Risk factors associated with stillbirth in sows

ABSTRACT: In this study, the risk factors associated with stillbirth in sows were identified and their odds ratio assessed. For this purpose, 587 farrowings on Farm A and 929 on Farm B were monitored, and the sow parity, body condition score, farrowing duration, total number of piglets born, numbers of live births, stillbirths, and mummified piglets, obstetric interventions, and piglet sex and weight were recorded. At the end of farrowing, piglets classified as stillborn were necropsied to confirm the diagnosis. Consequently, 5.49% of the piglets on Farm A and 5.10% of those on Farm B were stillborn. On both farms, sows with a high parity, prolonged farrowing, and a large litter size had the highest odds ratio of stillbirths. On Farm B, farrowing intervention through the use of vaginal palpation and oxytocin increased the odds of stillbirth by 1.7 and 2.5 times, respectively. Heavy litters increased the odds of stillbirth by 1.4 times. Additionally, low-birth-weight piglets were 2.3 and 3.1 times more likely than their medium-birth-weight and high-birth-weight counterparts, respectively, to be stillborn. In conclusion, on both farms, the risk factors associated with stillbirth were a high parity, a large litter size, and prolonged farrowing.


INTRODUCTION
In industrial pig farming, the annual number of piglets weaned per female is commonly used to express the reproductive efficiency index of the herd, which is mainly influenced by the number of live births and rate of pre-weaning mortality.Considering these factors, genetic selection through the introduction of hyperprolific sows (UDOMCHANYA et al., 2019) has been practiced to increase the number of piglets born, and hence weaned, per female (KOKETSU et al., 2017).However, the number of stillbirths increases with increasing litter size (BORGES et al., 2005).
Given the importance of knowing the risk factors for stillbirth and the impact of sows on herd efficiency, data on stillbirths in hyperprolific sows must be updated to define strategies for mitigating such occurrences in production systems.This study identified the risk factors associated with stillbirth in sows and to assess their odds ratio.

Animals and recorded parameters
This study was conducted on two commercial pig farms located in the southern region of Brazil.Farm A (24°55′S,50°05′W) houses 5,500 F1 Landrace × Large White sows (Camborough ® , Agroceres PIC ® , Patos de Minas, Minas Gerais, Brazil) in raised farrowing crates with a plastic floor (1.6 × 2.2 m) in air-conditioned negative-pressure rooms.On this farm, the study was performed between January and February 2018, with 587 farrowings monitored in total.Farm B (27°03′58.9″S,52°19′16.6″W)houses 2,500 F1 Landrace × Large White sows (TN70, Topigs Norsvin, Videira, Santa Catarina, Brazil) in raised farrowing crates with a plastic floor (1.6 × 2.2 m) in rooms fitted with the Ducto Fan ventilation system.On this farm, the study was performed between March and December 2019, with a total of 929 farrowing monitored.
Data on the age, breed, and parity of the sows were recorded once they had been transferred to the farrowing room.The body condition score of between 1 and 5 (1 = thin; 5 = fat), Caliper score, and ultrasound-assessed back fat thickness at P2 (Renco ® ) were recorded on Farm A only.
The farrowing was all supervised, with data on the sow parity, total farrowing duration, obstetric interventions (vaginal palpation and use of pharmacological drugs), total number of piglets born, numbers of live births, stillbirths, and mummified piglets, and piglet sex recorded.The farrowing duration was considered to have started with the birth of the first piglet and ended with fetal annexes expulsion.Obstetric interventions were performed when necessary, following the protocol: vaginal palpation was initially used when the sow had already farrowed at least one piglet and the interval between subsequent births was longer than 20 min.If uterine contractions were still absent despite the initial intervention, then 20 IU of oxytocin was administered intramuscular (Oxytocin, Biofarm, Brazil).On Farm B, the individual weight of each piglet and the mean litter weight were also recorded.The piglet sex and weight were recorded within 12 hours after birth.The stillborn piglets were necropsied for their classification as prepartum (piglets with organs showing variable autolysis, intracavitary fluid, and bluish corneas), intrapartum (piglets with meconium on the skin and colostrum in the stomach and without aerated lungs), or postpartum (piglets with aerated lungs, but no colostrum in the stomach, and variable stomach contents) stillbirths, according to the methodology proposed by LEENHOUWERS et al. (2003).On Farm B, the stillborn piglets were subjected to the lung test and classified as antepartum (without aerated lungs) or postpartum (with aerated lungs) stillbirths.

