J. Dairy Sci.
102:567–577
https://doi.org/10.3168/jds.2018-15070
© American Dairy Science Association®, 2019.
Using serum and plasma samples to assess failure
of transfer of passive immunity in dairy calves
I. Elsohaby,1,2* J. T. McClure,1 L. A. Waite,1 M. Cameron,1 L. C. Heider,1 and G. P. Keefe1
1
Department of Health Management, Atlantic Veterinary College, University of Prince Edward Island, Charlottetown, Prince Edward Island,
Canada C1A 4P3
2
Department of Animal Medicine, Division of Infectious Diseases, Faculty of Veterinary Medicine, Zagazig University, Zagazig City 44511,
Sharkia Province, Egypt
ABSTRACT
The objectives of this study were (1) to determine
the differences in IgG and total protein (TP) con-
tent of serum and plasma samples collected from the
same calves; (2) to evaluate the correlation between
calf serum and plasma IgG levels, Brix scores, and
TP concentrations; (3) to determine whether different
cut-off values should be used for plasma and serum to
assess failure of transfer of passive immunity (FTPI)
in dairy calves; and (4) to evaluate the level of agree-
ment between results obtained from using serum and
plasma samples of the same calves to assess FTPI using
optimal cut-off values. Blood samples (n = 217) were
collected from Holstein calves at 3 to 10 d of age on 30
commercial dairy farms in Nova Scotia and Newfound-
land, Canada. Paired serum and plasma samples were
analyzed for IgG concentration by the reference radial
immunodiffusion assay, transmission infrared (TIR)
spectroscopy, digital and optical Brix refractometers,
and optical TP refractometer. The IgG concentrations
measured by RID and TIR spectroscopy in serum were
similar to those in plasma. However, the Brix and
TP refractometer readings were significantly higher
in plasma than in serum. The prevalence of FTPI in
serum and plasma samples based on a RID-IgG concen-
tration <10 g/L was 43.3 and 46.5%, respectively. The
RID-IgG concentration was correlated with TIR-IgG
(r = 0.92 and 0.89), digital Brix (r = 0.80 and 0.80),
optical Brix (r = 0.77 and 0.77), and optical TP (r =
0.75 and 0.77) refractometers in serum and plasma, re-
spectively. The correlations between paired serum and
plasma IgG content were 0.85 by TIR spectroscopy,
0.80 by digital Brix, 0.77 by optical Brix, and 0.79 by
optical TP refractometer. The optimal cut-off values
Received May 15, 2018.
Accepted September 23, 2018.
*Corresponding author: ielsohaby@​upei​.ca
567
for TIR spectroscopy, digital Brix, optical Brix, and TP
refractometers to assess FTPI using serum were 13.1
g/L, 8.7% Brix, 8.4% Brix and 5.1 g/dL, respectively;
and the optimal cut-off values with plasma were 13.4
g/L, 9.4% Brix, 9.3% Brix and 5.8 g/dL, respectively.
When using these optimal cut-off values, the level of
agreement (88.1%) between results derived from testing
serum and plasma by TIR spectroscopy was substan-
tial, with a kappa (κ) value of 0.76. The results derived
from testing serum and plasma by digital Brix refrac-
tometer showed substantial agreement (83.4%), with
a κ value of 0.65, which is higher than the agreement
and κ value (74.7% and 0.51) reported for the opti-
cal Brix refractometer. Substantial agreement (81.6%)
between serum and plasma TP was also obtained when
using the optical TP refractometer, with a κ value of
0.63. In conclusion, serum or plasma samples can be
used interchangeably for measuring IgG concentrations
and assessing FTPI in dairy calves. However, differ-
ent cut-offs must be used to assess FTPI depending
on the sample matrix. Furthermore, results obtained
from serum samples showed higher agreement with the
reference RID assay than those obtained from plasma
samples.
Key words: calves, serum, plasma, immunoglobulin
G, radial immunodiffusion
INTRODUCTION
Dairy calves are born without any acquired immu-
nity because immunoglobulins cannot cross the placen-
tal structure of the cow during gestation (Borghesi et
al., 2014). Calves rely on the passive transfer of IgG
through consumption of maternal colostrum provided
within the first hours of life (McGrath et al., 2016).
Failure of calves to ingest or absorb sufficient amount
of colostral IgG results in failure of transfer of pas-
sive immunity (FTPI) (Weaver et al., 2000; Godden,
2008; Beam et al., 2009). There is a recognized asso-
ciation between FTPI (IgG <10 g/L) and short- and
long-term health and productivity in calves (Cuttance
568 ELSOHABY ET AL.
et al., 2018; Lora et al., 2018). Calves with FTPI are
more susceptible to infectious diseases and have higher
morbidity and mortality (Windeyer et al., 2014). Fur-
thermore, FTPI is associated with decreased milk yield
and increased culling rates in dairy heifers during the
first lactation (Chuck et al., 2018). Thus, monitoring
of calves for FTPI is essential to identify herd man-
agement deficiencies and ensure timely detection and
implementation of interventional measures, which can
reduce the morbidity, mortality, and production issues
associated with FTPI in the herd (Furman-Fratczak et
al., 2011).
Several methods have been developed to measure IgG
concentration in dairy calves either directly or indi-
rectly. Direct methods include radial immunodiffusion
(RID) assay (McBeath et al., 1971), ELISA (Gelsinger
et al., 2015), infrared spectroscopy (Elsohaby et al.,
2016), and an automated turbidimetric immunoassay
(Alley et al., 2012). Indirect methods include refrac-
tometry (Deelen et al., 2014), zinc sulfate (ZnSO4)
turbidity test, glutaraldehyde coagulation test, and
sodium sulfate turbidity test (Tyler et al., 1996; Parish
et al., 1997; Weaver et al., 2000). Although the RID
assay is the historical reference standard method for
direct quantification of IgG concentration in serum,
plasma, colostrum, and milk (Pritchett et al., 1994;
Tyler et al., 1996; Weaver et al., 2000; Bielmann et al.,
2010), it is a laboratory-based assay, takes 18 to 24 h
to obtain a result, requires skills to read the plates, has
a high cost, and uses reagents with a short shelf life
(Liu et al., 2007; Riley et al., 2007). Thus, ELISA and
infrared spectroscopy have been used recently for direct
quantification of serum and colostrum IgG concentra-
tion (Baumrucker et al., 2014; Gelsinger et al., 2015;
Elsohaby et al., 2016, 2017). Results from direct meth-
ods have been used in the development of the indirect
methods by adjusting the refractive index, Brix scores,
specific gravity, or total protein (TP) levels to the IgG
concentrations (Deelen et al., 2014; MacFarlane et al.,
2014; Villarroel et al., 2014; Hernandez et al., 2016).
