Table 2.

Multivariable GEE analysis of milk ELISA as a continuous (S/P % ELISA readings) and categorical (milk ELISA MAP positive/negative) dependent variable and independent variables.

	Continuous variable (S/P % ELISA readings)			Categorical variable (Milk ELISA MAP positive/negative)
Dependent variable Milk ELISA	Coefficient	P Value significant: P < 0.05	95% C.I.	Odds ratio	P value	95% C.I.	Model (P value <0.001)
Independent variable
Time point
June 20 vs. Maya	17.2	<0.001	14.3, 20.2	11.1	<0.001	5.8, 21.0
July 11 vs. May	5.43	<0.001	2.5, 8.4	2.7	0.004	1.4, 5.1
July 24 vs. May	0.29	0.850	−2.7, 3.3	1.1	0.819	0.5, 2.3
August 8 vs. May	0.94	0.537	−2.1, 3.9	1.1	0.831	0.5, 2.3	Sampling time point
August 21 vs. May	0.42	0.784	−2.6, 3.4	1.1	0.829	0.5, 2.3	Parity
Oct vs. May	1.51	0.322	−1.5, 4.5	1.8	0.105	0.9, 3.5	Breed
Nov vs. May	5.65	<0.001	2.6, 8.7	2.5	0.007	1.3, 4.9	Calving date
Parity
1b vs. 2	−11.2	<0.001	−5.4, −17.0	0.3	0.004	1.6, 10.3
2 vs. 3	11.2	0.001	17.5, 4.7	3.3	0.025	0.1, 0.9
2 vs. 4	8.3	0.002	13.5, 3.1	2.5	0.014	0.2, 0.8

^aMay is the ELISA sample taken pre SICCT. ^bParity 1: 1^st lactation. No significant interactions identified with other independent variables. C.I., confidence interval. Coefficient, difference across the sample population. Statistically significant P values highlighted in bold.