- Email: [email protected]
Foetal loss in dairy goats: an epidemiological study in 515 individual goats
REPmoN SCIENCE Animal Reproduction Science 49 (1997) 45-53
Foetal loss in dairy goats: an epidemiological study in 5 15 individual goats Inge Vogt Engeland a,*, Harald Waldeland Aage Tverdal ’ ” Department
qf Reproduction
and Forensic
Medicine,
Norwegian
P.O. Box 8146 Dep., Oslo NO-0033, h Depmtment
ofSheep
and Goat Research, Norwegian Sandnes NO-4301,
’ Department
of Morphology.
Genetics and Aquatic
b, (dystein Andresen a, College
qf Veterinary
Medicine,
Nornq
College of Veterinan
Medicine,
P.O. Box 264.
Norway
Biology,
Norwegian
P.O. Box X146 Dep., Oslo NO-0033,
College of Veterinu~
Medicine,
Norwq
Accepted 24 July I997
Abstract Effects of various factors pertaining to the individual goat on the reproductive performance were studied during one season in 515 pregnant dairy goats from seven herds. A significant difference in the risk of foetal loss was found between the herds. Advanced age, difficulty in conceiving, low social status, pregnancy with 2 3 foetuses and previous foetal loss were significantly associated with current loss. No association was found between foetal loss and the current goat index, milk production last year, daily milk production al the time of dry-off and the presence or absence of horns or dehorning of the goat. The risk of foetal loss was not related to mating with specific bucks. No indication of relationship between goats with foetal loss was found by pedigree analysis including the sire, dam and dam’s sire. However, the incidence of foetal loss was significantly higher in daughters of does that had suffered reproductive loss than in daughters of does without such a loss. This suggests a maternal or genetic influence on the occurrence of foetal loss. Culling of goats suffering non-infectious abortion and of their surviving progeny therefore seems reasonable. 0 1997 Elsevier Science B.V. Kewordsr
Goat: Foetal loss; Abortion: Age influence: Reproductive performance
* Corresponding author. Tel.: + 47 22 96 48 53; fax: +47 22 56 57 04; e-mail: [email protected]. 037%4320/97/$17.00 0 1997 Elsevier Science B.V. All rights reserved PII SO378-4320(97)00059-6
46
I. V. Engeland et al. /Animal Reproduction Science 49 (1997) 45-53
1. Introduction Foetal loss of non-identified causes is a serious problem in Norwegian goat husbandry (Melby et al., 1986; Loken, 1990; Waldeland and Loken, 1991). Engeland et al. (1997) studied the incidence of foetal loss in 22 goat herds, and found that 11.1% of the animals either aborted, delivered dead foetuses at full term, or were found empty after having been identified as pregnant. Listeria monocytogenes or Toxoplasma gondii was found to be the cause of foetal loss in only 12 (7.5%) out of the 160 goats with foetal loss. In nearly all herds the loss occurred in goats aged 2 3 yrs. In most cases a decomposed foetus was expelled during the last two months of pregnancy. Usually no other symptoms of disease were observed. Since infectious causes were found in relatively few cases of foetal loss in goats in Norway, Waldeland and Loken (1991) suggested that nutritional and environmental factors may be important. Engeland et al. (1997) reported that environmental conditions such as inferior natural lighting in the stable, pre-heated drinking water and a building design with a combined feeding and milking platform involving large and crowded pens were associated with a high incidence of foetal loss. The association between conditions in the herd and reproductive performance was further confirmed by the finding that a history of previous foetal loss in the herd was significantly linked to a high incidence of loss during the investigated pregnancy. Since the individual’s reactions to adverse conditions may vary, it was decided to study the information pertaining to the individual goat in selected herds. The intention was to relate individual traits to the outcome of pregnancy and thus provide insight into the causes and mechanisms behind foetal loss of non-identified causes.
2. Materials and methods
2.1. Goats and management Seven dairy herds comprising a total of 532 goats were selected from the region of Hallingdal and Valdres in Norway. The selection was based mainly on the incidence of foetal loss during each of the previous 3 years when five of the herds had a loss of > 15% (herd 1, 3, 4, 5 and 61, while the remaining two had a loss of < 10% (herd 2 and 7). The two flocks with a previous low rate of abortion were included to elucidate factors that may explain this difference in reproductive performance. All farmers kept one or two bucks in a separate pen during the whole breeding season and gestation period. During the first two months after the mating started, all goats were observed for oestrus in the presence of a buck twice daily. The goats were hand-mated and the dates of mating were registered. The goats were carefully observed for oestrus in the presence of a buck for 10 to 15 min twice daily 19-23 days after mating, and again at days 40-44. A goat was presumed to be pregnant if signs of oestrus were not observed in the latter two periods. Only goats presumed to be pregnant, totally 5 16, were included in this study. The herds were fed roughage ad libitum during pregnancy, and the main feed was grass silage in six herds and hay in one herd. In addition, the animals were fed equal rations of commercially available concentrates, and most rations also included salt stone.
