2014 US ReSincr - Diag gestação

Prévia do material em texto

4932
J. Dairy Sci. 97 :4932–4941
http://dx.doi.org/ 10.3168/jds.2013-7887 
© american Dairy Science association®, 2014 .
 aBStraCt 
 The aim of this study was to compare the reproduc-
tive performance of dairy cows subjected to early (ER) 
or late (LR) resynchronization programs after nonpreg-
nancy diagnoses based on either pregnancy-associated 
glycoproteins (PAG) ELISA or transrectal palpation, 
respectively. In addition, the accuracy of the PAG 
ELISA for early pregnancy diagnosis was assessed. 
Lactating Holstein cows were subjected to a Presynch-
Ovsynch protocol with timed artificial insemination 
(AI) performed between 61 and 74 DIM. On the day 
of the first postpartum AI, 1,093 cows were blocked by 
parity and assigned randomly to treatments; however, 
because of attrition, 452 ER and 520 LR cows were 
considered for the statistical analyses. After the first 
postpartum AI, cows were observed daily for signs of 
estrus and inseminated on the same day of detected 
estrus. Cows from ER that were not reinseminated in 
estrus received the first GnRH injection of the Ovsynch 
protocol for resynchronization 2 d before pregnancy 
diagnosis. On d 28 after the previous AI (d 27 to 34), 
pregnancy status was determined by PAG ELISA, and 
nonpregnant cows continued on the Ovsynch protocol 
for reinsemination. Pregnant cows had pregnancy status 
reconfirmed on d 46 after AI (d 35 to 52) by transrectal 
palpation, and those that lost the pregnancies were 
resynchronized. Cows assigned to LR had pregnancy 
diagnosed by transrectal palpation on d 46 after AI (d 
35 to 52) and nonpregnant cows were resynchronized 
with the Ovsynch protocol. Blood was sampled on d 
28 after AI (d 27 to 34) from cows in both treatments 
that had not been reinseminated on estrus and again 
on d 46 after AI (d 35 to 52) for assessment of PAG 
ELISA to determine the accuracy of the test. Cows 
were subjected to treatments for 72 d after the first 
insemination. Pregnancy per AI (P/AI) at first post-
partum timed AI did not differ between treatments and 
averaged 28.9%. The proportion of nonpregnant cows 
that were resynchronized and received timed AI was 
greater for ER than for LR (30.0 vs. 7.6%). Cows in 
ER had a shorter interval between inseminations when 
inseminated following spontaneous estrus (21.7 ± 1.1 
vs. 27.8 ± 0.8 d) or after timed AI (35.3 ± 1.2 vs. 
55.2 ± 1.4 d). Nevertheless, the ER did not affect the 
rate of pregnancy (adjusted hazard ratio = 1.23; 95% 
confidence interval = 0.94 to 1.61) or the median days 
postpartum to pregnancy (ER = 132 vs. LR = 140). 
A total of 2,129 PAG ELISA were evaluated. Overall, 
sensitivity, specificity, and positive and negative predic-
tive values averaged 95.1, 89.0, 90.1, and 94.5%, respec-
tively, and the accuracy was 92.1%. In conclusion, PAG 
ELISA for early diagnosis of pregnancy had acceptable 
accuracy, but early resynchronization after nonpreg-
nancy diagnosis with PAG ELISA did not improve the 
rate of pregnancy or reduce days open in dairy cows 
continuously observed for estrus. 
 Key words: dairy cow , pregnancy-associated glyco-
protein , reproduction , resynchronization 
 IntrODuCtIOn 
 Reproductive performance is a major component of 
the economic viability in dairy herds and shortening 
the reinsemination interval for nonpregnant cows is 
expected to reduce the time to pregnancy and improve 
the pregnancy rate (Meadows et al., 2005; De Vries, 
2006; Ribeiro et al., 2012). From 55 to 70% of lactat-
ing dairy cows fail to become pregnant in response to 
AI and need to be reinseminated (Santos et al., 2004). 
Programs for synchronization of the estrous cycle and 
timed AI have been widely used to increase submission 
to AI and to reduce the interval between inseminations 
(Fricke et al., 2003). However, nonpregnant cows need 
to be identified before being subjected to resynchroni-
zation protocols and the development of methods for 
early pregnancy diagnosis is critical to further reduce 
the interbreeding interval. 
 The diagnosis of pregnancy in cattle is traditionally 
performed starting around d 35 to 40 after insemina-
tion by transrectal palpation of the reproductive tract 
for identification of the amniotic vesicle (Warnick et 
 effect of early or late resynchronization based on different methods 
of pregnancy diagnosis on reproductive performance of dairy cows 
 L. D. p. Sinedino ,* F. S. Lima ,* r. S. Bisinotto ,* r. L. a. Cerri ,† and J. e. p. Santos *1
 * Department of animal Sciences, university of florida, gainesville 32611 
 † faculty of Land and food Systems, university of British Columbia, Vancouver, BC Vt6 1Z4, Canada 
 
 
 Received December 30, 2013.
 Accepted May 5, 2014.
 1 Corresponding author: jepsantos@ufl.edu 
Journal of Dairy Science Vol. 97 No. 8, 2014
ReSYNCHRoNIZatIoN PRogRaMS foR DaIRY CoWS 4933
al., 1995), or at approximately 27 d after insemination 
with the aid of ultrasonography (Silva et al., 2007). Al-
ternatively, chemical methods that detect the presence 
of conceptus-derived antigens can be used to identify 
pregnant cows early after AI (Green et al., 2009, 2011). 
Modern pregnancy-associated glycoproteins (PAG) 
are members of a large family of aspartic proteinases 
expressed by binucleate trophoblast cells in even-toed 
ungulates (Xie et al., 1991). These proteins are released 
into the uterine stroma after the fusion between bi-
nucleate and endometrial cells (Xie et al., 1991) and 
can be observed in the maternal circulation through-
out gestation (Sasser et al., 1986; Green et al., 2005). 
