J. Dairy Sci.
92:5677–5691
doi:10.3168/jds.2009-2378
© American Dairy Science Association, 2009.
Effects of anion supplementation to low-potassium prepartum diets
on macromineral status and performance of periparturient dairy cows
J. M. Ramos-Nieves,* B. J. Thering,*1 M. R. Waldron,†2 P. W. Jardon,‡3 and T. R. Overton*4
*Department of Animal Science, Cornell University, Ithaca, NY 14853
†Department of Animal Science, University of Vermont, Burlington 05405
‡West Central, Ralston, IA 51459
ABSTRACT
Data from multiparous Holstein cows (n = 43) were
used to determine whether supplementation of anions
to low-potassium (K) prepartum diets would improve
periparturient energy and macromineral status and af-
fect performance during the postpartum period. Begin-
ning 21 d before expected parturition, cows were fed a
control diet (1.29% K; +10 mEq/100 g; n = 21) or a low
dietary cation-anion difference (DCAD) diet (1.29% K;
−15 mEq/100 g; n = 22) with anions provided through
a combination of sulfate from calcium sulfate dihydrate
(0.40% S total ration) and chloride (1.17% Cl total
ration) from SoyChlor 16–7 (West Central, Ralston,
IA). All cows were fed the same postpartum diet from
parturition through 63 d postpartum. Feeding anions
decreased overall urine pH (8.17 vs. 6.70) during the
prepartum period. Overall, peripartum concentrations
of plasma Ca, P, and Mg were similar between treat-
ments; however, concentrations of plasma Ca tended
to be increased during the first 24 h postcalving in
cows fed the low DCAD diet. Overall, concentrations of
plasma P tended to be increased by feeding the anionic
diet prepartum; this effect was more pronounced during
the immediate peripartal period. Anionic supplementa-
tion did not affect incidence of clinical (<5 mg/dL) and
subclinical (5 to 8 mg/dL) hypocalcemia, clinical hypo-
phosphatemia (<2 mg/dL), or clinical (<1.1 mg/dL)
and subclinical (1.1 to 1.8 mg/dL) hypomagnesemia.
Nevertheless, subclinical hypophosphatemia (2 to 4
mg/dL) tended to be decreased at 16 h postcalving and
was decreased at d 2 postpartum for cows fed the an-
ionic diet prepartum. Anion supplementation decreased
prepartum dry matter intake (15.6 vs. 14.4 kg/d), but
did not affect postpartum dry matter intake (22.4 vs.
Received May 13, 2009.
Accepted July 21, 2009.
1
Current address: PO Box 263, McGraw, NY 13101.
2
Current address: Division of Animal Sciences, University of
Missouri, Columbia, MO 65211.
3
Current address: 815 Mockingbird Dr., Carroll, IA 51401.
4
Corresponding author: tro2@cornell.edu
5677
23.0 kg/d), milk yield (46.5 vs. 46.1 kg/d), or content
and yield of milk fat and true protein. Plasma con-
centrations of energy-related metabolites (glucose, non-
esterified fatty acids, β-hydroxybutyrate) were similar
for both groups during the prepartum and postpartum
periods. Glucose rate of appearance was determined
by continuous infusion of 6,6-dideuterated glucose in a
subset of cows between 6 and 10 d prepartum (control,
n = 12; low DCAD, n = 9) and 7 and 10 d postpartum
(control, n = 9; low DCAD, n = 8) periods. Glucose
rate of appearance was not affected by treatment dur-
ing the prepartum or postpartum periods. Overall, an-
ion supplementation of low K diets improved P status
during the early postpartum period, but did not affect
aspects of energy metabolism or periparturient perfor-
mance.
Key words: transition cow, dietary cation-anion dif-
ference, potassium, calcium
INTRODUCTION
Numerous experiments have demonstrated that ad-
justment of the DCAD is effective in reducing the in-
cidence of clinical hypocalcemia during the peripartum
period (Block, 1984; Oetzel et al., 1988; Gaynor et al.,
1989; Beede et al., 1991; Goff et al., 1991a). However,
as suggested by Overton and Waldron (2004), much of
this research has utilized animals with higher risk for
milk fever such as multiparous cows and Jersey cows.
Additionally, an important factor in DCAD research
has been the use of highly cationic diets as controls.
This involves the use of forages with high cationic loads
that exacerbate the difference between this control and
treatment groups in the cited experiments (Overton
and Waldron, 2004).
In some regions, the availability of forages with low K
concentrations has resulted in adoption of a strategy in
which the dietary DCAD is partially adjusted through
focus only on decreasing the K content of the prepar-
tum diet. Few studies have focused on whether further
reduction of the DCAD beyond that achieved through
use of low-K feeds would further improve macromineral
status. Moore et al. (2000) determined that reduction
5678 RAMOS-NIEVES ET AL.

