Independent report

Fortifying foods and drinks with vitamin D: main report

Published 23 May 2024

Applies to England

1. Introduction

1.1 Overview

Current government advice on vitamin D relates to protection of musculoskeletal health and is based on recommendations of the Scientific Advisory Committee on Nutrition (SACN), following publication of its 2016 report Vitamin D and health. In its report, SACN recognised that it was difficult to achieve recommended intakes of vitamin D from natural food sources alone and advised the government to consider strategies for the UK population to achieve the recommended intakes of vitamin D.

In spring 2022, the Department of Health and Social Care (DHSC) launched a review on vitamin D intakes to improve health outcomes and help tackle health disparities. The aim of the review was to promote the importance of vitamin D and identify ways to improve vitamin D intake across the population, including the potential options of dietary supplements and fortified foods and drinks.

As part of the DHSC review, SACN was asked to consider any gaps in the evidence. A SACN vitamin D fortification working group was established to provide scientific advice on the potential of mandatory vitamin D food fortification for the UK population to achieve UK dietary recommendations for vitamin D.

1.2 Terms of reference

The terms of reference for the Vitamin D fortification working group were to:

1. Consider the potential impact of mandatory fortification of foods with vitamin D, to include:

• (a) a review of the experiences from countries with existing vitamin D fortification programmes, and their impact on vitamin D intakes and vitamin D status
• (b) a review of the efficacy of different forms of vitamin D for use as a food fortificant
• (c) consideration of factors that could affect the potential efficacy of vitamin D fortification in the UK including habitual diet, life stage, inequalities, ethnicity and ultraviolet B (UVB) sunshine exposure
• (d) consideration and interpretation of models to assess the potential impact of mandatory vitamin D fortification on achievement of UK dietary recommendations and avoidance of intakes above tolerable upper intake levels (ULs)

2. Consider whether there is evidence published since the SACN Vitamin D and Health report (SACN, 2016) that would impact upon the existing recommendations for vitamin D intake and vitamin D status.

The work programme is being undertaken using a phased approach. This review considers points 1(a) and 1(b) of the terms of reference. Other points in part 1 of the terms of reference will be progressed when required by DHSC.

In relation to point 2 of the terms of reference, SACN agreed (at the horizon scan meeting in June 2022) to review the evidence on vitamin D requirements of dark-skinned population groups. The reference nutrient intake (RNI) for vitamin D, set by SACN in 2016 (see section ‘2.6 Current vitamin D recommendations in the UK’ below), was based on evidence from predominantly white-skinned population groups. Data at that time was insufficient to consider whether requirements differed for dark-skinned population groups. Since then, a number of relevant studies have been published. A decision on when to progress this aspect of the work will be made at a future date.

2. Background

2.1 Vitamin D

2.1.1 Vitamin D: sources and forms

Vitamin D is synthesised in the skin following exposure to sunlight containing UVB radiation (at wavelengths between 280 and 315 nanometres). Skin synthesis is the main source of vitamin D for most people.

Vitamin D can also be obtained from foods and dietary supplements. Dietary sources are essential when skin exposure to sunlight containing UVB radiation is limited, for example:

• during the autumn and winter months (from October to late March or early April in the UK)
• when habitually wearing clothing that covers most of the skin outdoors
• when confined indoors during the day

There are 2 major forms of vitamin D:

• vitamin D3 (also referred to as cholecalciferol), which is produced in skin and can also be obtained from the diet
• vitamin D2 (also referred to as ergocalciferol), which can only be obtained from the diet

2.1.2 Terms used in this review to express vitamin D intakes and amounts in fortified products

Vitamin D intake is expressed in micrograms (µg) or international units (IU), with 1µg of vitamin D3 equivalent to 40 IU. In this review, vitamin D intakes and amounts in fortified products are expressed as µg and the corresponding IU.

2.2 Assessment of vitamin D exposure

2.2.1 Terms used in this review to express vitamin D status

Plasma or serum 25-hydroxyvitamin D (25(OH)D) concentration is the principal indicator of total exposure to vitamin D (from skin synthesis and dietary intake) and the main indicator of vitamin D status.

In this review, the terms plasma or serum are used as they are reported in specific studies. Otherwise, for simplicity, the term serum is used in relation to 25(OH)D concentration.

Total serum 25(OH)D concentration comprises the sum of 25(OH)D2 and 25(OH)D3.

Serum 25(OH)D concentration is expressed as nanomoles (nmol) per litre or nanograms (ng) per millilitre (ml), with 2.5nmol per litre equivalent to 1ng per ml. In this review, serum 25(OH)D concentration is expressed as nmol per litre.

2.2.2 Measurement of serum 25(OH)D concentration

A variety of analytical methods are available to determine total serum 25(OH)D concentration but not all can distinguish between 25(OH)D2 and 25(OH)D3 concentrations (SACN, 2016). There is also considerable variation (15 to 20%) in measurements of serum 25(OH)D concentration obtained with different analytical methods (SACN, 2016).

The 2 most common types of methods are:

• antibody-based methods, which use a kit or an automated clinical chemistry platform
• liquid chromatography (LC) based methods with either ultraviolet or mass spectrometric (MS) detection

LC-based methods, including high performance liquid chromatography (HPLC) and liquid chromatography with tandem mass spectrometry (LC-MS/MS), can provide separate estimates of serum 25(OH)D2 and 25(OH)D3 concentrations, and detection of trace concentrations of serum 25(OH)D2. Use of LC-MS/MS methods that are traceable to reference measurement procedures with appropriate external quality assurance credentials such as the Vitamin D External Quality Assessment Scheme are preferred in research laboratories for transparent data reporting of vitamin D metabolite analysis since they have a low limit of detection and quantitation.

2.3 Dietary sources of vitamin D

The main dietary sources of vitamin D in the UK are:

• foods of animal origin such as fish, meat and eggs
• voluntarily fortified foods, such as breakfast cereals
• vitamin supplements

There are few naturally rich food sources of vitamin D. Foods that contain significant amounts (such as oily fish, fish liver oil or egg yolk) or small amounts (such as meat) contain vitamin D3. Vitamin D2 is present in low amounts in the diet.

Vitamin D3 used in supplements or fortified foods is primarily derived from lanolin extracted from sheep wool. Vitamin D2 is derived from fungi and UVB-irradiated yeast. Vegan vitamin D3 can be obtained from lichen.

2.4 Food fortification

Food fortification involves direct addition of micronutrients to food products, during their manufacture or processing, to increase their nutritional value. In the UK, the Bread and Flour Regulations 1998 stipulate mandatory addition of certain nutrients (iron, calcium, thiamin and niacin) to all wheat flour (except wholemeal flour). In September 2021, the UK government and devolved administrations announced their intention to proceed with mandatory fortification of non-wholemeal wheat flour with folic acid, to reduce the risk of neural tube defects during pregnancy.

In the UK, vitamin D fortification of food is voluntary. Foods commonly fortified with vitamin D on a voluntary basis include:

• fat spreads (such as margarines and spreads made from plant oils)
• breakfast cereals

Margarine was previously subject to mandatory fortification with vitamin D and vitamin A. The mandatory requirement was removed in 2013 as part of a government-wide initiative to reduce the number of regulations. This was done on the basis that only a few small producers in the UK made a fat spread that would legally qualify as margarine and almost all fat spreads made in England were fortified on a voluntary basis. For more information, see the consultation outcome for Spreadable fats and milk and milk products .

The vitamin D content of foods can also be increased through a process known as ‘biofortification’. This involves the addition of vitamin D to animal feed to increase the vitamin D content of produce, or exposure of animals, fungi (mushrooms) or yeast to UVB light (Hayes and Cashman, 2017).

Randomised controlled trials (RCTs) have clearly demonstrated that consuming foods fortified with vitamin D increases serum 25(OH)D concentrations (Cashman and others, 2020; Dunlop and others, 2021; Nikooyeh and Neyestani, 2021).

