Table 2. Boosted regression statistics.
| Biomarker | Original data set | Imputation | sets 1 to 20 | ||
|---|---|---|---|---|---|
| Mean | Std Dev | Min | Max | ||
| T.gondii antibodies (IU/ml) | 0.220 | 0.562 | 0.444 | 0.326 | 2.388 |
| Blood lead (ug/dL) | 1.482 | 1.658 | 0.066 | 1.537 | 1.753 |
| Mercury, total (ug/L) | 1.958 | 1.847 | 0.110 | 1.628 | 2.049 |
| Mercury, inorganic (ug/L) | 0.290 | 1.668 | 0.116 | 1.358 | 1.788 |
| White blood cell count (1000 cells/uL) | 1.243 | 1.126 | 0.065 | 1.000 | 1.277 |
| Lymphocyte percent (%) | 1.331 | 0.978 | 0.078 | 0.780 | 1.176 |
| Monocyte percent (%) | 1.996 | 1.595 | 0.172 | 1.371 | 1.904 |
| Segmented neutrophils percent (%) | 1.240 | 1.004 | 0.082 | 0.856 | 1.121 |
| Eosinophils percent (%) | 1.770 | 0.971 | 0.126 | 0.819 | 1.406 |
| Basophils percent (%) | 0.565 | 0.585 | 0.051 | 0.503 | 0.690 |
| Lymphocyte number (1000 cells/uL) | 0.754 | 0.912 | 0.171 | 0.691 | 1.559 |
| Monocyte number (1000 cells/uL) | 0.835 | 0.492 | 0.043 | 0.392 | 0.552 |
| Segmented neutrophils num (1000 cell/uL) | 1.138 | 1.103 | 0.075 | 0.950 | 1.282 |
| Eosinophils number (1000 cells/uL) | 0.229 | 0.129 | 0.026 | 0.052 | 0.157 |
| Basophils number (1000 cells/uL) | 0.069 | 0.084 | 0.012 | 0.056 | 0.100 |
| Red blood cell count (million cells/uL) | 0.829 | 1.083 | 0.148 | 0.898 | 1.595 |
| Hemoglobin (g/dL) | 1.717 | 2.942 | 0.152 | 2.667 | 3.210 |
| Hematocrit (%) | 0.847 | 2.798 | 0.148 | 2.533 | 3.103 |
| Mean cell volume (fL) | 0.494 | 0.943 | 0.082 | 0.853 | 1.154 |
| Mean cell hemoglobin (pg) | 0.848 | 1.307 | 0.045 | 1.251 | 1.432 |
| Mean Cell hemoglobin concentration (MCHC) (g/dL) | 5.139 | 3.380 | 0.105 | 3.195 | 3.610 |
| Red cell distribution width (%) | 3.423 | 1.918 | 0.105 | 1.685 | 2.058 |
| Platelet count (1000 cells/uL) | 2.395 | 2.632 | 0.112 | 2.443 | 2.842 |
| Mean platelet volume (fL) | 1.257 | 1.348 | 0.095 | 1.214 | 1.519 |
| Blood cadmium (nmol/L) | 5.306 | 4.159 | 0.136 | 3.908 | 4.447 |
| Glycohemoglobin (%) | 1.526 | 1.830 | 0.107 | 1.604 | 1.986 |
| C-reactive protein(mg/dL) | 1.877 | 2.264 | 0.107 | 2.024 | 2.405 |
| Direct HDL-Cholesterol (mg/dL) | 1.097 | 1.149 | 0.104 | 0.946 | 1.328 |
| RBC folate (ng/mL) | 0.536 | 0.779 | 0.097 | 0.588 | 1.021 |
| Serum folate (ng/mL) | 1.955 | 1.840 | 0.163 | 1.486 | 2.087 |
| Cotinine (ng/mL) | 2.011 | 1.982 | 0.277 | 1.542 | 2.739 |
| Urinary Total NNAL (ng/mL) | 3.226 | 5.579 | 0.496 | 4.544 | 6.459 |
| Albumin (g/dL) | 0.914 | 0.543 | 0.050 | 0.459 | 0.662 |
| Alanine aminotransferase ALT (U/L) | 0.843 | 0.836 | 0.033 | 0.791 | 0.889 |
| Aspartate aminotransferase AST (U/L) | 0.727 | 0.577 | 0.053 | 0.462 | 0.694 |
