Table 8.
Variables considered for inclusion in two regression models in the cohort study of post-weaning diarrhea
| Variables | Scale of measurement | Model 1 (Cox PH) | Model 2 (Mixed linear) |
|---|---|---|---|
| Producer | Dichotomous (A, B) | Covariate | Fixed effect |
| Pen in nursery | Nominal (n = 8) | Covariate | Fixed effect |
| Sow/pen before weaning | Nominal (n = 45) | Random effect | |
| Dam | Nominal (n = 30) | Shared frailty | Random effect |
| Dam parity | Discrete (1–7) | Covariate | Fixed effect |
| Total (still + live born) litter size, dam | Discrete (total n pigs 14–28) | Covariate | Fixed effect |
| Proportion of litter stillborn | Proportion (0–1) | Covariate | Fixed effect |
| Sex at insertion to nursery | Nominal (castrated male, intact male, female) | Covariate | Fixed effect |
| Weight at birth | Continuous (grams) | Covariate | Fixed effect |
| Weight at birth, dichotomized | Dichotomous (low birth weight, normal or high birth weight) | Covariate | |
| Moved in the farrowing unit | Dichotomous (stayed with dam, moved) | Covariate | Fixed effect |
| Time of weaning |
Dichotomous (weaned in the farrowing unit, weaned at insertion to nursery) A Time varying effect was added |
Covariate | Fixed effect |
| Weight at insertion | Continuous (grams) | Covariate | Fixed effect |
| Body condition score at insertion to nursery | Dichotomous (normal, below normal) | Covariate | Fixed effect |
| Days with hollow flanks | Discrete (n days with hollow flanks 0–14) | Fixed effect | |
| Days suffering from diarrhea | Nominal: (0 days, 1 day, 2 days, 3 days, 4–6 days) | Fixed effect |
Variables considered for inclusion in two regression models investigating risk factors for porcine post-weaning diarrhea (Model 1, Cox proportional hazards model), and factors associated with altered growth rate the first 14 days after insertion to the nursery unit (Model 2, mixed linear regression model)