2.5 Biological activity of vitamin D2 and vitamin D3

While both vitamin D2 and vitamin D3 increase total serum 25(OH)D concentrations, researchers continue to debate about their relative efficacy. This is because some studies have reported lower increases in total serum 25(OH)D concentrations following consumption of vitamin D2 compared with vitamin D3 (Tripkovic and others, 2012; Autier and others, 2012; Hammami and Yusuf, 2017; Tripkovic and others, 2017). SACN considered this debate in its 2016 report Vitamin D and health (see section ‘5.1 SACN’s previous assessment’ below).

Vitamins D2 and D3 differ structurally in their side chain. Vitamin D2 contains a double bond between carbon atoms 22 and 23, and an additional methyl group at C24 (Bikle, 2014). The presence of the methyl group may lower the affinity of vitamin D2 for the vitamin D binding protein, resulting in faster clearance of vitamin D2 from the circulation. The methyl group may also alter affinity of vitamin D2 to the main hydroxylases responsible for vitamin D metabolism (Bouillon and others, 2016).

Differences in gene expression have been reported in response to vitamins D2 and D3. A RCT that compared relative efficacy of vitamins D2 and D3 (15µg or 600 IU per day for 12 weeks) in raising serum 25(OH)D concentration (Tripkovic and others, 2017) also compared changes in gene expression in blood samples (using microarrays) from a subset of participants (n=97) (Durrant and others, 2022).

Extensive changes in gene expression were observed between samples taken at baseline and after 12 weeks. Although there was some overlap in the changes in gene expression between the vitamin D2 and D3 groups, most changes were specific to one form or the other of the vitamin. This suggests that vitamins D2 and D3 may have different effects on biological functions (cellular and metabolic).

The authors reported that many of the differentially expressed genes related to immune function with the potential to affect health. They noted, however, that the study was limited by low statistical power and that biological interpretation of the findings should be considered preliminary. A further limitation was that the study did not take account of potential changes in blood cell composition across seasons and across ethnicity.

2.6 Current vitamin D recommendations in the UK

In its 2016 report Vitamin D and health, SACN concluded that the risk of poor musculoskeletal health was increased at serum 25(OH)D concentrations less than (<) 25nmol per litre.

SACN made recommendations on vitamin D intake for the UK population based on protection of musculoskeletal health. The RNI for vitamin D was set at 10µg (400 IU) per day for the UK population aged 4 years and over. This is the amount of vitamin D needed for 97.5% of the population to maintain a serum 25(OH)D concentration equal to or above 25nmol per litre when UVB sunshine exposure is minimal.

The RNI set for the UK population (10µg or 400 IU per day) includes pregnant and lactating women and population groups at increased risk of having serum 25(OH)D concentrations below 25nmol per litre. This includes people:

• with dark skin
• with minimal sunshine exposure due to not spending time outdoors (for example, if they are housebound or living in care homes or other settings with limited outdoor access)
• who cover almost all their skin when outdoors

Data was insufficient to set RNIs for infants and children aged under 4 years. As a precaution, safe intakes were set at 8.5 to 10µg (340 to 400 IU) per day from birth up to 1 year (including exclusively and partially breast-fed infants) and 10µg (400 IU) per day for children aged 1 to 4 years.

Following a rapid review on vitamin D and acute respiratory tract infections (SACN, 2020a) and its subsequent update (SACN, 2020b), SACN concluded that a vitamin D intake of 10µg (400 IU) per day, as currently recommended, may also provide some additional benefit in reducing the risk of acute respiratory tract infections.

The UK government advises that, from late March and early April to the end of September, most people should be able to synthesise all the vitamin D they need from skin exposure to sunlight. In the autumn and winter months (October to late March, early April), when skin synthesis is minimal, everyone is advised to consider taking a daily vitamin D supplement (10µg or 400 IU) since it is difficult to meet the recommendation of 10µg (400 IU) per day from consuming foods containing vitamin D (naturally or fortified).

Population groups at risk of having a serum 25(OH)D concentration < 25nmol per litre (see above) are advised to consider taking a daily vitamin D supplement all year round.

It is recommended that children aged 1 to 4 years receive a daily vitamin D supplement of 10µg (400 IU) and that babies aged under 1 year receive a daily supplement of 8.5 to 10µg (340 to 400 IU). Babies that are formula-fed do not need any additional vitamin D (unless they consume less than 500ml) as infant formula is fortified with vitamin D.

2.7 Potential adverse effects of high exposures to vitamin D

Consuming too much vitamin D can lead to above normal concentrations of calcium in the blood (hypercalcaemia) (SACN, 2016). This can weaken bones and damage the kidneys and heart.

In the 2016 SACN report on Vitamin D and health, data on potential harmful effects of high doses of vitamin D was reviewed by the Committee on Toxicity of Chemicals in Food, Consumer Products and the Environment (COT, 2014). The COT endorsed the ULs set by the European Food Safety Authority (EFSA, 2012), which were:

• 100µg (4,000 IU) per day for adults and children aged 11 to 17 years
• 50µg (2,000 IU) per day for children aged 1 to 10 years
• 25µg (1,000 IU) per day for infants aged up to 1 year

The European Food Safety Authority reviewed the ULs in 2023. The recommendations remained the same for all age groups except infants aged 7 to 11 months (which increased to 35µg or 1,400 IU per day) (EFSA, 2023).

The ULs may not apply to people with medical conditions that predispose them to hypercalcaemia, including:

• normocalcaemic hyperparathyroidism
• granulomatous conditions, such as sarcoidosis and tuberculosis
• genetic predispositions, such as idiopathic infantile hypercalcaemia

2.8 Dietary vitamin D intakes and vitamin D status in the UK

Nationally representative data on current vitamin D intakes and vitamin D status of the general UK population was obtained from the National Diet and Nutrition Survey (NDNS) results from years 2016 to 2017 and 2018 to 2019 (Public Health England, 2020).

2.8.1 Vitamin D intakes

The NDNS shows that mean vitamin D intakes from dietary sources were below the RNI (10µg or 400 IU per day) in all age groups, ranging between:

• 2.2 to 3.2µg (88 to 128 IU) per day from food sources, excluding supplements
• 2.9 to 7.9µg (116 to 316 IU) per day from food sources, including supplements

The proportions of adults who reported taking vitamin D supplements were:

• 17% of adults aged 19 to 64 years
• 34% of adults aged 65 to 74 years
• 28% of adults aged 75 years and over

The low uptake of supplemental vitamin D in the UK suggests that recommendations for vitamin D supplement intake are not reaching the UK population, including groups at higher risk of having serum 25(OH)D concentrations below 25nmol per litre.

2.8.2 Vitamin D status

Vitamin D status varies by season. The NDNS reported that average plasma 25(OH)D concentrations were lowest in all age groups during January to March and more than 20nmol per litre higher during July to September.

The proportions in each age group with plasma 25(OH)D concentrations < 25nmol per litre (taking account of seasonal variation) were:

• 2% of children aged 4 to 10 years
• 19% of children aged 11 to 18 years
• 16% of adults aged 19 to 64 years
• 13% of adults aged 65 years and over

During January to March, the proportions with plasma 25(OH)D concentrations < 25nmol per litre were:

• 19% of children aged 4 to 10 years
• 37% of children aged 11 to 18 years
• 29% of adults

3. Methods

This review comprised 2 parts, which considered:

• experiences from countries with existing mandatory or voluntary vitamin D fortification policies, and their impact on vitamin D intakes and vitamin D status
• evidence published since the 2016 SACN report Vitamin D and health, comparing the relative efficacy of vitamins D2 and D3 in raising and maintaining total serum 25(OH)D concentration

The SACN framework for the evaluation of evidence was used as the basis to assess the evidence. The framework can be viewed on the SACN group page.

3.1 Experiences from countries with existing vitamin D fortification policies

3.1.1 Inclusion criteria

The following types of papers were included:

• peer-reviewed journal articles
• government publications (including legislation documents, and public health and survey data)
• grey literature (because it was recognised as a potential useful source of current information on vitamin D fortification policies and practices)
• national survey data for information on baseline and current vitamin D intakes and status

To assess the impact of fortification on population vitamin D intakes and status, only studies with nationally representative data and English language publications were included.

3.1.2 Literature search

A formal systematic approach was not used for the literature search because it was recognised that relevant information would largely be found in grey literature, which would not be identified using a systematic literature search. It is accepted that this approach meant some sources of information may have been missed.