The development of monoclonal antibodies specific for 
PAG produced during initial stages of gestation have 
allowed for the use of immunoassays such as ELISA 
for early pregnancy detection that can be performed in 
cows after 60 d postpartum (Silva et al., 2007; Green 
et al., 2009). Nevertheless, the concentrations of PAG 
in the blood of pregnant cows increase from d 24 to 30 
after AI and then decrease toward d 60 of gestation 
(Thompson et al., 2010), suggesting that the accuracy 
of determining pregnancy status varies depending on 
the day after AI in which the blood is collected (Szenci 
et al., 1998; Green et al., 2009).
Previous studies have shown that diagnosing preg-
nancy starting on d 25 after AI based on PAG concen-
trations resulted in acceptable accuracy (Silva et al., 
2007; Green et al., 2009) and reduced interbreeding 
interval and time to pregnancy in lactating cows re-
synchronized exclusively using timed AI (Silva et al., 
2009). Nonetheless, the combination of estrous detec-
tion and timed AI for reinsemination of nonpregnant 
cows is used in the majority of dairy herds, as it often 
results in greater submission to AI and reduced time to 
pregnancy (Ribeiro et al., 2012; Giordano et al., 2013). 
A large portion of cows are expected to be observed in 
estrus and reinseminated before d 32 after insemination 
(Chebel et al., 2003; Bartolome et al., 2005; Galvão 
et al., 2007); therefore, the need for early pregnancy 
diagnosis will vary according to the ability to identify 
nonpregnant cows that return to estrus before any 
pregnancy diagnosis testing.
The main hypothesis of the present study was that 
an early resynchronization (ER) program starting at 
28 d after AI using the PAG ELISA for pregnancy di-
agnosis and decision on resynchronization would reduce 
the interval between inseminations and, therefore, the 
time to pregnancy compared with the use of transrectal 
palpation in a late resynchronization protocol start-
ing 46 d after AI. A second hypothesis was that the 
PAG ELISA would result in acceptable accuracy to 
determine pregnancy status in lactating dairy cows. 
Therefore, the objectives were to compare the time to 
pregnancy in cows subjected to ER versus late resyn-
chronization (LR) programs based on nonpregnancy 
diagnosis using PAG ELISA or transrectal palpation, 
respectively, and to evaluate the accuracy of the PAG 
ELISA for pregnancy diagnosis.materIaLS anD metHODS
Animals, Housing, and Feeding
A total of 1,093 lactating Holstein cows from a com-
mercial dairy herd located in central California were en-
rolled in this experiment from December 2004 to April 
2005. Primiparous and multiparous cows were housed 
separately in freestall barns equipped with sprinklers 
and fans. Cows had ad libitum access to water and were 
fed a TMR to meet or exceed the requirements of lac-
tating Holstein cows weighing 650 kg and producing 45 
kg of milk per day with 3.5% fat and 3.2% true protein 
(NRC, 2001). Cows were fed multiple times daily and 
were milked 3 times daily.
Reproductive Management and Treatments
Weekly cohorts of cows were subjected to the 
Presynch-Ovsynch protocol (Moreira et al., 2001) for 
the first postpartum timed AI (Figure 1). Briefly, cows 
received an i.m. injection of PGF2α (Lutalyse sterile 
solution, 25 mg of dinoprost as tromethamine salt, Zo-
etis, Madison, NJ) given on d 32 (25 to 38) and again 
on d 46 (39 to 52) postpartum. The Ovsynch protocol 
was initiated at 58 (51 to 64) DIM with an i.m. injec-
tion of GnRH (Cystorelin, 100 μg of gonadorelin di-
acetate tetrahydrate; Merial Ltd., Duluth, GA), which 
was followed 7 d later by an injection of PGF2α and a 
final GnRH injection of PGF2α 48 h later. Cows were 
inseminated 24 h after the last GnRH injection (61 to 
74 DIM). The body condition of cows was scored using 
a 1 to 5 scale (0 = emaciated and 5 = obese; Ferguson 
et al., 1994) at 54 ± 3 DIM and those with BCS >2.00 
were considered eligible for enrollment in the study. For 
purposes of statistical analyses, cows were categorized 
as having low (≤2.75) or moderate (≥3.00) BCS. The 
study followed a randomized complete block design. On 
the day of the first AI postpartum, cows were blocked 
by parity and, within each block, assigned randomly 
to either an ER (n = 548) or LR (n = 545) program 
(Figure 1). After the first AI postpartum, cows from 
both programs were observed for signs of estrus based 
on removal of tail chalk and those in estrus were in-
seminated on the same day.
Cows assigned to ER had pregnancy diagnosed week-
ly and they received the first GnRH injection of the 
Ovsynch protocol for resynchronization, on average, on 
4934 SINeDINo et aL.
Journal of Dairy Science Vol. 97 No. 8, 2014
d 26 after the previous AI. Two days later, on d 28 after 
AI (d 27 to 34), pregnancy status was determined by 
PAG ELISA. Ultrasonography was performed on the 
same day of blood collection to serve as the gold stan-
dard for PAG accuracy determination. Because cows 
in ER were evaluated weekly, it was expected that all 
would have the blood test performed between 27 and 
32 d after AI. However, a small number of cows had to 
have a second blood sample collected 2 d later because 
of disagreement between the PAG ELISA and the ul-
trasonography. Blood was collected from the coccygeal 
vessels into 3 mL evacuated tubes containing K2 EDTA 
(Vacutainer; Becton Dickinson, Franklin Lakes, NJ). 
Tubes were placed on ice immediately after collection 
and shipped to Monsanto Co. (St. Louis, MO) using 
an overnight shipping courier (FedEx Corp., Memphis, 
TN). Cows diagnosed as nonpregnant based on the 
PAG assay received the remaining injections of the 
Ovsynch protocol and were reinseminated 8 d after the 
blood testing. Conversely, cows diagnosed as pregnant 
according to the PAG assay did not receive any further 
treatment and had pregnancy status reconfirmed on d 
46 after AI (d 35 to 52) by transrectal palpation of 
the reproductive tract and its contents. Throughout the 
manuscript, the first pregnancy diagnosis will be de-
fined as on d 28, which refers to d 27 to 34 after AI, and 
the final pregnancy diagnosis will be defined as on d 46, 
which refers to d 35 to 52 after insemination. Cows with 
pregnancies reconfirmed completed the study, whereas 
cows that lost the pregnancy were resynchronized with 
the Ovsynch protocol.