I.V. Engeland
et al. /Animal
Reproduction
Science 49 (1997145-53
17
The ration of concentrates was reduced for a short period in the last half of pregnancy to terminate milk production. Foot trimming was carried out once or twice during pregnancy. Treatment against nematodes and lice during the second month of pregnancy, shearing during the third month of pregnancy and vaccination against clostridial infections prior to kidding were undertaken in six of the seven herds. 2.2. Examination
of aborted material and serological exuminations
Examination of aborted material and serological described by Engeland et al. (1997). 2.3. Design
examinations
were performed
as
of the study
The study was based on the individual goat as the unit of analysis and with foetal loss (yes or no) as the outcome. Available information was recorded for each goat before the outcome of pregnancy was known. The recorded information included herd, age, the index of the goat which is mainly based on relatives and individuals milk production, the number of matings required to achieve pregnancy, identity of the buck, social status based on the farmers observations (high, medium, low), number of foetuses, previous individual reproductive history, the presence or absence of horns or dehorning, the total volume of milk produced the previous year and the daily volume of milk at the time of dry-off in the current investigation. Available information concerning the mother’s reproductive history was also recorded for 67 goats with foetal loss and for an equal number of controls, matched on herd and age. The pedigree of the 79 goats with foetal loss were compared with those of an equal number of unaffected animals matched on herd and age. The pedigree assessment included the individual goat’s sire, dam and dam’s sire. 2.4. Statistical analyses The analyses were performed on different numbers of goats according to the number of variables that were available: (1) all 515 goats, (2) 493 goats (excluding 22 goats with an undetermined number of foetusesl, (3) 362 goats (excluding 153 goats younger than 1 yr), (4) 134 goats with mother’s reproductive history as an additional variable, i.e. 67 pairs matched on age and herd, of which one in each of the pairs suffered reproductive loss while the other had a normal pregnancy. Each of the following variables was subdivided by quartiles before analysis: the goat index for breeding value, milk production during the previous year and daily milk production at the time of drying off. Chi-square tests were used both for pedigree analysis and in the analyses of univariate associations. Multivariate logistic regression were undertaken to adjust for potential confounders. Interacting effects were tested by the log likelihood ratio. In the case of matched pairs, conditional logistic regressions were performed.
I. V. Engeland et al. /Animal Reproduction Science 49 (1997) 45-53
48
3. Results L. monocytogenes was detected in one of the goats that aborted, and this individual was therefore omitted from the present material. No other significant organisms were detected by serological examination of goats that experienced foetal loss or by mi-
Table 1 Incidence loss
of foetal loss by herd, age, repeated mating,
Variables
social status, number of foetuses and previous
Number of goats With foetal loss
Incidence of foetal loss (%) Without foetal loss
Herd no.a 1 2 3 4 5 6 7
17 6 10 10 15 14 I
56 58 64 80 63 45 70
23.3 (reference) 9.4* 13.5* 1l.P 19.2* 23.7 9.P
Age fyrs,” 1 2 3 4 5 6 t1
12 12 17 16 9 5 8
141 91 71 55 29 27 16
7.8 (reference) 11.0 19.3* 22.5* 23.7* 15.6 33.3*
Pregnancy achieved on heat number” 1 64 2 9 23 6
364 60 12
15&eference) 13.0 33.3*
Social stat& Low Medium High
13 49 17
34 345 57
27.7Creference) 12.4* 23.0*
Number offoetusesb 1 19 2 29 23 9
167 246 23
10.2 (reference) 10.5 28.1*
Previous foetal loss’ No 43 Yes 24
260 35
14.2Creference) 40.7*
“Information from 515 goats. bInformation from 493 goats. ‘Information from 362 goats. *Logistic regression revealed statistical significantly
different from reference
category
foetal
1.V. En&and
et al. /Animal Reproduction Science 49 ( 1997) 45-53
49
croscopy or culture from the maternal placental tissues, foetal stomach contents or organs. Out of the remaining 515 goats, 79 (15%) lost their foetuses from non-identified causes. Thirty-eight (48%) of the goats aborted, 33 (42%) delivered dead foetuses at full term and eight (10%) were found empty after having been identified as pregnant. In six of the goats that aborted, only bloody vaginal discharges were observed. No symptoms of disease such as depression or anorexia were seen in connection with reproductive loss. Out of 89 foetuses examined, 18 (20%) were expelled in a relatively fresh state whereas 71 (80%) had a decomposed or mummified appearance. In eight of the latter cases, the number of foetuses was difficult to estimate due to advanced decomposition or maceration. A significant association was found between foetal loss and herd (Table 1). Goats aged 3 yrs or over exhibited significantly greater incidence of foetal loss than those aged under 3 yrs (21.7% vs 9.2%, P < 0.001). In the goats that did not mate successfully before the 3rd or later heats, the incidence of foetal loss was significantly increased compared to those that became pregnant on an earlier heat (33.3% vs 9.2%, P < 0.05). With respect to social hierarchy, goats with a low social status exhibited a significantly greater incidence of foetal loss than those with high or medium status (27.7% vs 14.1%, P < 0.02). The incidence of foetal loss in goats that had experienced previous foetal loss
Table 2 Incidence
of foetal loss by goat index, milk production
Variables
With foetal loss Goat index b < 98 98-102 103-106 2 107
Incidence of foetal loss (%I’
563-676 677-799 2 800
15.5 17.2 18.4 22.7
78 74 73 70
12.4 19.6 18.9 23.1
19 21 15 12
57 88 85 65
25.0 19.3 15.0 15.6
52 26
310 106 20
14.4 19.7 4.8
18 17 21 at dr$ng
production
Horn statusC Dehorned Polled Horned
Without foetal loss
60 96 71 68
11 20 16 20
Milk production lust year fkglb < 563 11
Dail! milk < 0.8 0.8-1.3 I .4- 1.9 2 2.0
and presence of horns
Number of goats
off (kgjb
1
“Logistic regression revealed no statistical bInformation from 362 goats. “Information from 5 15 goats.
