Search
Author
Title
Vol.
Issue
Year
1st Page

Journal of Animal Science - Animal Growth, Physiology, and Reproduction

Postpartum deaths: Piglet, placental, and umbilical characteristics1

 

This article in JAS

  1. Vol. 91 No. 6, p. 2647-2656
    unlockOPEN ACCESS
     
    Received: June 05, 2012
    Accepted: Feb 27, 2013
    Published: November 25, 2014


    2 Corresponding author(s): vibeke.rootwelt@nvh.no
 View
 Download
 Share

doi:10.2527/jas.2012-5531
  1. V. Rootwelt 2,
  2. O. Reksen*,
  3. W. Farstad* and
  4. T. Framstad*
  1. Department of Production Animal Clinical Sciences, Norwegian School of Veterinary Science, N-0033 Oslo, Norway

Abstract

The fetal growth of the piglet is highly dependent on its placenta, and the newborn piglet birth weight is highly associated with postpartum death. However, there is little information available in the literature on the assessment of the placenta in relation to postpartum death in piglets. The aim of this study was to evaluate the impact of the placental area and placental weight, status of the umbilical cord, and piglet birth characteristics, such as blood parameters, vitality score, and birth weight on postpartum death. All live born piglets in litters from 26 Landrace-Yorkshire sows were monitored during farrowing and the status of each was recorded, including placental area and placental weight and blood variables obtained from the piglets and umbilical veins. Out of the 386 live-born piglets, 16.8% died before weaning at 5 wk. Among these, 78.5% died within the first 3 d of life. Mean blood concentration of lactate was increased in piglets that did not survive to weaning (P = 0.003). Concentrations of hemoglobin and hematocrit were decreased (P < 0.001) compared with survivors. Piglets born with a broken umbilical cord had a reduced vitality score vs. piglets born with an intact umbilical cord (P = 0.021), and they had an increased probability of dying before weaning (P = 0.050). Mean birth weight, body mass index, placental area (P < 0.001), and placental weight (P = 0.020) were reduced in piglets that died before weaning vs. those that survived. Birth weight and placental area were furthermore negatively associated with live litter size. Blood concentrations of IgG and albumin recorded at d 1 were decreased in piglets that died before weaning (P < 0.01), and blood concentration of albumin was positively associated with placental area (P < 0.001). We conclude that placental area and placental weight, status of the umbilical cord, birth weight, body mass index, blood concentrations of lactate, hemoglobin, and hematocrit recorded at birth, and blood concentrations of IgG and albumin recorded at d 1 were associated with postpartum death in this study. These results may indicate that there is an upper uterine limitation of litter size and that placental area and placental weight influence postpartum survival.



INTRODUCTION

Postpartum death occurs in 13 to 15% of piglets (NAHMS, 2006; Agrovision, 2008; Ingris Animalia Norsvin, 2011) and is commonly defined as piglets born live, but which die before weaning at 5 wk. Although managerial factors may be most important for minimizing neonatal losses (Munsterhjelm et al., 2006; Andersen et al., 2009; Oliviero et al., 2010), birth weight is also highly associated with postpartum survival in pigs (Devillers et al., 2004; Baxter et al., 2008; Boulot et al., 2008). The placenta of the sow has a diffuse and epitheliochorial distribution (Senger, 2003; McGeady et al., 2006). The fetal growth of the piglet is highly dependent on its placenta, and fetal weight is found to be proportional to placental weight in several studies (Leenhouwers et al., 2002; Van Rens et al., 2005; Rampersad et al., 2011). As a consequence, placental weight is expected to be associated with postpartum death, but reports are somewhat contradictory (Leenhouwers et al., 2002; Van Rens et al., 2005; Baxter et al., 2008). Furthermore, increased litter size is associated with reduced birth weight (Boulot et al., 2008; Beaulieu et al., 2010; Rootwelt et al., 2012a). Due to this, placental area seems to be a better predictor of postpartum piglet death than placental weight. The latter has more possibilities for variations that are independent of uterine size (i.e., decreased or increased thickness of placental stroma), or variations of vasculogenesis or angiogenesis (Huppertz, 2011). Placental area is, on the other hand, limited by uterine size to a much larger extent in some commercial pig breeds (Knight et al., 1977; Vallet et al., 2011). Baxter et al. (2008), surprisingly, found no association between placental area and postpartum death. There is generally a relative paucity of information in the literature on the significance of the placenta in relation to postpartum death in piglets.

