Identification of lipid-defined profiles and their associations with glucose metabolism

Current evidence suggests that certain lipids are associated with specific parameters of glucose metabolism. However, a comprehensive understanding of these associations - especially regarding detailed lipid signatures - is still lacking. Therefore, the aim of this study was to identify lipid-defined profiles based on NMR-derived lipidomics data and to investigate their cluster-specific associations with glucose metabolism.

In this cross-sectional study, a total of 34 lipids measured in serum samples of 586 participants from the MEIA study were subjected to a preprocessing step involving principal component analysis followed by K-means clustering. Based on the resulting profiles, associations with the glucose-related parameters fasting glucose, 2-h post-load glucose from an oral glucose tolerance test, and HbA1c were investigated using a doubly robust estimation approach. Potential confounders were selected using a directed acyclic graph.

Following lipid-defined profiles were identified: (1) severe atherogenic, (2) hyperlipidemic, (3) phospholipid-enriched HDL-dominant, and (4) optimal. Considering the first profile as the reference, the fourth profile was associated with lower fasting glucose levels (β=-0.05; 95% CI [-0.08; -0.02]; P_adj=0.013) as well as lower HbA1c values (β=-0.07; 95% CI [-0.11; -0.03]; P_adj=0.005) on the log-scale. Additionally, the third profile was associated with 2-h glucose levels (β=-0.12; 95% CI [-0.21; -0.04]; P_adj=0.040). Sensitivity analyses confirmed these results.

Depending on the specific lipid-defined profile, differences in glucose metabolism were found. These findings suggest that lipid-defined patterns might capture clinically relevant variation in glucose metabolism beyond individual lipid parameters and may improve risk assessment for glucose metabolism disorders.