significant
differences
between the variable categories,
50
I.V. Engeland et al./Animl
Reproduction Science 49 (1997) 45-53
was significantly increased compared to the goats without such an experience (40.7% vs 14.2%, P < 0.001). In goats from which the number of foetuses could be recorded, animals pregnant with three or more foetuses exhibited a significantly greater incidence of foetal loss than in those with one or two foetuses (28.1% vs 10.4%, P < 0.01) (Table 1). Conditional logistic regression revealed that foetal loss was significantly higher in daughters of does that had suffered reproductive loss than in daughters of does without such a loss (P < 0.05). Multivariate logistic regression did not substantially alter the results. Models with interacting terms gave no better fit than models without. No association was found between foetal loss and the presence or absence of horns or dehorning or daily milk production at the time of dry-off. There was a trend for increasing incidence of foetal loss, although not statistically significant, with increasing goat index and milk production the last year (Table 2). The pedigree analysis did not reveal a different lineage in goats that experienced foetal loss compared with those with a normal pregnancy. Twenty-six bucks had mated the goats with foetal loss, and twenty-five of the bucks had also mated at least an equal number of goats with a normal pregnancy. The 26th buck was used only for a single goat that aborted.
4. Discussion The statistical analyses were performed on goats that showed foetal loss due to non-identified causes and on goats with a normal pregnancy. The exposure variables were mainly based on objective observations, and were recorded before the outcome of the pregnancy was known, to prevent biased misclassifications. Classification of social status was based on a subjective score performed by the farmer, and is not an ideal method. However, this is a commonly used and accepted approach in clinical studies. Since two of the seven herds (herd 2 and 7) had a history of previous low incidence of foetal loss (< 10%) while the incidence in the five other herds had been higher (> 15%), it was not surprising to find a significant association between foetal loss and herd during the investigated season (Table 1). It is possible that different environmental conditions could explain this difference in incidence of foetal loss as discussed by Engeland et al. (1997). No significant organisms were detected by serological examination of goats that experienced foetal loss or by microscopy or culture from the maternal placental tissues, foetal stomach contents or organs. The increased incidence of foetal loss among older goats supports the observation that the abortions were of a non-infectious nature. If infectious agents were involved, the rate of foetal loss should decrease with age due to acquired immunity. In South African Angora goats that aborted, Van Rensburg (1971) found an increase in the abortion rate with increasing age, due to altered adrenal status (hyperadrenalism) with increasing age. An immunological rejection of the foetus as demonstrated by Corbel (1972) in pregnant sheep infected in utero with Aspergillus
I. V. Engeland et al. /Animal Reproduction Science 49 (1997) 45-53
51
fumigatus, could account for repetitive abortions among older goats as discussed by Waldeland and Loken (19911. However, no evidence of such a rejection in Norwegian goats has been found (Waldeland and Loken, 199 1). The goats were mated on the first heat in the autumn, and only a few returned to service. The analyses revealed that if the goats conceived on the 3rd or later heat, this significantly increased the risk of foetal loss (Table 1). Various anomalies such as malformations, inferior luteal function or infections may be the cause of unsuccessful mating. Such abnormalities may also dispose the goat to foetal loss later during pregnancy. Low social status seemed to predispose the individual to foetal loss compared with high or medium status. It is possible that the basic reason for this loss could be chronic stress from social subordinance. In sheep, management procedures such as confinement, trucking: dipping and shearing have been found to increase the level of plasma cortisol (Kilgour and de Langen, 19701, which is commonly used as an indicator of stress. Experimental luteolysis has been obtained by prolonged cortisone administration in Angora goats (Wentzel and Roelofse, 19751 and with infusion of ACTH into goat foetuses (Thorbum et al., 1972). In South African Angora goats that aborted, Wentzel et al. (1975) found abnormally high maternal blood corticosteroid levels and extensive adrenal hyperplasia, which was suggested to create an unfavourable foetal environment inducing expulsion of foetuses which had an oedematous appearance. However, in the present study a large proportion of the foetuses were expelled in an decomposed or mummified condition. To investigate whether social stress may be involved in foetal loss, plasma levels of corticosteroids in goats under various conditions should be determined. Studies to clarify the genetic influence on the cortisol level should also be initiated. In stressed pregnant rats, an increased incidence of abortion in the offspring has been observed (Herrenkohl, 1979). In addition to a potential genetic influence, it is therefore also possible that an adverse environmental effect on the mother may have been passed on to the offspring and affected foetal survival in the daughter. In goats with three or more foetuses foetal loss occurred more often than in goats with one or two foetuses. It is possible that nutrition of each foetus via the placenta is reduced proportionally with