The aim of this study was to evaluate the impact of the fetal placental area and placental weight, status of the umbilical cord, and piglet birth characteristics, such as blood variables, vitality score, and birth weight on postpartum death.


MATERIALS AND METHODS

The experimental protocol for this study did not require approval by the Norwegian Animal Research Authority due to an exception for such procedures in the Norwegian regulations for animal testing (FOR 1996-01-15 no. 23, Regulation of animal testing, §2: Scope).

From this same sow material, results of the stillborn piglets are already published in Rootwelt et al. (2012b)

Animals

Landrace-Yorkshire sows (n = 26) were selected from a sow pool system as described by Dalin et al. (1997). All sows originated from the same multiplier herd and were housed in 1 farm in the southeastern part of Norway. Parity of the sows ranged from 1 to 8, and sows were grouped into 3 categories according to parity: first and second parity sows were grouped in parity group 1 (n = 7), third and fourth parity sows were grouped in parity group 2 (n = 11), and fifth to eighth parity sows were grouped in parity group 3 (n = 8).

The sows were inseminated with heterospermic semen from Landrace-Duroc boars by the same technician at standing estrus 1 d apart, with an insemination dose of 2.5 × 109 spermatozoa. Gilts and sows were loose housed and kept separated in 2 groups. At 3 wk before farrowing, the animals were transported to the test farm. All live-born piglets were included in the study, whereas stillborn piglets, defined by a veterinarian as piglets born with no respiration or palpable heartbeats, were excluded. The study period was from March 2010 to December 2011; every eighth week, 2 or 3 different gilts or sows were monitored during farrowing. One farrowing was supervised each day, and any random gilt or sow that started farrowing in the morning when the veterinarian was present was included in the study. The gilts or sows were included in the study each with 1 litter only.

Management

At the test farm, gilts and sows were kept individually loose housed and without fixation in standard farrowing pens without crates (7.3 m2) with a piglet creep area (0.8 m2), from 3 wk before expected farrowing until weaning at 5 wk. Each pen had a solid floor except for a slatted drainage floor at 1 end of the pen (2.8 m2). A commercial pelleted lactation dry feed (9.86 MJ NE kg−1, 8.26 g lysine kg−1) was offered twice daily to the sows before parturition and up to 1 wk after farrowing, 3 times a d in the second week, and then 4 times daily throughout the lactation period. All sows were given 0.5 kg of hay daily until farrowing and had ad libitum access to water. Farrowing was allowed to start naturally.

Recorded Variables

Gestation length in days from the last day of insemination until farrowing and number of live-born piglets was recorded. Within seconds after expulsion of each piglet, the newborn was given a birth vitality score consisting of a modified human appearance-pulse-grimace-activity-respiration (APGAR) score. This score was named by the first letter of each of the 3 variables chosen to represent vitality, which were quick and easy to record. These variables (respiration, meconium staining, and activity; RMA) were determined by evaluating the newborn piglet on a scale from 0 to 2, then summing up the 3 values thus obtained. The resulting RMA score ranged from 0 to 6, where 0 was no respiration or activity and gross meconium staining and 6 was normal respiration and activity and no meconium staining. Whether or not the umbilical cord was ruptured was recorded, and blood samples were taken from the intact umbilical vein. The cord was subsequently double ligated with a color code, and cut between the ligations. The piglets were immediately held in dorsal recumbency, and 0.5 mL of blood was evacuated from vena jugularis externa/interna/communis using 2-mL plain syringes with 23-gauge needles. Whole blood from the umbilical veins and the piglets was immediately analyzed for concentrations of glucose, lactate, oxygen pressure (pO2), carbon dioxide pressure (pCO2), pH, base excess, hemoglobin, hematocrit, sodium, potassium, and ionized calcium using a hand-held Epoc portable clinical analyzer (Epocal Inc., Ottawa, ON, Canada). Time until birth of each piglet, from expulsion of the first piglet in the litter, was recorded as well as birth intervals. The piglets were weighed on a scale with 10 g accuracy according to the manufacturer (Premium; EKS International SAS, Wittisheim, France). Birth weights were also categorized in 200 g intervals, yielding 11 groups in total. Body length was measured from the occipital bone to the root of the tail. Body weight and length were used for calculation of body mass index (BMI): [BW (kg)/length (m)2]. Gender was recorded, and the piglets were grouped in 3 categories according to birth order: first, middle, or last third of each separate litter.

