Insulin measurement is also inappropriate in paediatric obesity

To the Editor: I was pleased to read the article discouraging the measurement of insulin levels in the metabolic syndrome and obesity,1 as it concurs with my previous opinion.2 In parallel with the adult burden is the ever increasing paediatric obesity epidemic, which looms large across the world, with rates in Australia of up to 25%.3 This translates into a significant number of oral glucose tolerance tests being performed, often with measurement of insulin levels.

There is a distinct lack of evidence on the validity of this test in obese older children and adolescents in terms of defining cardiovascular and metabolic morbidity and mortality. The implementation and inter-pretation often stem from extrapolation of adult data.4 Furthermore, the administration of glucose to this age group is often impractical, as it is weight-based (1.75 g/kg, to a maximum dose of 75 g), and may induce morbidities such as nausea and vomiting, with subsequent failure to complete the test, multiple traumatic venepunctures, and unwarranted stress (both financially and emotionally) on the parents. Individual variations in gastric emptying and insulin secretion rate add to the poor accuracy and reproducibility of this so-called diagnostic test. If sufficient care is not taken, the analysis and processing of the insulin assay can give an incorrect result, and thus a low level can be misguiding when clinical findings indicate otherwise. A high level merely confirms the syndrome where body mass indices and waist circumferences are equally valid measurements.5 Either way, this test cannot be recommended as routine, and management should be based on clinical features. Although not flawless, perhaps fasting plasma glucose levels are more appropriate as diagnostic tools, with oral glucose tolerance tests reserved for high-risk and atypical groups. These tests should be done without measuring insulin levels and with the previously mentioned confounders in mind.

Although it is recognised that insulin resistance is central to the disease clustering seen in the metabolic syndrome, unless the syndromal terms that bear the subtext “insulin” are renamed, insulin testing in clinical practice will continue unabated at a costly rate. In contrast to international bodies6,7 and this well-founded opinion,1 the National Health and Medical Research Council (NHMRC) still recommends insulin measurement,8 albeit in selected circumstances.