increasing number, and that placental insufficiency in such cases may lead to foetal death (Matthews, 1991). Social rank was probably of little importance for nutrition of the does because roughages were given ad libitum, and the animals were fed equal rations of concentrated when the goats were fixed in feeding yokes. However, it is possible that both an elevated number of fertilised ova and multiple pregnancy may be associated with high social status and optimal body condition, and that stress from maintaining high social rank may also have an adverse effect. Data from the present study were analysed in different ways to investigate whether there was a possible genetic influence on the foetal loss. The possibility that some cases of previous loss in both the goats examined and their mothers may have been caused by infections and still included in these analyses cannot be excluded, since extensive examinations of aborted materials and serological examinations were not performed then. However, the chances are small and this would probably have little influence on the results, since both the present and other recent studies indicate that infections are
52
I.V. Engeland et al. /Animal
Reproduction Science 49 (1997) 45-53
found in only few cases of abortion in goats (Melby et al., 1986; Loken, 1990; Waldeland and Loken, 1991; Engeland et al., 1997). Although the pedigree analysis comprising the goat’s sire, dam and dam’s sire did not provide evidence of a genetic influence, there was a significant association between mother and daughter in the occurrence of reproductive problems (P < 0.05). Previous foetal loss was also associated with present loss (P < 0.05). These results suggest that there may be a hereditary predisposition for foetal loss. It may therefore be worthwhile to cull affected animals and to exclude their previous offspring from the breeding stock. Under natural conditions, a hereditary tendency of abortion may be self limiting. It is, however, possible that the farmers’ selection of breeding goats may have counteracted this, such as in the Angora goat where a hereditary predisposition to abortion has been found to be associated with production and quality of mohair (Van Rensburg, 1971). It should be noted that in the present work a tendency of increasing incidence of foetal loss with increasing goat index and milk production was found (Table 2). However, the power in detecting statistically significant associations between foetal loss and milk production in the present study is relatively small due to restricted number of animals. Therefore, an epidemiological study should be undertaken on a larger scale to elucidate whether there is a similar relationship between milk production and foetal loss as between mohair production and abortion in Angora goats. In the present study, the milk production potential of each goat was not fully demonstrated because of limitation imposed by the production quota allocated to each herd, and a possible stronger association between production and foetal loss may thereby have been concealed. 5. Conclusion
Advanced age, difficulty in conceiving, low social status, pregnancy with 2 3 foetuses and previous foetal loss were significantly associated with current loss. Foetal loss were significantly higher in daughters of does that had suffered reproductive loss themselves than in daughters of does without such a loss. This association suggests a hereditary influence on the occurrence of foetal loss. Goats with a history of foetal loss from non-identified causes or with a mother with such a loss should therefore be culled. The goat breeding index, milk production last year, daily milk production at the time of dry-off and the presence or absence of horns or dehorning were not associated with foetal loss. Acknowledgements
Financial support was received from the Research Council of Norway. References Corbel, M.J., 1972. Production of antibodies to placental antigens by pregnant ewes experimentally infected with Aspergillusfumigatus. Br. Vet. J. 128, xliv-xlvi. Engeland, I.V., Waldeland, H., Andresen, $3.. Llen, T., Bjcrkman, C., Bjerkls, I., 1997. Foetal loss in dairy goats: an epidemiological study in 22 herds. Small Rumin. Res. (in press).
I. V. Engeland
Herrenkohl,
L.R.,
1979. Prenatal
et al. /Animal
stress reduces
Reproduction
fertility
Science 49 119Y7145-53
and fecundity
in female
offspring.
53
Science
206,
1097- 1099.
Kilgour, R., de Langen, H., 1970. Stress in sheep resulting from management practices. Proc. N.Z. Sot. Anim. Prod. 30, 65-76. Loken. T., 1990. Pestivirus infection in Norway. Epidemiological studies in goats. .I. Comp. Pathol. 103, I-JO. Matthews, J.G., 1991. Abortion. In: Matthews, J.G. (Ed.), Outline of Clinical Diagnosis in the Goat. Writh. Oxford. pp. 17-32. Melby, H.P., Aursjo, J., Binde, M., Gronstol. H., 1986. Diseases and mortality in 27 dairy goat farms. Nord. Vet. Med. 38, 403-41 1. Thorburn, G.D., Nicol, D.H., Bassett, J.M., Shutt, D.A., Cox, R.I., 1972. Parturition in the goat and sheep: changes in corticosteroids, progesterone, oestrogens and prostaglandin. F. J. Reprod. Fertil. 16, 61-84. Suppl. Van Rensburg, S.J., 1971. Reproductive physiology and endocrinology of normal and habitual aborting Angora goats. Onderstepoort J. Vet. Rea. 38, I-62. Waldeland, H., Loken, T., 1991. Reproductive failure in goats in Norway: an investigation in 24 herds. Acta Vet. Stand. 32, 535-541. Wentzel, D., Roelofse, C.S., 1975. The habitually aborting Angora doe: VII. Induction of abortion by administration of cortisone acetate. Agroanimalia 7, 45-48. Wentzel, D., Morgenthal. J.C., van Niekerk, C.H., 1975. The habitually aborting Angora doe: IV. Adrenal function in normal and aborter does. Agroanimalia 7, 27-34.