Table 1. Ingredient and chemical composition of the prepartum diets lenge compared with cows fed cationic diets. It is not
Treatment known whether manipulation of DCAD in more moder-
ate ranges affects aspects of glucose metabolism.
Item Control Low DCAD Accordingly, the objectives of this experiment were
Ingredient (% of DM) to determine whether anion supplementation to a low-
Corn silage 47.3 47.1 K prepartum diet would affect macromineral status in
Wheat straw 15.7 15.7 transition cows. Furthermore, we sought to determine
Ground shelled corn 2.9 2.9
Soybean meal (47.5% CP) 5.9 5.9 whether DCAD strategy would affect aspects of energy
SoyPlus1 6.9 5.9 metabolism and performance during the transition pe-
Beet pulp 7.9 5.5 riod and early lactation.
Corn distillers grains 4.3 2.4
Soybean hulls pellets 5.9 5.7
Soy-Chlor 16–71 0.0 6.3
Calcium carbonate 2.0 1.1 MATERIALS AND METHODS
Calcium sulfate dihydrate 0.4 1.0
Magnesium oxide 0.3 0.1 Experimental Design and Treatments
Salt 0.2 0.3
Mineral and vitamin mix2 0.3 0.2 All procedures involving animals were approved
Chemical analysis3
[% of DM (SD)] before the experiment by the Cornell University Insti-
DM4 (% as-fed) 49.9 (3.7) 49.4 (4.4) tutional Animal Care and Use Committee. Forty-three
NEL5 (Mcal/kg) 1.48 (0.05) 1.48 (0.05) pregnant, nonlactating Holstein cows entering second
NDF (%) 46.1 (2.0) 44.6 (3.0)
NFC6 (%) 29.5 (2.8) 31.4 (2.7) or later lactation were moved into individual tiestalls
Starch (%) 17.5 (3.1) 18.0 (3.1) at the Cornell University Dairy Teaching and Research
Crude fat (%) 3.1 (0.5) 3.0 (0.3) Center at approximately d 28 before expected parturi-
CP (%) 15.9 (0.9) 15.9 (0.7)
Lignin (%) 3.7 (0.2) 4.0 (0.3) tion. Cows were fed a control prepartum diet (Table
Ca (%) 1.43 (0.26) 1.35 (0.24) 1) once daily for ad libitum intake. On d 21 before
K (%) 1.29 (0.14) 1.29 (0.15) expected parturition, cows were assigned to 1 of 2 treat-
Cl (%) 0.42 (0.04) 1.17 (0.08)
P (%) 0.31 (0.01) 0.32 (0.01) ment groups: control (n = 21) and low DCAD (n = 22)
Mg (%) 0.47 (0.08) 0.51 (0.04) diets fed once daily for ad libitum intake. Treatment
S (%) 0.31 (0.04) 0.4 (0.06) assignments were balanced based on parity, BCS, and
Na (%) 0.20 (0.01) 0.23 (0.04)
DCAD7 +11 −15 previous 305-d mature-equivalent milk yield.
1
Both experimental prepartum diets (Table 1) were
West Central (Ralston, IA).
2 formulated using CPM-Dairy version 3.0 (Cornell Uni-
Contained 3% Ca, 2.5% Mg, 6% K, 8.6% S, 8,600 mg/kg of Fe, 18,600
mg/kg of Zn, 6,000 mg/kg of Cu, 16,000 mg/kg of Mn, 110 mg/kg versity, the University of Pennsylvania, and the Miner
of Se, 330 mg/kg of Co, 570 mg/kg of I, 3,018.8 IU/g of vitamin A, Institute) and the forage component was based on corn
1,025.8 IU/g of vitamin D, and 20,242.3 IU/g of vitamin E. silage and wheat straw. The low DCAD diet contained
3
Based on 6 composite samples of TMR per group. additional calcium sulfate and a high-chloride (10.3%
4
Based on 21 (control) and 20 (low DCAD) weekly samples.
5 of DM) anionic supplement (SoyChlor 16–7, West Cen-
NEL (Mcal/kg) = 0.0245 × total digestible nutrients (%) – 0.12; Weiss
et al. (1992). tral, Ralston, IA). Both prepartum diets were fed once
6
NFC = 100 − [(NDF − neutral detergent insoluble CP) + CP + ash daily at 0800 h for ad libitum intake. After calving,
+ fat]. cows assigned to both treatments were fed the same
7
[(Na+ + K+) – (Cl− + S−2)]; mEq/100 g. lactation diet (Table 2) once daily for ad libitum intake
until d 63 postpartum.
of DCAD from +15 to 0 mEq/100 g of DM improved
plasma Ca status; further decreasing the DCAD to −15 Feeding, Milking Management, and Sampling
mEq/100 g of DM increased levels of total and ion-
ized plasma Ca compared with the partially adjusted Amounts of feed offered and refused were recorded
and control diets but only tended to improve Ca status daily, and samples of the forages, concentrate mixtures,
during the first day postcalving compared with the and TMR were obtained weekly. Dry matter content
partially adjusted diet. was determined by drying at 60°C until constant weight;
In addition, few studies have characterized effects of diets were adjusted using the weekly DM values for for-
DCAD adjustment on peripartal energy metabolism ages and concentrate mixtures. Dry matter intake was
and, more specifically, glucose metabolism. Bigner et calculated based upon the amounts of feed offered and
al. (1996) reported that nonpregnant, nonlactating refused together with the corresponding DM value for
cows fed very low DCAD diets (−40 mEq/100 g of the TMR. Dried samples of TMR were ground through
DM) had impaired insulin responses to glucose chal- a 2-mm screen in a Wiley mill, composited at 4-wk
Journal of Dairy Science Vol. 92 No. 11, 2009
 ANION SUPPLEMENTATION OF LOW-POTASSIUM DIETS FOR DRY COWS 5679
Table 2. Ingredient and chemical composition of the postpartum intervals, and subjected to wet chemistry analysis in
diet
a commercial laboratory (Cumberland Valley Ana-
Item Postpartum diet lytical Services Inc., Hagerstown, MD) for CP (method
Ingredient (% of DM)
990.03; AOAC, 2000), NDF (Van Soest et al., 1991),
TMR (%) 46.5 ether extract (method 920.39; AOAC, 1990), starch
Rye straw 1.2 (modification of Holm et al., 1986), lignin (Goering
Corn silage 27.7
Alfalfa hay 3.0
and Van Soest, 1970), and minerals (method 985.01;
High-moisture shell corn 13.4 AOAC, 2000).
Ground shelled corn 10.1 To confirm the presence of a metabolic acidosis in
Hay crop silage 18.9
Corn gluten feed 8.4
cows assigned to the low DCAD treatment, urine sam-
Wheat middlings 5.2 ples were collected from all cows twice weekly begin-
Corn distillers grains 4.9 ning during the week before assignment to treatment.
Blood meal, flash-dried 2.0
Animal fat 1.2
Midstream samples were collected at 5 to 7 h postfeed-
Limestone 1.1 ing and pH was measured with a calibrated portable
Sodium bicarbonate 0.8 pH meter (model UP-5 pH meter, Denver Instrument,
Expeller soybean meal 0.7
Fat supplement1 0.4
Denver, CO).
Urea 0.4 After parturition, cows were milked 3 times per day
Salt 0.4 and yields were recorded at each milking for the first
Calcium sulfate 0.1
Mepron M852 0.1
63 DIM. Milk samples were collected weekly from each
Trace mineral mix3 0.1 milking and composited into a single sample. The
Selenium 0.06% 050 0.04 composite samples were stored at 4°C with a milk pre-
Magnesium oxide 0.03
Gluten meal 60% 0.02
servative (Bronopol solution, Spectrum Chemicals &
Vitamins A, D, E4 0.02 Laboratory Products, Gardena, CA) for mid-infrared
Vitamin E (50% vitamin E acetate) 0.003 analysis of fat, protein, and lactose contents (AOAC,
Chemical analysis5 [% of DM (SD)]
DM6 (% as-fed) 46.5 (3.0)
2000; method 972.160) and SCC by optical fluorescent
NEL7 (Mcal/kg) 1.67 (0.05) analysis (AOAC, 2000: method 978.26; Dairy One
NDF (%) 37.1 (2.3) Cooperative Inc., Ithaca, NY). Calibration reference
NFC8 (%) 36.5 (2.6)
Starch (%) 24.3 (2.2)
methods used for the infrared milk analysis were as
Crude fat (%) 5.0 (0.4) follows: fat by ether extraction (AOAC, 2000; method
CP (%) 17.3 (0.8) 989.05), Kjeldahl true protein (AOAC, 2000; method
Lignin (%) 3.8 (0.3)
Ca (%) 0.91 (0.10)
991.22), total solids (AOAC, 2000; method 990.20), es-
P (%) 0.37 (0.03) timated NPN (0.19% on a protein basis), and estimated
K (%) 1.28 (0.07) ash [(true protein × 0.596) + 0.5379].
Mg (%) 0.30 (0.03)
Cl (%) 0.54 (0.05)
Body weight and BCS were recorded every week start-
S (%) 0.25 (0.02) ing 1 wk before treatment assignment and continuing
Na (%) 0.43 (0.03) throughout the experiment. The BCS was assessed on
DCAD9 20
a 5-point system (Wildman et al., 1982) using quarter-
1
Prilled saturated free fatty acids (99% fatty acid weight basis); Cargill point increments by 2 individuals, and the average of
Animal Nutrition proprietary blend (Elk River, MN). these 2 scores was used for statistical analysis. Health
2
Degussa Corporation (Kennesaw, GA). records with daily observations and treatments were
3
Contained 36% Ca, 0.01% P, 0.93% Mg, 0.84% S, 7,370 mg/kg of Zn,
1,273 mg/kg of Cu, 6,030 mg/kg of Mn, 110 mg/kg of Se, 167.5 mg/ maintained throughout the study.
kg of Co, 127.3 mg/kg of I, 1,202.5 IU/g of vitamin A, 225 IU/g of
vitamin D, and 2,302.5 IU/g of vitamin E.
4 Plasma Sampling
Contained 36,000 IU/g of vitamin A, 7,000 IU/g of vitamin D, and
70,000 IU/g of vitamin E.
5
Based on 7 composite samples of TMR.
Blood samples were collected via coccygeal venipunc-
6
Based on 29 weekly samples of TMR.
ture at 1400 h on the day before treatment assignment,
7
NEL (Mcal/kg) = 0.0245 × total digestible nutrients (%) – 0.12; Weiss once weekly during the prepartum treatment period,
et al. (1992). twice during the first 24 h postpartum (closest to 0700
8
NFC = 100 − [(NDF − neutral detergent insoluble CP) + CP + ash and 1500 h schedule), once daily for the next 5 d after
+ fat]. parturition, and every other day from d 5 to 15 after
9
[(Na+ + K+) – (Cl− + S−2)]; mEq/100 g.
parturition. Blood was collected in 10-mL heparinized
vacuum sampling tubes (143 USP units of sodium
heparin, Becton Dickinson, Franklin Lakes, NJ) and
plasma was harvested after centrifugation (2,800 × g,
Journal of Dairy Science Vol. 92 No. 11, 2009
5680 RAMOS-NIEVES ET AL.
15 min at 4°C), snap-frozen in liquid N2, and stored at
−20°C until subsequent analysis.
Glucose Kinetics