Pregnancy diagnosis in LR cows was performed every 
14 d by transrectal palpation of the uterus and con-
tents. Because a few cows did not have a definitive di-
agnosis, some were reevaluated a few days later, which 
resulted in an interval that was longer than 14 d (d 35 
to 52 after the previous AI). The 14-d interval reflected 
the management already implemented on the farm in 
which half of the herd was evaluated for pregnancy in 
one week, and the other half in the subsequent week. 
Throughout the manuscript, the day of pregnancy di-
agnosis for LR cows is defined as d 46, which refers 
to the average day for this treatment. Cows diagnosed 
as pregnant completed the study, whereas nonpregnant 
cows were resynchronized with the Ovsynch protocol 
starting on the day of nonpregnancy diagnosis. The 
study lasted 72 d from the first AI. A cow was consid-
ered pregnant for the study in both treatments based 
Figure 1. Diagram of the study. All cows underwent a Presynch-Ovsynch protocol and received the first AI at 68 DIM (range: 61–74 DIM; 
study d 0). Early Resynch: cows were subjected to early pregnancy diagnosis by pregnancy-associated glycoprotein (PAG) ELISA on d 28 af-
ter AI (27 to 34 d), and nonpregnant cows continued on the resynchronization protocol. Late Resynch: cows were subjected to late pregnancy 
diagnosis by transrectal palpation on d 46 after AI (35 to 52 d) and nonpregnant cows received the resynchronization protocol starting on the 
day of pregnancy diagnosis. Cows observed in estrus were inseminated on the same day. The study lasted 72 d from the first insemination. Cows 
were considered pregnant at the end of the study based on palpation performed on d 46 after AI. BS = blood sample collected for analysis of 
concentrations of progesterone; G = injection of gonadotropin-releasing hormone; PAG = PAG ELISA pregnancy diagnosis; PG = injection of 
PGF2α; TAI = timed AI; ○ = blood sample collected for measurements of progesterone concentrations in plasma; ● = blood sample collected 
for analysis of PAG by ELISA.
Journal of Dairy Science Vol. 97 No. 8, 2014
ReSYNCHRoNIZatIoN PRogRaMS foR DaIRY CoWS 4935
on the results of transrectal palpation performed on d 
46 after AI.
Progesterone Concentrations and Definition 
of Estrous Cyclic Status
Estrous cyclic status was defined based on the concen-
tration of progesterone in plasma collected on the day 
of the second PGF2α injection of the presynchronization 
(46 DIM) and the day of the first GnRH injection of 
the Ovsynch protocol (58 DIM). Blood was sampled by 
puncture of the coccygeal vessels into 10-mL evacuated 
tubes containing K2 EDTA (Vacutainer; Becton Dick-
inson). Samples were immediately placed on ice and 
arrived at the laboratory within 6 h of collection. Blood 
tubes were centrifuged at 3,000 × g for 15 min at 4°C 
for plasma separation. Plasma samples were frozen at 
−25°C until assayed. The concentration of progesterone 
was determined by an ELISA validated by Cerri et al. 
(2004); samples were analyzed in duplicate, with every 
microplate containing the standards 0.1, 0.2, 0.5, 1.0, 
2.0, 5.0, 10.0, and 20.0 ng/mL. The coefficient of varia-
tion was assessed with 2 known samples containing 1.5 
and 3.0 ng/mL. Intraassay and interassay coefficients of 
variation for all microplates were 5.7 and 9.1%, respec-
tively. Individual samples with a coefficient of variation 
greater than 15% were reanalyzed.
Cows were classified as estrous cyclic when the con-
centration of progesterone was ≥1 ng/mL in at least 
1 of the samples. Conversely, cows with progesterone 
concentrationpH 9.35) and incubated overnight at 4°C. On the next 
day, plates were washed 4 times with 200 μL of washing 
buffer (PBS with 0.05% Tween 20; pH 7.40) and then 
200 μL of blocking solution was added to each well. 
Plates were incubated for 1 h at 37°C, and the blocking 
solution was removed and wells washed 4 times with 
300 μL of washing buffer. After the last wash, 100 μL 
of either a standard with known PAG concentration or 
an unknown plasma sample was added to the plate in 
duplicate. Blocking buffer was used as the blank. Plates 
were incubated at 37°C for 1 h and then washed 4 times 
with 300 μL of washing buffer. One-hundred microliters 
of biotin-labeled PAG monoclonal antibody diluted in 
blocking buffer was added to each well, plates incu-
bated for 1 h at 37°C, and then wells washed 4 times 
with 300 μL of washing buffer. One-hundred microliters 
of horseradish peroxidase substrate solution was added 
to each well and plates were incubated at room tem-
perature for 15 min under constant shaking. All wells 
were added with 100 μL of a 1 M HCl solution to stop 
color development and absorbance was assessed using 
a spectrometer (SpectraMax Plus Microplate Reader; 
MDS Analytical Technologies, Sunnyvale, CA). Con-
centrations of PAG were regressed based on the stan-
dard curve of each plate in accordance to the software 
provided by the plate reader manufacturer.
The thresholds for classifying cows as being preg-
nant as well as the PAG concentrations for individual 
cows were proprietary to Monsanto Co. and only the 
pregnancy results were delivered to the research team. 
Pregnancy outcomes were received approximately 36 
h after blood samples were collected (3.5 d before the 
day of the injection of PGF2α of the Ovsynch protocol 
in nonpregnant cows). All the results from the ELISA 
had a concurrent transrectal ultrasonography (Sonovet 
2000, 7.5 MHz linear probe; Universal Medical Systems 
Inc., Bedford Hills, NY) performed on the same day. 
In case of disagreement between the PAG ELISA and 
ultrasonography, a second blood sample and ultrasound 
exam was performed 48 h after the initial test.
Assessment of Accuracy of the PAG ELISA 
as a Diagnostic Test for Pregnancy
In addition to the blood sample collected from ER 
cows on d 28 after AI, concentrations of PAG in plasma 
were also measured on d 28 in all LR cows that had 
not returned to estrus and in all cows on d 46 after 
AI in both experimental groups to generate additional 
data for the assessment of PAG ELISA accuracy. The 
pregnancy outcomes by transrectal ultrasonography or 
palpation of the reproductive tract performed concur-
rently with blood samplings served as a gold standard 
to test the accuracy of the PAG ELISA. In case of 
disagreement between the PAG ELISA and ultrasonog-
raphy, a second ultrasound examination was performed 
within 2 d of the first test to serve as final pregnancy 
outcome and gold standard for the chemical test.