Foetal loss in dairy goats: an epidemiological study in 5 15 individual goats Inge Vogt Engeland a,*, Harald Waldeland Aage Tverdal ’ ” Department
qf Reproduction
and Forensic
Medicine,
Norwegian
P.O. Box 8146 Dep., Oslo NO-0033, h Depmtment
ofSheep
and Goat Research, Norwegian Sandnes NO-4301,
’ Department
of Morphology.
Genetics and Aquatic
b, (dystein Andresen a, College
qf Veterinary
Medicine,
Nornq
College of Veterinan
Medicine,
P.O. Box 264.
Norway
Biology,
Norwegian
P.O. Box X146 Dep., Oslo NO-0033,
College of Veterinu~
Medicine,
Norwq
Accepted 24 July I997
Abstract Effects of various factors pertaining to the individual goat on the reproductive performance were studied during one season in 515 pregnant dairy goats from seven herds. A significant difference in the risk of foetal loss was found between the herds. Advanced age, difficulty in conceiving, low social status, pregnancy with 2 3 foetuses and previous foetal loss were significantly associated with current loss. No association was found between foetal loss and the current goat index, milk production last year, daily milk production al the time of dry-off and the presence or absence of horns or dehorning of the goat. The risk of foetal loss was not related to mating with specific bucks. No indication of relationship between goats with foetal loss was found by pedigree analysis including the sire, dam and dam’s sire. However, the incidence of foetal loss was significantly higher in daughters of does that had suffered reproductive loss than in daughters of does without such a loss. This suggests a maternal or genetic influence on the occurrence of foetal loss. Culling of goats suffering non-infectious abortion and of their surviving progeny therefore seems reasonable. 0 1997 Elsevier Science B.V. Kewordsr
Goat: Foetal loss; Abortion: Age influence: Reproductive performance
* Corresponding author. Tel.: + 47 22 96 48 53; fax: +47 22 56 57 04; e-mail: [email protected]. 037%4320/97/$17.00 0 1997 Elsevier Science B.V. All rights reserved PII SO378-4320(97)00059-6
46
I. V. Engeland et al. /Animal Reproduction Science 49 (1997) 45-53
1. Introduction Foetal loss of non-identified causes is a serious problem in Norwegian goat husbandry (Melby et al., 1986; Loken, 1990; Waldeland and Loken, 1991). Engeland et al. (1997) studied the incidence of foetal loss in 22 goat herds, and found that 11.1% of the animals either aborted, delivered dead foetuses at full term, or were found empty after having been identified as pregnant. Listeria monocytogenes or Toxoplasma gondii was found to be the cause of foetal loss in only 12 (7.5%) out of the 160 goats with foetal loss. In nearly all herds the loss occurred in goats aged 2 3 yrs. In most cases a decomposed foetus was expelled during the last two months of pregnancy. Usually no other symptoms of disease were observed. Since infectious causes were found in relatively few cases of foetal loss in goats in Norway, Waldeland and Loken (1991) suggested that nutritional and environmental factors may be important. Engeland et al. (1997) reported that environmental conditions such as inferior natural lighting in the stable, pre-heated drinking water and a building design with a combined feeding and milking platform involving large and crowded pens were associated with a high incidence of foetal loss. The association between conditions in the herd and reproductive performance was further confirmed by the finding that a history of previous foetal loss in the herd was significantly linked to a high incidence of loss during the investigated pregnancy. Since the individual’s reactions to adverse conditions may vary, it was decided to study the information pertaining to the individual goat in selected herds. The intention was to relate individual traits to the outcome of pregnancy and thus provide insight into the causes and mechanisms behind foetal loss of non-identified causes.
2. Materials and methods
2.1. Goats and management Seven dairy herds comprising a total of 532 goats were selected from the region of Hallingdal and Valdres in Norway. The selection was based mainly on the incidence of foetal loss during each of the previous 3 years when five of the herds had a loss of > 15% (herd 1, 3, 4, 5 and 61, while the remaining two had a loss of < 10% (herd 2 and 7). The two flocks with a previous low rate of abortion were included to elucidate factors that may explain this difference in reproductive performance. All farmers kept one or two bucks in a separate pen during the whole breeding season and gestation period. During the first two months after the mating started, all goats were observed for oestrus in the presence of a buck twice daily. The goats were hand-mated and the dates of mating were registered. The goats were carefully observed for oestrus in the presence of a buck for 10 to 15 min twice daily 19-23 days after mating, and again at days 40-44. A goat was presumed to be pregnant if signs of oestrus were not observed in the latter two periods. Only goats presumed to be pregnant, totally 5 16, were included in this study. The herds were fed roughage ad libitum during pregnancy, and the main feed was grass silage in six herds and hay in one herd. In addition, the animals were fed equal rations of commercially available concentrates, and most rations also included salt stone.
