Milk Progesterone to Predict Reproductive Status in a Commercial Dairy Herd I N A N C Y M. C O X 2 ,4, F. N. T H O M P S O N 2, and D. H. C U L V E R 3 University of Georgia Athens 30602 ABSTRACT
and remains elevated in the pregnant cow (7, 11, 22). Progesterone dropped within 24 h after removal o f the CL (14). Several studies have reported large differences in P between pregnant and nonpregnant cows at approximately 20 days after insemination (7, 12, 15, 17). Investigators have been more successful in identifying the nonpregnant cow than the pregnant cow via milk P (7, 8, 12, 15). This difference was due primarily to pathological changes in the uterus and embryonic loss that confounded assessments of pregnancy by causing elevated P. Our objectives were to evaluate P as an indicator of CL function in a commercial dairy herd and to obtain additional information on the incorrect positive diagnoses of pregnancy.
Progesterone was measured in milk strippings of cows to evaluate potential diagnosis of pregnancy and to estimate embryonic mortality. Progesterone in milk was measured at 20 days and at lO-day intervals thereafter through 90 days after artificial insemination when the cows were palpated for pregnancy. Mean progesterone for 24 nonpregnant cows in estrus was 4.5 ng/ml at 20 days after insemination. Cows were classified as pregnant or nonpregnant by a formula which provided a progesterone concentration for which the probability of nonpregnancy was .8. At 20 days after insemination, the success in identifying the nonpregnant and pregnant cows was 87% and 67%. Ten of 34 nonpregnant cows with low (nonpregnant range) progesterone in milk on day 20 were not seen in estrus. Progesterone at sampling intervals after 20 days was used to investigate incorrect assessments. Seventeen nonpregnant cows had progesterone indicative of pregnancy at day 20. Progesterone in milk decreased at various later sampling intervals in all 17 cows. Two of these cows were diagnosed as having pathology of their reproductive tracts. If the other 15 cows experienced death of their embryos, the frequency of embryonic mortality was 28%.
MATERIALS AND METHODS
INTRODUCTION
Progesterone (P) in milk reflects cyclic changes in activity of the corpus luteum (CL)
Received February 15, 1978. 1Journal Paper #1481, College of Veterinary Medicine, University of Georgia, Athens 30602. 2 Department of Physiology-Pharmacology. 3 Department of Statistics and Computer Science. * Department of Animal and Dairy Science. 1978 J Dairy Sci 61:1616-1621
Individual samples of milk strippings were obtained from 91 lactating Holstein cows in a commercial dairy herd and stored at 4 C without preservatives for not more than 5 days before assay. The cows were kept on pasture with housing available and were checked for estrus twice daily. Cows more than 60 days postpartum that had normal genital tracts were inseminated artificially at every estrus. Pregnancy was diagnosed by rectal palpation at 90 days after breeding. Immunoreactive P was quantified as described by Hoffman and Hamburger (14) except for the antiserum and the use of a petroleum ether-diethyl ether mixture (80:20) for extraction. Briefly, 100 or 200 ul of milk taken from a milk sample that had been mixed well on a metabolic shaker and was at room temperature were extracted with petroleum ether. The ether extract was incubated with methanol at - 5 C for 12 h followed by partitioning between petroleum ether and methanol. The antiprogesterone serum was $49#6 from G. E. Abraham (2). We found a 40% crossreactivity with 17~-hydroxyprogesterone for this antiserum. The only other
1616
MILK PROGESTERONE DURING PREGNANCY steroids that have crossreacted substantially with this antiserum are ll-desoxycortisol (90%), ll-desoxycorticosterone (35%), and cortisol (6.4%) (2). The radiolabeled P was [1,2,6,7-3H]p (New England Nuclear, Boston, MA) with a specific activity of 105 Ci/mM. Percentage binding of radiolabeled P (50 pg) to the antiserum was 47%, and 25 pg of unlabeled P caused reduction (P.05) between values before and after column chromatography. Before-column P values were 1.1 and 7.9 ng/ml and after-column P values (isoctane fraction) were 1.0 to 7.3 ng/ml for low and high P pools after correction to 100% recovery. In 11 of 13 instances, precolumn values were higher than postcolumn values. Mean recovery of radiolabeled P after column separation was 76.5 -+ 2% (-+ SE). Determinations of total milk fat on some of the milk samples were by the Babcock method
1617
(10). Progesterone as considered representative of nonpregnancy if it was confirmed by observation of the cows in estrus within + 3 days of day 20 sampling and -+ 5 days of sampling at the later 10-day intervals. "Pregnant" P was so classified if the cows were pregnant by palpation at 90 days after breeding. Student's t tests for pooled and paired samples were used where appropriate to compare sample means. Analysis of variance tested parallelism of standard curves for the assays. Pearson correlation coefficients were calculated for P for assay validation (21). R ESU LTS
Concentration of P, ~ + SE (n), of cows observed in estrus within 20 + 3 days of breeding and classified nonpregnant was 4.5 + .7 (24) ng/ml, with a range of .5 to 7.6 ng/ml. For cows which were pregnant at 90 days, mean -+ SE (n) 20-day P concentration was 27.4 + 2.6 (40) ng/ml with a range of 3.6 to 46.1 ng/ml. These P concentrations at 20 days after insemination for the pregnant and nonpregnant cows were significantly different (P
and remains elevated in the pregnant cow (7, 11, 22). Progesterone dropped within 24 h after removal o f the CL (14). Several studies have reported large differences in P between pregnant and nonpregnant cows at approximately 20 days after insemination (7, 12, 15, 17). Investigators have been more successful in identifying the nonpregnant cow than the pregnant cow via milk P (7, 8, 12, 15). This difference was due primarily to pathological changes in the uterus and embryonic loss that confounded assessments of pregnancy by causing elevated P. Our objectives were to evaluate P as an indicator of CL function in a commercial dairy herd and to obtain additional information on the incorrect positive diagnoses of pregnancy.
