RESULTS

Uterine cytology and recovery status

Endometrial cytology revealed distinct differences in PMN percentages between cows with suspected endometritis and those with subclinical endometritis (Table 1). Cows with suspected endometritis had a significantly lower mean PMN percentage than cows with subclinical endometritis (7.01 ± 0.28% vs. 11.68 ± 0.52%, respectively; p < 0.05).

Based on the initial cytological evaluation, 30 cows were classified with suspected endometritis and 30 with subclinical endometritis. Among cows with suspected endometritis, 25 recovered, and 5 did not, corresponding to a recovery rate of 83.3%. In contrast, among cows with subclinical endometritis, 15 recovered and 15 did not, resulting in a recovery rate of 50.0%. The recovery rate was significantly higher in cows with suspected endometritis than in those with subclinical endometritis (p < 0.05; Table 1). These findings indicate that greater cytological evidence of uterine inflammation was associated with a lower probability of recovery during the observation period.

Metabolic parameters

Serum concentrations of the metabolic biomarkers NEFAs and BHB at D0 and D30 are presented in Figure 1. Significant differences were observed between uterine health groups and sampling times (p < 0.05). At D0, cows with subclinical endometritis exhibited significantly higher NEFAs and BHB concentrations than cows with suspected endometritis, indicating a greater degree of metabolic imbalance (p < 0.05). In both groups, NEFAs and BHB concentrations decreased significantly from D0 to D30 (p < 0.05), suggesting partial improvement in metabolic status during the observation period.

Initial uterine status | PMN (%) | Total (n) | Recovered (n) | Not recovered (n) | Recovery rate (%)
Suspected endometritis | 7.01 ± 0.28ᵃ | 30 | 25 | 5 | 83.3ᵃ
Subclinical endometritis | 11.68 ± 0.52ᵇ | 30 | 15 | 15 | 50.0ᵇ

Figure 1: Distribution of metabolic biomarkers in dairy cows with suspected and subclinical endometritis at day 0 (D0) and day 30 (D30). (A) Non-esterified fatty acids (NEFAs). (B) β-hydroxybutyrate (BHB). Different superscripts indicate significant differences among groups (p < 0.05).

Biochemical and inflammatory parameters

Serum biochemical and inflammatory biomarkers are summarized in Table 2. Significant differences among uterine health groups and sampling times were observed for Ca, albumin, globulin, TP, and Hp (p < 0.05).

At D0, cows with subclinical endometritis had lower serum Ca and albumin concentrations than cows with suspected endometritis. Ca concentrations were 7.99 ± 0.18 mg/dL and 8.37 ± 0.21 mg/dL, respectively, whereas albumin concentrations were 2.95 ± 0.08 g/dL and 3.22 ± 0.11 g/dL, respectively.

Conversely, globulin, TP, and Hp concentrations were higher in cows with subclinical endometritis than in those with suspected endometritis. The highest Hp concentration was observed in cows with subclinical endometritis at D0 (1.11 ± 0.09 g/dL), followed by cows with suspected endometritis at D0 (0.84 ± 0.06 g/dL), cows with subclinical endometritis at D30 (0.62 ± 0.07 g/dL), and cows with suspected endometritis at D30 (0.43 ± 0.08 g/dL).

Between D0 and D30, Ca and albumin concentrations increased in both groups, whereas globulin and Hp concentrations decreased. Overall, cows with subclinical endometritis exhibited more pronounced biochemical and inflammatory disturbances at baseline, although partial physiological recovery was observed by D30.

Parameter | Suspected D0 | Suspected D30 | Subclinical D0 | Subclinical D30
Ca (mg/dL) | 8.37 ± 0.21ᶜ | 8.70 ± 0.22ᵃ | 7.99 ± 0.18ᵈ | 8.56 ± 0.18ᵇ
Albumin (g/dL) | 3.22 ± 0.11ᵇ | 3.33 ± 0.13ᵃ | 2.95 ± 0.08ᵈ | 3.11 ± 0.12ᶜ
Globulin (g/dL) | 3.67 ± 0.15ᶜ | 3.44 ± 0.17ᵈ | 4.04 ± 0.17ᵃ | 3.71 ± 0.14ᵇ
TP (g/dL) | 6.80 ± 0.05ᶜ | 6.79 ± 0.05ᵈ | 7.01 ± 0.05ᵃ | 6.85 ± 0.03ᵇ
Hp (g/dL) | 0.84 ± 0.06ᵇ | 0.43 ± 0.08ᵈ | 1.11 ± 0.09ᵃ | 0.62 ± 0.07ᶜ

Hormonal parameters

Serum concentrations of PGFM and cortisol at D0 and D30 are presented in Figure 2. Significant differences were observed among uterine health groups and sampling times (p < 0.05).

