J. Dairy Sci.

105:4370–4392
https://doi.org/10.3168/jds.2021-21246
© 2022, The Authors. Published by Elsevier Inc. and Fass Inc. on behalf of the American Dairy Science Association®.
This is an open access article under the CC BY license (http://creativecommons.org/licenses/by/4.0/).

Effects of restricted dietary phosphorus supply during the dry

period on productivity and metabolism in dairy cows
S. Wächter,1 I. Cohrs,2 L. Golbeck,1 T. Scheu,2 K. Eder,3 and W. Grünberg1*
1
Clinic for Cattle, University of Veterinary Medicine Hannover, Foundation, 30173 Hanover, Germany
2
Educational and Research Centre for Animal Husbandry, Hofgut Neumühle, 67728 Münchweiler an der Alsenz, Germany
3
Department of Animal Nutrition and Nutritional Physiology, Justus-Liebig-Universität-Giessen, 35392 Giessen, Germany

ABSTRACT were not observed, but plasma Ca was higher from 1

Phosphorus in bovine nutrition is under ongoing
scrutiny because of concerns with excessive amounts
of P excreted in manure contributing to environmental
pollution. Feeding rations with excessive P content,
however, still remains common practice, particularly
during the transition period, as limited P supply in late
gestation and early lactation is thought to present a
risk for health and productivity of high-yielding dairy
cows. The objectives of this study were to investigate
the effect of restricted P supply during the last 4 wk of
pregnancy on Ca and P homeostasis during the transi-
tion period in high-yielding dairy cows, and to iden-
tify possible effects on metabolism and productivity
throughout the following lactation. Thirty late-preg-
nant multiparous dairy cows were randomly assigned
to either a dry cow diet with low (LP) or adequate P
(AP) content [0.16 and 0.30% P in dry matter (DM),
respectively] to be fed in the 4 wk before calving. After
calving all cows received the same ration with adequate
P content (0.46% P in DM). Blood, milk, and liver
tissue samples were obtained during the dry period
and the following lactation, DM intake (DMI), body
weight, milk production, and disease occurrence were
monitored. Plasma was assayed for the concentrations
of P, Ca, Na, and K, metabolic parameters, and liver
enzyme activities. Liver tissue was analyzed for mineral,
triglyceride, cholesterol, and water contents. Repeated-
measures ANOVA was used to identify treatment, time,
and treatment × time interaction effects. Cows fed LP
had lower plasma P concentrations ([Pi]) than AP
cows during restricted P feeding, reaching a nadir of
1.1 mmol/L immediately before calving. After calving,
plasma [Pi] of LP cows was at or above the level of AP
cows and within the reference range for cattle. Symp-
toms assumed to be associated with hypophosphatemia
Received September 4, 2021.
Accepted February 4, 2022.
*Corresponding author: waltergruenberg@​yahoo​.com
wk before to 1 wk after calving in LP cows, which was
associated with a numerically lower incidence of clinical
and subclinical hypocalcemia in LP cows. Both treat-
ments had a similar 305-d milk yield (12,112 ± 1,298 kg
for LP and 12,229 ± 1,758 kg for AP cows) and simi-
lar DMI. Plasma and liver tissue biochemical analysis
did not reveal treatment effects on energy, protein, or
lipid metabolism. The results reported here indicate
that restricted dietary P supply during the dry period
positively affected the Ca homeostasis of periparturient
dairy cows but did not reveal negative effects on DMI,
milk production, or metabolic activity in the following
lactation. Restriction of P during the dry period was
associated with hypophosphatemia antepartum but
neither exacerbated postparturient hypophosphatemia,
which is commonly observed in fresh cows, nor was as-
sociated with any clinical or subclinical indication of P
deficiency in early lactation.
Key words: hypophosphatemia, calcium, transition
period, metabolic disease
INTRODUCTION
Phosphorus in ruminant nutrition has received
increased attention in the past decade, due to envi-
ronmental concerns about excessive amounts of this
mineral contained in manure. Official recommendations
for dietary P supply in cattle are continuously revisited
with the objective of reducing the P content in bovine
diets to the lowest possible amount without jeopardiz-
ing health and productivity of high-yielding dairy cows.
Many countries are currently implementing regulations
aiming at reducing the amount of P applied onto farm-
land with manure, which unavoidably also affects the
use of P in ruminant nutrition.
In dairy cows P deficiency is widely believed to be
associated with feed intake depression, as well as with
decreased productivity and fertility (Grünberg, 2014).
Furthermore, conditions commonly observed in fresh
cows, such as downer cow syndrome or intravascular

4370
 Wächter et al.: LONG-TERM EFFECTS OF RESTRICTED PHOSPHORUS FEEDING
hemolysis, have been associated with P deficiency and
hypophosphatemia on empirical grounds (Macwilliams
et al., 1982; Ménard and Thompson, 2007). Less well
known is the potentially deleterious effect of restricted
dietary P supply in cattle on the fermentation activ-
ity of the rumen microbiome. A marked reduction of
P content in rumen fluid was found to be associated
with altered microbial protein synthesis and with al-
tered ratios of the different fractions of volatile fatty
acids in rumen fluid (Harder et al., 2015; Köhler et al.,
2021). Another potential mechanism through which P
deprivation may affect the energy metabolism in cattle
is through a direct effect on the liver. Liver function
and the regenerative activity of liver tissue were found
to be impaired in states of P deficiency in humans with
liver disease, and oral supplementation of P in these
patients promoted faster convalescence and reduced the
risk of complications (Baquerizo et al., 2003). The ef-
fect of restricted P supply on liver function of cattle has
scarcely been studied (Grünberg et al., 2009, 2019). An
effect of P on liver function in cattle, as recognized in
humans, would certainly be of relevance, in particular
in transition and early lactating dairy cows, where liver
function disturbances caused by hepatic lipidosis are
common.
Phosphorus balance disorders in dairy cows are con-
sidered to be of particular concern during the transition
period, and consequently feeding rations in excess of
current recommendations to dry and fresh cows is still
common in the dairy industry (Plaizier et al., 2004).
The objective of producers, nutritionists, and veterinar-
ians advising and implementing the supply of P above
recommendations is to mitigate the severity and dura-
tion of the hypophosphatemic episode well recognized
to occur around parturition, and thereby to improve
herd health and productivity.
The metabolic relevance of this postparturient reduc-
tion of the plasma P concentration ([Pi]) in cattle is
currently not well understood. However, little evidence
supports concerns about abiding by currently recom-
mend dietary P contents for dry cow rations. Recent
studies investigating the effects of P deprivation in
transition cows found that feeding diets with P con-
tents well below current recommendations from the
last 4 wk of the dry period and for the first 4 wk of
gestation indeed had negative effects on DMI, health,
and productivity (Puggaard et al., 2014; Grünberg et
al., 2019b). However, restricted P supply during the
dry period was found to stimulate bone mobilization,
resulting in an important and metabolically relevant
release of Ca and P from bone during the last days of
gestation (Cohrs et al., 2018). This was associated with
more stable blood Ca concentrations in periparturient
Journal of Dairy Science Vol. 105 No. 5, 2022
4371
cows and decreased incidence of clinical and subclinical
hypocalcemia (Cohrs et al., 2018).
The objectives of this study were to explore the pos-
sible protracted effects of P deprivation during the dry
period on DMI, milk production, and metabolism in
the following lactation. We hypothesized that limiting
the period of restricted P supply to the last 4 wk of the
dry period would result in the beneficial effect on Ca
homeostasis reported earlier, without impairing liver
function and metabolism in high-yielding dairy cows.
MATERIALS AND METHODS
The study reported here is part of a multi-institu-
tional project, investigating effects of a restricted di-
etary P supply during the dry period on metabolism
and productivity in dairy cows. Whereas other working
groups on this project have investigated short-term ef-
fects on the Ca homeostasis during the periparturient
period, metabolomics, or effects on erythrocyte stabil-
ity (Wächter et al., 2022e), the study reported here em-
phasizes effects of P deprivation during late gestation
on productivity and metabolism during the lactation
following the dry period with restricted P supply.
Ethics Statement
The study was conducted at the Educational and
Research Centre for Animal Husbandry, Hofgut Neu-
mühle, Münchweiler an der Alsenz, Germany. All re-
lated procedures were approved by the Animal Welfare
and Ethics Committee of the government of Coblenz,
Rhineland Palatinate, Germany (permit no. 23-177-
07/G 19-20-008).
Study Design
The study was conceived as a prospective, random-
ized, and controlled study, and was conducted on 30
late-pregnant multiparous Holstein-Friesian dairy cows
entering their second, third, or fourth lactation. Cows of
the research farm expected to calve between November
2019 and April 2020 were blocked by lactation number
(LN) and, within each block, paired by first-lactation
305-d milk yield. One cow of each pair was then ran-
domly assigned to 1 of the 2 experimental treatments,
which were feeding a dry cow ration with either low
(LP) or adequate P content (AP) during the last 4
wk of gestation. Cows were enrolled approximately 6
wk before expected calving and at least 2 wk after dry-
off (Figure 1). The study consisted of different periods
with varying observation and sampling intensities, and
was initiated with a 2-wk acclimation period extending
 Wächter et al.: LONG-TERM EFFECTS OF RESTRICTED PHOSPHORUS FEEDING 4372

Figure 1. Outline of the study design. Each rectangle represents 1 wk of time. During acclimation, cows of both treatments received the same
ration with adequate P content for dry cows. Cows assigned to the adequate P (AP) treatment remained on this ration for the remaining dry
period, whereas cows assigned to the low P (LP) treatment were switched to the P-depleted ration. After calving, both treatments received the
same lactating cow ration with adequate P content for lactating cows. Small solid downward arrows represent times of analysis of the composite
weekly TMR samples. Large downward arrows mark weekly blood sampling times; the dashed arrow indicates monthly blood sampling times.
Bold upward arrows represent liver tissue sampling times T1 (end of acclimation), T2 (1 wk before expected calving), T3 (d 3 postpartum), T4
(d 14 postpartum), and T5 (d 42 postpartum). BL indicates the baseline sampling time point at the end of the acclimation, immediately before
initiating experimental feeding.

from 6 to 4 wk before expected calving. During this
phase, cows of both treatments received the same dry
cow ration with adequate P content (AP diet) and were
introduced to the restricted feeding protocol. Follow-
ing acclimation, AP cows remained on the AP dry cow
ration, but LP cows were switched to the P-deficient
LP dry cow ration until calving. From the moment of
calving, both treatments were fed the same standard
lactating cow ration.
Animal Housing and Milking
Dry and lactating cows were housed in separate areas
of a freestall barn with concrete flooring and manure
scraper; stalls were bedded with rubber mats covered
with a straw-lime mixture. Cows expected to calve
within a matter of hours were moved to individual
calving pens bedded with straw. A calving sensor sys-
tem was used to assist calving watch at night (Moocall
Ltd.). After parturition, cows with a DMI above 5 kg
during the previous 12 h and without apparent signs
of disease were moved to the lactating cow pen within
24 h after calving. Lactating cows were milked twice
daily between 0430 h and 0530 h and between 1530 h
and 1630 h.
Feeding and Experimental Rations
Feed was offered as TMR formulated to meet or
slightly exceed requirements of dry or lactating cows,
with the exception of the P content of the LP dry cow
ration (NRC, 2001). Feed was available ad libitum for
lactating cows but was restricted to 11.5 kg of DM
during the dry period for both treatments. Feed re-
striction in addition to a low-P diet was required to
maintain the daily P intake in LP cows below 20 g
of P/cow. This threshold for the dietary P supply to
dry cows, necessary to achieve a negative P balance,
was determined in a pilot study in preparation of this
experiment (Cohrs et al., 2021). Feed restriction was
implemented through the use of electronic feeding
gates and feed-weigh troughs (RIC System, Hokofarm
Group), programmed to gradually transition dry cows
from ad libitum feeding to a DMI capped at 11.5 kg/d
within the first week of acclimation, as described previ-
ously (Wächter et al., 2022e). Experimental cows were
fed from the feed-weigh troughs from the beginning of
the acclimation period until the sixth week of lacta-
tion, with the exception of the period when cows were
housed in the calving pens. While in these pens, feed
was offered from individual feed bunks, and feed intake
was recorded manually. After the sixth week of lacta-

