899

TRIENNIAL GROWTH SYMPOSIUM— Effects of dietary 25-hydroxycholecalciferol

and cholecalciferol on blood vitamin D and mineral status, bone turnover, milk composition, and reproductive performance of sows1,2

G. M. Weber,*3 A.-K. M. Witschi,† C. Wenk,† and H. Martens‡

*DSM Nutritional Products Ltd., Nutrition Innovation Center, 4002 Basel, Switzerland; †Institute of Animal Science, ETH Zurich, 8092 Zurich, Switzerland; and ‡Institute for Veterinary Physiology, Free University Berlin, 14163 Berlin, Germany

ABSTRACT: To evaluate the role of vitamin D3 during gestation and lactation of sows, 2 independent experi-ments were performed with the aim of investigating sow reproductive performance, milk composition (study 1 only), and changes in blood status of 25-hydroxycho-lecalciferol (25-OH-D3), 1,25-dihydroxycholecalcif-erol (1,25-(OH)2–D3; study 2 only), minerals, and bone markers of sows during gestation and lactation. Study 1 comprised 39 primi- and multiparous crossbred sows fed 1 of 3 barley meal-based diets fortified with 200 IU/kg vitamin D3 (NRC, 1998; treatment DL), 2,000 IU/kg vitamin D3 (cholecalciferol; treatment DN), or 50 μg 25-OH-D3 (calcidiol; treatment HD)/kg feed. This study was conducted over a 4-parity period under controlled conditions. Study 2, running over 1 parity only, was per-formed in a commercial farm with 227 primi- and mul-tiparous sows allocated to 2 dietary treatments: control (CON), receiving 2,000 IU vitamin D3/kg (equivalent to 50 μg/kg) feed (114 sows), and test (HYD), supple-mented with 50 μg 25-OH-D3/kg feed (113 sows). Blood samples of sows were collected at 84 and 110 d postcoitum and 1, 5, and 33 d postpartum (study 1)

and at insemination and 28 and 80 d postinsemination as well as d 5 and 28 postpartum (study 2). Colostrum and milk samples in study 1 were obtained at 1, 9, and 33 d of lactation after oxytocin administration. Plasma 25-OH-D3 concentrations were increased (P < 0.05) in sows receiving 25-OH-D3 (HD and HYD) at any time of sampling whereas circulating plasma concentra-tions of 1,25-(OH)2–D3, Ca, and P were not affected by treatment. Milk concentrations of Ca and P were simi-lar, but 25-OH-D3 content (except in colostrum) was clearly increased (P < 0.05) when 25-OH-D3 was fed. Most characteristics of sow reproductive performance responded similarly to the 2 sources and levels of vita-min D3, but weight gain of piglets between birth and weaning was decreased (P < 0.05) in offspring of DL and HD sows compared with animals of treatment DN (study 1). In study 2 total litter weight and birth weight per piglet were increased (P < 0.05) with 25-OH-D3 supplementation in comparison with the control (CON). Overall, feeding sows with 25-OH-D3 was considered to improve maternal supply with vitamin D3 and thereby maintain Ca homeostasis during gestation and lactation.

Key words: 25-hydroxycholecalciferol, plasma, reproduction, milk, sow, vitamin D3

© 2014 American Society of Animal Science. All rights reserved. J. Anim. Sci. 2014.92:899–909 doi:10.2527/jas2013-7209

INTRODUCTION

Vitamin D3 is essential for regular growth and devel-opment as well as for maintenance of a normal calcium–phosphate homeostasis in mammals by regulating the active intestinal absorption of those minerals required for bone mineralization (DeLuca, 1978). During gesta-tion and lactation, vitamin D3 metabolism in the female is subject to increasing requirements for Ca for fetal growth and milk production (Halloran et al., 1979). This exceptional demand is supposed to result in mobilization

1Presented at the Triennial Growth Symposium titled “Vitamin D – Establishing the basics to dispel the hype,” preceding the Joint Annual Meeting, July 8–12, 2013, Indianapolis, IN. The sympo-sium was sponsored, in part, by DSM Nutritional Products (Basel, Switzerland) and Zoetis Animal Health (Florham Park, NJ), with publication sponsored, in part, by the Journal of Animal Science and the American Society of Animal Science.

2The authors are grateful to Ulrich Höller and Anette Liesegang for analysis of feed, plasma and milk samples as well as to Peter Wenning for technical supervision and Rotraut Schoop for statistical analysis of the field study.

3Corresponding author: gilbert.weber@dsm.comReceived September 30, 2013.Accepted December 19, 2013.

Published November 24, 2014

Weber et al.900

of Ca from the skeleton. The bone content of Ca is known to be depleted during periods of reproduction (Giesemann et al., 1998), which causes weakness of the skeleton and leads to lameness or even bone fractures. Such problems with the musculoskeletal system are among the primary causes for culling of sows raised under commercial con-ditions (Kirk et al., 2005). It is generally accepted that 10 to 15% of the culls are attributable to leg problems but fig-ures of up to 40% have been reported (Kirk et al., 2005). Culling sows after a short production period represents a considerable loss to the pig industry (Kirk et al., 2005; Engblom et al., 2008a,b) but is also increasingly seen as a welfare issue because lameness is associated with severe pain (Jørgensen and Jørgensen, 1998).

The knowledge of vitamin D3 requirements of sows under current conditions is limited and very few studies have compared different sources of vitamin D3 in swine. In the present research, the efficacy of 2 dietary sources of vitamin D3 (i.e., cholecalciferol and 25-hydroxycholecal-ciferol) provided either at the NRC (1998) recommenda-tion or at the common practical level on reproductive per-formance as well as on plasma and milk concentrations of vitamin D3, minerals, and markers of bone turnover of sows was evaluated. Two experiments with sows were performed: 1 under controlled conditions over 4 consecu-tive reproductive cycles (study 1) and 1 on a commercial farm over 1 reproductive cycle (study 2).

