A multiple correspondence analysis of necropsy findings in non-caged laying hens that died during the production period

Causes of normal mortality in laying hens have received relatively little scientific attention, despite health, welfare, economic and environmental implications. Targeting control and preventative measures to currently relevant flock-specific mortality causes is essential for successful outcomes in commercial flocks. Monitoring of mortality causes is a prerequisite. However, diagnostic criteria are not consistently reported and terminology varied between previous studies. The objective of this study was to investigate associations between pathological findings in dead laying hens with an explorative approach applied to a previously obtained dataset on pathological findings in Danish laying hens from seven non-caged commercial flocks. This approach may provide new insights compared to descriptive analysis of predefined diagnoses. A dataset with pathological findings across 49 pathological variables in 1,648 laying hens was analyzed with multiple correspondence analysis to explore multidimensional associations between pathological findings in laying hens and how patterns in combinations of findings of individual hens may cluster as diagnoses or tentative causes of death. Unlike descriptive statistics, the multivariate approach enabled us to illustrate the complexity of pathological processes. The first four dimensions of the multiple correspondence analysis were retained, accounting for 20.2% of the variance. The results indicated clustering of hens suggesting diagnoses of similar expected chronicity and general etiology (infectious versus non-infectious). The most dominant clusters corresponded to the two most common causes of mortality diagnosed on the same hens and reported in previous studies: salpingitis-peritonitis and cannibalism. In addition, the results suggested at least two different clusters of hens that died due to cannibalism (acute or prolonged course with concurrent pathologies). These may point to relevant differences in etiology and pathogenesis that should be explored in future studies. We suggested recommendations for time-efficient field necropsies and preventative and control measures to target the most common causes of mortality in laying hens. The results may be used by farmers and their advisors to improve monitoring of health and welfare of laying hen flocks in non-cage housing systems.