Hematologic and biochemical profiles in Standardbred mares during peripartum

Theriogenology 81 (2014) 526–534

Contents lists available at ScienceDirect

Theriogenology journal homepage: www.theriojournal.com

Hematologic and biochemical profiles in Standardbred mares during peripartum Jole Mariella*, Alessandro Pirrone, Fabio Gentilini, Carolina Castagnetti Department of Medical Veterinary Sciences, University of Bologna, Ozzano dell’Emilia, Bologna, Italy

a r t i c l e i n f o

a b s t r a c t

Article history: Received 11 June 2013 Received in revised form 24 October 2013 Accepted 5 November 2013

The purposes of this study were to determine physiological changes occurring in hematologic and biochemical parameters in mares between the last month of gestation and the first week after parturition. If a significant change was observed with respect to the reference interval of an adult horse, a further aim of the study was to establish different reference intervals. Blood samples were collected from 62 healthy pregnant Standardbred mares. Seventeen nonpregnant and nonlactating mares were used as a control group. In pregnant mares, blood sampling was conducted every three days from 1 month before the expected foaling date (335 days after the last mating), at parturition, and 7 days after foaling. The barren mares in the control group were sampled once. Results from samples collected 20 and 10 days before parturition, at parturition, and 7 days after were considered in the statistical analysis. A parametric method for all the parameters studied was used to establish reference intervals. Results were compared by repeated measures ANOVA. When significant differences were observed in relation to sampling time, a post hoc analysis was performed (Tukey test). The one-way ANOVA test followed by Dunnett’s test was performed to evaluate the presence of a significant difference between each sampling time and the control group. Any significant difference in the blood count parameters at different sampling times was observed by repeated measure ANOVA. Hemoglobin (P < 0.01) and hematocrit (P < 0.01) 7 days after parturition and white blood cell count (P < 0.01) at parturition were significantly different from the control group. Erythrocyte indices and platelet count were within the normal reference intervals as established in the control group. In the biochemical panel, gamma-glutamyltransferase, creatinine, glucose, biliar acids, total protein, albumin-to-globulin ratio, and calcium were significantly different at different sampling times. Moreover, serum concentration of creatine kinase, aspartate aminotransferase, creatinine, blood urea nitrogen, glucose, lactate, total protein, albumin, albumin-to-globulin ratio, calcium, magnesium, sodium, chloride, potassium, and total, direct, and indirect bilirubin was different from that of the control group. Remarkable changes were not observed in alkaline phosphatase, triglyceride, and fibrinogen concentrations. Temporal changes in the hematologic and biochemical parameters observed in the present study in the peripartum and the differences with reference intervals made up for nonpregnant and nonlactating mares could be used to better evaluate the conditions of periparturient mares. Ó 2014 Elsevier Inc. All rights reserved.

Keywords: Mare Hematobiochemical profile Peripartum

1. Introduction

* Corresponding author. Tel.: þ39 (0)512097989; fax: þ39 (0)512097568. E-mail address: [email protected] (J. Mariella). 0093-691X/$ – see front matter Ó 2014 Elsevier Inc. All rights reserved. http://dx.doi.org/10.1016/j.theriogenology.2013.11.001

Each animal species needs specific reference intervals of hematologic and biochemical parameters for an appropriate interpretation of the results obtained from blood

J. Mariella et al. / Theriogenology 81 (2014) 526–534

samples. Moreover, each analyte could have a distinct reference interval for age, breed, and reproductive status. The hematologic and biochemical changes that occur throughout pregnancy could make the interpretation of laboratory results susceptible to misinterpretation. In women, it is well-known that, during pregnancy, reference intervals are different from the nonpregnant state [1,2]. Hematologic and biochemical changes in peripartum mares were studied little [3–6], although they could be useful to correctly identify metabolic diseases such as hyperlipidemia, and subclinical hepatopathy, reproductive disorders caused by malnutrition, emergencies during foaling, and postpartum diseases such as puerperal fever and metritis. Those studies evaluated the entire period of pregnancy: In one study, pregnancy was divided into three periods of similar duration and the third period ranging from 231 days to the end of gestation [6]; in other studies, the hematobiochemical analysis was repeated once a month [3–5]. No one of those studied thoroughly investigated the last month of pregnancy when most mares received the last veterinarian examinations before parturition. The purposes of this study were to determine whether significant changes occur in hematologic and biochemical parameters during the last month of gestation, at parturition, and at 7 days of lactation, and whether changes are substantial enough to establish specific reference intervals. 2. Materials and methods 2.1. Animals Blood samples were collected from 62 Standardbred mares. Thirty mares were housed at Scuderia Trio (Ozzano dell’Emilia, Bologna, Italy) and 32 were hospitalized at the Equine Perinatology Unit “Stefano Belluzzi” of the Department of Medical Veterinary Sciences, University of Bologna, during the 2006 to 2009 foaling seasons. The mares were hospitalized because the owners requested an attended parturition. The farms were located in the same geographic area (district of Bologna); therefore, the mares were in similar pastures and received similar quality of hay, although no attempt was made to strictly standardize feeding procedures and amounts. The experimental design was approved by the Ethic-Scientific Committee for Experiments on Animals of the University of Bologna, in accordance with Decree Law 116/92, and approved by the Ministry of Health. Orally informed consent for mares’ participation was given by the owners. The age of the mares ranged from 6 to 21 years. The mares received a complete clinical examination at admission and were diagnosed as clinically normal. All the mares foaled spontaneously, and remained healthy and clinically free of disease during the study period. 2.2. Sample collection and handling In pregnant mares, jugular blood samples were collected every 3 days from 1 month before the expected foaling date (335 days after the last mating), at parturition, and 7 days after foaling. In mares in the control group, jugular blood samples were collected once. To reduce

