Our data suggest that the non-invasive real-time BreathID® GBT reliably assesses changes in the liver glucose metabolism, and may serve as a tool for determining insulin resistance. High within-subject correlations between serum glucose and GBT were detected. The GBT parameters highly correlated with serum insulin levels.
In order to assess the sensitivity of the GBT for detecting insulin resistance, the study was performed in healthy volunteers after exercise. Exercise is a key component for the successful management of many obesity-related metabolic complications, including insulin resistance.19
Both chronic and acute endurance exercise has an effect on insulin action in obesity. Exercise-induced alterations in fatty acid partitioning within muscle cells affect insulin sensitivity [19
]. GBT and OGTT were repeated following physical exercise. High between-subjects correlations were found between the insulin levels ratio and the PDR 90 ratio before and after exercise, and between the ratio of glucose levels and the CPDR 60 ratio before and after exercise. Correlations were identified before exercise between the HOMA IR at 90 min and the PDR at 90 min, and between the HOMA B at 120 min and the CPDR at 30 min. Correlations were identified after exercise between the HOMA IR at 60 min and the peak time, and between the HOMA B at 150 min and the PDR at 150 min. This data suggests that the GBT may serve as a valid tool for detection of mild alterations in insulin resistance.
Currently available methods for assessment of insulin resistance are invasive and cumbersome, making them impractical for use at the point-of-care. The gold standard diagnostic test for insulin resistance is the hyperinsulinemic-euglycemic clamp, but this method is unsuitable for everyday clinical use. The hyperinsulinemic euglycemic clamp and the insulin suppression test, which are both labour and time intensive, directly assess insulin-mediated glucose utilization under steady-state conditions [20
]. The degree of insulin resistance is inversely proportional to the glucose uptake by target tissues during the procedure. A slightly less complex indirect method relies on minimal model analysis of a frequently sampled intravenous glucose tolerance test. The insulin sensitivity test (IST) and the Insulin tolerance test (ITT) measure the decline in serum glucose after an IV bolus of regular insulin. They primarily assess the insulin-stimulated uptake of glucose into skeletal muscle [21
]. ISI is calculated for fat-free body mass by dividing the glucose disposal rate by the average plasma insulin concentration [6
Surrogate indices for insulin sensitivity/resistance, including QUICKI, HOMA, 1/insulin, and Matusda index, are all derived from blood insulin and glucose concentrations under fasting conditions (steady state) or after an oral glucose load (dynamic) [6
]. Their relatively low sensitivity, the time required for their performance, and patient inconvenience make them unlikely to become point-of-care screening tests.
Oral glucose tolerance tests (OGTTs), are a mainstay for assessing insulin sensitivity [21
] in the non-invasive diagnosis of impaired glucose tolerance (IGT) and diabetes mellitus. A modified OGTT uses a 75 mg glucose load and measures glucose and insulin at various intervals over two to four hours. OGTT provides information on beta cell secretion and peripheral insulin action. Insulin sensitivity has been assessed by calculating insulin area under the curve (AUCinsulin
), by the AUCglucose
ratio, and by an insulin sensitivity index (ISI) that uses glucose and insulin values from 0 and 180 min in a mathematical formula. A more qualitative assessment of insulin resistance is the observation of one or more insulin values exceeding an upper, normal, limit at appropriate intervals [25
The search for simple and inexpensive quantitative tools to evaluate insulin sensitivity has led to development of fasting state (homeostatic) assessments. These tests are based on fasting glucose and fasting insulin, and a calculation to assess insulin sensitivity and beta cell function. The homeostasis model assessment (HOMA) and quantitative insulin sensitivity check index (QUICKI) have been devised, [3
] and may be applied to normoglycemic and hyperglycemic patients. Fasting insulin (I0), and the glucose/insulin (G/I) ratio are inexpensive assays for calculations of insulin resistance [6
]. We found a high correlation between the QUICKI score and the PDR at 120 minutes (before exercise), peak time (before exercise), and PDR 90 minutes (post exercise).
Homeostatic model assessment (HOMA) has been widely employed in clinical research to assess insulin sensitivity. Unlike the I0 and G/I ratio, HOMA calculations compensate for fasting hyperglycemia [6
]. HOMA values correlate well with the results of clamp techniques, and have been frequently used to assess changes in insulin sensitivity after treatment [26
]. However, this model assumes that beta cell function is normal and does not apply to patients with type 2 diabetes. One of the weaknesses of these models is that they assume the relationship between glucose and insulin is linear, when in fact it's parabolic. Another weakness is that they assume that beta cell function is normal and do not apply to patients with type 2 diabetes. These indices of insulin resistance necessitate serum insulin and glucose measurements, may require complex calculations, and have not yet made major inroads into general medical practice. None of these can be performed at the point-of-care, and their sensitivity varies over patient populations and BMIs.
Our pre exercise data shows high correlations between HOMA IR 90 min and PDR 90. Our post exercise data shows high correlations between HOMA B 120 and CPDR 30 min, between HOMA IR 60 min and peak time post exercise, and between HOMA B 150 min and PDR 150 min.
Previous studies suggested the use of GBT in patients with diabetes [28
]. A comparison of the 13
C GBT using mass spectrometry with the hyperinsulinemic-euglycemic clamp to determine insulin resistance has been performed, and suggests a high correlation between the results of the two tests [28
]. High correlations between 13
C GBT parameters and glucose metabolism and insulin sensitivity indices from insulin clamp measurements were described. The magnitudes of these correlations compared favourably with QUICKI and were superior to the homeostasis model assessment. However, the use of mass spectrometer in this setting is cumbersome and does not enable the use of the test as a daily decision making tool.
GBT using an office-based device offers several advantages over the currently available techniques for assessment of insulin resistance. It is a non-invasive, point-of-care test. It is not operator dependent, and its sensitivity seems higher than the currently used tests. Currently available methods are either invasive, or not sensitive enough to detect insulin resistance at early stages, or to follow-up treatment. Target populations for screening may include patients with metabolic syndrome, pregnant women, patients with NAFLD, patients at early stages of diabetes, and chronic HCV patients without overt diabetes [29
Although we studied a relative small population of subjects, the data of the present study suggests that the GBT can serve as a non-invasive tool for dynamic evaluation of glucose metabolism, for early diagnosis and for follow-up of patients in groups at high-risk of insulin resistance.