Hypoglycaemia is associated with characteristic changes in the EEG with the appearance of slow frequency waves. In a recent study the investigators have shown that these changes can be recorded from subcutaneous electrodes and processed by an automated mathematical algorithm based on non-linear spectral analysis, and that changes are present before the occurrence of severe hypoglycaemia in type 1 diabetes patients. An alarm device based on real-time analysis of continuous EEG-recordings may thus be possible. For many diabetes patients nocturnal hypoglycaemia is a feared complication which may thus be preventable.

The different sleep stages are associated with specific EEG-changes of high complexity with the occurrence of slow frequency waves during stages of deep sleep.

The aim of this study is to assess EEG changes during insulin-induced hypoglycaemia in type 1 diabetes patients in the different stages of sleep. The core questions will be:

(i) Will the pathological hypoglycaemia-related EEG-changes dominate over the physiological sleep-related changes when hypoglycaemia occurs during sleep? (ii) Is it possible to refine the mathematical algorithm to an extend, where EEG-changes during hypoglycaemia can be distinguished from sleep-related EEG-changes in all sleep stages.

(iii) If so, at what blood-glucose level will hypoglycaemia associated EEG-changes be detectable and (iv) Will patients be able to react adequately by ingestion of carbohydrates if an alarm can is given at the time of hypoglycaemia associates EEG-changes.

Twelve patients with type 1 diabetes will be studied. EEG will be recorded during graded hypoglycaemia achieved by insulin infusion and frequent glucose measurements. EEG will be analysed by the automated algorithm and by visual analysis to address the questions (i), (ii) and (iii). To address question (iv) real-time EEG-analysis with a predefined threshold defining hypoglycaemia will be performed and an alarm will aim to warn the patients of impeding hypoglycaemia. Patients will be asked to consume carbohydrates at alarm.

During the experiments the patients will be under continuous observation.

Inclusion Criteria:

• Type 1 diabetes patients
• Hypoglycemia unawareness
• aged 18-70

Exclusion Criteria:

• Severe cardiac disease

  • History of myocardial infarction
  • Cardiac arrhythmia

• Previous stroke or cerebral haemorrhage and any other structural cerebral disease

• Active cancer or cancer diagnosis within the past five years

• Uremia defined as s-creatinine above 3 times upper reference value

• Liver disease defined as s-ALAT above 3 times upper reference interval

• Inability to understand the informed consent

• Epilepsy

• Use of antiepileptic drugs for any purposes

• Clinical important hearing impairment

• Use of active implantable medical device including

  • Pacemaker and ICD-unit
  • Cochlear implant

• Use of following drugs

  • Chemotherapeutic drugs of any kind
  • Methotrexate
  • Third generation antipsychotic drugs (aripiprazole, quetiapine, clozapine, ziprasidone, paliperidone, risperidone, sertindole, amisulpride, olanzapine)

• Abuse of alcohol (defined as consumption of > 250g alcohol (in Danish: 21 "genstande") per week or abuse of any other neuroactive substances

• Infection at the site of device-implantation

• Any hemorrhagic disease

• Diving (snorkel diving allowed) or parachute jumping

• Patients that are judged incapable of understanding the patient information or who are not capable of carrying through the investigation

Type 1 diabetes