The expelled fetal placentas (i.e., the chorioallantoic sacs), were kept at room temperature until examination on the same or the next day, when they were rinsed in water, each placenta separated from 1 another, and they were left to remove excess water for 1 h. Attached amniotic membrane, necrotic parts of the tip of each avascularized chorion, and the umbilical cord at the junction where it joins the placenta and splits into its tributaries were all removed. Each chorioallantoic sac was then weighed wet on a digital scale with 1 g accuracy according to the manufacturer (Z17489; Silvercrest, Milomex Ltd., Bedfordshire, UK), spread on solid paper, and the circumference was cut out with a sharp pair of scissors. The paper was numbered, dried, and the area in cm2 was later recorded by a planimeter (Lasico 42P, B-90899; Los Angeles Scientific Instrument Co., Inc., Los Angeles, CA) and multiplied by 2 to approximate the surface area. As for piglet birth weights, placental areas and weights were also each categorized into 11 groups. Placental area and placental weight were recorded blindly with regard to piglet identity, but with known sow identity.

At d 1 (24 ± 2 h later, measured from midtime between expulsion of the first and last piglet of each litter) each live piglet had 0.5 mL whole blood collected in tubes containing EDTA as anticoagulant. The blood sample was subsequently stored at room temperature until analyzed the next working day at the Central Laboratory, Norwegian School of Veterinary Science. The plasma was separated by centrifugation at 2,000 × g for 10 min at 20°C and analyzed for total plasma protein, which was assayed with an ADVIA 1650 system (Siemens Medical Solutions Diagnostics Inc., Tarrytown, NY). Albumin, α1 globulin, α2 globulin, β1 globulin, β2 globulin, and γ globulin were separated by electrophoresis using Capillarys 2 (Sebia, Lisses, France). Plasma was then frozen at −40°C until analysis of IgG with an in-house Single Radial Immunodiffusion test kit (VMRD Inc., Pullman, WA). Values below the lower detection limit 3.8 g/L were defined as 0. At d 1, d 2, d 3, and at weaning at 5 wk, the piglets were recorded as either dead or alive.

Statistical Analyses

Means of live-born litter size at birth and weaning were calculated using the statistical software JMP 8 (SAS Inst. Inc., Cary, NC). One way ANOVA was used to study the association of all blood parameters recorded at birth and d 1 between dead vs. live piglets at weaning. The same statistical procedure was used to study the association of time until birth from expulsion of the first piglet in the litter, birth interval, birth weight, BMI, placental area, and placental weight between dead vs. live piglets at weaning. Similarly, ANOVA was used for the association between RMA immediately after birth and the status of the umbilical cord, and between hemoglobin in the piglet and the status of the umbilical cord. Chi square analysis was used to assess the association between the status of the umbilical cord at birth and whether the piglet died before weaning or not. Univariate linear regression analyses were used for assessing the association between piglet birth weight and live-born litter size, for RMA and birth weight, and for blood parameters recorded in the piglet at d 1 and placental area and placental weight.

For the outcome variables birth weight, BMI, placental area, and placental weight separate multivariable generalized linear models (GLM) were conducted using the xtreg option in Stata SE11 (StataCorp LP, College Station, TX). Explanatory variables which clinically might influence these piglet outcomes (i.e., parity group, birth order group, gender, live-born litter size, and gestation length) were all simultaneously included in the models. Sow was included as a random effects variable to account for clustering at the sow level. A backward elimination procedure was employed, and explanatory variables with an association to the outcome variable yielding a P > 0.10 were omitted from the final models. Dead vs. alive at weaning was forced into all models for the comparison. Recorded blood parameters at d 1 with association to placental area and placental weight, that is, albumin, was run in a similar GLM procedure. For the outcome dead vs. alive at weaning, the status of the umbilical cord was included as an explanatory variable in a logistic regression analysis in Stata SE11, with sow as a random effects variable to account for clustering.

Overall statistical significance of the models was assessed by the type III F-test in Stata. Homoscedasticity and normality of the residuals were assessed using plots of standardized residuals.


