Vitamin D deficiency increases the risk of bone fracture due to osteoporosis and decreases muscle strength. Recent investigations show a relationship between vitamin D deficiency and other afflictions such as cancer, reduced immune defence and CVDReference Zittermann1–Reference Giovannucci, Liu and Rimm3. During the summer period, the primary source of vitamin D for man exposed to sunlight is the metabolism of 7-dehydrocholesterol to pre-vitamin D3 in the skin by UV B radiation (290–315 nm), whereas vitamin D in food is the secondary source. In winter oral intake of vitamin D may be the primary source, as absorption through the skin is limited at latitudes above 35°, e.g. for 4 months in Boston, USA (42°N) and for 6 months in Bergen, Norway (61°N)Reference Holick4. Similarly, oral intake of vitamin D is the primary source all year round for people not exposed to sunlight due to confinement indoors or clothing.
Estimation of dietary intake of vitamin D is essential for investigating the influence of vitamin D on health parameters in a population as well as in human intervention studies. Such calculations are based on dietary intake data from dietary surveys combined with the content of nutrients available in food composition tables. Until 10–15 years ago vitamin D data in food composition tables was mainly derived from biological assays, which used the ability of vitamin D to cure rickets in vitamin D-deficient rats5, Reference Busck6. For the last 10 years, food composition tables have included specific values for vitamin D3 (vitD3) and 25-hydroxyvitamin D3 (25OHD3)Reference Clausen, Jakobsen, Leth and Ovesen7–12. To calculate the total vitamin D content, the relative activity between 25OHD3 and vitD3 is required.
The studies conducted to assess this factor were performed 30–40 years ago in deficient rats, in which the estimated values were between 1·4 and 5Reference Blunt, Tanaka and Deluca13–Reference Tanaka, Frank and Deluca16. However, to date, no consensus has been establishedReference Ovesen, Brot and Jakobsen17.
The aim of the present study was to investigate the relative activity between vitD3 and 25OHD3 in pigs, as a model for man. The end-points were plasma 25OHD3, and the content of vitD3 and 25OHD3 in the pork loin and liver. Vitamin Dtotal throughout the paper is defined as the sum of vitD3 and 25OHD3.
Materials and methods
Pigs
The twenty-four pigs selected for the present study were a subgroup of 3225 healthy pigs used in a feeding-trial conducted to investigate whether the productivity of the pigs was affected when vitD3 was replaced by 25OHD3. The feeding-trial was conducted at an ordinary Danish farm in stables with partially slatted floors and cover. A computer-controlled system distributed the feed to the separate double pens. Each double pen contained forty-five pigsReference Maribo, Nielsen and Jakobsen18. The pigs were fed via a tube feeder with nipple drinks as well as drinking bowls. The pigs were raised without exposure to sunlight. The stable lighting was produced with lamps (F36W/T8/33-630; Osram, Sylvania, MA, USA).
Experimental design
The study was performed as a supplementation study and consisted of a parallel trial with three treatments of vitamin D. Vitamin D was given at equal levels as vitD3, as a mixture of vitD3 and 25OHD3, or as 25OHD3, from weaning at an age of 5 weeks to slaughter at an age of approximately 5½ months. The feedstuff used was produced by DSM Nutritional Products (Copenhagen, Denmark) and DLG (Dansk Landbrugs Grovvareselskab, Copenhagen, Denmark). Detailed information of the content of the feed is given elsewhereReference Maribo, Nielsen and Jakobsen18.
Sampling
The amount fed to each pig was calculated per pen for each of the three periods: 6–7 weeks, 8–12 weeks, 13–23 weeks (weight of 32·5 kg, and until slaughter at approximately 100 kg). For the same periods feed was sampled under the principle of the theory of sampling, and analysed for vitD3 and 25OHD3Reference Gy19.
The day before slaughter a blood sample was drawn, and processed to EDTA–plasma and stored at − 80°C until analysis. At slaughter the liver was sampled and the next day boneless loin with rind was separated. The loin was subsequently carefully dissected into lean meat without any subcutaneous fat (lean meat), subcutaneous fat without any lean meat (fat), and skin without any subcutaneous fat (skin). All the samples were packed in plastic bags and frozen at − 20°C until analysis, which was performed within 8 months. Before analysis the liver, lean meat and fat were slowly thawed and separately ground in a homogenizer (1094 Homogenizer; Tecator, Paris, France) for 2 min, while the skin was slowly thawed and manually cut into pieces of 10–15 mm2.