The reliability of the PAG ELISA for pregnancy 
diagnosis was assessed with regard to its sensitivity 
(Se), specificity (Sp), positive predictive value (PPV), 
negative predictive value (NPV), and accuracy in 
comparison with transrectal ultrasonography or palpa-
tion of the reproductive tract. The Se of the assay was 
expressed as the proportion of pregnant cows detected 
as pregnant by the PAG ELISA [(number of pregnant 
4936 SINeDINo et aL.
Journal of Dairy Science Vol. 97 No. 8, 2014
cows correctly diagnosed by PAG ELISA/number of 
pregnant cows diagnosed by ultrasonography or palpa-
tion) × 100]. The Sp was calculated as the proportion 
of nonpregnant cows identified as nonpregnant by the 
assay [(number of nonpregnant cows correctly diag-
nosed by PAG ELISA/number of nonpregnant cows di-
agnosed by ultrasonography or palpation) × 100]. The 
PPV was calculated as the proportion of cows testing 
pregnant that were truly pregnant [(number of cows 
correctly testing pregnant/number of total pregnant 
cows by PAG ELISA) × 100], whereas the NPV was 
calculated as the proportion of cows testing nonpreg-
nant that were truly nonpregnant [(number of non-
pregnant cows correctly testing nonpregnant/number 
of total nonpregnant cows diagnosed by PAG ELISA) 
× 100]. Test accuracy was defined as the proportion of 
correct outcomes (pregnant and nonpregnant) provided 
by the assay [(number of correct PAG ELISA tests/
number of total tests performed) × 100].
Statistical Analysis
Continuous data, such as the interval between in-
seminations, were analyzed by ANOVA using PROC 
GLIMMIX of SAS (version 9.3; SAS Institute Inc., 
Cary, NC), fitting a normal distribution. Models in-
cluded the fixed effects of treatment (ER vs. LR), par-
ity, type of AI (timed AI vs. insemination on estrus), 
estrous cyclic status (cyclic vs. anovular), BCS at study 
enrollment (≤2.75 vs. ≥3.00), AI number, and inter-
actions between treatment and type of AI, treatment 
and parity, and the random effect of cow nested within 
treatment and block.
Binary data were analyzed by logistic regression us-
ing PROC GLIMMIX of SAS. Initial models included 
the fixed effects of treatment, parity, BCS at study 
enrollment, estrous cyclic status, and the interactions 
between treatment and parity, treatment and BCS, and 
treatment and cyclic status, and the random effect of 
block. A backward stepwise elimination method was 
used to continuously remove independent variables with 
P > 0.10. Treatment was forced in all final models.
Time to pregnancy was analyzed by the Cox propor-
tional hazard model with PROC PHREG of SAS. The 
time variable considered was the number of days be-
tween the first timed AI and the day of AI that resulted 
in pregnancy or the day in which cows were censored 
from the analyses (i.e., cows that were sold, died, or 
remained nonpregnant until reaching the end of study). 
Models included the fixed effects of treatment, parity, 
BCS, estrous cyclic status, and the interactions be-
tween treatment and covariates. A backward stepwise 
elimination method was used to continuously remove 
independent variables with P > 0.10 and treatment was 
forced in all final models. The median and mean days 
to pregnancy were obtained from PROC LIFETEST 
of SAS. A survival plot was generated with MedCalc 
software (version 12.7.7.0; MedCalc Software bvba, 
Mariakerke, Belgium).
Estimates between the agreement of pregnancy 
outcomes of transrectal ultrasonography or palpation 
and PAG ELISA were calculated by kappa statistic 
in PROC FREQ of SAS and 95% confidence intervals 
were determined. Treatment differences with P ≤ 0.05 
were considered significant and those with 0.05 0.30) between ER 
(35.5% and 2.23 ± 0.06) and LR groups (38.2% and 
2.16 ± 0.06). Similarly, the proportion of cows classified 
as having low BCS and the average BCS did not differ 
(P > 0.60) between ER (61.2% and 2.84 ± 0.01) and 
LR groups (60.5% and 2.84 ± 0.01). The proportion of 
cows classified as estrous cyclic at the initiation of the 
Ovsynch protocol for the first postpartum AI did not 
differ (P = 0.13) between treatments(ER = 64.7 vs. 
LR = 69.3%).
Accuracy of PAG ELISA for Early 
Pregnancy Diagnosis
A total of 2,129 test diagnostics were performed at 
different intervals after AI. The overall accuracy of the 
PAG ELISA for early diagnosis of pregnancy was high 
and showed little variation with the day after insemina-
tion when the blood was sampled (Table 1). The overall 
kappa value was 0.82 (95% CI = 0.79–0.84). The Se 
averaged 95.1% and the Sp averaged 89.0%.
Journal of Dairy Science Vol. 97 No. 8, 2014
ReSYNCHRoNIZatIoN PRogRaMS foR DaIRY CoWS 4937
Pregnancy per AI, Pregnancy Loss, 
and Time to Pregnancy
Pregnancy per AI (P/AI) at the first AI diagnosed 
on d 46 after insemination did not differ between 
treatments and averaged 28.9% (Table 2). Cows with 
moderate BCS had greater (P = 0.003) P/AI than 
those with low BCS (34.6 vs. 25.3%). Estrous cyclic 
cows tended to have (P = 0.06) greater P/AI compared 
with those that were anovular at the initiation of the 
Ovsynch protocol (31.0 vs. 24.5%). Pregnancy to the 
first AI postpartum was not affected by parity or by the 
interactions between treatment and parity, BCS, and 
estrous cyclic status. The risk of pregnancy loss at first 
postpartum insemination between the initial pregnancy 
diagnosis using transrectal ultrasound on 28 and the 
transrectal palpation on d 46 of gestation did not differ 
between treatments (Table 2). Multiparous cows had 
greater (P = 0.02) pregnancy loss than primiparous 
cows (19.7 vs. 12.5%). Estrous cyclic status and BCS 
did not influence (P > 0.40) the risk of pregnancy loss 
at first AI.
