Table 4.
Multivariable model results for atorvastatin (ATV) analyte ratios or sum total
| Model characteristic | 2‐OH ATV/ATV | 2‐OH ATV L/ATV L | ATV L/ATV | TOTAL | ||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Variable | B (SE) | I | P‐value | B (SE) | I | P‐value | B (SE) | I | P‐value | B (SE) | I | P‐value |
| ATV dose | ‐ | ‐ | ‐ | ‐ | ‐ | ‐ | ‐ | ‐ | ‐ | 0.215 (0.046) | 1 | 3.70x10–6 |
| Sex (F vs M) | 0.124 (0.022) | 1 | <2.0x10–16 | 0.090 (0.022) | 1 | .000038 | ‐ | ‐ | ‐ | ‐ | ‐ | ‐ |
| Age | 0.002 (0.0009) | 0.41 | .013 | 0.003 (0.001) | 1 | .00050 | ‐ | ‐ | ‐ | 0.005 (0.001) | 1 | 3.91x10–6 |
| Body mass index | −0.004 (0.002) | 0.39 | .013 | ‐ | ‐ | ‐ | ‐ | ‐ | ‐ | −0.005 (0.002) | 0.65 | .0061 |
| Smoking | −0.067 (0.022) | 0.89 | .0020 | −0.058 (0.021) | 0.86 | .0049 | 0.102 (0.021) | 1 | 6.90x10–7 | ‐ | ‐ | ‐ |
| Diabetes mellitus | ‐ | ‐ | ‐ | ‐ | ‐ | ‐ | −0.055 (0.024) | 0.35 | .020 | ‐ | ‐ | ‐ |
| Aspirin | ‐ | ‐ | ‐ | 0.096 (0.040) | 0.33 | .016 | ‐ | ‐ | ‐ | ‐ | ‐ | ‐ |
| P2Y12 inhibitor | 0.076 (0.026) | 0.87 | .0035 | 0.077 (0.026) | 0.76 | .0031 | −0.055 (0.026) | 0.11 | .036 | ‐ | ‐ | |
| ACEI/ARB | 0.068 (0.024) | 0.07 | .0052 | 0.058 (0.024) | 0.06 | .015 | ‐ | ‐ | ‐ | ‐ | ‐ | ‐ |
| Loop diuretic | ‐ | ‐ | ‐ | ‐ | ‐ | ‐ | ‐ | ‐ | ‐ | 0.093 (0.032) | 1 | .0039 |
| Proton pump inhibitor | ‐ | ‐ | ‐ | ‐ | ‐ | ‐ | 0.046 (0.019) | 0.41 | .014 | 0.087 (0.024) | 1 | .00023 |
| CYP3A inhibitor | −0.144 (0.048) | 1 | .0025 | −0.120 (0.047) | 0.46 | .011 | ‐ | ‐ | ‐ | ‐ | ‐ | ‐ |
| Amiodarone | −0.186 (0.083) | 0.28 | .025 | −0.208 (0.082) | 0.43 | .011 | ‐ | ‐ | ‐ | ‐ | ‐ | ‐ |
| Sample storage duration | −0.0001 (0.00002) | 1 | 1.00x10–7 | −0.00007 (0.00002) | 1 | .00087 | 0.0002 (0.00002) | 1 | <2.0x10–16 | ‐ | ‐ | ‐ |
| Time since last ATV | 0.0002 (0.00008) | 1 | .0060 | 0.0001 (0.00008) | 1 | .21 | 0.0002 (0.00008) | 1 | .0047 | −0.0008 (0.0001) | 1 | <2.0x10–16 |
| Num of models averaged | 10 | 12 | 5 | 2 | ||||||||
| R2 | 0.18 | 0.14 | 0.24 | 0.21 | ||||||||
I = importance; SE = standard error. Total = ATV + 2‐OH ATV + ATV L + 2‐OH ATV L.
All subsets linear regression was performed using the (maximum) model containing all considered covariates for each endpoint, limited by the requirement that all subsetted models include time since last ATV dose. For each endpoint, regression models were ranked according to BIC, and those with BIC < 2 from the lowest ranked model were averaged to produce the averaged coefficients and standard errors of the final model, displayed in these tables. The relative importance of each covariate included in the final model for a given endpoint was assessed by summing the BIC‐derived model weight of all models that underwent model averaging and contain that covariate (1 indicates high importance). P‐values are provided for reference only; they were not used during covariate selection.
ACEI/ARB, angiotensin‐converting enzyme inhibitor/angiotensin II receptor blocker