The highest PGFM concentration was detected in cows with subclinical endometritis at D0, followed by cows with suspected endometritis at D0, cows with subclinical endometritis at D30, and cows with suspected endometritis at D30. This pattern suggests a greater prostaglandin-mediated uterine response in cows with more severe uterine inflammation during the initial assessment.

Figure 2: Serum concentrations of (A) prostaglandin F₂α metabolite (PGFM) and (B) cortisol in dairy cows with suspected and subclinical endometritis at day 0 (D0) and day 30 (D30). Different superscripts indicate significant differences among groups (p < 0.05).

A similar trend was observed for cortisol concentrations. Cows with subclinical endometritis had significantly higher serum cortisol concentrations than cows with suspected endometritis, particularly at D0 (p < 0.05). In both groups, PGFM and cortisol concentrations decreased significantly between D0 and D30 (p < 0.05), indicating reduced endocrine activation and physiological stress during the observation period. Nevertheless, cows with subclinical endometritis consistently maintained higher hormonal concentrations than cows with suspected endometritis throughout the study.

Reproductive performance

Reproductive performance data are presented in Table 3. Significant differences were observed between cows with suspected endometritis and those with subclinical endometritis for all evaluated reproductive variables (p < 0.05).

Cows with suspected endometritis exhibited significantly higher estrus intensity scores than cows with subclinical endometritis (2.33 ± 0.76 vs. 1.50 ± 0.51; p < 0.05). Conversely, cows with subclinical endometritis required significantly more services per conception than cows with suspected endometritis (2.50 ± 0.51 vs. 1.67 ± 0.76; p < 0.05). Likewise, DO were significantly longer in cows with subclinical endometritis than in cows with suspected endometritis (136.20 ± 23.36 vs. 104.27 ± 17.06 days; p < 0.05).

Overall, cows with subclinical endometritis exhibited poorer reproductive performance, characterized by weaker estrus expression, increased insemination requirements, and delayed conception.

Parameter | Suspected endometritis | Subclinical endometritis
Estrus intensity (score) | 2.33 ± 0.76ᵃ | 1.50 ± 0.51ᵇ
SC | 1.67 ± 0.76ᵇ | 2.50 ± 0.51ᵃ
DO (days) | 104.27 ± 17.06ᵇ | 136.20 ± 23.36ᵃ

Correlation between physiological biomarkers and reproductive performance

Pearson correlation analysis between physiological biomarkers and reproductive performance indicators is presented in Figure 3. Several metabolic, biochemical, inflammatory, and hormonal biomarkers showed significant associations with estrus intensity, SC, and DO.

Among all evaluated biomarkers, Hp demonstrated the strongest associations with reproductive performance. Hp was strongly negatively correlated with estrus intensity (r = −0.84, p < 0.01) and strongly positively correlated with SC (r = 0.84, p < 0.01) and DO (r = 0.87, p < 0.01).

Metabolic biomarkers also demonstrated significant, although comparatively weaker, associations with reproductive performance. NEFAs and BHB were negatively correlated with estrus intensity (r = −0.30 to −0.34, p < 0.05) and positively correlated with SC and DO (r = 0.31–0.42, p < 0.05).

Among the biochemical biomarkers, globulin and TP showed positive correlations with SC and DO (r = 0.26–0.45, p < 0.05), whereas albumin was positively correlated with estrus intensity (r = 0.31, p < 0.05) and negatively correlated with SC and DO (r = −0.33 to −0.39, p < 0.01).

Similarly, hormonal biomarkers were associated with reproductive outcomes. PGFM was positively correlated with SC (r = 0.50, p < 0.01) and DO (r = 0.53, p < 0.01), whereas cortisol exhibited weaker positive correlations with SC and DO (r = 0.28–0.29, p < 0.05).

Overall, the correlation analysis demonstrated that poorer reproductive performance was associated with greater metabolic stress, systemic inflammation, and endocrine activation, with Hp exhibiting the strongest association among all evaluated physiological biomarkers.

Figure 3: Heatmap of Pearson correlation coefficients between physiological biomarkers and reproductive performance indicators in dairy cows. Values represent correlation coefficients (r). *p < 0.05; **p < 0.01.