The stable isotope 6,6-dideuterated glucose (D2-
glucose) was obtained from Isotec (Miamisburg, OH).
Procedures for preparation of infusates were described
previously by Waldron et al. (2006). Prepartum and
postpartum bolus infusates contained 8.5 and 14
μmol of D2-glucose/kg of BW, respectively, whereas
infusates for continuous infusion contained 7 and 11.5
μmol of D2-glucose/kg of BW for the prepartum and
postpartum experimental procedures, respectively. The
D2-glucose for the continuous infusion was diluted into
a total volume of 200 mL using sterile 0.9% NaCl solu-
tion.
The experimental day for the basal glucose kinetics
protocol was between 7 and 10 d before expected calv-
ing date (9.5 ± 2.6 d and 6.7 ± 3.6 d for control and
low DCAD treatment groups, respectively; P = 0.05)
and between 7 and 10 d after parturition (9.6 ± 1.2 d
and 8.1 ± 0.6 d for control and low DCAD treatment
groups, respectively; P = 0.01). Preparation of cows
and sampling procedures during the infusion protocol
were as described previously by Waldron et al. (2006).
On the basal glucose kinetics experimental day, a base-
line (0 min) blood sample was obtained via a jugular
catheter at approximately 1130 h (approximately 3 h
after feeding). Following this blood sample, the priming
dose of D2-glucose was administered via the infusion
catheter in a bolus (<10 s) infusion. The bolus infusate
was flushed immediately from the catheter with 10 mL
of sterile physiological saline. With exception of the 0
min blood sample drawn immediately before bolus infu-
sion, all subsequent blood samples were timed starting
at the beginning of the D2-glucose bolus infusion. At
30 min after bolus infusion, a Plum XL infusion pump
(Abbott Laboratories, North Chicago, IL) was used to
initiate a continuous infusion (at a rate of 7.0 or 11.5
μmol of D2-glucose/kg of BW per hour during the pre-
partum and postpartum periods, respectively) of D2-
glucose through a low-protein-binding, 0.22-μm, inline
sterile Millex GV filter unit (SLGV033RS, Durapore,
Billerica, MA). Blood was sampled at 30, 60, 90, 120,
150, 165, 180, 195, and 210 min after bolus infusion.
The continuous infusion of isotope was terminated after
the 210 min blood sample. Access to feed and fresh
water was maintained at all times during the infusion
period.
Plasma Analyses
Plasma glucose concentrations were determined using
an enzymatic assay (glucose oxidase) in a commercial
Journal of Dairy Science Vol. 92 No. 11, 2009
kit (510A, Sigma Chemical, St. Louis, MO). Plasma
NEFA concentrations were analyzed by enzymatic
methods using a commercial kit (Wako Pure Chemical
Industries, Osaka, Japan) with modifications reported
by McCutcheon and Bauman (1986). Plasma BHBA
concentrations were analyzed by enzymatic methods
(BHBA dehydrogenase) using a commercial kit (310-
UV, Sigma Chemical). All blood metabolite assays were
conducted using 96-well microplates (Costar, Corning
Inc., Acton, MA) and a Versamax tunable microplate
reader spectrophotometer (Molecular Devices, Sunny-
vale, CA). Plasma total Ca, P, and Mg were analyzed
by spectrophotometric methods using standard proce-
dures for the Hitachi 917 Chemistry Analyzer (Roche
Diagnostics, Switzerland) at the Clinical Pathology
Laboratory of the Cornell University College of Veteri-
nary Medicine (Ithaca, NY).
Energy Balance
Prepartum and postpartum energy balance calcula-
tions were determined according to NRC (2001) equa-
tions. Weekly values for prepartum calculated energy
balance were determined as follows:
Prepartum NEL (Mcal/d) balance = energy intake
(Mcal of NEL/d) – [maintenance requirement (Mcal of
NEL/d) + pregnancy requirement (Mcal of NEL/d)],
where energy intake (Mcal/d) = weekly DMI average
(kg/d) × diet NEL (Mcal/kg of DM); maintenance
requirement (Mcal) = week metabolic BW (MBW;
kg0.75) × 0.08 (Mcal/kg0.75 per day); and pregnancy
requirement (Mcal) = (0.00318 × day of gestation –
0.0352) × (1/0.218).
Weekly values for postpartum calculated energy bal-
ance were determined as follows:
Postpartum NEL (Mcal/d) balance = energy intake
(Mcal of NEL/d) – [maintenance requirement (Mcal of
NEL/d) + lactation requirement (Mcal of NEL/d)],
where energy intake (Mcal/d) = weekly DMI average
(kg/d) × diet NEL (Mcal/kg of DM); maintenance re-
quirement (Mcal) = week MBW (kg0.75) × 0.08 (Mcal/
kg0.75 per day); and lactation requirement (Mcal) =
0.0929 × fat percentage + 0.0563 × true protein per-
centage + 0.0395 × lactose percentage.
Statistical Analysis
Before statistical analysis, abnormal milk produc-
tion data from 6 cows in the low DCAD group were
 ANION SUPPLEMENTATION OF LOW-POTASSIUM DIETS FOR DRY COWS
Figure 1. Least squares means and standard errors for urine pH
during the prepartum period for cows fed either a control (n = 21)
or low DCAD (n = 22) diet. Overall means were 8.2 and 6.7 for the
control and low DCAD groups, respectively. The P-value for the effect
of treatment was <0.0001, and the P-value for the interaction of treat-
ment and time was 0.08. Cov = covariate urinary pH values obtained
before introduction of treatment diets.
eliminated after 30 d or more postpartum because of
mastitis problems in those cows. Data for postpartum
production (DMI and milk yield) were reduced to
weekly means before analysis. Data were analyzed as
a completely randomized design with repeated mea-
sures using the MIXED procedures of SAS (version
2001; SAS Institute, Cary, NC). Terms in the model
were treatment, time, and the interaction of treatment
and time. Previous 305-d mature-equivalent milk yield
values and pretreatment (collected during the week be-
fore assignment to treatment) values for DMI, plasma
variables, BW, and BCS were used as covariates during
their respective analyses. Five covariance structures
were tested (compound symmetry, compound symme-
try with heterogeneous variance, autoregressive order
1, autoregressive order 1 with heterogeneous variance,
and antedependence 1); the structure with the lowest
Akaike’s information criterion was chosen. The propor-
tion of cows with subnormal levels of Ca, P, and Mg
were analyzed using Fisher’s exact test for nonpara-
metric data. Significance was declared at P < 0.05 and
trends at 0.05 < P < 0.15. Least squares means are
presented throughout.
RESULTS
Supplementation of anions to cows fed the low DCAD
treatment during the prepartum period resulted in
5681
changes in acid-base balance as reflected by lower urine
pH compared with controls depicted in Figure 1. Pre-
partum urine pH averaged 8.2 and 6.7 for cows assigned
to the control and low DCAD treatments, respectively.
Results for plasma concentrations of Ca, P, and Mg
during the periparturient period as affected by prepar-
tum treatment are presented in Table 3 and Figure 2.
Plasma concentrations of Ca and Mg during the prepar-
tum period were not affected by treatment (P > 0.15;
Table 3); however, prepartum plasma P concentrations
tended to be increased by feeding the low DCAD diet
(5.36 vs. 5.05; P = 0.09). Overall concentrations of Ca
and Mg in plasma samples collected during the first
15 d postpartum were not affected by treatment (P >
0.15; Table 3); however, a treatment by time interac-
tion existed (P < 0.05) for plasma P such that cows fed
the low DCAD diet during the prepartum period main-
tained higher plasma P concentrations during the first
7 d postpartum (Figure 3). This interaction contributed
to an overall trend for increased plasma P concentra-
tions during the postpartum period for cows fed the low
DCAD diet prepartum (5.26 vs. 4.88 mg/dL, P = 0.11;
Table 3). Although Ca concentrations were not affected
by treatment during the entire postpartum period (P >
0.15; Table 3), Ca concentrations during the first 24 h
postpartum tended (P = 0.07) to be slightly increased
in cows fed the low DCAD diet during the prepartum
period (7.99 ± 0.32 vs. 7.12 ± 0.33 mg/dL; Figure 4).
In addition to determining the effects of prepartum
DCAD treatment on plasma macromineral concentra-
tions, we evaluated the effects of treatment on the fre-
quency of cows experiencing clinically or subclinically
low concentrations of Ca, P, and Mg at different time
points during the early postpartum period; results are
presented in Figures 5, 6, and 7. The proportion of cows
experiencing clinical (<5 mg/dL) or subclinical (5–8
mg/dL) hypocalcemia was not affected by treatment
during the day after parturition and during the first 5 d
postpartum (Figure 5). With the exception of controls
at 8 h postpartum (~20% clinically hypocalcemic), the
proportion of clinically hypocalcemic cows remained
at 10% or less for both treatments at all time points
evaluated (Figure 5a). The proportion of cows catego-
rized as subclinically hypocalcemic was approximately
50% during each of the time points evaluated during
the first 3 d following parturition; however, prepartum
dietary treatment did not affect the proportion of cows
categorized as subclinically hypocalcemic (Figure 5b).
Despite these categorizations determined after the ex-
periment, only one cow from each treatment developed
clinical signs of paresis, as diagnosed by the herd health
staff, and it was treated using intravenous Ca infusion
during the first 5 d postpartum; a total of 3 cows from
the control treatment and 2 cows from the low DCAD
Journal of Dairy Science Vol. 92 No. 11, 2009
5682 RAMOS-NIEVES ET AL.

