Associations of postpartum health with progesterone after insemination and endocrine signaling during early pregnancy in dairy cows

Abstract

Listen

Translate article

Our objective was to investigate associations of postpartum health with serum progesterone (P4) concentrations after first artificial insemination (AI), expression of interferon-stimulated genes (ISG) in blood leukocytes, and serum pregnancy-associated glycoprotein (PAG) during early pregnancy in dairy cows. In this observational study, Holstein cows (n = 475) from 2 commercial herds were enrolled 3 wk before expected parturition. Body condition score (BCS) and lameness were assessed at enrollment, and total Ca, haptoglobin, and nonesterified fatty acids (NEFA) were measured in serum at 2 and 6 ± 2 d in milk (DIM). Blood β-hydroxybutyrate and metritis were evaluated at 4, 8, 11, and 15 ± 2 DIM, and purulent vaginal discharge (PVD) and endometritis based on endometrial cytology were diagnosed at 35 DIM. Onset of cyclicity was assessed by biweekly P4 measurements. The BCS was assessed at enrollment and at 63 DIM, and lameness at enrollment and at 21 and 49 DIM. First AI was based on estrus detection until ∼75 DIM, or synchronization for timed AI thereafter. Serum P4 was measured at d 8 and 12 after first AI; expression of ISG (ISG15 and RTP4) in blood leukocytes was assessed at d 19; and PAG in serum of pregnant cows was measured at d 29, 33, and 40. Multivariable linear regression models were built including health variables (with metabolites categorized via receiver operating characteristic curve analysis) and covariates (season, milk yield, AI method, DIM), accounting for pregnancy status (for ISG and PAG models), repeated measures (for P4 and PAG models), and herd as random. Results are reported comparing cows affected versus unaffected by each predictor. Serum P4 concentrations were lesser at d 8 after AI in cows that had total Ca ≤2.09 mM (3.6 vs. 4.0 ± 0.2 ng/mL) at 2 DIM or a clinical disease (3.7 vs. 4.0 ± 0.2 mg/mL), and P4 tended to be lesser at d 8 in cows that had PVD (3.6 vs. 4.0 ± 0.2 ng/mL) or any uterine disease (3.7 vs. 4.1 ± 0.2 ng/mL). Among cows pregnant at d 29, relative ISG15 expression (fold change) at d 19 was greater in cows that had lameness (5.10 [4.18-6.18] vs. 3.14 [3.10-3.19]) but tended to be lesser in cows that had endometritis (3.20 [2.89-3.54] vs. 4.29 [4.20-4.35]), and relative expression of RTP4 tended to be lesser in cows that had displaced abomasum (1.02 [0.62-1.67] vs. 1.75 [1.73-1.78]). Serum PAG was lesser at d 29 (3.8 vs. 4.2 ± 0.2 sample-to-positive ratio [S:P]) in cows that had NEFA ≥0.73 mM at 6 DIM, and lesser at d 33 (3.9 vs. 4.5 ± 0.2 S:P) and 40 (3.1 vs. 3.7 ± 0.2 S:P) in cows that had prepartum BCS ≥3.75. These findings suggest long-term effects of metabolic or reproductive tract disorders on luteal function after first AI, pregnancy recognition signaling, and placental function during early pregnancy. These associations may partially explain some of the pathways linking postpartum health and fertility in dairy cows.