| Alkaline phosphotase (U/L) | 2.992 | 1.966 | 0.086 | 1.825 | 2.184 |
| Blood urea nitrogen (mg/dL) | 0.840 | 0.875 | 0.088 | 0.718 | 1.061 |
| Total calcium (mg/dL) | 0.671 | 0.610 | 0.053 | 0.528 | 0.753 |
| Cholesterol (mg/dL) | 0.373 | 0.793 | 0.086 | 0.680 | 1.076 |
| Bicarbonate (mmol/L) | 3.769 | 1.840 | 0.089 | 1.661 | 2.035 |
| Creatinine (mg/dL) | 3.286 | 3.041 | 0.189 | 2.491 | 3.370 |
| Gamma glutamyl transferase (U/L) | 1.023 | 0.689 | 0.045 | 0.628 | 0.823 |
| Glucose, serum (mg/dL) | 4.911 | 2.487 | 0.168 | 2.140 | 2.742 |
| Iron, refigerated (ug/dL) | 1.752 | 1.443 | 0.088 | 1.288 | 1.559 |
| Lactate dehydrogenase (U/L) | 1.050 | 1.102 | 0.066 | 0.973 | 1.202 |
| Phosphorus (mg/dL) | 0.556 | 0.587 | 0.040 | 0.514 | 0.668 |
| Total bilirubin (mg/dL) | 3.366 | 2.555 | 0.291 | 1.841 | 2.853 |
| Total protein (g/dL) | 0.639 | 0.930 | 0.064 | 0.820 | 1.055 |
| Triglycerides (mg/dL) | 0.987 | 1.285 | 0.170 | 1.126 | 1.867 |
| Uric acid (mg/dL) | 2.598 | 2.401 | 0.116 | 2.232 | 2.675 |
| Sodium (mmol/L) | 0.957 | 0.299 | 0.034 | 0.242 | 0.366 |
| Potassium (mmol/L) | 0.654 | 0.664 | 0.059 | 0.559 | 0.778 |
| Chloride (mmol/L) | 2.810 | 1.629 | 0.065 | 1.460 | 1.726 |
| Osmolality (mmol/Kg) | 0.749 | 1.165 | 0.061 | 1.019 | 1.263 |
| Globulin (g/dL) | 0.501 | 0.679 | 0.062 | 0.529 | 0.778 |
| Total Cholesterol (mg/dL) | 0.457 | 0.609 | 0.058 | 0.488 | 0.714 |
| Albumin, urine (ug/mL) | 1.247 | 1.459 | 0.085 | 1.326 | 1.680 |
| Creatinine, urine (umol/L) | 0.963 | 0.992 | 0.108 | 0.769 | 1.193 |
| First albumin creatinine ratio (mg/g) | 0.723 | 0.787 | 0.118 | 0.636 | 1.110 |
| Second albumin (ug/mL) | 0.975 | 2.314 | 1.184 | 1.071 | 5.309 |
| Second creatinine (mg/dL) | 1.768 | 2.056 | 0.525 | 1.459 | 3.329 |
| Second albumin creatinine ratio (mg/g) | 1.176 | 3.321 | 0.825 | 1.783 | 5.326 |
| The volume of urine collection #1 | 1.485 | 2.082 | 0.092 | 1.948 | 2.336 |
| Urine #1 Flow Rate | 3.262 | 2.921 | 0.720 | 1.743 | 4.456 |
| Urine osmolality (mOsm/kg) | 1.308 | 1.675 | 0.179 | 1.352 | 2.242 |
| Hepatitis A Antibody | 0.031 | 0.089 | 0.028 | 0.046 | 0.151 |
| Hepatitis B surface antibody | 0.034 | 0.051 | 0.018 | 0.023 | 0.082 |
Note: Highlighted indicate biomarker selected for univariate logistic regression. Validation on original plus each imputed data set. Random splitting of 60:40 training:validation, λ = 0.001, 50% bagging, 6 maximum number of boosting interactions. Original pseudo-R² = 0.032, imputed data set pseudo-R² ranged from 0.044 to 0.052. Variables selected at this step accounted for more than 50% of the total relative importance: original data was 53.85%; mean of 20 imputation sets was 53.33%.