Earlier research studies have used serum or plasma
IgG and TP concentrations to assess FTPI in dairy
calves (MacFarlane et al., 2014; Villarroel et al., 2014).
However, to our knowledge, no direct comparisons
have been made between the performance of RID, TIR
spectroscopy, Brix, and TP refractometers using serum
and plasma samples from the same calves. Thus, the
objectives of our study were (1) to determine whether
IgG concentration, Brix scores, and TP values differ in
serum and plasma samples of the same calves; (2) to
evaluate the correlation between calf serum and plasma
IgG levels, Brix scores, and TP concentrations; (3) to
determine the optimal cut-off values and the diagnostic
Journal of Dairy Science Vol. 102 No. 1, 2019
test characteristics of the TIR spectroscopy, Brix, and
TP refractometers to assess FTPI using serum and
plasma of the same calves; and (4) to evaluate the level
of agreement between results obtained using serum and
plasma samples.
MATERIALS AND METHODS
Calf Enrollment and Sampling
Holstein calves (n = 217) from 30 commercial dairy
farms in Nova Scotia (n = 24) and Newfoundland (n =
6), Canada, were sampled between April and Septem-
ber 2015. Herds were selected by veterinary clinics in
the study area, and each farm delivered between 5 and
11 samples. At each farm, whole blood samples were
collected from 3- to 10-d-old calves by jugular veni-
puncture using a 20-gauge, 1-inch hypodermic needle
(BD Vacutainer Precision Glide, Becton Dickinson Co.,
Franklin Lakes, NJ), into a sterile, plastic Vacutainer
tube with and without lithium heparin (BD Vacu-
tainer, Becton Dickinson Co.) to obtain plasma and
serum samples, respectively. Samples were labeled with
the farm name, calf identification number, and collec-
tion date, and then stored on ice for transportation
to the Maritime Quality Milk Laboratory, University
of Prince Edward Island (UPEI). Serum and plasma
samples were separated by centrifugation at 1,500 ×
g for 15 min at ~20°C. Three aliquots of serum and
plasma were collected and stored at −80°C for later
analysis. In the end, 217 paired serum and plasma
samples were available for analysis. This study was
conducted in accordance with the Canadian Council on
Animal Care guidelines (CCAC, 2009) under a protocol
(#15-001) approved by the Animal Care Committee at
UPEI. The sample size calculation for the study was
based on finding a difference of 0.2 g/L between serum
and plasma IgG concentrations, considering that the
standard deviation of the IgG concentration 1.0 g/L
(Deelen et al., 2014; Elsohaby et al., 2015), and α and
β errors of 0.05 and 0.20, respectively. A sample size
of 198 calves was targeted for enrollment in this study.
The sample size calculation was performed using Java
Applets for Power and Sample Size (www​.stat​.uiowa​
.edu/​~rlenth/​Power; Lenth, 2006–2009).
RID Analysis
Serum and plasma samples were thawed at room
temperature (20–24°C) and vortexed for 10 s. A com-
mercial RID assay (Bovine IgG RID Kit; Triple J
Farms, Bellingham, WA) was used as the reference
method to measure serum and plasma IgG concentra-
 FAILURE OF TRANSFER OF PASSIVE IMMUNITY IN DAIRY CALVES
tions. The RID assay was performed according to the
manufacturer’s instructions, using 5 μL of undiluted
sample in each well tested alongside the manufactur-
er’s standards. Diameters of precipitating rings were
measured using a handheld caliper after 18 to 24 h
of incubation at room temperature. Serum and plasma
samples with IgG concentrations greater than the
manufacturer’s stated performance range for the assay
(>30 g/L) were diluted (1:1) with deionized sterile wa-
ter and retested. All samples and assay standards were
tested in duplicate. The average of the assay standards
was used to build a calibration curve to determine the
IgG concentration for each serum and plasma sample.
The final IgG concentration for each sample was de-
termined by calculating the average of the 2 replicates.
The results were considered acceptable if the coefficient
of determination of the calibration curve derived from
the standards was ≥0.97.
TIR Spectroscopy Analysis
Infrared spectra were acquired for serum and
plasma samples using a transmission infrared (TIR)
spectrometer (Tensor 37, Bruker Optics, Milton, ON,
Canada) equipped with a deuterium tryglycine sulfate
detector and controlled by proprietary software (Opus
ver. 6.5, Bruker Optics). Thawed serum and plasma
samples were diluted (1:1) with deionized sterile water
and tested in replicates of 6 by evenly spreading 10-μL
aliquots into 5-mm-diameter wells within an adhesive-
masked, 96-well silicon microplate (Riley et al., 2007).
An empty well served as the background reference for
each microplate. A total of 1,302 (217 serum or plasma
samples × 6 replicate) spectra were collected, over a
wavenumber range between 4,000 and 400 cm−1 with a
nominal resolution of 4 cm−1, with 512 scans collected
for data acquisition. Collected spectra were converted
into printable (PRN) format (Grams/AI version 7.02,
Thermo Fisher Scientific, Waltham, MA) and imported
into MATLAB (MathWorks R2016b, Natick, MA). A
previously developed partial least squares regression
model built for prediction of serum IgG concentration
from infrared spectra (Elsohaby et al., 2014) was used
to predict serum and plasma IgG concentrations. The
IgG concentration was predicted from each spectrum
and, subsequently, the IgG concentration for each se-
rum and plasma sample was calculated as the average
of the 6 replicate IgG values.
Refractometer Analyses
Serum and plasma samples were allowed to thaw at
room temperature (20 to 24°C), and analysis of Brix
569
scores was performed using a digital Brix refractometer
(PAL-1, Atago Co. Ltd., Bellevue, WA), with a scale
from 0 to 52% Brix, and an optical Brix refractometer
(model 300001; SPER Scientific, Scottsdale, AZ), with
a scale from 0 to 32% Brix. Samples were also analyzed
for TP using an optical handheld refractometer (RHC-
00ATC handheld refractometer, Westover, Woodinville,
WA). For the digital refractometer, approximately 250
μL of serum or plasma was used, and the Brix score was
determined by transmitting light through the sample
in the prism and recording the reading on a digital
scale. For the optical refractometers, approximately 50
μL of serum or plasma was placed on the prism and
held up to a light source; the result was read through
the viewfinder at the interface between light and dark
areas. Before each analysis, the refractometers were
cleaned and calibrated with distilled water at room
