Table 2.

Logistic regression models considering HPR as endpoint

	p-value	OR (95% CI)
Model 1 (pseudoR2: 0.499, p < 0.001)
HbA1c	0.012	7.25 (1.55–33.86)
Model 2 (pseudoR2: 0.757, p < 0.0001)
HbA1c	0.019	13.21 (1.52–114.19)
MAGE	0.034	1.094 (1.007–1.188)
Model 3 (pseudoR2: 0.618, p < 0.0001)
HbA1c	0.018	9.13 (1.46–56.79)
CV	0.071	1.22 (0.98–1.53)

HPR high platelet reactivity, Hb1Ac glycated hemoglobin, MAGE mean amplitude glucose excursions, CV coefficient of variation

In the multivariate analysis (hierarchical enter method) HPR was entered in the model as dependent variable and as independent variables were included only the variables with p < 0.10 at the bivariate regression analysis: HbA1c, MAGE and CV. Excluded variables: fasting blood glucose, BMI, age, gender, weight, left ventricle ejection fraction, clopidogrel bolus (600 mg), chronic clopidogrel therapy, haemoglobin, haematocrit, platelet count number, white blood cells, total cholesterol, HDL, LDL, triglycerides and all the other glycaemic variability indexes (glycaemic average, SD, CONGA1, CONGA2, CONGA4, MAGE UP, MAGE DOWN); *p < 0.05