Members of the working group initially identified publications providing either background information or reviewing national vitamin D fortification programmes. The secretariat then identified further relevant papers from the reference lists of these publications.

Between March and June 2022, the secretariat conducted online searches for information about vitamin D fortification practices. PubMed (an online database) was searched for any studies that had assessed the impact of national vitamin D fortification programmes on vitamin D intakes and vitamin D status at a population level. The following search terms were used: (‘vitamin D’ OR ‘25(OH)D’ or ‘25-hydroxy*) AND (‘food fortification’).

After identifying countries with vitamin D fortification policies, the secretariat:

• conducted web searches for any survey data on vitamin D intakes and status
• reviewed public health websites and policy legislation documents for each country

3.1.3 Study selection

Working group members initially identified 26 papers. Two members of the secretariat assessed the full texts of the 26 papers for relevance. Eleven of the 26 papers met the inclusion criteria. The secretariat subsequently identified 20 further publications (6 journal articles, 9 legislation documents and 5 public health publications) from reference lists, additional literature searching on PubMed, public health webpages and policy legislation documents.

In total, 31 publications met the inclusion criteria.

A quality assessment of the selected literature was not conducted because the eligible papers included diverse types of publications, including grey literature.

3.1.4 Data extraction

The following details were extracted from the eligible papers, where available:

• year that fortification was introduced
• whether fortification was mandatory or voluntary
• the products fortified
• form of vitamin D (D2 or D3) used for fortification
• level of fortification
• uptake of voluntary fortification policies
• whether there was any evaluation of the fortification programme on subsequent vitamin D intakes and status of the population
• whether there was any consideration of vitamin D intakes above upper recommended levels

Any limitations to the fortification policies were also noted, including barriers to the use of fortified foods.

The online translation programme Google Translate was used to translate details of vitamin D fortification policies that were not published in English (Finland, Norway and Sweden).

3.1.5 Countries with vitamin D fortification policies

Vitamin D fortification policies were identified in 6 countries. Policies were:

• mandatory in Australia, Canada and Sweden
• voluntary in Finland, Norway and the USA

The evidence sources for each country are summarised in annex 1.

An overview of the vitamin D fortification policies and their evaluations (where available) for each country is in section ‘4. Experiences from countries with existing vitamin D fortification policies’ below.

Following the original literature search (conducted between March and June 2022), working group members and the SACN secretariat identified 4 additional countries with vitamin D fortification policies (in October 2022). These were:

• Belgium
• Chile
• Ethiopia
• Pakistan

The policies in these countries are described briefly in section 4 below.

A working group member subsequently identified the Global Fortification Data Exchange (GFDx) in November 2023. The GFDx aggregates data on 5 commonly fortified foods (maize flour, oil, rice, salt and wheat flour). It also lists additional countries with mandatory or voluntary vitamin D fortification policies (further details are available on the GFDx website). Policies in these countries were not considered in this review because they were identified at a late stage in the process. However, the GFDx data may be useful for any future considerations. Not all countries submit data to GFDx and Australia, Belgium, Canada and Sweden are not included.

3.2 Relative efficacy of vitamin D2 and vitamin D3

3.2.1 Inclusion criteria

The following types of studies were included:

• systematic reviews and meta-analyses of RCTs that compared the relative efficacy of vitamins D2 and D3 in raising serum 25(OH)D concentration
• those published since the 2016 SACN report on vitamin D and health

3.2.2 Literature search

The SACN secretariat conducted a search of PubMed in June 2022 using the following search terms: (‘vitamin D2 and D3’ OR ‘ergocalciferol and cholecalciferol’) AND (‘25 hydroxyvitamin D’ OR ‘25(OH)D’).

The search identified one systematic review with meta-analysis (Balachandar and others, 2021) which included RCTs published up to 19 June 2021.

A further literature search was then conducted for any relevant RCTs and non-RCTs published since the systematic review with meta-analysis by Balachandar and others (2021). The UK Health Security Agency’s (UKHSA) knowledge and library services conducted a literature search on 11 August 2022 (for studies published between 1 June 2021 and the date of search). The eligibility criteria, adapted from Balachandar and others (2021), are detailed in annex 2.

UKHSA searched the following electronic databases:

• Ovid Medline
• Ovid Embase
• PubMed (top-up search excluding Medline)

UKHSA also searched Scopus, which covers a number of sources, to identify any pre-print articles. The search terms for each database are detailed in annex 3.

3.2.3 Study selection

The literature searches identified 265 records. After de-duplication in EndNote (reference management software), the secretariat screened 219 records in EPPI-Reviewer (a systematic review database), by title and abstract against the eligibility criteria.

Two members of the secretariat independently screened 22 records (10%), identified in the literature search, by title and abstract, against the eligibility criteria. Agreement between both about eligibility was 100%. Both members of the secretariat then independently screened the titles and abstracts of the remaining publications, and subsequently assessed the full texts against the eligibility criteria.

Full texts of 4 potentially relevant publications were assessed by both members of the secretariat. All 4 publications were excluded. The reasons for their exclusion are provided in annex 4.

Only the systematic review with meta-analysis by Balachandar and others (2021) was included for detailed assessment.

3.2.4 Quality assessment

Two members of the secretariat independently assessed the quality of the systematic review by Balachandar and others (2021) using AMSTAR 2 (a measurement tool to assess systematic reviews) (Shea and others, 2017). There was no disagreement between the secretariat members’ assessments. A summary of the AMSTAR 2 assessment is provided in annex 5.

Overall confidence in the findings of the Balachandar and others (2021) systematic review was rated as ‘low’. The main reason for this rating was that the authors did not account for risk of bias in the primary studies in the interpretation and discussion of the results.

4. Experiences from countries with existing vitamin D fortification policies

The purpose of this part of the review was to assess the likely effectiveness of a national vitamin D fortification programme in the UK.

This was done by considering the experiences from countries with existing vitamin D fortification policies, including:

• the food products that are fortified
• levels of fortification
• the form of vitamin D used (D2 or D3)

This review does not consider the dose-response relationship between the amount of a fortified food that would need to be consumed for a specific increase in serum 25(OH)D concentration. Therefore, it does not report data on average population consumption of foods fortified with vitamin D.

Evidence from genetic association studies suggest that common polymorphisms in genes that are involved in vitamin D metabolism may influence the dose-response to vitamins D2 and D3 (Ahn and others, 2010; Wang and others, 2010) and therefore the impact of vitamin D fortification. However, genetic effects are not considered in this review.

4.1 Vitamin D fortification policies

Vitamin D fortification policies were initially identified in 6 countries. Those with mandatory vitamin D fortification policies are considered first (Australia, Canada and Sweden) followed by those with voluntary fortification policies (Finland, Norway and the USA).

The policies are summarised in annex 6. Table A6.1 summarises countries with mandatory vitamin D fortification and table A6.2 summarises countries with voluntary vitamin D fortification. The following details are included in the tables:

• the foods fortified
• fortification level
• form of vitamin D used
• any evaluations of their impact on vitamin D intakes and status

In studies and surveys that reported data on vitamin D intakes and status, the terms referring to ethnicity of population groups are those used by the authors.

Limited details were available on the vitamin D fortification policies subsequently identified in 4 additional countries (Belgium, Chile, Ethiopia and Pakistan). These are briefly described in section ‘4.4 Other countries with vitamin D fortification policies’ below.

4.2 Countries with mandatory fortification policies

4.2.1 Australia

Fortification policy

In Australia, fortification of margarine and edible oil spreads with not less than 55μg (2,200 IU) of vitamin D per kilogram (kg) has been mandatory since 1987 (Food Standards Australia New Zealand; Jayaratne and others, 2013). The form of vitamin D is not specified.

Voluntary vitamin D fortification (0.8µg or 31 IU per 100ml) is permitted for fluid milk products and their alternatives (such as plant-based beverages) (Dunlop and others, 2022).

Impact on vitamin D intakes and status

No trend data or studies were identified that had specifically evaluated the impact of vitamin D fortification on the vitamin D intakes and status of the population.

