Skip to main content

Table 1 Predictors selected as candidates for multivariable modeling

From: Farm characteristics and management routines related to cow longevity: a survey among Swedish dairy farmers

Variable

n

Continuous

mean, SD (min, max)

Categorical

Overall

Categories

n (%)

P-valuea

Farmer characteristics

 Age of farmer (year)

225

48.7, 11.3 (20, 79)

  

< 0.001

Herd characteristics

 Herd size (no. of cows)

213

87.6, 84.6 (26, 1000)

  

0.192

 Housing type

227

 

Tie stall

100 (56)

0.017

  

Free stall

127 (44)

 

 Region

228

 

South Sweden

32 (14)

0.089

  

East Sweden

62 (27)

 
  

West Sweden

74 (32)

 
  

Middle Sweden

20 (9)

 
  

North Sweden

40 (18)

 

Farmer attitudes

 Expected future herd size

224

 

Same or fewer

Expand

163 (73)

61 (27)

0.046

 Expectation that the farm will still be operating in 5 year

227

7.6, 2.7 (1, 10)

  

0.144

 Interest in animal breeding (genetic selection)

223

7.9, 2.0 (0.5, 10)

  

0.145

Management factors

 Composition of workforce

222

 

Only family members

Employee(s)

94 (42)

128 (58)

0.141

 Used herd health advice during the last year

182

 

No

Yes

82 (45)

100 (55)

0.029

 Used breeding advice during the last year

204

 

No

Yes

55 (27)

149 (73)

0.123

 Calving occurs in individual calving pens

227

 

Yes

No

78 (34)

149 (66)

0.017

 Time required for drying off (d)

214

7.8, 4.3 (0, 21)

  

0.019

 Call veterinarian for lame cows

210

4.4, 3.1 (0, 10)

  

0.003

 Move lame cows to isolation pen

202

5.3, 3.2 (0, 10)

  

0.029

 Call veterinarian for unhealthy cows

219

7.5, 2.4 (0.1, 10)

  

0.011

 Move unhealthy cows to isolation pen

213

6.1, 3.2 (0, 10)

  

0.121

 Initiate treatment of unhealthy cows on my own

215

5.0, 3.4 (0, 10)

  

0.148

  1. Predictors selected as candidates for multivariable modeling of associations with average herd longevity according to the categories identified in the causal diagram
  2. aUnivariable P-values for the confounders (age, herd size, region), but multivariable P-values for the predictors