RESULTS

In total, 386 live piglets were born and mean live-born litter size was 14.9 ± 0.60 piglets. Mean litter size at weaning was 12.3 ± 0.54. Loss from birth to weaning was 65 piglets or 16.8%. Of these 65 dead piglets, 78.5% died within the first 3 d of life: 34 piglets by d 1, 43 piglets by d 2, and a total of 51 piglets by d 3.

Of all piglets born, RMA was recorded in 330. Piglets born with a broken umbilical cord (n = 66) had a decreased RMA score of 4.4 ± 0.16 vs. 4.8 ± 0.08 in piglets born with an intact umbilical cord (n = 249; P = 0.021). Piglets born with a broken umbilical cord had an increased probability of dying before weaning, also when sow was included in the model to account for clustering (P = 0.050). The RMA was negatively associated with time to birth from the expulsion of the first piglet (P < 0.001). There was no association between RMA and piglet birth weight (P = 0.503).

At birth, recorded mean blood concentrations of lactate were increased in piglets that did not survive to weaning vs. those that did, whereas mean concentrations of hemoglobin and hematocrit were decreased (Table 1). Mean blood concentration of hemoglobin was decreased in piglets born with a broken umbilical cord vs. intact; 88.8 g/L and 97.0 g/L, respectively (P = 0.016). Mean time to birth from the expulsion of the first piglet in the litter and birth interval were not different between piglets that died before weaning vs. piglets that survived. Birth weight ranged between 0.41 and 2.53 kg and was negatively associated with live-born litter size (n = 358; R2 adj = 0.05; P < 0.001). The number of postpartum deaths for each birth weight category is presented in Fig. 1. The numbers of postpartum deaths for each placenta area category and placenta weight category are presented in Fig. 2 and 3, respectively. Birth weight, BMI, placental area, and placental weight were associated with survival to weaning (Table 2). After adjusting for significant explanatory variables, birth weight, BMI, placental area, and placental weight remained decreased in piglets that died before weaning. Placental area was negatively associated with live-born litter size (Table 3).


View Full Table | Close Full ViewTable 1.

Comparison of blood parameters at birth from piglets and umbilical veins of piglets that died before weaning vs. piglets that survived1

 
Piglet
Umbilical vein
Item Dead at weaning Alive at weaning Dead at weaning Alive at weaning
Glucose, mmol/L 2.7 ± 0.19 (45) 3.0 ± 0.09 (212) 2.7 ± 0.27 (18) 2.9 ± 0.12 (96)
Lactate, mmol/L 7.08 ± 0.46a (40) 5.58 ± 0.21c (189) 5.33 ± 0.63 (17) 4.60 ± 0.27 (93)
pO2, mmHg 28.0 ± 2.06 (46) 25.0 ± 0.96 (211) 36.1 ± 2.98 (19) 30.33 ± 1.24 (110)
pCO2, mmHg 60.1 ± 1.41 (46) 58.7 ± 0.66 (211) 47.3 ± 2.41 (19) 48.8 ± 1.00 (110)
pH 7.30 ± 0.01 (47) 7.33 ± 0.01 (213) 7.40 ± 0.02 (19) 7.41 ± 0.01 (107)
Base Excess, mmol/L 3 ± 0.9 (47) 5 ± 0.4 (212) 4 ± 1.2 (19) 6 ± 0.5 (107)
Hemoglobin, g/L 86 ± 3.0a (47) 98 ± 1.4d (215) 81 ± 4.6 (18) 90 ± 1.9 (107)
Hematocrit, L/L 0.25 ± 0.01a (47) 0.29 ± 0.00d (215) 0.24 ± 0.01 (18) 0.27 ± 0.01 (107)
Sodium, mmol/L 134.6 ± 0.52a (47) 135.9 ± 0.24b (217) 131.2 ± 1.76 (18) 133.5 ± 0.71 (109)
Potassium, mmol/L 4.7 ± 0.14 (47) 4.9 ± 0.06 (220) 4.4 ± 0.27 (19) 4.3 ± 0.11 (109)
Ionized calsium, mmol/L 1.56 ± 0.02 (46) 1.52 ± 0.01 (220) 1.69 ± 0.05 (19) 1.58 ± 0.02 (109)
a,bBetween columns for umbilical vein or piglet, means differ (P < 0.05).
a,cBetween columns for umbilical vein or piglet, means differ (P < 0.01).
a,dBetween columns for umbilical vein or piglet, means differ (P < 0.001).
1Values are mean ± SE (number).
Figure 1.
Figure 1.