The Australian Health Survey 2011 to 2012 (Australian Government, 2019) reported that 23% of Australian adults had serum 25(OH)D concentrations below 50nmol per litre, the cut-off for vitamin D deficiency in Australia (in the UK, the cut-off for risk of deficiency is 25nmol per litre). Serum 25(OH)D concentrations below 50nmol per litre were more common among people born in Southern and Central Asia, North-East Asia, South-East Asia, North Africa and the Middle East. The Australian Aboriginal and Torres Strait Islander Health Survey reported that, in 2012 to 2013, 26.5% of adults had serum 25(OH)D concentrations below 50nmol per litre (Australian Government, 2019).

4.2.2 Canada

Fortification policy

In Canada, vitamin D fortification of cow’s milk and margarine has been mandatory since the 1970s. The Food and Drug Regulations were updated in 2022 to approximately double the vitamin D amounts in cow’s milk, goat’s milk and margarine (Government of Canada, 2022b). The regulations now:

• require cow’s milk to contain 2µg (80 IU) per 100ml
• permit goat’s milk to contain 2µg (80 IU) per 100ml
• require margarine to contain 26µg (1040 IU) per 100 grams (g)

Either vitamin D2 or vitamin D3 may be used as the source of vitamin D for food fortification (Government of Canada, 2023).

In 1997, Health Canada published an interim marketing authorisation (IMA) permitting vitamin D fortification (0.85µg or 34 IU per 100ml) of plant-based beverages (Government of Canada, 1997). If fortified, plant-based beverages were required to meet the compositional amounts specified in the IMA. The IMA expired before Health Canada was able to incorporate it into the regulations. However, an interim policy was implemented in 2017 to support continued fortification of plant-based beverages until the necessary regulatory amendments could be made.

In 2022, Health Canada updated the interim policy to allow fortified plant-based beverages to contain the new vitamin D amount for cow’s and goat’s milk (Government of Canada, 2022a). Under the interim policy, manufacturers may add the original vitamin D amount in the expired IMA (0.85µg or 34 IU per 100ml) or the new amount included in the 2022 interim policy (2µg or 80 IU per 100ml).

Impact on vitamin D intake and status

No trend data or studies were identified that had specifically evaluated the impact of vitamin D fortification on vitamin D intakes and status of the population.

An analysis of nationally representative dietary data from the 2007 to 2009 Canadian Health Measures Survey (CHMS) (5,306 participants, age range: 6 to 79 years) reported that, compared with non-consumers, consumption of at least one serving per day of fortified milk was associated with a difference in plasma 25(OH)D concentrations of at least 6nmol per litre (from November to April) (Calvo and Whiting, 2013; Langlois and others, 2010). Less than 0.5% of the population in the CHMS (2007 to 2009) had plasma 25(OH)D concentrations above 220nmol per litre (Langlois and others, 2010).

Average plasma 25(OH)D concentrations were higher for white participants (71.2nmol per litre) compared with those from other racial backgrounds (52.3nmol per litre). The mean difference between groups was approximately 19nmol per litre.

The second cycle of the CHMS (2009 to 2011) (7,830 participants, age range: 3 to 79 years) reported that individuals who consumed milk once or more per day had a higher average plasma 25(OH)D concentration (68nmol per litre) than those who consumed milk less than once per day (59nmol per litre) (Statistics Canada, 2015). Of those individuals who consumed milk once or more per day, 75% had plasma 25(OH)D concentrations above 50nmol per litre compared with 60% of those who consumed milk less than once per day. The survey did not consider groups at risk of vitamin D deficiency (see section 2.6 above).

4.2.3 Sweden

Fortification policy

In Sweden, low-fat milks and solid margarines have been fortified with vitamin D (form not specified) on a mandatory basis since 2007. In 2018, the mandatory fortification policy was expanded (Itkonen and others, 2021; Swedish Food Agency, 2018; Swedish Food Agency, 2007) to include:

• less than 3% fat milk
• sour milk products
• lactose-free products
• vegetable-based alternatives
• liquid margarines

The regulations (Swedish Food Agency, 2018) stipulate that:

• less than 3% fat milk should contain between 0.95 to 1.10μg (38 to 44 IU) of vitamin D per 100g
• less than 3% fermented milk products should contain between 0.75 to 1.10μg (30 to 44 IU) per 100g
• margarine, fat spreads and fluid margarine should contain between 19.5 to 21.0μg (780 to 840 IU) per 100g

Impact on vitamin D intake and status

Nälsén and others (2020) assessed dietary vitamin D intakes and serum or plasma 25(OH)D concentrations in 2 nationally representative cross-sectional studies of:

• school children (206 participants, age range: 10 to 12 years, blood samples collected between March and May 2014)
• adults (1,797 participants, age range: 18 to 80 years, 268 adults provided blood samples, collected between May 2010 and July 2011)

Mean dietary intakes of vitamin D were 7.0μg (280 IU) per day in adults and 5.3μg (212 IU) per day in children, which were below the daily average requirement (AR) of 7.5μg (300 IU) in both adults (33% met AR) and children (16% met AR). In adults, plasma 25(OH)D concentrations below 30nmol per litre were reported in 1% and 9% of participants during the summer and winter periods, respectively. In children, 5% had serum 25(OH)D concentrations below 30nmol per litre from March to May.

No adult participants had vitamin D intakes above the UL (100μg or 4,000 IU per day) from the diet alone; one adult had an intake above the UL when vitamin D intake from supplements was included. One adult had a plasma 25(OH)D concentration over 125nmol per litre. The authors did not consider groups at risk of vitamin D deficiency (see section 2.6 above).

Summerhays and others (2020) examined time trends in serum 25(OH)D concentrations between 1986 and 2014 using data from 7 population-based surveys (the Northern Sweden MONICA Study) of adults (11,129 participants, age range: 25 to 74 years). All surveys were conducted during the same months (between January and April). The authors reported no clear upward or downward trend in serum 25(OH)D concentrations between 1986 and 2014. Mean concentrations (nmol per litre) across 7 time points were:

• 46 in 1986
• 52.8 in 1990
• 55 in 1994
• 49 in 1999
• 42.5 in 2004
• 48.5 in 2009
• 52.8 in 2014

Using the 1986 survey as the reference category, the multivariable-adjusted mean difference (nmol per litre) was:

• 6.75 in 1990
• 8.0 in1994
• 4.0 in 1999
• −5.0 in 2004
• 2.5 in 2009
• 7.75 in 2014

The authors did not consider groups at risk of vitamin D deficiency (see section 2.6).

4.3 Countries with voluntary fortification policies

4.3.1 Finland

Fortification policy

In Finland, a voluntary vitamin D fortification policy was introduced in 2003 recommending addition of vitamin D3 to fluid milk products (0.5μg or 20 IU per 100g) and fat spreads (10μg or 400 IU per 100g) (Tylavsky and others, 2006; Ministry of Trade and Industry of Finland, 2002; Lamberg-Allardt and others, 2003).

In 2010, the recommended fortification amounts for these products were doubled (1μg or 40 IU per 100g for fluid milk products; 20μg or 800 IU per 100g for fat spreads) (Finnish National Nutrition Council, 2010). Although the fortification policy was voluntary, it was adopted widely by the food industry, resulting in mass fortification (Pilz and others, 2018; Lips and others, 2019).

Impact on vitamin D intake and status

Jääskeläinen and others (2017) assessed the impact of the vitamin D fortification policy on vitamin D intakes and vitamin D status in a nationally representative adult population (aged 30 years and above) from the Finnish Health Survey in 2000 (before fortification, 6,134 participants) and 2011 (after fortification, 4,051 participants).

Since serum 25(OH)D concentrations in the 2000 and 2011 surveys were measured using different methods, the results were standardised according to the Centers for Disease Control and Prevention’s Vitamin D Standardization-Certification Program.