temperature. The reading of the optical refractometers
was determined first, to avoid any bias in the results by
the technician.
Statistical Analysis
Statistical analysis was performed using Stata sta-
tistical software (version 15.0; StataCorp, 2017), with
results considered significant at P < 0.05. Descriptive
statistics for the serum and plasma IgG, Brix scores,
and TP concentrations were calculated, and normality
was assessed by application of the Shapiro-Wilk test.
A paired Student’s t-test was performed to evaluate
differences between serum and plasma IgG, Brix scores,
and TP concentrations. Serum and plasma TIR-IgG
values, Brix scores, and TP concentrations were plotted
against each other and the reference RID-IgG values.
From these plots, correlation coefficients (r-values)
were determined. The receiver operating characteristic
(ROC) curve was created to plot the true positive rate
against the false positive rate for TIR spectroscopy,
Brix and TP refractometers. The area under the curve
in the ROC plot for each assay with serum and plasma
samples was calculated. The optimal cut-points for
each assay were determined (cutpt command in Stata;
Clayton, 2013). The diagnostic test characteristics—
sensitivity (Se), specificity (Sp), positive predictive
value (PPV), negative predictive value (NPV), and
accuracy—of TIR spectroscopy and Brix and TP refrac-
tometers were calculated to assess FTPI in dairy calves,
using a serum and plasma RID-IgG concentration of 10
g/L as the cut-off value. Sensitivity was defined as the
proportion of calves with FTPI (<10 g/L) that were
correctly identified as such, and Sp was defined as the
proportion of calves without FTPI (≥10 g/L) that were
correctly identified as such. The PPV was defined as
Journal of Dairy Science Vol. 102 No. 1, 2019
570 ELSOHABY ET AL.
the probability that a calf with a positive test result
has FTPI, whereas NPV describes the probability that
a calf with a negative test result has adequate transfer
of passive immunity. Accuracy was defined as the pro-
portion of calves that were correctly classified. Bland-
Altman plots were used to examine the difference and
interchangeability between serum and plasma TIR-IgG,
Brix scores, and TP concentrations (Bland and Alt-
man, 1995). The level of agreement between TIR, Brix,
and TP refractometers for detecting FTPI was assessed
for paired serum and plasma samples, followed by cal-
culation of Cohen’s kappa statistic (κ).
RESULTS AND DISCUSSION
Descriptive Statistics
The distribution of serum and plasma IgG concentra-
tions measured by RID, TIR spectroscopy, and Brix
and TP refractometers was normal (P > 0.05). Descrip-
tive statistics (mean, SD, minimum, and maximum) of
the IgG concentration in calf serum and plasma ob-
tained by RID, TIR spectroscopy, and Brix and TP
refractometers are presented in Table 1. In this study,
we detected no significant differences between serum
and plasma IgG concentrations measured by RID (P
= 0.073) and TIR spectroscopy (P = 0.131), which is
similar to previous reports in the literature (McEwan
et al., 1970). Conversely, the differences between serum
and plasma Brix scores and TP concentrations mea-
sured by refractometers were significant (P = 0.001),
which is consistent with the results reported in other
studies (Naylor and Kronfeld, 1977; MacFarlane et al.,
2014; Villarroel et al., 2014).
Approximately half of the serum (94/217) and
plasma (101/217) samples had RID-IgG concentra-
tions <10 g/L, generating an FTPI prevalence of 43.3
and 46.5%, respectively. The prevalence of calves with
FTPI in this study is consistent with what has been
previously reported in Atlantic Canada by Elsohaby
et al. (2014; 51%) and higher than the prevalence re-
Table 1. Descriptive statistics of serum and plasma IgG concentrations, Brix scores (%Brix), and total protein (g/dL) in 217 dairy calves
Serum
1
Method Mean SD Minimum
RID (g/L) 13.34 9.25 1.64
TIR spectroscopy (g/L) 14.00 8.57 −4.80
Digital Brix (% Brix) 8.64 0.91 6.8
Optical Brix (% Brix) 8.75 0.99 6.3
Optical total protein (g/dL) 5.24 0.83 3.6
1
RID = radial immunodiffusion assay; TIR = transmission infrared spectroscopy.
2
Paired t-test differed significantly if P-value <0.05.
Journal of Dairy Science Vol. 102 No. 1, 2019
ported in other Canadian provinces by Deelen et al.
(2014; 4.75%) and Atkinson et al. (2017; 16%). The
difference in the prevalence of FTPI in the present and
previous studies may be related to the age of calves at
sampling and number of farms enrolled in each study,
as well as the management practices on the farm (God-
den, 2008). In the present study, only 34 out of 217
calves were sampled between 6 and 10 d of age. These
samples may not represent a valid prevalence estimate
for FTPI in these herds. However, the 30 farms used in
this study had previously reported a high proportion
of poor-quality colostrum (48% of colostrum had IgG
<50 g/L; Elsohaby et al., 2017), so FTPI estimates are
likely higher than would be expected, compared with
farms feeding high-quality colostrum to calves.
Correlation Coefficients
The correlation coefficient between serum IgG levels
was 0.92, as estimated by TIR spectroscopy and RID
assay (Figure 1A), similar to the correlation (r = 0.94)
reported between RID and TIR spectroscopy for dairy
calves (Elsohaby et al., 2016). The correlations between
serum RID-IgG concentrations and Brix scores mea-
sured by digital and optical Brix refractometers were
0.80 (Figure 1B) and 0.77 (Figure 1C), respectively.
The serum TP measured by optical TP refractometer
was also correlated with serum RID-IgG concentration
(r = 0.75; Figure 1D). Although similar correlations
between serum Brix scores or serum TP and RID-IgG
concentration have been reported (Elsohaby et al.,
2015; Hernandez et al., 2016), others have found nu-
merically higher correlation coefficients from 0.87 to
0.93 between refractometry and RID assay (Morrill et
al., 2013; Deelen et al., 2014). Plasma IgG concentra-
tion measured by RID assay was positively correlated
with IgG concentration obtained by TIR spectroscopy
(r = 0.89; Figure 2A), Brix scores obtained by digital
Brix (r = 0.80; Figure 2B) and optical Brix (r = 0.77;
Figure 2C) refractometers, and TP concentration mea-
sured by optical TP refractometer (r = 0.77; Figure
Plasma
P-value:
Maximum   Mean SD Minimum Maximum t-test2
51.42   12.56 9.18 1.77 59.36 0.073
41.37   13.54 7.87 −1.3 41.38 0.131
12.3   9.15 0.92 7.1 13.1 0.001
12.2   9.25 0.97 7.1 13.4 0.001
8.6   5.64 0.85 3.8 9.2 0.001
 FAILURE OF TRANSFER OF PASSIVE IMMUNITY IN DAIRY CALVES 571

