Reference intervals and variability of urinary traits in lactating Holstein cows

Monitoring urinary composition can provide valuable information on animal health, physiological status, nutritional deficiencies, and potential renal dysfunction. Although total urine collection over 24 h is considered the gold standard, its application under commercial farming conditions is impractical because of housing constraints, collection devices, and animal stress. Spot urine sampling represents a feasible alternative; however, normalization for urine concentration, for instance by specific gravity as a proxy, is required to remove this type of individual variation. This study aimed to (i) establish reference intervals for urinary traits in clinically healthy lactating Holstein cows, (ii) evaluate the effects of lactation stage and parity on urinary trait variability, and (iii) assess relationships between urinary and milk traits. Spot urine and milk samples were collected from 1,271 cows across 17 farms in northern Italy. Urinary pH, specific gravity, urea, protein, creatinine, allantoin, uric acid, glucose, and the protein-to-creatinine ratio were determined using reference methods, whereas milk composition was analyzed using infrared spectroscopy and flow cytometry. Generalized linear mixed models including urinary specific gravity as a covariate were used to evaluate sources of variation in urinary traits, with parity and stage of lactation as fixed effects and herd-test-date as a random effect. Pearson correlations were computed among urinary traits and between urinary and milk traits. Reference intervals were established in all urine traits investigated by calculating the 2.5th to 97.5th percentiles (95% CI). Urea was the predominant nitrogenous compound, followed by allantoin, creatinine, and uric acid. Urinary traits exhibited wide phenotypic variation, with the protein-to-creatinine ratio showing the highest variability, whereas pH and specific gravity were the least variable traits. Specific gravity was significantly affected by stage of lactation and parity and explained a substantial proportion of the variability of all urinary traits, supporting its usefulness in adjusting for differences in urine dilution. Stage of lactation influenced all urinary traits except pH and uric acid, whereas parity significantly affected only creatinine concentration. Urine concentration and most N-compounds were at the highest levels in early lactation and followed distinct temporal patterns throughout lactation. Herd-test-date explained a considerable proportion of the variance, particularly for urinary pH and urea concentration. Strong correlations were observed between urinary N-compounds, while associations between urinary and milk traits were generally weak; milk urea concentration showed the strongest positive correlations with urinary total N and urea. Results of the present study provide comprehensive reference intervals for urinary traits in lactating Holstein cows and identify key sources of variability, leading the foundations for using spot urine samples for diagnostic and herd management applications.