Journal of Dairy Science Vol. 105 No. 5, 2022

 Wächter et al.: LONG-TERM EFFECTS OF RESTRICTED PHOSPHORUS FEEDING
tion, cows were fed from regular feed bunks accessible
through conventional head gates. Feed was offered once
daily to dry cows and twice daily to milking cows. Feed
intake per cow and day was recorded from the onset of
the study until wk 6 of lactation. Water was available
ad libitum.
The experimental TMR for dry cows was based on
corn silage, pressed beet pulp, hay, and straw. Pelleted
concentrate, specifically formulated for each treat-
ment, was added to obtain the targeted P content of
the ration of each treatment (Supplemental Table S1;
https:​/​/​www​.researchgate​.net/​publication/​359025502​
_Supplementary ​ _ Table ​ _ S1 ​ _ Ration ​ _ composition;
Wächter et al., 2022b). The LP and AP rations were
formulated to contain 0.15% P and 0.35% P in DM,
respectively. Mono-ammoniumphosphate (Windmill
Monamphos FG, Aliphos Rotterdam BV) was used
as P supplement in the AP concentrate to obtain the
targeted dietary P content in the AP ration. Urea was
added to the LP concentrate to equalize the N content
of both diets.
Sample Collection
Feed Samples. Samples of TMR from both experi-
mental rations were collected daily and composited to
weekly samples to determine the mean weekly DM con-
tent of the experimental diets, as described previously
(Wächter et al., 2022e).
Blood Sampling. Blood samples were obtained
twice weekly from the end of the acclimation phase to 6
wk after calving, and once a month thereafter until the
42nd week of lactation. Blood was obtained between
0800 h and 1000 h by puncture of a jugular vein, and
collected in Li-heparin tubes (LH Vacuette, Greiner
Bio-One).
Milk Sampling. Cows were first milked within 1
h of calving. The total solid content of colostrum was
determined with a digital Brix refractometer (Digital
Pocket Refractometer, Model PAL-S, Atago), and the
colostrum yield at the first milking was recorded. Milk
samples for a DHIA equivalent analysis were obtained
once a week throughout the entire lactation. For this
purpose, composited milk samples of an evening and
the following morning milking were analyzed, and the
combined milk yield of these 2 consecutive milkings re-
corded. The daily milk yield at the time of the weekly
milk sampling was used as mean daily milk yield for the
corresponding week of lactation.
Urine Sampling. Samples of spontaneously voided
urine were obtained 3 times a week during the first 6
wk of lactation.
Liver Tissue Collection. Liver tissue specimens
were obtained at the end of the acclimation period
Journal of Dairy Science Vol. 105 No. 5, 2022
4373
(T1), approximately 1 wk before expected calving (T2),
and on d 3 (T3), 14 (T4), and 42 (T5) postpartum
(Figure 1), as described previously (Grünberg et al.,
2019b). Briefly, the procedure, which was conducted
aseptically and under local anesthesia, obtained liver
tissue transcutaneously using an automated biopsy
needle (BIP-Evocore EC2215, Biomed. Instrumente
und Produkte GmbH). Three liver tissue specimens,
each of approximately 0.1 g, were collected at each
sampling time, immediately submerged in liquid nitro-
gen, and stored at −80°C until processing, as described
subsequently.
Body weight was determined on an electronic scale at
the times of liver tissue sampling (T1–T5, Figure 1) and
thereafter in monthly intervals at the times of blood
sampling.
Sample Processing and Analysis
Feed Sample Analysis. The DM of composited feed
samples was determined by oven drying at 100°C for 24
h, and the dietary P content measured by inductively
coupled plasma mass spectrometry. The amount of
fresh feed offered per day was adjusted weekly in case
of change in dietary DM content, to minimize variation
in the dietary P content.
Blood Sample Analysis. Blood in Li-heparin tubes
was centrifuged at room temperature at 1,730 × g for
15 min (Jouan CR422, Thermo Fisher Scientific) within
20 min of collection, and plasma was stored at −21°C
until analysis. Total protein (TP) content in plasma
was measured by refractometry (RHC-200ATC, C &
A Scientific) once a week at the time of harvesting.
All plasma samples were assayed for [Pi] (ammonium
molybdate method, Cobas Mira Plus CC, Hoffmann-
La Roche AG) and total Ca concentration ([TCa],
Arsenazo III method, AU680, Beckman Coulter Inc.).
One weekly sample from each cow was analyzed for the
concentrations of sodium ([Na]) and potassium ([K],
both by direct potentiometry), creatinine ([Crea],
enzymatic method), BHB ([BHB], UV method),
and total bilirubin ([TBil], dichloraniline method).
Furthermore the activities of γ-glutamyl transferase
(GGT, glycyle glycine method) and glutamate dehy-
drogenase (GLDH, UV method with α-ketoglutarate)
were determined. The concentration of nonesterified
fatty acids ([NEFA], acetyl-CoA-synthetase-acetyl-
CoA-oxydase method), and the activity of alkaline
phosphatase (AlP, photometric p-nitrophenyl phos-
phate method; all conducted on Cobas Mira Plus CC,
Hoffmann-La Roche AG) were also determined once a
week until the sixth week of lactation. All analyses were
conducted at the Diagnostic Veterinary Laboratory of
the Clinic for Cattle, University of Veterinary Medi-
 Wächter et al.: LONG-TERM EFFECTS OF RESTRICTED PHOSPHORUS FEEDING
cine Hannover, Foundation (Hanover, Germany), or
a commercial veterinary clinical pathology laboratory
(Synlab Holding Deutschland GmbH, Augsburg, Ger-
many). Intra- and interassay coefficients of variation for
these analyses are provided in Supplemental Table S2
(https:​/​/​www​.researchgate​.net/​publication/​359135082​
_Supplementary​ _ Table​ _ S2​ _ Intra​ - and​ _ interassay​
_coefficients​ _ of​ _ variation​ _ for​ _ plasma​ _ biochemical​
_laboratory​_analyses; Wächter et al., 2022c).
Milk Sample Analysis. Milk samples obtained as
described previously were immediately submitted to a
commercial milk laboratory (Landeskontrollverband
Rheinland-Pfalz-Saar e.V., Bad Kreuznach, Germany),
conducting the DHIA equivalent analysis to determine
the milk fat content (%), milk protein (%), MUN
(mg/L), milk lactose (%), and SCC (cells/mL) via in-
frared analyzer (MilkoScan FT-6000, Foss Analytical).
The linear SCS was calculated by log-transforming the
SCC.
Urine Analysis. Freshly voided urine was visually
inspected for overt discoloration suggestive of hemoglo-
binuria. The concentration of acetoacetate in urine was
determined semiquantitatively with a commercial urine
dipstick assay (Medi-Test Keton, Macherey-Nagel). In
case of suspected hemoglobinuria, based on discolor-
ation of urine, the sample in question was tested for the
presence of blood with a commercial urine dipstick test
(Medi-Test Combi, Macherey-Nagel).
Liver Tissue Analysis. Liver tissue DM was deter-
mined by atmospheric oven drying to constant weight at
85°C, as described earlier (Grünberg et al., 2019b). The
P, K, and Mg content of each sample was determined in
the previously dried specimens by inductively coupled
optical emission spectrometry (Vista Pro, Varian). The
detected emission wavelengths were 185.878 nm for P,
766.491 nm for K, and 279.553 nm for Mg. Both K and
Mg, like P, are predominantly intracellular electrolytes,
and were analyzed to determine whether changes in
liver tissue P content over time were specific to P or
also occurred with other intracellular electrolytes. The
electrolyte contents determined in liver tissue were
expressed as amounts in liver tissue DM (designated
herein as subscript DM), wet weight (subscript WW),
and fat-free liver tissue (subscript FF; Grünberg et al.,
2019b). The liver triacylglycerol (TAG) and cholesterol
(Chol) contents were determined after lipid extraction
from liver tissue specimens with a mixture of n-hexane
and isopropanol (3:2, vol/vol; Hara and Radin, 1978).
Aliquots of the lipid extracts were first concentrated
and then dissolved using a 1:1 mixture of chloroform
and Triton X-100, as described earlier (Gessner et al.,
2015). After vacuum drying at 42°C for 2 h, liver TAG
and Chol contents were measured using enzymatic re-
agent kits (Fluitest CHOL, cat. no. 4241, and Fluitest
Journal of Dairy Science Vol. 105 No. 5, 2022
4374
TG, cat. no. 5741, Analyticon Biotechnologies AG).
The intraday coefficients of variation were 3.3 and
6.2% for TAG and Chol, respectively. All liver tissue
specimens were analyzed on the same day.
Animal Health
Activity and demeanor of study animals were as-
sessed daily, and a physical examination was conducted
weekly on every cow from the beginning of the study
until the sixth week of lactation. Researchers in charge
of monitoring and sampling study cows were not blinded
to treatment for reasons of practicality. Animal obser-
vation was intensified during the periparturient period,
with an emphasis on the early detection of signs of com-
mon periparturient disorders, as previously described
(Wächter et al., 2022e). Health events and related
therapeutic interventions were recorded for every cow
throughout the study. Study animals did not receive
any standard preventive treatments such as drenches or
boluses. Only plain fresh water was offered ad libitum
from large buckets immediately after calving.
For the purposes of this study, we used the following
case definitions and treatment protocols for common
periparturient disorders. Cows with clinical signs sug-
gestive of milk fever such as dullness, coolness to the
touch, and difficulty standing or rising (Constable et
al., 2016a), were immediately tested for blood con-
centration of ionized Ca ([Ca2+]) using a point-of-care
unit (EPOC Host and Reader, Siemens Healthineers)
that was previously validated for this purpose in our
laboratory (Wächter et al., 2022e). Clinical signs, as
mentioned previously, in combination with [Ca2+] below
1.10 mmol/L were classified as clinical hypocalcemia.
Due to close monitoring and early detection of charac-
teristic symptoms, clinical hypocalcemia never reached
stage 2 (recumbency) in the present study (Constable
et al., 2016a). The corresponding treatment protocol
consisted of subcutaneous administration of 11 g of Ca
(Calcitat S50, Livisto) in combination with Ca salts ad-
ministered orally as a bolus (Bovikalc, Boehringer In-
gelheim Vetmedica). Treatment was repeated as needed
until full recovery, defined as resolution of clinical signs
and restoration of normocalcemia (blood [Ca2+] ≥ 1.10
mmol/L). The 2 twice-weekly blood samples following
the last Ca treatment, either subcutaneous or oral,
were excluded from statistical analysis. The diagnosis
of subclinical hypocalcemia in animals included in
this experiment was retrospectively obtained from the
data set generated in a separate study of this proj-
ect, investigating short-term effects of P deprivation
on Ca homeostasis (Wächter et al., 2022e). Subclinical
hypocalcemia in that study was defined as [Ca2+] below
1.10 mmol/L within the first 7 d of lactation, without
 Wächter et al.: LONG-TERM EFFECTS OF RESTRICTED PHOSPHORUS FEEDING
apparent clinical signs consistent with milk fever. No
treatment of any sort was administered in such cases.
The protocol at the research farm defined subclinical
ketosis as the occurrence of an acetoacetate concen-
tration in urine above 2.5 mmol/L, determined semi-
quantitatively as described previously, that was not
associated with any clinical signs (Andersson, 1988).
The corresponding standard treatment consisted of an
oral administration of a single daily dose of 200 mL
of propylene glycol (PG) until resolution of ketonuria.
Ketonuria, as described earlier, in combination with
clinical signs such as dullness, feed intake depression,
and decreased rumen fill or rumen motility, was de-
fined as clinical ketosis (Baird, 1982). Affected cows
were treated with intravenous dextrose (500 mL, 40%
dextrose solution, G-40, B. Braun) in combination with
oral PG, as previously described. Cases unresponsive to
treatment within 24 h were further treated with dexa-
methasone administered intravenously (40 mg/cow,
Rapidexon Albrecht, Dechra Veterinary Products). The
weekly blood sample following the last dextrose admin-
istration was excluded from data analysis. The defini-
tions of clinical and subclinical ketosis, as previously
described, were used for convenience, as this allowed
cow-side identification of animals thought to require
treatment. For the purposes of this study, however, di-
agnoses of clinical and subclinical ketosis were based on
plasma [BHB] greater than or equal to 1.2 mmol/L for
subclinical (without clinical signs) and clinical ketosis
(with clinical signs such as anorexia, decreased rumen
motility, or dullness) rather than acetoacetate in urine.
Uterine disease diagnosed during this study included
retained fetal membranes, metritis, and endometritis.
Retention of fetal membranes was defined as failure to
pass the placenta within 24 h of calving (Risco et al.,
1994). Metritis was defined as fetid uterine discharge
within 21 d postpartum, and endometritis as uterine
discharge occurring after d 21 postpartum, as diagnosed
by vaginal examination (Sheldon et al., 2006). Metri-
tis was further subdivided into nonfebrile and febrile,
depending on rectal temperature greater than 39.5°C.
The standard protocol for the treatment of retained
fetal membranes, as well as metritis and endometritis,
consisted of parenteral administration of amoxicillin
(10 mg/kg of BW i.m. once daily; Duphamox, Zoetis),
in combination with a nonsteroidal anti-inflammatory
drug (meloxicam, 0.5 mg/kg once daily; Meloxidyl, Ceva
Tiergesundheit GmbH) until complete detachment of
fetal membranes or resolution of uterine discharge.
Clinical mastitis was defined as the presence of milk
with abnormal appearance or signs of inflammation of
the mammary gland (Federation, 1999), whereas sub-
clinical mastitis, for the purposes of this study, was
diagnosed retrospectively based on weekly DHIA equiv-
Journal of Dairy Science Vol. 105 No. 5, 2022
4375
alent results, with a threshold for SCC above 100,000
cells/mL (Thorberg et al., 2009).
Gait abnormalities were recorded whenever observed
by farm personnel during their daily routine. Affected
study cows were diagnosed and treated in a foot trim-
ming chute within 24 h of a lameness alert. Causes of
lameness were categorized into digital dermatitis, heel
horn erosion, interdigital lesions, and claw horn lesions
or deformations. Interdigital lesions included interdigi-
tal hyperplasia, interdigital dermatitis, and foot rot.
Digestive tract disorders diagnosed during this study
included feed intake depression of undetermined cause
but unrelated to ketosis, diarrhea, hardware disease,
acute rumen acidosis, abomasal displacement, and
hemorrhagic enteritis. These diagnoses were made by a
veterinarian and treated accordingly.
Other conditions not fitting into any of the afore-
mentioned categories were summarized under “other
diseases” and are also enumerated in Table 1.
Data Analysis
Baseline values of studied parameters were defined as
values obtained at the end of the acclimation period.
Weekly, twice weekly, and monthly sampling times were
retrospectively expressed as sampling times in weeks
relative to calving. The daily DMI was calculated for
each day of the study period from the end of acclima-
tion (−4 wk relative to calving) until 6 wk after calving
from the recorded daily intakes (as-fed basis) and the
DM contents of the composited feed samples of the cor-
responding week.
The milk yield of each cow was studied based on the
weekly recording at the time of the DHIA equivalent
examination. The 305-d milk yield, as well as the aver-
age fat and protein contents for the 305-d milk yield,
were calculated for each cow from the DHIA equivalent
records.
To study a possible effect of treatment with PG on
some of the plasma biochemical parameters included
in this study, the variable PG was introduced in the
data set. The variable PG was set as 1 for weekly blood
samples obtained between the first and within 1 wk of
the last oral treatment with PG, and as 0 for all other
samples.
Statistical Analysis
Unless stated otherwise, results are expressed as
LSM ± SEM or as median and interquartile range for
variables not meeting the assumption of normality. The
statistical significance level was set at P < 0.05. Data
were tested for normal distribution and homogeneity
of variance, and log-transformed whenever necessary
 Wächter et al.: LONG-TERM EFFECTS OF RESTRICTED PHOSPHORUS FEEDING 4376

