Table 4.
Multivariable log-linear modeling for the association of TNF-α with potential explanatory variables
| Variables | Multivariable model for log10-TNF-α level | ||
|---|---|---|---|
| B coefficient | 95% CI | P value | |
| IL-10* | 0.561 | (0.343, 0.420) | < 0.0001 |
| IL-6* | 0.176 | (0.106, 0.207) | < 0.0001 |
| Sex (male) | −0.038 | (−0.081, 0.016) | 0.149 |
| Village (Vuga) | −0.024 | (−0.088, 0.039) | 0.126 |
| 11–12 years old | 0.135 | (0.078, 0.244) | < 0.0001 |
| Wasted† | −0.063 | (−0.134, 0.009) | 0.087 |
| Stunted‡ | 0.020 | (−0.041, 0.077) | 0.545 |
| Malaria | 0.026 | (−0.038, 0.105) | 0.356 |
| S. haematobium§ | 0.056 | (−0.006, 0.105) | 0.080 |
| Hookworm | −0.025 | (−0.103, 0.040) | 0.389 |
Final models adjusted for sex, age category, location, wasting, stunting, anemia, individual infections (S. haematobium, malaria, hookworm), and significant interactions for infection by two or more parasites.
*
log10 of IL-10 and IL-6.
†
Wasting: Body mass index-for-age Z score (BAZ) < −2 SD.
‡
Stunting: Height-for-age Z score (HAZ) < −2 SD.
§
Positive infection status is defined by S. haematobium egg +/antibody +, egg –/antibody +, or egg +/antibody –.