This study investigates the influence of different food matrices on the bioavailability of vitamin D.

Although most vitamin D comes from skin synthesis in response to sun exposure, dietary intake is also important - especially during winter time where there is no endogenous production of vitamin D in Denmark. A way to maintain an adequate vitamin D status is to supplement either as tablets/droplets or as fortified food. However, there seems to be an inter-individual variation in response to supplementation.

This study aims to investigate whether this variation in absorption of vitamin D may depend on delivery system.

BACKGROUND The current project is a part of the vitamin D fortification with enhanced bioavailability study program (acronym: DFORT) which is an interdisciplinary project including research groups from Denmark, Spain, and the Netherlands supported by the Danish Innovation Foundation. The overall aim of DFORT is to develop more efficient strategies for vitamin D fortification by studying the influence of the delivery matrix on the bioavailability of vitamin D. DFORT is organized into four scientific work packages (WP).

The first two WPs have aimed to study whether complex formation (nano-encapsulation) of vitamin D with different proteins may enhance the stability of vitamin D (WP 1 lead by prof. Daniel Otzen, AU-iNANO) and the effect of complex formation in real food systems including investigations on the stability during storage, light- and heat-exposure (WP 2 lead by associate professor Trine Kastrup Dalsgaard, AU-FOOD). WP 1+2 have shown that vitamin D can be stabilized by complex formation with whey protein and that the encapsulation may cause less oxidative degradation thereby improving the stability of vitamin D in different food systems.

In the current study (WP 3 lead by prof. Lars Rejnmark, AU-Health), the bioavailability of vitamin D in different food matrices (including complex formation with whey protein) will be studied in humans. Biological samples will be collected in WP 3 allowing for metabolomics studies on possible associations between vitamin D supplementation through different food matrices and metabolic phenotype (WP 4 lead by prof. Hanne C. Bertram, AU-FOOD).

Although most of the total body vitamin D is synthesized in the skin after exposure to UV light (wavelength of 290-315 nm), most individuals require at least some dietary vitamin D to maintain a replete vitamin D status. This is especially true during wintertime. With a latitude of 56°N in Denmark, there is no endogenous synthesis of vitamin D in the months extending from October to April, which means that the inhabitants have to rely on food sources in order to maintain a replete vitamin D status. Cholecalciferol (vitamin D3 [D3]) is the main dietary source of vitamin D, but it is only present in a limited number of food items (such as fatty fish) making it difficult to achieve the recommended intake of 10 µg D3 per day.

Vitamin D status may be improved in response to an increased intake of vitamin D in terms of either supplementation with tablets or food fortification. Numerous studies have shown increased 25-hydroxy vitamin D (25OHD) levels in response to an increased intake of vitamin D. It is generally assumed that mean 25OHD concentrations increase by 0.7 nmol/L in response to an increased long-term intake of 1 µg vitamin D per day although the relative increase per microgram supplemented may be higher if baseline levels are low. Despite this well-known dose-response relationship in groups of people, several studies have documented that the change in serum 25OHD levels in response to vitamin D supplementation varies widely.

Several reasons may account for the inter-individual variation in response to vitamin D supplementation. In gross terms, the variation may be due to dosing inaccuracies (inconsistencies between claimed and actual values of vitamin D) and variation in bioavailability of vitamin D.

Inconsistencies between claimed and measured values of vitamin D content in vitamin D tablets and food-fortified products may be due to inconsistencies in dose used for fortification or to the instability of the vitamin per se. Discrepant results have been reported on the stability of vitamin D in different food matrices and when exposed to different physiochemical hazards. Some investigators have reported vitamin D to be unstable whereas others have found it to be remarkable stable when exposed to oxidation, light, and to acid and alkali.

Only few studies have searched for factors responsible for the inter-individual variation in 25OHD levels in response to vitamin D supplementation. These studies have suggested that body composition (including fat mass content), genetic variants of the vitamin D binding protein (VDBP), and the ratio of serum 24,25-dihydroxy vitamin D (24,25(OH)2D) to 25OHD may contribute to variation in serum 25OHD levels. However, in a recent study only 47% of the variations in the response to vitamin D supplementation could be explained by accounting for factors of known importance to changes in 25OHD levels.