MATERIALS AND METHODS

Experimental Design: Study 1This study was performed on the research farm

Chamau (Hünenberg, Switzerland) of the Swiss Federal Institute of Technology, Zurich; the experimental use of animals and procedures applied were approved by the Committee for Animal Care and Use of Kanton Zug. Thirty-nine primiparous and multiparous crossbred sows of predominantly Large White breeding were randomly allotted to 3 dietary treatments on the day of mating and were kept on those treatments for 4 reproductive cycles. Mixed groups of 4 to 6 animals of all treatment entered the trial staggered in intervals of 21 d. The sows were kept in 1 large herd during gestation and fed individually with a computerized feeding system (Schauer Compident VI, Prambachkirchen, Austria), which was located in a pen and recorded feed intake and body weight daily. At 110 d postcoitum, pregnant dams were placed in indi-vidual farrowing pens that contained straw as bedding and heated nests for the piglets according to Swiss ani-mal welfare guidelines. Two different types of basal sow diets (gestation and lactation), with barley, corn, wheat, and soybean meal as main ingredients, were used in the present experiment. The gestation diet was fed from mat-

ing until d 110 of gestation followed by the lactation diet, which was provided throughout the lactation period until weaning on d 35 postpartum (PP). The diets were offered in a pelleted form twice daily during both gestation (2.2 to 2.5 kg/d) and lactation (5.2 to 5.6 kg/d), depending on the requirements of the individual animal. The composi-tion of the diets as well as the analyzed nutrient content is given in Table 1. During both gestation and lactation, ani-mals had free access to water. At breeding, the sows were randomly allotted to 1 of 3 experimental treatments. The diet of treatment (DL) was supplemented with 200 IU cholecalciferol (vitamin D3; ROVIMIX D3-500; DSM Nutritional Products, Basel, Switzerland)/kg feed, rep-resenting the minimum level according to recommenda-tions of the NRC (1998). The diet for treatment (DN) was fortified with 2,000 IU vitamin D3/kg feed (ROVIMIX D3-500; equivalent to 50 μg/kg feed), corresponding to the level used commercially in Switzerland (Boltshauser et al., 1993) and for treatment (HD) with 50 μg 25-hy-droxycholecalciferol (25-OH-D3; ROVIMIX Hy-D 1.25%; DSM Nutritional Products)/kg feed. Considering the substantial variability of the analytical method, feed analysis confirmed the correct mixing of the test products (DL = 280 IU/kg vitamin D3, DN = 1,680 IU/kg vitamin D3, and HD = 36 μg/kg 25-OH-D3).

At approximately 84 and 110 d postcoitum as well as at 1, 5, and 33 d PP (before weaning) of each repro-

Table 1. Composition of the experimental diets of sows and piglets (as-fed basis) for study 1Variable Gestation LactationIngredient, %

Barley 25.00 16.80Corn 3.20 15.00Wheat 3.00 16.00Soybean meal 43 14.80 8.05Fat 1.30 4.00l-Lysine HCl – 0.23dl-Methionine – 0.05Limestone 1.59 1.12Monocalcium phosphate 0.51 1.07Iodized salt 0.45 0.57Vitamin and mineral premix1 0.60 1.20

Calculated DE and analyzed nutrient contentDE, MJ/kg 11.6 14.0Dry matter, g/kg 896 897Crude protein, g/kg 150 175Crude fat, g/kg 35.5 58.5NDF, g/kg 264 175ADF, g/kg 105 65Ca, g/kg 7.6 8.1P, g/kg 6.6 6.9

1Supplied per kilogram diet (gestation + lactation): 18,000 IU vitamin A (acetate), 4 mg vitamin K, 18 mg d-pantothenic acid, 5 mg riboflavin, 34 mg niacin, 2.7 mg folacin, 0.1 mg d-biotin, 29 μg vitamin B12, and 0.98 g choline.

Vitamin D3 sources on sow reproduction 901

ductive cycle, 10 mL of plasma and 10 mL of EDTA-stabilized whole-blood samples were collected from the external jugular vein of each dam. Blood was centri-fuged (3,500 × g for 15 min at 4°C) and plasma was harvested, frozen (–80°C), and analyzed at a later time for 25-OH-D3, Ca, P, crosslaps (CL), and osteocalcin (OC). Within 24 h after parturition, a 100-mL sample of colostrum was collected by hand milking from ev-ery functional mammary gland after intramuscular in-jection with 1.5 mL oxytocin (LongActon; Vital AG, Oberentfelden, Switzerland). Further milk samples were collected in the same manner at 9 and 33 d PP. Colostrum and milk samples were frozen at –80°C and analyzed later for 25-OH-D3, Ca, and P. Colostrum and milk samples were only collected in parities 1, 2, and 3.

Experimental Design: Study 2

This study was performed at the pig farm “Sauen-zucht Kölsa” in Kölsa (Germany), which housed approxi-mately 2,000 sows. The multiparous sows belonged to a local breed (LeiCoMa) and the primiparous sows were coming from the farm’s own breeding stock. A total of 227 primiparous and multiparous sows were used for this trial: the control group contained 114 animals (32 primip-arous and 82 multiparous sows) and the test group 113 animals (28 primiparous and 85 multiparous sows). The sows were allocated to 1 of 2 treatments: control (CON; 2,000 IU vitamin D3/kg [equivalent to 50 μg/kg] feed [ROVIMIX D-500; DSM Nutritional Products]) or test (HYD; 50 μg 25-OH-D3/kg [ROVIMIX Hy-D 1.25%; DSM Nutritional Products]). All sows, inseminated with-in 1 wk, were allocated to a single treatment group (first week: CON and second week: HYD), so that after 2 wk all sows had been enrolled in the study. The sows were housed in single stands with slatted flooring during early pregnancy (1 to 85 d), flat decks with slatted flooring dur-ing late pregnancy (85 to 110 d), and farrowing decks with slatted flooring during farrowing and lactation (110 to 140 d). Temperature on the farm was kept between 20 and 25°C through heating, but humidity was not controlled.

During early pregnancy, all sows receiving the same treatment were kept in 1 room and were given the same feed in a common trough, which later in the day was used to supply drinking water. The feed was provided dry in mash form. During late pregnancy, the sows were given liquid feed in a common trough. The dry feed from the silo was suspended in water and released directly into the feeding trough. Drinking water was supplied separately. The farrowing decks were all supplied with feed from 1 silo only. This feed contained vitamin D3 and was given to the control (i.e., CON) animals while the feed containing 25-OH-D3 (i.e., HYD) was added by hand to the feeders of the test group.

Three feeds were used on this farm: 1) feed for early pregnancy (dry), 2) feed for late pregnancy (same com-position as 1 but liquid), and 3) lactation feed. The basal feeds contained wheat, barley, soybean meal, and sor-ghum as main ingredients, but the exact diet composition was not disclosed by the farm. Feed analysis revealed the following compositions: feeds 1 and 2: ME = 12.1 to 12.3 MJ/kg, DM = 88.2%, CP = 136 to 143 g/kg DM, crude fiber = 56 to 57 g/kg DM, crude fat = 32 to 34 g/kg DM, Ca = 6.0 to 6.2 g/kg DM, and P = 6.1 to.6.5 g/kg DM and feed 3: ME = 13.2 to 13.3 MJ/kg, DM = 88.8 to 89.2%, CP = 179 to 180 g/kg DM, crude fiber = 40 to 41 g/kg DM, crude fat = 45 to 46 g/kg DM, Ca = 8.2 to 8.4 g/kg DM, and P = 5.5 to 6.0 g/kg DM. A pooled sample of each experimental feed was analyzed for the presence of the 2 vitamin D3 sources (CON: vitamin D3 = 2.460 IU/kg and 25-OH-D3 = below detection limit; HYD: vitamin D3 = 890 IU/kg and 25-OH-D3 = 42.6 μg/kg).