I.V. Engeland
et al. /Animal
Reproduction
Science 49 (1997145-53
17
The ration of concentrates was reduced for a short period in the last half of pregnancy to terminate milk production. Foot trimming was carried out once or twice during pregnancy. Treatment against nematodes and lice during the second month of pregnancy, shearing during the third month of pregnancy and vaccination against clostridial infections prior to kidding were undertaken in six of the seven herds. 2.2. Examination
of aborted material and serological exuminations
Examination of aborted material and serological described by Engeland et al. (1997). 2.3. Design
examinations
were performed
as
of the study
The study was based on the individual goat as the unit of analysis and with foetal loss (yes or no) as the outcome. Available information was recorded for each goat before the outcome of pregnancy was known. The recorded information included herd, age, the index of the goat which is mainly based on relatives and individuals milk production, the number of matings required to achieve pregnancy, identity of the buck, social status based on the farmers observations (high, medium, low), number of foetuses, previous individual reproductive history, the presence or absence of horns or dehorning, the total volume of milk produced the previous year and the daily volume of milk at the time of dry-off in the current investigation. Available information concerning the mother’s reproductive history was also recorded for 67 goats with foetal loss and for an equal number of controls, matched on herd and age. The pedigree of the 79 goats with foetal loss were compared with those of an equal number of unaffected animals matched on herd and age. The pedigree assessment included the individual goat’s sire, dam and dam’s sire. 2.4. Statistical analyses The analyses were performed on different numbers of goats according to the number of variables that were available: (1) all 515 goats, (2) 493 goats (excluding 22 goats with an undetermined number of foetusesl, (3) 362 goats (excluding 153 goats younger than 1 yr), (4) 134 goats with mother’s reproductive history as an additional variable, i.e. 67 pairs matched on age and herd, of which one in each of the pairs suffered reproductive loss while the other had a normal pregnancy. Each of the following variables was subdivided by quartiles before analysis: the goat index for breeding value, milk production during the previous year and daily milk production at the time of drying off. Chi-square tests were used both for pedigree analysis and in the analyses of univariate associations. Multivariate logistic regression were undertaken to adjust for potential confounders. Interacting effects were tested by the log likelihood ratio. In the case of matched pairs, conditional logistic regressions were performed.
I. V. Engeland et al. /Animal Reproduction Science 49 (1997) 45-53
48
3. Results L. monocytogenes was detected in one of the goats that aborted, and this individual was therefore omitted from the present material. No other significant organisms were detected by serological examination of goats that experienced foetal loss or by mi-
Table 1 Incidence loss
of foetal loss by herd, age, repeated mating,
Variables
social status, number of foetuses and previous
Number of goats With foetal loss
Incidence of foetal loss (%) Without foetal loss
Herd no.a 1 2 3 4 5 6 7
17 6 10 10 15 14 I
56 58 64 80 63 45 70
23.3 (reference) 9.4* 13.5* 1l.P 19.2* 23.7 9.P
Age fyrs,” 1 2 3 4 5 6 t1
12 12 17 16 9 5 8
141 91 71 55 29 27 16
7.8 (reference) 11.0 19.3* 22.5* 23.7* 15.6 33.3*
Pregnancy achieved on heat number” 1 64 2 9 23 6
364 60 12
15&eference) 13.0 33.3*
Social stat& Low Medium High
13 49 17
34 345 57
27.7Creference) 12.4* 23.0*
Number offoetusesb 1 19 2 29 23 9
167 246 23
10.2 (reference) 10.5 28.1*
Previous foetal loss’ No 43 Yes 24
260 35
14.2Creference) 40.7*
“Information from 515 goats. bInformation from 493 goats. ‘Information from 362 goats. *Logistic regression revealed statistical significantly
different from reference
category
foetal
1.V. En&and
et al. /Animal Reproduction Science 49 ( 1997) 45-53
49
croscopy or culture from the maternal placental tissues, foetal stomach contents or organs. Out of the remaining 515 goats, 79 (15%) lost their foetuses from non-identified causes. Thirty-eight (48%) of the goats aborted, 33 (42%) delivered dead foetuses at full term and eight (10%) were found empty after having been identified as pregnant. In six of the goats that aborted, only bloody vaginal discharges were observed. No symptoms of disease such as depression or anorexia were seen in connection with reproductive loss. Out of 89 foetuses examined, 18 (20%) were expelled in a relatively fresh state whereas 71 (80%) had a decomposed or mummified appearance. In eight of the latter cases, the number of foetuses was difficult to estimate due to advanced decomposition or maceration. A significant association was found between foetal loss and herd (Table 1). Goats aged 3 yrs or over exhibited significantly greater incidence of foetal loss than those aged under 3 yrs (21.7% vs 9.2%, P < 0.001). In the goats that did not mate successfully before the 3rd or later heats, the incidence of foetal loss was significantly increased compared to those that became pregnant on an earlier heat (33.3% vs 9.2%, P < 0.05). With respect to social hierarchy, goats with a low social status exhibited a significantly greater incidence of foetal loss than those with high or medium status (27.7% vs 14.1%, P < 0.02). The incidence of foetal loss in goats that had experienced previous foetal loss
Table 2 Incidence
of foetal loss by goat index, milk production
Variables
With foetal loss Goat index b < 98 98-102 103-106 2 107
Incidence of foetal loss (%I’
563-676 677-799 2 800
15.5 17.2 18.4 22.7
78 74 73 70
12.4 19.6 18.9 23.1
19 21 15 12
57 88 85 65
25.0 19.3 15.0 15.6
52 26
310 106 20
14.4 19.7 4.8
18 17 21 at dr$ng
production
Horn statusC Dehorned Polled Horned
Without foetal loss
60 96 71 68
11 20 16 20
Milk production lust year fkglb < 563 11
Dail! milk < 0.8 0.8-1.3 I .4- 1.9 2 2.0
and presence of horns
Number of goats
off (kgjb
1
“Logistic regression revealed no statistical bInformation from 362 goats. “Information from 5 15 goats.