In 2011:

• mean daily vitamin D intake per day from diet alone was almost twice as high (men: 14µg or 560 IU; women: 12µg or 480 IU) than in 2000 (men and women: 7µg or 280 IU)
• 74% of men and 58% of women achieved the recommendation for vitamin D intake (10µg or 400 IU per day)
• fish, fluid milk products and fat spreads contributed 38%, 34% and 10% respectively to vitamin D dietary intake for both men and women (the corresponding proportions in 2000 were 57%, 4% and 9%)

The prevalence of serum 25(OH)D concentrations < 30nmol per litre - the cut-off for vitamin D deficiency defined by the US Institute of Medicine (IOM, 2011) (renamed National Academy of Medicine in 2015) - decreased from 12% to < 1%. Mean serum 25(OH)D concentrations increased by 17.8nmol per litre (from 47.6nmol per litre in 2000 to 65.4nmol per litre in 2011).

The authors reported that this increase was mainly explained by fortification, especially of fluid milk products, but observed that vitamin D supplement use also increased (from 11 to 41%) during this period (explained by changes in supplementation recommendations). When analyses were restricted to individuals who did not take supplements, the mean increase in serum 25(OH)D concentration from 2000 to 2011 was 6nmol per litre higher in individuals who consumed fluid milks products compared with non-consumers.

The authors noted that part of the increase in serum 25(OH)D concentration (about 10nmol per litre) that occurred over the period could not be explained either by fortification or by increased use of vitamin D supplements. They suggested that it might be due to a difference in UVB availability between the years (the mean UV radiation index in southern Finland from June to August 2011 was 18% higher than during the same period in 2000). The authors also noted that, although data was adjusted for the month in which blood samples were taken, only 53% were taken in the same season for both surveys (percentage of blood samples taken during the winter was 65% in 2000 and 37% in 2011).

The authors did not consider the prevalence of vitamin D intakes above recommended upper levels (100µg or 4,000 IU per day) but reported that no participants from the 2000 survey sample and only 8 participants from the 2011 survey sample (of whom only one did not use supplements) had serum 25(OH)D concentrations above 125nmol per litre. Groups at risk of vitamin D deficiency (see section 2.6 above) were not considered.

4.3.2 Norway

Fortification policy

In Norway, vitamin D fortification of butter and margarine (10μg or 400 IU per 100g) and some types of low-fat milks (0.4μg or 16 IU per 100g) is on a voluntary basis (Norwegian National Nutrition Council, 2018). The form of vitamin D is not specified.

In 2018, the Norwegian National Nutrition Council recommended changes to the vitamin D fortification policy to reach at-risk groups. It suggested that a wider range of products be fortified with moderate amounts of vitamin D rather than few products fortified with high amounts (Norwegian National Nutrition Council, 2018; Itkonen and others, 2021). It also recommended that fortification should be extended to all fluid milk products and plant-based alternatives, and that fortification of juice, bread and cooking oils should also be considered.

Impact on vitamin D intake and status

No trend data or studies were identified that had specifically evaluated the impact of vitamin D fortification on vitamin D intakes and status of the population.

Oberg and others (2014) assessed data from the Tromsø Study (2010 to 2011), which was conducted in healthy teenagers (890 participants, age range: 15 to 18 years) in northern Norway where UVB radiation is below the limit of skin vitamin D production for approximately 6 months of the year. Data was collected during the school year (not in the summer months).

They reported that 60.2% of participants had serum 25(OH)D concentrations < 50nmol per litre and 16.5% had serum 25(OH)D concentrations < 25nmol per litre (the cut-off for deficiency in Norway). Consumption of vitamin D-fortified milk was associated with higher serum 25(OH)D concentrations in boys only. Groups at risk of vitamin D deficiency (see section 2.6 above) were not considered.

4.3.3 USA

Fortification policy

In the USA, foods that can be fortified with vitamin D (D2 or D3) on a voluntary basis (Calvo and others, 2004) include:

• fluid milks (up to 1.1µg or 44 IU per 100g)
• yogurts (up to 2.23µg or 89.2 IU per 100g)
• margarine (up to 8.23µg or 329.2 IU per 100g)
• breakfast cereals (up to 8.75µg or 350 IU per 100g)

In 2016, the US Food and Drug Administration (FDA, 2022; FDA, 2016) approved:

• an increase in the amount of vitamin D3 that could be added to milk (up to 2.1µg or 84 IU per 100g)
• the addition of vitamin D2 to plant-based beverages intended as milk alternatives (up to 2.1µg or 84 IU per 100g) and plant-based yogurt alternatives (up to 2.23µg or 89 IU per 100g)

Impact on vitamin D intake and status

Fulgoni and others (2011) analysed vitamin D intakes from food sources (both naturally occurring and fortified) in the National Health and Nutrition Examination Survey (NHANES) 2003 to 2006 (16,110 participants; age: 2 years and over). They estimated the total mean vitamin D intake from food sources to be 4.9µg (196 IU) per day. Naturally occurring food sources contributed 40.8% (2µg or 80 IU per day) and foods fortified with vitamin D contributed 59.2% (2.9µg or 116 IU per day) of the total vitamin D intake from food sources. The percentage of the population with vitamin D intakes above the UL (100µg or 4,000 IU) was less than 3%. Groups at risk of vitamin D deficiency (see section 2.6 above) were not considered.

Moore and others (2014) reported that, according to data from NHANES 2007 to 2010 (9,719 participants; age: 19 years and over), fortified milk and milk products made the largest contribution (43.7%) to dietary vitamin D intakes of adults in the USA.

Schleicher and others (2016) used NHANES data to assess temporal trends in serum 25(OH)D concentrations (standardised for different measurement methods) of the USA population (aged 12 years and over) between 1988 and 2010. The authors identified no time trends between 1988 and 2006 (38,700 participants) but found mean serum 25(OH)D concentrations increased (by 5 to 6nmol per litre) between 2007 and 2010 (12,446 participants). They also noted that this coincided with an increase in the use of higher doses of supplemental vitamin D. Between 1988 and 2010, the proportions of the population with serum 25(OH)D concentrations below 40nmol per litre (defined by the IOM as the concentration that meets the needs of half the population) were:

• 14 to 18% overall
• 46 to 60%, non-Hispanic blacks
• 21 to 28%, Mexican Americans
• 6 to 10%, non-Hispanic whites

Calvo and others (2004) reported that vitamin D fortification in the USA and Canada was not effective for increasing vitamin D intakes in all population groups. Data from NHANES (2007 to 2008) indicated serum 25(OH)D concentrations were highest in white Americans, followed by Mexican Americans and lowest in black Americans (Calvo and Whiting, 2013).

Although efficiency of skin vitamin D synthesis may be reduced in population groups with darker skin colour (Clemens and others, 1982), lower consumption of foods fortified with vitamin D may also contribute to the differences in serum 25(OH)D concentrations by ethnicity. Fluid milk and ready-to-eat cereals were the major contributors to vitamin D intakes in the USA, but consumption of these foods differed across various racial groups and was lowest for black American adults. In addition to differing food preferences, another barrier to consumption of fortified foods was the cost of fortified products.

Calvo and Whiting (2013) also reported a large discrepancy between the number of foods eligible for vitamin D fortification, and the number and variety of vitamin D-fortified foods that were available.

4.4 Other countries with vitamin D fortification policies

This section briefly describes vitamin D fortification policies in Belgium, Chile, Ethiopia and Pakistan. Vitamin D fortification policies in these countries were identified after the literature search (between March and June 2022). Details of the policies in these countries were limited.

Chile, Ethiopia and Pakistan implemented mandatory vitamin D fortification policies in 2022, with vitamin D3 being used in all 3 countries. It is too soon to evaluate their impact on vitamin D intakes and status.

4.4.1 Belgium

In Belgium, vitamin D fortification has been mandatory for spreadable fats and margarine (6.5 to 7.5μg or 260 to 300 IU per day) since 1980 and is voluntary for other products (such as milk, milk substitutes, dairy desserts, cereals, biscuits, chocolate powder, and fruit juices) (Moyersoen and others, 2019). The form of vitamin D was not specified.

No trend data or studies were identified that had specifically evaluated the impact of vitamin D fortification on vitamin D intakes and status of the population.

4.4.2 Chile

Chile introduced a policy of vitamin D fortification of liquid milk, milk powder and flour in May 2022 (Chile Ministry of Health, 2022).

Liquid milk is required to be fortified with a minimum of 1μg (40 IU) of vitamin D per 100ml, which may be exceeded by up to 40% (reaching 1.4μg or 65 IU per 100ml).