Related to reproductive characteristics the percent-age of pregnant sows, the number and weight of piglets born (alive, or dead), and the total weight of weaned piglets were determined. Blood samples were taken by venipuncture of the vena cava cranialis on the day of AI, 28 and 80 d postinsemination, and on d 5 and 28 PP from 25 sows of each treatment group. The concentrations of the vitamin D3 metabolites 25-OH-D3, 1,25-dihydroxy-cholecalciferol (1,25-(OH)2–D3), Ca, and P as well as the concentration of the bone markers OC and CL were determined in blood plasma. Due to sows dying or being culled during the experiment or insufficient sample size, not all planned analyses could be performed.

Analytical Methods

Determination of 25-OH-D3 in plasma and feed samples was performed by DSM Nutritional Prod-ucts. The method was based on an isotope dilution assay using an Agilent 1100 reversed-phase HPLC-mass spectrometry system (Basel, Switzerland) with a trapping column for quantification. In colostrum and milk samples, 25-OH-D3 concentration was analyzed using a RIA kit (25-Hydroxy Vitamin D; IDS Immu-nodiagnostic Systems Ltd., Boldon, Tyne, and Wear, UK). Blood concentration of 1,25-(OH)2–D3 was de-termined from 250 μL plasma using a commercial RIA (1,25-dihydroxy Vitamin D RIA, catalog number AA-54; Immunodiagnostic Systems). The assay comprised an immune extraction for purification and an enrich-ment of 1,25-(OH)2–D3 followed by RIA with 125I as tracer. Calcium and P in plasma were determined by colorimetry with a UV–visible recording spectropho-tometer (UV-160A; Shimadzu Corporation, Kyoto, Japan) using commercial kits (Axon Lab AG, Baden, Switzerland). Analyses were based on the methylthy-

Weber et al.902

mol blue method for Ca (Gindler and King, 1972) and the phosphomolybdate method for P (Quadri and Sriv-astava, 1980). Plasma concentrations of OC and CL were measured using commercially available ELISA test kits (Metra OC; Quidel Corporation, San Diego, CA; Crosslaps; IDS Immunodiagnostic Systems Ltd.). For the determination of milk mineral content, the sam-ples were prepared according to the method described by Park et al. (1994) and then Ca and P were measured as in plasma samples.

Statistical Analysis

Statistical analysis for study 1 was performed us-ing the generalized least squares procedure (GLS) of R 2.8.1 (Venables et al., 2006) with treatment and parity as main effects. Values in tables represent means and the maximum SEM is stated. When significant (P < 0.05) dependent variable effects were found, mean values were separated using Wilcoxon’s sign rank test.

Statistical analyses for study 2 were performed with the statistical software R, version 3.0.1 (Venables et al., 2006) using additional packages nlme (version 3.1-109 [Pinheiro et al., 2013]) and car (version 2.0-18 [Fox and Weisberg, 2011]). All blood characteris-tics (except 1,25-(OH)2–D3) were log-transformed to base 10 because their distributions were skewed. To investigate the effect of the supplementation, a linear mixed-effects model with the fixed-effects treatment group, sampling time point (as factor), treatment × time interaction, and the random effect animal was used. The

reported treatment × time point interaction P-values de-scribe the additional influence of the supplementation at the respective time point adjusted for baseline differ-ences between groups and accounting for multiple mea-surements per sow. Fertility analyses were done using the exact Fisher test for proportions, ANOVA (adjusted for lactation period), and Poisson regression (adjusted for lactation period). A P-value of ≤0.05 was consid-ered significant whereas a P-value ≤0.10 but >0.05 was taken as an indication for a near-significant trend.

RESULTS

Blood Characteristics of Sows: Study 1The effects of treatment on plasma 25-OH-D3 lev-

els at the various measurement periods are presented in Fig. 1. Animals on diet HD had increased (P < 0.05) 25-OH-D3 plasma concentrations (between +40.5% in parity 3 [d 5 PP] and 332% in parity 1 [d 33 PP]) over the DL-group at any time of sampling while HD did not differ statistically from the DN group. Concentra-tions reached their maximum immediately before par-turition and decreased during the course of lactation. In treatment groups DL and DN, the least plasma levels of 25-OH-D3 were found at weaning whereas levels in HD animals were increasing again after the first week of lactation to the level of the PP samples.

Regarding circulating concentrations of Ca in plas-ma of sows at the various time points of measurement (Fig. 2), differences (P < 0.05) between treatments oc-

Table 2. Effect of dietary vitamin D supplementation and parity on sow reproductive performance in study 1

Variable

Treatment1 P-value2

Parity P-value

SEMDL DN HD 1 2 3 4

No. of farrowings 38 43 38 38 33 29 21Mean litter number 5 4 4 3 4 5 5Birth

Total born, no. 12.9 11.5 11.1 0.24 11.2 11.5 11.7 13.8 0.13 0.65Stillborn, no. 1.2 0.6 0.8 0.26 0.8 0.8 0.9 1.1 0.79 0.35Live, no. 11.6 10.9 10.2 0.15 10.2 10.7 10.8 12.7 0.36 0.69Pig wt., kg 1.4 1.6 1.5 0.14 1.5 1.6 1.5 1.6 0.25 0.07Litter wt., kg 17.1 17.7 15.4 0.30 15.9 16.2 16.8 19.4 0.08 1.18

Weaning, 35 dLive pigs, no. 10.0 9.6 8.8 0.10 9.2b 8.9b 9.5b 11.0a 0.04 0.45Loss,3 no. 1.8 1.6 2.0 0.71 1.5 1.8 2.1 2.1 0.10 0.52Pig wt., kg 7.0 8.7 7.3 0.47 7.0 8.2 7.8 8.2 0.21 0.40Gain,4 kg 5.4b 7.1a 5.8b <0.01 5.5b 6.4a 6.4a 6.6a <0.01 0.46Litter wt., kg 68.5 82.7 62.9 0.16 63.1c 71.2b 70.7b 89.3a 0.02 5.15

a–cLetters indicate a significant treatment effect (P < 0.05).11DL = 200 IU/kg vitamin D3; DN = 2,000 IU/kg vitamin D3; HD = 50 μg/kg 25-OH-D3.2From a generalized least squares procedure analysis including the effects of treatment and parity as fixed effects. 30 to 35 d.4Weight gain between 0 and 35 d.

Vitamin D3 sources on sow reproduction 903

curred only on d 84 of gestation during parity 3, where levels of Ca in the DN group decreased significantly by 17.4% compared with the DL group. In all treatment groups, there was a decrease in circulating levels of Ca during gestation and a subsequent increase after farrow-ing, which lasted until midlactation. Over time, a de-cline in plasma Ca levels between parity 1 and 4 was ob-served. Plasma P concentrations were similar between treatments (Fig. 3). Only at d 110 of gestation in parity 1, circulating P levels were greater by 15.1 and 17.2%, respectively (P < 0.05), in DN and HD sows than in the DL-group. Common to all treatment groups, P concen-trations clearly declined with advancing parity.