527

circadian variations, all samples were collected in the morning, between 9 and 11 hours, except for the one collected at parturition. Blood samples were divided into four aliquots collected in 2.5-mL tubes containing K3EDTA, in 5-mL tubes with gel clotting activator, in 5-mL tubes containing sodium citrate, and in 1.5-mL tubes containing sodium fluoride and potassium oxalate (NaF/KOx), respectively. We used EDTA tubes for hematologic studies, gel clotting activator tubes for clinical biochemistry, citrate tubes for fibrinogen, and NaF/KOx tubes for lactate (Lac). All the K3EDTA tubes were refrigerated (0  C–4  C) and analyzed within 24 hours from withdrawal. Gel clotting activator tubes and citrate and NaF/KOx tubes were centrifuged at 3000 rpm for 10 minutes after withdrawal. Citrate and NaF/KOx plasma and serum were harvested and transferred into plastic tubes. Finally, the samples were stored at 20  C and analyzed within 2 months after collection. All analyses were performed at the Veterinary Clinical Pathology Service of the Department of Veterinary Medical Sciences, University of Bologna. 2.3. Complete blood count The K3EDTA tubes were first placed on Vortex (Reamix 2789; Hecht Assistent, Sondheim/Rhon, Germany); then, the samples were analyzed with an automated cell counter (CELL-Dyn 3500 R, Abbott Laboratories, Santa Clara, CA, USA). The red blood cell count (1012/L), hematocrit (Hct proportion of 1.0), hemoglobin concentration (g/L), white blood cell (WBC) count (109/L), mean corpuscular volume (fL), mean corpuscular hemoglobin (pg), mean corpuscular hemoglobin concentration (g/L), platelet count (109/L), and mean platelet volume (fL) were measured. Specimens containing clots or grossly hemolyzed were excluded. The quality control of the cell counter was performed every day. 2.4. Biochemical and fibrinogen analyses In each sample, the concentrations of creatine kinase (CK;

mkat/L), aspartate aminotransferase (AST; mkat/L), alkaline phosphatase (ALP; mkat/L), gamma-glutamyltransferase (GGT; mkat/L), creatinine (Cre; mmol/L), blood urea nitrogen (BUN; mmol/L), Lac (mmol/L), glucose (Glc; mmol/L), total, direct, and indirect bilirubin (Bt, Bd, Bi, respectively; mmol/L), biliar acids (mmol/L), triglyceride (mmol/L), total protein (TP; g/L), albumin (Alb; g/L), Alb-to-globulin ratio (A/G), ionized calcium (Ca; mmol/L), magnesium (Mg; mmol/L), sodium (Na; mmol/L), potassium (K; mmol/L), and chloride (Cl; mmol/L) were measured using an automated clinical chemistry analyzer (Chemistry Analyzer Olympus AU400, Beckman Coulter, Milan, Italy). Fibrinogen (mmol/L) was measured with a turbidimetric assay (Turbidimetric Fibrinogen, INstruchemie BV, Delfzijl, The Netherlands). Specimens containing clots or that were grossly hemolyzed were excluded. 2.5. Statistical analyses Results from the samples collected 20 days (T  20) and 10 days (T  10) before parturition, at parturition (Tp), and 7 days (T þ 7) after parturition were considered for

528

J. Mariella et al. / Theriogenology 81 (2014) 526–534

statistical analysis. The Kolmogorov–Smirnov test was used to test if the distribution of the data was Gaussian. If the distribution was not Gaussian, the data were logtrasformed and the distribution was reassessed. The statistical method for the analysis of a reference interval was selected based on the number of samples and distribution of the data. Because a limited number of samples were collected, reference intervals were calculated by robust (distribution-independent) or parametric (if Gaussianity can be established) methods and 90% of confidence interval was calculated, as recommended by the International Federation of Clinical Chemistry and the American Society of Veterinary Clinical Pathology [7]. Results were compared by ANOVA for repeated measures. When significant differences were observed in relation to sampling time, a post hoc analysis was performed (Tukey test) on all pairs. The one-way ANOVA test followed by Dunnett’s test was performed to evaluate the presence of a significant difference between each sampling time and the control group. A Pvalue less than 0.05 was considered significant. All analyses were carried out using commercial software (MedCalc

Software Version 12.3.0, Broekstraat 52, 9030 Mariakerke, Belgium). 3. Results Not all samples collected were used for the study because, as mentioned, some samples contained clots or were grossly hemolyzed. Moreover, not all mares were sampled at each time point. The correct numbers of the sample for each time point and each parameter were indicated in Tables 1 and 2. However, because the study dealt with a particular physiological state, the number of animals was sufficient to establish a valid reference range as suggested by the Quality Assurance and Laboratory Standards Committee in the guidelines for the determination of reference intervals in veterinary species [7]. Outliers were not detected during the data analysis. The distribution of the data was Gaussian and for some parameters, the logtransformation was necessary. The Hct, red blood cell count, and mean corpuscular volume at T  10; WBC count at T  20 and T þ 7; AST, Bt, and Bi at T  20, serum ALP,

Table 1 Value of hematologic parameters in periparturient mares and in control group. Parameter

T  20

Hemoglobin concentration (g/L) Reference interval 146  2 (10) 90% CI for lower limit 119–131 90% CI for upper limit 160–173 No. of samples 23 Hematocrit (proportion of 1.0) Reference interval 0.41  2 (0.03) 90% CI for lower limit 0.34–0.37 90% CI for upper limit 0.44–0.48 No. of samples 23 Red blood cells (1012/L) Reference interval 8.5  2 (0.7) 90% CI for lower limit 6.9–7.7 90% CI for upper limit 9.4–10.2 No. of samples 20 White blood cells (109/L) Reference interval 9.3  2 (2.1) 90% CI for lower limit 4.1–6.5 90% CI for upper limit 12.1–14.6 No. of samples 24 Platelets (109/L) Reference interval 14.4  2 (3.8) 90% CI for lower limit 4.7–9.2 90% CI for upper limit 19.6–24.1 No. of samples 24 Mean corpuscular hemoglobin (pg) Reference interval 17.1  2 (0.9) 90% CI for lower limit 14.8–15.9 90% CI for upper limit 18.4–19.5 No. of samples 24 Mean corpuscular hemoglobin concentration (g/L) Reference interval 358  2 (7) 90% CI for lower limit 339–348 90% CI for upper limit 369–378 No. of samples 24 Mean corpuscular volume (fL) Reference interval 47.8  2 (2.5) 90% CI for lower limit 41.4–44.4 90% CI for upper limit 51.2–54.1 No. of samples 24

