Abstract
Animal welfare is increasingly recognised as a core component of sustainable dairy production, yet objective assessment at the herd level remains challenging. This study evaluated whether milk biomarkers can serve as non-invasive indicators of cow welfare. Thirty-seven dairy farms were assessed using the Welfare Quality(®) protocol and various milk analysis parameters. As a first line of results, Spearman correlations revealed strong associations between milk biomarkers and welfare indicators. For example, a higher fat-to-protein ratio was linked to better feeding, lower prevalence of hunger, and improved human-animal relationships. In contrast, elevated somatic cell count and differential somatic cell count were associated with mastitis, lameness, dirtiness, and reduced emotional well-being. Using Principal Component Analysis (PCA), three dimensions were identified, health-hygiene, socio-behavioural, and metabolic stress, explaining 44.7% of variance. K-means clustering distinguished three herd profiles: feeding-metabolic balance, behavioural-comfort, and clinical-hygiene risk. These findings demonstrated that routine milk biomarkers provide integrated, non-invasive information on herd health, behaviour and, comfort. Incorporating routine milk analysis into welfare assessments can support the early detection of issues, facilitate evidence-based decision-making, and promote sustainable dairy management.