Table 4.

Results of the final multivariable model, using four-tooth (30-months-old at breeding) predictors, showing the OR (95% CI) for risk of four-tooth mortality between pregnancy diagnosis (PD) and weaning

Variable	Category	Odds ratio	95% CI	p-value
Cohort
	Farm A 2010-born	1	-	-
	Farm A 2011-born	0.537	0.404–0.712	< 0.001
	Farm B	0.528	0.417–0.667	< 0.001
Four-tooth PD BCS
	2.0	1	-	-
	2.5	0.512	0.361–0.725	< 0.001
	3.0	0.534	0.369–0.772	< 0.001
	3.5	0.801	0.489–1.312	0.378
Effect of PD result at average weight1
Litter size (PD result)
	Single-bearing	1	-	-
	Multiple-bearing	1.487	1.160–1.906	0.002
Effect of weight at different PD results2
Litter size (PD result)
	Single-bearing	0.918	0.886–0.952	< 0.001
	Multiple-bearing	0.960	0.336–2.745	0.939

¹There was a significant interaction between weight and PD result. Therefore, this shows, for a ewe of average weight, the effect of PD result (litter size)

²There was a significant interaction between PD (litter size) and weight. Therefore, this shows, for each litter size (single- vs. multiple-bearing), the effect of increasing the weight by 1 kg