Figure 1. Scatter plots of serum IgG concentrations obtained by radial immunodiffusion (RID) assay and (A) serum IgG concentration
predicted by transmission infrared (TIR) spectroscopy; (B) serum Brix scores obtained by digital Brix refractometer; (C) serum Brix scores
obtained by optical Brix refractometer; and (D) serum total protein measured by optical total protein refractometer for 217 serum samples.

2D). These findings are higher than the correlation
previously reported between refractometry and RID
(McBeath et al., 1971; r = 0.72) and between ELISA
and RID (Gelsinger et al., 2015; r = 0.59) using plasma
samples.
The wide range of reported correlation coefficients
could be explained by variations in the source of se-
rum and plasma samples and the instrument variation
between refractometers used in these different studies
(Vandeputte et al., 2011). In the current study, 217
samples were collected from 30 farms, whereas Deelen
et al. (2014) and MacFarlane et al. (2014) collected 400
serum samples from 5 farms and 61 plasma samples
from 7 farms, respectively. Furthermore, variations in
the correlation coefficients in these studies could be due
to the differences in non-IgG contents in serum and
plasma samples (Pfeiffer et al., 1977). Refractometry
measures IgG content indirectly by measuring serum
TP and plasma TP, which are affected by a calf’s health
status (dehydration; Tyler et al., 1999). Consequently,
a number of studies have reported a poor correlation
between plasma TP and IgG concentration because of
variable fibrinogen content in plasma samples (Naylor
and Kronfeld, 1977). Others have reported that the
nonimmune components of serum proteins might lead
to poor correlation with IgG concentration (Pfeiffer et
al., 1977). Thus, the use of caprylic acid to separate
IgG from other serum proteins has been investigated
(Morrill et al., 2013).
The correlation coefficient between paired serum and
plasma IgG concentrations was 0.85 for TIR spectros-
copy (Figure 3A). The correlations between serum and
plasma Brix scores estimated by the digital and optical
Brix refractometers were 0.80 and 0.77, respectively