Vitamin D3 and 25-hydroxyvitamin D3 in the feed
DSM Laboratory (Basel, Switzerland) carried out analyses of vitD3 and 25OHD3 in the feed. For quantification of 25OHD3, 10 g feed was added to 500 ng d6-hydroxyvitamin D3 (synthesized by Prof. Mourino, University of Santiago di Compostela, SpainReference Mascarenas, Perezsestelo, Castedo and Mourino20) as the internal standard, and 60 ml water. The sample was gently swirled into a slurry and sonicated at 50°C for 10 min. The vitamins were extracted with 40 ml tert-butyl methyl ether by shaking and sonication for 5 min followed by centrifuging for 3 min at 3000 rpm. Supernatant (10 ml) was evaporated and the residue dissolved in 2 ml mobile phase for preparative HPLC (2-propanol–ethyl acetate–isooctane, 1:10:89). For clean-up 100 μl were injected into a preparative HPLC system equipped with a silica-column (Si60, 3 μm, 150 × 4·6 mm; Hypersil, Shandon Products, Runcorn, UK). The fraction of 25OHD3 and the internal standard with a retention time of 14–16 min was collected. Subsequently, the organic solvent was evaporated and dissolved in 700 μl methanol and 300 μl water. The quantitative determination was performed by injection of 90 μl into the HPLC–atmospheric pressure chemical ionization–MS equipment (Agilent 1946C LC/MSD single-quadrupole mass specific detector equipped with an atmospheric pressure chemical ionization unit; Agilent Technologies AG, Basel, Switzerland). Additionally, the HPLC system consisted of a C18 column (Aquasil C18 (Aquasil, Thermo Fisher Scientific, Waltham, MA, USA), 3 μm, 2·0 × 100 mm) and the mobile phase was a gradient of methanol–water (99·95 : 0·05).
The quantification of vitD3 in the feed was determined by using vitamin D2 (vit D2) as the internal standard according to EN1282121.
Vitamin D3 and 25-hydroxyvitamin D3 in the meat
The analytical method and the equipment used to determine vitD3 and 25OHD3 in the meat are previously describedReference Jakobsen, Clausen, Leth and Ovesen22. Minor modifications were made as 25-hydroxyvitamin D2 (Sigma-Aldrich, Buchs, Switzerland) was used as the internal standard for 25OHD3 similar to the utilization of vitD2 as the internal standard for vitD3. Briefly, the internal standards of vitD2 and 25-hydroxyvitamin D2 were added to the meat samples and saponified with ethanolic potassium hydroxide. The unsaponifiable matter was extracted with diethyl ether–petroleum ether (1 : 1). The solution was then purified on a silica solid-phase extraction column and further cleaned by preparative HPLC equipped with silica and amino columns. Analysis of the liver samples included an extra preparative HPLC procedure, which consisted of a cyano column (Luna, Cyano, 3 μm, 150 × 4·6 mm) from Phenomenex (Torrance, CA, USA), and a mobile phase of 2-propanol–n-heptane (1·5 : 98·5). The fraction of vitD2 and vitD3 co-eluted with a retention time of 5 min at a flow rate of 1 ml/min. This fraction was collected and evaporated to dryness using a gentle stream of nitrogen, and finally dissolved in methanol–acetonitrile (20 : 80). Another fraction containing 25-hydroxyvitamin D2 with retention time at 16 min as well as 25OHD3 with retention time at 21 min was collected in the same vial, evaporated and dissolved in acetonitrile–water (90 : 10). These two fractions were injected into the analytical HPLC system described earlierReference Jakobsen, Clausen, Leth and Ovesen22.
Content of fat in the meat
Content of fat in the meat was determined by the gravimetric method following a modified Schmid–Bondzynski–Ratslaff method23. Briefly, the sample was boiled with hydrochloric acid followed by the addition of ethanol and extraction of the lipids with diethyl ether–petroleum ether (1 : 1). After evaporation of the solvent, the fat was weighed.
Plasma 25-hydroxyvitamin D
The quantification of 25OHD3 in plasma was performed by the HPLC method described previouslyReference Andersen, Mølgaard and Skovgaard24. Briefly, plasma proteins were precipitated with ethanol and the supernatant was cleaned by a MFC18 solid-phase extraction. The 25OHD3 in the solution was separated, detected and measured by analytical HPLC equipped with a diode array detector (220–320 nm) and a UV detector (265 nm) and external calibration.
Data analysis
Based on previously assessed variation of content of vitD3 and 25OHD3 in meat, six pigs should be included in each feeding group to detect a relative activity of 1·5 for 25OHD3 compared to vitD3 with a power of 80 % and a significance level of 5 %Reference Clausen, Jakobsen, Leth and Ovesen7.
To test the effect of the content of 25OHD3 and vitD3 in feed on 25OHD3 in plasma and on 25OHD3 and vitD3 in meat and liver, regression analysis was performed. In the regression model, 25OHD3 in plasma, meat and liver, and vitD3 in meat and liver were dependent variables, and the total content of vitamin D in feed (vitamin Dtotal) and the difference between 25OHD3 and vitD3 (vitamin Ddiff) were independent variables. Furthermore, ANOVA was performed with feed as an independent fixed variable to test and estimate differences between feeding groups. Association between determinants and variables were assessed with Pearson's correlation coefficients. Data are expressed as means and their standard errors. SAS version 9.1 (SAS Institute, Cary, NC, USA) was used for all statistical analyses, with a significance level of 0·05.