Mean OC concentrations did not differ between treat-ments at any point of sampling, but a decline was ob-served throughout gestation in all treatment groups fol-lowed by an increase in the subsequent lactation (Fig. 4).

Over time, there was a general decrease in plasma OC concentrations from parity 1 to 4. The course of the curve of the circulating CL concentrations was similar for all treatment groups with a drop at birth followed by a more or less pronounced increase during lactation (Fig. 5). On d 1 of parity 4, a significant decrease of CL by 11.1% (P < 0.05) was observed in HD sows in comparison to DL.

Vitamin D Concentration of Colostrum and Milk: Study 1

The supplementation of 25-OH-D3 clearly affected 25-OH-D3 concentrations in milk (Fig. 6). In the first parity, colostrum 25-OH-D3 concentration was greatest (P < 0.05) in samples of HD animals (12.3% above DL) whereas 25-OH-D3 concentrations in colostrum of the following parities did not differ between treatment groups. Later milk samples of HD sows contained between 42.4

Table 3. Concentration of the vitamin D metabolites 25-hydroxycholecalciferol (25-OH-D3) and 1,25-dihydroxycholecal-ciferol (1,25-(OH)2–D3), the bone markers crosslaps and osteocalcin, and the minerals calcium and phosphorus in plasma of sows in study 2 (means ± SD). Results of the statistical analyses as described in the text are reported as P-values.

Variable

Treatment1

Parturition LactationInsemination 28 d PI2 80 d PI2 5 d PP2 28 d PP2

25-OH-D3, ng/mL

CON 14.0 ± 5.65 22.6 ± 8.43 17.1 ± 5.47 11.6 ± 4.98 15.5 ± 4.27(n = 25) (n = 19) (n = 15) (n = 16) (n = 15)

HYD 23.0 ± 7.64 48.0 ± 10.6 41.3 ± 9.10 26.3 ± 5.78 43.0 ± 16.1(n = 25) (n = 20) (n = 15) (n = 16) (n = 16)

P-value <0.0001 0.048 0.013 0.019 0.0021,25-(OH)2–D3, pg/mL

CON 115.6 ± 44.8 131.2 ± 18.1 151.0 ± 27.1 127.1 ± 28.1 149.9 ± 32.7(n = 24) (n = 19) (n = 15) (n = 16) (n = 15)

HYD 121.3 ± 45.5 146.0 ± 30.7 155.6 ± 26.7 189.6 ± 40.0 153.8 ± 37.5(n = 25) (n = 20) (n = 16) (n = 16) (n = 16)

P-value 0.584 0.507 0.888 <0.001 0.998Calcium, mmol/L

CON 2.47 ± 0.11 2.31 ± 0.14 2.41 ± 0.09 2.41 ± 0.12 2.33 ± 0.12(n = 25) (n = 20) (n = 16) (n = 15) (n = 15)

HYD 2.49 ± 0.13 2.30 ± 0.12 2.41 ± 0.10 2.31 ± 0.12 2.39 ± 0.14(n = 25) (n = 20) (n = 16) (n = 16) (n = 16)

P-value 0.718 0.626 0.826 0.028 0.436Phosphorus, mmol/L

CON 1.99 ± 0.35 2.04 ± 0.44 1.95 ± 0.37 1.61 ± 0.24 1.46 ± 0.28(n = 25) (n = 20) (n = 16) (n = 12) (n = 10)

HYD 2.10 ± 0.59 2.11 ± 0.36 2.09 ± 0.53 1.74 ± 0.15 1.57 ± 0.21(n = 25) (n = 20) (n = 16) (n = 7) (n = 9)

P-value 0.468 0.967 0.778 0.544 0.526Osteocalcin, ng/mL

CON 173.7 ± 54.6 178.4 ± 43.2 156.3 ± 64.5 104.1 ± 45.1 126.9 ± 34.1(n = 17) (n = 16) (n = 15) (n = 16) (n = 15)

HYD 203.0 ± 74.8 172.7 ± 41.5 169.4 ± 68.8 117.4 ± 27.5 145.3 ± 46.3(n = 19) (n = 16) (n = 15) (n = 16) (n = 16)

P-value 0.247 0.238 0.745 0.678 0.984Crosslaps, ng/mL

CON 0.95 ± 0.29 1.03 ± 0.27 1.02 ± 0.31 2.20 ± 0.40 3.64 ± 1.02(n = 17) (n = 16) (n = 15) (n = 16) (n = 15)

HYD 0.74 ± 0.24 0.99 ± 0.29 1.21 ± 0.22 2.15 ± 0.49 3.31 ± 0.94(n = 19) (n = 16) (n = 15) (n = 16) (n = 16)

P-value 0.010 0.122 0.001 0.10 0.25

1CON = 2,000 IU/kg vitamin D3; HYD = 50 μg/kg 25-OH-D3.2PI = postinsemination; PP = postpartum.

Weber et al.904

and 129% more (P < 0.05) 25-OH-D3 than those from DL and DN animals in parities 2 and 3 but not in parity 1. There was no effect of dietary treatment on Ca con-centrations of colostrum and milk, but contents generally increased from the start of lactation to weaning (Fig. 7). Concentrations of P were similar at the various sampling times except for d 9 of lactation in parity 3, where sam-ples of DN sows contained significantly more (+29.5%) of that mineral (P < 0.05; Fig. 8) than the DL sows.

Reproductive and Litter Performance: Study 1

The main effects of treatment and parity on parturi-tion and litter performance are reported in Table 2. Mean litter number between the groups was similar and treat-ment had no effect on the number of piglets born, the number of stillborn piglets, or on litter weight at birth. At weaning, litter gain, loss of piglets during the suck-ling period, and individual pig weight were not affected by vitamin D source and level whereas there was a trend (P = 0.10) towards a smaller number of piglets weaned in treatment HD. Weight gain of the suckling piglets, however, was greatest (P < 0.01) when diet DN was pro-vided. The number of piglets born, the number of piglets

weaned, and weight gains of individual piglets and lit-ters increased (P < 0.05) with advancing parity of dams.

Blood Characteristics of Sows: Study 2

Supplementation of the diet with 25-OH-D3 (HYD) resulted in an increase (P < 0.05) in this vitamin D3 metabolite in blood by 36.5 to 177.4% compared with the CON-group (Table 3) but did not influence the con-centration of 1,25-(OH)2–D3 during pregnancy. In both groups, the plasma concentration of 1,25-(OH)2–D3 in-creased before parturition to almost the same extent (Ta-ble 3). Yet 5 d PP the concentration of 1,25-(OH)2–D3 was significantly greater in treatment HYD compared with CON (P < 0.05) whereas on d 28 PP no difference among the treatments was observed again.

