Table 1 Characteristics of the data included in the meta-analysis of type 2 diabetes mellitus
Author-year-Journal | Study design | Population | Cases N | Control N | Biological sample | Detection method | Analytical method | Covariates | Metabolites |
|---|---|---|---|---|---|---|---|---|---|
Zhu et al., 2011, Talanta [10] | Case-control study | Chinese | 30 | 30 | plasma | NPLC-TOF/MS | PCA, PLS-DA and ANOVA | Sex | (↑) LPC (16:0), LPC (18:2), LPC (18:0) |
Liu et al., 2016, Sci Rep [11] | Case-control study | Chinese | 15 | 15 | plasma | 1H NMR spectroscopy | PCA, PLS-DA, HCA and ROC curve analysis | Gender, age, BMI, SBP, DBP, total cholesterol, BUN and serum creatinine | (↑) Glucose, glycine (↓) Valine, leucine, alanine, isoleucine, arginine, proline, glutamine, threonine, tyrosine, creatine |
Wang-sattler et al., 2017, Mol Syst Biol [22] | KORA S4 cross-sectional study | Augsburg and the surrounding towns and villages | KORA S4: 91 EPIC: 133 Prospective: 91 | KORA S4: 866 EPIC: 1253 Prospective: 91 | serum | LC-FIA-MS | Multivariate logistic regression and linear regression | Age, sex, BMI, PA, alcohol intake, smoking, SBP and HDL | (↓) Glycine, LPC (18:2) |
Ha et al., 2012, Clin Endocrinol [23] | Case-control study | South Korean (men) | 26 | 27 | serum | UPLC/Q-TOF-MS | PLS-DA and ROC curve analysis | Age and BMI | (↑) Myristic acid (14:0) (−) Palmitic acid (16:0), stearic acid (18:0), arachidonic acid (20:4 ω6) |
Menni et al., 2013, Diabetes [24] | Case-control study | British | 115 | 1897 | plasma | NA | Random intercept logistic regressions analysis | Age, BMI, batch effect and family relatedness | (↑) Valine, leucine, isoleucine, proline, mannose, glucose,2-hydroxybutyrate, lactate (↓) Citrulline, myristate (14:0),1,5-anhydroglucitol |
Kujala et al., 2016, Front Med [25] | Case-control study | Older Finnish men | 126 | 214 | serum | NMR platform | Mann-Whitney U-test or Kruskal-Wallis test | Age | (↑) Valine, isoleucine, leucine |
Peddinti et al., 2017, Diabetologia [26] | Botnia Prospective Study | Finland | 146 | 397 | serum | UHPLC-MS/MS and GC-MS | Multivariate logistic regression analysis | Age, sex, BMI, fasting insulin level and family history | (↑) Valine, isoleucine, glutamate, mannose, glucose (↓) Histidine, glutamine |
Okekunle et al., 2017, Diabetes Res Clin Pract [27] | Case-control study | Chinese (Harbin) | 50 | 50 | serum | UPLC-TQ-MS | ANOVA and covariance analysis | Age, sex, BMI and insulin resistance | (↑) Glutamic acid, ornithine (↓) Serine, asparagine, glycine, threonine (−) Valine, leucine, isoleucine, tyrosine, phenylalanine, lysine, alanine, histidine, proline, citrulline, arginine, ethionine, glutamine |
Abu Bakar et al., 2017, Mol Biosyst [28] | Nested-case control study | Malaysian | 37 | 32 | plasma | LC-MS and HPLC | The Kruskal-Wallis test, PLS-DA and ROC curve analysis | Age, gender, BMI and SBP | (↑) Valine, leucine, tyrosine, soleucine, phenylalanine, glutamine, glutamate, lysine, proline, palmitic acid, arachidonic acid, myristic acid (↓) Alanine, stearic acid |
Andersson-Hall et al., 2018, J Diabetes Res [29] | 6-years follow-up study | Women in Gothenburg | 44 | 139 | serum | NMR Spectroscopy | ANOVA and ANCOVA | BMI | (↑) Valine, leucine, tyrosine, isoleucine, glucose, mannose, acetoacetate, glycerol, 3-hydroxy-isobutyrate (↓) Glycine (−) Phenylalanine |
Lee et al., 2016, Metabolomics [30] | KARE cohort (prospective) study | South Korean | 517 | 924 | serum | LC/MS/MS and LC-FIA-MS | Multivariable logistic regression and linear regression | Age, sex, BMI and HDL | (↓) Glycine, LPC a (18:2) |
Floegel et al., 2013, Diabetes [31] | EPIC-Potsdam case-cohort study | The area of Potsdam in eastern Germany | 800 | 2282 | serum | FIA-MS/MS | Cox proportional hazards regression and PCA | Age, sex, alcohol intake, smoking, education, coffee intake, BMI and waist circumference | (↑) Phenylalanine (↓) Glycine, LPC a (18:2) |