Statistical analysis
The data were analyzed using Statistical Analysis System software (v.9.4,SAS Institute Inc., Cary, NC, USA).Descriptive statistics (mean and standard deviation) and normality and frequency distributions were generated to characterize the occurrence of stillbirths (UNIVARIATE and FREQ procedures).Spearman's rank correlation analysis (CORR procedure) was performed to determine whether the variables under study were associated with stillbirths.All variables that correlated significantly with stillbirths were then categorized and included in the logistic regression model; namely, parity (1, 2-4, and 5-6 on Farm A; and1, 2, and 6-7 on Farm B); use of oxytocin during farrowing (yesor no), vaginal palpation (yes or no), and obstetric interventions (yes or no); farrowing duration (< 300 min and ≥ 300 min); total number of piglets born (<12 piglets and ≥12 piglets on Farm A; < 14 piglets and ≥ 14 piglets on Farm B); and litter weight (< 20.3 kg and ≥ 20.3 kg) and mean piglet weight (<1.308 kg, 1.309-1.625kg, and ≥ 1.626 kg) on Farm B. Logistic regression analysis was performed using the GLIMMIX procedure with binary distribution.The categorized variables were considered fixed effects.In another approach, the occurrence of stillbirths was analyzed using the GLIMMIX procedure with binomial distribution, with continuous variables categorized by quartiles.Significance was defined as P < 0.05 for all tests, and differences with P-values between 0.05 and 0.1 were considered significant.

RESULTS
The percentages of stillbirths were similar between the farms under study, reaching 5.49%and 5.10% on farms A and B, respectively (Table 1).Of the total percentage of stillbirths on Farm A, 1.73% were classified as prepartum, 3.30% as intrapartum, and 0.45% as postpartum (Table 1).On Farm B, the percentage of postpartum stillbirths (0.41%) was similar to that observed on Farm A. There was no difference between the percentages of antepartum and intrapartum stillbirths on Farm A, with the two categories combined accounting for 5.03% stillbirths.On Farm B, all live (1.46 ± 0.38 kg) and stillborn (1.20 ± 0.41 kg) piglets were weighed, where up on the stillborn were found to be 260 g lighter at birth (P < 0.001).
The variables significantly associated with stillbirth on both farms were identified using Spearman's rank correlation method.Some variables, such as parity, farrowing duration, and total number of piglets born, were significantly correlated with stillbirth (P ≤ 0.05) on both farms.Variables such as oxytocin use, vaginal palpation, and a high litter weight were also significantly (P < 0.01) associated with stillbirth on Farm B, reaching 50.76%, 60.00%, and 43.99%, respectively.Broadly speaking, the correlations were weak.Nonetheless, the significant variables were analyzed in detail to understand their effect on stillbirth (Table 2).
On both farms, sows with a parity number higher than 5 and between 6 and 7 were respectively 1.6 and 1.8 times more likely to farrow stillborn piglets than sows with parity ranging from 2 to 4.On Farm B, oxytocin use, obstetric interventions, and vaginal palpation increased the likelihood of piglets being stillborn by 1.7, 1.8, and 2.5 times, respectively.However, the effects of those interventions on the percentage of stillbirth were not identified on Farm A (Table 3).Sows with prolonged farrowing (≥300 min) were 1.8 times more likely to birth stillborn piglets than sows with short farrowing (<300 min).Large litters (≥12piglets on Farm A and ≥14 piglets on Farm B) increased the likelihood of stillbirths by 3.9 and 3.3 times, respectively.On Farm B, heavy litters (≥20.3 kg) were 1.4 times more likely to include stillborn piglets.However, for piglets with a low birth weight (<1.308kg), the risks of being stillborn were 2.3 and 3.1 times higher than for those with a medium (1.309-1.625 kg) and high birth weight (≥1.626 kg), respectively (Table 4).The frequencies of farrowing with stillbirths on farms A and B were 49.40% (parity 5-6) and 48.00% (parity 6-7), respectively.The percentages of stillbirths from farrowing lasting 300 min or longer were 56.00% on Farm A and 48.52% Table 1 -Description of sow parameters, farrowing performance, and litter characteristics (mean ± SD) in the farms under study.
The total number of piglets born also significantly affected the risk of stillbirth, as described in previous studies in which sows with a litter of more than 9 (NAAM & SUKON, 2020a) or 12 (BORGES et al., 2005;LUCIA JR et al., 2002) piglets had a higher risk of farrowing stillborn.The increased occurrence of stillbirths in larger litters is mainly due to the longer farrowing duration (NAAM & SUKON, 2020a), as farrowing lasting longer than 300 min subject the piglets to unfavorable conditions (ALONSO-SPILSBURY et al., 2005).In this study, prolonged farrowing (≥ 300 min) increased the risk of stillbirth by 1.8 times.This may be related to the farrowing process itself, as intense uterine contractions may promote rupture of the umbilical cord prior to its expulsion, cutting off the blood flow to the placenta and leading to piglet asphyxia and hypoxia (MOTA-ROJAS et al., 2006).
Sows with a parity number ranging from 5 to 7 had a higher risk of farrowing stillborns than those with parities ranging from 2 to 4. Older sows tend to have lower muscle tone (LE COZLER et al., 2002), resulting in poor expulsion of the fetuses and

