Predicting periparturient disease risk is of immense value to the dairy industry. Periparturient diseases are interrelated with each other; however, predicting the onset risk of these diseases has predominantly been based on a single blood parameter for a single disease. This study examined a new diagnostic method to predict the risk of periparturient diseases. We conducted cluster analysis of multiple blood constituents from 20 Holstein cattle at 1 week post‐partum, and the cattle were divided into two groups, A or B. We then compared the periparturient and early‐lactation blood constituents of these groups. Group B had significantly higher 3‐hydroxybutyric acid concentrations and were suspected to have subclinical ketosis. Group B also had significantly lower calcium concentrations, with a tendency for subclinical hypocalcemia. We also performed discriminant analysis using blood parameters at 1 week post‐partum, which grouped the population into the same two groups as the cluster analysis based on three variables: inorganic phosphorus, calcium, and either phospholipids or total cholesterol. We further showed that these discriminant functions could be used to predict the risk of periparturient disease even before parturition. Our results indicate that cluster analysis with multiple blood constituents is useful for predicting periparturient disease risks.

中文翻译：

---

聚类分析评估围产期奶牛的疾病风险。

预测围产期疾病风险对乳制品行业具有巨大价值。围产期疾病相互关联；然而，预测这些疾病的发病风险主要是基于单一疾病的单一血液参数。本研究检验了一种预测围产期疾病风险的新诊断方法。我们对产后 1 周的 20 头荷斯坦牛的多种血液成分进行了聚类分析，并将牛分为 A 组或 B 组。然后我们比较了这些组的围产期和泌乳早期血液成分。B 组的 3-羟基丁酸浓度显着升高，怀疑患有亚临床酮症。B 组的钙浓度也显着降低，有亚临床低钙血症的倾向。我们还使用产后 1 周的血液参数进行了判别分析，根据三个变量将人群分为与聚类分析相同的两组：无机磷、钙以及磷脂或总胆固醇。我们进一步表明，这些判别函数甚至可以在分娩前用于预测围产期疾病的风险。我们的结果表明，多种血液成分的聚类分析可用于预测围产期疾病风险。我们进一步表明，这些判别函数甚至可以在分娩前用于预测围产期疾病的风险。我们的结果表明，多种血液成分的聚类分析可用于预测围产期疾病风险。我们进一步表明，这些判别函数甚至可以在分娩前用于预测围产期疾病的风险。我们的结果表明，多种血液成分的聚类分析可用于预测围产期疾病风险。