Table 3. Least squares means for plasma total Ca, P, and Mg concentrations during the prepartum and
postpartum period for cows fed either a control or low DCAD diet during the prepartum period

Treatment (Trt)1 P-value

Item Control Low DCAD SEM Trt Trt × day

2
Prepartum
Plasma total Ca (mg/dL) 8.69 8.72 0.15 0.88 0.51
Plasma P (mg/dL) 5.05 5.36 0.13 0.09 0.06
Plasma Mg (mg/dL) 1.53 1.51 0.03 0.65 0.44
Postpartum3
Plasma total Ca (mg/dL) 8.42 8.50 0.19 0.80 0.40
Plasma P (mg/dL) 4.88 5.26 0.16 0.11 0.05
Plasma Mg (mg/dL) 1.51 1.48 0.04 0.51 0.27
1
Cows assigned to control (n = 21) treatment received a prepartum diet with DCAD value of +11 mEq/100
g of DM and those assigned to low DCAD (n = 22) received a prepartum diet with a DCAD value of −15
mEq/100 g of DM.
2
Data collected weekly from assignment to treatment until parturition.
3
Data collected at 8 and 16 h on the day of parturition and on d 1, 2, 3, 4, 5, 7, 9, 11, 13, and 15 postpar-
tum.

treatment were treated with intravenous Ca during the
first 21 d postpartum.
The proportions of cows assigned to the 2 prepartum
treatments that experienced clinical (<2 mg/dL) or
subclinical (2–4 mg/dL) hypophosphatemia during the
first 24 h postpartum and first 5 d thereafter are pre-
sented in Figure 6. Consistent with the improvement in
plasma P concentrations during the first several days
postpartum in cows fed the low DCAD diet during
the prepartum period, cows fed low DCAD prepartum
tended (P = 0.11) to have decreased clinical hypophos-
phatemia at 8 h (Figure 6a) and decreased (P < 0.09)
subclinical hypophosphatemia at 16 h after parturition
(Figure 6b). In addition, cows fed the low DCAD diet
prepartum had decreased (P < 0.02) hypophosphatemia
on d 2 postpartum (Figure 6b).
Figure 7 depicts the proportion of cows categorized as
clinically (<1.1 mg/dL) or subclinically (1.1 to 1.8 mg/
dL) hypomagnesemic during the day of parturition and
first 5 d postpartum. Differences between treatments
were not apparent at any of the time points evaluated;
however, it is interesting to note that nearly 20% of
cows were categorized as clinically hypomagnesemic
during d 2 to 5 postpartum, and the percentage of
cows categorized as subclinically hypomagnesemic ap-
proached 80% during the same timeframe. These clas-
sifications occurred apparently without existing clinical
signs in cows that warranted treatment and in cows
that were fed Mg at or above NRC (2001) recommenda-
tions during the prepartum and postpartum periods in
diets that contained relatively low concentrations of K.
These results suggest that the dietary Mg sources used
in the present study had low bioavailability or that the
NRC (2001) requirements for dietary Mg supply merit
reevaluation.
Effects of prepartum anion supplementation on DMI
of cows during the prepartum and postpartum periods
are described in Table 4 and Figure 8. Cows fed the
low DCAD treatment had decreased DMI during the
prepartum period (14.4 vs. 15.6 kg/d, P < 0.01; Table
4). Although treatment by day interactions were not
significant (P = 0.48), it is apparent from Figure 8 that
decreased DMI early in the treatment period was a ma-
jor contributor to this overall effect. Despite the effects
of prepartum treatment on prepartum DMI, milk yield
averaged 46.3 kg/d during the first 63 d postpartum
and was not affected by prepartum treatment (Table
4). Similarly, none of the characteristics of milk com-
position measured in this experiment were affected by
prepartum treatment (Table 4).
Despite decreased DMI for cows fed the low DCAD
diet during the prepartum period, calculated net en-
ergy balance was not different (5.2 vs. 7.5 Mcal/d; P
= 0.12) during the prepartum period (Table 5). Cows
assigned to both prepartum treatments maintained a
substantial positive net energy balance (approximately
148 and 136% of calculated requirements for the con-
trol and low DCAD treatments, respectively) during
the prepartum period (Table 5). Postpartum calculated
net energy balance (Table 5) was not affected by treat-
ment (P > 0.70). Prepartum BW was not affected by
treatment (P = 0.19; Table 5); however, a treatment
by time interaction existed (P = 0.04) for postpartum
BW such that cows fed the low DCAD treatment had
slightly lower BW compared with controls during much
of the postpartum period (Figure 9). Similar to these
effects, BCS of cows fed the low DCAD treatment pre-
partum tended to be decreased (P = 0.11) compared
with controls during the postpartum period (Table 5),
and an interaction of treatment and time (P = 0.02) for

Journal of Dairy Science Vol. 92 No. 11, 2009

 ANION SUPPLEMENTATION OF LOW-POTASSIUM DIETS FOR DRY COWS 5683

postpartum BCS (Figure 10) was comparable to that

measured for BW.
Effects of prepartum anion supplementation on
energy-related metabolites in plasma during the pre-
partum and postpartum periods are presented in Table
6. Overall plasma concentrations of glucose, NEFA,
and BHBA were not affected by treatment during the
prepartum period; however, a trend for a treatment by
time interaction existed (P = 0.11) for plasma NEFA
such that plasma NEFA concentration was slightly
higher in controls during the week before calving. Over-
all concentrations of glucose, NEFA, and BHBA during
the postpartum period were not affected by prepartum
treatment.
A subset of cows in this study was subjected to
measurement of glucose kinetics during the prepartum
and postpartum periods. Glucose rate of appearance
(Figure 11) was measured in control (n = 11) and low
DCAD (n = 9) at (mean ± SD) 9.5 ± 2.6 d and 6.7 ±
3.6 d before actual calving; differences between the 2
groups in day of study relative to calving were statis-
tically significant (P = 0.05). During the postpartum
period, glucose rate of appearance was measured in
8 cows per treatment and on d 9.6 ± 1.2 and 8.1 ±
0.6 postpartum for cows assigned to control and low
DCAD groups, respectively; again, differences between
the 2 groups in day of study relative to calving were
statistically significant (P = 0.01). Glucose rate of ap-
pearance (millimoles per hour per kilogram of MBW)
was higher during the postpartum period (4.85 ± 0.20
and 4.58 ± 0.20 mmol/h per kg of BW0.75, for control
and low DCAD groups, respectively) than during the
prepartum period (3.24 ± 0.10 and 3.08 ± 0.11 mmol/h
per kg of BW0.75, for control and low DCAD groups
respectively), but was not affected by treatment during
either period. Figure 2. Least squares means and standard errors for plasma to-
tal Ca (panel a), P (panel b), and Mg (panel c) concentrations during
the periparturient period for cows fed either a control (----; n = 21)
DISCUSSION or low DCAD (—Δ—; n = 22) diet during the prepartum period. The
effect of day was significant (P < 0.0001) for all variables.
As expected, cows fed the low DCAD diet had lower
urine pH compared with controls during the prepartum
period (Figure 1), reaching the 6.2 to 6.8 range that is ria in ruminants (Fredeen et al., 1988), and an increase
commonly recommended for dry cows fed low DCAD in dietary Ca concentration commonly is recommended
diets in practice (Goff, 2004) and for sufficient duration to improve Ca homeostasis with the use of anionic
(DeGaris and Lean, 2008). This suggests that applica- supplements (Oetzel et al., 1988). The Ca content of
tion of the anionic diet treatment in this experiment both prepartum diets (1.43 vs. 1.35% of DM for control
was successful in inducing changes in systemic acid-base and low DCAD, respectively) can be characterized as
balance that should improve the metabolic response to modestly high relative to the concentration used in
parathyroid hormone signaling under high Ca demand various studies (Oetzel et al., 1988; Goff and Horst,
(Goff and Horst, 2003b). 1997; Chan et al., 2006) and were likely sufficient for
Plasma concentrations of Ca during the prepartum maintenance of Ca homeostasis.
period were not different between cows fed the control As has been observed in numerous studies (Block,
and low DCAD treatments (Table 3). It has been re- 1984; Oetzel et al., 1988; Goff et al., 1991a; Goff and
ported that acidification of plasma induces hypercalciu- Horst, 1998; Moore et al., 2000), the sudden increase
Journal of Dairy Science Vol. 92 No. 11, 2009
5684 RAMOS-NIEVES ET AL.