Dead vs. live piglets at weaning, when categorized by birth weight.

 
Figure 2.
Figure 2.

Dead vs. live piglets at weaning, when categorized by placental area.

 
Figure 3.
Figure 3.

Dead vs. live piglets at weaning, when categorized by placental weight.

 

View Full Table | Close Full ViewTable 2.

Comparison of time until birth from expulsion of the first piglet, birth interval, birth weight, body mass index, placental area, and placental weight between piglets that died before weaning vs. piglets that survived1

 
Item Dead at weaning Alive at weaning P-value
Time until birth from expulsion of the first piglet, min 117 ± 13.6 (65) 138 ± 6.2 (317) 0.162
Birth interval, min 15 ± 3.0 (65) 17 ± 1.3 (317) 0.607
Birth weight, kg 1.16 ± 0.04 (63) 1.46 ± 0.02 (295) < 0.001
Body mass index 18.5 ± 0.35 (63) 19.7 ± 0.17 (285) 0.002
Placental area, cm2 1633 ± 98.9 (28) 2052 ± 44.5 (138) < 0.001
Placental weight, g 178 ± 12.6 (28) 209 ± 5.6 (140) 0.026
1Values are mean ± SE (number).

View Full Table | Close Full ViewTable 3.

Associations between birth weight, body mass index, placental area, and placental weight and piglets that died before weaning vs. piglets that survived

 
Item1 n Parameter estimate2 SE P-value
Birth weight, kg
    Dead at weaning 63
    Alive at weaning 295 0.275 0.041 < 0.001
    Parity group 1
    Parity group 2 0.333 0.078 < 0.001
    Parity group 3 0.220 0.090 0.014
    Birth order group 1
    Birth order group 2 −0.084 0.036 0.019
    Birth order group 3 −0.007 0.037 0.843
    Male piglet
    Female piglet −0.103 0.031 0.001
    Live litter size −0.030 0.011 0.005
    Gestation length, days −0.058 0.034 0.087
    Constant 8.273 3.934 0.035
Body mass index
    Dead at weaning 63
    Alive at weaning 285 1.411 0.366 < 0.001
    Birth order group 1
    Birth order group 2 −0.616 0.321 0.055
    Birth order group 3 −0.342 0.334 0.320
    Male piglet
    Female piglet −0.714 0.279 0.011
    Constant 18.979 0.456 <0.001
Placental area, cm x cm
    Dead at weaning 28
    Alive at weaning 138 357.909 100.464 <0.001
    Live litter size −50.877 22.912 0.026
    Constant 2471.833 367.846 <0.001
Placental weight, g
    Dead at weaning 28
    Alive at weaning 140 28.741 12.033 0.017
    Birth order group 1
    Birth order group 2 23.376 9.560 0.014
    Birth order group 3 32.281 10.852 0.003
    Live litter size −5.572 3.111 0.073
    Constant 247.709 49.671 <0.001
1Piglets are grouped into 3 categories according to birth order groups: first, middle, or last third of each litter.
2Parameter estimate = slope of multivariable regression for explanatory variables.

At d 1, mean concentrations of IgG and albumin were decreased in piglets that died before weaning vs. those that survived (Table 4). Blood concentration of IgG was not associated with placental area or placental weight, but albumin was associated with both placental variables (Table 5). After adjusting for significant explanatory variables, albumin remained positively associated with placental area (Table 6).


View Full Table | Close Full ViewTable 4.

Comparison of blood variables at d 1 from piglets that died before weaning vs. piglets that survived1

 
Piglet
Item Dead at weaning Alive at weaning
Total protein, g/L 48 ± 1.9a (26) 52 ± 0.6b (256)
IgG, g/L 16.5 ± 1.51a (26) 20.6 ± 0.48c (256)
Albumin, g/L 5.3 ± 0.29a (26) 6.2 ± 0.09c (256)
α1 globulin, g/L 9.4 ± 0.20a (26) 8.9 ± 0.06b (256)
α2 globulin, g/L 2.0 ± 0.08a (26) 2.2 ± 0.03b (256)
β1 globulin, g/L 5.6 ± 0.26 (26) 6.0 ± 0.08 (248)
β2 globulin, g/L 8.1 ± 0.48a (26) 9.2 ± 0.15b (256)
γ globulin, g/L 17.3 ± 1.36 (26) 19.5 ± 0.43 (255)
a,bBetween columns, means differ (P < 0.05).
a,cBetween columns, means differ (P < 0.01).
1Values are mean ± SE (number).

