Table 3.
Linear regression and quartile analysis of lipid levels as determinants for TG parametersdetermined using PPP reagent.
| Cholesterol mg/dL) | Triglycerides (mg/dL) | LDL cholesterol (mg/dL) | HDL cholersterol (mg/dL) | ||
|---|---|---|---|---|---|
| Lag time | β-coef. UVR | 0.002*** | 0.001*** | 0.003*** | −0.005*** |
| β-coef. MR | COL | 0.001*** | 0.003*** | −0.005*** | |
| Q1 | 203 (52) | 97 (67) | 121 (43) | 58 (20) | |
| Q2 | 211 (54)*** | 107 (73)*** | 128 (44)*** | 57 (20)*** | |
| Q3 | 216 (56)*** | 115 (79)*** | 134 (45)*** | 55 (19)*** | |
| Q4 | 221 (59)*** | 127 (90)*** | 137 (49)*** | 52 (18)*** | |
| ETP | β-coef. UVR | 1.97*** | 0.50*** | 2.17*** | |
| β-coef. MR | 1.13*** | 0.51*** | 0.80** | ||
| Q1 | 200 (52) | 98 (68) | 120 (44) | 56 (19) | |
| Q2 | 209 (54)*** | 107 (73)*** | 126 (43)*** | 56 (20) | |
| Q3 | 215 (53)*** | 112 (79)*** | 132 (45)*** | 56 (19) | |
| Q4 | 220 (58)*** | 117 (82)*** | 136 (48)*** | 56 (19) | |
| Peak | β-coef. UVR | 0.29*** | 0.10*** | 0.27*** | 0.15*** |
| β-coef. MR | COL | 0.20*** | 0.22*** | 0.38*** | |
| Q1 | 204 (56) | 99 (70) | 124 (47) | 55 (20) | |
| Q2 | 208 (53)*** | 105 (74)** | 126 (44)*** | 56 (19) | |
| Q3 | 212 (54)*** | 111 (74)*** | 129 (44)*** | 56 (20) | |
| Q4 | 219 (56)*** | 119 (84)*** | 134 (47)*** | 56 (19) | |
| Time-to-peak | β-coef. UVR | 0.003*** | 0.001*** | 0.004*** | −0.006*** |
| β-coef. MR | COL | 0.001*** | 0.005*** | −0.006*** | |
| Q1 | 202 (50) | 101 (70) | 119 (42) | 57 (20) | |
| Q2 | 211 (54)*** | 107 (74)*** | 128 (45)*** | 57 (19)* | |
| Q3 | 214 (55)*** | 111 (77)*** | 132 (46)*** | 55 (20)*** | |
| Q4 | 220 (59)*** | 120 (88)*** | 137 (49)*** | 54 (19)*** | |
| Velocity index | β-coef. UVR | 0.05*** | 0.04*** | 0.06* | |
| β-coef. MR | 0.01 | 0.06*** | 0.15** | ||
| Q1 | 211 (57) | 128 (47) | 105 (74) | 55 (19) | |
| Q2 | 210 (55) | 128 (45) | 104 (73) | 56 (20)** | |
| Q3 | 209 (54) | 127 (45) | 108 (75)** | 56 (20)* | |
| Q4 | 214 (54)*** | 130 (45)*** | 117 (82) | 56 (19) |
Linear regression results are presented as β-coefficients and their associated p-values. Lipid levels were quantified as the median and IQR for the quartiles of each TG parameter. (Bonferroni corrected) p-values were indicated as *p <0.05, **p < 0.01, ***p < 0.001. Sex was coded as ‘'0” for men and ‘'1” for women, hence a negative β-coefficient indicates a decrease in outcome (TG) variable in women compared to men. Smoking was coded as ‘'0” for non-smokers and ‘'1” for current smokers, hence a negative β-coefficient indicates a decrease in outcome (TG) variable in smokers compared to non-smokers. Significant variables shown in the table were included in the final model. β-coef., β-coefficient; ETP, Endogenous thrombin potential; UVR, univariate regression analysis; MR, multiple regression analysis; COL, variable excluded from the model because of collinearity (VIF ≥ 10), Q, quartile.