Milk powder is required to be fortified with a minimum of 10μg (400 IU) of vitamin D per 100g, which may be exceeded by up to 40% (reaching 14μg or 560 IU per 100g).

Flour is required to be fortified with a minimum of 2.2μg (88 IU) per 100g, which may be exceeded by up to 40% (reaching 3.15μg or 140 IU per 100g).

4.4.3 Ethiopia

Ethiopia introduced vitamin D fortification of edible oils in July 2022 (Ethiopian Standards Agency, 2022). Edible oils are required to contain a minimum level of 0.0167μg (0.668 IU) per 100g and a minimum regulatory level of 0.015μg (0.6 IU) per 100g.

4.4.4 Pakistan

Pakistan introduced vitamin D fortification of ghee and edible oil in June 2022 (Provincial Assembly of Khyber Pakhtunkhwa, 2022). These products are required to contain a minimum of 75μg (3,000 IU) per kg and a maximum of 112.5μg (4,500 IU) per kg.

4.5 Limitations of scoping review

The main limitations of the scoping review are summarised below.

The literature search was carried out within a limited timeframe and a formal systematic approach was not used. It is possible, therefore, that some sources of information may have been missed.

Only 3 countries specified the form of vitamin D (D2 or D3) to use for fortification (Canada, Finland and the USA). In the countries that did not specify the form of vitamin D to be used for fortification (Australia, Norway and Sweden), it was assumed that both vitamins D2 and D3 are permitted.

Data varied between countries in terms of definitions of low vitamin D status, the food products that were fortified and fortification levels, making it difficult to compare data between countries.

Evidence was limited on the effectiveness and impact of the vitamin D fortification policies that were identified. Finland is the only country that has evaluated the impact of its vitamin D fortification policy on vitamin D intakes and vitamin D status of the population.

Evidence was limited on the impact of vitamin D fortification on population groups at risk of vitamin D deficiency, including those:

• with dark skin
• with minimal sunshine exposure due to not spending time outdoors (for example, if they are housebound or living in settings with limited outdoor access)
• covering almost all skin when outdoors

Information on vitamin D intakes and vitamin D status was obtained from national survey data. Since dietary surveys rely on self-reported measures of intake, misreporting of food consumption may have affected estimates of vitamin D intakes.

Serum 25(OH)D concentrations are influenced by several factors including those that affect skin synthesis (such as latitude and equivalent UV dose, time of year the blood sample was taken or genetics) and the analytical method used for measurement, which can vary considerably (15 to 20%) (SACN, 2016). The national surveys considered in this review were not designed to assess these factors, so it was not possible to consider how they might have affected serum 25(OH)D concentrations reported in the surveys.

Studies generally did not consider the proportion of the population exceeding upper recommended levels of vitamin D intakes or any potential adverse effects of fortification.

5. Relative efficacy of vitamin D2 and vitamin D3

5.1 SACN’s previous assessment

In its report Vitamin D and health, SACN noted that:

• both vitamins D2 and D3 raise serum 25(OH)D concentration, and both prevent and treat rickets associated with vitamin D deficiency
• results from studies comparing the effectiveness of vitamins D2 and D3 in raising serum 25(OH)D concentrations were inconsistent

SACN considered a systematic review with meta-analysis of RCTs that compared effects of vitamin D2 and vitamin D3 supplementation on serum 25(OH)D concentrations (Tripkovic and others, 2012). Daily doses ranged from 25 to 100µg (1,000 to 4,000 IU) and bolus doses ranged from 1,250 to 7,500µg (50,000 to 300,000 IU). The meta-analysis indicated a significantly greater effect of vitamin D3 compared with vitamin D2 in raising total serum 25(OH)D concentration (mean difference (MD): 15.23nmol per litre; 95% confidence interval (CI): 6.12 to 24.34; p=0.001; 7 RCTs, 344 participants) but the authors noted that heterogeneity between studies was high (I²=81%).

In separate analyses by dosing frequency, there was a significantly larger increase in total serum 25(OH)D concentrations with vitamin D3 compared with vitamin D2 in studies with single bolus doses (MD: 34.10nmol per litre; 95% CI: 16.38 to 51.83; p=0.0002; 3 RCTs, 96 participants) but not in those with daily doses (MD: 4.83nmol per litre; 95% CI: −0.98 to 10.64; p=0.10; 5 RCTs, 248 participants). Heterogeneity was higher in studies using a bolus dose (I²=77%) compared with those administering daily doses (I²=44%).

SACN was unable to draw conclusions on any differences in biological activity between vitamins D2 and D3 from this meta-analysis because of the limitations, which included:

• small number and size of studies
• variability in methodology used to measure serum 25(OH)D concentrations
• differences in dose size and frequency
• differences in duration

SACN also noted that the studies used very high vitamin D doses and that effects may be different at lower doses.

5.2 Evidence published since the SACN vitamin D report (2016)

One systematic review with meta-analysis (Balachandar and others, 2021) was identified by the literature search (hereafter referred to as the ‘Balachandar systematic review’ for simplicity).

5.2.1 Assessment of the Balachandar systematic review

The Balachandar systematic review evaluated the relative efficacy of vitamins D2 and D3 in raising serum concentrations of total 25(OH)D, 25(OH)D2 and 25(OH)D3. The inclusion criteria were randomised and non-randomised controlled studies investigating the relative efficacy of vitamin D2 and vitamin D3 intervention (by either supplementation or food fortification) in apparently healthy participants. Studies of participants with either acute or chronic conditions were excluded.

The authors reported that all studies meeting the inclusion criteria involved random allocation of participants to receive either vitamin D2 or vitamin D3 and that 24 studies had been identified. However, 2 publications (Biancuzzo and others, 2013; Glendenning and others, 2013) included the same data set as 2 earlier studies (Biancuzzo and others, 2010; Glendenning and others, 2009). This means that 24 publications from 22 primary studies were actually included in the systematic review.

The meta-analysis included data from 21 primary studies. The authors reported that 2 studies were not included in the meta-analysis because quantitative data was presented graphically (Hammami and Yusuf, 2017; Thacher and others, 2010). However, the forest plot analysis shows that one of these studies (Thacher and others, 2010) was included in the meta-analysis.

Characteristics of studies included in meta-analysis

The meta-analysis included 1,277 participants (age range: 18 to over 90 years) from 12 countries. Sample sizes of individual studies ranged from 18 to 270 participants and included:

• healthy adults or children (19 studies)
• women residing in a nursing home (1 study)
• hospital inpatients with hip fracture (1 study)

Studies were heterogeneous with respect to:

• frequency and dose of vitamins D2 and D3
• duration of the intervention (in studies with daily or weekly doses)
• follow-up time (in studies that administered a single bolus dose)

Eleven studies administered daily doses only, ranging from 5 to 100µg (200 to 4,000 IU), with intervention durations ranging from 14 days to 1 year.

Five studies administered a single bolus dose, ranging from 1,250 to 15,000µg (50,000 to 600,000 IU), with follow-up time ranging from 4 to 24 weeks. Also, one of these studies administered the vitamin D2 and D3 doses through different routes (intramuscular and orally respectively).

Dosing regimens in the 5 remaining studies were:

• daily (40µg or 1,600 IU) or monthly (1,250µg or 50,000 IU) doses (for 1 year)
• weekly doses (1,250µg or 50,000 IU for 12 weeks)
• twice weekly doses (1,250µg or 50,000 IU for 5 weeks)
• monthly doses (2,500µg or 100,000 IU for 4 months)
• a single dose (2,500µg or 100,000 IU) at the start followed by a daily dose (120µg or 4,800 IU per day from day 7 for 14 days, follow-up time 77 days)

Risk of bias

The risk of bias in the Balachandar systematic review was assessed for the following domains:

• random sequence generation
• allocation concealment
• blinding of participants and personnel
• blinding of outcome assessment
• incomplete outcome data
• selective reporting

The domains were rated as ‘low’, ‘unclear’ or ‘high’ risk as described in the Cochrane Handbook for Systematic Reviews of Interventions (Higgins and others, 2019).