The plasma concentrations of Ca and P varied little during pregnancy and lactation and did not differ be-tween the 2 treatment groups, except on d 5 PP, where Ca was less in the HYD than in the CON group (P = 0.028; Table 3). The plasma levels of OC showed consid-erable variations, but no consistent differences between the treatments were recorded. In both groups, there was a trend of a reduction of OC levels during parturition to decrease at 5 d PP and then to increase to the end of lactation (28 d PP). The CL concentrations increased in both treatments subsequent to parturition and were greatest at the end of lactation. The plasma levels of CL were reduced (P = 0.01) in the HYD group at insemina-tion and tended to be reduced (P = 0.10) during early lactation (5 d PP), but CL concentration was greater (P = 0.001) in HYD pigs during late pregnancy in compari-son with CON pigs (80 d postinsemination; Table 3).

Table 4. Effect of dietary vitamin D supplementation on reproductive performance of sows in study 2 (count, proportion [in %], or mean ± SD) Variable

CON1

HYD1

Statistical analysis method

P-valueTotal number of sows 114 113Total pregnant sows 101 104Proportion of pregnancies 89% 92% Fisher test 0.502Total farrowed sows 89 93Proportion of farrowed sows 78% 82% Fisher test 0.506Total piglets born alive 966 1,058Number of piglets born alive per farrowed sow

10.85 ± 3.05 11.38 ± 3.60 ANOVA2 0.43

Total piglets born dead3 92 120Number of piglets born dead per farrowed sow

1.03 ± 1.55 1.29 ± 1.80 Poisson regression2

0.163

Total piglets losses4 115 155Proportion of lost piglets of piglets born alive

12% 15% Fisher test 0.077

Total piglets weaned 837 880Proportion of weaned piglets of piglets born alive

87% 83% Fisher test 0.03

Total litter weight, kg 1,383 1,622 ANOVA2 0.027Birth weight per piglet, kg 1.44 ± 0.27 1.55 ± 0.24 ANOVA2 0.008Total weaning weight, kg 5,800 6,473Weaning weight per piglet, kg 7.10 ± 1.34 7.53 ± 1.50 ANOVA2 0.15

1CON = 2,000 IU/kg vitamin D3; HYD = 50 μg/kg 25-OH-D3.2Analysis adjusted for lactation period.3Including mummified embryos.4Losses within 3 d postpartum.

Figure 1. Effect of vitamin D3 source and level on plasma 25-hydroxy-cholecalciferol (25-OH-D3) concentrations (mean ± SE) during gestation (G) and lactation (L) over 4 consecutive parities in study 1. Circulating 25-OH-D3 concentrations differed between treatments DL (200 IU/kg vitamin D3) and HD (50 μg/kg 25-OH-D3; *P < 0.05) at any time of sampling. A minimum of 7, 9, and 7 sows were used for the treatments DL, DN (2,000 IU/kg vitamin D3) and HD, respectively.

Vitamin D3 sources on sow reproduction 905

Reproductive and Litter Performance: Study 2

From 114 sows of the CON-group that were artifi-cially inseminated, 89% became pregnant and 78% even-tually farrowed (Table 4). In the HYD group, 92% of 113 inseminated sows became pregnant and 82% of them far-rowed. These proportions were not different between the 2 treatments. Although the absolute number of piglets born alive and of piglets weaned was numerically greater in the HYD group, the proportion of piglets born alive that were weaned was less (P = 0.03) in HYD when compared with CON, due to a trend of a greater (P = 0.077) proportion of lost piglets to piglets born alive. However, the total litter weight (P = 0.027) as well as the birth weight per piglet (P = 0.008) was greater than the control (CON) in the supple-mented group (HYD; Table 4).

DISCUSSION

The aim of the present studies was to investigate the vitamin D3 requirements of sows under current husbandry conditions and to compare supplementation of vitamin D3 with 25-OH-D3 in swine. Using different sources and levels of vitamin D3 did not strongly affect any mater-nal plasma measurement except 25-OH-D3. Circulating 25-OH-D3 levels are known to be directly related to di-etary vitamin D3 intake (Hollis and Wagner, 2006) and, therefore, regarded as a good index of nutritional vitamin D3 status (Salle et al., 2000). Plasma 25-OH-D3 levels of sows in both studies showed greatest concentrations in animals receiving 25-OH-D3 as the vitamin D3 source. As reported by Horst and Littledike (1982), 25-OH-D3 circu-lates in most species at 30 to 50 ng/mL. In study 1, plasma concentrations of DN and DL animals were found within that range whereas circulating levels of 25-OH-D3 in HD sows exceeded that range at any time of sampling due

to dietary 25-OH-D3 administration. In the more practi-cal study 2, all 25-OH-D3 concentrations were consider-ably less with levels from 11.6 to 22.6 ng/mL in the CON and 23.0 to 48.0 ng/mL in the HYD group. For humans, Holick (2007) considered 25-OH-D3 plasma levels below 20 ng/mL as insufficient and over 30 ng/mL as adequate. Assuming a similar situation in swine, the CON sows in the commercial farm were deficient in vitamin D3 al-though they received 2,000 IU vitamin D3/kg feed, which represents the maximum legal dietary vitamin D3 level in the European Union. Only the HYD animals, which were supplemented with 25-OH-D3 via the feed, were suffi-ciently supplied with vitamin D3 activity.

Hughes et al. (1977) reported that hydroxylation of 25-OH-D3 to 1,25-(OH)2–D3, the active form of vitamin D3, was not enhanced by increasing the dietary intake of vitamin D3, but it was rather limited by the availability of its substrate. It seems likely that increasing Ca de-mand during pregnancy and lactation should result in greater requirements for the precursor of 1,25-hydrox-ylation and, thus, may cause a decrease in plasma 25-OH-D3. In this respect the observation of a significantly smaller 1,25-(OH)2–D3 concentration during early lac-tation in the CON-group compared with the HYD sows in study 2 was interesting. This finding could indicate that at the onset of milk production, the demand for Ca was increased to such an extent that in CON sows not enough 25-OH-D3 may have been available to sustain the required production of 1,25-(OH)2–D3. Furthermore, it was remarkable that, in contrast to animals of the oth-er treatment groups, plasma 25-OH-D3 concentration in sows supplemented with this vitamin D3 metabolite (HD and HYD) increased already shortly after parturi-tion and reached prepartum levels at weaning. This ob-servation indicates that dietary 25-OH-D3 is more read-

Figure 2. Effect of vitamin D3 source and level of on plasma Ca con-centrations (mean ± SE) in study 1. A minimum of 7, 9, and 7 sows were used for the treatments DL (200 IU/kg vitamin D3), DN (2,000 IU/kg vitamin D3) and HD (50 μg/kg 25-OH-D3), respectively. *indicates a significant difference between the treatments DN and DL (P < 0.05). G = gestation; L = lactation.