Journal of Dairy Science Vol. 102 No. 1, 2019

572 ELSOHABY ET AL.

Figure 2. Scatter plots of plasma IgG concentrations obtained by radial immunodiffusion (RID) assay and (A) plasma IgG concentration
predicted by transmission infrared (TIR) spectroscopy; (B) plasma Brix scores obtained by digital Brix refractometer; (C) plasma Brix scores
obtained by optical Brix refractometer; and (D) serum total protein measured by optical total protein refractometer for 217 plasma samples.

(Figure 3B). However, the serum and plasma TP con-
centrations obtained by optical TP refractometer were
positively correlated (r = 0.79; Figure 3C). Correla-
tions reported in the present study are lower than those
reported between serum and plasma TP by MacFarlane
et al. (2014; r = 0.98) and Villarroel et al. (2014; r =
0.91). Pearson correlation coefficients between serum
and plasma IgG, Brix scores, and TP concentrations
are given in Supplemental Table S1 (https:​/​/​doi​.org/​10​
.3168/​jds​.2018​-15070).
Diagnostics Test Characteristics
Receiver operator characteristic curves were cre-
ated to determine the optimal cut-off values for TIR
spectroscopy and Brix and TP refractometers to detect
FTPI (RID-IgG <10 g/L) using serum (Figure 4A) or
plasma (Figure 4B) samples of the same calves. The
test characteristics (Se, Sp, PPV, NPV, and accuracy)
associated with these optimal cut-points are shown in
Table 2 for each test with serum and plasma. These
results showed that the optimal cut-off values for as-
sessing FTPI using calf serum were different than those
for plasma. Similar results were previously reported by
MacFarlane et al. (2014), who used different cut-points
for serum TP (5.2 g/dL) and plasma TP (5.6 g/dL)
to assess FTPI in the same calves. The TIR spectros-
copy in the present study showed higher Se and lower
Sp than what has been previously reported for both
transmission and attenuated total reflectance infrared
spectroscopies (Elsohaby et al., 2015). Conversely, the
Brix and TP refractometers showed similar Se, Sp, and
accuracy to that reported by Elsohaby et al. (2015),
lower Sp than that reported by Deelen et al. (2014), and

Journal of Dairy Science Vol. 102 No. 1, 2019

 FAILURE OF TRANSFER OF PASSIVE IMMUNITY IN DAIRY CALVES
Figure 3. Scatter plots comparing (A) serum and plasma IgG con-
centration obtained by transmission infrared (TIR) spectroscopy; (B)
serum and plasma Brix scores measured by digital and optical Brix
refractometers; and (C) serum and plasma total protein obtained by
optical total protein refractometer for 217 paired serum and plasma
samples.
573
lower Se than that reported in other studies (Thornhill
et al., 2015; Hernandez et al., 2016). Furthermore, our
TIR spectroscopy and Brix and TP refractometry re-
sults showed higher Sp and slightly lower Se in serum
than in plasma of the same calves.
For each assay, using serum samples resulted in a
higher number of calves being correctly classified as
FTPI positive and negative than when plasma samples
were used. Furthermore, when plasma was used, the as-
says, particularly the refractometers, showed relatively
lower PPV compared with when serum was used. This
means that when plasma is used with these assays, a
substantial number of calves with adequate transfer of
passive immunity will be classified as having FTPI (i.e.,
false positives). The cost of a false-positive FTPI diag-
nosis at calf level may be relatively low; however, the
cost could be significantly higher if misdiagnosis results
in changes in the management of calves and colostrum
on the farm. The PPV and NPV of each assay with
serum and plasma samples across populations with dif-
ferent FTPI prevalences are presented in Supplemental
Figure S1 (https:​/​/​doi​.org/​10​.3168/​jds​.2018​-15070).
Agreement Among Tests
The level of agreement among TIR spectroscopy, Brix
and TP refractometry, and the reference RID assay
for detection of FTPI was higher with serum samples
than with plasma samples of the same calves, as as-
sessed using the κ statistic (Table 2). A similar level of
agreement with RID assays has been reported for TIR
and attenuated total reflectance infrared spectroscopy
(Elsohaby et al., 2016) and Brix and TP refractometers
(Elsohaby et al., 2015).
Bland-Altman plots (Figure 5) revealed that the
mean values of the difference between serum and plasma
IgG concentrations provided by TIR spectroscopy was
0.46 g/L (Figure 5A). The level of agreement (88.1%)
between results derived from testing serum and plasma
by TIR spectroscopy was substantial, with a κ value of
0.76 (Table 3). This revealed that using either serum
or plasma samples to measure IgG levels for the as-
sessment of FTPI would result in similar classification
of dairy calves with and without FTPI, which is con-
sistent with the results previously published for calves
(McEwan et al., 1970) and foals (Ujvari et al., 2017).
The agreement between serum and plasma Brix
scores and TP concentrations was assessed using
Bland-Altman plots and showed no obvious system-
atic bias between serum and plasma results obtain by
digital Brix (Figure 5B), optical Brix (Figure 5C), and
TP (Figure 5D) refractometers. However, the results
derived from testing serum and plasma by digital Brix
refractometer showed a substantial agreement (83.4%),
Journal of Dairy Science Vol. 102 No. 1, 2019
574 ELSOHABY ET AL.