to achieve normal distribution. Transformed variables
were SCC, plasma [NEFA], [BHB], [TBil], AlP, GLDH,
GGT, liver tissue DM, TAG, and Chol.
Differences between treatments for numerical and
categorical variables were tested with t-test and chi-
squared statistics, respectively. Repeated-measures
ANOVA, with animal identification number as subject,
was used to determine fixed effects of treatment, time,
LN, and the interaction between treatment and time,
with time as repeated factor, using PROC MIXED
(SAS version 9.4, SAS Institute Inc.). The most ap-
propriate covariance structure was chosen based on
the lowest Akaike information criterion. Bonferroni-
adjusted P-values were used to assess differences be-
tween treatments at specific sampling times and differ-
ence between sampling times whenever the F-test was

Table 1. Incidence of health events recorded from 4 wk before to 305 d after calving in absolute terms and
percentage (in parentheses) for cows on adequate (AP) and restricted P supply (LP) during the dry period;
30 multiparous dairy cows were fed a dry cow ration with either AP (n = 15, 0.30% P in DM) or LP (n = 15,
0.16% P in DM) during the last 4 wk of gestation

Incidence, n (%)

Item   Designation AP LP
1
Hypocalcemia Subclinical 10 (67) 8 (53)
Clinical2 3 (20) 1 (7)
Ketosis Subclinical3 4 (27) 3 (20)
Clinical4 1 (7) 5 (33)
Uterine disease Retained fetal membranes5 3 (20) 1 (7)
Metritis I6 2 (13) 1 (7)
Metritis II7 2 (13) 1 (7)
Endometritis8 1 (7) 1 (7)
Udder health9 Subclinical10 3 (20) 4 (27)
Clinical11 6 (40) 1 (7)
Indigestion12,13   6 (40) 4 (27)
Lameness Digital dermatitis 12 (80) 11 (73)
Heel horn erosion 12 (80) 10 (67)
Interdigital lesions14 2 (13) 4 (27)
Claw horn lesions or deformations 13 (87) 12 (80)
Ovarian cyst15   4 (27) 5 (33)
Other16   5 (33) 4 (27)
1
Subclinical hypocalcemia was defined as blood [Ca2+] <1.1 mmol/L within the first 7 d of calving without
apparent clinical signs.
2
Clinical hypocalcemia was defined as blood [Ca2+] <1.10 mmol/L with clinical signs.
3
Subclinical ketosis was defined as plasma [BHB] >1.2 mmol/L without apparent clinical signs.
4
Clinical ketosis was defined as plasma [BHB] >1.2 mmol/L with clinical signs.
5
Retention of fetal membranes was defined as failure to pass the placenta within 24 h of calving.
6
Metritis I was defined as fetid uterine discharge within 21 d postpartum with rectal temperature ≤39.5°C.
7
Metritis II was defined as fetid uterine discharge within 21 d postpartum with rectal temperature >39.5°C.
8
Endometritis was defined as uterine discharge occurring after d 21 postpartum with rectal temperature
≤39.5°C.
9
One LP cow was affected twice by subclinical mastitis; 1 AP cow was affected twice and 1 AP cow 3 times
by clinical mastitis.
10
Subclinical mastitis was defined as SCC >100,000 cells/mL.
11
Clinical mastitis was defined as presence of milk with abnormal appearance or signs of inflammation of the
mammary gland.
12
Indigestion included feed intake depression of undetermined cause but unrelated to ketosis (AP n = 4, LP n =
1), diarrhea (AP n = 2), hardware disease (AP n = 2, LP n = 1), acute rumen acidosis (LP n = 1), abomasal
displacement (AP n = 1), and hemorrhagic enteritis (LP n = 1).
13
One AP cow was affected consecutively by feed intake depression, abomasal displacement, hardware disease,
and diarrhea.
14
Interdigital lesions included interdigital hyperplasia, interdigital dermatitis, and foot rot.
15
Follicular cyst was defined as cystic structure on 1 ovary with a diameter of at least 2.5 cm, diagnosed ultra-
sonographically at 2 consecutive exams at least 10 d apart.
16
“Other” included unspecific fever (AP n = 1, LP n = 2), pneumonia (AP n = 1, LP n = 1), swelling of head
and neck of unknown cause (AP n = 1), actinomykosis (LP n = 1), intramammary bleeding (AP n = 1),
noninflammatory udder edema (AP n = 1), tibial nerve paralysis (AP n = 1), and non-periparturient clinical
hypocalcemia (AP n = 1).

Journal of Dairy Science Vol. 105 No. 5, 2022

 Wächter et al.: LONG-TERM EFFECTS OF RESTRICTED PHOSPHORUS FEEDING
statistically significant. The same model including the
variable PG was run to explore the effect of PG treat-
ment on metabolic parameters.
Pearson or Spearman correlation analyses and mul-
tiple stepwise regression analyses were conducted to
identify possible associations between plasma and liver
biochemical parameters obtained at the same time.
Correlation and regression analyses were conducted
for each sampling time separately. Each regression
model was checked for variance inflation by screening
tolerance, variance inflation factors, and eigenvalue in
combination with the condition index of each variable
included in the model. Cut-off values for variance of
inflation and tolerance used were 10 and 0.1, respec-
tively. In case of suspected multicollinearity between
some variables, the variable potentially affected by col-
linearity with the lowest coefficient of determination
(R2) was removed and the analysis rerun to obtain the
final model.
The required sample size for this study was estimated
on the basis of results obtained from an earlier study
investigating the effects of dietary P deprivation in
transition dairy cows (Grünberg et al., 2019a,b). We
anticipated a difference for one of the key outcome
variables, plasma [TCa], of approximately 0.4 mmol/L
between treatments around parturition with a standard
deviation of 10%, and furthermore expected a 15%
dropout rate due to early periparturient disease poten-
tially exacerbated by P deprivation. Fifteen cows by
treatment provided 80% power while controlling for a
5% type I error to identify the previously mentioned
effect size. Similarly, the smallest effect sizes on daily
DMI and milk yield with a statistical power of 0.8 and
an α of 0.5 detectable with the sample size of this study
were 15% and 10%, respectively. Retrospective power
analyses were conducted using PROC POWER and
PROC GLMPOWER to assess the statistical power of
the analysis of DMI and milk yield data. All analyses
were conducted with SAS software (SAS version 9.4,
SAS Institute Inc.).
RESULTS
Animals and Feeding
Respectively 7 and 8 AP and LP cows included in
this study entered the second lactation; 5 and 4 cows
the third; and 3 cows of each treatment the fourth lac-
tation. The average P content of experimental rations
determined throughout the study were 0.30 ± 0.05%
and 0.16 ± 0.01% P (mean ± SD) in DM for dry cow
ration for AP and LP cows, respectively. The mean P
content of the lactation cow ration was 0.46% P in DM.
Journal of Dairy Science Vol. 105 No. 5, 2022
4377
Health Events
Health events are reported in a descriptive manner
only. Two cows of the LP treatment did not complete
the study. One cow was found dead unexpectedly,
without prior signs of disease, in the fifth week of lac-
tation. Postmortem examination did not reveal an ap-
parent cause of death. This animal was retrospectively
eliminated from the study and replaced, because a
late-pregnant animal of the same LN and similar first-
lactation milk yield was available in the herd. Another
LP cow developed hemorrhagic enteritis, deteriorated
despite of symptomatic and supportive treatment, and
died within a matter of hours on d 125 of lactation.
Acute rumen acidosis was diagnosed in 1 LP cow on d 3
postpartum, after engorging 22 kg of DM immediately
after calving, when switched from restricted to ad libi-
tum feeding at calving. This animal required intensive
therapy but eventually made a full recovery; data from
this cow obtained between calving and the sixth week
of lactation were excluded from the data set. One AP
cow was diagnosed with abomasal displacement to the
left on d 10 of lactation, immediately underwent sur-
gery, and recovered uneventfully, and thus remained in
the study. The occurrence of clinical and subclinical
hypocalcemia in experimental cows has been reported
previously (Wächter et al., 2022e). Briefly, 13 of 15 AP
cows, and 9 of 15 LP cows were diagnosed with peri-
parturient hypocalcemia. Stage 1 clinical hypocalcemia
was diagnosed and treated in 3 AP and 1 LP cows
(Constable et al., 2016a). Further, 10 AP and 8 LP
cows were diagnosed with subclinical hypocalcemia as
defined earlier (Table 1; Wächter et al., 2022e). One
AP cow previously diagnosed with subclinical hypo-
calcemia during the first week of lactation developed
clinical hypocalcemia on d 95 of lactation and required
parenteral and oral Ca supplementation. However, this
incident was not counted toward the cases of hypocal-
cemia as it did not occur in the periparturient period,
but was rather included in the category of “other dis-
eases” (Table 1).
The occurrence of other health events, including ke-
tosis, uterine, udder, and reproductive tract disease,
and lameness is summarized in Table 1.
Comparing the on-farm diagnosis of clinical ketosis,
based on urine acetoacetate concentration (urine dip-
stick), with the presumed gold standard—that is, the
use of plasma [BHB]—the urine dipstick produced 6 (1
AP, 5 LP) true positive, 1 false positive (1 AP), and
0 false negative diagnoses for clinical ketosis. For the
diagnosis of subclinical ketosis, 7 true positive (4 AP, 3
LP), 7 false positive (4 AP, 3 LP), and 0 false negative
diagnoses were produced when categorizing based on
 Wächter et al.: LONG-TERM EFFECTS OF RESTRICTED PHOSPHORUS FEEDING 4378

urine acetoacetate concentration. Overall using urine
dipsticks with the threshold of 2.5 mmol/L acetoac-
etate in urine overestimated the incidence of subclinical
ketosis by approximately 50%. Accordingly, 8 cows (5
AP and 3 LP) of this study must be considered as hav-
ing received PG without indication.
None of the study animals developed clinical signs
often thought to be associated with P deprivation,
such as recumbency unresponsive to treatment with Ca
salts, or overt hemoglobinuria.
0.0001) and LN effects (P = 0.0002), but no treatment,
treatment × time interaction, or treatment × LN inter-
action effects, were identified. The mean body weight
of both treatments remained constant until calving and
declined thereafter, reaching a nadir at 6 wk after calv-
ing. Cows of LN 2 (663 ± 13 kg) had a lower body
weight than cows of LN 3 (762 ± 17 kg, P = 0.0002)
and LN 4 (736 ± 21 kg, P = 0.01).
Milk Yield and Milk Composition

Dry Matter Intake and Body Weight
The DMI from 4 wk antepartum to 6 wk postpartum,
stratified by treatment, is presented in Figure 2. Time
(P < 0.0001) and LN (P = 0.03), but no treatment,
treatment × time, or treatment × LN interaction ef-
fects, were identified. Feed consumption was identical
for both treatments during the dry period with re-
stricted access to feed until calving. From the nadir at
calving, the DMI increased continuously in both treat-
ments at a similar pace, reaching the highest values 6
wk after calving, when recording of DMI was discon-
tinued (Figure 2). The reported LN effect did reflect a
higher average daily DMI of LN 4 (14.8 ± 0.2 kg) than
LN 3 (13.8 ± 0.2 kg, P = 0.009). The statistical power
determined with retrospective power analysis based on
the observed effect size and degree of variation with an
α of 0.5 was only 0.06.
The body weights of cows over time, stratified by
treatment, are presented in Figure 2. Time (P <
Daily milk production, represented by weekly mea-
surements throughout the first 305 d of lactation
stratified by treatment, as well as by treatment and
LN, is depicted in Figure 3. Milk yield was subject to
statistically significant time (P < 0.0001), LN (P =
0.005), and treatment × LN (P = 0.002) effects. Nei-
ther treatment nor treatment × time interaction effects
were statistically significant. The daily milk yield rose
at a similar pace in both treatments, reaching a peak in
the range of 49 kg/d in the ninth week of lactation, and
declining thereafter at a similar pace to values in the
range of 30 kg/d in the 45th week of lactation (Figure
3). As expected, the average daily milk yield of cows
of LN 4 was higher than that of cows of LN 2 (42.6 ±
1.1 kg vs. 38.1 ± 0.7 kg, P = 0.002). When stratified
by lactation, LP cows in their second lactation had
higher average daily milk yield than AP cows (39.9
± 1.0 kg/d vs. 36.4 ± 1.0 kg/d, P = 0.02), whereas
for the fourth lactation, LP cows produced less milk
than AP cows (39.1 ± 1.5 kg/d vs. 46.1 ± 1.5 kg/d,

Figure 2. LSM ± SEM of DMI (a, left panel) and body weight (b, right panel) for adequate P (AP, blue, dashed line, open squares) and low
P (LP, red, solid line, closed circles) treatments. The vertical dashed lines mark parturition. Thirty multiparous dairy cows were fed a dry cow
ration with either adequate (AP, n = 15, 0.30% P in DM) or low P content (LP, n = 15, 0.16% P in DM) during the last 4 wk of gestation. BL
indicates the baseline sampling time point at the end of the acclimation, before initiating experimental feeding. Time points of both treatments
are slightly offset on the time axes to improve readability.