T  10

Tp

Tþ7

Control group

147  2 (15) 110–124 170–183 44

148  2 (14) 112–126 169–182 40

134  2 (16)a 92–113 155–176 20

147  2 (8) 127–137 157–168 17

0.41  2 (0.04) 0.34–0.36 0.47–0.50 44

0.42  2 (0.04) 0.34–0.37 0.47–0.50 40

0.36  2 (0.04)a 0.28–0.33 0.44–0.49 20

0.42  2 (0.02) 0.35–0.39 0.44–0.48 17

8.6  2 (0.9) 6.3–7.1 10.0–10.1 42

8.9  2 (1.0) 6.5–7.4 10.5–11.4 40

8.0  2 (1.0) 5.3–6.7 9.4–10.7 20

8.9  2 (0.6) 7.4–8.2 9.6–10.4 17

9.4  2 (1.6) 5.6–7.0 11.8–13.2 43

10.2  2 (2.6)a 4.1–6.4 14.1–16.4 40

9.1  2 (1.9) 4.2–6.7 11.6–14.0 20

7.9  2 (1.3) 4.3–6.2 9.5–11.4 17

11.4  2 (4.1) 1.6–5.2 17.6–21.2 42

12.5  2 (4.1) 2.6–6.3 18.6–22.3 40

13.2  2 (3.2) 4.9–9.0 17.4–21.6 20

12.1  2 (2.6) 5.1–8.9 15.2–19.0 17

17.1  2 (0.9) 14.9–15.7 18.5–19.3 43

16.9  2 (1.1) 14.3–15.2 18.5–19.5 40

16.8  2 (0.6) 15.2–16 17.6–18.4 20

16.6  2 (0.7) 14.8–15.8 17.4–18.4 17

356  2 (8) 336–343 369–377 43

354  2 (9) 332–341 368–376 41

356  2 (5) 342–349 362–369 20

355  2 (3) 346–351 359–363 17

47.9  2 (2.7) 41.4–43.8 51.9–54.3 43

47.7  2 (2.7) 41.1–43.6 51.7–54.2 40

47.4  2 (1.9) 42.5–44.9 49.9–52.3 20

46.7  2 (1.9) 41.5–44.2 49.1–51.9 17

Abbreviations: CI, confidence interval; T þ 7, 7 days after parturition; T  10, 10 days before parturition; T  20, 20 days before parturition; Tp, at parturition. a Significantly different from the control group.

J. Mariella et al. / Theriogenology 81 (2014) 526–534

529

Table 2 Value of biochemical parameters in periparturient mares and in control group. Parameter

T  20

Creatine kinase (mkat/L) Reference interval 2.3  2 (0.6)a 90% CI for lower limit 0.7–1.4 90% CI for upper limit 3.1–3.8 No. of samples 27 Serum alkaline phosphatase (mkat/L) Reference interval 4.1  2 (2.1) 90% CI for lower limit 0–1.1 90% CI for upper limit 7–9.3 No. of samples 28 Aspartate aminotransferase (mkat/L) Reference interval 4.4  2 (1.5) 90% CI for lower limit 2.1–2.8 90% CI for upper limit 6.3–8.6 No. of samples 27 Gamma-glutamyltransferase (mkat/L) Reference interval 0.23  2 (0.06) 90% CI for lower limit 0.11–0.15 90% CI for upper limit 0.33–0.46 No. of samples 28 Creatinine (mmol/L) Reference interval 150.3  2 (26.5)a,b 90% CI for lower limit 79.6–114.9 90% CI for upper limit 185.6–221 No. of samples 28 Blood urea nitrogen (mmol/L) Reference interval 13.4  2 (3)a 90% CI for lower limit 5.7–8.9 90% CI for upper limit 17.9–21.1 No. of samples 28 Total bilirubin (mmol/L) Reference interval 44.5  2 (20.5)a 90% CI for lower limit 15.4–23.9 90% CI for upper limit 68.4–104.3 No. of samples 28 Indirect bilirubin (mmol/L) Reference interval 34.2  2 (20.5) 90% CI for lower limit 10.3–15.4 90% CI for upper limit 59.9–99.2 No. of samples 28 Direct bilirubin (mmol/L) Reference interval 10.3  2 (3.4)a 90% CI for lower limit 1.7–5.1 90% CI for upper limit 13.7–17.1 No. of samples 28 Biliar acids (mmol/L) Reference interval 18.4  2 (11)b 90% CI for lower limit 0–2.9 90% CI for upper limit 33.5–46 No. of samples 26 Glucose (mmol/L) Reference interval 5.3  2 (1.1)b 90% CI for lower limit 2.6–3.7 90% CI for upper limit 6.9–8 No. of samples 32 Lactate (mmol/L) Reference interval 1.1  2 (0.8) 90% CI for lower limit 0–0.1 90% CI for upper limit 2.2–3 No. of samples 31 Total protein (g/L) Reference interval 66  2 (5)a,b 90% CI for lower limit 54–59 90% CI for upper limit 73–78 No. of samples 28 Albumin (g/L) Reference interval 35  2 (4)a 90% CI for lower limit 26–30 90% CI for upper limit 40–45