In addition to the above-mentioned indices, factors of importance to the intestinal absorption of vitamin D as well as the food matrix by which vitamin D supplementation is provided may contribute to inter-individual variations in 25OHD responses. However, only few studies are available on the bioavailability of vitamin D from different food matrices and the intestinal absorption of vitamin D, including the intraluminal fate, and molecular mechanisms facilitating the absorption are still only partially understood.

As vitamin D is a fat-soluble molecule, it has generally been assumed that vitamin D is absorbed in the small intestine by simple passive diffusion with vitamin D being incorporated into the micelle and transported by chylomicrons via lymph veins to the liver. This is in alignment with studies showing an increased risk of low 25OHD levels in patients with fat malabsorption. Accordingly, it has been suggested that ingestion of vitamin D with a meal rich in fat may increase the release of bile, allowing an increased incorporation of vitamin D in the bile salt micelle thereby improving the bioavailability of vitamin D. However, discrepant results have been reported, on whether the composition of the food matrices (and its fat content) by which vitamin D is ingested influence its bioavailability.

In a randomized, controlled trial by Raimundo et al., the mean change in 25OHD levels two weeks after the treatment with a single large oral dose of 50,000 IU D3 was larger, when the meal had at least 15 g of fat compared to a fat-free meal. In contrast, the fat content of the food matrices was not found to influence the time-concentration profile as measured by vitamin D2 levels in plasma 2, 4, 8, 12, 48, and 72 h after ingestion of a single dose of 25,000 IU D2 added to either whole milk, skim milk or dissolved in 0.1 mL corn oil and applied to toast. However, both of these studies are limited by the use of very high (pharmacological) doses of vitamin D, which may override any physiological effects of the composition of the food matrices.

A lack of an effect of the fat content of the food by which vitamin D is ingested is also supported by studies on vitamin D fortification of orange juice. Comparing the bioavailability of vitamin D added to orange juice or supplemented as capsules showed a similar increase in 25OHD concentrations in response to 11 weeks of supplementation with 1000 IU vitamin D per day and the increase was significant compared to placebo. The fact that vitamin D may be sufficiently absorbed following a fat-free meal (such as orange juice) may be explained by recent findings on the mechanism by which vitamin D is absorbed. It seems that vitamin D is not only absorbed by simple passive diffusion (by incorporation into the micelle), as cholesterol membrane transporters, such as SR-BI, CD36, or NPC1L1, have been shown to be involved in the absorption. Differences in expression levels and the existence of functional polymorphisms in the genes encoding these proteins may also contribute to the large inter-individual variation in postprandial responses to vitamin D.

Only very few studies are available on the time-plasma concentration profile of vitamin D after intake of an oral dose. Denker et al. studied the pharmacokinetic profile of vitamin D3 after administration of a single D3 dose of either 2800 or 5600 IU, showing that plasma D3 levels increased steadily after the intake and peaked at 9±2.3 h with concentrations returning to near baseline values by 72 h. It is unknown whether the food matrix (including complex formation of vitamin D by encapsulation with whey proteins) affects the bioavailability of vitamin D as assessed by the plasma-time concentration profiles and whether this may influence the inter-individual variability in response to vitamin D supplementation.

The importance of calcium intake, and especially calcium intake from milk products and tablets (supplements) has been investigated in a number of studies, showing discrepant results. A Cochrane meta-analysis has suggested an overall beneficial effect of increased calcium intake from milk products and calcium supplements. However, a recent trial has suggested an increase in blood pressure in the hours following intake of 1000 mg of calcium citrate compared with placebo. It has so far not been investigated whether milk intake causes similar effects on indices of cardiovascular health, including blood pressure and arterial stiffness.

AIM The overall aim of the study is to investigate the influence of different food matrices (including complex-formation with whey proteins) on the bioavailability of vitamin D, as assessed by maximum concentration profiles (Cmax) and the time-concentration curve of D3 in plasma and thereby whether the inter-individual variation in the absorption of vitamin D may depend on delivery system.

Co-primary (null-) hypothesis:

• The food matrix by which D3 is delivered does not affect Cmax of D3 as determined 10h post-dosing.
• The absorption profile (time-concentration curve in terms of Area Under the Curve from 0h to 12h [AUC0-12h]) does not differ according to the food matrix by which D3 is delivered.