Journal of Dairy Science Vol. 105 No. 5, 2022

 Wächter et al.: LONG-TERM EFFECTS OF RESTRICTED PHOSPHORUS FEEDING 4379

P = 0.003, Figure 3). None of the cows in LN 2 were .net/​publication/​359025854​_Supplementary​_Table​_S3​

treated with oral or parenteral Ca around calving. Of 9
cows in LN 3 (5 AP and 4 LP), 1 LP cow received Ca
treatment in the first week of lactation. Three cows of
each treatment were in LN 4, of which 2 AP but no LP
cow were treated with Ca around calving. Treatment
with Ca was, on average, administered on d 3 postpar-
tum (range: d 1 to d 7 postpartum). The mean 305-d
milk yield was 12,112 ± 1,298 kg for LP and 12,229
± 1,758 kg for AP cows (mean ± SD). The retrospec-
tive power analysis conducted for the t-test comparing
the mean 305-d milk production yielded a power of
0.58 when controlling for a 5% type I error. For the
repeated-measures ANOVA, a power of only 0.5 was
determined based on the effect size and degree of varia-
tion observed in this study, with an α of 0.5. Results
of the weekly DHIA equivalent analysis are provided
in Supplemental Table S3 (https:​/​/​www​.researchgate​
_Milk​_composition​#fullTextFileContent; Wächter et
al., 2022d). Milk fat and lactose contents showed time
(P < 0.0001) and LN effects (P = 0.0003 and P = 0.03,
respectively) but no treatment, treatment × time, or
treatment × LN interaction effects. Milk fat content
of second-lactation cows (4.2 ± 0.1%) was higher than
of third- and fourth-lactation cows (4.0 ± 0.1%, P =
0.02 and 3.8 ± 0.1%, P < 0.0001, respectively). A time
effect (P < 0.0001) but no treatment, LN, treatment ×
time, or treatment × LN interaction effects were identi-
fied for milk protein content and MUN (Supplemental
Table S3).
The linear SCS showed time (P = 0.005), LN (P
= 0.02), and treatment × LN interaction effects but
no treatment or treatment × time interaction effects.
Cows of LN 4 had higher linear scores than cows of LN
3 (1.92 ± 0.06 vs. 1.69 ± 0.05, P = 0.007). The average

Figure 3. LSM ± SEM of milk yield of all study animals (a, upper left panel); mean ± SD milk yield of second-lactation cows (b, upper right
panel), third-lactation cows (c, lower left panel), and fourth-lactation cows (d, lower right panel) for adequate P (AP, blue, dashed line, open
squares) and low P (LP, red, solid line, closed circles) treatments. LN = lactation. Thirty multiparous dairy cows were fed a dry cow ration with
either adequate (AP, n = 15, 0.30% P in DM) or low P content (LP, n = 15, 0.16% P in DM) during the last 4 wk of gestation. Time points of
both treatments are slightly offset on the time axes to improve readability.

Journal of Dairy Science Vol. 105 No. 5, 2022

 Wächter et al.: LONG-TERM EFFECTS OF RESTRICTED PHOSPHORUS FEEDING
linear score of second-lactation cows was higher in AP
compared with LP cows (1.89 ± 0.06 vs. 1.61 ± 0.05, P
= 0.001). For the fourth lactation, AP cows had lower
linear score than LP cows (1.67 ± 0.08 vs. 2.17 ± 0.08;
P = 0.0004).
No treatment, LN, or treatment × LN effects were
identified for the colostrum yield and total solid con-
tent. The AP cows had 8.4 ± 1.4 kg and LP cows 6.8
± 1.4 kg of colostrum. The total solid contents in co-
lostrum were 25.4 ± 1.4% and 27.8 ± 1.3% for AP and
LP cows, respectively. The retrospective power analysis
yielded 86% power while controlling for a 5% type I
error for the observed effect size.
Plasma Biochemical Analysis
Minerals. Plasma [Pi] time curves stratified by
treatment throughout the study are presented in Figure
4a. Plasma [Pi] showed treatment (P = 0.001), time (P
< 0.0001), LN (P = 0.0009), and treatment × time in-
teraction effects (P < 0.0001). The LP cows had lower
[Pi] than AP cows at wk −4 and from wk −2 until
calving. Mean values were below the reference range of
1.4 mmol/L from wk −3 to wk −0.5 in LP cows, and
around calving in AP cows (Constable et al., 2016b).
Cows of LN 2 had higher [Pi] (1.7 ± 0.0 mmol/L) than
cows of LN 3 and LN 4 (1.6 ± 0.0 mmol/L, P = 0.001).
Plasma [TCa] time curves stratified by treatment are
presented in Figure 4b. This parameter revealed treat-
ment (P = 0.01), time (P < 0.0001), LN (P = 0.04),
and treatment × time interaction effects (P = 0.0004).
The LP cows had higher plasma [TCa] than AP cows
from wk −1 to wk +1. Fourth-lactation cows had lower
[TCa] compared with second-lactation cows (2.45 ±
0.01 mmol/L vs. 2.50 ± 0.01 mmol/L, P = 0.004).
Values for AlP showed time (P < 0.0001), LN (P
= 0.0008), and treatment × time interaction (P =
0.007) effects. Values determined between wk −4 to
wk +1 were higher than values between wk +2 and
wk +6 in both treatments (Supplemental Figure S1;
https:​/​/​www​.researchgate​.net/​publication/​359025856​
_Supplementary​_Figure​_S1​_Alkaline​_Phosphatase;
Wächter et al., 2022a). The LP cows had higher AlP
activities (169 U/L, interquartile range: 127–221 U/L)
compared with AP cows (122 U/L, 95–143 U/L, P =
0.01) in wk −1 (Supplemental Figure S1). Furthermore,
cows in their second lactation had higher AlP activities
(135 IU/mL, 97–166 IU/mL) compared with third- and
fourth-lactation cows (85 IU/mL, 71–117 IU/mL, P
= 0.0003, and 99 IU/mL, 80–122 IU/mL, P = 0.01,
respectively).
Energy Metabolism. Plasma [NEFA] and [BHB]
are presented in Figure 5. For both parameters a statis-
Journal of Dairy Science Vol. 105 No. 5, 2022
4380
tically significant time effect (P < 0.0001) and a treat-
ment × LN interaction effect were observed for [BHB],
with higher concentrations in second-lactation LP
cows (0.67 mmol/L, 0.52–0.87 mmol/L), than second-
lactation AP cows (0.57 mmol/L, 0.45–0.71 mmol/L,
P = 0.03).
Parameters Related to Liver Function and
Liver Injury. The values for plasma [TBil] and ac-
tivities of GGT and GLDH are presented in Figure 6.
For these parameters, a time effect (all P < 0.0001)
but no treatment effect was determined. For [TBil],
a sudden and marked increase was apparent around
calving, which was numerically more pronounced in
AP than LP cows, and values then gradually returned
to precalving levels over the first 10 wk of lactation
(Figure 6). The activities of GGT increased continu-
ously in both treatments from calving to the end of the
observation period. Analysis of GLDH revealed LN (P
= 0.0005) and treatment × LN interaction effects (P
= 0.02). Activities of GLDH were higher in third- than
in second-lactation cows (11.1 IU/L, 7.4–21.0 IU/L, vs.
17.5 IU/L, 11.4–28.6 IU/L, P = 0.0002), and, within
the third lactation, LP cows had higher GLDH activi-
ties than AP cows (18.1 IU/L, 10.1–26.5 IU/L, vs. 8.7
IU/L, 6.8–13.5 IU/L, P = 0.001).
Treatment with PG in the days before blood sam-
pling was associated with higher plasma [TBil] com-
pared with samples obtained from untreated animals
(P = 0.005). The values for plasma [TBil] and activity
of GLDH were within or slightly above the reference
range for cattle at all sampling times and for both
treatments, whereas values for the activity of GGT
were above the reference range throughout the study in
cows of both treatments (Kraft and Dürr, 2014; Con-
stable et al., 2016b).
Parameters Related to Water Balance and
Protein Metabolism. The concentration time curves
for plasma [TP], [Crea], [Na], and [K], stratified by
treatment, are presented in Figure 7. For all param-
eters, time (P = 0.002 for [Na], P < 0.0001 for the
remaining parameters) but no treatment effects were
identified. The [TP] showed a decline over the last 4 wk
of gestation in the range of 10 g/L, reaching a nadir at
calving, followed by a continuous increase over the first
6 wk of lactation. The development of plasma [Crea]
mirrored that of [TP], with a continuous increase until
calving, followed by a continuous decline, reaching a
nadir below the prepartum level at 10 wk of lactation.
Values of [Na] and [K] showed continuous decline from
calving to the end of the observation period. For [Na]
a treatment × LN interaction (P = 0.0001) effect was
identified. Although the average [Na] of cows of LN
3 was higher in LP cows (144.1 ± 0.7 mmol/L) than
 Wächter et al.: LONG-TERM EFFECTS OF RESTRICTED PHOSPHORUS FEEDING 4381

AP (140.8 ± 0.6 mmol/L, P = 0.001), for LN 4 lower The values for plasma [TP], [Crea], [Na], and [K]
[Na] was measured in LP cows (140.0 ± 0.8 mmol/L) were within or slightly below the reference range for
compared with AP cows (144.6 ± 0.8 mmol/L, P = cattle at all sampling times and for both treatments
0.0003). (Constable et al., 2016b).

Figure 4. LSM ± SEM of plasma inorganic P ([Pi]; a, upper panel) and plasma total Ca concentrations ([TCa]; b, lower panel) for adequate
P (AP, blue, dashed line, open squares) and low P (LP, red, solid line, closed circles) treatments. Thirty multiparous dairy cows were fed a dry
cow ration with either adequate (AP, n = 15, 0.30% P in DM) or low P content (LP, n = 15, 0.16% P in DM) during the last 4 wk of gestation.
BL indicates the baseline sampling time point at the end of the acclimation, before initiating experimental feeding. The horizontal dotted lines
represent the lower reference ranges for plasma [Pi] and [TCa] in cattle (Constable et al., 2016b). The vertical dashed lines mark parturition.
Time points labeled with an asterisk differ statistically significantly between treatments (P < 0.05). Time points of both treatments are slightly
offset on the time axes to improve readability.

Journal of Dairy Science Vol. 105 No. 5, 2022

 Wächter et al.: LONG-TERM EFFECTS OF RESTRICTED PHOSPHORUS FEEDING 4382

Liver Tissue Biochemical Analysis
Liver tissue TAG, Chol, and DM at the different
sampling times are presented in Figure 8. For all 3 pa-
rameters, time (P < 0.0001) but no treatment or treat-
ment × time interaction effects were identified. The
liver TAG increased in both treatments from below 1%
at T1 and T2 to values in the range of 5% and 12% TAG
in liver wet weight at the sampling times T3 and T4,
respectively. A parallel but more moderate increase was
observed for liver DM, from a mean value below 28% at
T1 and T2 to mean peak values in the range of 34% of
liver wet weight at T4. A more moderate increase was
observed in both treatments for liver Chol, from values
in the range of 2 mg/g liver wet weight before calving
to peak values around 4 mg/g liver wet weight at T4
(Figure 8).
The P, Mg, and K contents, expressed as amounts
in liver DM and fat-free wet weight at the different
sampling times, are presented in Figure 9. Again, a
time effect (P < 0.0001) but neither treatment nor
treatment × time interaction effects were observed.
Values of PDM, MgDM, and KDM declined in the range
of 18 to 25% between T3 and T4 in both treatments
and increased again moderately by 13 to 16% between
T4 and T5. A similar although more moderate pattern
was observed for PWW, MgWW, and KWW, with declines
of mineral contents between T3 and T4 in the range of
13 to 19% (data not shown), and for PFF, MgFF, and
KFF with declines of 4 to 12% in the same time interval
(Figure 9).
The correlation analysis between liver tissue param-
eters, stratified by sampling time, revealed strong posi-
tive associations between PDM, KDM, and MgDM (rho >
0.90, P < 0.0001), and more moderate associations be-
tween PWW, KWW, and MgWW (rho > 0.68, P < 0.0001)
and PFF, KFF, and MgFF (rho > 0.66, P < 0.0001) at
sampling times T3 to T5.
As liver TAG increased, the association of liver TAG
with liver DM increased from no association at T1 and
T2, over moderate associations at T3 (rho = 0.44, P =
0.02), to strong associations at T4 (rho = 0.64, P =
0.0002) and T5 (rho = 0.65, P = 0.0002). With increas-
ing liver TAG, the negative association of liver TAG
with PDM rose from rho = −0.48 (P = 0.008) at T3, over
rho = −0.85 (P < 0.0001) at T4, to rho = −0.79 (P <
0.0001) at T5. Similar developments were observed for
the associations of liver TAG with KDM (rho = −0.42,
P = 0.02; rho = −0.88 and rho = −0.74, both P <
0.0001) and MgDM (rho = −0.42, P = 0.02; rho = −0.89
and rho = −0.77, both P < 0.0001) at sampling times
T3, T4, and T5, respectively, increasing in strength over
time. With increasing correlation of liver TAG with
liver DM, a concomitant negative association of PDM,
KDM, and MgDM with liver DM also became apparent
at T4 and T5.