The proportion of nonpregnant cows that were re-
synchronized and received timed AI was greater (P 35 94.4 (474/502) 85.2 (167/196) 94.2 (474/503) 85.6 (167/195) 91.8 (641/698) 0.79 (0.74–0.84)
Overall 95.1 (1,042/1,096) 89.0 (919/1,033) 90.1 (1,042/1,156) 94.5 (919/973) 92.1 (1,961/2,129) 0.82 (0.79–0.84)
1Sensitivity = (number of pregnant cows correctly diagnosed by PAG ELISA/number of pregnant cows diagnosed by ultrasonography or palpa-
tion) × 100; specificity = (number of nonpregnant cows correctly diagnosed by PAG ELISA/number of nonpregnant cows diagnosed by ultra-
sonography or palpation) × 100; positive predictive value = (number of cows correctly testing pregnant/number of total pregnant cows by PAG 
ELISA) × 100; negative predictive value = (number of cows correctly testing nonpregnant/number of total nonpregnant cows by PAG ELISA) 
× 100; accuracy = (number of correct PAG ELISA/number of total tests performed) × 100.
Table 2. Pregnancy per AI and pregnancy loss at the first timed AI, submission to timed AI after the first 
insemination, and interval between AI in Holstein cows subjected to early or late resynchronization protocols 
Parameter
Treatment1
P-valueEarly Resynch Late Resynch
Cows 452 520 —
Pregnant at first postpartum timed AI,2 %
 d 28 35.4 34.6 0.80
 d 46 28.3 29.4 0.78
Pregnancy loss, 28 to 46 d,3 % 20.0 15.3 0.31
Submission to resynchronized timed AI,4 % 30.0 7.6 0.20) P/AI after 
the first insemination.
The rate of pregnancy after the first postpartum 
insemination did not differ (P = 0.12) between treat-
ments [adjusted hazard ratio (AHR) = 1.23; 95% CI 
= 0.94 to 1.61], which resulted in minor numerical dif-
ferences in interval from calving to pregnancy or from 
study enrollment to pregnancy (Figure 3). As a conse-
quence of the similar pregnancy rates, the proportions 
of pregnant cows after 72 d of breeding did not differ 
between treatments (52.1 for ER and 50.0% for LR).
Cows with moderate BCS became pregnant faster (P 
= 0.04) than those with low BCS (AHR = 1.34; 95% CI 
= 1.02 to 1.76). Similarly, primiparous cows tended (P 
= 0.09) to became pregnant at a faster rate compared 
with multiparous cows (AHR = 1.29; 95% CI = 0.96 to 
1.72); however, estrous cyclic status did not affect (P = 
0.19) the rate of pregnancy after first AI (AHR = 1.23; 
95% CI = 0.90 to 1.67).
DISCuSSIOn
Timely reinsemination of nonpregnant cows requires 
either adequate detection of estrus or an accurate meth-
od of pregnancy diagnosis that can be used soon after 
Figure 2. Day after insemination when pregnancy diagnosis was 
performed for early Resynch (ER) by pregnancy-associated glycopro-
tein (PAG) ELISA (A) or late Resynch (LR) by transrectal palpation 
(B). For ER and LR, the mean (±SD) day was 28.0 ± 1.8 and 45.6 ± 
3.5 after insemination, respectively.
Figure 3. Survival curves for interval to pregnancy after calving 
(A) or after the first postpartum AI (B) in lactating Holstein cows 
subjected to early (ER; solid line) or late (LR; dotted line) resyn-
chronization (Resynch) protocols based on nonpregnant diagnosis by 
pregnancy-associated glycoproteins ELISA or transrectal palpation, 
respectively. During the experimental periods of 72 d, cows were ob-
served for estrus and those in estrus were inseminated on the same day. 
The adjusted hazard ratio for ER compared with LR was 1.23 (95% 
CI = 0.94 to 1.61; P = 0.12). The median (95% CI) days postpartum 
to pregnancy were 132 (113 to 145) and 140 (125 to 143) for ER and 
LR, respectively. The median days from first AI to pregnancy were 65 
and 72 d for ER and LR, respectively. After 72 d of breeding, 52.1 and 
50.0% of the ER and LR cows, respectively, became pregnant.
Journal of Dairy Science Vol. 97 No. 8, 2014
ReSYNCHRoNIZatIoN PRogRaMS foR DaIRY CoWS 4939
insemination (Green et al., 2011). Resynchronization 
programs are regularly used to manage reproduction in 
dairycattle with the aim of improving reinsemination 
rates (Fricke et al., 2003; Galvão et al., 2007), such that 
time to pregnancy can be shortened. Therefore, per-
forming a pregnancy test shortly after AI is expected to 
benefit reproductive performance because it should re-
duce the interval between breedings and, consequently, 
reduce days open (De Vries, 2006; Ribeiro et al., 2012; 
Giordano et al., 2013). As anticipated, cows assigned 
to the ER program had a shorter interval to pregnancy 
diagnosis, which reduced the mean interval between in-
seminations by 13 d. However, this reduction in interval 
between AI was not sufficient to alter the hazard of 
pregnancy in the first 72 d of breeding, and resulted in 
only a numerical reduction of 8 d open. This difference 
in time to pregnancy was less than that described by 
Silva et al. (2009), who reported a 14.5-d reduction in 
cows resynchronized following a nonpregnancy diagno-
sis by PAG ELISA on d 27 compared with cows exam-
ined by transrectal ultrasonography on d 32 after AI. 
The smaller difference observed in the current study 
was caused by the large proportion of cows reinsemi-
nated following detection of estrus and before a non-
pregnancy diagnosis. In general, the benefit of an early 
resynchronization should be observed when detection of 
spontaneous estrus and reinsemination are low. When 
different programs for resynchronization were imple-
mented in the same herd, the high detection of estrus 
before nonpregnancy diagnosis limited the benefit of 
an early resynchronization that combined insemination 
on estrus followed by timed AI 3 d after detection of 
nonpregnancy (Galvão et al., 2007). On the other hand, 
when all cows were subjected to timed insemination, 
then an earlier diagnosis of nonpregnancy and concur-
rent resynchronization shortened the interval between 
inseminations and reduced the time to pregnancy in 
dairy cows (Silva et al., 2009). Therefore, benefits of an 
early pregnancy diagnosis and resynchronization seem 
to be dependent on the proportion of the herd that 
is subjected to timed insemination. When detection of 
estrus is high, such that a large portion of the cows 
are reinseminated before pregnancy diagnosis, then the 
potential benefits of an early method for pregnancy 
diagnosis and resynchronization are largely diminished.