T  10

Tp

Tþ7

Control group

2.1  2 (0.7)a 0.5–1 3.2–3.7 55

3  2 (1.3) 0–1 5.1–6.2 50

2.9  2 (1) 0.4–1.5 4.3–5.5 27

3.4  2 (1.3) 0.1–1.9 5–6.7 17

3.9  2 (2.5) 0 7.9–9.8 55

4.4  2 (2.9) 0.7–1.2 10.7–19.1 49

4.9  2 (2.6) 0–1.3 8.5–11.4 26

3.4  2 (0.8) 1.2–2.3 4.4–5.6 17

4.2  2 (1.3)a 1.2–2.2 6.2–7.3 53

4.8  2 (1.4) 1.5–2.6 7–8.1 48

4.9  2 (2.6) 1.1–2.7 6.8–8.5 26

5.5  2 (2) 0.1–2.9 8–10.9 17

0.25  2 (0.13)b 0.08–0.11 0.41–0.56 53

18  2 (0.11)b 0.03–0.11 0.46–0.56 49

0.25  2 (0.06) 0.05–0.15 0.35–0.45 27

0.25  2 (0.13) 0–0.08 0.43–0.61 17

150.3  2 (26.5)a,b 88.4–106.1 185.6–212.2 55

159.1  2 (26.5)a,b 97.2–114.9 194.5–221 50

106.1  2 (17.7) 61.9–79.6 141.4–159.1 27

123.8  2 (8.8) 97.2–114.9 141.4–159.1 17

13.4  2 (2.7)a 6.8–8.9 17.9–20 55

12.6  2 (2.6)a 6.4–8.6 16.8–18.9 50

13  2 (3.4)a 4.5–8.3 6.4–21.4 27

10.2  2 (1.6) 6.1–8.2 12.1–14.3 17

46.2  2 (20.5)a 0–12 78.7–95.8 55

51.3  2 (29.1)a 15.4–22.2 94.1–135.1 50

32.5  2 (10.3) 6.8–17.1 46.2–58.2 27

25.7  2 (5.1) 12–18.8 30.8–37.6 17

35.9  2 (20.5) 0–3.4 68.4–83.8 55

41  2 (29.1)a 8.6–13.7 82.1–131.7 50

22.2  2 (8.6) 0–10.3 35.9–46.2 27

20.5  2 (3.4) 10.3–15.4 25.7–30.8 17

10.3  2 (3.4)a 3.4–6.8 15.4–17.1 55

10.3  2 (3.4)a 5.1–6.8 15.4–17.1 50

10.3  2 (1.7)a 3.4–6.8 12–15.4 27

5.1  2 (1.7) 1.7–3.4 6.8–8.6 17

17.6  2 (7.8) 0–5.1 29.9–36.7 46

13.5  2 (10.8)b 3.7–5.9 25.2–39.9 38

13.5  2 (4.7) 2–7.3 19.6–25 24

12.5  2 (2) 7.1–10 15.2–18.1 17

5.4  2 (1.3)b 2.4–3.4 7.5–8.5 56

7.7  2 (2.1)a,b 2.8–4.5 11–12.7 52

4.6  2 (0.8) 2.6–3.4 5.7–6.6 29

4.8  2 (0.6) 3.1–4 5.6–6.5 17

1  2 (0.5) 0–0.3 1.8–2.1 55

1.5  2 (0.9)a 0.5–0.7 3.2–4.8 49

1  2 (0.3) 0.1–0.5 1.4–1.8 26

0.7  2 (0.1) 0.4–0.6 0.9–1 17

64  2 (5)a,b 51–56 72–77 55

72  2 (7) 54–60 83–89 50

63  2 (7)a,b 46–53 72–80 27

70  2 (4) 58–64 75–81 17

35  2 (4)a 27–29 41–44

39  2 (5)a 29–32 47–50

34  2 (4)a 24–28 40–44

31  2 (5) 18–25 38–44 (continued on next page)

530

J. Mariella et al. / Theriogenology 81 (2014) 526–534

Table 2 (continued ) Parameter

T  20

T  10

Tp

Tþ7

Control group

No. of samples A/G ratio Reference interval 90% CI for lower limit 90% CI for upper limit No. of samples Triglycerides (mmol/L) Reference interval 90% CI for lower limit 90% CI for upper limit No. of samples Fibrinogen (mmol/L) Reference interval 90% CI for lower limit 90% CI for upper limit No. of samples Caþþ Reference interval 90% CI for lower limit 90% CI for upper limit No. of samples Naþ Reference interval 90% CI for lower limit 90% CI for upper limit No. of samples Kþ Reference interval 90% CI for lower limit 90% CI for upper limit No. of samples ClReference interval 90% CI for lower limit 90% CI for upper limit No. of samples Mgþþ Reference interval 90% CI for lower limit 90% CI for upper limit No. of samples

27

55

51

27

17

1.2  2 (0.2)b 0.6–0.9 1.4–1.7 28

1.2  2 (0.2) 0.7–0.9 1.6–1.7 55

1.2  2 (0.2) 0.7–0.9 1.6–1.9 50

1.2  2 (0.2)b 0.6–0.9 1.5–1.8 27

1  2 (0.1) 0.6–0.8 1.1–1.3 17

0.9  2 (1) 0–0.1 3–11.7 27

0.6  2 (0.7) 0.1–0.2 1.5–2.6 55

0.7  2 (0.9) 0–0.1 1.8–3.7 49

0.3  2 (0.5) 0–0.1 0.5–1.2 28

0.3  2 (0.1) 0–0.2 0.4–0.6 17

9.1  2 (3.1) 1.4–4.9 13.4–16.9 26

9.5  2 (2.2) 4.3–6.1 12.9–14.7 55

10.3  2 (2.1) 5.2–7.1 13.5–15.4 39

10.3  2 (3) 2.6–6.3 14.4–18.1 23

10.8  2 (1.7) 6.3–8.7 12.9–15.3 17

2.9  2 (0.2)b 2.5–2.7 3.2–3.45 28

2.9  2 (0.2) 2.5–2.6 3.2–3.4 55

2.7  2 (0.2)a 2.2–2.4 3.1–3.3 50

2.9  2 (0.2)b 2.4–2.6 3.2–3.5 27

3  2 (0.1) 2.7–2.9 3.2–3.3 17

137.8  2 (3.8)a 128.4–132.4 143–147 28

137.7  2 (3.3)a 130–132.6 142.8–145.4 51

139.5  2 (4.9)a 127.9–131.9 147–151 50

135.7  2 (5.9) 120.8–127.4 144.1–150.7 27

134.3  2 (1.5) 130.2–132.3 136.2–138.3 17

4.1  2 (0.8)a 2–3 5.2–6.2 22

3.9  2 (0.7)a 2.2–2.8 5.1–5.7 41

3.8  2 (0.5)a 2.5–3 4.6–5.1 36

3.7  2 (0.6)a 2.1–2.8 4.5–5.3 23

3.1  2 (0.5) 1.8–2.5 3.7–4.4 17

101.1  2 (3.3)a 92.7–96.3 105.8–109.5 28

101.2  2 (3)a 94.1–96.4 106–108.3 55

103  2 (4.2)a 93–96.4 109.5–112.9 50

97.8  2 (4) 87.7–92.1 103.5–108 27

98.2  2 (1.4) 94.5–96.4 99.9–101.9 17

1  2 (0.2) 0.7–0.8 1.2–1.3 28

1  2 (0.15) 0.7–0.8 1.2–1.35 55

1.05  2 (0.2)a 0.6–0.75 1.4–1.55 50

1.1  2 (0.15)a 0.7–0.85 1.3–1.45 27

0.95  2 (0.1) 0.7–0.85 1.05–1.2 17

Abbreviations: T  20, 20 days before parturition; T  10, 10 days before parturition; Tp, at parturition; T þ 7, 7 days after parturition. a Significantly different from the control group. b Significant difference between the different sampling times.

biliar acid, Bt, Bi, Bd, and Lac at Tp; GGT at T  10 and T  20; and triglyceride at T  20, T  10, Tp, and T þ 7, were backtransformed after logarithmic transformation. A parametric method was used to establish reference intervals for all the parameters studied. Parametric methods encompass slightly more than the central 95th percentile of the data and establish the lower and upper reference limits as mean minus 2 standard deviations and mean plus 2 standard deviations, respectively. Ninety percent confidence intervals around the reference limits were determined using a parametric method. Reference intervals for each parameters and the presence of a significant difference among different sampling times and the control group are indicated in Tables 1 and 2. 3.1. Hematologic data The hematologic parameters did not show any difference over time during the last 20 days of pregnancy and the first week after parturition. Significant variations against

the control group were observed in the values of hemoglobin concentration, Hct, and WBC count. Hemoglobin and Hct were lower (P < 0.01) compared with the control group at T þ 7, and WBC count was significantly higher (P < 0.01) compared with the control group at parturition (Table 1). There was no difference in any other hematologic parameter measured. 3.2. Biochemical parameters On comparing the biochemical results obtained during the last 20 days of pregnancy with those during the first week of lactation, several significant changes were found. At different sampling times, GGT, Cre, Glc, biliar acid, TP, A/G, and Ca were significantly different (Table 2). The GGT activity was higher (P < 0.01) at T  10 and at Tp than at the other sampling times. Creatinine was higher (P < 0.001) during the last 20 days of pregnancy, and then decreased at T þ 7. Glucose and TP were higher (P < 0.001) at Tp than at the other sampling times. Glucose measured at T þ 7 was