Figure 4. Receiver operating characteristic (ROC) curve analysis of IgG concentration measured by transmission infrared (TIR) spectros-
copy, digital Brix refractometer (DBR), optical Brix refractometer (OBR), and optical total protein refractometer (OPR) compared with the
reference radial immunodiffusion (RID) assay for (left) serum and (right) plasma samples collected from 217 dairy calves.

with a κ value of 0.65, which higher than the agree-
ment and κ value (74.7% and 0.51) reported for the
optical Brix refractometer. The agreement (81.6%)
between serum and plasma TP obtained by optical TP
refractometer was slightly lower than that of the digital
Brix refractometers (Table 3). The lower agreement
between plasma and serum TP concentration might
be attributed to the significant and consistent higher
concentration of plasma TP than serum TP (Villarroel
et al., 2014). Furthermore, plasma contains fibrinogen
and other coagulation proteins, which are lost during
serum processing (George, 2001). Plasma may also
contain excessive disodium EDTA or lithium heparin
due to underfilled blood collection tubes, which falsely

Table 2. Test characteristics of transmission infrared (TIR) spectroscopy and Brix and total protein (TP) refractometers for detecting failure
of transfer of passive immunity in serum and plasma samples collected from 217 dairy calves using optimal cut-off values compared with IgG
(<10 g/L) determined by radial immunodiffusion (RID) assay

Test characteristics1 (%)

RID Test
Method   Cut-off Se Sp PPV NPV Accuracy κ2 high/low3 high/low4
Serum samples                
  TIR spectroscopy 13.1 g/L 93.6 86.2 83.8 94.6 89.4 0.79 123/94 112/105
  Digital Brix 8.7% Brix 91.5 70.7 70.5 91.6 79.7 0.60 123/94 95/122
  Optical Brix 8.4% Brix 72.3 86.2 80.0 80.3 80.2 0.59 123/94 132/85
  Optical TP 5.1 g/dL 75.5 77.2 71.7 80.5 76.5 0.52 123/94 118/99
Plasma samples                
  TIR spectroscopy 13.4 g/L 95.0 83.6 83.5 95.1 88.9 0.78 116/101 102/115
  Digital Brix 9.4% Brix 93.1 58.6 66.2 90.7 74.7 0.50 116/101 75/142
  Optical Brix 9.3% Brix 84.2 64.7 67.5 82.4 73.7 0.48 116/101 91/126
  Optical TP 5.8 g/dL 81.2 68.1 68.9 80.6 74.2 0.49 116/101 98/119
1
Se = sensitivity; Sp = specificity; PPV = positive predictive value; NPV = negative predictive value; accuracy = percentage of correctly clas-
sified samples.
2
κ = Cohen’s kappa value.
3
RID high/low = number of samples declared high (IgG ≥10 g/L) or low (IgG <10 g/L) by RID assay.
4
Test high/low = number of samples declared high (IgG ≥10 g/L) or low (IgG <10 g/L) by each assay.

Journal of Dairy Science Vol. 102 No. 1, 2019

 FAILURE OF TRANSFER OF PASSIVE IMMUNITY IN DAIRY CALVES 575
Table 3. Level of agreement between results obtained from using serum and plasma samples for assessing of failure of transfer of passive
immunity in the same 217 dairy calves

Method1   Serum cut-off   Plasma cut-off Agreement (%) κ2 P-value

TIR spectroscopy 13.1 g/L 13.4 g/L 88.1 0.76 0.0001
Digital Brix 8.7% Brix 9.4% Brix 83.4 0.65 0.0001
Optical Brix 8.4% Brix 9.3% Brix 74.7 0.51 0.0001
Optical total protein 5.1 g/dL 5.8 g/dL 81.6 0.63 0.0001
1
TIR = transmission infrared spectroscopy.
2
κ = Cohen’s kappa value.

increases plasma TP concentration (Dubin and Hunt,
1978).
CONCLUSIONS
Our results suggest that serum or plasma samples
could be used interchangeably to measure IgG concen-
trations for assessing FTPI in dairy calves. However,
different cut-off values should be used for serum and
plasma to assess FTPI when using TIR or refractom-
eters. Optimal cut-off values for TIR spectroscopy,
digital Brix, optical Brix, and TP refractometers to as-
sess FTPI using serum were 13.1 g/L, 8.7% Brix, 8.4%
Brix, and 5.1 g/dL, whereas those for plasma were 13.4

Figure 5. Bland-Altman plots of the differences in serum and plasma IgG concentrations, Brix scores (%Brix), and total protein measured
by (A) transmission infrared (TIR) spectroscopy; (B) digital Brix refractometer; (C) optical Brix refractometer; and (D) optical total protein
refractometer in 217 paired samples.

