Table 6. Multivariate regression parameters for testosterone and other co-explanatory variables for the model of ‘cumulative blood platelet reactivity’ as a dependent variable.
| variable | beta coefficient | std error of beta coefficient | -95%CI | +95%CI | determination coefficient [R2] | partial correlation | semipartial correlation | significance |
| testosterone | -0.644 | 0.270 | -1.179 | -0.110 | 0.938 | -0.194 | -0.161 | 0.018 |
| m -0.172 | m 0.087 | m -0.314 | m -0.030 | m 0.243 | m -0.165 | m -0.147 | m 0.043 | |
| f -0.223 | f 0.077 | f -0.406 | f -0.041 | f 0.165 | f -0.203 | f -0.172 | f 0.004 | |
| uric acid | -0.259 | 0.080 | -0.417 | -0.102 | 0.282 | -0.261 | -0.220 | 0.001 |
| m -0.345 | m 0.078 | m -0.484 | m -0.206 | m 0.126 | m -0.352 | m -0.329 | m <<0.0001 | |
| f -0.212 | f 0.084 | f -0.381 | f -0.043 | f 0.306 | f -0.208 | f -0.177 | f 0.011 | |
| platelet count | 0.257 | 0.076 | 0.108 | 0.407 | 0.203 | 0.272 | 0.230 | 0.001 |
| m 0.271 | m 0.090 | m 0.105 | m 0.437 | m 0.335 | m 0.245 | m 0.221 | m 0.003 | |
| f 0.309 | f 0.074 | f 0.164 | f 0.454 | f 0.137 | f 0.313 | f 0.273 | f <<0.0001 | |
| glucose | 0.232 | 0.077 | 0.081 | 0.384 | 0.223 | 0.244 | 0.205 | 0.003 |
| m 0.285 | m 0.082 | m 0.123 | m 0.447 | m 0.199 | m 0.288 | m 0.263 | m 0.001 | |
| f 0.188 | f 0.082 | f 0.034 | f 0.342 | f 0.256 | f 0.188 | f 0.159 | f 0.023 | |
| homocysteine | 0.200 | 0.074 | 0.053 | 0.347 | 0.175 | 0.218 | 0.182 | 0.008 |
| m 0.240 | m 0.082 | m 0.080 | m 0.400 | m 0.214 | m 0.237 | m 0.214 | m 0.004 | |
| f 0.188 | f 0.074 | f 0.015 | f 0.362 | f 0.147 | f 0.208 | f 0.177 | f 0.013 | |
| age | -0.140 | 0.072 | -0.282 | 0.002 | 0.117 | -0.160 | -0.131 | 0.053 |
| m -0.088 | m 0.081 | m -0.223 | m 0.046 | m 0.191 | m -0.097 | m -0.085 | m 0.272 | |
| f -0.216 | f 0.073 | f -0.371 | f -0.061 | f 0.138 | f -0.245 | f -0.209 | f 0.003 | |
| cholesterol | -0.091 | 0.076 | -0.242 | 0.060 | 0.219 | -0.098 | -0.080 | 0.237 |
| m -0.161 | m 0.080 | m -0.297 | m -0.024 | m 0.159 | m -0.168 | m -0.149 | m 0.046 | |
| f -0.053 | f 0.076 | f -0.207 | f -0.102 | f 0.181 | f -0.082 | f -0.068 | f 0.488 | |
| leukocyte count | 0.038 | 0.078 | -0.116 | 0.191 | 0.243 | 0.040 | 0.033 | 0.627 |
| m -0.082 | m 0.096 | m -0.254 | m 0.090 | m 0.413 | m -0.068 | m -0.060 | m 0.389 | |
| f 0.111 | f 0.079 | f -0.017 | f 0.239 | f 0.220 | f 0.124 | f 0.103 | f 0.162 | |
| haemoglobin | 0.033 | 0.088 | -0.141 | 0.208 | 0.417 | 0.031 | 0.026 | 0.706 |
| m 0.094 | m 0.091 | m -0.057 | m 0.245 | m 0.336 | m 0.090 | m 0.079 | m 0.295 | |
| f 0.002 | f 0.073 | f -0.144 | f 0.148 | f 0.120 | f -0.014 | f -0.011 | f 0.983 | |
| sex | 0.320 | 0.274 | -0.222 | 0.862 | 0.939 | 0.096 | 0.079 | 0.245 |
The coefficients presented as the bootstrap-boosted standardized beta coefficients and their standard errors; n = 155. The dependent variable (‘cumulative blood platelet reactivity’) contained the van der Waerden normal scores of Amax cumulated through the agonists used (AA, collagen, ADP). Per analogiam, the bootstrap-boosted multiple regression models, including the same variables, were iterated separately for men (m) and for women (f) by means of the bootstrap resampling with replacement adjusted for sample size of the complete group (n =155) (10000 iterations). The multiple correlation coefficient (R) and the corrected overall determination coefficient (R2corr) for the model were: R = 0.582 and R2corr = 0.294, P<< 0.0001, for the overall group (n=155); R = 0.484 and (R2corr) = 0.187, P<< 0.0001, for men (nresampled =155); R = 0.558 and (R2corr) = 0.269, P<< 0.0001, for women (nresampled =155).