Abstract P3052: Ttitle: Multi-Omic Responses to Low-Calorie Diet Intervention for Remission of Type 2 Diabetes and Individual Variability Short title: Multi-Omics, Low-Calorie Diet, and Diabetes Remission

Introduction: Remission of type 2 diabetes (T2D) can be achieved by a low-calorie diet (LCD) intervention, but molecular mechanisms remain poorly understood. Furthermore, whether responders for T2D remission and non-responders to an LCD intervention have distinct molecular responses and genetic determinants is yet clear. Methods: A 6-month LCD intervention (815-835 kcal/d) was conducted among Chinese patients with T2D who had BMI of 24-45 kg/m 2 and HbA1c level of 6.5-12.0%. Clinical biomarkers, genome, serum proteome, metabolome (serum and feces), and gut microbiome were measured at baseline and weeks 1, 12, and 24. Linear mixed models were used to identify significant changes in multi-omic markers during the 24-week intervention. A noise-augmented von Mises-Fisher mixture model was used to cluster the trajectories of significant omic markers. Different multi-omic responses between responders for T2D remission and non-responders were identified using linear mixed models, adjusting for age, sex, T2D duration, visit time, and baseline BMI as fixed effects and participant ID as random effect. Genome-wide association analysis (GWAS) was used to find single nucleotide polymorphisms (SNPs) associated with T2D remission. We used the Benjamini-Hochberg procedure to control the false discovery rate < 5% for all analyses. Results: Thirty-two participants were analyzed (mean age: 39.8 years, BMI: 29.8 kg/m 2 , HbA1c: 8.0%; 34% women). Significant changes in omic markers post intervention were observed in 46 proteins, 106 serum metabolites, 222 fecal metabolites, and 153 gut bacteria, clustering into 6-8 distinct trajectories per omics. Pathway enrichment analysis mapped these signatures to cholesterol metabolism, glycolysis, and TCA cycle pathways. Between responders and non-responders, 261 proteins, 197 serum metabolites, 212 fecal metabolites, and 171 gut bacteria differed; among these markers, 11 proteins, 27 serum metabolites, 37 fecal metabolites, and 23 gut bacteria also changed significantly post the LCD intervention. However, no SNP was associated with T2D remission. Significant associations of multi-omic markers with muscle mass, liver fat, glycemic control, insulin resistance, and insulin secretion were found. Conclusions: Dramatical changes in multi-omic markers, enriched in glucose, lipid, and energy metabolism pathways, were observed following an LCD intervention. Multi-omic differences between responders and non-responders for T2D remission were delineated.