J. Mariella et al. / Theriogenology 81 (2014) 526–534

lower (P < 0.001) than at the other sampling times. Biliar acids and A/G were higher (P < 0.05) at T  20 than at T þ 7. Ionized calcium was significantly lower (P < 0.05) at Tp than at T  20 and T þ 7. Serum concentrations of CK, AST, Cre, BUN, Bt, Bd, Bi, Glc, Lac, TP, Alb, A/G, Ca, Mg, Na, Cl, and K were different compared with the control group. The CK activity was lower (P < 0.001) than control at T  20 and T  10 and the AST activity was lower (P < 0.05) than control at T  10. Creatinine was significantly higher (P < 0.001) than control during late gestation, but after 1 week of gestation no significant difference was observed compared with control. Significant changes were observed in BUN levels, which remained higher (P < 0.001) than those in the control at all sampling times. Serum Bt was higher (P < 0.001) than control in the last 20 days of gestation; at parturition, Bd was higher (P < 0.001) than control at all sampling time and Bi was higher (P < 0.001) at delivery. Total protein was lower (P < 0.001) than control except at delivery, although Alb was higher (P < 0.001) than control at all sampling times. Glucose and Lac were higher (P < 0.001) than control at delivery. As regards serum electrolytes concentration, Ca was lower (P < 0.001) than control at delivery; Na and Cl were higher (P < 0.001) than control at T  20, T  10, and at delivery and K was higher (P < 0.001) than control at all sampling times. Magnesium was higher (P < 0.01) than control at delivery and at T þ 7. Remarkable changes were not observed in serum ALP, triglyceride and fibrinogen concentrations. 4. Discussion The deep knowledge of hematologic and biochemical dynamics in periparturient mares is crucial to promptly and correctly diagnose mare’s diseases that could jeopardize both the mare’s and the fetus’ life, and to monitor mare’s response to treatments. In the present study, hematologic and biochemical profiles are thoroughly investigated beginning from the last month of pregnancy. In the clinical practice, pregnant mares often undergo the last veterinarian examination and obstetric ultrasonography during the last month of gestation, because it could be useful for the practitioners to know the specific reference range and the changes that occur during this period. As in a previous study [4], pregnant mares enrolled in this study have a wide age range (6–21 years) but in the authors’ opinion it does not influence the results. It is supported by a study about biochemical parameters in purebred Arabian mares that revealed no differences among three age groups: 6 to 12, 12 to 20, and older than 20 years [8]. In the present study, Hct and hemoglobin measured after 7 days from parturition are lower than those measured in the nonpregnant and nonlactating mares. No other studies measured the hematologic parameters in the early lactation, but a slightly lower Hct was reported to occur in mares, when foals were allowed to suckle longer than 4 months [9]. During pregnancy, most domestic animals and humans develop a mild anemia because of a relative hemodilution that is an exception in the horse [10–12]. Because in the horse endometrial bleeding is not possible (the placenta is

531

noncaducous), blood loss after parturition is usually the result of uterine, cervical, or vaginal damage, but none of these events happened in the mares of this study. In agreement with previous studies [3,5], erythrocyte indices (mean corpuscular volume, mean corpuscular hemoglobin, and mean corpuscular hemoglobin concentration) are within the normal reference intervals of adult horses. The lower Hct and hemoglobin at T þ 7 could suggest a relative anemia, probably resulting from a large increase in water ingestion with the beginning of lactation and subsequent overhydration and erythrocyte dilution. Lactating mares require from 50% to 70% additional water [13]. Increased intake is likely because of a combination of factors, principally the fluid losses associated with milk secretion and the increased consumption of feed to support milk production. Other factors that could influence water intake are diet composition and ambient temperature. So, a lactating mare may drink up to 75 L/d, whereas a 500-kg horse at rest consumes between 15 and 35 L [13]. In agreement with previous studies [3,5,14], WBC count is higher at parturition than at the other sampling times. Cortisol and catecholamine release during parturition [15] influenced WBC count and other parameters. In the present study, a differential leucocyte count was not performed, but it was reported that the increased number of WBCs at parturition is the result of increased number of blood neutrophils [3]. In the present study, WBC count at T þ 7 is within the normal reference intervals. Blanchard, et al. [14] reported that the mean WBC count gradually declined on Day 2 postpartum and returned to normal baseline values by Day 3 postpartum. In agreement with previous studies [3,5], platelet count is within the normal reference interval of adult horses. Aspartate aminotransferase activity in horses in late gestation (T  10) is significantly lower than in barren mares, as reported also by Flisinska-Bojanowska, et al. [16]. The reason is unclear. As other authors suggest, this could be because of a predominance of anabolic processes during pregnancy over catabolic ones [15]. In the present study, ALP does not change during late gestation and the beginning of lactation, as reported in a previous study [17]. In draft mares, GGT, ALP, and AST activities increased around the delivery and decreased gradually after foaling [5]. The authors suggested that it might have been a physiological load on the liver in the perinatal period, and this could be true for mares of heavy draft breeds. In the present study, GGT activity is higher at T  10 and at parturition than at the other sampling times. No other authors reported the presence of significant changes in serum GGT activity during pregnancy, except the one previously mentioned. Serum GGT is a valuable marker of disorders of the hepatobiliary system [18], and it is widely used for diagnosing hepatic disorders in animals. In horses, high levels of serum GGT activity are associated with a variety of hepatic disorders, such as toxic hepatic failure and subclinical hepatopathy, hyperlipemia, and cholestasis. Rather than a physiological load on the liver, it could be possible that the presence of cholestasis because of the large pregnant uterus causes the increase of direct and Bt during late pregnancy found in the present study. Engelking [19] reported that cholestasis

