Establishing age-specific reference intervals for peripheral blood lymphocyte counts in lactating Holstein dairy cows...

Lymphocyte count (LC) is an important hematological parameter in dairy cows. Historical studies establishing LC reference intervals often included cows that were not tested for bovine leukosis virus (BLV). As a result, BLV-infected cows showing subclinical lymphocytosis were possibly included in those studies, leading to higher LC reference intervals. Because BLV (among other pathogens or processes) can cause persistent lymphocytosis, including BLV-infected cows when establishing LC reference intervals could lead to their overestimation. Our objectives were to propose updated LC reference intervals for Holstein cows and to estimate the most discriminant LC threshold for differentiating cows belonging to the low- versus high-LC distributions. We conducted a cross-sectional observational study to collected blood samples from 631 Holstein cows, irrespective of their BLV status, across 8 herds in Québec, Canada. Blood samples were analyzed on the day of collection to determine LC in 10⁹ cells/L. Using Gaussian finite mixture models, we identified the 2 most probable Gaussian LC distributions representing healthy versus abnormal cows within the complete dataset (i.e., all sampled lactating cows analyzed together, regardless of age), and within 2 age groups: young (2-4 yr-old lactating cows) and older cows (>4-year-old lactating cows). We used the identified distributions to propose LC reference intervals for Holstein cows. Then, we determined the most discriminant LC thresholds to distinguish cows belonging to the low- versus high-LC distributions (the latter being presumed to be affected by either a short-term or chronic immunological alteration). To assess persistent lymphocytosis, we collected a second set of blood samples 4 mo later from cows with an initial LC ≥5.5 × 10⁹ cells/L. The LC reference intervals were 1.5 to 5.5 × 10⁹ cells/L in the complete dataset, and 1.9 to 6.0 and 1.6 to 4.5 × 10⁹ cells/L, in young and older cows, respectively. The most discriminant LC thresholds were ≥5.5 × 10⁹ cells/L in the complete dataset, ≥5.5 × 10⁹ cells/L for young cows, and ≥4.5 ×  10⁹ cells/L for older cows. The probability of belonging to the Gaussian LC distribution representing presumably abnormal values from cows affected by either a short-term or chronic immunological alteration increased markedly with increasing LC but was also high for cows with a very low LC (<1.5 × 10⁹ cells/L). Persistent lymphocytosis was observed in 77% (49/64) of cows in the young group and 98% (55/56) of cows in the older group. The reference LC intervals proposed are lower than those reported in previous studies and reflect an expected age-related variation. The proposed LC thresholds provide a practical marker of chronic lymphocytosis.