Figure 3. Effect of vitamin D3 source and level on plasma P concentra-tions (mean ± SE) in study 1. A minimum of 7, 9, and 7 sows were used for the treatments DL (200 IU/kg vitamin D3), DN (2,000 IU/kg vitamin D3) and HD (50 μg/kg 25-OH-D3), respectively. *indicates a significant difference between the treatments DL and DN or HD(P < 0.05). G = gestation; L = lactation.

Weber et al.906

ily available to rapidly reestablish a steady state in the sows after an exceptional demand.

In all treatments of the present studies, plasma Ca concentrations were within the normal reference range, as described by Friendship et al. (1984) and Reese et al. (1984), and were not affected by dietary treatment. Mater-nal adaptations of Ca homeostasis are known to differ be-tween pregnancy and lactation. According to Meissonier et al. (1980) and Lingaas et al. (1992), plasma Ca of sows decreased with progressing parity, which was confirmed by Liesegang et al. (2005) and which is consistent with the observations in study 1. Likewise in the rat, plasma Ca has been reported to decrease during the last several days of pregnancy (Garner et al., 1988). Kovacs and Kro-nenberg (1997) suggested that maternal losses of Ca to a litter of rapidly growing fetuses probably exceed the maternal Ca stores (Garel, 1987). Since plasma Ca con-centrations in sows on both treatments remained stable, these animals apparently absorbed enough Ca from the feed to keep a balanced mineral homeostasis.

Mean plasma P concentration in pigs varies with the amount of P available in the diet (McDowell, 1992). Be-cause in both studies P concentrations in plasma were found to be similar to the reference values obtained by Verheyen et al. (2007) in commercial sow herds, the diet composition was adequate for the P requirement of sows. Nevertheless, circulating concentrations of P in the present studies varied during the course of the reproductive cycle. The decrease at the beginning of lactation in most treatment groups was probably attributable to the sudden increase in P demand for milk production. The subsequent increase between d 5 of lactation and weaning (only in study 1) may be an indica-tion of the increased skeletal resorption and decreased renal excretion (Kovacs and Kronenberg, 1997). Furthermore, as reported by Meissonnier et al. (1980) and Girard et al.

(1996), plasma P concentrations were greater in first parity sows than in sows during later parities. It seems likely that lactation intensity, as described above, was responsible for the decrease of both plasma Ca and P.

In women, circulating levels of bone formation markers have been found to be decreased in early ges-tation and to rise slightly by term (Cross et al., 1995). This was not the case in our studies where circulat-ing OC concentrations decreased in late gestation and reached the least peak at parturition or early lactation. The subsequent increase in bone formation during lacta-tion, however, has been described previously for wom-en (Cross et al., 1995; Salle et al., 2000) and for sows (Liesegang et al., 2005; Verheyen et al., 2007). Con-sistent with previous reports (Friendship et al., 1984; Tumbleson et al., 1986; Verheyen et al., 2007), plasma OC concentrations were greater in first parity sows than in older animals (study 1), which may be associated with the achievement of skeletal maturity.

As in humans (Cross et al., 1995), levels of the bone resorption marker, CL, in these studies increased dur-ing late gestation and lactation, indicating that maternal skeletal Ca stores were mobilized during the time of rap-id fetal accretion of Ca (Kovacs and Kronenberg, 1997). In study 1, circulating levels of CL were less during the first lactation compared with later lactations, indicating that the extra Ca necessary for milk production was mo-bilized from the skeleton of first parity sows to a smaller extent than of older sows. Except for a near-significant trend for decreased CL during treatment HYD (P = 0.10) in study 2 (5 d PP), which could indicate reduced bone resorption, the CL concentrations were not different among the treatments. Nevertheless, it is probable that supplementing the diet of particularly multiparous sows with a more readily available source of vitamin D3 may

Figure 4. Effect of vitamin D3 source and level on plasma osteocalcin (OC) concentrations (mean ± SE) in study 1. A minimum of 7, 9, and 7 sows were used for the treatments DL (200 IU/kg vitamin D3), DN (2,000 IU/kg vitamin D3) and HD (50 μg/kg 25-OH-D3), respectively. G = gestation; L = lactation.

Figure 5. Effect of vitamin D3 source and level on plasma crosslaps (CL) concentrations (mean ± SE) in study 1. A minimum of 7, 9, and 7 sows were used for the treatments DL (200 IU/kg vitamin D3), DN (2,000 IU/kg vitamin D3) and HD (50 μg/kg 25-OH-D3), respectively. *indicates a significant difference be-tween the treatments HD and DL or DN (P < 0.05). G = gestation; L = lactation.

Vitamin D3 sources on sow reproduction 907

have a sparing effect on skeletal Ca stores by enhancing alternative mechanisms of Ca acquisition, such as intes-tinal absorption and renal resorption.

Brommage and DeLuca (1984) previously reported that although milk production is reduced in vitamin D-deficient rats, no severe deficiency in any essen-tial nutrient other than vitamin D3 could be detected. Even though the transfer of vitamin D3 metabolites into breast milk is rather limited (Hollis and Wagner, 2004), milk 25-OH-D3 concentrations were found to corre-spond with maternal plasma 25-OH-D3 concentrations and maternal vitamin D3 intake (Specker et al., 1985). Consistent with that, sows receiving 25-OH-D3 as the vitamin D3 source (i.e., treatment HD in study 1) se-creted greater concentrations of 25-OH-D3 into their milk but not into their colostrum. This is consistent with observations in rats where, during the very early stages of lactation, the vitamin was present mainly in the form of vitamin D3 (not determined in our study), which declined within several days so that the predominant metabolite was 25-OH-D3 (Clements and Fraser, 1988). Because of reported cases of rickets in piglets due to a lack of vitamin D in the milk (Hartmann and Holmes, 1989), enhancing vitamin D3 content of milk by supple-menting the sow’s diet might be a possible strategy to help maintain Ca homeostasis in the suckling piglet.