Variable
----------------------------Farm A (587 farrowing)---------------------------SEM P-value  a prolonged farrowing duration (VANDERHAEGHE et al., 2013;NAAM & SUKON, 2020a).Obstetric intervention, whether through the use of oxytocin or vaginal palpation, increased the risk of stillbirth.Although, oxytocin is a drug used to promote uterine contractions in sows, its use remains controversial with regard to the rate of stillbirths (NAAM & SUKON, 2020b).Oxytocin administration has been shown to reduce the farrowing interval between piglets and the farrowing duration, decreasing the risk of stillbirth (UDOMCHANYA et al., 2019).However, if the dose exceeds a specific amount or is administered with insufficient cervical dilation, then the risk of farrowing problems and the number of stillborn piglets increase (LE COZLER et al., 2002).In this study, the use of oxytocin increased the likelihood of stillbirth by 1.7 times, possibly as a result of the drug activity, which increases the frequency of uterine contractions but consequently also reduces blood flow through compression or tearing of the umbilical cord, thereby resulting in intrauterine piglet hypoxia (MOTA-ROJAS et al., 2006).Farrowing supervision provides the help needed in cases of dystocia (LE COZLER et al., 2002;VANDERHAEGHE et al., 2013). However, HOLM et al. (2004) observed that prolonged or dystocic farrowing were associated with a greater need for farrowing supervision and a higher percentage of stillbirths.In the present study, the risk of stillbirth increased 2.5 times when vaginal palpation was applied.However, the procedure was performed only when farrowing assistance was required, suggesting that it may not be the causative factor of stillbirth.

CONCLUSION
High parity, a large litter size, and prolonged farrowing were the risk factors associated with stillbirths in sows on two farms in Brazil.Among these risk factors, a large litter size had the highest odds ratio for stillbirths.

Table 3 -
Percentage of sows with stillborn piglets, and results from the logistic regression analysis of the association of risk factors with stillbirth (Farm A).

Table 2 -
Spearman's rank correlation coefficient of stillbirths by sow parameter, farrowing performance, and litter characteristics in the farms under study.

Table 4 -
Percentage of sows with stillborn piglets, and results from the logistic regression analysis of the association of risk factors with stillbirth (Farm B).

Table 5 -
Stillbirth rates on farms A and B, according to different variables classified by quartile distribution.