Rajala P J, Gröhn Y T
Department of Population Medicine and Diagnostic Sciences, College of Veterinary Medicine, Cornell University, Ithaca, NY 14853, USA.
J Dairy Sci. 1998 Dec;81(12):3172-81. doi: 10.3168/jds.s0022-0302(98)75883-7.
The effects of dystocia, retained placenta, and metritis on milk yield were studied in 37,776 Finnish Ayrshire cows that calved during 1993; cows were studied for one lactation. Monthly test day milk yields were treated as repeated measurements within a cow in a mixed models analysis. Index categories were created to relate the time of disease occurrence to the test dates and to capture the short-term effects of diseases on milk yield. Diseases other than dystocia, retained placenta, or metritis were categorized into two groups: diseases occurring within 42 d after calving or diseases occurring later than that. The statistical models for each parity included calving season, stage of lactation, and disease variables as fixed effects. For parities 2, 3, and 4 or higher, the cows were grouped into four categories based on previous lactation milk yield, and the disease effect was studied separately for each yield level. An autoregressive covariance structure was used to model the association among the repeated measurements. Models with a 305-d yield as the outcome were also run; the diseases were treated as binary covariates in these models. Dystocia, retained placenta, and early metritis significantly affected milk yield, as indicated by monthly test day milk yields. Late metritis was not associated with milk loss. The impact of the diseases differed across parities and also across different levels of milk yield. Using 305-d milk yield as the milk measure, no diseases were associated with reduced milk yield.
对1993年产犊的37776头芬兰艾尔夏奶牛研究了难产、胎盘滞留和子宫炎对产奶量的影响；对奶牛进行了一个泌乳期的研究。在混合模型分析中，将每月测定日的产奶量视为同一头奶牛的重复测量值。创建索引类别以关联疾病发生时间与测定日期，并捕捉疾病对产奶量的短期影响。难产、胎盘滞留或子宫炎以外的疾病分为两组：产犊后42天内发生的疾病或产犊后42天以后发生的疾病。每个胎次的统计模型包括产犊季节、泌乳阶段和疾病变量作为固定效应。对于第2、3和4胎次及以上胎次，根据前一个泌乳期的产奶量将奶牛分为四类，并对每个产奶量水平分别研究疾病的影响。使用自回归协方差结构对重复测量值之间的关联进行建模。还运行了以305天产奶量为结果的模型；在这些模型中，将疾病视为二元协变量。如每月测定日的产奶量所示，难产、胎盘滞留和早期子宫炎显著影响产奶量。晚期子宫炎与产奶量损失无关。疾病的影响在不同胎次以及不同产奶量水平之间存在差异。以305天产奶量作为衡量产奶量的指标时，没有疾病与产奶量降低相关。
