Skip to main content
. 2014 Jan 8;9(1):e84758. doi: 10.1371/journal.pone.0084758

Table 3. Univariate models.

Variable OR (95% CI) QIC
Age at the test 1.024 (1.0021.046) 562.28
Previous Lyme diagnosis 7.437 (4.41912.515) 426.69
BEHAVIOR
Hours in vegetation 1.469 (1.0811.996) 562.22
Owning a dog 1.149 (0.719–1.835)
Owning a cat 1.294 (0.788–2.125)
Owning a different pet 1.960 (0.826–4.649)
Frequency of deer seen on property 0.976 (0.727–1.310)
Tick bite within the past year 1.111 (0.737–1.674)
Tick bite within the past yearon Block Island 1.136 (0.659–1.960)
Use of any protective measure 0.726 (0.454–1.161)
Repellant 0.683 (0.337–1.386)
Protective clothing 0.456 (0.2730.761) 505.14
Avoiding brush 0.801 (0.490–1.309)
Tick checking 0.786 (0.495–1.247)
Landscape-related tickcontrol measures 1.001 (0.605–1.654)
Occupational exposureto tick habitat 0.639 (0.263–1.556)
LANDSCAPE
Lawn largest patch index 0.899 (0.699–1.156)
Lawn edge density 1.009 (0.809–1.258)
Shrub class area 1.152 (0.941–1.410)
Shrub largest patch index 1.204 (0.965–1.503)
Shrub total edge 1.109 (0.938–1.310)
Shrub edge density 1.283 (1.0151.621) 566.68
Shrub landscape shape index 1.158 (0.944–1.419)

Univariate logistic regression models of the association between human behaviors and landscape metrics and positive Lyme disease serology. Statistically significant results at p<0.05 are indicated in bold.