532

J. Mariella et al. / Theriogenology 81 (2014) 526–534

should be suspected when the conjugated bilirubin concentration is greater than 30% of the total value. In women, the liver received a lower percentage of cardiac output during pregnancy, but little effect is seen on common tests of liver function. Moreover, there are a number of liver disorders unique to pregnancy; the most common is cholestasis of pregnancy [1,20]. It is not known if that condition can affect also the pregnant mare. An increase in the placental ALP activity has been detected in the serum of pregnant queens [21], and probably in pregnant cows [22] but not in pregnant bitches [23]. In women, ALP activity increased dramatically after 24 weeks of gestation because of placental isoenzyme. Although placental ALP was found in the serum of pregnant women [1], it has not been yet found in the serum of pregnant mares [15]. The differences among species could be explained by differences in placentation and it will be interesting to evaluate the isoenzymes in further studies in mares. In the present study, CK activity is lower at T  20 and T  10 than the normal reference interval. As for AST activity, it is difficult to explain the decrease of this enzyme. Aoki and Ishii [5] found greater CK activity at parturition than during pregnancy and lactation, suggesting skeletal muscle damage during delivery, because some mares lay down for a long time. This could be true for heavy draft mares with large muscle masses, but in mares with eutocic delivery, which lasts no more than 20 minutes [24], there is no evidence of muscle damage. Also in women, CK activity was higher during labor, but it is related to the length of the labor, the kind of parturition, the parity, and the weight of the newborn [25]. To the authors’ knowledge, there are no studies about the CK isoenzymes at parturition in horses. In the present study, serum Cre and BUN are higher in late gestation and at delivery. At T þ 7, Cre concentration returns to within the normal reference range, whereas BUN remained elevated. Changes in serum Cre and BUN might reflect an increase in energy demand and a higher request of amino acids for anabolic process. A high concentration of Cre is also produced by the fetus, but it was excreted by the mare during pregnancy. After delivery, Cre returns to within the normal reference intervals may be because of the absence of fetal production, whereas BUN remains elevated, probably because of the high energy demand at the beginning of lactation. In heavy draft mares, there were no changes in serum BUN levels, whereas serum Cre levels decreased after parturition [5]. In accord with a previous study [4], Aoki and Ishii [5] suggested that these findings might be related to changes in energy metabolism rather than in renal function. In our opinion, a physiological overload of the kidney is not a plausible explanation. Studies in women demonstrated that renal plasma flow and glomerular filtration rate are 50% higher than those during the nonpregnant status [1]. Studies about it in mares are lacking. In accord with a previous study [5], Glc levels are higher at parturition. Aoki and Ishii [5] hypothesized that it was related to physical stress associated with foaling. Physical stress increases the cortisol level [26] and cortisol promotes gluconeogenesis. In contrast with previous studies [4,5] Glc concentration at T þ 7 is lower than that in late pregnancy.

Because Glc metabolism is well studied in horses [27–29], it was supposed that it is the effect of the development of insulin resistance. Glucose regulation is altered during pregnancy and lactation in many species, including humans [30], laboratory animals [31], sheep [32], cattle [33], and pigs [34]. Changes in Glc regulation during pregnancy include progressive development of insulin resistance, which allows for improved placental transfer of Glc to meet the increasing demands of the fetus. Insulin resistance during late gestation facilitates the Glc supply to the fetus at the expense of maternal tissues, through a shift in substrate utilization from carbohydrates to fatty acids, and decreases Glc utilization in peripheral tissues [30,33]. In horses, changes in carbohydrate metabolism and pancreatic b-cell function during pregnancy may contribute to insulin resistance [28]. Lactate is higher at parturition. It was measured at parturition only in one previous study [24], in which blood Lac concentration was evaluated every 12 hours in the first 72 hours postpartum. As reported also in a preliminary study from the same working group [35], Lac concentration was higher at parturition and decreased significantly after 12 hours, both after normal delivery and dystocia. In a study from Pirrone, et al. [24], mares were hospitalized; thus, the length of stage II labor was relatively short, even in those experiencing dystocia, and they received supportive therapy after parturition when necessary. It could have influenced the Lac concentration [24] in women. Nordström, et al. [36] suggested that the source of maternal Lac production during parturition could be the skeletal muscles. It is presumably true also in mares. Total protein levels are lower in late pregnancy and early lactation compared with parturition and the control group. Otherwise, Alb concentration was higher than the control group at all sampling times. No other study reported these findings. Because A/G remains constant and TP is the sum of globulins and Alb, globulins were supposed to be diminished. It could happen because globulins are concentrated in the mammary gland for the production of colostrum in the last period of pregnancy. As regards immunoglobulin G in milk, the major portion comes from the serum [37]. To the authors’ knowledge, there is no study about the correlation between mare’s globulins and the colostrum quality. The higher TP value at parturition could be because of a transient hemoconcentration, as suggested by other authors [5]. Hemoconcentration is the most common reason for hyperalbuminemia, but concurrent hyperproteinemia, and perhaps erythrocytosis, are expected. True overproduction of Alb has not been known to occur in other animals. It was reported that the increase in Alb concentration could be because of an increased production induced by glucocorticoid drugs or hormones, such as thyroid hormones and insulin, or possibly increased Alb life span [38]. Serum calcium concentration was investigated in a large number of studies. In the present study, serum calcium concentration is lower than the control group at parturition, and after 1 week it returns within the normal reference interval. It is probably related to Ca utilization for muscles and uterine contractions during active delivery. In vitro studies of myometrial tissue from mares emphasized the

J. Mariella et al. / Theriogenology 81 (2014) 526–534

importance of ionized Ca-related mechanisms in the control of uterine contractions [39]. Clinically relevant postpartum hypocalcemia is rare in horses [40]. An early study [41] reported no difference in total serum calcium concentration in pregnant versus nonpregnant mares, although it was studied in a small number of pregnant mares and the authors did not differentiate the ionized and complex fractions, which are more clinically relevant. Another study reported a transient decrease of serum calcium values associated with an increase in the parathyroid serum concentration on the postpartum Day 2 [42]. A more recent study measured total and ionized Ca in 26 mares in the third trimester of pregnancy and after 7 days of lactation and found it to be significantly lower in third trimester pregnant mares compared with adult horses [43]. In accord with the latter study, Ca serum concentration in lactating mares is not different from the control group. Some authors proposed that an accelerated fetal skeletal development and mineralization with increased Ca demand in the third trimester and placental transfer of Ca from mare to fetus may account for these findings [44,45]. Remarkable changes were not observed in Ca and Mg levels in other previous studies [4,5]. Serum Ca is also implicated in the detachment of the placenta in the early postpartum, and a study confirmed that mares with retained placenta had lower calcium concentration within 12 hours after foaling than mares without retained placenta [46]. It is also to be noticed that the present study measured the Ca concentration at parturition in a larger number of mares (n ¼ 50) than previously. On the contrary, Mg serum concentration was found to be significantly higher than in the control group at parturition and after 7 days. Magnesium is the fourth most common cation in the body and the second most common intracellular cation after K. It is involved in several processes, including muscle contraction, and in many of its actions it has been likened to a physiological calcium antagonist. In the authors’ opinion, the rise in Mg concentration at parturition could be because of movement of Mg from cells to bloodstream, probably mediated by the release of catecholamine and the stimulation of betaadrenergic receptors [47] during parturition. This finding has never been reported by other authors. The other electrolytesdNa, K, and Cldwere found to be constantly higher throughout the last period of pregnancy and at delivery. Although Na and Cl returned within the reference interval at T þ 7, K remained higher than the control group. This is probably because of an increased renal Na and Cl retention mediated by aldosterone (ALD) release during pregnancy. Recently, the components of the renin-angiotensin-ALD system were studied in Spanish broodmares [48], and it was found that pregnancy caused significant modifications in the renin and ALD concentrations: the former showed a large increase in the second and third periods of pregnancy and the latter showed a 12 times greater concentration than control horses in the second period of pregnancy. The authors suggested that ALD opposes the natriuretic effect of progesterone at distal tubule level, avoiding the loss of Na and allowing its gradual accumulation in the fetoplacental compartment and in maternal extracellular fluids. It was also reported that arginine vasopressin increases after the onset of second-

