Biomarkers of Morbid Obesity and Prediabetes byMetabolomic

16 Metabolomic studies aimed to dissect the connection between the development of type 2 diabetes and 17 obesity are still scarce. In the present study, fasting serum from sixty-four adult individuals classified 18 into four sexmatched groups by their BMI [non-obese versus morbid obese] and the increased risk of 19 developing diabetes [prediabetic insulin resistant state versus non-prediabetic non-insulin resistant] 20 was analyzed by LCand FIAESIMS/MS–driven metabolomic approaches. Altered levels of 21 [lyso]glycerophospholipids was the most specific metabolic trait associated to morbid obesity, 22 particularly lysophosphatidylcholines acylated with margaric, oleic and linoleic acids [lysoPC C17:0: 23 R = −0.56, p = 0.0003; lysoPC C18:1: R = −0.61, p = 0.0001; lysoPC C18:2 R = −0.64, p b 0.0001]. 24 Several amino acidswere biomarkers of risk of diabetes onset associated to obesity. For instance, 25 glutamate significantly associatedwith fasting insulin [R=0.5, p=0.0019] and HOMA-IR [R=0.46, 26 p=0.0072],while glycine showed negative associations [fasting insulin: R = −0.51, p = 0.0017; 27 HOMA-IR: R = −0.49, p = 0.0033], and the branched chain amino acid valine associated to 28 prediabetes and insulin resistance in a BMI-independentmanner [fasting insulin: R=0.37, 29 p=0.0479;HOMA-IR: R=0.37, p=0.0468].Minority sphingolipids including specific 30 [dihydro]ceramides and sphingomyelins also associated with the prediabetic insulin resistant state, 31 hence deserving attention as potential targets for early diagnosis or therapeutic intervention. 32