approximates the 10.7% improvement in plasma Ca

predicted by the equation.
Differences in plasma Ca concentrations beyond the
first 24 h postcalving were not significant in this experi-
ment. Goff et al. (1986) reported that intestinal absorp-
tion of dietary Ca is stimulated and bone mobilization
is significantly activated after 24 and 48 h, respectively,
of parathyroid hormone stimulation during periods of
high Ca demand. Cows assigned to both control and
low DCAD treatment groups averaged normal plasma
Ca (>8 mg/dL) concentrations by 48 h after parturi-
tion.
Feeding a low DCAD prepartum diet did not affect
the incidence of clinical or subclinical hypocalcemia
during the first 24 h postcalving or the 5 d thereafter
(Figure 5). These results are consistent with other re-
Figure 3. Least squares means and standard errors for plasma ports that have suggested that reductions of DCAD of
P concentration during the postpartum period for cows fed either a <30 mEq/100 g of DM can increase plasma concentra-
control (----; n = 21) or low DCAD (—Δ—; n = 22) diet during the
prepartum period. The P-value for the effect of treatment was 0.11,
tions of Ca or ionized Ca but do not affect the incidence
and the P-value for the interaction of treatment and time was 0.05. of parturient paresis (Oetzel et al., 1988; Tucker et al.,
1992; Delaquis and Block, 1995; Moore et al., 2000;
Kurosaki et al., 2007).
Results from this experiment suggest that large dif-
in Ca demand at the onset of lactation reduced plasma ferences (~1.5 units) in prepartum urine pH do not nec-
Ca concentrations at parturition (Figure 2) followed by essarily result in large differences in periparturient Ca
an eventual recovery during the first week of lactation. status. Tucker et al. (1992) and Kurosaki et al. (2007)
Prepartum dietary treatment did not affect overall were effective in decreasing urinary pH in the range of 6
plasma Ca concentrations during the postpartum pe- to 7 and reach acidification by feeding low DCAD diets
riod (Table 3); however, cows fed the low DCAD diet
prepartum tended to have modestly increased plasma
concentrations of Ca during the first 24 h postpartum
(Figure 4).
According to several studies (Oetzel et al., 1988;
Joyce et al., 1997; Moore et al., 2000) and a recent
meta-analysis (Charbonneau et al., 2006), both ionized
Ca and total Ca in plasma increase at parturition when
dietary DCAD is reduced during the prepartum period.
The modest response of plasma Ca during the first 24 h
postcalving and lack of response thereafter likely relates
to the comparatively small difference in dietary DCAD
in this study compared with others in which large re-
sponses of plasma Ca have been measured in response
to decreased prepartum dietary DCAD (Block, 1984;
Goff et al., 1991a; Joyce et al., 1997; Gant et al., 1998;
Goff and Horst, 1998). Furthermore, it is worth noting
that many of these previous studies were conducted
using cows at higher risk for hypocalcemia, such as
multiparous cows and Jerseys, and with control diets
containing higher DCAD than in the current study.
Interestingly, the difference of 12% in total plasma Ca Figure 4. Least squares means and standard errors for plasma Ca
in the first 24 h after calving, with a reduction of 23.3 concentration during the first 24 h postpartum for cows fed either a
mEq/100 g of DM {using the most relevant equation control (n = 21) or low DCAD (n = 22) diet during the prepartum
period. Values represent overall means for plasma samples collected at
[(Na+ + K+) + (Cl− − 0.6S–)] from the meta-analysis approximately 8 and 16 h after parturition. The P-value for the effect
conducted by Charbonneau et al. (2006)} in this study of treatment was 0.07.

Journal of Dairy Science Vol. 92 No. 11, 2009

 ANION SUPPLEMENTATION OF LOW-POTASSIUM DIETS FOR DRY COWS 5685

Figure 5. Frequency of hypocalcemia during the early postpartum period for cows fed either a control (DCAD = 11 mEq/100 g of DM) or
low DCAD (DCAD = −15 mEq/100 g of DM) diet during the prepartum period. a) Proportion of cows categorized as clinically (<5 mg/dL)
hypocalcemic; and b) proportion of cows categorized as subclinically (5–8 mg/dL) hypocalcemic. n = 20 for control group at 3 d; n = 21 for
control group at 8 h, 16 h, 1 d, 2 d, 4 d, 5 d, and low DCAD group at 16 h and 1 d; n = 22 for low DCAD group at 8 h, 2 d, 3 d, 4 d, and 5 d.
Treatment differences were assessed using Fisher’s exact test.

prepartum, but did not improve plasma Ca or reduce
parturient paresis. Similarly, Goff and Horst (1997)
reported that one of their high prepartum DCAD treat-
ment groups with a clearly alkaline urine pH (8.23)
presented higher plasma Ca concentrations during the
first 12 h postcalving compared with treatment groups
with lower urine pH. These concepts are consistent
with the findings of Charbonneau et al. (2006), who
reported that dietary DCAD correlated more highly
with urine pH than with hypocalcemia incidence. Seifi
et al. (2004) analyzed the use of urinary pH during
the prepartum period as a predictive tool to identify
animals at high risk for clinical hypocalcemia during
the postpartum period. The statistically significant,
but numerically low, difference in urinary pH within 48
h prepartum between healthy animals and those that
subsequently presented clinical hypocalcemia (recum-
bent animals 25th percentile = 8.38 vs. nonrecumbent
animals 75th percentile = 8.25) suggests that farm-
specific conditions may contribute to alkalosis that will
predispose cattle to clinical hypocalcemia during the
periparturient period.
Similar to plasma Ca concentrations, plasma P con-
centrations decreased during the immediate peripartu-
rient period (Figure 2) and prepartum P concentrations
were not affected by treatment. However, cows fed the
low DCAD diet during the prepartum period had sig-
nificantly higher plasma P concentrations compared
with controls during the postpartum period (Table 3);
these results were particularly evident during the day
of calving and for the 5 d thereafter (Figure 3). In
addition, the increased plasma P concentrations were
sufficient to result in significant decreases and trends
for significant decreases in clinical and subclinical
hypophosphatemia during the day of calving and 2 d
thereafter (Figure 6).
Most of the studies related to DCAD effect that ob-
served an improved plasma Ca or ionized Ca concentra-
tion did not report differences in P metabolism during
the pre- or postpartum period (Eppard et al., 1996;
Joyce et al., 1997; Gant et al., 1998; Goff and Horst,
1998; Oetzel et al., 1988; Moore et al., 2000). However,
Block (1984) and Goff and Horst (1997) observed im-
proved P status when DCAD was decreased below zero;

Journal of Dairy Science Vol. 92 No. 11, 2009

5686 RAMOS-NIEVES ET AL.

Figure 6. Frequency of hypophosphatemia during early postpartum period for cows fed either a control (DCAD = 11 mEq/100 g of DM) or
low DCAD (DCAD = −15 mEq/100 g of DM) diet during the prepartum period. a) Proportion of cows categorized as clinically (<2 mg/dL)
hypophosphatemic; and b) proportion of cows categorized as subclinically (2–4 mg/dL) hypophosphatemic. n = 20 for control group at 3 d; n
= 21 for control group at 8 h, 16 h, 1 d, 2 d, 4 d, 5 d, and low DCAD group at 16 h and 1 d; n = 22 for low DCAD group at 8 h, 2 d, 3 d, 4 d,
and 5 d. Treatment differences were assessed using Fisher’s exact test.

