TABLE 1.

Regression model results showing the association between numbers of VFs and log SCC and inflammation severity

VF gene type	No. of VF genes	Linear regressiona			Logistic regressionb
		Coefficient	Standard error	P value	Coefficient	Standard error	P value
Adherence	28	0.067	0.036	0.065	−0.020	0.039	0.607
Exoenzyme	21	0.098	0.055	0.078	0.076	0.060	0.199
Host immune evasion	20	0.045	0.023	0.054	0.074	0.025	0.003
Iron uptake	29	0.011	0.029	0.714	−0.006	0.031	0.858
Toxin	93	0.067	0.034	0.045	0.019	0.040	0.638
Total	191	0.024	0.010	0.017	0.015	0.011	0.168

^aMixed-effects linear regression using log SCC as the outcome. The coefficient represents the increase in log SCC (log number of cells/milliliter) for each additional VF detected.

^bMixed-effects ordinal logistic regression using inflammation severity as the outcome (measured as low, medium, and high SCC and clinical mastitis in order of increasing severity). The coefficient represents the increase in odds of the inflammation being more severe for each additional VF detected.