Uhaa I J, Riemann H P, Thurmond M C, Franti C E
Department of Epidemiology and Preventive Medicine, University of California, Davis 95616.
Vet Res Commun. 1990;14(6):461-70. doi: 10.1007/BF00367058.
In a cross-sectional study of bluetongue virus (BTV) and Mycoplasma bovis (M. bovis) infections, a sample of 572 California dairy cows was tested for the presence of antibodies to answer the question: Is it possible to identify and to assess quantitatively the associations between positive antibody test and production? Serum samples collected from these cows during December 1986 were tested for the presence of antibodies to BTV and M. bovis using the enzyme-linked immunosorbent assay (ELISA). Data on milk production were extracted from individual cow sheets of the California Dairy Herd Improvement Association (DHIA) record-keeping system and interfaced with percentage ELISA results for analysis. Univariate and multivariate statistical analyses, using the chi 2 test for categorical variables or Student's t-test for continuous variables and multiple logistic regression respectively, were carried out to evaluate for possible associations between positive antibody tests to each agent and each production variable of interest. Complete data on all variables studied were obtained for 289 (50.5%) cows for M. bovis and 423 (74%) cows for BTV. For cows with complete data on all variables, estimates of the point prevalence of antibodies to BTV and M. bovis were 70.5% and 66.1%, respectively. Results of this study indicated that Guernsey cows were more likely to have a positive BTV test than Holstein cows and that cows in higher lactations were more likely to test positive to BTV ELISA than those in lower lactations (p less than 0.05). Because all cows except those on one farm were Holstein, our confidence in the effect of breed is limited. The association between lactation number and BTV seropositive test may be an age factor identified earlier in the study. For M. bovis, the results of the analysis indicated that seropositive cows were more likely to produce less milk, on a mature equivalent basis (ORadj = 0.96, p = 0.034), and that they had less extended 305 day milk production potential (ORadj = 0.90, p less than 0.0001) than seronegative cows.
在一项关于蓝舌病病毒（BTV）和牛支原体（M. bovis）感染的横断面研究中，对572头加利福尼亚奶牛进行了抽样检测，以确定是否存在抗体，从而回答以下问题：是否有可能识别并定量评估抗体检测呈阳性与生产之间的关联？1986年12月从这些奶牛采集的血清样本，采用酶联免疫吸附测定（ELISA）检测是否存在针对BTV和M. bovis的抗体。牛奶产量数据从加利福尼亚奶牛群改良协会（DHIA）记录保存系统的个体奶牛记录表中提取，并与ELISA结果百分比进行对接以便分析。分别使用卡方检验（用于分类变量）或学生t检验（用于连续变量）以及多元逻辑回归进行单变量和多变量统计分析，以评估针对每种病原体的抗体检测呈阳性与每个感兴趣的生产变量之间可能存在的关联。对于M. bovis，获得了289头（50.5%）奶牛的所有研究变量的完整数据；对于BTV，获得了423头（74%）奶牛的完整数据。对于所有变量都有完整数据的奶牛，针对BTV和M. bovis的抗体点患病率估计分别为70.5%和66.1%。本研究结果表明，与荷斯坦奶牛相比，娟姗奶牛更有可能BTV检测呈阳性，并且处于较高泌乳期的奶牛比处于较低泌乳期的奶牛更有可能BTV ELISA检测呈阳性（p < 0.05）。由于除了一个农场的奶牛外，所有奶牛均为荷斯坦奶牛，因此我们对品种影响的置信度有限。泌乳次数与BTV血清学阳性检测之间的关联可能是该研究中较早确定的一个年龄因素。对于M. bovis，分析结果表明，与血清学阴性奶牛相比，血清学阳性奶牛在成熟当量基础上产奶量更有可能较低（调整后OR = 0.96，p = 0.034），并且其305天的产奶潜力更低（调整后OR = 0.90，p < 0.0001）。