Results
Pigs performance
The vitamin D in each of the three diets was given as the same vitamin D source(s) but due to the different feed and consumption levels during the growth period, the mean daily intake differed. vitD3 and 25OHD3 in each of the three feeding periods are presented in Fig. 1. Carcass weights are presented in Table 1. No significant difference was detected between the diets or between the subgroup and the pigs included in the main feeding study (n 3225).
The mean daily amount of vitamin D (Vitamin Dtotal) in each of the three diet groups (eight pigs in each) and divided into the three different age groups. Vitamin D source in the feed was either 25-hydroxyvitamin D3 (25OHD3, 
) or vitamin D3 (vitD3, □). The diet groups A, B and C are vitD3, a mixture of vitD3 and 25OHD3, and 25OHD3, respectively.
Weight of the carcass and the effect of the diet groups on 25-hydroxyvitamin D3 (25OHD3) in the serum, liver and separated parts of the loin and vitamin D3 (vitD3) in the liver and separated parts of the loin at slaughter* (Mean values with their standard errors)
a,b,cMean values within a row with unlike superscript were significantly different (P < 0·05).
* Diets A, B and C are vitD3, a mixture of vitD3 and 25OHD3, and 25OHD3, respectively.
Effect of 25-hydroxyvitamin D3 and vitamin D3 as vitamin D source
Results for vitamin D status assessed as plasma 25OHD3, and the contents of 25OHD3 and vitD3 in the liver, and in the three separated parts of the loin including the content of fat are presented in Table 1.
The results show that 25OHD3 in the plasma, in the liver and in the three separated parts of loin did not depend on the vitamin D source, but on the daily intake of vitamin Dtotal, as content of 25OHD3 in the separated cuts was significantly associated with vitamin Dtotal (R 0·42–0·56; P < 0·05).
In contrast, vitD3 in the liver and in the three separated parts of loin depended on the vitamin D source (P < 0·001), and showed no association with vitamin Dtotal. However, content of vitD3 was significantly associated with vitD3 in the feed (R 0·65–0·89; P < 0·001).
Discussion
Previous work has shown that vitD3 and 25OHD3 in meat is positively associated with the content of fatReference Clausen, Jakobsen, Leth and Ovesen7. Therefore it is essential that comparison of meat derived from pigs fed different diets does not differ in the content of fat. No significant difference was shown for subcutaneous fat, lean meat and skin. The present results show that the intention to produce three similar separated cuts from each pig was fulfilled regarding the content of fat.
The significant effect of the feeding level of vitamin Dtotal on plasma 25OHD3 in these pigs was in line with the positive association between dietary intake of vitamin Dtotal and serum 25OHD3 shown in women and menReference Heaney, Davies, Chen, Holick and Barger-Lux25, Reference Viljakainen, Palssa, Karkkainen, Jakobsen and Lamberg-Allardt26.
The present study was originally designed to investigate whether the productivity of the pigs was affected when the vitD3 was replaced by 25OHD3 in the feed. No difference was found in the present study which included 3225 pigsReference Maribo, Nielsen and Jakobsen18. As the twenty-four pigs selected for the present nutritional study did not differ from the whole group concerning weight, and growth rate, the mean daily intake of vitamin D calculated from the whole group is applied.
Plasma 25OHD determined as the sum of plasma 25-hydroxyvitamin D2 and plasma 25OHD3 is accepted as the biomarker for vitamin D intake in the absence of sun exposure27. The observed effect that plasma 25OHD achieves a steady state if supplemented at the same level for an adequate time period has previously been used to study differences between natural and synthetic vitD2, vitD3 in fortified bread, juice and supplement, as well as different levels of vitD3 supplementReference Heaney, Davies, Chen, Holick and Barger-Lux25, Reference Outila, Mattila, Piironen and Lamberg-Allardt28–Reference Tangpricha, Koutkia, Rieke, Chen, Perez and Holick30. In human intervention studies supplementation levels of 5–10 μg/d for 4 weeks was shown to be adequate to reach a steady state for vitamin D status (J Jakobsen, unpublished results). The half-life of vitamin D in man is 1 monthReference Clements, Davies and Hayes31. Though the differences in the rate of metabolism between pigs and man are unknown for vitamin D, the applied period of 16 weeks to reach a steady state is assumed to be adequate. As the metabolism in pigs and man regarding fat-soluble vitamins is rather similar, it is assumed that the vitamin D status and vitamin D in the liver and in the meat at slaughter was not influenced by the feed given in the earliest stage of growth up to an age of 12 weeks.
