The Usefulness of Diagnostic Panels Based on Circulating Adipocytokines/Regulatory Peptides, Renal Function Tests, Insulin Resistance Indicators and Lipid-Carbohydrate Metabolism Parameters in Diagnosis and Prognosis of Type 2 Diabetes Mellitus with Obesity

Abstract

The quantitative analysis of selected regulatory molecules, i.e., adropin, irisin, and vaspin in the plasma of obese patients with newly diagnosed, untreated type 2 diabetes mellitus, and in the same patients after six months of using metformin, in relation to adropinemia, irisinemia and vaspinemia in obese individuals, was performed. The relationship between plasma concentration of the adipocytokines/regulatory peptides and parameters of renal function (albumin/creatinine ratio-ACR, estimated glomerular filtration rate-eGFR), values of insulin resistance indicators (Homeostatic Model Assessment of Insulin Resistance (HOMA-IR2), Homeostatic Model Assessment of Insulin Sensitivity (HOMA-S), Homeostatic Model Assessment of β-cell function (HOMA-B), quantitative insulin sensitivity check index (QUICKI), insulin), and parameters of carbohydrate-lipid metabolism (fasting plasma glucose-FPG, glycated hemoglobin-HbA_1C, estimated glucose disposal rate-eGDR, fasting lipid profile, TG/HDL ratio) in obese type 2 diabetic patients was also investigated. Circulating irisin and vaspin were found significantly different in subjects with metabolically healthy obesity and in type 2 diabetic patients. Significant increases in blood levels of both analyzed adipokines/regulatory peptides were observed in diabetic patients after six months of metformin treatment, as compared to pre-treatment levels. The change in plasma vaspin level in response to metformin therapy was parallel with the improving of insulin resistance/sensitivity parameters. An attempt was made to identify a set of biochemical tests that would vary greatly in obese non-diabetic subjects and obese patients with type 2 diabetes, as well as a set of parameters that are changing in patients with type 2 diabetes under the influence of six months metformin therapy, and thus differentiating patients' metabolic state before and after treatment. For these data analyses, both statistical measures of strength of the relationships of individual parameters, as well as multidimensional methods, including discriminant analysis and multifactorial analysis derived from machine learning methods, were used. Adropin, irisin, and vaspin were found as promising regulatory molecules, which may turn out to be useful indicators in the early detection of T2DM and differentiating the obesity phenotype with normal metabolic profile from T2DM obese patients. Multifactorial discriminant analysis revealed that irisin and vaspin plasma levels contribute clinically relevant information concerning the effectiveness of metformin treatment in T2D patients. Among the sets of variables differentiating with the highest accuracy the metabolic state of patients before and after six-month metformin treatment, were: (1) vaspin, HbA1c, HDL, LDL, TG, insulin, and HOMA-B (ACC = 88 [%]); (2) vaspin, irisin, QUICKI, and eGDR (ACC = 86 [%]); as well as, (3) vaspin, irisin, LDL, HOMA-S, ACR, and eGFR (ACC = 86 [%]).

Keywords: adipocytokines; adropin; irisin; metformin therapy; obesity; type 2 diabetes mellitus; vaspin.

Figure 1

Three-parameter discriminant analysis, which allow to differentiate subjects with metabolically healthy obesity and T2DM obese patients: (a) Adropin, irisin and vaspin data set obtained from all examined individuals (b) Graph of the discriminatory curve which based on adropin, irisin and vaspin allows for separation between healthy obesity and T2DM patients with ACC = 94 [%] (c) Irisin, QIUCKI, and HOMA-IR data set obtained from all examined individuals (d) Graph of the discriminatory curve which based on irisin, QIUCKI and HOMA-IR allows for separation between healthy obesity and T2DM patients with ACC = 90 [%].

Figure 1

Three-parameter discriminant analysis, which allow to differentiate subjects with metabolically healthy obesity and T2DM obese patients: (a) Adropin, irisin and vaspin data set obtained from all examined individuals (b) Graph of the discriminatory curve which based on adropin, irisin and vaspin allows for separation between healthy obesity and T2DM patients with ACC = 94 [%] (c) Irisin, QIUCKI, and HOMA-IR data set obtained from all examined individuals (d) Graph of the discriminatory curve which based on irisin, QIUCKI and HOMA-IR allows for separation between healthy obesity and T2DM patients with ACC = 90 [%].

Figure 2

Three-parameter discriminant analysis, which allow to differentiate metabolic state of T2DM patient before treatment and after 6-month metformin therapy: (a) adropin, irisin, and vaspin data set obtained diabetic patients (b) graph of the discriminatory curve which based on adropin, irisin and vaspin allows for separation between metabolic state od diabetic patients before and after metformin treatment; ACC = 69 [%] (c) irisin, HbA1c, and eGDR data set obtained from diabetic patients; and, (d) graph of the discriminatory curve, which, based on irisin, HbA1c, and eGDR, allows for separation between metabolic state od diabetic patients before and after metformin treatment; ACC = 80 [%].

Figure 2

Three-parameter discriminant analysis, which allow to differentiate metabolic state of T2DM patient before treatment and after 6-month metformin therapy: (a) adropin, irisin, and vaspin data set obtained diabetic patients (b) graph of the discriminatory curve which based on adropin, irisin and vaspin allows for separation between metabolic state od diabetic patients before and after metformin treatment; ACC = 69 [%] (c) irisin, HbA1c, and eGDR data set obtained from diabetic patients; and, (d) graph of the discriminatory curve, which, based on irisin, HbA1c, and eGDR, allows for separation between metabolic state od diabetic patients before and after metformin treatment; ACC = 80 [%].

Figure 3

A cumulative graph of ACC value changes for all variable combinations of selected variables (HbA1c, HDL, LDL, TG, ACR, eGFR, glucose, insulin, QUICKI, HOMA-IR, HOMA-S, HOMA-B, and eGDR) with always present vaspin. The number of all analyzed combinations of selected parameters was specified as 2¹⁴ − 1 = 16383 classifications, among which the diagnostic panels with the highest ACC value were selected.