significant
differences
between the variable categories,
50
I.V. Engeland et al./Animl
Reproduction Science 49 (1997) 45-53
was significantly increased compared to the goats without such an experience (40.7% vs 14.2%, P < 0.001). In goats from which the number of foetuses could be recorded, animals pregnant with three or more foetuses exhibited a significantly greater incidence of foetal loss than in those with one or two foetuses (28.1% vs 10.4%, P < 0.01) (Table 1). Conditional logistic regression revealed that foetal loss was significantly higher in daughters of does that had suffered reproductive loss than in daughters of does without such a loss (P < 0.05). Multivariate logistic regression did not substantially alter the results. Models with interacting terms gave no better fit than models without. No association was found between foetal loss and the presence or absence of horns or dehorning or daily milk production at the time of dry-off. There was a trend for increasing incidence of foetal loss, although not statistically significant, with increasing goat index and milk production the last year (Table 2). The pedigree analysis did not reveal a different lineage in goats that experienced foetal loss compared with those with a normal pregnancy. Twenty-six bucks had mated the goats with foetal loss, and twenty-five of the bucks had also mated at least an equal number of goats with a normal pregnancy. The 26th buck was used only for a single goat that aborted.
4. Discussion The statistical analyses were performed on goats that showed foetal loss due to non-identified causes and on goats with a normal pregnancy. The exposure variables were mainly based on objective observations, and were recorded before the outcome of the pregnancy was known, to prevent biased misclassifications. Classification of social status was based on a subjective score performed by the farmer, and is not an ideal method. However, this is a commonly used and accepted approach in clinical studies. Since two of the seven herds (herd 2 and 7) had a history of previous low incidence of foetal loss (< 10%) while the incidence in the five other herds had been higher (> 15%), it was not surprising to find a significant association between foetal loss and herd during the investigated season (Table 1). It is possible that different environmental conditions could explain this difference in incidence of foetal loss as discussed by Engeland et al. (1997). No significant organisms were detected by serological examination of goats that experienced foetal loss or by microscopy or culture from the maternal placental tissues, foetal stomach contents or organs. The increased incidence of foetal loss among older goats supports the observation that the abortions were of a non-infectious nature. If infectious agents were involved, the rate of foetal loss should decrease with age due to acquired immunity. In South African Angora goats that aborted, Van Rensburg (1971) found an increase in the abortion rate with increasing age, due to altered adrenal status (hyperadrenalism) with increasing age. An immunological rejection of the foetus as demonstrated by Corbel (1972) in pregnant sheep infected in utero with Aspergillus
I. V. Engeland et al. /Animal Reproduction Science 49 (1997) 45-53
51
fumigatus, could account for repetitive abortions among older goats as discussed by Waldeland and Loken (19911. However, no evidence of such a rejection in Norwegian goats has been found (Waldeland and Loken, 199 1). The goats were mated on the first heat in the autumn, and only a few returned to service. The analyses revealed that if the goats conceived on the 3rd or later heat, this significantly increased the risk of foetal loss (Table 1). Various anomalies such as malformations, inferior luteal function or infections may be the cause of unsuccessful mating. Such abnormalities may also dispose the goat to foetal loss later during pregnancy. Low social status seemed to predispose the individual to foetal loss compared with high or medium status. It is possible that the basic reason for this loss could be chronic stress from social subordinance. In sheep, management procedures such as confinement, trucking: dipping and shearing have been found to increase the level of plasma cortisol (Kilgour and de Langen, 19701, which is commonly used as an indicator of stress. Experimental luteolysis has been obtained by prolonged cortisone administration in Angora goats (Wentzel and Roelofse, 19751 and with infusion of ACTH into goat foetuses (Thorbum et al., 1972). In South African Angora goats that aborted, Wentzel et al. (1975) found abnormally high maternal blood corticosteroid levels and extensive adrenal hyperplasia, which was suggested to create an unfavourable foetal environment inducing expulsion of foetuses which had an oedematous appearance. However, in the present study a large proportion of the foetuses were expelled in an decomposed or mummified condition. To investigate whether social stress may be involved in foetal loss, plasma levels of corticosteroids in goats under various conditions should be determined. Studies to clarify the genetic influence on the cortisol level should also be initiated. In stressed pregnant rats, an increased incidence of abortion in the offspring has been observed (Herrenkohl, 1979). In addition to a potential genetic influence, it is therefore also possible that an adverse environmental effect on the mother may have been passed on to the offspring and affected foetal survival in the daughter. In goats with three or more foetuses foetal loss occurred more often than in goats with one or two foetuses. It is possible that nutrition of each foetus via the placenta is reduced proportionally with