Figure 5. Plasma nonesterified fatty acids ([NEFA], left panel) and BHB concentrations ([BHB], right panel) for adequate P (AP, blue
striped boxes) and low P (LP, red open boxes) treatments. Thirty multiparous dairy cows were fed a dry cow ration with either adequate (AP,
n = 15, 0.30% P in DM) or low P content (LP, n = 15, 0.16% P in DM) during the last 4 wk of gestation. Box and whisker plots represent
median (horizontal line), lower and upper quartiles (bottom and top of box, respectively), and 10th and 90th percentiles (lower and upper ends
of whiskers, respectively). BL indicates the baseline sampling time point at the end of the acclimation, before initiating experimental feeding.
The horizontal dotted lines represent the upper reference ranges for plasma [NEFA] and [BHB] in cattle (Constable et al., 2016b). The vertical
dashed lines mark parturition.

Journal of Dairy Science Vol. 105 No. 5, 2022

 Wächter et al.: LONG-TERM EFFECTS OF RESTRICTED PHOSPHORUS FEEDING 4383

Figure 6. Plasma total bilirubin concentration ([TBil]; a, upper panel) and activity of γ-glutamyl transferase (GGT; b, middle panel) and
glutamate dehydrogenase (GLDH; c, lower panel) for adequate P (AP, blue striped boxes) and low P (LP, red open boxes) treatments. Thirty
multiparous dairy cows were fed a dry cow ration with either adequate (AP, n = 15, 0.30% P in DM) or low P content (LP, n = 15, 0.16% P
in DM) during the last 4 wk of gestation. Box and whisker plots represent median (horizontal line), lower and upper quartiles (bottom and top
of box, respectively), and 10th and 90th percentiles (lower and upper ends of whiskers, respectively). BL indicates the baseline sampling time
point at the end of the acclimation, before initiating experimental feeding. The horizontal dotted lines represent the reference ranges for plasma
[TBil], GGT, and GLDH in cattle (Kraft and Dürr, 2014; Constable et al., 2016b). The vertical dashed lines mark parturition.

Journal of Dairy Science Vol. 105 No. 5, 2022

 Wächter et al.: LONG-TERM EFFECTS OF RESTRICTED PHOSPHORUS FEEDING
Figure 7. LSM ± SEM of plasma total protein ([TP]; panel a),
plasma creatinine ([Crea]; panel b), plasma sodium ([Na]; panel c),
and plasma potassium ([K]; panel d) concentrations for adequate P
(AP, blue, dashed line, open squares) and low P (LP, red, solid line,
closed circles) treatments. Thirty multiparous dairy cows were fed a
dry cow ration with either adequate (AP, n = 15, 0.30% P in DM)
or low P content (LP, n = 15, 0.16% P in DM) during the last 4 wk
of gestation. BL indicates the baseline sampling time point at the
end of the acclimation, before initiating experimental feeding. The
horizontal dotted lines represent the reference ranges for plasma [TP],
[Na], and [K] and the lower reference range for plasma [Crea] in cattle
(Constable et al., 2016b). The vertical dashed lines mark parturition.
Time points of both treatments are slightly offset on the time axes to
improve readability.
Journal of Dairy Science Vol. 105 No. 5, 2022
4384
Association of Parameters Related to Liver Tissue
With Plasma Biochemical Parameters
Plasma [Pi] was not associated with any parameter
determined in liver tissue, other than liver TAG at T4
(rho = −0.38, P = 0.04). Correlations were identified
between the liver P content and activities of GLDH and
GGT in plasma at different sampling times. Specifi-
cally, GLDH was negatively associated with PDM (rho
= −0.51, P = 0.04), PWW (rho = −0.42, P = 0.02), and
PFF (rho = −0.43, P = 0.02), as well as with KDM (rho
= −0.41, P = 0.02), KWW (rho = −0.36, P = 0.05),
and KFF (rho = −0.37, P = 0.04) at T1. Furthermore,
GLDH was negatively associated with PDM (both rho
= −0.43, P = 0.02) and PWW (rho = −0.52, P = 0.004
and rho = −0.43; P = 0.02), as well as with KDM (both
rho = −0.39, P = 0.02) and KWW (rho = −0.45, P
= 0.01, and rho = −0.43, P = 0.02) at T4 and T5,
respectively. At T4 and T5 GLDH was also associated
with TAG (rho = 0.40, P = 0.03, and rho = 0.51, P =
0.004, respectively).
The activity of GGT in plasma was negatively as-
sociated with PDM at T5 (rho = −0.43, P = 0.02) but
showed even stronger associations with liver DM (rho =
0.59, P = 0.0009) and TAG (rho = 0.43, P = 0.02). In
the regression analyses, GLDH and GGT were removed
from all final regression models in all instances due to
suspected collinearity.
DISCUSSION
The objective of this study was to determine effects
of restricted dietary P supply to late-pregnant high-
yielding dairy cows on productivity and metabolism
during and in the months following dietary P depri-
vation. The positive effect of the restricted P feeding
protocol used in this study on Ca homeostasis during
the periparturient period has been described earlier
(Wächter et al., 2022e). Feeding a P-deficient dry cow
diet during the last 4 wk of gestation resulted in higher
plasma [TCa] from 1 wk before to 1 wk after calving,
compared with cows on a diet with adequate P content
during the dry period. Upregulation of osteoclast activ-
ity in response to a negative P balance was proposed as
the primary underlying mechanism causing this effect
(Cohrs et al., 2018; Köhler et al., 2021; Wächter et al.,
2022e). The data set of the present study also included
plasma activity of AlP, an enzyme used, among oth-
ers, as biomarker to assess bone metabolic activity in
cattle and other species (Sato et al., 2002; Fernandez
and Kidney, 2007). The AlP activities before calving
were higher than after calving in both treatments, and
were furthermore higher in P-deprived cows compared
with dry cows on adequate dietary P supply. This cor-
 Wächter et al.: LONG-TERM EFFECTS OF RESTRICTED PHOSPHORUS FEEDING 4385

Figure 8. Liver tissue triacylglycerol (TAG; a, upper panel), cholesterol (Chol; b, middle panel) and DM (c, lower panel) contents for ad-
equate P (AP, blue striped boxes) and low P (LP, red open boxes) treatments. Thirty multiparous dairy cows were fed a dry cow ration with
either adequate (AP, n = 15, 0.30% P in DM) or low P content (LP, n = 15, 0.16% P in DM) during the last 4 wk of gestation. Box and whisker
plots represent median (horizontal line), lower and upper quartiles (bottom and top of box, respectively), and 10th and 90th percentiles (lower
and upper ends of whiskers, respectively). Liver sampling times were T1 (end of acclimation), T2 (1 wk before expected calving), T3 (d 3 post-
partum), T4 (d 14 postpartum), and T5 (d 42 postpartum). The vertical dashed lines mark parturition. Time points with different lowercase
letters differ statistically significantly within AP treatment; time points with different capital letters differ statistically significantly within LP
treatment (P < 0.05, Bonferroni corrected).