View Full Table | Close Full ViewTable 5.

Univariate associations between placental area and placental weight, and recorded blood variables in the piglet at d 11

 
Item Placental area Placental weight
Total protein, g/L 134, 0.967 136, 0.330
IgG, g/L 134, 0.610 136, 0.427
Albumin, g/L 134, 0.032 136, 0.038
α1 globulin, g/L 134, 0.132 136, 0.037
α2 globulin, g/L 134, 0.799 136, 0.530
β1 globulin, g/L 131, 0.488 133. 0.019
β2 globulin, g/L 134, 0.334 136, 0.047
γ globulin, g/L 133, 0.776 135, 0.260
1Values are number of animals and P-value.

View Full Table | Close Full ViewTable 6.

Associations between piglet blood concentration of albumin at d 1 and placental area

 
Item1 n Parameter estimate2 SE P-value
Albumin, g/L
    Placental area, cm2 134 0.001 0.000 < 0.001
    Birth order group 1
    Birth order group 2 −0.571 0.207 0.006
    Birth order group 3 −0.395 0.233 0.090
    Constant 4.889 0.438 < 0.001
1Piglets were grouped into 3 categories according to birth order: first, middle, or last third of each litter.
2Parameter estimate = slope of multivariable regression for explanatory variables.

DISCUSSION

Live born litter size and neonatal loss were increased in this study compared with national reports (NAHMS, 2006; Agrovision 2008; Ingris Animalia Norsvin 2011). By a constant supervision of the farrowing, as in this study, more piglets could be registered as being alive at birth than generally reported from swine production. Consequently, a severely exhausted piglet, which in this study was recorded as being alive, had a greater risk of dying later during the neonatal period. The increased litter size could also contribute to increased neonatal mortality, as these in several studies are reported to be associated (Marchant et al., 2000; Lay et al., 2002; Andersen et al., 2011). The increased mortality in this study could also be due to increased risk of crushing by the loose housed sow, although studies of the association between loose housing and increased risk of being crushed are contradictory (Weber et al., 2007; Pedersen et al., 2011).

The human APGAR was devised in 1952 with the purpose of quickly determining whether a newborn needs immediate medical care (Apgar, 1953). Because some of these variables are difficult to determine in the pig, other vitality tests have been described (Randall, 1971; Herpin et al., 1996; Mota-Rojas et al., 2005). In our study, a RMA test was developed, a simplified APGAR scoring system which was quick and easy to perform. The decreased score found in piglets born with a broken umbilical cord in this study, combined with an association between a broken umbilical cord and survival at weaning, may indicate that RMA is a practical score of piglet vitality that may be used to determine whether newborn piglets are at risk and in need of immediate care, such as open airways, warmth, protection against mechanical injury, and colostrum. The lack of association between RMA and birth weight is in accordance with several studies that report that being born small is not associated with being born less vital (Sonntag et al., 1996; Leenhouwers et al., 2001, 2002; Rootwelt et al., 2012b). The RMA only reflects the piglet status immediately after birth and does not include birth weight as a characteristic. A compromised piglet at birth may therefore not necessarily need special care for several days.

Recorded mean blood concentrations of lactate, an indicator of hypoxic stress, were increased in piglets which died before weaning vs. piglets that survived. This is in accordance with earlier studies (Herpin et al., 1996; Alonso-Spilsbury et al., 2005; Rootwelt et al., 2012c). Base excess, an indicator of longer lasting hypoxia (Ross and Gala, 2002; Van Dijk et al., 2006), was not significantly reduced in piglets which died before weaning. This is not surprising because the piglets in this study were already preselected as alive at birth. The majority of the piglets died within the first 3 d of life and is in accordance with the review of preweaning survival in swine by Lay et al. (2002). The review of Alonso-Spilsbury (2005) reports association between asphyxia and increased peripartum risk of being crushed by the sow. Miller and Miller (1965) also documented that although piglets survived up to 5 min during asphyxia conditions, shorter periods of asphyxia led to brain damage and were presumed to be a contributor to morbidity during early neonatal life. Yet, one may speculate on the exact cause of a somewhat delayed death after a birth with severe hypoxia because a delayed energy failure develops in piglets 24 to 48 h after birth (Lorek et al., 1994), and neuronal death is reported to occur several days after the insult (Pulsinelli et al., 1982).