To our knowledge, this is the first study investigating the effect of 25OHD3 and vitD3 in healthy mammals.
Thirty to 40 years ago the difference between 25OHD3 and its parent vitD3 was tested in vitamin D-deficient rats either by testing the effect on intestinal calcium absorption measured by the everted gut sac technique, serum calcium and body weight, or by the ability to cure rickets. By the everted gut sac technique an equal effect of the two compounds was shown after 24 h, though 25OHD3 acted more rapidlyReference Blunt, Tanaka and Deluca13, Reference Martin and Deluca32, Reference Winter, Morava, Simon and Gyüre33. In 1973, 25OHD3 was shown to be five times as active as vitD3 in the maintenance of serum calcium and growthReference Tanaka, Frank and Deluca16. In the ability to cure rickets 25OHD3 had an effect 1·4–2 times the activity of vitD3 in three different studies, but in another study the effect was estimated to be 5 times as activeReference Blunt, Tanaka and Deluca13–Reference Tanaka, Frank and Deluca16.
Today, the factor of 5 for the activity between 25OHD3 and vitD3 is widely used in recommended dietary allowances as well as its implementation in food composition tables10–12. However, the documentation for the factor of 5 seems limited due to the non-standardized methods usedReference Jakobsen, Burckhardt, Dawson-Hughes and Heaney34.
The data obtained for the relative activity between 25OHD3 and vitD3 in pigs need to be verified in a similar study in man, as well as further investigation of the possible difference of the effect of vitamin D derived from pork and from supplements. For mushrooms, no difference was shown between natural vitD2 and vitD2 given as a supplement in the ability to increase vitamin D status, investigated in a human intervention studyReference Outila, Mattila, Piironen and Lamberg-Allardt28. However, the vitamin D activity in meat may not be reflected only by vitD3 and 25OHD3. The content of 1,25-dihydroxyvitamin D3 and other dihydroxyvitamin D3 compounds are unknown, but may contribute to the vitamin D activity of meat.
The effect of 25OHD3 and its parent form vitD3 on plasma 25OHD3 should be regarded equal in the diet for pigs. However, for the nutritional value of pork meat, 25OHD3 in pig feed should be regarded as rather low compared with vitD3, as the content of vitD3 depended on the vitamin D source. The use of 25OHD3 only in the feed instead of vitD3 produced meat and liver with significantly lower content of vitD3. That the pigs fed solely on 25OHD3 did not produce meat and liver with vitD3 is not surprising, as vitD3 is not synthesized in the pigs. The hydroxylation of vitD3 to 25OHD3 by 25-hydroxylase is not a reversible reaction.
Additionally, the present study shows that the concentration of vitD3 as a vitamin D source in the feed determines the concentration of vitD3 in meat and liver even at small differences in the feeding levels, which was previously shown in pigs fed super nutritional levels at 1000 μg vitD3/kgReference Wilborn, Kerth, Owsley, Jones and Frobish35. Additional feeding trials are necessary to investigate fully the possibility of pork meat bio-fortified with vitamin D.
Presently, a human intervention study is being conducted to evaluate whether human subjects respond to supplements of vitD3 and 25OHD3 in a similar fashion to pigs. However, more research on the relative bioactivity of vitamin Dtotal from animal products compared to supplements of vitD3 is an important issue for the calculation of dietary vitD intake.
Conclusion
The findings of the present study showed that 25OHD3 and vitD3 equally affect 25OHD3 in plasma, meat and liver. However, for the benefit of human nutrition, 25OHD3 in pig feed should be regarded as lower than vitD3, as meat and liver produced by feeding the pigs exclusively 25OHD3 had a significantly lower content of vitD3. Regardless of the vitamin D source the present study identified a dose–response effect between vitamin Dtotal in the meat and in the liver with vitamin Dtotal in the feed, which for meat and liver indicate the possibility to produce meat and liver bio-fortified with vitamin D.
Acknowledgements
The authors thank Kirsten Pinndal, Heddi Koch Jacobsen, Astrid Kvindebjerg from the National Food Institute, Technical University of Denmark for technical skills in the analysis of serum and liver, Peter Hoffman, DSM, Basel for the analysis of feed, Ole Lund Svendsen and Gilbert Weber, DSM Nutritional Product for the approval of this nutritional study attached to the original feeding trial, which were financed by DSM Nutritional Product, and Grethe Andersen, Danish Meat Association for providing finances used for the butchering of the pigs. J. J. designed the nutritional part of the feeding trial and was responsible for the analyses of vitamin D, and wrote the manuscript. H. M. designed, planned and conducted the main feeding trial. A. B. was responsible for the analyses of fat. H. M. S. and O. H. were responsible for the statistical test performed. All authors read and contributed to finalization of the manuscript. The sponsors of the study and the authors had no conflict of interests.