Journal of Dairy Science Vol. 105 No. 5, 2022

 Wächter et al.: LONG-TERM EFFECTS OF RESTRICTED PHOSPHORUS FEEDING
Figure 9. LSM ± SEM of liver tissue P (a, upper panel), Mg (b,
middle panel), and K (c, lower panel) contents, expressed as amount
in DM for adequate P (AP, blue, dashed line, open squares) and low P
(LP, red, solid line, closed circles) treatments, as well as expressed as
amount in fat-free liver tissue for adequate P (AP, blue, dashed-dotted
line, open circles) and low P (LP, red, dotted line, closed triangles)
treatments. Thirty multiparous dairy cows were fed a dry cow ration
with either adequate (AP, n = 15, 0.30% P in DM) or low P content
(LP, n = 15, 0.16% P in DM) during the last 4 wk of gestation. Liver
sampling times were T1 (end of acclimation), T2 (1 wk before ex-
pected calving), T3 (d 3 postpartum), T4 (d 14 postpartum), and T5
(d 42 postpartum). The vertical dashed lines mark parturition. Time
points with different lowercase letters differ statistically significantly
within AP treatment; time points with different capital letters differ
statistically significantly within LP treatment (P < 0.05, Bonferroni
corrected).
Journal of Dairy Science Vol. 105 No. 5, 2022
4386
roborates the concept of enhanced bone mobilization
in response to dietary P deprivation. Low plasma AlP
activity before calving was proposed as a prognostic
indicator for increased risk of periparturient hypocal-
cemia by Furrl (2012), but other authors found this
parameter to be of limited value for this purpose (Sato
et al., 2005). When pooling cows of both treatments
and grouping them by the occurrence of clinical and
subclinical hypocalcemia after calving, the AlP activi-
ties in cows that later developed clinical hypocalcemia
were numerically lower from 4 wk before calving until
calving, a difference that did not reach significance
level due to the large degree of interindividual variation
in association with the small sample size (Supplemental
Figure S1).
Feeding a diet with either adequate or low P content
during the dry period resulted in lower plasma [Pi] in
P-deprived cows during the entire P deprivation period.
The mean plasma [Pi] of this treatment was below the
reference range for cattle (Constable et al., 2016b),
reaching a nadir in the range of 1 mmol/L immediately
before calving. Remarkably, the level of hypophos-
phatemia achieved in this study was considerably less
pronounced than in an earlier study that used a very
similar feeding protocol (Grünberg et al., 2019b). In
this previous study, plasma [Pi] reached a nadir below
0.5 mmol/L immediately before calving. Importantly,
the results presented here indicate that restricting the
dietary P supply during the dry period does not ag-
gravate or prolong the transient hypophosphatemic
period, which is well recognized to occur in dairy cows
in the first hours and days after calving (Goff et al.,
2002; Grünberg, 2014; Megahed et al., 2018). To the
contrary, it appears that the negative P balance during
the dry period enhanced the capacity to absorb dietary
P from the digestive tract, which resulted in a prompt
increase of plasma [Pi] after switching P-deprived cows
to the lactating cow diet with adequate P content.
Earlier studies conducted on cows in various stages of
lactation showed a similarly rapid effect of switching
previously P-deprived cows to a diet with adequate P
content (Grünberg et al., 2015b, 2019b).
The metabolic and clinical relevance of the hypo-
phosphatemia observed in fresh cows is uncertain
(Grünberg, 2014). Concerns are often based on the
erroneous assumption that subnormal plasma [Pi] in
fresh cows is indicative of P deficiency, as well as on
the empirical association of the occurrence of recum-
bency or intravascular hemolysis in early lactation with
hypophosphatemia (Macwilliams et al., 1982; Ménard
and Thompson, 2007). During the study reported here,
none of the experimental cows developed signs or symp-
toms thought to be associated with hypophosphatemia
or P deprivation, such as muscle function disturbances
 Wächter et al.: LONG-TERM EFFECTS OF RESTRICTED PHOSPHORUS FEEDING
or intravascular hemolysis. Intravascular hemolysis, but
no signs indicative of clinically relevant impairment of
muscle function, was observed in a similar study where
P deprivation extended beyond the dry period into the
first weeks of lactation (Grünberg et al., 2019a). It is
uncertain whether the primary reason for the occur-
rence of postparturient hemoglobinuria in this earlier
study is the prolonged duration of dietary P depriva-
tion or, rather, the specific P deprivation during early
lactation.
Feed intake depression associated with lower milk
production and impaired fertility are the main con-
cerns associated with sustained P deprivation in dairy
cows (Kincaid et al., 1981; Call et al., 1987; Valk and
Sebek, 1999; Wu et al., 2000). In the present study
no treatment effects on DMI or milk production were
identified. However, the limited statistical power of this
study, with 30 cows, was only sufficient to identify a
minimum effect size of 10% on milk yield and 15% on
DMI. Furthermore, feed intake depression in cows on
restricted dietary P supply that may have occurred
during the dry period may have gone unnoticed, due
to the restricted feeding during the dry period. Studies
using feeding protocols restricting the dietary P supply
of dairy cows not only during the dry period but also
during the first weeks of lactation reported negative
effects of P deprivation on DMI and milk production
well above 20% (Puggaard et al., 2014; Grünberg et al.,
2019b). Again it is uncertain whether the more pro-
nounced effects on DMI of fresh cows reported in ear-
lier studies is attributable to prolonged P deprivation
or is, rather, associated with P deprivation specifically
occurring during the fresh cow period. One study in-
vestigating effects of long-term P deprivation on health
and productivity, and initiating P restricted feeding in
mid-lactation, noticed a decrease in DMI during the dry
period and early lactation after over 20 wk of restricted
dietary P supply (Valk and Sebek, 1999). In contrast,
Puggaard et al. (2014) and Grünberg et al. (2019b),
who initiated P deprivation 3 and 4 wk before expected
calving, respectively, reported important decreases in
DMI within the first 2 wk of lactation; thus, after only
3 to 5 wk of dietary P deprivation. It thus remains to
be determined whether restricting the P supply during
the fresh period only would suffice to induce feed intake
depression in fresh cows.
As for DMI, the body weight time curves showed the
characteristic development for dairy cattle over time,
with similar shape and amplitude for both treatments.
This indicates that neither body weight in late gesta-
tion or in early lactation, nor the development of the
body condition of experimental cows, were affected by
the restricted P supply. Loss of body condition in P-
deprived dairy cows, compared with cows on adequate
Journal of Dairy Science Vol. 105 No. 5, 2022
4387
P supply, reported in earlier studies, was associated
with pronounced feed intake depression in P-deprived
animals (Call et al., 1987; Grünberg et al., 2019b).
As for DMI, an effect of P deprivation on milk pro-
duction could not be identified either, a result meriting
confirmation in larger-scale studies sufficiently powered
to identify smaller effect sizes. A negative effect on milk
yield while on a P-deficient diet has been reported by
several authors and was, again, always associated with
feed intake depression (Call et al., 1987; Valk and Sebek,
1999; Puggaard et al., 2014). In one study, restoring
the dietary P supply to adequate levels in previously
P-deprived dairy cows in the fifth week of lactation
resulted in a return to normal DMI within a matter
of days, but was not associated with an increase in
milk production (Grünberg et al., 2019b). Remarkably,
in the present study we identified a treatment × LN
interaction effect, with an approximately 10% higher
daily milk yield of LP compared with AP cows in the
second lactation (Figure 3). Inversely, for the cohort of
fourth-lactation cows, AP cows, with an average daily
milk yield of over 46 kg, produced 15% more milk than
LP cows. It is noteworthy that the cohort of fourth-lac-
tation cows was small, with 3 cows for each treatment,
of which 2 AP but no LP cows received Ca salts around
parturition, as they had developed clinical milk fever.
We deem it conceivable that the observed treatment ×
LN interaction effect in the fourth-lactation cows was
confounded by this Ca treatment. Several studies have
reported a beneficial effect of Ca supplementation in
the first week of lactation on milk production in dairy
cows with above-average productive potential (Oetzel
and Miller, 2012; Martinez et al., 2016; Leno et al.,
2018).
The contents of protein, fat, lactose, and urea ni-
trogen in milk followed the same pattern for both
treatments, as well as of the rest of the lactating herd
of the research farm over time. Again, no effect of P
deprivation during the dry period on milk composition
was apparent. As for milk yield, a LN × treatment
interaction effect was observed for SCC, with higher
cell counts among AP cows in the second lactation
and lower counts in AP cows in their fourth lactation
when compared with LP cows with corresponding LN.
Because a negative association of SCC with milk yield,
presumably attributable to dilution, is well established,
we deem it probable that the observed treatment × LN
interaction effect for SCC is the result of collinearity
between milk yield and SCC, rather than a direct effect
of P deprivation on udder health (Green et al., 2006).
The mean colostrum yield was numerically lower,
whereas the total solid content was numerically higher
in LP compared with AP cows. Expressing the colos-
trum yield as total solids in grams for the first milking
 Wächter et al.: LONG-TERM EFFECTS OF RESTRICTED PHOSPHORUS FEEDING
yielded similar amounts for both treatments. The nega-
tive association between the total solids content, which
is commonly used as a crude proxy for colostrum IgG
concentration (Godden et al., 2019), on one hand, and
colostrum yield on the other, is again well established
and considered to be a result of dilution (Morin et al.,
2010; Godden et al., 2019). The colostrum yield at first
milking was within the range of what is commonly ob-
tained per cow at the research farm. The mean total
solid contents of above 25% qualified the colostrum of
both treatments as high quality (Godden et al., 2019).
The energy metabolism and metabolic health of study
cows were assessed based on a combination of produc-
tion, clinical, and biochemical parameters. Apart from
milk yield, DMI, and body weight, we also measured
blood biochemical parameters that are indicative of en-
ergy and protein metabolism, liver function, and liver
cell injury, as well as parameters characterizing liver
tissue composition.
Plasma [NEFA] and [BHB], 2 parameters reflecting
the status of energy metabolism of dairy cows, did
not display treatment-related effects (van Dorland et
al., 2009; Roberts et al., 2012). The numerically more
elevated plasma [BHB] of LP cows in their second lac-
tation compared with AP cows of the corresponding
lactation might be attributable to the 10% higher milk
yield of LP than AP cows in this age cohort (Ospina
et al., 2010). As expected, the highest plasma [NEFA]
values were measured within the first month of lac-
tation, with numerically higher concentrations in AP
compared with LP cows. Moderately elevated plasma
[BHB], with a large degree of interindividual variation,
was observed throughout the first 6 wk of lactation in
both treatments.
Liver enzyme activities of GGT and GLDH, as well as
concentrations of TBil in plasma, were measured as in-
dicators for liver cell injury and liver function. Although
activities of both enzymes were within or slightly above
the reference range for cattle, increases over time were
observed for both treatments. Increasing activities of
GGT and GLDH in cattle as lactation progresses have
been reported (Busato et al., 2002; Kupczyński and
Chudoba-Drozdowska, 2002; Kim et al., 2020). Such
increases are usually attributed to hepatocyte injury,
resulting in translocation of these enzymes to the ex-
tracellular space. However, increased enzyme activities
in plasma can also be the result of liver hypertrophy, a
mechanism that deserves consideration as an underly-
ing cause in cows in early lactation. The liver of dairy
cows increases in mass by over 30% in the first 20 wk of
lactation, while the fraction of tissue water and ash as
percentage of liver wet weight remains constant, which
indicates an increase of the number of cells and thus
hypertrophy (Gibb et al., 1992). An increase of the liver
Journal of Dairy Science Vol. 105 No. 5, 2022
4388
mass relative to the total body mass, and thus also
relative to the total extracellular volume, will result in
increased enzyme activity in plasma that is not driven
by liver cell injury.
The sudden increase of [TBil] in plasma observed at
calving, which persisted over the first weeks of lactation,
was of similar degree for both treatments. Increases of
plasma [TBil] at the onset of lactation are well recog-
nized to occur in dairy cows, and are considered a sequel
of negative energy balance and ensuing liver lipid ac-
cumulation (Kauppinen, 1984; Busato et al., 2002). Re-
markably, in this study, [TBil] was strongly associated
with plasma [NEFA] at the postpartum sampling times,
but not with the activity of GGT or GLDH in plasma,
nor with any of the parameters determined in liver tis-
sue, and specifically not with the liver TAG content. By
contrast, the activities of GGT and GLDH were associ-
ated with liver tissue parameters that were affected by
the increase in liver TAG at the sampling times after
calving. This suggests that liver enzyme activities are
more indicative of morphological changes occurring in
liver tissue, such as liver lipid accumulation and reduc-
tion of tissue water, whereas [TBil] is more reflective
of metabolic changes, such as lipid mobilization and
ensuing increase of plasma [NEFA]. This interpretation
is supported by the observation that treatment with
PG, triggered by the presumed diagnosis of ketosis, was
associated with higher plasma [TBil].
