Table 4.

Negative binomial regression model in a study of herd-level risk factors for cow mortality in 6,898 Swedish dairy herds from September (Sep.) 2002 to August (Aug.) 2010.

Variable	Category	MR1 ratio	95% CI	P-value2
Calving interval, d	<389.3	1		<0.0001
	389.3 to 403.5	1.05	1.03, 1.08
	403.6 to 421.9	1.14	1.11, 1.18
	≥422.0	1.26	1.22, 1.30
Herd size × season3,4	20 to 29.9, autumn-winter	1		<0.0001
	20 to 29.9, winter-spring	1.36	1.29, 1.43
	20 to 29.9, summer	1.35	1.28, 1.42
	30 to 49.9, autumn-winter	1.09	1.03, 1.14
	30 to 49.9, winter-spring	1.38	1.31, 1.45
	30 to 49.9, summer	1.42	1.35, 1.49
	50 to 99.9, autumn-winter	1.31	1.24, 1.38
	50 to 99.9, winter-spring	1.47	1.39, 1.55
	50 to 99.9, summer	1.57	1.49, 1.65
	≥100, autumn-winter	1.69	1.59, 1.79
	≥100, winter-spring	1.72	1.62, 1.82
	≥100, summer	1.89	1.79, 2.00
Milk yield, kg of ECM	<8,525	1		<0.0001
	8,525 to 9,290	0.97	0.94, 0.99
	9,291 to 9,980	0.90	0.87, 0.93
	≥9,981	0.88	0.85, 0.91
Predominant breed	Swedish Holstein	1		<0.0001
	Swedish Red	0.78	0.76, 0.81
	Mix/other breeds5	0.90	0.88, 0.93
Region6	Södra Götaland (1)	1		<0.0001
	Östra Götaland (2)	1.10	1.05, 1.15
	Västra Götaland (3)	0.99	0.94, 1.04
	Östra Svealand (4)	1.05	0.98, 1.12
	Västra Svealand (5)	1.04	0.97, 1.11
	Norrland (6)	1.13	1.06, 1.19
Year	Sep. 2002 to Aug. 2003	1		<0.0001
	Sep. 2003 to Aug. 2004	1.04	1.01, 1.06
	Sep. 2004 to Aug. 2005	1.15	1.12, 1.18
	Sep. 2005 to Aug. 2006	1.17	1.14, 1.20
	Sep. 2006 to Aug. 2007	1.17	1.14, 1.20
	Sep. 2007 to Aug. 2008	1.10	1.07, 1.13
	Sep. 2008 to Aug. 2009	1.20	1.17, 1.23
	Sep. 2009 to Aug. 2010	1.22	1.18, 1.25

¹Mortality rate.

²Significance level (Wald test) of main effects or combined effects.

³Autumn-winter (September 1 to December 31), winter-spring (January 1 to April 30), and summer (May 1 to August 31).

⁴The combined effect of the main variables.

⁵Crossbreeds, Swedish Polled, and Swedish Jersey.

⁶Numbers in parentheses refer to different regions identified in Figure 1.