. In 8/13 episodes of nocturnal hypoglycaemia, arrhythmias (excluding sinus tachycardia) included sinus bradycardia, ventricular or atrial ectopic beats, and P-wave abnormalities. The QTc interval (a measure of ventricular depolarisation and repolarisation) was prolonged.9 Among 42 people with type 2 diabetes who had parallel continuous glucose monitoring and ECG monitoring for five days, 30 people were on insulin and/or sulphonylurea and 12 were on metformin and/or DPP-4 inhibitors. All 12 of the people with glucose <3.1mmol/L were on insulin and/or sulphonylurea. People with hypoglycaemia had more ventricular ectopic beats, couplets and triplets than those without; there were five episodes of ventricular tachycardia. Patients felt few of these dysrhythmias which were mostly at night.10 In rat studies, severe hypoglycaemia prolongs QTc interval, followed by atrial and ventricular ectopic beats, non-sustained ventricular tachycardia, increasing heart block, and cardiac arrest.11 In clinical studies, QTc prolongation may be increased by the duration of hypoglycaemia and the rate at which plasma glucose falls, and longer hypoglycaemia (for example, at night) may be associated with bradydysrhythmias. Cardiac autonomic neuropathy, previous hypoglycaemia, beta blockers, and duration of type 1 diabetes appear to reduce the prolongation of QTc, favouring ventricular arrhythmias.12 Hypoglycaemia is frequent among people with diabetes on insulin, and may occur in those on sulphonylureas. Hypoglycaemia is disturbing, disabling, and dangerous. Up to 10% of deaths among people with type 1 diabetes are due to hypoglycaemia. Educate and support patients to avoid hypos especially at night.13 In people with an acute myocardial infarct, hyperglycaemia on admission (with or without known diabetes) increases the risk of early ventricular tachycardia. A US study of 4140 patients with acute myocardial infarction found on multivariate analysis that those with a serum glucose of ≥140mg/dl (5.7mmol/L) had an adjusted OR of 1.48 (1.23–1.78) for ventricular tachycardia compared with those with lower glucose levels.14 People with diabetes are two to four times more likely to have a sudden cardiac arrest out of hospital than people without diabetes. The most common cause is ventricular fibrillation. Among people with diabetes the sudden cardiac arrest incidence rate per 1000 patient-years in those without previously clinically-recognised heart disease was 3.15 (2.39–4.08), compared with 13.80 (11.96–15.84) among people with clinically-recognised heart disease. After correcting for multiple risk factors, people with type 2 diabetes with a prolonged QT interval in the upper quartile had a 2.8 times risk of sudden cardiac arrest compared with people with diabetes in the lowest quartile.15 Cardiac autonomic neuropathy (CAN) is underdiagnosed. It causes persistent resting tachycardia (>90 bpm), postural hypotension, and arrhythmias. A Canadian Consensus panel has published diagnostic guidance.16 The prevalence of CAN varies from 17–90% in patients with type 1 diabetes and 20–73% in patients with type 2 diabetes, depending on groups studied and diagnostic techniques used. Age, diabetes duration, poor glycaemic control, obesity, smoking, hypertension, and the presence of other diabetes tissue damage increase the risk of CAN.17 CAN may cause atrial dysrhythmias such as atrial fibrillation, but also life-threatening ventricular dysrhythmias and sudden death. In a meta-analysis, people with diabetic CAN had an increased pooled relative mortality risk of 2.14 (1.83–2.51) compared with people without evidence of CAN.18 The Canadian group state: ‘Suitable candidates for CAN screening are asymptomatic type 2 diabetic patients at diagnosis and type 1 diabetic patients after 5 years of disease, in particular those at greater risk for CAN due to a history of poor glycaemic control (haemoglobin A1c >7% [53mmol/mol]), or the presence of one major cardiovascular risk factor (among hypertension, dyslipidaemia, and smoking), or the presence of macro- or microangiopathic complications (level B). CAN screening may be also required in asymptomatic patients for pre-operative risk assessment before major surgical procedures.’16 The deaths of people with type 1 diabetes aged <50 years who had died suddenly and unexpectedly were studied: ‘The most puzzling group were 22, aged 12–43 years, most of whom had gone to bed in apparently good health and been found dead in the morning. Nineteen of the 22 were sleeping alone at the time of death and 20 were found lying in an undisturbed bed. Most had uncomplicated diabetes and in none were anatomical lesions found at autopsy.’ This has been called ‘dead in bed’ syndrome. It still happens. While hypoglycaemia would seem the most likely cause, no proof of this was found. Nowadays, many people with type 1 diabetes use continuous/flash glucose monitoring so, hopefully, nocturnal deaths due to hypoglycaemia will reduce.19 Having diabetes increases the likelihood of heart block. Sleep apnoea is common among people with type 2 diabetes; night-time desaturations may be associated with heart block or ventricular arrhythmias. These problems are improved with CPAP. Prevent hypoglycaemia: among other dangers, it prolongs QTc and may cause bradycardias, ventricular arrhythmias, or death. People with diabetes have a worrying risk of sudden death. Prevention and early detection of coronary artery disease are important. Also, prevent, seek, and treat arrhythmias, and other cardiac disease. Seek cardiac autonomic neuropathy – patients with this are at particular risk of sudden death. Check your patients’ pulse, and teach them how to do it themselves.