In the present study we observed a marked overes-
timation of the incidence of clinical and subclinical
ketosis when recurring to the cow-side urine dipstick
for diagnosis of ketonuria. Overestimation of the occur-
rence of ketosis with urine dipsticks is well recognized,
and has been attributed to the fact that this test does
not account for changes in urine density (Nielen et al.,
1994). However, the observed overestimation in the
range of 50% for subclinical ketosis is well above what
has been reported earlier (Nielen et al., 1994).
The liver TAG contents in this study showed the
characteristic development during the transition pe-
riod of dairy cows (Bobe et al., 2004; Grünberg et al.,
2019b). Liver TAG did not differ between treatments,
but peak values recorded in the present study were
above contents previously reported in early lactating
dairy cows (Bobe et al., 2004; Grünberg et al., 2019b).
The productivity of the cows in this experiment, with
a 305-d milk yield of over 12,000 kg, was well above
that of most of previous studies reporting liver TAG
contents, which could explain a more pronounced level
of negative energy balance, and thus of liver TAG accu-
mulation. However, it is conceivable that the restricted
feeding protocol used in this experiment contributed
to the relatively high level of liver lipid accumulation
observed in the cows of both treatments. Should this
 Wächter et al.: LONG-TERM EFFECTS OF RESTRICTED PHOSPHORUS FEEDING
have been the case, this would have affected both treat-
ments equally.
Dietary P restriction did not alter the liver P content
during or following P deprivation. A marked decline
of the liver P content, when expressed as amount of
P in DM, of over 20%, was observed during the first
2 wk of lactation in both treatments. A similar de-
cline was, however, observed for other predominantly
intracellular electrolytes, such as K and Mg, indicating
that this effect is not specific to P. The strong nega-
tive associations of the liver TAG content, not only
with liver P but also with K and Mg in liver DM at
the postpartum sampling times, as well as the positive
association of liver TAG with liver DM at these sam-
pling times, indicate that this apparent decline of the
content of predominantly intracellular electrolytes in
liver tissue may be driven by liver lipid accumulation.
A positive correlation of the liver TAG content with
liver tissue DM has been reported earlier in cattle and
other species (Häussinger and Lang, 1991; Grünberg
et al., 2009, 2019b). The association is thought to be
the result of a compensatory reduction of the cytosol
volume of hepatocytes as they accumulate lipids, with
the objective of preventing excessive cell volume expan-
sion, which may jeopardize liver cell and organ function
(Häussinger and Lang, 1991; Dunkelberg et al., 2001;
Grünberg et al., 2009). When reducing their cytosol
volume, cells must also reduce the contents of osmoti-
cally active solutes, to maintain intracellular osmolarity
and thus the osmotic equilibrium between intra- and
extracellular space. Nearly constant mineral contents,
when expressed as amount per fat-free liver wet weight,
as observed in this and earlier studies, corroborate the
proposed hypothesis (Grünberg et al., 2009, 2019b).
The liver cholesterol content was studied as an in-
dicator for the activity of hepatic liver metabolism in
experimental cows. Cholesterol values in liver tissue
determined at the different phases of the transition
period were in good agreement with values reported in
an earlier study in transition dairy cows (Schlegel et al.,
2012). Increased liver cholesterol at the time of parturi-
tion, as observed here, can be the result of increased he-
patic cholesterol synthesis, enhanced cholesterol uptake
from plasma, or impaired cholesterol excretion with the
bile. An earlier study found a marked upregulation of
hepatic cholesterol synthesis at the onset of lactation,
associated with the observed increase of the liver cho-
lesterol content around calving, making this pathway
a plausible mechanism behind increases of liver tissue
cholesterol in early lactation (Schlegel et al., 2012). Up-
regulating the hepatic cholesterol synthesis is deemed a
meaningful adaptation process to lactation, to provide
sufficient substrate for the enhanced formation of bile
acids, and to promote formation of cell membranes to
Journal of Dairy Science Vol. 105 No. 5, 2022
4389
support early lactation liver hypertrophy mentioned
earlier, and to provide cholesterol required for lactogen-
esis (Smith et al., 1998). The results reported here do
not suggest impaired lipid metabolism associated with
restricted P supply during the dry period.
The plasma [Pi] measured at the liver tissue sampling
times were not associated with the liver tissue P con-
tent. This finding is in agreement with several earlier
studies that failed to identify associations between plas-
ma [Pi] and contents of P in liver tissue, muscle tissue,
or red blood cells (Grünberg et al., 2015a, 2019a,b).
These studies all indicate that subnormal plasma [Pi]
is not associated with subnormal P contents of various
tissues. Accordingly, the plasma [Pi] must be consid-
ered unsuitable to assess or quantify deficiencies of this
predominantly intracellular electrolyte.
Several blood biochemical parameters were ana-
lyzed as indicators for the protein metabolism, and
to crudely assess changes in the water balance of the
cows on study over time. None of these parameters were
measurably affected by the experimental treatment.
The plasma protein concentration showed a transient
decline around parturition in both treatments, which
is in line with observations of numerous earlier studies
(Bruckmaier and Blum, 1998; Grünberg et al., 2011;
Piccione et al., 2011). This decline, which was not as-
sociated with a concomitant decline of plasma albumin
concentration in these earlier studies, has been attrib-
uted to the excretion of globulin with colostrum at the
onset of lactation (Piccione et al., 2011).
We further observed a continuous decline of plasma
[Crea] over the first 6 wk of lactation, which, again, is
in agreement with the literature and has been attrib-
uted to enhanced protein metabolization in early lacta-
tion in response to negative energy balance (Megahed
et al., 2018). Remarkably, we observed an increase of
plasma [Crea] over the last weeks of gestation that,
to the best of our knowledge, has not been reported
earlier. Creatinine is a breakdown product of muscle
metabolism, which is produced at a constant rate as
long as muscle mass remains constant, and is excreted
exclusively through the kidneys. Assuming that muscle
mass of the dam does not change in late gestation, and
further assuming that the placenta is impermeable to
fetal creatinine, the observed increase of plasma [Crea]
over the last 4 wk of pregnancy would be most plausibly
explained by a decrease of the extracellular volume and
an ensuing reduction of the renal glomerular filtration
rate. The assumption of extracellular volume reduction
in the days and weeks before calving would also explain
the numerical increase of the plasma [Na] before and
the decline after calving observed in this study. Such a
reduction of extracellular volume could, in theory, be
caused by decrease of the protein content in plasma.
 Wächter et al.: LONG-TERM EFFECTS OF RESTRICTED PHOSPHORUS FEEDING
We propose, however, that this transient increase of
plasma [Crea] may be an artifact of the study design
with restricted feeding during the dry period, when the
observed increase in plasma [Crea] occurred. Voluntary
water intake is highly correlated with DMI in cattle,
and feed intake furthermore contributes to water con-
sumption through water contained in feed (Kume et
al., 2010). We hypothesize that switching study cows
from ad libitum to restricted feeding halfway through
the dry period may have resulted in a reduction of daily
water intake, thereby affecting the hydration status of
these animals. If applicable, this effect would have af-
fected cows of both treatments similarly.
Disease incidences, with exception of issues related
to claw health, were low and, in combination with the
relatively small sample size, did not allow us to draw
the unambiguous conclusion that P deprivation during
the dry period is innocuous to bovine health. Other
studies with phases of dietary P deprivation extending
into the first weeks of lactation reported a dramatic dif-
ference in disease occurrence between P-deprived cows
and animals on diets with adequate P content (Valk
and Sebek, 1999; Puggaard et al., 2014; Grünberg et
al., 2019a).
An important limitation of the present study arises
from the restricted feeding protocol, which was re-
quired to induce a negative P balance in dry cows.
Although the study design was conceived to control for
the restricted DMI, and although key parameters such
as DMI or milk yield did not differ from the rest of
the herd at the research facility, it is still conceivable
that certain parameters such as liver TAG or plasma
creatinine concentration may have been affected by
the restricted access to feed. A pilot trial conducted
in preparation for this study showed that the daily P
intake during the dry period needed to be capped at 20
g of P per cow to consistently achieve a state of nega-
tive P balance (Cohrs et al., 2021). With available feed
ingredients, we were unable to formulate an otherwise
balanced dry cow ration with a P content below 0.16%
in DM. With the dry cows of the research farm consum-
ing on average 14 to 16 kg of DM per day, limiting the
DMI was the only available option. Inducing a negative
P balance in dry cows by lowering the dietary P content
without resorting to feed restriction was attempted in
earlier studies, without success (Peterson et al., 2005).
These findings underscore that a negative P balance in
dry cows is highly unlikely to occur in regions with soils
that are not P deficient.
Another weakness of the study was associated with
the use of urine dipsticks to diagnose ketosis, which
resulted in the inappropriate treatment with PG of sev-
eral study cows. This protocol was employed because it
allowed identification of ketotic cows by farm personnel
Journal of Dairy Science Vol. 105 No. 5, 2022
4390
on the spot, knowingly taking into account a rate of
false positive results in the range of 10%, as reported
in the literature (Nielen et al., 1994). The number of
false positive results in the range, as observed here,
was unexpected; however, an attempt was made to
retrospectively identify potentially confounding effects
of the PG treatment on blood biochemical parameters.
CONCLUSIONS
Restricting the dietary P content in dairy cows dur-
ing the dry period to a maximum intake of 20 g of P per
day is effective in statistically significantly improving
the Ca homeostasis during the first week of lactation,
and in mitigating the severity and duration of postpar-
turient hypophosphatemia, which is known to occur in
the majority of high-yielding dairy cows. Restriction of
P during the dry period did not have pronounced effects
on DMI or milk yield in highly productive cows. No
metabolic effects associated with P deprivation, other
than on Ca homeostasis, were identified during the dry
period or the following lactation. Although comparison
of disease incidence did not yield any indication of a
deleterious effect of P deprivation on health, these re-
sults should be confirmed in a larger-scale study.
ACKNOWLEDGMENTS
The authors acknowledge the assistance of Christian
Koch, Educational and Research Centre for Animal
Husbandry, Hofgut Neumühle (Münchweiler an der
Alsenz, Germany) in formulating the experimental ra-
tions, and of the staff of the dairy unit of Hofgut Neu-
mühle for their dedicated technical assistance. The feed
additive mono-ammoniumphosphate was generously
provided by Aliphos Rotterdam BV (Vlaardingen, the
Netherlands). The authors furthermore acknowledge
the financial support of the H. Wilhelm Schaumann
Foundation (Hamburg, Germany) through a graduate
student grant. The authors have not stated any con-
flicts of interest.
REFERENCES
Andersson, L. 1988. Subclinical ketosis in dairy cows. Vet. Clin. North
Am. Food Anim. Pract. 4:233–251. https:​/​/​doi​.org/​10​.1016/​S0749​
-0720(15)31046​-X.
Baird, G. D. 1982. Primary ketosis in the high-producing dairy cow:
Clinical and subclinical disorders, treatment, prevention, and
outlook. J. Dairy Sci. 65:1–10. https:​/​/​doi​.org/​10​.3168/​jds​.S0022​
-0302(82)82146​-2.
Baquerizo, A., D. Anselmo, C. Shackleton, T. W. Chen, C. Cao, M.
Weaver, J. Gornbein, S. Geevarghese, N. Nissen, D. Farmer, A.
Demetriou, and R. W. Busuttil. 2003. Phosphorus as an early
predictive factor in patients with acute liver failure. Transplan-
tation 75:2007–2014. https:​/​/​doi​.org/​10​.1097/​01​.TP​.0000063219​
.21313​.32.
 Wächter et al.: LONG-TERM EFFECTS OF RESTRICTED PHOSPHORUS FEEDING
Bobe, G., J. W. Young, and D. C. Beitz. 2004. Invited review: Pa-
thology, etiology, prevention, and treatment of fatty liver in dairy
cows. J. Dairy Sci. 87:3105–3124. https:​/​/​doi​.org/​10​.3168/​jds​
.S0022​-0302(04)73446​-3.
Bruckmaier, R. M., and J. W. Blum. 1998. Oxytocin release and milk
removal in ruminants. J. Dairy Sci. 81:939–949. https:​/​/​doi​.org/​10​
.3168/​jds​.S0022​-0302(98)75654​-1.
Busato, A., D. Faissler, U. Küpfer, and J. W. Blum. 2002. Body con-
dition scores in dairy cows: Associations with metabolic and en-
docrine changes in healthy dairy cows. J. Vet. Med. A Physiol.
Pathol. Clin. Med. 49:455–460. https:​/​/​doi​.org/​10​.1046/​j​.1439​
-0442​.2002​.00476​.x.
Call, J. W., J. E. Butcher, J. L. Shupe, R. C. Lamb, R. L. Boman,
and A. E. Olson. 1987. Clinical effects of low phosphorus concen-
trations in feed given to lactating dairy cows. Am. J. Vet. Res.
48:133–136.
Cohrs, I., M. R. Wilkens, and W. Grunberg. 2018. Short communica-
tion: Effect of dietary phosphorus deprivation in late gestation
and early lactation on the calcium homeostasis of periparturient
dairy cows. J. Dairy Sci. 101:9591–9598. https:​/​/​doi​.org/​10​.3168/​
jds​.2018​-14642.
Cohrs, I., M. R. Wilkens, S. Wächter, L. Golbeck, T. Scheu, and W.
Grünberg. 2021. Estimation of dietary phosphorus requirements
of dairy cows during the dry period by means of a biomarker for
bone resorption. Page 54 in Proc. Soc. Nutr. Physiol. DLG-Verlag.