533

stage labor [49]. The higher arginine vasopressin release at parturition probably promoted the return of Na and Cl within the adult reference intervals. In the authors’ opinion, the high K concentration during pregnancy is probably because of the augmented reabsorption of the ion in the gastrointestinal tract, rather than a reduction in renal blood flow caused by the enlarged pregnant uterus. There are no studies about it in horses. In a previous study [5], higher concentrations of Na and Cl and a lower concentration of K were found at parturition and the authors presumed that it could happen because during parturition there is a certain grade of hemoconcentration. Remarkable changes were not observed by the same authors in Mg and Ca levels. In contrast with previous studies [4,5,50], tryglicerides did not show significant differences compared with the control group. The increase of triglyceride concentration was found in pregnant Shetland ponies when measured near term, and in heavy draft mares it was significantly lower than those measured before foaling [5,50]. The discrepancy with the results of those studies could be because of a greater fat content and fat mobilization in pony and draft horses than in light breed horses. Because in the present study the biochemical analysis was performed 1 week after parturition, it could be possible that the triglyceride concentration lowered after a longer period of lactation as showed in the study of Harvey, et al. [4]. Fibrinogen also did not show significant differences against the control group. Fibrinogen is an acute phase protein that increases in response to inflammation within 24 to 72 hours of an inflammatory stimulus. It could be useful in the peripartum period to detect conditions such as peritonitis or metritis, but in the authors’ opinion, its relatively slow acute phase response seriously hampers its clinical utility [51]. In a previous study [52], various hemostatic analytes were evaluated for 4 months prepartum and 5 months postpartum in 14 healthy mares. The plasma fibrinogen concentration and both factor VIII:C and von Willebrand factor activity showed gradual increases from midgestation and reached maximal, or near maximal activity at parturition. As regards breed-specific changes, in the authors’ opinion, Standardbred can be considered as a light breed horses and those results can be useful to the equine population as a whole, except for pony and heavy draft horses. 4.1. Conclusions The temporal changes in the hematologic and biochemical parameters observed in the present study in the peripartum and the differences with reference intervals made up for nonpregnant and nonlactating mares could be used to better evaluate the conditions of periparturient mares. Veterinarians treating late-term pregnant and parturient mares must be aware of the physiological changes that occur during peripartum to avoid misinterpretation of results of laboratory investigation leading to erroneous diagnoses and hence to incorrect treatment or unjustified abstinence from treatment. It is worth noting that the main changes were observed at parturition, when the most life-threatening conditions could affect the mares.

534

J. Mariella et al. / Theriogenology 81 (2014) 526–534

Acknowledgments [26]

The authors thank Dr. Emiliana Antenucci and Dr. Lara Brunori who provided samples for the analysis and Dr. Laura Ingrà and Dr. Elisa Brini for their assistance in biochemical analysis.

[28]

References

[29]

[1] Lockitch G. Clinical biochemistry of pregnancy. Crit Rev Clin Lab Sci 1997;34:67–139. [2] Ashwood ER, Knight GJ. Clinical chemistry of pregnancy. In: Burtis CA, Ashwood ER, Bruns DE, editors. Teitz textbook of clinical chemistry and molecular diagnostics. Fourth edition. St. Louis, MO: Elsevier Saunders; 2006. p. 2153–206. [3] Harvey JW, Asquith RL, Pate MG, Kivipelto J, Chen CL, Ott EA. Haematological findings in pregnant, postparturient and nursing mares. Comp Hematol Int 1994;4:25–9. [4] Harvey JW, Pate MG, Kivipelto J, Asquith RL. Clinical biochemistry of pregnant and nursing mares. Vet Clin Pathol 2005;34:248–54. [5] Aoki T, Ishii M. Hematological and biochemical profiles in peripartum mares and neonatal foals (Heavy draft horses). J Equine Vet Sci 2012;32:170–6. [6] Satué K, Montesinos P. Plasma biochemistry profile in pregnant Spanish purebred broodmares. Comp Clin Pathol 2013;22:113–7. [7] Friedrichs K, Barnhart K, Blanco J, et al. ASVCP Quality Assurance and Laboratory Standards Committee (QALS) guidelines for the determination of reference intervals in veterinary species and other related topics. American Society for Veterinary Clinical Pathology. 2005. http://www.asvcp.org/pubs/pdf/RI%20Guidelines%20For%20ASVCP% 20website.pdf. [8] Gurgoze SY, Icen H. The influence of age on clinical biochemical parameters in Pure-bred Arabian mares. J Equine Vet Sci 2010;30: 569–74. [9] Hansen MF, Todd AC, McGee WR. Blood picture of lactating and non-lactating Throughbred mares. Vet Med 1950;45:228–30. [10] Allard RL, Carlos AD, Faltin EC. Canine hematologic changes during gestation and lactation. Compan Anim Pract 1989;19:3–6. [11] Berman E. Hemogram of the cat during pregnancy and lactation and after lactation. Am J Vet Res 1974;35:457–60. [12] Wells MY, Decobecq CP, Decouvelaere DM, Justice C, Guittin P. Changes in clinical pathology parameters during gestation in the New Zealand white rabbit. Toxicol Pathol 1999;27:370–9. [13] Cymbaluck NF. Water. In: Geor RJ, Harris PA, Coenen M, editors. Equine applied and clinical nutrition: health, welfare, and performance. New York: Elsevier Health Sciences; 2013. p. 80–95. [14] Blanchard TL, Orsini JA, Garcia MC, Elmore RG, Youngquist RS, Bierschwal CJ. Influence of dystocia on white blood cell and blood neutrophil counts in mares. Theriogenology 1986;25:347–52. [15] Davies Morel MC. Equine reproductive physiology, breeding and stud management. Ipswich, UK: Farming Press Books; 1993. [16] Flisinska-Bojanowska A, Komosa M, Gill J. Influence of pregnancy on diurnal and seasonal changes in glucose level and activity of FDPA, AlAT and AspAT in mares. Comp Biochem Physiol A 1991; 98:31–5. [17] Meuten DJ, Kociba G, Therelfall WR, Nagode LA. Serum alkaline phosphatase in pregnant mares. Vet Clin Pathol 1980;9: 27–30. [18] Braun JP, Benard P, Burgat V, Rico AG. Gamma glutamyl transferase in domestic animals. Vet Res Commun 1983;6:77–90. [19] Engelking LR. Evaluation of equine bilirubin and bile acid metabolism. Compend Cont Educ Pract Vet 1989;11:328. [20] Fagan EA. Intrahepatic cholestasis of pregnancy. Br Med J 1994;309: 1243–4. [21] Everett RM, Duncan JR, Prasse KW. Alkaline phosphatases in tissues and sera of cats. Am J Vet Res 1977;38:1533–8. [22] Yokus B, Cakir UD. Seasonal and physiological variations in serum chemistry and mineral concentrations in cattle. Biol Trace Elem Res 2006;109:255–66. [23] Allard RL, Carlos AD, Faltin EC. Canine hematologic changes during gestation and lactation. Compan Anim Pract 1989;19:3–6. [24] Pirrone A, Mariella J, Gentilini F, Castagnetti C. Amniotic fluid and blood lactate concentrations in mares and foals in the early postpartum period. Theriogenology 2012;78:1182–9. [25] Chemnitz G, Nevermann L, Schmidt E, Schmidt FW, Lobers J. Creatine kinase (EC-No.2.7.3.2) and creatine kinase isoenzymes during

