Constructing a metabolic integral score model for the quantification of metabolic dysfunction and tendency

The binary nature of metabolic syndrome (MetS) cannot quantitatively describe the severity of metabolic abnormalities. We aim to establish a metabolic integral score (MIS) model to quantify the severity and polarity of metabolic disorders and their relationship with insulin sensitivity and secretion.

We performed factor analysis on 9950 participants from a cross-sectional study conducted in China. The MIS model was established using 10 variables including body mass index (BMI), waist circumference, hip circumference, glycosylated hemoglobin (HbA1c), fasting and 2-h plasma glucose (FPG, 2h-PG), systolic blood pressure (SBP), diastolic blood pressure (DBP), high-density lipoprotein (HDL) and triglyceride (TG) levels. Four common factors were identified as "glucose factor," "obesity factor," "blood pressure factor," and "lipid factor," respectively, in MIS model (KMO = 0.755, P < 0.001). MIS = 0.433 × Factor 1 + 0.267 × Factor 2 + 0.172 × Factor 3 + 0.128 × Factor 4. Insulin sensitivity and β-cell function decreased with the increase of MIS (P < 0.001). We classified four metabolic tendencies according to factor quartiles. Individuals in Tendency 1 (severe hyperglycemia) had the worst β-cell function. Tendency 3 (severe hypertension) had the best insulin sensitivity. Tendency 4 (severe dyslipidemia) had preferable β-cell function (P < 0.05).

Our MIS model provides a quantitative scoring system to assess various patterns of metabolic abnormality that indicate different underlying pathophysiology.