The reproductive performance of the animals was not substantially affected by dietary treatment in study 1. Even when offered the minimum vitamin D3 supplemen-tation (i.e., DL group) over a period of 4 parities, sows were able to reproduce and numbers of live and dead born piglets as well as piglet birth weights were similar among the 3 treatment groups. Results from experiments in rats (Gray et al., 1979; Halloran and DeLuca, 1980)

and humans (Kovacs and Kronenberg, 1997) confirmed that vitamin D3 is not directly necessary for reproduction because, due to homeorhetic mechanisms, the fetal-pla-cental unit is able to meet its needs, irrespective of ma-ternal Ca or vitamin D3 levels. Moreover, because only maternal 25-OH-D3 but not 1,25-(OH)2–D3 crosses the placenta (Haddad et al., 1971), circulating levels of 1,25-(OH)2–D3 in the fetus must be derived from fetal sources. At parturition, Ca metabolism of the fetus undergoes sub-stantial changes. After cutting of the umbilical cord, the placental Ca infusion is abruptly lost and, therefore, the adaptive goals of the neonate are to quickly turn on para-thyroid hormone and 1,25-(OH)2–D3 synthesis, which in turn upregulates intestinal Ca absorption (Kovacs and Kronenberg, 1997). Therefore, the importance of vitamin D3 in maintaining normal neonatal weight gain, plasma mineral concentrations, and normal skeletal develop-ment is negligible before the suckling period and in-creases progressively during the later stages of lactation (Brommage and Neuman, 1981; Halloran and DeLuca, 1979, 1981). It was therefore concluded that the growth retardation of DL progeny in the present study probably was not started until the later stages of lactation, when vitamin D3–mediated Ca transport mechanisms become activated. It has been shown previously that growth of piglets is predominantly limited by the amount of milk produced by the sow (Pluske and Dong, 1998), which has been shown to be significantly correlated with the size of the litter suckled, the litter weight at weaning, and the change in weight of the sows during lactation. It is therefore concluded that differences in piglet growth rate are in the first line caused by differences in the amount of milk produced by their dam. Brommage and DeLuca (1984) reported reduced milk production and, thus, re-tarded pup growth in vitamin D-deficient rats.

Figure 6. Colostrum (d 1) and mature milk (d 9 and 33) 25-hydroxy-cholecalciferol (25-OH-D3) concentrations (mean ± SE) during 3 consecu-tive lactations in study 1; n = 9 for all parities. a–cLetters indicate significant treatment effects (P < 0.05). DL = 200 IU/kg vitamin D3; DN = 2,000 IU/kg vitamin D3; HD = 50 μg/kg 25-OH-D3.

Figure 7. Colostrum (d 1) and mature milk (d 9 and 33) Ca concen-trations (mean ± SE) of 3 consecutive lactations in study 1; n = 9 for all parities. DL = 200 IU/kg vitamin D3; DN = 2,000 IU/kg vitamin D3; HD = 50 μg/kg 25-OH-D3.

Weber et al.908

The results of the field trial (i.e., study 2) indicated an interesting effect of the dietary source of vitamin D3 on the productivity of sows. Although there were no differences among the treatments in the number of pig-lets born alive per farrowed sow, the total litter weight and the birth weight per piglet were significantly in-creased in sows supplemented with 25-OH-D3 (HYD) when compared with the controls (CON) fed an equiva-lent level of vitamin D3. It is suggested that maternal 25-OH-D3 beneficially supported the intrauterine de-velopment of the embryos.

SUMMARY AND CONCLUSIONS

In summary, the supplementation of the diet of sows during gestation and lactation with 25-OH-D3 clearly improved blood vitamin D3 status and also 25-OH-D3 content of milk. Accordingly, minimum maternal supply with vitamin D3 resulted in decreased concentration of 25-OH-D3 in blood and milk. These results indicate that an adequate vitamin D3 supplementation of the sow helps to maintain maternal Ca homeostasis during periods of strong demand such as gestation and lactation. Further-more, supplementation of the sow with 25-OH-D3 im-proved birth weight of the piglets, but the gestational vita-min D3 status of the mother had no influence on neonatal development. Although milk concentration of 25-OH-D3 increased when 25-OH-D3 was fed, the greater vitamin D3 content did not markedly influence piglet growth. The tendency towards diminished bone turnover in multipa-rous sows receiving 25-OH-D3 is promising.

LITERATURE CITEDBoltshauser, M., M. Jost, J. Keller, and P. Stoll. 1993. Fütterungsemp-

fehlungen und Nährwerttabellen für Schweine. (In German.) Landwirtschaftliche Lehrmittelzentrale, Zollikofen, Switzerland.

Brommage, R., and H. F. DeLuca. 1984. Vitamin D-deficient rats produce reduced quantities of nutritionally adequate milk. Am. J. Physiol. Endocrinol. Metab. 246:E221–E226.

Brommage, R., and W. F. Neuman. 1981. Growth failure in vitamin D-deficient rat pups. Calcif. Tissue Int. 33:277–280.

Clements, M. R., and D. R. Fraser. 1988. Vitamin D supply to the rat fetus and neonate. J. Clin. Invest. 81:1768–1773.

Cross, N. A., L. S. Hillmann, S. H. Allen, and G. F. Krause. 1995. Changes in bone mineral density and markers of bone remod-eling during lactation and postweaning in women consuming high amounts of calcium. J. Bone Miner. Res. 10:1312–1320.

DeLuca, H. F. 1978. Vitamin D. In: H. F. De Luca, editor, Handbook of lipid research. Plenum Press, New York. p. 69–132.

Engblom, L., L. Eliasson-Selling, N. Lundheim, K. Belák, K. An-dersson, and A.-M. Dalin. 2008a. Post mortem findings in sows and gilts euthanised or found dead in a large Swedish herd. Acta Vet. Scand. 50:25–34.

Engblom, L., N. Lundeheim, E. Strandberg, M. P. Schneider, A. M. Da-lin, and K. Andersson. 2008b. Factors affecting length of produc-tive life in Swedish commercial sows. J. Anim. Sci. 86:432–441.

Friendship, R. M., J. H. Lumsden, I. McMillan, and M. R. Wilson. 1984. Hematology and biochemistry values for Ontario swine. Can. J. Comp. Med. 48:390–393.

Fox, J., and S. Weisberg. 2011. An {R} companion to applied regres-sion, second ed. Sage, Thousand Oaks, CA.

Garel, J.-M. 1987. Hormonal control of calcium metabolism during the reproductive cycle in mammals. Physiol. Rev. 67:1–66.

Garner, S. C., T. C. Peng, and S. U. Toverud. 1988. Modulation of plasma parathyroid hormone and ionized calcium concentra-tions during reproduction in rats fed a low calcium diet. J. Bone Miner. Res. 3:319–323.

Giesemann, M. A., A. J. Lewis, P. S. Miller, and M. P. Akhter. 1998. Effects of reproductive cycle and age on calcium and phosphorus metabolism and bone integrity in sows. J. Anim. Sci. 76:796–807.

Gindler, E. M., and J. D. King. 1972. Rapid colorimetric determina-tion of calcium in biological fluids with methylthymol blue. Am. J. Clin. Pathol. 58:376–382.

Girard, C. L., S. Robert, J. J. Mattte, C. Farmer, and G.-P. Martineau. 1996. Plasma concentrations of micronutrients, packed cell vol-ume, and blood hemoglobin during the first two gestations and lactations of sows. Can. J. Vet. Res. 60:179–185.

