Table 1. Multiple regression analysis of the anti-F1 IgM and IgG titers.
| (1) | (2) | (3) | (4) | |
| Explained parameter | Day 13 anti-F1 IgM | Day 25 anti-F1 IgG | ||
| Nb rats | 82/118 | 65/118 | 84/118 | 67/118 |
| R Multiple | 0,413 | 0,316 | 0,329 | 0,339 |
| F | 3,13 | 1,07 | 1,89 | 1,3 |
| R2 adjusted (p) | 0,116(0,012) | 0,007(0,385) | 0,051(0,104) | 0,026(0,270) |
| Beta coefficient | ||||
| Dose | 0,151 | 0,215 | 0,274 | 0,313 |
| Sex (M/F) | 0,107 | 0,021 | 0,084 | 0,043 |
| Weight (g) | 0,049 | −0,13 | 0,056 | −0,02 |
| Nb fleas | −0,12 | −0,13 | −0,12 | −0,04 |
| Erythrocyte count (d0) | −0,04 | −0,03 | ||
| Leukocytes count (d0) | −0,12 | 0,031 | ||
| Village | −0,31 | −0,06 | ||
four separate logistic regressions (1) to (4) were conducted on the set of 118 rats from Ambohimasina and Maromanana studied for short term follow up of IgM and IgG, to check effect of parameters (beta coefficient are reported when parameters are used in the regression and plotted in bold when significant at less than p = 0.05).