| Date/Time: | 8/29/2026 14:15 |
| Author: | Dan Lin |
| Clinic: | Cornell University |
| City, State, ZIP: | Ithaca, NY 14850 |
Dan Lin, PhD
1
;
Jessica McArt, DVM, PhD
1
;
1Department of Population Medicine and Diagnostic Sciences, College of Veterinary Medicine, Cornell University, Ithaca, NY, 14850
Postpartum inflammation drives transition cow disease, yet early identification of at-risk cows remains challenging. We hypothesized that serum haptoglobin (Hp) trajectory clustering could define distinct postpartum inflammation phenotypes linked to disease risk, metabolic dysregulation, and detectable non-invasive signals.
Forty-five periparturient Holstein cows were monitored from d −14 through d 14 postpartum for blood inflammatory and metabolic markers (sampled at 8-h intervals on d 1–4), milk composition (8-h resolution, d 1–14), and sensor-derived rumination and activity. K-means clustering on Hp trajectories defined high-inflammation (HI, n = 11) and low-inflammation (LI, n = 34) phenotypes. Adjusted risk ratios for transition diseases were estimated via modified Poisson regression, and single-timepoint Hp cutoffs were externally validated in an independent herd (n = 56).
HI cows exhibited elevated rectal temperature, lower albumin, and a left shift in neutrophils from d 2, preceding the Hp rise by approximately 1 d, followed by significantly lower serum tCa, DMI, and higher NEFA and total bilirubin from d 2–4 (all FDR p < 0.05). Within HI cows, Hp co-varied with tCa (r = −0.88), DMI (r = −0.64), and TBILI (r = 0.67), forming a tightly integrated inflammatory–metabolic syndrome absent in LI cows. HI cows had markedly elevated risk for metritis and retained placenta (adj. RR = 32.1, p < 0.001 at d 4) and dyscalcemia (adj. RR = 5.7, p = 0.003), with disease events clustering within the same individuals, a comorbidity pattern not captured by hypocalcemia alone. These associations replicated in the validation herd: discovery-derived Hp cutoffs yielded adj. RR = 18.6 (p = 0.002) for dyscalcemia at d 3, and SAA concentrations diverged sharply between groups from d 2 onward (p < 0.001). Non-invasively, HI cows showed depressed rumination and activity at d 2–4, lower milk lactose from d 3–14, and elevated milk NEFA and acetone from d 3–5, mirroring blood markers with an approximately 1 d lag.
Hp trajectory clustering identifies a postpartum high-inflammation phenotype characterized by coordinated metabolic dysregulation and elevated multi-disease risk, validated across herds and detectable through non-invasive milk and sensor signals within the first days of lactation.