Journal of Dairy Science Vol. 102 No. 1, 2019

576 ELSOHABY ET AL.
g/L, 9.4% Brix, 9.3% Brix, and 5.8 g/dL, respectively.
Results obtained from using serum samples showed
higher agreement with the reference RID assay than
those obtained from using plasma samples.
ACKNOWLEDGMENTS
The authors thank summer program students, par-
ticipating dairy farmers and veterinarians, as well
as Theresa Andrews, Cynthia Mitchell, and Natasha
Robinson (Maritime Quality Milk Laboratory, Char-
lottetown, PEI, Canada) for their technical assistance
and data collection. This research was funded by Zoetis
(Kirkland, QC, Canada) and Atlantic Canada Oppor-
tunities Agency (AIF: 195174). I. Elsohaby is supported
by a Mitacs Elevate Postdoctoral Fellowship (IT09473).
REFERENCES
Alley, M. L., D. M. Haines, and G. W. Smith. 2012. Short communica-
tion: Evaluation of serum immunoglobulin G concentrations using
an automated turbidimetric immunoassay in dairy calves. J. Dairy
Sci. 95:4596–4599.
Atkinson, D. J., M. A. G. von Keyserlingk, and D. M. Weary. 2017.
Benchmarking passive transfer of immunity and growth in dairy
calves. J. Dairy Sci. 100:3773–3782.
Baumrucker, C. R., A. Stark, O. Wellnitz, C. Dechow, and R. M.
Bruckmaier. 2014. Short communication: Immunoglobulin varia-
tion in quarter-milked colostrum. J. Dairy Sci. 97:3700–3706.
Beam, A. L., J. E. Lombard, C. A. Kopral, L. P. Garber, A. L. Win-
ter, J. A. Hicks, and J. L. Schlater. 2009. Prevalence of failure of
passive transfer of immunity in newborn heifer calves and associ-
ated management practices on US dairy operations. J. Dairy Sci.
92:3973–3980.
Bielmann, V., J. Gillan, N. R. Perkins, A. L. Skidmore, S. Godden,
and K. E. Leslie. 2010. An evaluation of brix refractometry instru-
ments for measurement of colostrum quality in dairy cattle. J.
Dairy Sci. 93:3713–3721.
Bland, J. M., and D. G. Altman. 1995. Comparing two methods of
clinical measurement: A personal history. Int. J. Epidemiol. 24:S7–
S14.
Borghesi, J., L. C. Mario, M. N. Rodrigues, P. O. Favaron, and M.
A. Miglino. 2014. Immunoglobulin transport during gestation in
domestic animals and humans—A review. Open J. Anim. Sci.
4:323–336.
CCAC (Canadian Council on Animal Care). 2009. The care and use
of farm animals in research, teaching and testing. CCAC, Ottawa,
ON, Canada.
Chuck, G. M., P. D. Mansell, M. A. Stevenson, and M. M. Izzo. 2018.
Early-life events associated with first-lactation performance in
pasture-based dairy herds. J. Dairy Sci. 101:3488–3500.
Clayton, P. 2013. CUTPT: Stata module for empirical estimation of
cutpoint for a diagnostic test. Statistical Software Components
S457719. Department of Economics, Boston College, Newton, MA.
Cuttance, E. L., W. A. Mason, R. A. Laven, and C. V. C. Phyn. 2018.
The relationship between failure of passive transfer and mortality,
farmer-recorded animal health events and body weights of calves
from birth until 12 months of age on pasture-based, seasonal calv-
ing dairy farms in New Zealand. Vet. J. 236:4–11.
Deelen, S. M., T. Ollivett, D. Haines, and K. Leslie. 2014. Evalua-
tion of a brix refractometer to estimate serum immunoglobulin G
concentration in neonatal dairy calves. J. Dairy Sci. 97:3838–3844.
Dubin, S., and P. Hunt. 1978. Effect of anticoagulants and glucose on
refractometric estimation of protein in canine and rabbit plasma.
Lab. Anim. Sci. 28:541–544.
Journal of Dairy Science Vol. 102 No. 1, 2019
Elsohaby, I., J. T. McClure, M. Cameron, L. C. Heider, and G. P.
Keefe. 2017. Rapid assessment of bovine colostrum quality: How
reliable are transmission infrared spectroscopy and digital and op-
tical refractometers? J. Dairy Sci. 100:1427–1435.
Elsohaby, I., J. T. McClure, and G. P. Keefe. 2015. Evaluation of
digital and optical refractometers for assessing failure of transfer of
passive immunity in dairy calves. J. Vet. Intern. Med. 29:721–726.
Elsohaby, I., J. T. McClure, C. B. Riley, R. A. Shaw, and G. P. Keefe.
2016. Quantification of bovine immunoglobulin G using transmis-
sion and attenuated total reflectance infrared spectroscopy. J. Vet.
Diagn. Invest. 28:30–37.
Elsohaby, I., C. B. Riley, S. Hou, J. T. McClure, R. A. Shaw, and G.
P. Keefe. 2014. Measurement of serum immunoglobulin G in dairy
cattle using fourier-transform infrared spectroscopy: A reagent free
approach. Vet. J. 202:510–515.
Furman-Fratczak, K., A. Rzasa, and T. Stefaniak. 2011. The influence
of colostral immunoglobulin concentration in heifer calves’ serum
on their health and growth. J. Dairy Sci. 94:5536–5543.
Gelsinger, S. L., C. Jones, and A. Heinrichs. 2015. Effect of colostrum
heat treatment and bacterial population on immunoglobulin G ab-
sorption and health of neonatal calves. J. Dairy Sci. 98:4640–4645.
George, J. W. 2001. The usefulness and limitations of hand-held re-
fractometers in veterinary laboratory medicine: An historical and
technical review. Vet. Clin. Pathol. 30:201–210.
Godden, S. 2008. Colostrum management for dairy calves. Vet. Clin.
North Am. Food Anim. Pract. 24:19–39.