The authors reported that none of the studies were biased by incomplete or selective reporting of outcome. They also reported that the majority of studies were at low risk of bias of random sequence generation and blinding of participants and personnel, but at unclear risk of allocation concealment and blinding of outcome assessment. However, a table detailing the risk of bias assessments for all domains by study did not match the narrative description for selective reporting of outcome. According to the table, risk of bias for this domain was unclear for over half of all studies.

Results of meta-analysis

The meta-analysis was conducted using a random-effects model in anticipation of heterogeneity among the included primary studies.

Total serum 25(OH)D concentration

Effects of vitamin D3 versus D2 supplementation on net changes in total serum 25(OH)D concentrations: vitamin D3 raised total serum 25(OH)D concentrations to a greater extent than vitamin D2 (MD: 15.69nmol per litre; 95% CI: 9.46 to 21.93; p<0.00001; 22 publications, 1,277 participants). However, heterogeneity among the included studies was reported as ‘very high’ (I²=94%). In addition, 2 of the publications contributing to the meta-analysis included the same data set (Glendenning and others, 2013; Glendenning and others, 2009).

Subgroup analysis by frequency of supplementation: heterogeneity and effect size were reduced and 95% CIs were narrower with studies that administered a daily dose (MD: 9.62nmol per litre; 95% CI: 5.82 to 13.43; p<0.00001; I²=67%; 13 publications, 934 participants; 2 publications included the same data set) compared with studies that administered a single bolus dose (MD: 25.06nmol per litre; 95% CI: 3.92 to 46.19; p=0.02; I²=97%; 5 studies, 166 participants) or other dosing regimen (MD: 21.97nmol per litre; 95% CI: 4.39 to 39.56; p=0.01, I²=94%; 5 studies, 177 participants). Heterogeneity and effect size were also lower in studies that administered intervention doses between 5 and 25µg (200 and 1,000 IU) per day (MD: 8.36nmol per litre; 95% CI: 3.70 to 13.03; p=0.0004; I²=66%; 9 publications, 724 participants; 2 publications included the same data set).

Lower heterogeneity was also observed in subgroup analyses of studies that used HPLC (I²=9%) and LC-MS (I²=64%) methods to measure total serum 25(OH)D concentrations compared with those that used radioimmunoassay (RIA) (I²=97%) or another analytical method (I²=98%). Heterogeneity was not reduced in subgroup analyses according to participant age (under 65 versus 65 years or over) or baseline serum 25(OH)D concentrations (< 50 versus 50nmol per litre or above).

Serum 25(OH)D2 and 25(OH)D3 concentrations

Ten studies (679 participants) measured effects of vitamins D2 and D3 supplementation on net changes in serum 25(OH)D2 and 25(OH)D3 concentrations separately. Out of these, vitamin D doses were administered daily in 9 studies (ranging from 5 to 100µg or 200 to 4,000 IU per day) and monthly in one study (2,500µg or 100,000 IU per month).

Both vitamin D2 and vitamin D3 interventions resulted in higher serum concentrations of their respective 25(OH)D metabolites. However, heterogeneity among studies was reported as very high (I² =97% or over). Meta-analysis comparing effects of vitamin D3 versus D2 supplementation on net changes in serum:

• 25(OH)D2 favoured vitamin D2 (MD: −27.5nmol per litre; 95% CI: −34.24 to −20.76; p<0.00001; I²=98%)
• 25(OH)D3 favoured vitamin D3 (MD: 40.85nmol per litre; 95% CI: 31.52 to 50.17; p<0.00001; I²=97%)

The authors reported that subgroup analyses were not possible because too few studies reported serum 25(OH)D2 and 25(OH)D3 concentrations separately.

5.2.2 Limitations of the Balachandar systematic review

The authors acknowledged the limitations of the systematic review and meta-analysis, including the:

• heterogeneity of the primary studies in terms of participants, variable doses and frequency of vitamin D administration
• use of different analytical methods for measurement of serum 25(OH)D concentrations

In addition, there were considerable differences in the duration of studies with daily or weekly interventions (ranging from 14 days to 1 year) and in follow-up time (ranging from 4 to 24 weeks) of studies with a single bolus dose.

The authors reported that only 2 studies (Fisk and others, 2012; Tripkovic and others, 2017) provided clear descriptions of the methods and were considered to be of high quality. They regarded the remaining studies as moderate quality because of incomplete description of methods.

5.2.3 Overall assessment of the Balachandar systematic review

Overall, the Balachandar systematic review suggests that vitamin D3 is more efficacious than vitamin D2 for increasing total serum 25(OH)D concentrations. However, due to the limitations, the pooled estimate of the mean difference between vitamins D3 and D2 (MD: 15.69nmol per litre; 95% CI: 9.46 to 21.93; p<0.00001; I²=94%) may not be robust.

Heterogeneity was higher in studies that administered a single bolus dose (I²=97%) compared with studies that supplemented daily (I²=67%) or studies that administered doses between 5 and 25µg (200 and 1,000 IU) per day (I²=66%). This suggests greater consistency in the daily dosing studies and those administering doses between 5 and 25µg (200 and 1,000 IU) per day. These studies reported lower mean differences (with narrower CIs) between vitamins D3 and D2 in raising total serum 25(OH)D concentration, and suggest an advantage of vitamin D3 over D2 of about 8nmol per litre (95% CI: 4 to 13nmol per litre).

5.3 Consideration of primary studies in the Balachandar systematic review that administered daily doses of vitamin D

The primary studies that administered daily doses of vitamin D are more representative of the likely increases in vitamin D intakes with food fortification. Therefore, the impact of vitamins D2 and D3 on serum concentrations of 25(OH)D, 25(OH)D2 and 25(OH)D3 in these studies was examined further.

In total, there were 12 primary studies (reported in 14 publications) that administered daily doses of vitamins D2 and D3. Out of the 12 studies:

• 8 administered daily doses between 5 and 25µg (200 and 1,000 IU)
• 4 administered doses between 40 and 100µg (1,600 to 4,000 IU)

The vitamin D2 and D3 interventions were provided in various forms including as capsules or added to foods (such as bread or biscuits) or drinks (such as milk or juice).

The studies that administered daily doses between 5 and 25µg (200 and 1,000 IU) are considered in greater detail since these doses are more representative of the increases in vitamin D intakes that might be achieved through food fortification.

5.3.1 Studies with daily vitamin D doses between 5 and 25µg (200 and 1,000 IU)

Details of the 8 studies (reported in 10 publications) are provided in annex 7 (in Table A7.1).

Four out of the 10 publications related to 2 primary studies (Biancuzzo and others, 2010; Biancuzzo and others, 2013; Glendenning and others, 2009; Glendenning and others, 2013). The later publications used the same serum samples as the earlier papers for further analysis. Biancuzzo and others (2013) included data on serum 25(OH)D2 and 25(OH)D3 concentrations, so is included here for consideration. Glendenning and others (2013) used a smaller number of serum samples from the primary data to calculate the ‘free’ bioavailable serum 25(OH)D concentration from published equations. This publication was not included for further assessment since the additional analyses were not considered relevant.

Six of the studies were double-blind controlled trials, one was a single-blind controlled trial, and one was unblinded and uncontrolled.

Out of the 8 studies:

• 1 administered 5µg (200 IU) and 10µg (400 IU) doses of vitamins D2 and D3 (Fisk and others, 2012)
• 1 administered 10µg (400 IU) (Nimitphong and others, 2013)
• 1 administered 15µg (600 IU) (Tripkovic and others, 2017)
• 5 administered 25µg (1,000 IU) (Biancuzzo and others, 2010; Glendenning and others, 2009; Holick and others, 2008; Itkonen and others, 2016; Logan and others, 2013)

Study duration ranged from 4 to 25 weeks. The number of participants ranged between 34 and 270 (age range: 15 to 91 years). Only 4 out of the 8 studies provided information on study power. Seven studies were conducted in healthy free-living populations while participants in one study were hospital inpatients with a hip fracture (Glendenning and others, 2009).

All except 2 studies were conducted in the winter months to avoid the influence of UVB exposure on serum 25(OH)D concentrations. Out of the 2 studies not conducted in winter, one was conducted in New Zealand at the end of summer. This study compared effects of vitamins D3 and D2 on maintaining serum 25(OH)D concentrations over the winter months (Logan and others, 2013). Participants in the other study were hospitalised so would not have been exposed to sunlight (Glendenning and others, 2009).

