It is difficult to determine a frequency of hypoglycaemia without first defining what is meant by 'hypoglycaemia'. In many studies, hypoglycaemia is documented by self-reporting, which may be very unreliable (Heller et al., 1995). Retrospective analyses suffer from problems of recall, and accurate documentation of hypoglycaemia is obtained only in prospective research studies that require biochemical verification of low blood glucose concentrations (see Chapter 3).
Hypoglycaemia can be categorised by its symptomatology and its severity, but no real consensus exists. 'Mild' hypoglycaemia is usually defined as an episode that a person recognises and treats themselves and does not significantly disrupt daily living; 'severe' hypo-glycaemia is an episode in which blood glucose has fallen to a level where the patient has become so disabled that assistance is required from another person (The Diabetes Control and Complications Trial Research Group, 1991). Alternatively, 'severe' hypoglycaemia may be defined by the requirement for parenteral treatment (intramuscular glucagon or intravenous dextrose), with or without hospital admission, or by the development of coma (The Diabetes Control and Complications Trial Research Group, 1987). A category of 'moderate' hypoglycaemia, in which an individual requires external assistance but which falls short of requiring parenteral therapy or developing a coma, or the division of hypoglycaemia into grades of severity has also been used (Limbert et al., 1993). Obviously the definition used will affect the estimate of incidence, and if severe hypoglycaemia is defined solely as coma, rates will be lower than if all episodes requiring assistance are included.
Various levels of blood glucose concentration have been used to define mild and moderate hypoglycaemia biochemically, but there is now recognition that the blood glucose concentrations which have been used arbitrarily to define pathological 'spontaneous' hypoglycaemia (such as < 2.2 mmol/l) are unsuitable for defining hypoglycaemia in people with diabetes. An arterialised plasma glucose concentration of around 3.6 mmol/l may be sufficient to cause physiological autonomic responses in healthy volunteers (Chapter 1). Subtle changes in cognitive function can initially be detected by formal testing below 4.0 mmol/l, although clinically relevant cognitive impairment does not occur until the arterialised plasma glucose concentration has fallen to approximately 3.0 mmol/l (Chapter 2). There is evidence that if plasma glucose falls below 3.0 mmol/l for a period of time, this can reduce the symptomatic responses to a further episode of hypoglycaemia occurring within the following 24 hours (Heller and Cryer, 1991), so-called antecedent hypoglycaemia (Chapter 7). The continuous avoidance of low blood glucose concentrations (below 3.0 mmol/l) can restore normal hormonal and symptomatic responses to hypoglycaemia (Fanelli et al., 1993; Cranston et al., 1994), and it follows that values below 3.0 mmol/l can be considered to be hypoglycaemic. Diabetes UK coined the phrase 'make four the floor' to protect against potentially dangerous hypoglycaemia and, most recently, a panel convened by the American Diabetes Association concluded in favour of a concentration of 3.9 mmol/l (Working Group on Hypoglycemia, American Diabetes Association, 2005). This was based on research data, showing evidence of counterregulatory responses at blood glucose concentrations (often arterial or arterialised) of 3.9 mmol/l and the demonstration of a small reduction in the glucagon response to hypoglycaemia induced immediately afterwards. However, this defines 'hypoglycaemia' at blood glucose concentrations that commonly occur in healthy non-diabetic people and may therefore encourage over-treatment. It is probably more appropriate to differentiate between targets for adjusting therapy, which may properly be above 4.0 mmol/l, and 'hypoglycaemia', which implies a pathological cause and requires intervention.
Was this article helpful?