Secondary (null-)hypotheses

• Compared with vitamin D provided as droplets, the absorption of D3 is not enhanced by delivery through each of the tested food matrices (i.e., increased Cmax).
• Compared with vitamin D added to juice, the absorption of D3 is not enhanced by whey protein complex-bound D3 (i.e., increased Cmax).
• Treatments do not affect plasma levels of parathyroid hormone (PTH) and ionized calcium.
• The variability to vitamin D supplementation in terms of Cmax is lower if vitamin D is complex-bound to whey proteins as compared to the other tested supplementation methods.
• Arterial stiffness as assessed by tonometry is not affected by milk intake.

Explanatory hypotheses In order to allow for further investigations on indices of importance to responses to vitamin D supplementation, data will be collected on body composition, genetic polymorphisms, cholesterol status, and habitual dietary habits.

MATERIALS AND METHODS

STUDY DESIGN The study is performed as a multiple cross-over study using a balanced latin-square design. This design allows for each participant to function as her own control thereby counterbalancing risk of an adverse influence on results of the order of treatment or other factors such as effect of period, as well as inter-individual variations attributable to e.g., genetic variations, body weight etc. By randomization, each participant will be allocated to receive all the five treatment regimes in a pre-specified order with a 10-21 days wash-out period in-between each of the treatment arms.

The treatment sequences are:

Treatment sequence 1: A B E C D

Treatment sequence 2: B C A D E

Treatment sequence 3: C D B E A

Treatment sequence 4: D E C A B

Treatment sequence 5: E A D B C

Treatment sequence 6: D C E B A

Treatment sequence 7: E D A C B

Treatment sequence 8: A E B D C

Treatment sequence 9: B A C E D

Treatment sequence 10: C B D A E

PROCEDURES FOR HANDLING VITAMIN D SUPPLEMENTATION The supplement will be acquired commercially and stored at the Osteoporosis Clinic, Aarhus University Hospital and kept away from other medication and supplementation. Sub-investigator is responsible for correct handling and dispensing of vitamin D supplement, as well as securing that the supplement will only be used as described in the protocol and that the participants are instructed to take it correct.

PROCEDURES FOR RANDOMIZATION Randomization will be done using a computer generate list. Treatments will not be blinded for the investigator. In terms of comparing juice with or without whey proteins bound-complexes, a single-blind design will be applied, as participants will not be told which of the treatments they are receiving. Each treatment sequence will be allocated to the same number of patients - e.g. 3 participants will be in treatment sequence 1, 3 in treatment sequence 2 etc.

POPULATION Thirty participants will be recruited from the general background population by direct mailing using a list of randomly selected individuals living in the area of Aarhus generated by "Research services" at Statens Serum Institut. The study will be performed during wintertime (November-April).

WITHDRAWAL AND DROPOUT Any participant can at any point drop out of the study without any explanation and will not have to go through a final examination. The investigator can withdraw a participant if this seems necessary for the participant's safety. Dropouts and withdrawals will be noted and explained in the CRF.

Withdrawal will happen in case of one of the following criteria is fulfilled:

• Change in vitamin D supplementation
• Ionized calcium ≥1.40 mmol/L
• Disease or new medication that will influence the study
• Serious adverse effects/symptoms that is expected to be caused by vitamin D supplementation Diseases that occur within 7 days of treatments can be a possible cause of participation in the study. Sub-investigator can in this time frame be contacted in order to investigate whether it is a cause of the vitamin D supplementation. In case it is, the symptoms or disease will be followed until it is cured or have become chronic.

EXAMINATIONS Participants will be examined 5 times over a time period of 6 to 12 weeks. At each visit, the participants will arrive fasting before 9am and will stay at the department until blood sampling at 12 hours is taken. Hereafter, the participants is free to go home and come back the following day for the 24 hours blood sampling and delivering the urine samples or stay the night at the hospital. During the 12 hours at the department, the participants will get standardized food.

Basic health information and questionnaires:

Participants will answer questionnaires regarding their general health as well as dietary habits and sun exposure.

Biochemistry:

Blood samples will be collected at different time points (0, 2, 4, 6, 8, 10, 12, and 24 hours).

