Accessibility Skip to Global Navigation Skip to Local Navigation Skip to Content Skip to Search Skip to Site Map Menu

Assessing hyperglycaemia in patients with chronic liver disease: HbA1c versus interstitial glucose

A 2019/2020 Summer Studentship research project

In the general population, HbA1c is thought to be a good ‘mirror’ of average glucose values and is therefore used as the standard HbA1c test to screen for diabetes. The HbA1c test may not however be valid in patients with severe chronic liver disease and might give falsely normal and reassuring results in people who do in fact have diabetes requiring treatment. Patients with severe liver disease may therefore need to be screened for diabetes using alternative approaches. The ‘gold standard’ oral glucose tolerance test is however time consuming for the patient and many patients find the glucose drink needed for this test has an unpleasant taste. Could interstitial glucose results be used as an alternative way of making a diagnosis of abnormally elevated glucose levels?

Student: Cicely Barron
Supervisors: Professor Helen Lunt, Dr Jeffery Ngu, Gastroenterologist; Dr Huan Chan, Endocrinologist; Helen Heenan, Diabetes research nurse; Judith McLaughlin, Gastroenterology research nurse.
Sponsor: Diabetes Christchurch Inc & the Diabetes Training and Research Trust


Chronic liver disease including cirrhosis may be associated with an HbA1c test (a diabetes blood test that looks indirectly at long term glucose values) that underestimates true (venous or interstitial) glucose. In patients on anti-retroviral agents, this finding is likely to be due to haemolysis and associated changes in red cell turnover. In patients with cirrhosis but not on anti-retrovirals, the mechanism for this ‘artefactual’ low HbA1c is poorly understood. Currently, has been suggested that the 30% of patients with cirrhosis and diabetes (much of which is undiagnosed), be diagnosed using an oral glucose tolerance test, rather than an HbA1c blood test. A better understanding of possible discordance between HbA1c and venous glucose value in patients with chronic liver disease, will aid the diagnosis and management of pre-diabetes and diabetes in this population. This feasibility study also explores whether or not interstitial glucose can be used as an alternative method of diagnosing diabetes and impaired glucose tolerance, in this population.


The primary aim of this feasibility study is to determine how well HbA1c correlates with; i) the ‘gold standard’ oral glucose tolerance test and also ii) two weeks of interstitial glucose values, in a group of patients with severe chronic liver disease (biopsy Stage 4 or ≥12.5kpa on Fibroscan). Gastroenterology patients will be recruited who have no diagnosis of diabetes, but have some laboratory evidence of high-normal glucose values.


Twenty-five participants with evidence of elevated glucose levels on recent laboratory tests, will be selected from a local database of 600+ patients with well-characterised cirrhosis, undergoing regular cirrhosis surveillance screening at the gastroenterology clinic. Patients with co-morbidities known to affect red cell turnover and patients on anti-retroviral agents will be excluded from the study.

Participants will spend half a day in clinic, undergoing laboratory tests including an oral glucose tolerance test, haptoglobin measurements and HbA1c. A Libre flash interstitial glucose sensor, ‘Libre Pro’, will be inserted. This device was chosen because it requires no patient/participant maintenance after insertion and, if functioning well, provides two weeks of continuous glucose values. At the end of two weeks of self-monitoring of interstitial glucose, participants will return the disposable sensor to the clinic by post (mail). Interstitial glucose values will be downloaded from the sensor’s hardware, for analysis. The magnitude of the difference between estimated HbA1c glucose values, the values from the oral glucose tolerance test and also interstitial glucose, will be compared. Also, the utility of the Libre Pro will be assessed, for example how many sensors provided a full two weeks of glucose results?

Student researcher’s component of the study

The student will be expected to ‘project manage’ the study but will receive practical help from the research nurses, for example in relation to the insertion of the Libre Pro. (This task is simple. To place it in context, most patients starting on a ‘sister’ Libre device self-manage insertion, after reviewing a YouTube instruction clip). The student will: a) identify patients on the cirrhosis database with elevated laboratory glucose values, b) lead the recruitment of these patients and c) coordinate participant visits and data management, with clinical staff. The student will have support from staff in the Departments of Gastroenterology and Endocrinology/Diabetes. The student will help with data analysis, which will be semi descriptive and will be done in conjunction with a biostatistician.

Student Prerequisites

Medical Student

How to apply