Progesterone was measured in milk strippings of cows to evaluate potential diagnosis of pregnancy and to estimate embryonic mortality. Progesterone in milk was measured at 20 days and at lO-day intervals thereafter through 90 days after artificial insemination when the cows were palpated for pregnancy. Mean progesterone for 24 nonpregnant cows in estrus was 4.5 ng/ml at 20 days after insemination. Cows were classified as pregnant or nonpregnant by a formula which provided a progesterone concentration for which the probability of nonpregnancy was .8. At 20 days after insemination, the success in identifying the nonpregnant and pregnant cows was 87% and 67%. Ten of 34 nonpregnant cows with low (nonpregnant range) progesterone in milk on day 20 were not seen in estrus. Progesterone at sampling intervals after 20 days was used to investigate incorrect assessments. Seventeen nonpregnant cows had progesterone indicative of pregnancy at day 20. Progesterone in milk decreased at various later sampling intervals in all 17 cows. Two of these cows were diagnosed as having pathology of their reproductive tracts. If the other 15 cows experienced death of their embryos, the frequency of embryonic mortality was 28%.
MATERIALS AND METHODS
INTRODUCTION
Progesterone (P) in milk reflects cyclic changes in activity of the corpus luteum (CL)
Received February 15, 1978. 1Journal Paper #1481, College of Veterinary Medicine, University of Georgia, Athens 30602. 2 Department of Physiology-Pharmacology. 3 Department of Statistics and Computer Science. * Department of Animal and Dairy Science. 1978 J Dairy Sci 61:1616-1621
Individual samples of milk strippings were obtained from 91 lactating Holstein cows in a commercial dairy herd and stored at 4 C without preservatives for not more than 5 days before assay. The cows were kept on pasture with housing available and were checked for estrus twice daily. Cows more than 60 days postpartum that had normal genital tracts were inseminated artificially at every estrus. Pregnancy was diagnosed by rectal palpation at 90 days after breeding. Immunoreactive P was quantified as described by Hoffman and Hamburger (14) except for the antiserum and the use of a petroleum ether-diethyl ether mixture (80:20) for extraction. Briefly, 100 or 200 ul of milk taken from a milk sample that had been mixed well on a metabolic shaker and was at room temperature were extracted with petroleum ether. The ether extract was incubated with methanol at - 5 C for 12 h followed by partitioning between petroleum ether and methanol. The antiprogesterone serum was $49#6 from G. E. Abraham (2). We found a 40% crossreactivity with 17~-hydroxyprogesterone for this antiserum. The only other
1616
MILK PROGESTERONE DURING PREGNANCY steroids that have crossreacted substantially with this antiserum are ll-desoxycortisol (90%), ll-desoxycorticosterone (35%), and cortisol (6.4%) (2). The radiolabeled P was [1,2,6,7-3H]p (New England Nuclear, Boston, MA) with a specific activity of 105 Ci/mM. Percentage binding of radiolabeled P (50 pg) to the antiserum was 47%, and 25 pg of unlabeled P caused reduction (P.05) between values before and after column chromatography. Before-column P values were 1.1 and 7.9 ng/ml and after-column P values (isoctane fraction) were 1.0 to 7.3 ng/ml for low and high P pools after correction to 100% recovery. In 11 of 13 instances, precolumn values were higher than postcolumn values. Mean recovery of radiolabeled P after column separation was 76.5 -+ 2% (-+ SE). Determinations of total milk fat on some of the milk samples were by the Babcock method
1617
(10). Progesterone as considered representative of nonpregnancy if it was confirmed by observation of the cows in estrus within + 3 days of day 20 sampling and -+ 5 days of sampling at the later 10-day intervals. "Pregnant" P was so classified if the cows were pregnant by palpation at 90 days after breeding. Student's t tests for pooled and paired samples were used where appropriate to compare sample means. Analysis of variance tested parallelism of standard curves for the assays. Pearson correlation coefficients were calculated for P for assay validation (21). R ESU LTS
Concentration of P, ~ + SE (n), of cows observed in estrus within 20 + 3 days of breeding and classified nonpregnant was 4.5 + .7 (24) ng/ml, with a range of .5 to 7.6 ng/ml. For cows which were pregnant at 90 days, mean -+ SE (n) 20-day P concentration was 27.4 + 2.6 (40) ng/ml with a range of 3.6 to 46.1 ng/ml. These P concentrations at 20 days after insemination for the pregnant and nonpregnant cows were significantly different (P