Gray, T. K., G. E. Lester, and R. S. Lorene. 1979. Evidence for extra-renal 1α-hydroxylation of 25-hydroxyvitamin-D3 in pregnancy. Science 204:1311–1313.

Haddad, J. G., Jr., V. Boisseau, and L. V. Avioli. 1971. Placental transfer of vitamin D3 and 25-hydroxycholecalciferol in the rat. J. Lab. Clin. Med. 77:908–915.

Halloran, B. P., E. Barthell, and H. F. DeLuca. 1979. Vitamin D me-tabolism during pregnancy and lactation in the rat. Proc. Natl. Acad. Sci. USA 76:5549–5553.

Halloran, B. P., and H. F. DeLuca. 1979. Vitamin D deficiency and reproduction in rats. Science 204:73–74.

Halloran, B. P., and H. F. DeLuca. 1980. Effect of vitamin D defi-ciency on fertility and reproductive capacity in the female rat. J. Nutr. 110:1573–1580.

Halloran, B. P., and H. F. DeLuca. 1981. Effect of vitamin D defi-ciency on skeletal development during early growth in the rat. Arch. Biochem. Biophys. 209:7–14.

Figure 8. Colostrum (d 1) and mature milk (d 9 and 33) P concentrations (mean ± SE) of 3 consecutive lactations in study 1; n = 9 for all parities. a,bLetters indicate significant treatment effects (P < 0.05). DL = 200 IU/kg vitamin D3; DN = 2,000 IU/kg vitamin D3; HD = 50 μg/kg 25-OH-D3.

Vitamin D3 sources on sow reproduction 909

Hartmann, D. A., and M. A. Holmes. 1989. Sow lactation. In: L. Bar-nett and D. P. Hennessy, editors, Manipulating pig production. V. J. Aust. Pig Sci. Assoc., Werribee, Australia. p. 72–97.

Holick, M. F. 2007. Vitamin D deficiency. N. Engl. J. Med. 357:266–281.Hollis, B. W., and C. L. Wagner. 2004. Assessment of dietary vi-

tamin D requirements during pregnancy and lactation. Am. J. Clin. Nutr. 79:717–726.

Hollis, B. W., and C. L. Wagner. 2006. Nutritional vitamin D status dur-ing pregnancy: Reasons for concern. Can. Med. Assoc. J. 174:-1290.

Horst, R. L., and E. T. Littledike. 1982. Comparison of plasma con-centrations of vitamin D and its metabolites in young and aged domestic animals. Comp. Biochem. Physiol. 73:485–489.

Hughes, M. R., D. J. Baylink, W. A. Gonnerman, S. U. Toverud, W. K. Ramp, and M. R. Haussler. 1977. Influence of dietary vita-min D3 on the circulating concentration of its active metabolites in the chick and rat. Endocrinology 100:799–806.

Jørgensen, B., and M. T. Jørgensen. 1998. Different rearing intensi-ties of gilts: II. Effects on subsequent leg weakness and longev-ity. Livest. Prod. Sci. 54:167–171.

Kirk, R. K., B. Svensmark, L. P. Ellegard, and H. E. Jensen. 2005. Locomotive disorders associated with sow mortality in Danish pig herds. J. Vet. Med., A 52:423–428.

Kovacs, C. S., and H. M. Kronenberg. 1997. Maternal-fetal calcium and bone metabolism during pregnancy, puerperium and lacta-tion. Endocr. Rev. 18:832–872.

Liesegang, A., L. Loch, E. Bürgi, and J. Risteli. 2005. Influence of phy-tase added to a vegetarian diet on bone metabolism in pregnant and lactating sows. J. Anim. Physiol. Anim. Nutr. 89:120–128.

Lingaas, F., E. Brun, and E. Aarskaug. 1992. Biochemical blood pa-rameters in pigs. 2. Estimates of heritability of 20 blood param-eters. J. Anim. Breed. Genet. 109:281–290.

McDowell, L. R. 1992. Minerals in animal and human nutrition. Ac-ademic Press, San Diego, CA.

Meissonnier, E., O. Ursache, and L. Chevrier. 1980. Profils biochi-miques et hématologiques chez les truies reproductrices: Influ-ence du stade physiologique et du numéro de cycle de reproduc-tion. (In French.) Journ. Rech. Porcine Fr., [C. R.] 12:317–326.

NRC. 1998. Nutrient requirements of swine. 10th ed. National Acad-emy Press, Washington, DC.

Park, Y. W., M. Kandeh, K. B. Chin, W. G. Pond, and L. D. Young. 1994. Concentrations of inorganic elements in milk of sows selected for high and low plasma cholesterol. J. Anim. Sci. 72:1399–1402.

Pinheiro, J., D. Bates, S. DebRoy, D. Sarkar, and the R Development Core Team. 2013. nlme: Linear and nonlinear mixed effects models. R package version 3.1–109.

Pluske, J. R., and G. Z. Dong. 1998. Factors influencing the utili-zation of colostrum and milk. In: M. W. A. Verstegen, P. J. Moughan, and J. W. Schrama, editors, The lactating sow. Wa-geningen Pers, Wageningen, The Netherlands. p. 45–70.

Quadri, M. A., and R. K. Srivastava. 1980. Micro-determination of inorganic phosphorus and alkaline phosphatase activity in bio-logical fluids. Arch. Exp. Veterinarmed. 34:667–672.

Reese, D. E., E. R. Peo, J. Lewis, and A. Hogg. 1984. Plasma chemi-cal values of gestating and lactating swine: Reference values. Am. J. Vet. Res. 45:978–980.

Salle, B. L., E. E. Delvin, A. Lapillonne, N. J. Bishop, and F. H. Glo-rieux. 2000. Perinatal metabolism of vitamin D. Am. J. Clin. Nutr. 71:1317S–1324S.

Specker, B. L., R. C. Tsang, and B. W. Hollis. 1985. Effect of race and diet on human milk vitamin D and 25-hydroxyvitamin D. Am. J. Dis. Child. 139:1134–1137.

Tumbleson, M. E., D. A. Schmidt, and E. Scholl. 1986. Hematology and clinical chemistry. In: A. D. Leman, B. Straw, R. D. Glock, W. L. Mengeling, R. C. H. Penny, and E. Scholl, editors, Diseases of swine, 6th ed. Iowa State Univ. Press, Ames, IA. p. 27–43.

Venables, W. N., D. M. Smith, and the R Development Core Team. 2006. A language and environment for statistical computing. R Foundation for Statistical Computing, Vienna, Austria. www.r-project.org/ (accessed 26 Sep. 2013).

Verheyen, A. J. M., D. G. D. Maes, B. Mateusen, P. Deprez, G. P. J. Jans-sens, L. DeLange, and G. Counotte. 2007. Serum biochemical ref-erence values for gestating and lactating sows. Vet. J. 174:92–98.