Gogna et al., 2015, Mol Biosyst [32] | Case-control study | South Indian Asians | 165 | 128 | serum | COSY, HSQC, HMQC, CPMG NMR spectra | PCA and PLS-DA | Age, sex and BMI | (↑) Valine, leucine, isoleucine, lysine, glutamine, phenylalanine, proline, threonine, histidine, glucose, lactate, glycerol |
Lu et al., 2016, Diabetologia [33] | Prospective cohort study | Chinese | 197 | 197 | serum | LC-MS and GC-MS | OPLS-DA, ROC and conditional logistic regression analysis | Age, sex, BMI, smoking, status and history of hypertension | (↑) Valine, leucine, glycine, isoleucine, threonine, palmitic acid (16:0), stearic acid (18:0) (↓) Ornithine, proline, serine, glycerol |
Palmer et al., 2015, J Clin Endocrinol Metab [34] | 5-year follow-up study | European American, Hispanic and African American | 76 | 70 | plasma | MS/MS | Logistic regression analysis | Age, sex, ethnicity, recruitment site and BMI | (↑) Valine, leucine, isoleucine, tyrosine, phenylalanine, glutamine and glutamate (↓) Glycine, alanine (−) Serine, proline, histidine, methionine, ornithine, citrulline, arginine |
Ng et al., 2012, Diabetologia [35] | Singapore Diabetes Cohort Study (SDCS) | Singaporean | 44 | 46 | urine | GC/MS and LC/MS | OPLS-DA, PCA and LASSO | Multiple hypotheses testing by controlling for FDR | (↑) L-serine, creatinine |
Liu et al., 2017, Metabolomics [36] | 14-years follow-up study | Southwest of the Netherlands | 137 | 1434 | plasma | LC-MS, NMR-COMP and NMR-LIPO | LASSO and ROC curve analysis | Age, sex, family history and BMI | (↑) Isoleucine, methionine, tyrosine, 1,5-anhydroglucitol, 2-hydroxybutyrate, lactate, glycerol |
Merino et al., 2018, Diabetologia [37] | Prospective study | Framingham | 95 | 1055 | plasma | LC-MS/MS | LASSO and ROC curve analysis | Age, sex, BMI, fasting glucose and triacylglycerols | (↑) Phenylalanine (↓) Glycine |
Li et al., 2017, Mol Biosyst [38] | Case-control study | Chinese | 25 | 20 | urine | GC-TOF-MS | PCA, OPLS-DA and ROC curve analysis | Age | (↓) Glycine (−) Serine, stearic acid, palmitic acid, 1,5-anhydroglucitol |
Chou et al., 2018, J Chromatogr B [39] | Case-control study | Chinese (Harbin) | 47 | 48 | serum | GC-MS | ANOVA, ROC, PCA and PLS-DA analysis | Age, sex, smoking and alcohol consumption | (↑) Lactate (−) α-hydroxybutyrate |
Wolak-Dinsmore et al., 2018, Clin Biochem [40] | Cross-sectional study | Groningen cohort-white | 67 | 56 | serum | LC-MS/MS and NMR spectroscopy | Multivariable linear regression analyses | Age and sex | (↑) Valine, leucine, isoleucine |
Lu et al., 2019, Metabolites [41] | Prospective study | Chinese | 144 | 144 | serum | LC-MS | t-test, chi-square test and logistic regression | BMI, history of hypertension, smoking status, HDL-cholesterol and triglycerides | (↑) Valine, leucine, Isoleucine, phenylalanine, lysine, methionine, alanine (−) Threonine, histidine, glutamine, glycine, tyrosine, serine, proline |
Lu et al., 2018, J Clin Endocrinol Metab [42] | Follow-up study | Singapore Chinese | 144 | 144 | serum | HPLC-QQQ-MS/GC | ROC and conditional logistic regression analysis | BMI, history of hypertension, smoking, physical activity, fasting status, triglycerides and HDL cholesterol | (↑) Myristic acid (14:0), palmitic acid (16:0), stearic acid (18:0), arachidonic acid (20:4n-6) |
Friedrich et al., 2015, Metabolomics [43] | Longitudinal cohort study | North-east area of Germany | Men (87), Women (50) | Men (1266), Women (1306) | urine | NMR spectroscopy | Logistic regression and ROC curve analysis | Age and waist circumference | Women: (↑) alanine, glycine, glucose, lactate (↓) Creatinine Men: (↑) valine, glucose, lactate (−) Glycine |