1698-P: Partial Least Squares Discrimination Analysis of Cytokine Profiles in T Cells with Inflammation in Subjects with Obesity

Introduction & Objective: We aims to use Partial Least Squares-Discrimination Analysis (PLS-DA) to analyze cytokine profiles in T cells in the context of obesity.

Methods: PLS-DA is a robust statistical method that facilitates the identification of intricate patterns within complex cytokine data sets. We applied PLS-DA to investigate combinatorial cytokine profiles that differentiate T cells of old and young subjects with obesity. Additionally, we evaluated the effect of metformin on these profiles, focusing on its potential impact across age groups in individuals with obesity.

Results: Our PLS-DA analysis distinguished cytokine profiles produced by T cells from old and young subjects with obesity with a 74% accuracy. PLS-DA indicated a borderline effect of metformin on cytokine profiles generated by T cells from older subjects with obesity, with a model accuracy of 67%. Lastly, PLS-DA indicated no impact of metformin on cytokine profiles produced by T cells from younger subjects with obesity, with a model accuracy of 52%. The outcomes contrast with the strong impact of metformin on cytokine profiles produced by T cells from lean older and younger adults (Bharath et al, Cell Metab 2020).

Conclusion: PLS-DA analysis revealed differences in cytokine profiles among old and young subjects with obesity. This analytical method highlighted the differential impact of metformin on cells from these groups, showing a moderate effect on cells from older subjects with obesity and no significant impact on cells from younger subjects with obesity. These findings emphasize the importance of age and obesity status in cytokine profile alterations and impact of these variables on efficacy of treatments like metformin.