Log-linear methodology was used to examine relations among caprine arthritis-encephalitis virus (CAEV) seroreactivity and host/management factors in a cross-sectional study of 2,826 goats on 13 California dairies. The CAEV serologic status was associated with age and feeding method (pasteurized/unpasteurized milk), but not with breed. Data from a prevalence survey of 321 goats from 2 additional dairies demonstrated very good fit of the selected log-linear model (P = 1.00), indicating that the model was very appropriate to describe the relations. Odds of seropositivity and odds ratios were generated by use of a logit model derived from the log-linear model. Goats raised by the unpasteurized feeding method were estimated to have been 3.3 times more likely to be CAEV-seropositive than goats fed by the pasteurized method, when adjusted for the effects of age. Goats aged 2, 3, 4, and 5 or greater years were estimated to have been 1.7, 2.6, 4.5, and 5.7 times, respectively, more likely to be CAEV-seropositive than were yearling goats when ratios were adjusted for pasteurization status. Breed, gender, and herd of origin were not associated with CAEV seroreactivity when the effects of other factors were considered. Estimated odds of CAEV seroreactivity and the associated odds ratios for combinations of factor levels are reported. In this study, the magnitude and direction of the associations among CAEV serologic status, age, and pasteurized feeding methods were demonstrated.

Red blood cells from 6 Pygmy goats were determined to be significantly (P less than 0.01) more susceptible to osmotic lysis and mechanical stress than were RBC from 6 Toggenburg goats. Differences in RBC size and shape and adenosine 5'-triphosphate concentration between the 2 breeds were not significant. The differences observed in the in vitro tests may be attributable to differences in RBC membrane composition.