Impaired Glucose Regulation In Patients With Metabolomics Results With Tcm Symptoms Of The Preliminary Study

Objective:To explore the correlations between the clinical manifestations and metabonomics profiles in primary type 2 diabetes mellitus (T2DM) patients.Methods:In the 1539 persons who accepted the health examination in Hang-xing Community,143 cases with Impaired fasting glucose (fasting venous blood glucose ranged from 6.1 to 6.9 mmol/L) were detected and recruited in this study. After then an OGTT was performed for every patient, blood serum sample was collected at 2 time spots:fasting and 2 hours after 75g glucose being administrated orally. T2DM relative routine laboratory parameters were tested, and body measured data and diet survey for proxima luce were collected at the same time with OGTT. According to the OGTT, the patients were divided into 3 groups with diagnosis of health, pre-diabetic state, and diabetes mellitus, respectively.35 cases in each group were included and matched each other by age and gender. For the metabolomics investigations blood serum samples were collected after an overnight fasting and 2 hours after glucose load under standardized conditions and immediately stored in aliquots at-80 C. The UPLC system was coupled to a q TOF-MS (Micromass, Manchester, UK) equipped with an electrospray source operat-ing in either positive or negative ion mode, produced by Waters Corporation in USA. The pre-processed UPLC-qTOF-MS data were exported into Soft Indepen-dent Modelling of Class Analogy (SIMCA)-P (version 11.0, Umetrics AB, Umea, Sweden) for analysis and visualization by multivariate statistical methods. After Pareto scaling and OSC-filtering, principal components analysis (PCA), partial least squares-discriminant analysis (PLS-DA), and orthogonal signal correction PLS-DA (OPLS-DA) were applied. The predictive ability of the model was assess-ed by internal validation using 7-fold cross-validation and response permutation testing. The correlation between clinical manifestations and matabonomics profil-es was analyzed with logistic regression approach, performed on SAS9.1.3 soft-ware (No.195557). Statistical significance was set at P\0.05. For the detection of metabolite ion masses with major influence on the group membership the S-plot was used according to. The metabolite heat map was generated using MultiExperi-ment View V4.1 (www.tm4.org). Results:The baseline data showed no significant difference in gender among the 3 groups, but significantly different in age, BMI, HbA1c, fasting plasma glucose, OGTT 2 hours plasma glucose.The metabonomics analysis revealed that:PCA can't distinguish the 3 groups clearly by the fasting and OGTT 2 hours blood serum samples. The healthy and diabetic individuals could be distinguished by PLS-DA and OPLS-DA with fasting blood samples but negatively discerned with OGTT 2 hours blood serum samples; the pre-diabetic state cases can't be distinguished from neither health cases nor DM cases. PCA approach indicated there existed similar shifting directions from fasting to OGTT 2 hours state in all 3 groups. Two subgroups were identified by PCA in T2DM groups, further analysis proved that between the 2 subgroups, the principle distinctive metabolites were LPC, tryptophan and phenylalanine. No positive discovery was obtained in healthy and T2DM groups analysis by PCA, PLS-DA and OPLS-DA.Logistic regression analysis manifested that there were significant difference in the occurrence of TCM symptoms combined with "fearing of cold in low back and legs" and "frequent nocturia". These two symptoms were regarded as the symbol of Kidney Yang deficiency pattern according to TCM theory. In the sub-group with kidney Yang deficiency pattern diagnosis, tryptophan and pheny-lalanine were down regulated, and LPC was up regulated.Conclusions:1) OGTT can tell the difference among the 3 groups, yet metabonomics analysis approach can't distinguish the pre-diabetic patients from healthy and T2DM patients, neither by fasting blood serum samples nor OGTT 2 hours samples. This result posed us a hypothesis that although diagnosed with different diseases period, the patients might have similar metabolic state due to similar disease states and environment.2) The metabolic states were similar in the cases of healthy and pre-diabetic groups, non-effective approaches were explored to find differences between these 2 groups. However, primary T2DM led to different metabolic state, which could be detected by metabonomics analysis, and PCA could distinguish T2DM individuals into 2 subgroups by the fasting blood samples. The two subgroups could be identified as kidney Yang deficiency group and non-kidney Yang deficiency group according to TCM theory.3) Between the 2 subgroups of T2DM cases, there were 3 metabolites showed different distribution: LPC, phenylalanine and tryptophan. In kidney Yang deficiency group, LPC was up regulated and phenylalanine and tryptophan were down regulated. Phenylalanine acts as the substrate in synthesis of catecholamines and thyroid hormones, trypto-phan is substrate in synthesis of melatonin, which supported the results that there were correlation between the down-regulation of phenylalanine and tryptophan with kidney Yang deficiency pattern in biomedical way. In all, there should exist the metabolic change in the development of kidney Yang deficiency pattern in T2DM patients.