Mean time until birth from the expulsion of the first piglet in the litter was not different between piglets that died before weaning vs. those that survived. These results may seem contradictory to studies that show that increased birth time is associated with stillbirth (Friend et al., 1962; Randall, 1972; Rootwelt et al., 2012b), with studies that report that piglets born in the last third of a litter are more asphyxic compared with those born earlier in the litter (Herpin et al., 1996; Van Dijk et al., 2006; Rootwelt et al., 2012c), and with results that show that increased birth order is unrewarding with regard to postpartum survival (Baxter et al., 2008). Nevertheless, increased birth time was associated with a decreased RMA score. Additionally, increased birth time also incorporates piglets most probably having had a cranial uterine location. Such piglets are hypothesized by Stanton and Carroll (1974) to be heavier compared with piglets with locations in the middle of the uterine horns, although some anatomical differences exist between species. A review of Leman et al. (1979) and recent results of our study group (Rootwelt et al., 2012b) also indicate correlation between piglet weight and uterine location. Birth weight is unanimously positively associated with postpartum survival (Devillers et al., 2004; Baxter et al., 2008; Boulot et al., 2008), as also found in this study. Interestingly though, a reduced BMI found in the nonsurviving piglets indicates that not only are these piglets smaller, but they are also relatively thinner than their litter mates. Reduced energy reservoirs perhaps could be the cause of this, as supported by Randall (1979) who reported reduced liver and cardiac glycogen concentrations in asphyxiated piglets than in litter mates. In our study, a birth weight of 1 kg seemed to be a threshold of increased postpartum survival, which is supported from results of Vallet et al. (2012). A high percentage of postpartum deaths for piglets from 1 of the upper BW ranges of this study probably is due to few piglets of this group or the confounding effect of the association between larger piglets having a more cranial uterine location.

Concentrations of hemoglobin and hematocrit were both reduced in piglets not surviving to weaning. This is in accordance with earlier studies (Herpin et al., 1996; Rootwelt et al., 2012c). Leenhouwers et al. (2002), on the other hand, found no association between hematocrit at birth and postpartum survival, although a normal hematocrit concentration does not exclude the possibility of blood loss through a recently ruptured umbilical cord. The decreased hemoglobin values recorded are only about 10% reduced in piglets that did not survive to weaning. This suggests that only minor differences have strong impact on vitality. Yet, interestingly, the hemoglobin concentrations recorded in the nonsurviving piglets in this study does not indicate critical anemia (Egeli, 2008) and could therefore rather be a symptom instead of a contributing cause of mortality. In this study, the concentration of hemoglobin was reduced in piglets born with a broken umbilical cord. In the horse, blood is commonly transferred from the placenta to the fetus through the umbilical cord also postpartum. This is by many considered as beneficial for the newborn, although significantly larger blood volume was not found in foals whose cords were left intact for 15 min vs. ligated at delivery (Lopate et al., 2003). In humans, delayed umbilical cord clamping at birth is reported to reduce the prevalence of anemia in human newborns (Jaleel et al., 2009). Although the primary and most acute function of the umbilical cord is oxygen delivery to the fetus, one may speculate if a final blood delivery interrupted through a prepartum broken umbilical cord may be just the threshold difference that is sufficient to cause death in the compromised piglet.

The umbilical vein supports the piglet with nutrients and oxygen rich blood from the placenta (Huppertz, 2011) and is therefore a determinant of most blood concentrations of the piglet. Samples of the umbilical veins revealed no differences in any of the recorded blood parameters between piglets that died before weaning vs. those that survived, although the numerical values showed the same tendency as in the piglets. As expected, all recorded parameters indicative of oxygen availability were favorable in the blood of the umbilical vein vs. in the piglet, except for the concentrations of hemoglobin and hematocrit. The explanation for a decreased hemoglobin and hematocrit concentration in the umbilical vein vs. piglet is unclear but may reveal a sudden hemoconcentration of the newborn piglets or too few successful blood samples of the umbilical veins to represent “true” values.