additionally, Goff and Horst (1997) observed that P
levels during the prepartum period were influenced by
dietary Ca levels and observed a higher P plasma con-
centration in low dietary Ca (~0.5%) compared with
high dietary Ca (~1.5%).
Plasma P levels in this experiment clearly decreased
at parturition and reached a postpartum peak at 2 and
3 d postcalving for control and low DCAD groups, re-
spectively. Interestingly, plasma P decreased in cows
assigned to both treatments around 9 d postpartum;
this phenomenon has been observed in previous work
(Romo et al., 1991; Goff and Horst, 1998; Peterson et
al., 2005), but the physiological basis or significance is
unknown.
Plasma concentrations of P normally result from the
balance between dietary absorption, excretion, and
reabsorption of P; they are not thought to be subject
to hormonal control. Nevertheless, when P and Ca
concentrations decrease at parturition, P excretion is
exacerbated by parathyroid hormone during the first
stage of Ca recovery, followed by recovery of both Ca
and P levels due to an increase in renal and intestinal
absorption induced by higher levels of 1–25-(OH)2D3
(Horst, 1986; Goff, 1999).
Consistent with the patterns of Ca and P in plasma
during the periparturient period, plasma Mg decreased
sharply following parturition (Figure 2), although
the decrease in Mg appeared to lag that of Ca and
P. Prepartum DCAD treatment did not affect plasma
Mg status in this experiment; however, it is interesting
to note that the incidence of clinical and subclinical
hypomagnesemia was generally higher (Figure 7) than
that of hypocalcemia or hypophosphatemia. This oc-
curred despite relatively high concentrations of Mg in
the prepartum diets (0.47 and 0.51% for control and
low DCAD diets, respectively) that were above those
recommended by Goff (2004) to induce ruminal passive
absorption and overcome any possible active transport
impairment by mineral interactions, and concentrations
of Mg in the postpartum diet (0.30%) were above those
recommended by NRC (2001). Based upon the lack of
clinical outcomes, the low plasma Mg was apparently
without consequence; however, the low Mg status in
the face of apparently adequate dietary supply suggests

Journal of Dairy Science Vol. 92 No. 11, 2009

 ANION SUPPLEMENTATION OF LOW-POTASSIUM DIETS FOR DRY COWS 5687

Figure 7. Frequency of hypomagnesemia during early postpartum period for cows fed either a control (DCAD = 11 mEq/100 g of DM) or
low DCAD (DCAD = −15 mEq/100 g of DM) diet during the prepartum period. a) Proportion of cows categorized as clinically (<1.1 mg/dL)
hypomagnesemic; and b) proportion of cows categorized as subclinically (1.1–1.8 mg/dL) hypomagnesemic. n = 20 for control group at 3 d; n
= 21 for control group at 8 h, 16 h, 1 d, 2 d, 4 d, 5 d, and low DCAD group at 16 h and 1 d; n = 22 for low DCAD group at 8 h, 2 d, 3 d, 4 d,
and 5 d. Treatment differences were assessed using Fisher’s exact test.

that either the dietary Mg sources used in this study
had poor bioavailability or the NRC (2001) require-
ments warrant reevaluation.
Supplementing anions to the prepartum diet de-
creased prepartum DMI in this experiment (Table 4).
The decrease in DMI was modest and confined mostly
to the early part of the treatment period (Figure 8).
Furthermore, calculated energy balance (Table 5) re-
mained substantially positive during the prepartum pe-
riod for both treatments. Some researchers (Oetzel and
Barmore, 1993; Horst et al., 1997; Joyce et al., 1997;
Moore et al., 2000) have reported that supplementa-
tion of anionic sources significantly reduced prepartum
intake; however, others have reported an increase or no
effect of anion supplementation on intake during the
prepartum period (Goff and Horst, 1997, 1998; Oetzel
et al., 1988; Vagnoni and Oetzel, 1998). The reduction
of 1.2 kg of DM/d with a DCAD difference between
treatments of 23.3 mEq/100 g DM [using the most rel-
evant equation in Charbonneau’s meta-analysis: (Na+
+ K+) + (Cl− − 0.6S–)] is close to that predicted by
Charbonneau et al. (2006); they predicted a reduction of
1.3 kg/d of DM when DCAD is reduced by 30 mEq/100
g of DM. Anionic supplements based upon hydrochlo-
ric acid, such as the primary supplement used in this
experiment, have been reported to have less effect on
DMI than other anionic salts (Goff and Horst, 2003a;
Goff, 2004); however, results from the current experi-
ment suggest that modest reduction of DMI is possible
with anion supplementation, regardless of source.
Despite the effect of anion supplementation on pre-
partum DMI, DMI and milk yield and composition
during the first 63 DIM were not affected by treatment
(Table 4). Few studies have reported milk production
for cows fed diets with differing DCAD during the pre-
partum period (Block, 1984; Joyce et al., 1997; Moore
et al., 2000; Roche et al., 2002, 2003). Only Block
(1984) reported an increase in milk yield in the cows
supplemented with anions prepartum because of the
high incidence of paresis that affected the cows fed
the cationic diet. When data from the 2-yr crossover
experiment were analyzed, a difference of 6.8% in milk
yield was observed; however, when paretic cows in the
cationic group were compared against their nonparetic

Journal of Dairy Science Vol. 92 No. 11, 2009

5688 RAMOS-NIEVES ET AL.

Table 4. Least squares means for DMI and milk yield and composition for cows fed a control or low DCAD
diet during the prepartum period

Treatment (Trt)1 P-value

Item Control Low DCAD SEM Trt Trt × week

DMI (kg/d)
Prepartum2 15.6 14.4 0.3 0.01 0.48
Postpartum3 22.4 23.0 0.4 0.36 0.67
Milk yield4 (kg/d) 46.2 46.4 1.4 0.93 0.47
3.5% FCM5 46.9 46.5 1.3 0.83 0.63
ECM6 (kg/d) 46.1 45.6 1.2 0.79 0.43
ECM/DMI 2.14 2.11 0.07 0.78 0.85
Fat7 (%) 3.65 3.67 0.12 0.92 0.32
Fat (kg/d) 1.65 1.63 0.05 0.79 0.60
True protein7 (%) 2.90 2.88 0.06 0.81 0.38
True protein (kg/d) 1.31 1.29 0.03 0.67 0.33
Lactose7 (%) 4.68 4.63 0.04 0.35 0.16
Lactose (kg/d) 2.18 2.14 0.07 0.67 0.55
Total solids7 (%) 12.13 12.08 0.16 0.82 0.67
Total solids (kg/d) 5.60 5.54 0.15 0.80 0.42
SCC7 (× 1000 cells/mL) 192 151 42 0.49 0.73
Linear SCS 2.34 2.38 0.25 0.91 0.67
MUN7 (mg/dL) 12.3 12.2 0.4 0.89 0.38
1
Cows assigned to the control treatment (n = 21) were fed a prepartum diet with DCAD value of +11 mEq/100
g of DM and cows assigned to the low DCAD treatment (n = 22) were fed a prepartum diet with a DCAD
value of −15 mEq/100 g of DM.
2
Data collected daily during last 21 d prepartum.
3
Data collected daily from calving day to 63 d postpartum.
4
Data collected daily from calving to 63 d postpartum and then reduced to weekly means for analysis.
5
3.5% FCM = [(kg of milk × 0.4324) + (16.216 × kg of fat)].
6
ECM = [(0.327 × kg of milk) + (12.95 × kg of fat) + (7.2 × kg of protein)].
7
Data collected weekly from wk 1 through 9 postpartum.

Figure 8. Temporal pattern of DMI during the prepartum period

for cows fed a control (DCAD = +11 mEq/100 g of DM) diet or a
low DCAD diet (DCAD = −15 mEq/100 g of DM) during the prepar-
tum period. Values are least squares means and error bars represent
standard errors; n = 21 and 22 for control and low DCAD group,
respectively. The DMI averaged 15.6 and 14.4 kg/d for control and
low DCAD group, respectively. The P-values for the effects of treat-
ment and treatment by day were 0.01 and 0.48, respectively, for the
interaction of treatment and day. Covariate values for DMI averaged
15.7 and 15.2 kg/d for cows fed the control and low DCAD treatments,
respectively, and were not affected by treatment (P = 0.56).
Figure 9. Temporal pattern of BW for cows fed a control (DCAD
= +11 mEq/100 g of DM) diet or a low DCAD diet (DCAD = −15
mEq/100 g of DM) during the periparturient period. Values are least
squares means and error bars represent standard errors; n = 21 and
22 for control and low DCAD groups, respectively. The P-value for the
interaction of treatment and week for the postpartum period was 0.04.
Asterisk indicates within-timepoint difference of P < 0.05.