As anticipated, the ER group had a shorter AI inter-
val (approximately 13 d) and this difference occurred 
in both cows inseminated following spontaneous estrus 
or timed AI. Because cows in LR had a longer period 
for detection of estrus (approximately 2 wk), they 
had, on average, 6 additional days to reinsemination 
by detected estrus and almost 3 additional weeks for 
reinsemination by timed AI. This was expected, as the 
design of the study allowed a longer period for detec-
tion of estrus and 3 additional weeks to initiate the 
timed AI protocol to resynchronize nonpregnant cows. 
When cows were subjected only to timed AI, then 
early nonpregnancy detection with the PAG ELISA 
followed by a resynchronized timed insemination re-
duced days between inseminations and the interval to 
pregnancy (Silva et al., 2009). Because P/AI usually 
did not change with the day after insemination when 
the resynchronized timed AI was initiated (Silva et al., 
2009), it is anticipated that any improvement in the 
rate of pregnancy will be dependent on the interval 
between breedings. When cows were subjected to an 
ER protocol following a rapid PAG ELISA on d 25 
after AI with reinsemination on d 28 compared with 
reinsemination on d 35 after nonpregnancy diagnosis 
performed by transrectal ultrasonography, then the 7 d 
was not sufficient to reduce the interval to pregnancy 
(Green et al., 2011). It is possible that the 7-d interval 
between reinseminations or the limited number of cows 
(n = 203) might have precluded those authors from 
finding differences between treatments. In the current 
study, the ER protocol was not capable of improving 
the reproductive performance of dairy cows, despite 
initiating the resynchronization protocol 2 d before 
an early pregnancy diagnosis by the PAG ELISA, and 
the fact that the PAG ELISA occurred, on average, 14 
d before the transrectal palpation. More than 90% of 
the LR cows were reinseminated following detection of 
spontaneous estrus, which likely defeated the benefit 
of an earlier resynchronization protocol (Galvão et 
al., 2007). Furthermore, it is possible that the 72 d of 
reinsemination was not sufficient to allow differences 
in interval to pregnancy to be detected between the 2 
treatments implemented.
The measures of reliability of the PAG ELISA were 
high and similar to those reported by others that used 
the same (Silva et al., 2007) or similar assays (Green et 
al., 2009; Thompson et al., 2010). The sensitivity of the 
PAG ELISA ranged from 94.4 to 98.7%, indicating that 
1.3 to 5.6% of the pregnant cows were not detected as 
pregnant and continued the resynchronization protocol 
for timed AI. These cows likely aborted following the 
PGF2α injection of the Ovsynch protocol during resyn-
chronization, which was expected to reduce the rate 
of pregnancy in the ER group. Despite the estimated 
overall 4.9% iatrogenic abortion in ER, the interval to 
pregnancy and the proportion of pregnant cows by the 
end of the 72-d breeding period did not differ between 
treatments. Therefore, the risk of iatrogenic abortion 
was partially offset by the earlier resynchronization, 
which resulted in a numerical reduction in median days 
to pregnancy. It is important to note that both tran-
srectal ultrasonography and palpation are not 100% ac-
curate (Silva et al., 2007), although they continue to be 
4940 SINeDINo et aL.
Journal of Dairy Science Vol. 97 No. 8, 2014
the gold standard methods for detection of pregnancy 
in cattle. Nevertheless, the kappa value of 0.82 indi-
cates a high level of agreement between the diagnostic 
methods used as gold standard and the PAG ELISA 
(Martin et al., 1987). Typically, a kappa value of 1 
denotes perfect agreement, whereas a kappa value of 0 
denotes no agreement beyond chance. When comparing 
tests, a kappa >0.60 indicates a high level of agreement 
(Martin et al., 1987).
The measures of reliability of the PAG ELISA varied 
slightly according to the day after insemination when 
blood was sampled, and was highest between 31 to 
35 d after AI. Szenci et al. (1998) and Green et al. 
(2009) also demonstrated that measures of reliability 
of chemical tests varied slightly with day after AI. The 
differences with day after insemination probably oc-
curred because of the changes in PAG concentration 
throughout gestation (Green et al., 2005; Thompson 
et al., 2010; Giordano et al., 2012), and that very early 
pregnancy diagnosis typically results in inflated values 
for pregnancy loss (Santos et al., 2004). Collectively, 
these factors would likely influence the accuracy of test 
diagnostics. Green et al. (2005) reported that PAG 
became detectable by ELISA as early as d 22 of gesta-
tion in a small percentage of the pregnant cows. Green 
et al. (2005) also showed that, from d 24 to 28, PAG 
concentrations increased rapidly, reaching an average 
of 8.75 ± 3.04 ng/mL by d 28 of gestation. In the same 
study, the average concentration of PAG increased to 
12.3 ± 4.08 ng/mL in wk 5 and then declined to 6.8 
± 3.8 ng/mL in wk 8 of gestation. A similar pattern 
was observed by Thompson et al. (2010), who followed 
the plasma profile of pregnant dairy cows and observed 
that PAG concentrations increased from d 24 to 30, 
reaching approximately 5.0 ng/mL, and then decreased 
to approximately 2.0 ng/mL by d 60 after AI. Gior-
dano et al. (2012) also followed the dynamics of PAG 
concentration from 1 to 49 d of gestation and observed 
a rapid increase from 22 to 32 d, reaching a peak of 
4.4 ± 1.8 ng/mL at approximately d 42, followed by a 
steadily decline to reach the lowest value of 2.5 ± 0.9 
ng/mL on d 49 aftertimed AI. From these studies, it 
seems plausible that the accuracy of PAG as a method 
of pregnancy diagnosis will suffer from small variations 
in the timing of blood sampling and testing.