increasing number, and that placental insufficiency in such cases may lead to foetal death (Matthews, 1991). Social rank was probably of little importance for nutrition of the does because roughages were given ad libitum, and the animals were fed equal rations of concentrated when the goats were fixed in feeding yokes. However, it is possible that both an elevated number of fertilised ova and multiple pregnancy may be associated with high social status and optimal body condition, and that stress from maintaining high social rank may also have an adverse effect. Data from the present study were analysed in different ways to investigate whether there was a possible genetic influence on the foetal loss. The possibility that some cases of previous loss in both the goats examined and their mothers may have been caused by infections and still included in these analyses cannot be excluded, since extensive examinations of aborted materials and serological examinations were not performed then. However, the chances are small and this would probably have little influence on the results, since both the present and other recent studies indicate that infections are
52
I.V. Engeland et al. /Animal
Reproduction Science 49 (1997) 45-53
found in only few cases of abortion in goats (Melby et al., 1986; Loken, 1990; Waldeland and Loken, 1991; Engeland et al., 1997). Although the pedigree analysis comprising the goat’s sire, dam and dam’s sire did not provide evidence of a genetic influence, there was a significant association between mother and daughter in the occurrence of reproductive problems (P < 0.05). Previous foetal loss was also associated with present loss (P < 0.05). These results suggest that there may be a hereditary predisposition for foetal loss. It may therefore be worthwhile to cull affected animals and to exclude their previous offspring from the breeding stock. Under natural conditions, a hereditary tendency of abortion may be self limiting. It is, however, possible that the farmers’ selection of breeding goats may have counteracted this, such as in the Angora goat where a hereditary predisposition to abortion has been found to be associated with production and quality of mohair (Van Rensburg, 1971). It should be noted that in the present work a tendency of increasing incidence of foetal loss with increasing goat index and milk production was found (Table 2). However, the power in detecting statistically significant associations between foetal loss and milk production in the present study is relatively small due to restricted number of animals. Therefore, an epidemiological study should be undertaken on a larger scale to elucidate whether there is a similar relationship between milk production and foetal loss as between mohair production and abortion in Angora goats. In the present study, the milk production potential of each goat was not fully demonstrated because of limitation imposed by the production quota allocated to each herd, and a possible stronger association between production and foetal loss may thereby have been concealed. 5. Conclusion
Advanced age, difficulty in conceiving, low social status, pregnancy with 2 3 foetuses and previous foetal loss were significantly associated with current loss. Foetal loss were significantly higher in daughters of does that had suffered reproductive loss themselves than in daughters of does without such a loss. This association suggests a hereditary influence on the occurrence of foetal loss. Goats with a history of foetal loss from non-identified causes or with a mother with such a loss should therefore be culled. The goat breeding index, milk production last year, daily milk production at the time of dry-off and the presence or absence of horns or dehorning were not associated with foetal loss. Acknowledgements
Financial support was received from the Research Council of Norway. References Corbel, M.J., 1972. Production of antibodies to placental antigens by pregnant ewes experimentally infected with Aspergillusfumigatus. Br. Vet. J. 128, xliv-xlvi. Engeland, I.V., Waldeland, H., Andresen, $3.. Llen, T., Bjcrkman, C., Bjerkls, I., 1997. Foetal loss in dairy goats: an epidemiological study in 22 herds. Small Rumin. Res. (in press).
I. V. Engeland
Herrenkohl,
L.R.,
1979. Prenatal
et al. /Animal
stress reduces
Reproduction
fertility
Science 49 119Y7145-53
and fecundity
in female
offspring.
53
Science
206,
1097- 1099.
Kilgour, R., de Langen, H., 1970. Stress in sheep resulting from management practices. Proc. N.Z. Sot. Anim. Prod. 30, 65-76. Loken. T., 1990. Pestivirus infection in Norway. Epidemiological studies in goats. .I. Comp. Pathol. 103, I-JO. Matthews, J.G., 1991. Abortion. In: Matthews, J.G. (Ed.), Outline of Clinical Diagnosis in the Goat. Writh. Oxford. pp. 17-32. Melby, H.P., Aursjo, J., Binde, M., Gronstol. H., 1986. Diseases and mortality in 27 dairy goat farms. Nord. Vet. Med. 38, 403-41 1. Thorburn, G.D., Nicol, D.H., Bassett, J.M., Shutt, D.A., Cox, R.I., 1972. Parturition in the goat and sheep: changes in corticosteroids, progesterone, oestrogens and prostaglandin. F. J. Reprod. Fertil. 16, 61-84. Suppl. Van Rensburg, S.J., 1971. Reproductive physiology and endocrinology of normal and habitual aborting Angora goats. Onderstepoort J. Vet. Rea. 38, I-62. Waldeland, H., Loken, T., 1991. Reproductive failure in goats in Norway: an investigation in 24 herds. Acta Vet. Stand. 32, 535-541. Wentzel, D., Roelofse, C.S., 1975. The habitually aborting Angora doe: VII. Induction of abortion by administration of cortisone acetate. Agroanimalia 7, 45-48. Wentzel, D., Morgenthal. J.C., van Niekerk, C.H., 1975. The habitually aborting Angora doe: IV. Adrenal function in normal and aborter does. Agroanimalia 7, 27-34.