Journal of Dairy Science Vol. 92 No. 11, 2009

 ANION SUPPLEMENTATION OF LOW-POTASSIUM DIETS FOR DRY COWS 5689
Table 5. Least squares means for energy balance, BW, and BCS for cows fed a control or low DCAD diet
during the prepartum period

Treatment (Trt)1 P-value

Item Control Low DCAD SEM Trt Trt × week

2
Energy balance
Prepartum3 (Mcal/d) 7.5 5.2 1.0 0.12 0.13
Prepartum3 (%) 148.3 135.8 6.8 0.18 0.11
Postpartum4 (Mcal/d) −4.5 −3.9 1.1 0.70 0.89
Postpartum4 (%) 89.7 90.9 2.5 0.73 0.92
BW5
Prepartum (kg) 790.7 783.1 4.1 0.19 0.79
Postpartum (kg) 678.1 662.5 10.2 0.28 0.04
BCS5
Prepartum 3.44 3.44 0.05 0.93 0.10
Postpartum 3.08 2.97 0.05 0.11 0.02
1
Cows assigned to the control treatment (n = 21) were fed a prepartum diet with DCAD value of +11 mEq/100
g of DM and cows assigned to the low DCAD treatment (n = 22) were fed a prepartum diet with a DCAD
value of −15 mEq/100 g of DM.
2
Calculated according to NRC (2001) as net energy balance.
3
Data collected daily for 21 d before parturition and reduced to weekly basis for analysis.
4
Data collected daily from calving to 63 d postpartum and reduced to weekly basis for analysis.
5
Data collected weekly from wk 3 before parturition to wk 9 after parturition.

group mates, a difference of 14% in milk yield was de-
tected. Feeding the low DCAD diet prepartum in this
experiment resulted in slightly better maintenance of
BW and BCS during the postpartum period (Table 5;
Figures 9 and 10). Calculated energy balance (Table
5) during the postpartum period was not affected by
prepartum treatment; therefore, reasons for the differ-
ences are not apparent.
Consistent with the lack of effect of prepartum DCAD
strategy on most aspects of performance, overall con-
centrations of glucose, NEFA, and BHBA in plasma
were not different between treatment groups during the

Figure 10. Temporal pattern of weekly BCS for cows fed a control
(DCAD = +11 mEq/100 g of DM) diet or a low DCAD diet (DCAD
= −15 mEq/100 g of DM) during the periparturient period. Values are
least squares means and error bars represent standard errors; n = 21
and 22 for control and low DCAD groups, respectively. The P-value for
the interaction of treatment and week for the postpartum period was
0.02. Asterisk indicates within-timepoint difference of P < 0.05.
Figure 11. Least squares means and standard errors for glucose
rate of appearance in plasma determined at plateau enrichment of
plasma glucose with dideuterated glucose with plasma samples col-
lected at 165, 180, 195, and 210 min after initiation of continuous
infusion of stable isotope. Prepartum basal glucose kinetics procedures
were performed at 9.5 ± 2.6 d and 6.7 ± 3.6 d for control and low
DCAD treatment groups, respectively; and at 9.6 ± 1.2 d and 8.1 ±
0.6 d for control and low DCAD treatment groups, respectively, during
the postpartum period. The P-values for the effects of treatment were
0.30 and 0.35 for prepartum and postpartum procedures, respectively.

Journal of Dairy Science Vol. 92 No. 11, 2009

5690 RAMOS-NIEVES ET AL.

Table 6. Least squares means for concentrations of energy-related metabolites in plasma from cows fed a
control or low DCAD diet during the prepartum period

Treatment (Trt)1 P-value

Item Control Low DCAD SEM Trt Trt × day

2
Prepartum
Glucose (mg/dL) 66.5 65.9 0.7 0.58 0.91
NEFA (μEq/L) 229 225 34 0.94 0.11
BHBA (mg/dL) 10.1 9.9 0.2 0.59 0.47
Postpartum3
Glucose (mg/dL) 53.4 51.0 1.4 0.23 0.92
NEFA (μEq/L) 581 604 50 0.74 0.75
BHBA (mg/dL) 14.5 15.3 1.2 0.61 0.46
1
Cows assigned to the control treatment (n = 21) were fed a prepartum diet with DCAD value of +11 mEq/100
g of DM and cows assigned to the low DCAD treatment (n = 22) were fed a prepartum diet with a DCAD
value of −15 mEq/100 g of DM.
2
Data collected weekly during the first 21 d prepartum.
3
Data collected at 8 h, 16 h, and d 1, 2, 3, 4, 5, 7, 9, 11, 13, and 15 postpartum.

prepartum and postpartum periods (Table 6). A trend
for a treatment by time interaction for plasma NEFA
during the prepartum period corresponded to the ef-
fects of prepartum DCAD strategy on prepartum DMI,
although differences were small. Moore et al. (2000)
reported a lack of effect of prepartum DCAD (+15, 0,
or −15 mEq/100 g of DM) on plasma NEFA in mul-
tiparous cows. Collectively, these results suggest that
effects of DCAD on circulating NEFA concentrations
are minimal and are confined to situations in which
there are effects on DMI.
This experiment was the first to quantitatively mea-
sure glucose metabolism in dairy cows as affected by
varying the DCAD of the prepartum diet. Not surpris-
ingly, glucose rate of appearance was higher in cows
at 7 to 10 d postpartum compared with 7 to 10 d pre-
partum (Figure 11), likely as a result of the increased
total splanchnic production of glucose measured in
cows during this timeframe (Reynolds et al., 2003) and
decreased insulin-dependent glucose utilization (Bell,
1995). Rates of glucose appearance during the post-
partum period were comparable with those measured
previously in our laboratory in cows at a similar time
point postcalving (5.09 mmol/h per kg of BW0.75 for
saline-infused animals; Waldron et al., 2006). Prepar-
tum anion supplementation did not affect glucose rate
of appearance during the prepartum or postpartum pe-
riods, further supporting the lack of effect of treatment
on aspects of energy metabolism in this experiment.
CONCLUSIONS
Supplementation of anions to a low-K prepartum diet
did not change the incidence of clinical or subclinical
hypocalcemia. Feeding low DCAD diets prepartum in-
creased P plasma concentrations during the pre- and
postpartum periods and, in general, tended to reduce
the incidence of clinical and subclinical hypophos-
phatemia during the first 5 d postpartum. Cows fed the
low DCAD diet had slightly lower prepartum DMI; how-
ever, none of the aspects of postpartum performance or
periparturient energy metabolism studied were affected
by prepartum treatment in this experiment.
ACKNOWLEDGMENTS
The assistance of the following students and staff at
Cornell University (Ithaca, NY) in implementing the
study is gratefully acknowledged: D. Gentile, J. Lukas,
S. Stebulis, R. Ehrhardt, and the staff at the Cornell
University Dairy Teaching and Research Center. We are
very appreciative of the helpful comments of Jesse Goff
during preparation of this manuscript. We gratefully
acknowledge the Consejo Nacional de Ciencia y Tec-
nología de Mexico (CONACYT) for their scholarship
support of the MS program of J. M. Ramos-Nieves and
West Central (Ralston, IA) for their financial support
of this research.
REFERENCES
AOAC. 1990. Official Methods of Analysis. 15th ed. AOAC, Arlington,
VA.
AOAC. 2000. Official Methods of Analysis. 17th ed. AOAC, Arlington,
VA.
Beede, D. K., C. Wang, G. A. Donovan, L. F. Archivald, and W.
K. Sanchez. 1991. Dietary cation-anion difference (electrolyte
balance) in late pregnancy. Page 98 in Florida Dairy Prod. Conf.,
Univ. Florida, Gainesville.
Bell, A. W. 1995. Regulation of organic nutrient metabolism during
transition from late pregnancy to early lactation. J. Anim. Sci.
73:2804–2819.
Bigner, D. R., J. P. Goff, M. A. Faust, J. L. Burton, H. D. Tyler, and R.
L. Horst. 1996. Acidosis effects on insulin response during glucose
tolerance tests in Jersey cows. J. Dairy Sci. 79:2182–2188.
Block, E. 1984. Manipulating dietary anions and cations for prepartum
dairy cows to reduce incidence of milk fever. J. Dairy Sci. 67:2939–
2948.