[27]

[30] [31] [32]

[33]

[34]

[35]

[36]

[37] [38] [39]

[40] [41]

[42]

[43]

[44] [45]

[46]

[47] [48]

[49]

[50]

[51] [52]

pregnancy and labor and in the cord blood. Clin Biochem 1979;12: 277–81. Wong CW, Smith SE, Thong YH, Opdebeeck JP, Thornton JR. Effects of exercise stress on various immune functions in horses. Am J Vet Res 1992;53:1414–7. Evans JW. Effect of fasting, gestation, lactation and exercise on glucose turnover in horses. J Anim Sci 1971;33:1001–4. Fowden AL, Comline RS, Silver M. Insulin secretion and carbohydrate metabolism during pregnancy in the mares. Equine Vet J 1984;16:239–46. Hoffman RM, Kronfeld DS, Cooper WL, Harris PA. Glucose clearance in pregnant mares is affected by diet, pregnancy and lactation. J Anim Sci 2003;81:1764–71. Boden G. Fuel metabolism in pregnancy and in gestational diabetes mellitus. Obstet Gynecol Clin North Am 1996;23:1–10. Leturque A, Huguel P, Ferré P, Girard J. Glucose metabolism in pregnancy. Biol Neonate 1987;51:64–9. Petterson JA, Dunshea FR, Ehrhardt RA, Bell AW. Pregnancy and undernutrition alter glucose metabolic responses to insulin in sheep. J Nutr 1993;123:1286–95. Bell AW. Regulation of organic nutrient metabolism during transition from late pregnancy to early lactation. J Anim Sci 1995;73: 2804–19. Pére MC, Etienne M, Dourmad JY. Adaptations of glucose metabolism in multiparous sows: effects of pregnancy and feeding level. J Anim Sci 2000;78:2933–41. Castagnetti C, Mariella J, Pirrone A. Monitoring of lactate during the peripartum period in the equine species. Proceedings of the VII SIRA Congress. Ozzano dell’Emilia (BO), July, 2009. p. 68–71. Nordström L, Achanna S, Naka K, Arulkumaran S. Fetal and maternal lactate increase during active second stage labor. Br J Obstet Gynaecol 2001;108:263–8. Hurley WL, Theil PK. Perspectives on immunoglobulins in colostrum and milk. Nutrients 2011;3:442–74. Rothschild MA, Oratz M, Schreiber SS. Serum albumin. Hepatology 1988;8:385–401. Coruzzi G, Poli E, Bertaccini G. Effect of calcium-channel blockers and salbutamol on the isolated mares uterus interactions with the calcium agonist Bay K 8644. J Vet Pharmacol Ther 1989;12:404–10. Richardson JD, Harrison LJ, Edwards GB. Two horses with hypocalcemia. Vet Rec 1991;129:98. Holley DC, Evans JW. Determination of total and ultrafilterable calcium, and magnesium in normal equine serum. Am J Vet Res 1977;38:259–62. Martin KL, Hoffman RM, Kronfeld DS, Ley WB, Warnick LD. Calcium decreases and parathyroid hormone increases in serum of periparturient mares. J Anim Sci 1996;74:834–9. Berlin D, Aroch I. Concentrations of ionized and total magnesium and calcium in healthy horses: effects of age, pregnancy, lactation, pH and sample type. Vet J 2009;181:305–11. Care AD. Fetal calcium homeostasis. Equine Vet J Suppl 1997;24:59– 61. Wooding FBP, Morgan G, Fowden AL, Allen WR. Separate sites and mechanisms of placental transport of calcium, iron and glucose in the equine placenta. Placenta 2000;21:635–45. Sevinga M, Barkema HW, Hesselink JW. Serum calcium and magnesium concentrations and the use of a calcium-magnesium borogluconate solution in the treatment of Friesian mares with retained placenta. Theriogenology 2002;57:941–7. Romani AMP, Scarpa A. Regulation of cellular magnesium. Front Biosci 2000;5:720–34. Satué K, Domingo R, Redondo JI. Relationship between progesterone, oestrone sulphate and cortisol and the components of renin angiotensin aldosterone system in Spanish purebred broodmares during pregnancy. Theriogenology 2011;76:1404–5. Vivrette SL, Kindahl H, Munro CJ, Roser JF, Stabenfeldt GH. Oxytocin release and its relationship to dihydro-15-keto PGF2a and arginine vasopressin release during parturition and to suckling in postpartum mares. J Reprod Fertil 2000;119:347–57. Watson TD, Burns L, Packard CJ, Shepherd J. Effects of pregnancy and lactation on plasma lipid and lipoprotein concentrations, lipoprotein composition and post-heparin lipase activities in Shetland pony mares. J Reprod Fertil 1993;97:563–8. Crisman MV, Scarratt WK, Zimmerman KL. Blood protein and inflammation in the horse. Vet Clin Equine 2008;24:285–97. Gentry PA, Feldman BF, O’Neill SL, Madigan JE, Zinkl JG. Evaluation of the haemostatic profile in the pre and post parturient mare, with particular focus on the perinatal period. Equine Vet J 1992;24: 33–6.