A discriminant analysis of plasma metabolomics for the assessment of metabolic responsiveness to red raspberry consumption

The steady-state basal plasma glucose and insulin concentrations are determined by their interaction in a feedback loop. A computer-solved model has been used to predict the homeostatic concentrations which arise from varying degrees beta-cell deficiency and insulin resistance. Comparison of a patient's fasting values with the model's predictions allows a quantitative assessment of the contributions of insulin resistance and deficient beta-cell function to the fasting hyperglycaemia (homeostasis model assessment, HOMA). The accuracy and precision of the estimate have been determined by comparison with independent measures of insulin resistance and beta-cell function using hyperglycaemic and euglycaemic clamps and an intravenous glucose tolerance test. The estimate of insulin resistance obtained by homeostasis model assessment correlated with estimates obtained by use of the euglycaemic clamp (Rs = 0.88, p less than 0.0001), the fasting insulin concentration (Rs = 0.81, p less than 0.0001), and the hyperglycaemic clamp, (Rs = 0.69, p less than 0.01). There was no correlation with any aspect of insulin-receptor binding. The estimate of deficient beta-cell function obtained by homeostasis model assessment correlated with that derived using the hyperglycaemic clamp (Rs = 0.61, p less than 0.01) and with the estimate from the intravenous glucose tolerance test (Rs = 0.64, p less than 0.05). The low precision of the estimates from the model (coefficients of variation: 31% for insulin resistance and 32% for beta-cell deficit) limits its use, but the correlation of the model's estimates with patient data accords with the hypothesis that basal glucose and insulin interactions are largely determined by a simple feed back loop.

The advent of high throughput technologies has led to a wealth of publicly available ‘omics data coming from different sources, such as transcriptomics, proteomics, metabolomics. Combining such large-scale biological data sets can lead to the discovery of important biological insights, provided that relevant information can be extracted in a holistic manner. Current statistical approaches have been focusing on identifying small subsets of molecules (a ‘molecular signature’) to explain or predict biological conditions, but mainly for a single type of ‘omics. In addition, commonly used methods are univariate and consider each biological feature independently. We introduce mixOmics, an R package dedicated to the multivariate analysis of biological data sets with a specific focus on data exploration, dimension reduction and visualisation. By adopting a systems biology approach, the toolkit provides a wide range of methods that statistically integrate several data sets at once to probe relationships between heterogeneous ‘omics data sets. Our recent methods extend Projection to Latent Structure (PLS) models for discriminant analysis, for data integration across multiple ‘omics data or across independent studies, and for the identification of molecular signatures. We illustrate our latest mixOmics integrative frameworks for the multivariate analyses of ‘omics data available from the package.

Several methods have been proposed to evaluate insulin sensitivity from the data obtained from the oral glucose tolerance test (OGTT). However, the validity of these indices has not been rigorously evaluated by comparing them with the direct measurement of insulin sensitivity obtained with the euglycemic insulin clamp technique. In this study, we compare various insulin sensitivity indices derived from the OGTT with whole-body insulin sensitivity measured by the euglycemic insulin clamp technique. In this study, 153 subjects (66 men and 87 women, aged 18-71 years, BMI 20-65 kg/m2) with varying degrees of glucose tolerance (62 subjects with normal glucose tolerance, 31 subjects with impaired glucose tolerance, and 60 subjects with type 2 diabetes) were studied. After a 10-h overnight fast, all subjects underwent, in random order, a 75-g OGTT and a euglycemic insulin clamp, which was performed with the infusion of [3-3H]glucose. The indices of insulin sensitivity derived from OGTT data and the euglycemic insulin clamp were compared by correlation analysis. The mean plasma glucose concentration divided by the mean plasma insulin concentration during the OGTT displayed no correlation with the rate of whole-body glucose disposal during the euglycemic insulin clamp (r = -0.02, NS). From the OGTT, we developed an index of whole-body insulin sensitivity (10,000/square root of [fasting glucose x fasting insulin] x [mean glucose x mean insulin during OGTT]), which is highly correlated (r = 0.73, P < 0.0001) with the rate of whole-body glucose disposal during the euglycemic insulin clamp. Previous methods used to derive an index of insulin sensitivity from the OGTT have relied on the ratio of plasma glucose to insulin concentration during the OGTT. Our results demonstrate the limitations of such an approach. We have derived a novel estimate of insulin sensitivity that is simple to calculate and provides a reasonable approximation of whole-body insulin sensitivity from the OGTT.