Sensitivity, which is the ability of the test to identify 
a pregnant cow, was highest (98.7%) when blood was 
collected from 31 to 35 d after AI. In the same period, 
the NPV was 99.1%, which represented the proportion 
of nonpregnant outcomes provided by PAG ELISA that 
were from truly nonpregnant cows. A possible reason 
for the high NPV is that temporal PAG profile that 
peaks around the fifth week of gestation (Green et al., 
2005; Thompson et al., 2010), possibly allowing for a 
more accurate diagnosis around this time. These results 
indicate that a PAG ELISA would likely identify almost 
all pregnant cows from 31 to 35 d after AI. They also 
indicate that, from every 100 nonpregnancy diagnosis, 
only 0.9% would be represented by a pregnant cow. 
The PPV averaged 90.1%, meaning that 9.9% of the 
pregnant outcomes by the PAG ELISA were actually 
nonpregnant cows that would require a later reconfir-
mation of pregnancy to identify these misdiagnosed 
animals. It is likely that a portion of these misdiagnoses 
were the result of pregnancy loss. In fact, not every 
cow diagnosed as pregnant by ultrasonography is truly 
pregnant (Silva et al., 2007), and some cows that have 
signs of a viable pregnancy will fail to maintain gesta-
tion in the following days (Santos et al., 2004; Giordano 
et al., 2012). Santos et al. (2004) summarized numerous 
studies and found 12.8% late-embryonic mortality in 
lactating dairy cows in the first 15 d after the initial 
pregnancy diagnosis starting on d 27 after AI. Gior-
dano et al. (2012) reported that although concentra-
tions of PAG declined within 1 to 2.5 d after induction 
of pregnancy loss with PGF2α or intrauterine infusion 
of hypertonic saline solution, it took an average of 9.5 
d for concentrations to become equal to those of non-
pregnant cows. Therefore, is it expected that during the 
process of pregnancy loss, some cows might be falsely 
diagnosed as pregnant because the amount of PAG in 
plasma is still elevated and above the threshold value 
for detection of pregnancy. In fact, only 5 d after induc-
tion of embryonic mortality, the concentration of PAG 
was below the cut-off value for pregnancy (Giordano 
et al., 2012). This delay is in part because the half-life 
for some PAG in the maternal circulation is 2.7 to 3.9 
d, and induction of pregnancy loss caused the PAG 
concentration to decrease below the cut-off value for 
pregnancy only after 4 to 8 d (Szenci et al., 2003). On 
the other hand, the false-negative results are caused by 
some pregnant cows with PAG concentrations below the 
threshold for pregnancy detection. Indeed, some cows 
exhibit individual variations with low PAG concentra-
tion (Szenci et al., 1998; Green et al., 2005; Giordano 
et al., 2012). Nevertheless, these cases are exceptions. 
The majority of the cows exhibit a temporal pattern of 
PAG expression during gestation (Green et al., 2005; 
Giordano et al., 2012).
COnCLuSIOnS
Implementing an early diagnosis of pregnancy fol-
lowed by resynchronization in cows observed for estrus 
resulted in a small proportion of resynchronized in-
seminations performed by timed AI, which reduced the 
interval between breedings, but was unable to reduce 
days open or the interval from first insemination to 
Journal of Dairy Science Vol. 97 No. 8, 2014
ReSYNCHRoNIZatIoN PRogRaMS foR DaIRY CoWS 4941
pregnancy. The lack of benefit from early resynchro-
nization was likely caused by the high detection of 
estrus before the day of pregnancy diagnosis. The PAG 
ELISA for early pregnancy diagnosis had high mea-
sures of reliability at different intervals after AI, with 
the highest Se and NPV observed between 30 and 35 
d after AI, indicating that this interval resulted in the 
lowest proportion of missed pregnancies and the lowest 
risk for inducing an iatrogenic abortion with the use of 
prostaglandins. When herds have high detection of es-
trus, the interval after AI when pregnancy is diagnosed 
seems to have little effect on time to pregnancy.
aCKnOWLeDGmentS
The authors thank the owner and staff of Rancho 
Teresita Dairy (Tulare, CA) for the use of their cows 
and facilities. Our appreciation is extended to Mon-
santo Co. (St. Louis, MO) for performing the chemical 
tests. Financial support for this study was provided by 
a grant from Monsanto Co.
reFerenCeS
Bartolome, J. A., A. Sozzi, J. McHale, P. Melendez, A. C. M. Arteche, 
F. T. Silvestre, D. Kelbert, K. Swift, L. F. Archbald, and W. W. 
Thatcher. 2005. Resynchronization of ovulation and timed insemi-
nation in lactating dairy cows. II: Assigning protocols according to 
stages of the estrous cycle, or presence of ovarian cysts or anestrus. 
Theriogenology 63:1628–1642.
Cerri, R. L. A., J. E. P. Santos, S. O. Juchem, K. N. Galvão, and R. 
C. Chebel. 2004. Timed artificial insemination with estradiol cypi-
onate or insemination at estrus in high-producing dairy cows. J. 
Dairy Sci. 87:3704–3715.
Chebel, R. C., J. E. P. Santos, R. L. A. Cerri, K. N. Galvão, S. O. 
Juchem, and W. W. Thatcher. 2003. Effect of resynchronization 
with GnRH on day 21 after artificial insemination on pregnancy 
rate and pregnancy loss in lactating dairy cows. Theriogenology 
60:1389–1399.
De Vries, A. 2006. Economic value of pregnancy in dairy cattle. J. 
Dairy Sci. 89:3876–3885.
Ferguson, J. D., D. T. Galligan, and N. Thomsen. 1994. Principal 
descriptors of body condition score in Holstein cows. J. Dairy Sci. 
77:2695–2703.
Fricke, P. M., D. Z. Caraviello, K. A. Weigel, and M. L. Welle. 2003. 
Fertility of dairy cows after resynchronization of ovulation at 
three intervals following first timed insemination. J. Dairy Sci. 
86:3941–3950.
Galvão, K. N., J. E. P. Santos, R. L. A. Cerri, R. C. Chebel, H. M. 
Rutigliano, R. G. Bruno, and R. C. Bicalho. 2007. Evaluation of 
methods of resynchronization for insemination in cows of unknown 
pregnancy status. J. Dairy Sci. 90:4240–4252.
