A pharmacist's perspective on health and metabolic disease
I presented my thesis at the AUT Postgraduate Symposium (2015). Here is the abstract of my presentation and a link to the video that was posted to Youtube .
“Hyperinsulinaemia, or chronically high levels of plasma insulin, heralds the onset of many non-communicable diseases including type 2 diabetes, vascular diseases, certain cancers, and dementia. This condition may affect a substantial proportion of population with no obvious signs or symptoms making the hyperinsulinaemia a “silent” condition. The difficulty with hyperinsulinaemia is the absence of a simple and reliable diagnostic test; fasting insulin is notoriously variable – even under controlled conditions. This presentation reports on my completed empirical research in the area understanding the risk of hyperinsulinaemia and recommending a reliable diagnostic test under the supervision of Prof Grant Schofield and Drs Caryn Zinn and Mark Wheldon.
During the course of my thesis I have conducted extensive literature reviews, secondary data analysis and primary data collection to draw together an understanding of hyperinsulinaemia and how it should be diagnosed.
My final study, and the topic of this presentation, considered whether the hyperinsulinaemia testing process could be simplified. The current testing process requires a fasting blood test, followed by consuming 100g glucose, then further blood tests at 30, 60, 120 and 180 minutes following the glucose load; a process with significant cost and time demands. Using the results of more than 7000 people who had this test performed, I could dichotomise these people into normal or high insulin tolerance patterns. Then, using current clinical stratification, followed by sensitivity and specificity modelling, I developed an algorithm that determines the likelihood that a person is hyperinsulinaemic.
The results show that if a person has an elevated glycosylated haemoglobin (HbA1c) or fasting plasma insulin > 30 µU/mL, hyperinsulinaemia can be assumed. In people with normal glucose tolerance and fasting insulin ≤ 30 µU/mL, a 2-hour plasma insulin level ≥ 30 µU/mL should be considered diagnostic for hyperinsulinaemia (99% sensitivity and 62% specificity). Using a 2-hr insulin cut-off of 50 µU/mL the sensitivity and specificity became 72% and 99% respectively.
In the absence of an elevated HbA1c or fasting plasma insulin, a 2-hour plasma insulin > 30 µU/mL should be used to diagnose hyperinsulinaemia. Although the 62% specificity means that we may erroneously diagnose healthy people, the first line treatment is lifestyle change. This is preferred compared to wrongly concluding people with hyperinsulinaemia are healthy.
A 2-hr, 100g, oral glucose tolerance test with insulin assays should be the preferred test for clinical practice and future research in hyperinsulinaemia related conditions.”