Obesity plays an important role in the development of insulin resistance and diabetes, but the molecular mechanism that links obesity and diabetes is still not completely understood. Here, we used 146 targeted metabolomic profiles from the German KORA FF4 cohort consisting of 1715 participants and associated them with obesity and type 2 diabetes. In the basic model, 83 and 51 metabolites were significantly associated with body mass index (BMI) and T2D, respectively. Those metabolites are branched-chain amino acids, acylcarnitines, lysophospholipids, or phosphatidylcholines. In the full model, 42 and 3 metabolites were significantly associated with BMI and T2D, respectively, and replicate findings in the previous studies. Sobel mediation testing suggests that the effect of BMI on T2D might be mediated via lipids such as sphingomyelin (SM) C16:1, SM C18:1 and diacylphosphatidylcholine (PC aa) C38:3. Moreover, mendelian randomization suggests a causal relationship that BMI causes the change of SM C16:1 and PC aa C38:3, and the change of SM C16:1, SM C18:1, and PC aa C38:3 contribute to T2D incident. Biological pathway analysis in combination with genetics and mice experiments indicate that downregulation of sphingolipid or upregulation of phosphatidylcholine metabolism is a causal factor in early-stage T2D pathophysiology. Our findings indicate that metabolites like SM C16:1, SM C18:1, and PC aa C38:3 mediate the effect of BMI on T2D and elucidate their role in obesity related T2D pathologies.

Aims: A data‐driven cluster analysis in a cohort of European individuals with type 2 diabetes (T2D) has previously identified four subgroups based on clinical characteristics. In the current study, we performed a comprehensive statistical assessment to (1) replicate the above‐mentioned original clusters; (2) derive de novo T2D subphenotypes in the Kooperative Gesundheitsforschung in der Region Augsburg (KORA) cohort and (3) describe underlying genetic risk and diabetes complications. Methods: We used data from n = 301 individuals with T2D from KORA FF4 study (Southern Germany). Original cluster replication was assessed forcing k = 4 clusters using three different hyperparameter combinations. De novo clusters were derived by open k ‐means analysis. Stability of de novo clusters was assessed by assignment congruence over different variable sets and Jaccard indices. Distribution of polygenic risk scores and diabetes complications in the respective clusters were described as an indication of underlying heterogeneity. Results: Original clusters did not replicate well, indicated by substantially different assignment frequencies and cluster characteristics between the original and current sample. De novo clustering using k = 3 clusters and including high sensitivity C‐reactive protein in the variable set showed high stability (all Jaccard indices >0.75). The three de novo clusters ( n = 96, n = 172, n = 33, respectively) adequately captured heterogeneity within the sample and showed different distributions of polygenic risk scores and diabetes complications, that is, cluster 1 was characterized by insulin resistance with high neuropathy prevalence, cluster 2 was defined as age‐related diabetes and cluster 3 showed highest risk of genetic and obesity‐related diabetes. Conclusion: T2D subphenotyping based on its sample's own clinical characteristics leads to stable categorization and adequately reflects T2D heterogeneity.

Aims: This study investigated the role of plasma proteins in obesity to identify predictive biomarkers and explore underlying biological mechanisms. Methods: In the Cooperative Health Research in the Region of Augsburg (KORA) FF4 study, 809 proteins were measured in 2045 individuals (564 obese and 1481 non‐obese). Multivariate logistic regression adjusted for confounders (basic and full models) was used to identify obesity‐associated proteins. Priority‐Lasso was applied for feature selection, followed by machine learning models (support vector machine [SVM], random forest [RF], k‐nearest neighbour [KNN] and adaptive boosting [Adaboost]) for prediction. Correlation and enrichment analyses were performed to elucidate relationships between protein biomarkers, obesity risk factors and perturbed pathways. Mendelian randomisation (MR) assessed causal links between proteins and obesity. Results: A total of 16 proteins were identified as significantly associated with obesity through multivariable logistic regression in the basic model and subsequent Priority‐Lasso analysis. Enrichment analyses highlighted immune response, lipid metabolism and inflammation regulation were linked to obesity. Machine learning models demonstrated robust predictive performance with area under the curves (AUC) of 0.820 (SVM), 0.805 (RF), 0.791 (KNN) and 0.819 (Adaboost). All 16 proteins correlated with obesity‐related risk factors such as blood pressure and lipid levels. MR analysis identified AFM, CRP and CFH as causal and potentially modifiable proteins. Conclusions: The protein signatures identified in our study showed promising predictive potential for obesity. These findings warrant further investigation to evaluate their clinical applicability, offering insights into obesity prevention and treatment strategies.