Giordano, J. O., P. M. Fricke, and V. E. Cabrera. 2013. Economics 
of resynchronization strategies including chemical tests to identify 
nonpregnant cows. J. Dairy Sci. 96:949–961.
Giordano, J. O., J. N. Guenther, G. Lopes Jr., and P. M. Fricke. 2012. 
Changes in serum pregnancy-associated glycoprotein, pregnancy-
specific protein B, and progesterone concentrations before and af-
ter induction of pregnancy loss in lactating dairy cows. J. Dairy 
Sci. 95:683–697.
Green, J. A., T. E. Parks, M. P. Avalle, B. P. Telugu, A. L. McLain, 
A. J. Peterson, W. McMillan, N. Mathialagan, R. R. Hook, S. Xie, 
and R. M. Roberts. 2005. The establishment of an ELISA for the 
detection of pregnancy-associated glycoproteins (PAGs) in the se-
rum of pregnant cows and heifers. Theriogenology 63:1481–1503.
Green, J. C., E. M. Newsom, and M. C. Lucy. 2011. Incorporation of 
a rapid pregnancy-associated glycoprotein ELISA into a CIDR-
Ovsynch resynchronization program for a 28 day re-insemination 
interval. Theriogenology 75:320–328.
Green, J. C., D. H. Volkmann, S. E. Poock, M. F. McGrath, M. Eh-
rhardt, A. E. Moseley, and M. C. Lucy. 2009. Technical note: A 
rapid enzyme-linked immunosorbent assay blood test for pregnan-
cy in dairy and beef cattle. J. Dairy Sci. 92:3819–3824.
Martin, S. W., A. H. Meek, and P. Willeberg. 1987. Measurement 
of disease frequency and production. Pages 62–76 in Veterinary 
Epidemiology: Principles and Methods. 1st ed. Iowa State Univ. 
Press, Ames.
Meadows, C., P. J. Rajala-Schultz, and G. S. Frazer. 2005. A spread-
sheet-based model demonstrating the nonuniform economic effects 
of varying reproductive performance in Ohio dairy herds. J. Dairy 
Sci. 88:1244–1254.
Moreira, F., C. Orlandi, C. A. Risco, R. Mattos, F. Lopes, and W. 
W. Thatcher. 2001. Effects of presynchronization and bovine so-
matotropin on pregnancy rates to a timedartificial insemination 
protocol in lactating dairy cows. J. Dairy Sci. 84:1646–1659.
NRC. 2001. Nutrient Requirements of Dairy Cattle. 7th rev. ed. Natl. 
Acad. Press, Washington, DC.
Ribeiro, E. S., K. N. Galvão, W. W. Thatcher, and J. E. P. Santos. 
2012. Economic aspects of applying reproductive technologies to 
dairy herds. Anim. Reprod. 9:370–387.
Santos, J. E. P., W. W. Thatcher, R. C. Chebel, R. L. A. Cerri, and 
K. N. Galvão. 2004. The effect of embryonic death rates in cattle 
on the efficacy of estrus synchronization programs. Anim. Reprod. 
Sci. 82–83:513–535.
Sasser, R. G., C. A. Ruder, K. A. Ivani, J. E. Butler, and W. C. 
Hamilton. 1986. Detection of pregnancy by radioimmunoassay of a 
novel pregnancy-specific protein in serum of cows and a profile of 
serum concentrations during gestation. Biol. Reprod. 35:936–942.
Silva, E., R. A. Sterry, D. Kolb, N. Mathialagan, M. F. McGrath, 
J. M. Ballam, and P. M. Fricke. 2007. Accuracy of a pregnancy-
associated glycoprotein ELISA to determine pregnancy status of 
lactating dairy cows twenty-seven days after timed artificial in-
semination. J. Dairy Sci. 90:4612–4622.
Silva, E., R. A. Sterry, D. Kolb, N. Mathialagan, M. F. McGrath, J. 
M. Ballam, and P. M. Fricke. 2009. Effect of interval to resynchro-
nization of ovulation on fertility of lactating Holstein cows when 
using transrectal ultrasonography or a pregnancy-associated gly-
coprotein enzyme-linked immunosorbent assay to diagnose preg-
nancy status. J. Dairy Sci. 92:3643–3650.
Szenci, O., J. F. Beckers, P. Humblot, J. Sulon, G. Sasser, M. A. M. 
Taverne, J. Varga, R. Baltusen, and G. Schekk. 1998. Comparison 
of ultrasonography, bovine pregnancy-specific protein B, and bo-
vine pregnancy-associated glycoprotein 1 tests for pregnancy de-
tection in dairy cows. Theriogenology 50:77–88.
Szenci, O., J. F. Beckers, J. Sulon, M. M. Bevers, L. Börzsönyi, L. 
Fodor, F. Kovács, and M. A. M. Taverne. 2003. Effect of induction 
of late embryonic mortality on plasma profiles of pregnancy associ-
ated glycoprotein 1 in heifers. Vet. J. 165:307–313.
Thompson, I. M., R. L. A. Cerri, I. H. Kim, J. A. Green, J. E. P. San-
tos, and W. W. Thatcher. 2010. Effects of resynchronization pro-
grams on pregnancy per artificial insemination, progesterone, and 
pregnancy-associated glycoproteins in plasma of lactating dairy 
cows. J. Dairy Sci. 93:4006–4018.
Warnick, L. D., H. O. Mohammed, M. E. White, and H. N. Erb. 1995. 
The relationship of the interval from breeding to uterine palpation 
for pregnancy diagnosis with calving outcomes in Holstein cows. 
Theriogenology 44:811–825.
Xie, S., B. G. Low, R. J. Nagel, K. K. Kramer, R. V. Anthony, A. P. 
Zoli, J. F. Beckers, and R. M. Roberts. 1991. Identification of the 
major pregnancy-specific antigens of cattle and sheep as inactive 
members of the aspartic proteinase family. Proc. Natl. Acad. Sci. 
USA 88:10247–10251.
Copyright of Journal of Dairy Science is the property of Elsevier Science and its content may
not be copied or emailed to multiple sites or posted to a listserv without the copyright holder's
express written permission. However, users may print, download, or email articles for
individual use.