Hernandez, D., D. Nydam, S. Godden, L. Bristol, A. Kryzer, J. Ra-
num, and D. Schaefer. 2016. Brix refractometry in serum as a
measure of failure of passive transfer compared to measured im-
munoglobulin G and total protein by refractometry in serum from
dairy calves. Vet. J. 211:82–87.
Lenth, R. V. 2006–2009. Java Applets for Power and Sample Size. Ac-
cessed Feb. 25, 2015. http:​/​/​www​.stat​.uiowa​.edu/​~rlenth/​Power.
Liu, K. Z., A. Man, T. C. Dembinski, and R. A. Shaw. 2007. Quanti-
fication of serum apolipoprotein B by infrared spectroscopy. Anal.
Bioanal. Chem. 387:1809–1814.
Lora, I., F. Gottardo, B. Contiero, B. D. Ava, L. Bonfanti, A. Stefani,
and A. Barberio. 2018. Association between passive immunity and
health status of dairy calves under 30 days of age. Prev. Vet. Med.
152:12–15.
MacFarlane, J. A., D. H. Grove-White, M. D. Royal, and R. F. Smith.
2014. Use of plasma samples to assess passive transfer in calves
using refractometry: Comparison with serum and clinical cut-off
point. Vet. Rec. 174:303.
McBeath, D. G., W. J. Penhale, and E. F. Logan. 1971. An examina-
tion of the influence of husbandry on the plasma immunoglobulin
level of the newborn calf, using a rapid refractometer test for as-
sessing immunoglobulin content. Vet. Rec. 88:266–270.
McEwan, A. D., E. Fisher, I. Selman, and W. Penhale. 1970. A turbid-
ity test for the estimation of immune globulin levels in neonatal
calf serum. Clin. Chim. Acta 27:155–163.
McGrath, B. A., P. F. Fox, P. L. McSweeney, and A. L. Kelly. 2016.
Composition and properties of bovine colostrum: A review. Dairy
Sci. Technol. 96:133–158.
Morrill, K. M., J. Polo, A. Lago, J. Campbell, J. Quigley, and H.
Tyler. 2013. Estimate of serum immunoglobulin G concentration
using refractometry with or without caprylic acid fractionation. J.
Dairy Sci. 96:4535–4541.
Naylor, J. M., and D. S. Kronfeld. 1977. Refractometry as a measure
of the immunoglobulin status of the newborn dairy calf: Compari-
son with the zinc sulfate turbidity test and single radial immuno-
diffusion. Am. J. Vet. Res. 38:1331–1334.
Parish, S. M., J. W. Tyler, T. E. Besser, C. C. Gay, and D. Kryten-
berg. 1997. Prediction of serum IgG1 concentration in Holstein
calves using serum gamma glutamyltransferase activity. J. Vet.
Intern. Med. 11:344–347.
Pfeiffer, N. E., T. C. McGuire, R. B. Bendel, and J. M. Weikel. 1977.
Quantitation of bovine immunoglobulins: comparison of single ra-
dial immunodiffusion, zinc sulfate turbidity, serum electrophoresis,
and refractometer methods. Am. J. Vet. Res. 38:693–698.
 FAILURE OF TRANSFER OF PASSIVE IMMUNITY IN DAIRY CALVES
Pritchett, L. C., C. C. Gay, D. D. Hancock, and T. E. Besser. 1994.
Evaluation of the hydrometer for testing immunoglobulin G 1 con-
centrations in Holstein colostrum. J. Dairy Sci. 77:1761–1767.
Riley, C. B., J. T. McClure, S. Low-Ying, and R. A. Shaw. 2007. Use of
Fourier-transform infrared spectroscopy for the diagnosis of failure
of transfer of passive immunity and measurement of immunoglobu-
lin concentrations in horses. J. Vet. Intern. Med. 21:828–834.
StataCorp. 2017. Stata Statistical Software: Release 15. StataCorp LP,
College Station, TX.
Thornhill, J. B., G. Krebs, and C. Petzel. 2015. Evaluation of the brix
refractometer as an on-farm tool for the detection of passive trans-
fer of immunity in dairy calves. Aust. Vet. J. 93:26–30.
Tyler, J. W., D. D. Hancock, S. M. Parish, D. E. Rea, T. E. Besser, S.
G. Sanders, and L. K. Wilson. 1996. Evaluation of 3 assays for fail-
ure of passive transfer in calves. J. Vet. Intern. Med. 10:304–307.
Tyler, J. W., S. M. Parish, T. E. Besser, D. C. Van Metre, G. M.
Barrington, and J. R. Middleton. 1999. Detection of low serum im-
munoglobulin concentrations in clinically ill calves. J. Vet. Intern.
Med. 13:40–43.
577
Ujvari, S., C. Schwarzwald, N. Fouché, J. Howard, and A. Schoster.
2017. Validation of a point-of-care quantitative equine IgG turbi-
dimetric immunoassay and comparison of IgG concentrations mea-
sured with radial immunodiffusion and a point-of-care IgG ELISA.
J. Vet. Intern. Med. 31:1170–1177.
Vandeputte, S., J. Detilleux, and F. Rollin. 2011. Comparison of four
refractometers for the investigation of the passive transfer in beef
calves. J. Vet. Intern. Med. 25:1465–1469.
Villarroel, A., T. B. Miller, J. K. Ward, E. D. Johnson, and K. R.
Noyes. 2014. Differences in total protein concentration between
fresh and frozen serum and plasma samples used to assess failure
of passive transfer in dairy calves. J. Adv. Dairy Res. 2:1–5.
Weaver, D. M., J. W. Tyler, D. C. VanMetre, D. E. Hostetler, and G.
M. Barrington. 2000. Passive transfer of colostral immunoglobulins
in calves. J. Vet. Intern. Med. 14:569–577.
Windeyer, M. C., K. E. Leslie, S. M. Godden, D. C. Hodgins, K. D.
Lissemore, and S. J. LeBlanc. 2014. Factors associated with mor-
bidity, mortality, and growth of dairy heifer calves up to 3 months
of age. Prev. Vet. Med. 113:231–240.
Journal of Dairy Science Vol. 102 No. 1, 2019