Journal of Dairy Science Vol. 92 No. 11, 2009

 ANION SUPPLEMENTATION OF LOW-POTASSIUM DIETS FOR DRY COWS
Chan, P. S., J. W. West, and J. K. Bernard. 2006. Effect of prepartum
dietary calcium on intake and serum and urinary mineral
concentrations of cows. J. Dairy Sci. 89:704–713.
Charbonneau, E., D. Pellerin, and G. R. Oetzel. 2006. Impact of
lowering dietary cation-anion difference in nonlactating dairy
cows: A meta-analysis. J. Dairy Sci. 89:537–548.
DeGaris, P. J., and I. J. Lean. 2008. Milk fever in dairy cows: A review
of pathophysiology and control principles. Vet. J. 176:58–69.
Delaquis, A. M., and E. Block. 1995. Dietary cation-anion difference,
acid-base status, mineral metabolism, renal function, and milk
production of lactating cows. J. Dairy Sci. 78:2259–2284.
Eppard, P. J., J. J. Veenhuizen, W. J. Cole, P. G. Comens-Keller, G.
F. Hartnell, R. L. Hintz, L. Munyakazi, P. K. Olsson, R. H. Sorbet,
T. C. White, C. A. Baile, R. J. Collier, J. P. Goff, and R. L. Horst.
1996. Effect of bovine somatotropin administered to periparturient
dairy cows on the incidence of metabolic disease. J. Dairy Sci.
79:2170–2181.
Fredeen, A. H., E. J. DePeters, and R. L. Baldwin. 1988a. Effects of
acid-base disturbances caused by differences in dietary fixed ion
balance on kinetics of calcium metabolism in ruminants with high
calcium demand. J. Anim. Sci. 66:174–184.
Gant, R. G., W. Sanchez, and R. L. Kincaid. 1998. Effect of anionic
salts on selenium metabolism in nonlactating, pregnant dairy cows.
J. Dairy Sci. 81:1637–1642.
Gaynor, P. J., F. J. Mueller, J. K. Miller, N. Ramsey, J. P. Goff,
and R. L. Horst. 1989. Parturient hypocalcemia in jersey cows fed
alfalfa haylage-based diets with different cation to anion ratios. J.
Dairy Sci. 72:2525–2531.
Goering, H. K., and P. J. Van Soest. 1970. Forage Fiber Analysis.
USDA Agricultural Research Service. Handbook number 379.
USDA, Washington, DC.
Goff, J. P. 1999. Treatment of calcium, phosphorus, and magnesium
balance disorders. Vet. Clin. North Am. Food Anim. Pract.
15:619–639.
Goff, J. P. 2004. Macromineral disorders of the transition cow. Vet.
Clin. North Am. Food Anim. Pract. 20:471–494.
Goff, J. P., and R. L. Horst. 1997. Effects of the addition of potassium
or sodium, but not calcium, to prepartum ratios on milk fever in
dairy cows. J. Dairy Sci. 80:176–186.
Goff, J. P., and R. L. Horst. 1998. Use of hydrochloric acid as a source
of anions for prevention of milk fever. J. Dairy Sci. 81:2874–
2880.
Goff, J. P., and R. L. Horst. 2003a. Milk fever control in the United
States. Acta Vet. Scand. Suppl. 97:145–147.
Goff, J. P., and R. L. Horst. 2003b. Role of acid-base physiology on
the pathogenesis of parturient hypocalcaemia (milk fever)—The
DCAD theory in principle and practice. Acta Vet. Scand. Suppl.
97:51–56.
Goff, J. P., R. L. Horst, F. J. Mueller, J. K. Miller, G. A. Kiess,
and H. H. Dowlen. 1991a. Addition of chloride to a prepartal diet
high in cations increases 1,25-dihydroxyvitamin D response to
hypocalcemia preventing milk fever. J. Dairy Sci. 74:3863–3871.
Goff, J. P., E. T. Littledike, and R. L. Horst. 1986. Effect of synthetic
bovine parathyroid hormone in dairy cows: Prevention of
hypocalcemic parturient paresis. J. Dairy Sci. 69:2278–2289.
Goff, J. P., T. A. Reinhardt, and R. L. Horst. 1991b. Enzymes and
factors controlling vitamin D metabolism and action in normal
and milk fever cows. J. Dairy Sci. 74:4022–4032.
Holm, J., I. Bjorck, A. Drews, and N.-G. Asp. 1986. A rapid method
for the analysis of starch. Starch/Die Starke 7:224–226.
Horst, R. L. 1986. Regulation of calcium and phosphorus homeostasis
in the dairy cow. J. Dairy Sci. 69:604–616.
Horst, R. L., J. P. Goff, T. A. Reinhardt, and D. R. Buxton. 1997.
Strategies for preventing milk fever in dairy cattle. J. Dairy Sci.
80:1269–1280.
Joyce, P. W., W. K. Sanchez, and J. P. Goff. 1997. Effect of anionic
salts in prepartum diets based on alfalfa. J. Dairy Sci. 80:2866–
2875.
5691
Kurosaki, N., O. Yamato, F. Mori, S. Imoto, and Y. Maede. 2007.
Preventing effect of mildly altering dietary cation-anion difference
on milk fever in dairy cows. J. Vet. Med. Sci. 69:185–192.
McCutcheon, S. N., and D. E. Bauman. 1986. Effect of chronic growth
hormone treatment on responses to epinephrine and thyrotropin-
releasing hormone in lactating cows. J. Dairy Sci. 69:44–51.
Moore, S. J., M. J. VandeHaar, B. K. Sharma, T. E. Pilbeam, D.
K. Beede, H. F. Bucholtz, J. S. Liesman, R. L. Horst, and J. P.
Goff. 2000. Effects of altering dietary cation-anion difference on
calcium and energy metabolism in peripartum cows. J. Dairy Sci.
83:2095–2104.
NRC. 2001. Nutrient Requirements of Dairy Cattle. 7th rev. ed.
National Academy Press, Washington, DC.
Oetzel, G. R., and J. A. Barmore. 1993. Intake of a concentrate mixture
containing various anionic salts fed to pregnant, nonlactating dairy
cows. J. Dairy Sci. 76:1617–1623.
Oetzel, G. R., J. D. Olson, C. R. Curtis, and M. J. Fettman. 1988.
Ammonium chloride and ammonium sulfate for prevention of
parturient paresis in dairy cows. J. Dairy Sci. 71:3302–3309.
Overton, T. R., and M. R. Waldron. 2004. Nutritional management of
transition dairy cows: Strategies to optimize metabolic health. J.
Dairy Sci. 87(E Suppl.):E105–E119.
Peterson, A. B., M. W. Orth, J. P. Goff, and D. K. Beede. 2005.
Periparturient responses of multiparous Holstein cows fed different
dietary phosphorus concentrations prepartum. J. Dairy Sci.
88:3582–3594.
Reynolds, C. K., P. C. Aikman, B. Lupoli, D. J. Humphries, and D.
E. Beever. 2003. Splanchnic metabolism of dairy cows during the
transition from late gestation through early lactation. J. Dairy
Sci. 86:1201–1217.
Roche, J. R., D. Dalley, P. Moate, C. Grainger, M. Rath, and F.
O’Mara. 2003. A low dietary cation-anion difference precalving
and calcium supplementation postcalving increase plasma calcium
but not milk production in a pasture-based system. J. Dairy Sci.
86:2658–2666.
Roche, J. R., J. Morton, and E. S. Kolver. 2002. Sulfur and chlorine
play a non-acid base role in periparturient calcium homeostasis. J.
Dairy Sci. 85:3444–3453.
Romo, G. A., R. O. Kellems, K. Powell, and M. V. Wallentine. 1991.
Some blood minerals and hormones in cows fed variable mineral
levels and ionic balance. J. Dairy Sci. 74:3068–3077.
Seifi, H. A., M. Mohri, and J. Kalamati Zadeh. 2004. Use of pre-
partum urine pH to predict the risk of milk fever in dairy cows.
Vet. J. 167:281–285.
Tucker, W. B., J. F. Hogue, G. D. Adams, M. Aslam, I. S. Shin, and
G. Morgan. 1992. Influence of dietary cation-anion balance during
the dry period on the occurrence of parturient paresis in cows fed
excess calcium. J. Anim. Sci. 70:1238–1250.
Vagnoni, D. B., and G. R. Oetzel. 1998. Effects of dietary cation-
anion difference on the acid-base status of dry cows. J. Dairy Sci.
81:1643–1652.
Van Soest, P. J., J. B. Robertson, and B. A. Lewis. 1991. Methods for
dietary fiber, neutral detergent fiber, and nonstarch polysaccharides
in relation to animal nutrition. J. Dairy Sci. 74:3583–3597.
Waldron, M. R., A. E. Kulick, A. W. Bell, and T. R. Overton. 2006.
Acute experimental mastitis is not causal toward the development
of energy-related metabolic disorders in early postpartum dairy
cows. J. Dairy Sci. 89:596–610.
Weiss, W. P., H. R. Conrad, and N. R. S. Pierre. 1992. A theoretically
based model for predicting total digestible nutrient values of
forages and concentrates. Anim. Feed Sci. Technol. 39:95–110.
Wildman, E. E., G. M. Jones, P. E. Wagner, H. F. Boman, H. F.
Troutt, and T. N. Lesch. 1982. A dairy cow body condition scoring
system and its relationship to selected production characteristics.
J. Dairy Sci. 65:495–501.
Journal of Dairy Science Vol. 92 No. 11, 2009