All measurements will be performed when all material from all 30 participants have been collected in order to avoid variation in results. Blood samples will be stored in a biobank for a maximum of 15 years after the end of the study.

Urine samples:

Urine will be collected in 3 batches at the first day of each dosing i.e., from 0-4 hours, from 4-8 hours and from 8-24 hours.

All measurements will be performed when all material from all 30 participants have been collected.

Bone scans:

Dual-Energy X-ray absorptiometry (DXA) and High-Resolution peripheral Quantitative Computed Tomography (HRpQCT):

DXA scanning with the Hologic QDR Discovery scanner. Bone mineral density (BMD) will be measured in lumbar spine (L1-L4), femoral neck, and the distal forearm. Furthermore, total body composition will be determined, including fat- and lean-tissue mass.

A HRpQCT bone scan of the distal radius and tibia will be performed using an Xtreme CT-scanner (SCANCO Medical AG, Switzerland). This will allow for assessment of volumetric BMD for cortical and trabecular bone, bone structure and geometry (including cortical and trabecular thickness, trabecular separation etc.) and bone strength.

Blood pressure measurements and tonometry:

Blood pressure and measurements of arterial stiffness (tonometry) are performed twice in each participant in relation to treatment regimes "C" and "D".

On both occasions, measurements are performed in the morning with the participant in the fasting state. After the measurements are performed, the participant will be provided the intervention together with a breakfast meal. After this, the participant will be fasting until next measurement is performed four hours later.

Office blood pressure (BP) is measured in a sitting position after 5 minutes of rest on the right upper arm using a digital automatic BP monitor. Three BP readings will be performed with 2 minutes of rest in-between. The average of the last two measurements is recorded.

Arterial stiffness and pulse wave velocity (PWV) will be assessed by tonometry using the SphygmoCor system (Xcel; AtCor Medical, Sydney, NSW, Australia). For measurements of carotid-to-femoral PWV, an inflated femoral cuff placed on the right upper thigh combined with carotid applanation tonometry will be used. Measurements are performed in a quiet room. The participant will be resting for 10 minutes in a supine position prior to test start. Brachial BP is measured on the right upper arm and two consecutive BP readings are performed. If BP readings do not differ by > 5 mmHg, the last one is recorded. If BP readings differ by > 5 mmHg, four BP readings are obtained. The average of the last two measurements is recorded. AIx is assessed as the ratio of wave reflection amplitude to central pulse pressure. The mean of two measurements are used in the analyses. Carotid-femoral PWV is assessed as the distance travelled divided by the transit time using the direct carotid-to-cuff distance as measured with a non-stretchable tape (infantometer). A minimum of two measurements is performed. If measurements differs < 0.5 m/s the average of the two measurements is used for analyses. If PWV differs by > 0.5 m/s a third measurement is obtained and the median value is used for analyses. According to general recommendations, the direct carotid-to-cuff distance mean PWV is multiple with 0.8.

PERSPECTIVES The study will provide insight into the bioavailability of vitamin D3 supplementation, including sources of variation. Since Denmark is a country with low latitude and high prevalence of vitamin D insufficiency and fortification of food items is not common or legislated, this study may lead to way to fortifying food items in Denmark.

Inclusion Criteria:

• Postmenopausal
• Caucasian
• Total plasma 25-hydroxy vitamin D < 50 nmol/L
• Understand oral and written Danish
• Able to consent

Exclusion Criteria:

• Known allergic reaction/intolerance to Vitamin D supplementation / milk products / juice
• Known chronic kidney disease (creatinine > 90 µmol/L), previous kidney transplantation or known kidney artery stenosis
• Known liver disease
• Known gastrointestinal malabsorption
• Current malignant disease
• Hypercalcemia (ionised calcium ≥ 1.33 mmol/L)
• Treatment with diuretics, lithium or current use of steroids
• Current use of calcium and/or vitamin D supplementation
• Planned travel during the intervention period to areas where sun exposure is expected
• Use of solarium
• Treatment with beta-blockers
• Overt cardiovascular disease such as known severe heart failure (NYHA III-IV), previous major heart surgery, pacemaker, arrhythmias (e.g. atrial fibrillations or flutter, second- and third-degree atrioventricular block)

Participants will not know if they receive juice with complex-bound vitamin D3 or not.