On screening 104 consecutive individuals at increased risk for T2DM referred to the ongoing prevention program for T2DM at the outpatient clinic of the Division of Diabetology, Department of Endocrinology, Medical University – Sofia, we recruited 25 subjects with NGT (age 44.8±9.6 years; body mass index (BMI) 32.3±6.9 kg/m2) and 27 subjects with IGT (47.6±11.8 years; 31.0±6.5 kg/m2). Subjects’ eligibility was based on their medical history. The exclusion criteria were defined as follows: subjects with known type 1 or type 2 diabetes, hypothyroidism, alcohol use (>20–50 g daily for women and >30–60 g daily for men), other causes of chronic liver disease, use of anti-hyperglycemic medications, methotrexate, amiodarone, tetracycline, glucocorticoids, tamoxifen, chemotherapeutics, and nucleoside analogs, impaired kidney function, arrhythmias, cardiovascular events, recent acute illness, and pregnancy.

All subjects were informed about the aims, methods and risks of participating in the study and provided written informed consent for participation in accordance with the Declaration of Helsinki and the rules of Good Clinical Practice before entering the study.

All participants underwent a careful assessment of anthropometric and clinical parameters before undergoing assessment of glucose tolerance, autonomic nervous system, and liver elastography.

All measurements were obtained after an overnight fast and included recording of height (m) and weight (kg) for BMI calculation, and waist circumference taken at midline between the superior border of the iliac crest and the inferior margin of the ribs in the standing position after exhalation; visceral fat area (VFA, cm2) and total body fat (TBF) by the InBody 720 bioimpedance device (BioSpace, USA); blood pressure measurement (performed twice under standard conditions after 5 min sitting).

After these initial evaluation, subjects underwent a 75 g oral glucose tolerance test (OGTT) and a mixed meal tolerance test (MMTT: 75 g glucose dissolved in 250 mL of water, 50 g cheese, one boiled egg and one slice of bread; 600 kcal: 68% carbohydrates, 16% proteins, and 16% fat). The tests were performed on 2 consecutive days after a 12-hour overnight fast. Participants were asked to be on a diet regimen with 150 g of carbohydrate daily intake during the 72 hours prior to the first OGTT test. Venous blood samples were taken at −20 to –10, 0, 10, 20, 30, 45, 60, 90, and 120 min during OGTT and at −20 to –10, 0, 10, 20, 30, 45, 60, 90, 120, 150, and 180 min during MMTT. Plasma glucose (hexokinase method), insulin and C-peptide (electrochemiluminescence method) concentrations were measured at each time points on Cobas Integra (Roche, Switzerland).

On the day after the second tolerance test, autonomic nervous system function was evaluated with the АNX V.3.0 autonomic monitoring system (ANSAR Medical Technologies, Philadelphia, Pennsylvania, USA). The ANX V.3.0 software computes SNS and PSNS activity simultaneously and independently using a “frequency-domain” analysis of standard clinical tests: deep breathing, Valsalva maneuver, and standing from a seated position. The method, focused on the low-frequency range of the spectrum between 0.04 and 0.15 Hz, computes parasympathetic (respiratory frequency area) activity and indirectly sympathetic (low frequency area) activity via spectral analysis of respiratory activity with concomitant spectral analysis of HRV.11 12 All assessments were performed between 08:00 and 11:00, at least 30 min after breakfast and at least 24 hours after the last dose of any medications potentially affecting autonomic function such as antihypertensives, tricyclic antidepressants, and selective serotonin reuptake inhibitors (SSRIs), and refraining from coffee and smoking for the 12 hours prior to the test.

Transient elastography (TE) was performed within a month after the laboratory tests by Fibroscan 502 TOUCH (Echosense, France). The procedure was performed in a fasting state by a blinded experienced gastroenterologist and radiologist with participant in a standard supine position. All measurements were made in the expiratory phase of a quiet respiration with a transducer applied on the body surface without pressure. Depending on BMI, transient elastography was performed with M probe for BMI <30 kg/m2—the majority of the participants—and with XL probe for BMI >30 kg/m213 14 to determine shear wave propagation speed and calculate equivalent liver stiffness. Fibrosis was also defined in kPa.

Based on OGTT and according to the American Diabetes Association (ADA) criteria the study population was divided in subjects with NGT and IGT.15 MAFLD was defined as CAP >233 dB/m16 and overweight or obesity, or evidence of other metabolic dysregulation.17 Four subgroups were then identified: (1) NGT without MAFLD, (2) NGT with MAFLD, (3) IGT without MAFLD, and (4) IGT with MAFLD.

Areas under the curve (AUCs) during OGTT and MMTT for glucose, insulin and C-peptide were calculated by the trapezoidal rule. The Insulinogenic Index (IGI),18 Homeostatic model assessment for beta-cell function (HOMA-b),19 Insulin Secretion-Sensitivity Index-2 (ISSI-2) and B-cell Function Index ((ISSI-2/320) × 100)20 were calculated to explore beta-cell function. The Matsuda Insulin Sensitivity Index (Matsuda ISI),21 Homeostatic model assessment for insulin resistance (HOMA-IR),19 the Quantitative Insulin Sensitivity Check Index (Quicky-IS),22 and the Triglycerides-Glucose (TyG) Index23 were calculated for assessment of insulin sensitivity.

Statistical analyses were performed with the SPSS V.23.0 software (IBM Corporation, Chicago, Illinois, USA). Data normal distribution was evaluated by Kolmogorov-Smirnov test and variables with skewed distribution were log-transformed before statistical analysis. Differences across groups were tested using descriptive analysis and a single-factor dispersion analysis of variance (ANOVA). Pearson’s correlation analysis was performed to determine the relationships between CAP, autonomic and beta-cell function and insulin action parameters. Logistic regression analysis using forward stepwise method and controlling for age, BMI, and χ2 test were performed to establish the independent relationship between CAP and autonomic and beta-cell function indices. Receiver operating characteristic (ROC) curve was constructed to estimate the “cut-off” prediction values of VFA and AUC-insulinMMTT with highest sensitivity and specificity for the presence of MAFLD in the studied cohort. All variables are shown as mean±SD or median and IQR, depending on their distribution. A p value (two tailed) <0.05 was considered statistically significant.