Article Text

Download PDFPDF

Diabetes unmasked by electric shock
  1. M Kane,
  2. P G F Swift
  1. Leicester Royal Infirmary Children's Hospital, Leicester LE1 5WW, UK
  1. Correspondence to:
    Dr P G F Swift, Leicester Royal Infirmary Children's Hospital, Leicester LE1 5WW, UK;


Following an accidental electric shock, a boy with no previous symptoms developed hyperglycaemia, rapidly evolving into diabetes. He was aglycosuric for 24 hours after the shock. Islet cell antibodies were shown shortly after the accident. Although destined to develop diabetes, it seems likely that the electric shock unmasked impaired glucose tolerance. Glucose homoeostasis should be monitored in children following significant electric shocks.

  • type 1 diabetes
  • electric shock
  • BG, blood glucose
  • CNS, central nervous system
  • ECT, electroconvulsive therapy
  • ICA, islet cell antibodies

Statistics from

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.

In acute illnesses and injury of the central nervous system, raised blood glucose (BG) concentrations may occur and can be important prognostically.1, 2 Similarly, electric shocks to the brain alter the neuroendocrine axis, resulting in hyperglycaemia.3 We have found no previous reports of electric shocks precipitating diabetes in children.


A 13 year old boy scout who was mowing his neighbour's lawn, severed the lawn mower's electric cable, causing an electric shock to throw him to the ground five feet away. He sustained an electric burn to his hand. He was healthy before the accident with no significant medical or family history. He was slim and in early puberty.

No cardiac arrhythmia was recorded. Urinalysis for myoglobin was negative, as were glucose tests on two occasions the day following admission. Thirty six hours after admission, evening urinalysis showed notable glycosuria. Capillary BG was 20 mmol/l. The following morning glycosuria was absent and fasting laboratory BG was 7.6 mmol/l. The evening BG again rose to 16.4 mmol/l. Glycosylated haemoglobin (HbA1) was 7.3% (normal range 4–8.5%).

He was discharged home with instructions to monitor urine and capillary BG. Initially he remained asymptomatic but recorded intermittent glycosuria without ketonuria. Capillary BG concentrations ranged between 6 and 11 mmol/l. Over succeeding weeks he often had no glycosuria in the morning, but intermittent glucose excretion during the day with variable BG concentrations. Six weeks after the original insult he developed thirst and polyuria. After four months he lost weight, experienced nocturia, and postprandial BG concentrations were raised. The HbA1 concentration rose to 8.9% and then 9.7%. He was commenced on insulin.

A retrospective analysis of serum taken 48 hours after the electric shock showed pancreatic islet cell antibodies (ICA); when remeasured one and six weeks later, the ICA titre was 80 JDFU. Six weeks after presentation a mid morning serum insulin was 13.3 mU/l accompanied by a C peptide concentration of 1.3 nmol/l (normal range 0.14–1.39). Serum amylase and thyroid function were normal.

Two months after starting insulin (0.3 units/kg/day) he had regained 8 kg weight. Insulin dose was reduced, but subsequently increased to concentrations consistent with total insulin dependency. HbA1 returned to normal.


Can type 1 diabetes be precipitated by electric shock? This boy did not have diabetes at the time of the shock and he was aglycosuric in the initial 24 hour monitoring period. There was a rapid progression of glycaemic deterioration in association (in retrospect) with raised islet autoantibody markers, but symptoms consistent with diabetes only developed later.

High titres of complement fixing ICA (markers of β cell destruction) in the general population or in siblings of patients with diabetes have strong predictive value for disease development.4, 5 An ICA titre of 80 JDFU has a predictive value of 100% within 10 years in first degree relatives of patients with type 1 diabetes.6

The presence of high titres of ICA presumably predated the electric shock, suggesting that he was predestined to develop diabetes, although the timing and predictability were uncertain. Almost certainly the electric shock precipitated hyperglycaemia and glycosuria.

Interestingly, Claude Bernard, in 1855, showed that transfixing the brain medulla with a metal probe caused acute hyperglycaemia and glycosuria in rats, so called “picure” diabetes.7 Recently it has been postulated that thrombotic strokes in specific areas of the brain have metabolic consequences, leading to diabetes mellitus.8

Electric shock can disturb hypothalamic neuroendocrine control of glucose homoeostasis, acting via autonomic outflow to the viscera.3 These complex mechanisms are responsible for acute hyperglycaemia in response to certain central stimuli, seen in acutely ill children with central nervous system (CNS) infections, seizures, and head trauma.2 The effect of neurogenic influences on the development of diabetes is further illustrated by the occurrence of diabetes in a number of predominantly neurological genetic disorders,9 and the association of depression with insulin resistance and impaired glucose homoeostasis.10

Little is known about the effect of accidental electric shock on blood glucose control. There is controversy about the effects of electroconvulsive therapy (ECT) on glucose homoeostasis and diabetes.11 When ECT is performed on individuals without diabetes, transient hyperglycaemia may occur12; it causes minimal alterations in type 2 diabetes, but in patients with pre-existing type 1 diabetes, it may provoke dangerous hyperglycaemia.11

Hyperglycaemia following ECT has been reported in an adult without known pre-existing diabetes, but with known risk factors, who became insulin dependent.13 That case may represent the unmasking of disease as in our case.

In conclusion, it may be more than coincidence that this young boy developed type 1 diabetes after an electric shock. Electric shocks to the CNS are known to impair glucose homoeostasis, and in a predisposed person it may be sufficient to precipitate diabetes. We think it prudent, therefore, in all patients who have had a significant electric shock to monitor urinary glucose at each voiding for 48 hours, and blood glucose before and after morning and evening meals if glycosuria occurs.

Whether the risk of electric shock applies only to those already at risk of diabetes (for example, with circulating antibodies or a positive family history) remains to be seen.