Constable, P. D., K. W. Hinchcliff, S. H. Done, and W. Grünberg.
2016a. Metabolic and endocrine diseases. Pages 1662–1757 in Vet-
erinary Medicine: A Textbook of the Diseases of Cattle, Horses,
Sheep, Pigs, and Goats. 11th ed. Elsevier.
Constable, P. D., K. W. Hinchcliff, S. H. Done, and W. Grünberg.
2016b. Reference laboratory values. Pages 2217–2219 in Veterinary
Medicine: A Textbook of the Diseases of Cattle, Horses, Sheep,
Pigs, and Goats. 11th ed. Elsevier.
Dunkelberg, J. C., A. P. Feranchak, and J. G. Fitz. 2001. Liver cell
volume regulation: Size matters. Hepatology 33:1349–1352. https:​
/​/​doi​.org/​10​.1053/​jhep​.2001​.24750.
Federation, I. D. 1999. Suggested interpretation of mastitis terminol-
ogy. Pages 3–26 in IDF Bulletin No. 338. International Dairy Fed-
eration.
Fernandez, N. J., and B. A. Kidney. 2007. Alkaline phosphatase: Be-
yond the liver. Vet. Clin. Pathol. 36:223–233. https:​/​/​doi​.org/​10​
.1111/​j​.1939​-165X​.2007​.tb00216​.x.
Fürll, M. 2012. Gibt es Fortschritte bei der Früherkennung und Thera-
pie von Festliegen? Pages 65–69 in Proc. 6. Leipziger Tierärztekon-
gress, Leipzig, Germany.
Gessner, D. K., C. Koch, F. J. Romberg, A. Winkler, G. Dusel, E.
Herzog, E. Most, and K. Eder. 2015. The effect of grape seed and
grape marc meal extract on milk performance and the expression
of genes of endoplasmic reticulum stress and inflammation in the
liver of dairy cows in early lactation. J. Dairy Sci. 98:8856–8868.
https:​/​/​doi​.org/​10​.3168/​jds​.2015​-9478.
Gibb, M. J., W. E. Ivings, M. S. Dhanoa, and J. D. Sutton. 1992.
Changes in body components of autumn-calving Holstein-Friesian
cows over the first 29 weeks of lactation. Anim. Sci. 55:339–360.
https:​/​/​doi​.org/​10​.1017/​S0003356100021036.
Godden, S. M., J. E. Lombard, and A. R. Woolums. 2019. Colostrum
management for dairy calves. Vet. Clin. North Am. Food Anim.
Pract. 35:535–556. https:​/​/​doi​.org/​10​.1016/​j​.cvfa​.2019​.07​.005.
Goff, J. P., K. Kimura, and R. L. Horst. 2002. Effect of mastectomy
on milk fever, energy, and vitamins A, E, and β-carotene status at
parturition. J. Dairy Sci. 85:1427–1436. https:​/​/​doi​.org/​10​.3168/​
jds​.S0022​-0302(02)74210​-0.
Green, L. E., Y. H. Schukken, and M. J. Green. 2006. On distinguish-
ing cause and consequence: Do high somatic cell counts lead to
lower milk yield or does high milk yield lead to lower somatic
cell count? Prev. Vet. Med. 76:74–89. https:​/​/​doi​.org/​10​.1016/​j​
.prevetmed​.2006​.04​.012.
Grünberg, W. 2014. Treatment of phosphorus balance disorders. Vet.
Clin. North Am. Food Anim. Pract. 30:383–408. https:​/​/​doi​.org/​
10​.1016/​j​.cvfa​.2014​.03​.002.
Journal of Dairy Science Vol. 105 No. 5, 2022
4391
Grünberg, W., S. S. Donkin, and P. D. Constable. 2011. Periparturient
effects of feeding a low dietary cation-anion difference diet on acid-
base, calcium, and phosphorus homeostasis and on intravenous
glucose tolerance test in high-producing dairy cows. J. Dairy Sci.
94:727–745. https:​/​/​doi​.org/​10​.3168/​jds​.2010​-3230.
Grünberg, W., J. A. Mol, and E. Teske. 2015a. Red blood cell phos-
phate concentration and osmotic resistance during dietary phos-
phate depletion in dairy cows. J. Vet. Intern. Med. 29:395–399.
https:​/​/​doi​.org/​10​.1111/​jvim​.12497.
Grünberg, W., P. Scherpenisse, I. Cohrs, L. Golbeck, P. Dobbelaar, L.
M. van den Brink, and I. D. Wijnberg. 2019a. Phosphorus content
of muscle tissue and muscle function in dairy cows fed a phos-
phorus-deficient diet during the transition period. J. Dairy Sci.
102:4072–4093. https:​/​/​doi​.org/​10​.3168/​jds​.2018​-15727.
Grünberg, W., P. Scherpenisse, P. Dobbelaar, M. J. Idink, and I. D.
Wijnberg. 2015b. The effect of transient, moderate dietary phos-
phorus deprivation on phosphorus metabolism, muscle content
of different phosphorus-containing compounds, and muscle func-
tion in dairy cows. J. Dairy Sci. 98:5385–5400. https:​/​/​doi​.org/​10​
.3168/​jds​.2015​-9357.
Grünberg, W., R. Staufenbiel, P. D. Constable, H. M. Dann, D. E.
Morin, and J. K. Drackley. 2009. Liver phosphorus content in
Holstein-Friesian cows during the transition period. J. Dairy Sci.
92:2106–2117. https:​/​/​doi​.org/​10​.3168/​jds​.2008​-1897.
Grünberg, W., S. Witte, I. Cohrs, L. Golbeck, J. F. Brouwers, A. E.
Muller, and M. Schmicke. 2019b. Liver phosphorus content and
liver function in states of phosphorus deficiency in transition dairy
cows. PLoS One 14:e0219546. https:​/​/​doi​.org/​10​.1371/​journal​
.pone​.0219546.
Hara, A., and N. S. Radin. 1978. Lipid extraction of tissues with a
low-toxicity solvent. Anal. Biochem. 90:420–426. https:​/​/​doi​.org/​
10​.1016/​0003​-2697(78)90046​-5.
Harder, H., A. Khol-Parisini, B. U. Metzler-Zebeli, F. Klevenhusen,
and Q. Zebeli. 2015. Treatment of grain with organic acids at 2
different dietary phosphorus levels modulates ruminal microbial
community structure and fermentation patterns in vitro. J. Dairy
Sci. 98:8107–8120. https:​/​/​doi​.org/​10​.3168/​jds​.2015​-9913.
Häussinger, D., and F. Lang. 1991. Cell volume in the regulation
of hepatic function: A mechanism for metabolic control. Bio-
chim. Biophys. Acta 1071:331–350. https:​/​/​doi​.org/​10​.1016/​0304​
-4157(91)90001​-D.
Kauppinen, K. 1984. ALAT, AP, ASAT, GGT, OCT activities and
urea and total bilirubin concentrations in plasma of normal and
ketotic dairy cows. Zentralbl. Veterinarmed. A 31:567–576. https:​/​
/​doi​.org/​10​.1111/​j​.1439​-0442​.1984​.tb01316​.x.
Kim, S. B., S. H. Jung, Y. H. Do, C. Choe, S. Ha, H. Y. Joeng, A. Cho,
S. I. Oh, E. Kim, J. G. Yoo, and S. Kim. 2020. Haemato-chemical
and immune variations in Holstein cows at different stages of lac-
tation, parity, and age. Veterinární Medicína 65:95–103. https:​/​/​
doi​.org/​10​.17221/​110/​2019​-VETMED.
Kincaid, R. L., J. K. Hillers, and J. D. Cronrath. 1981. Calcium and
phosphorus supplementation of rations for lactating cows. J. Dairy
Sci. 64:754–758. https:​/​/​doi​.org/​10​.3168/​jds​.S0022​-0302(81)82644​
-6.
Köhler, O. M., W. Grünberg, N. Schnepel, A. S. Muscher-Banse, A.
Rajaeerad, J. Hummel, G. Breves, and M. R. Wilkens. 2021. Di-
etary phosphorus restriction affects bone metabolism, vitamin
D metabolism and rumen fermentation traits in sheep. J. Anim.
Physiol. Anim. Nutr. (Berl.) 105:35–50. https:​/​/​doi​.org/​10​.1111/​
jpn​.13449.
Kraft, W., and U. M. Dürr. 2014. Organsysteme und Stoffwechselpa-
rameter. Pages 79–556 in Klinische Labordiagnostik in der Tier-
medizin. 7th ed. Schattauer.
Kume, S., K. Nonaka, T. Oshita, and T. Kozakai. 2010. Evaluation of
drinking water intake, feed water intake and total water intake in
dry and lactating cows fed silages. Livest. Sci. 128:46–51. https:​/​/​
doi​.org/​10​.1016/​j​.livsci​.2009​.10​.012.
Kupczyński, R., and B. Chudoba-Drozdowska. 2002. Values of selected
biochemical parameters of cows’ blood during their drying-off and
the beginning of lactation. Electron. J. Pol. Agric. Univ. 5:1–10.
 Wächter et al.: LONG-TERM EFFECTS OF RESTRICTED PHOSPHORUS FEEDING
Leno, B. M., R. C. Neves, I. M. Louge, M. D. Curler, M. J. Thomas,
T. R. Overton, and J. A. A. McArt. 2018. Differential effects of a
single dose of oral calcium based on postpartum plasma calcium
concentration in Holstein cows. J. Dairy Sci. 101:3285–3302. https:​
/​/​doi​.org/​10​.3168/​jds​.2017​-13164.
Macwilliams, P. S., G. P. Searcy, and J. E. Bellamy. 1982. Bovine
postparturient hemoglobinuria: A review of the literature. Can.
Vet. J. 23:309–312.
Martinez, N., L. D. P. Sinedino, R. S. Bisinotto, R. Daetz, C. Lopera,
C. A. Risco, K. N. Galvao, W. W. Thatcher, and J. E. P. Santos.
2016. Effects of oral calcium supplementation on mineral and acid-
base status, energy metabolites, and health of postpartum dairy
cows. J. Dairy Sci. 99:8397–8416. https:​/​/​doi​.org/​10​.3168/​jds​.2015​
-10527.
Megahed, A. A., M. W. H. Hiew, S. A. El Badawy, and P. D. Con-
stable. 2018. Plasma calcium concentrations are decreased at least
9 hours before parturition in multiparous Holstein-Friesian cattle
in a herd fed an acidogenic diet during late gestation. J. Dairy Sci.
101:1365–1378. https:​/​/​doi​.org/​10​.3168/​jds​.2017​-13376.
Ménard, L., and A. Thompson. 2007. Milk fever and alert downer
cows: Does hypophosphatemia affect the treatment response? Can.
Vet. J. 48:487–491.
Morin, D. E., S. V. Nelson, E. D. Reid, D. W. Nagy, G. E. Dahl, and
P. D. Constable. 2010. Effect of colostral volume, interval between
calving and first milking, and photoperiod on colostral IgG con-
centrations in dairy cows. J. Am. Vet. Med. Assoc. 237:420–428.
https:​/​/​doi​.org/​10​.2460/​javma​.237​.4​.420.
Nielen, M., M. G. Aarts, A. G. Jonkers, T. Wensing, and Y. H. Schuk-
ken. 1994. Evaluation of two cowside tests for the detection of
subclinical ketosis in dairy cows. Can. Vet. J. 35:229–232.
NRC. 2001. Nutrient Requirements of Dairy Cattle. 7th rev. ed. Natl.
Acad. Press.
Oetzel, G. R., and B. E. Miller. 2012. Effect of oral calcium bolus
supplementation on early-lactation health and milk yield in com-
mercial dairy herds. J. Dairy Sci. 95:7051–7065. https:​/​/​doi​.org/​
10​.3168/​jds​.2012​-5510.
Ospina, P. A., D. V. Nydam, T. Stokol, and T. R. Overton. 2010. Asso-
ciations of elevated nonesterified fatty acids and β-hydroxybutyrate
concentrations with early lactation reproductive performance and
milk production in transition dairy cattle in the northeastern Unit-
ed States. J. Dairy Sci. 93:1596–1603. https:​/​/​doi​.org/​10​.3168/​jds​
.2009​-2852.
Peterson, A. B., M. W. Orth, J. P. Goff, and D. K. Beede. 2005. Peri-
parturient responses of multiparous Holstein cows fed different di-
etary phosphorus concentrations prepartum. J. Dairy Sci. 88:3582–
3594. https:​/​/​doi​.org/​10​.3168/​jds​.S0022​-0302(05)73043​-5.
Piccione, G., V. Messina, A. Schembari, S. Casella, C. Giannetto, and
D. Alberghina. 2011. Pattern of serum protein fractions in dairy
cows during different stages of gestation and lactation. J. Dairy
Res. 78:421–425. https:​/​/​doi​.org/​10​.1017/​S0022029911000562.
Plaizier, J. C., T. Garner, T. Droppo, and T. Whiting. 2004. Nu-
tritional practices on Manitoba dairy farms. Can. J. Anim. Sci.
84:501–509. https:​/​/​doi​.org/​10​.4141/​A03​-115.
Puggaard, L., P. Lund, A. Liesegang, and J. Sehested. 2014. Long term
effect of reduced dietary phosphorus on feed intake and milk yield
in dry and lactating dairy cows. Livest. Sci. 159:18–28. https:​/​/​doi​
.org/​10​.1016/​j​.livsci​.2013​.10​.009.
Risco, C. A., L. F. Archbald, J. Elliott, T. Tran, and P. Chavatte.
1994. Effect of hormonal treatment on fertility in dairy cows with
dystocia or retained fetal membranes at parturition. J. Dairy Sci.
77:2562–2569. https:​/​/​doi​.org/​10​.3168/​jds​.S0022​-0302(94)77197​
-6.
Journal of Dairy Science Vol. 105 No. 5, 2022
4392
Roberts, T., N. Chapinal, S. J. LeBlanc, D. F. Kelton, J. Dubuc, and
T. F. Duffield. 2012. Metabolic parameters in transition cows as
indicators for early-lactation culling risk. J. Dairy Sci. 95:3057–
3063. https:​/​/​doi​.org/​10​.3168/​jds​.2011​-4937.
Sato, J., M. Kanata, J. Yasuda, R. Sato, K. Okada, Y. Seimiya, and
Y. Naito. 2005. Changes of serum alkaline phosphatase activity in
dry and lactational cows. J. Vet. Med. Sci. 67:813–815. https:​/​/​doi​
.org/​10​.1292/​jvms​.67​.813.
Sato, J., K. Okada, S. Fukuda, R. Sato, J. Yasuda, and Y. Naito.
2002. Serum activities of tartrate-resistant acid phosphatase and
bone specific alkaline phosphatase as indices of bone metabolism
in the cow. J. Vet. Med. Sci. 64:653–655. https:​/​/​doi​.org/​10​.1292/​
jvms​.64​.653.
Schlegel, G., R. Ringseis, J. Keller, F. J. Schwarz, and K. Eder. 2012.
Changes in the expression of hepatic genes involved in cholesterol
homeostasis in dairy cows in the transition period and at different
stages of lactation. J. Dairy Sci. 95:3826–3836. https:​/​/​doi​.org/​10​
.3168/​jds​.2011​-5221.
Sheldon, I. M., G. S. Lewis, S. LeBlanc, and R. O. Gilbert. 2006.
Defining postpartum uterine disease in cattle. Theriogenology
65:1516–1530. https:​/​/​doi​.org/​10​.1016/​j​.theriogenology​.2005​.08​
.021.
Smith, J. L., S. R. Lear, T. M. Forte, W. Ko, M. Massimi, and S. K.
Erickson. 1998. Effect of pregnancy and lactation on lipoprotein
and cholesterol metabolism in the rat. J. Lipid Res. 39:2237–2249.
https:​/​/​doi​.org/​10​.1016/​S0022​-2275(20)32479​-2.
Thorberg, B. M., M. L. Danielsson-Tham, U. Emanuelson, and K.
Persson Waller. 2009. Bovine subclinical mastitis caused by dif-
ferent types of coagulase-negative staphylococci. J. Dairy Sci.
92:4962–4970. https:​/​/​doi​.org/​10​.3168/​jds​.2009​-2184.
Valk, H., and L. B. J. Sebek. 1999. Influence of long-term feeding of
limited amounts of phosphorus on dry matter intake, milk produc-
tion, and body weight of dairy cows. J. Dairy Sci. 82:2157–2163.
https:​/​/​doi​.org/​10​.3168/​jds​.S0022​-0302(99)75459​-7.
van Dorland, H. A., S. Richter, I. Morel, M. G. Doherr, N. Castro, and
R. M. Bruckmaier. 2009. Variation in hepatic regulation of metab-
olism during the dry period and in early lactation in dairy cows. J.
Dairy Sci. 92:1924–1940. https:​/​/​doi​.org/​10​.3168/​jds​.2008​-1454.
Wächter, S., I. Cohrs, L. Golbeck, T. Scheu, K. Eder, and W. Gru-
enberg. 2022a. Supplementary figure S1: Alkaline phosphatase.
https:​/​/​doi​.org/​10​.13140/​RG​.2​.2​.30534​.14402.
Wächter, S., I. Cohrs, L. Golbeck, T. Scheu, K. Eder, and W. Gruen-
berg. 2022b. Supplementary table S1-Ration composition. https:​/​
/​doi​.org/​10​.13140/​RG​.2​.2​.36615​.88487.
Wächter, S., I. Cohrs, L. Golbeck, T. Scheu, K. Eder, and W. Gruen-
berg. 2022c. Supplementary table S2: Intra- and interassay coef-
ficients of variation for plasma biochemical laboratory analyses.
https:​/​/​doi​.org/​https:​/​/​doi​.org/​10​.13140/​RG​.2​.2​.14933​.29924.
Wächter, S., I. Cohrs, L. Golbeck, T. Scheu, K. Eder, and W. Gruen-
berg. 2022d. Supplementary table S3: Milk composition. https:​/​/​
doi​.org/​10​.13140/​RG​.2​.2​.23823​.25768.
Wächter, S., I. Cohrs, L. Golbeck, M. R. Wilkens, and W. Grünberg.
2022e. Effects of restricted dietary phosphorus supply to dry cows
on periparturient calcium status. J. Dairy Sci. 105:748–760. https:​
/​/​doi​.org/​10​.3168/​jds​.2021​-20726.
Wu, Z., L. D. Satter, and R. Sojo. 2000. Milk production, reproductive
performance, and fecal excretion of phosphorus by dairy cows fed
three amounts of phosphorus. J. Dairy Sci. 83:1028–1041. https:​/​/​
doi​.org/​10​.3168/​jds​.S0022​-0302(00)74967​-8.