Of the electrolytes recorded in this study, only sodium was different between groups. Sodium depletion almost invariably is associated with excessive losses of sodium-containing fluid, as for instance blood loss (Carlson and Bruss, 2008). It seems unlikely that blood loss has occurred for a long enough time to result in relative water excess and thus reduced plasma sodium in the newborn piglet. Furthermore, the mean sodium concentration of the nonsurviving piglets was in this study only numerically reduced with 1.3 mmol/L in the live-born piglets vs. with 5 mmol/L in stillborn piglets (Rootwelt et al., 2012b). Maintenance of a transmembrane sodium gradient by the sodium pump requires continual metabolism and generation of ATP (Hornbuckle et al., 2008). Because of the positive associations between placental area and weight, birth weight, and BMI and survival, it is therefore tempting to speculate if the increased sodium concentration in the surviving piglets is not a cause of increased survival but a result.

In our study, blood concentrations of IgG at d 1 were increased in piglets that survived to weaning. Because piglets are born virtually without immunoglobulins (Prokesova et al., 1969; Martin et al., 2005), concentrations of IgG at this time point reflect a measure of colostral uptake and could thus be an indicator of vitality. Also, IgG protects the piglets against environmental infections before weaning. Blood concentrations of albumin increase more than threefold during the first day of life (Rootwelt et al., 2012c) and colostrum thus seems to be the source of this. Albumin contributes to maintain the colloid osmotic pressure and blood volume. Although it constitutes about 50% of the total protein mass in circulation, it is responsible for 80% of the colloid osmotic pressure. Albumin also serves as a transport protein, and several metabolites circulate in the blood bound to this protein (Eckersall, 2008). In this study, blood concentration of albumin was positively associated with survival at weaning and, interestingly, was also the only recorded blood variable on d 1 which was associated with placental area and placental weight. This result seems supported by those of Stone and Christenson (1982) and Herpin et al. (1996), who found a correlation between blood concentrations of albumin and birth weight. Because increased placental area is highly associated with increased birth weight (Rootwelt et al., 2012b) and heavier piglets more easily gain access to food through the constant success in competition within a litter, an increased concentration of albumin may only be an indicator of survival rather than a contributor. Nevertheless, IgG is also a measure of colostrum uptake and was increased in piglets that survived to weaning, but this parameter was neither associated with placental area nor placental weight.

Large litters are a prerequisite for profitability in swine production, and there is a constant ethical question of when an increased litter size compromises animal welfare to an unacceptable level. There is strong association between placental area and birth weight (Rootwelt et al., 2012b), between birth weight and litter size, and between placental area and litter size (Knight et al., 1977; Vallet et al., 2011), as also found in this study. These results combined with numerous studies which show association between birth weight and postpartum survival, indicate that with regard to postpartum survival, uterine endometrial contact surface is a major limiting factor. Because of the high correlation between birth weight, placental weight, and placental surface area (Biensen et al., 1998; Rootwelt et al., 2012b), it is challenging to make good statistical models where a placental property is exclusively distinguished from birth weight. The figures visualizing number of dead piglets related to placental area and placental weight did not add more information than when related to birth weight. Placental measurements, which are time consuming to make, cannot be included in a vitality score, nor do they add more predictive information than the simple birth weight. They may, however, be part of the pathophysiological explanation as to why increased litter size is negatively associated with birth weight and postpartum survival.

We conclude that placental area and placental weight, status of the umbilical cord, birth weight, BMI, and several blood variables recorded at birth and d 1 were associated with postpartum death in this study. Birth weight and placental area were negatively associated with live litter size. Placental area and placental weight were positively associated with blood concentration of albumin at d 1, a variable associated with survival to weaning. With regard to postpartum survival, in contrast to stillbirth, these results support the theory of an upper uterine limitation in litter size. Litter sizes above a certain threshold may lead to an increase in postpartum piglet death if the compromised neonate is not provided supportive care.

 

References

Footnotes


Comments
Be the first to comment.



Please log in to post a comment.
*Society members, certified professionals, and authors are permitted to comment.