All studies used LC-MS/MS or HPLC to analyse serum 25(OH)D concentration. Mean baseline total serum 25(OH)D concentrations in the 8 studies ranged between 30 (plus or minus (±) 29) to 80 (± 18) nmol per litre. Mean baseline serum 25(OH)D2 concentrations were < 10nmol per litre in all except one study (Glendenning and others, 2009), which reported concentrations of 12.7 (± 8.9) and 13.3 (± 12.1) nmol per litre in the vitamin D2 and D3 groups respectively. Mean baseline serum 25(OH)D3 concentrations ranged between 24.2 (± 13.4) and 63.1 (± 18.8) nmol per litre.

A comparison of the impact of vitamins D2 and D3 on total serum 25(OH)D, 25(OH)D2 and 25(OH)D3 concentrations (change from baseline) is provided in annex 8 (data extracted from the forest plots in the Balachandar systematic review).

5.3.2 Impact of vitamins D2 and D3 on total serum 25(OH)D concentrations

All 8 studies reported effects of vitamin D2 and D3 on total serum 25(OH)D concentrations. Of these:

• 6 studies reported that serum 25(OH)D concentrations were higher in the vitamin D3 intervention groups compared with the vitamin D2 groups (Fisk and others, 2012; Itkonen and others, 2016; Glendenning and others, 2009; Logan and others, 2013; Nimitphong and others, 2013; Tripkovic and others, 2017)
• 2 studies reported slightly higher serum 25(OH)D concentrations with the D2 intervention (Holick and others, 2008; Biancuzzo and others, 2010)

Mean differences in serum 25(OH)D concentrations between vitamin D3 versus D2 groups ranged from −3.75nmol per litre (95% CI: −14.35 to 6.85) to 18.12nmol per litre (95% CI: 7.43 to 28.81).

Out of these 8 studies:

• 5 reported that the mean differences in total serum 25(OH)D concentrations between the vitamin D2 and D3 interventions were not significant, ranging from −3.75nmol per litre (95% CI: −14.35 to 6.85) to 8.29nmol per litre (95% CI: 3.91 to 12.67) (Biancuzzo and others, 2010; Fisk and others, 2012; Holick and others, 2008; Itkonen and others, 2016; Nimitphong and others, 2013)
• 3 reported that serum 25(OH)D concentrations were significantly higher following vitamin D3 supplementation compared with vitamin D2 supplementation, with mean differences ranging from 14.07nmol per litre (95% CI: 1.33 to 26.81) to 18.12nmol per litre (95% CI: 7.43 to 28.81) (Logan and others, 2013; Glendenning and others, 2009; Tripkovic and others, 2017)

5.3.3 Impact of vitamins D2 and D3 on serum 25(OH)D2 and 25(OH)D3 concentrations

All 8 studies reported effects of vitamins D2 and D3 on serum 25(OH)D2 and 25(OH)D3 concentrations although some data was presented graphically or reported narratively or incompletely.

All studies reported that both vitamin D2 and D3 interventions resulted in greater increases in concentrations of their respective 25(OH)D metabolites. Increases in serum 25(OH)D2 concentrations following the vitamin D2 interventions ranged between 9.2 (± 4.43) and 31.3 (± 0.01) nmol per litre. Increases in serum 25(OH)D3 concentrations following the vitamin D3 interventions ranged between 12.0 (± 11.06) and 34.19 (± 41.18) nmol per litre.

Three studies reported that the vitamin D2 intervention did not affect serum 25(OH)D3 concentrations (Biancuzzo and others, 2013; Fisk and others, 2012; Holick and others, 2008). The remaining 5 studies reported that the vitamin D2 intervention significantly lowered serum 25(OH)D3 concentration with decreases ranging between −4.54 (± 18.1) and −21.7 (± 0.1) nmol per litre (Glendenning and others, 2009; Itkonen and others, 2016; Logan and others, 2013; Nimitphong and others, 2013; Tripkovic and others, 2017). However, 2 of these studies did not include a placebo group so it was not possible to assess the decrease against the expected seasonal decrease in serum 25(OH)D3 concentrations (Nimitphong and others, 2013; Glendenning and others, 2009).

One study included an additional intervention group that received a combination of vitamin D2 (12.5µg or 500 IU) and D3 (12.5µg or 500 IU) (Holick and others, 2008). The increase in total serum 25(OH)D concentration in this group was the same as that reported for the groups that received either 25µg (1,000 IU) of vitamin D2 or D3 separately. The authors also reported comparable increases in serum concentrations of 25(OH)D2 (14.25 ± 11.25nmol per litre) and 25(OH)D3 (15.25 ± 10.75nmol per litre).

Most studies were not able to assess whether vitamin D3 supplementation caused a decrease in serum 25(OH)D2 concentrations because baseline concentrations were below limits of detection. Two studies reported that vitamin D3 had no effect on serum 25(OH)D2 (Fisk and others, 2012; Itkonen and others, 2016). One study reported a decline in serum 25(OH)D2 concentration in the group that received vitamin D3 (in juice) (Tripkovic and others, 2017).

5.4 Studies with daily vitamin D doses above 25µg (1,000 IU)

Details of the 4 studies that administered daily doses above 25 µg (1,000 IU) are provided in annex 7 (in table A7.2). The studies and vitamin D doses were:

• Binkley and others (2011) (40µg or 1,600 IU)
• Lehmann and others (2013) (50µg or 2,000 IU)
• Hartwell and others (1987) (100µg or 4,000 IU)
• Trang and others (1998) (100µg or 4,000 IU)

The number of participants in the studies ranged from 18 to 107 (age range: 19 to 88 years). In one study, participant numbers in the intervention groups were not evenly matched (D2 group, n=17; D3 group, n=55) and all were not randomly assigned (Trang and others, 1998). Only 1 out of the 4 studies reported a power calculation (Lehmann and others, 2013). Study duration ranged from 14 days to one year. Except for the study with a duration of one year, all were conducted in winter.

Three studies used LC-based methods (2 used HPLC, 1 used LC-MS) and 1 used RIA to measure serum 25(OH)D concentrations. Baseline total serum 25(OH)D concentrations ranged from 37.6 (± 13.3) to 80 (± 5.25) nmol per litre.

All 4 studies reported that vitamin D3 was more effective than vitamin D2 in raising total serum 25(OH)D concentrations with mean differences between vitamin D3 versus D2 interventions ranging from 7.8 (95% CI: 2.29, 13.31) to 19.49 (95% CI: 10.47 to 28.51) nmol per litre (see annex 8).

Three of the studies also measured serum 25(OH)D2 and 25(OH)D3 concentrations (Binkley and others, 2011; Lehmann and others, 2013; Hartwell and others, 1987). Baseline serum concentrations of 25(OH)D2 were below the limits of detection of the analytical methods used in all studies. Baseline serum concentrations of 25(OH)D3 were reported in 2 studies and ranged from 36.4 (± 13.3) to 77.5 (± 5.2) nmol per litre. All 3 studies reported that the vitamin D2 intervention decreased serum 25(OH)D3 concentrations. However, 2 of the 3 studies did not include a placebo group for comparison. The study that did include a placebo group (Lehmann and others, 2013) reported that the decrease in serum 25(OH)D3 concentrations observed in the vitamin D2 group was significantly greater than the seasonal decrease observed in the placebo group.

The effect of vitamin D3 on serum 25(OH)D2 concentrations could not be determined because most participants had baseline concentrations below detectable limits.

6. Overall summary and conclusions

In spring 2022, DHSC launched a review to promote the importance of vitamin D and identify ways to improve vitamin D intake across the population, including through fortification of foods and drinks.

As part of this review, SACN was asked to consider any gaps in the evidence and to provide scientific advice on the potential of mandatory vitamin D food fortification for the UK population to achieve UK dietary recommendations for vitamin D.

This review considers the following 2 elements of SACN’s terms of reference (see section 1.2 above):

• experiences from countries with existing vitamin D fortification programmes, and their impact on vitamin D intakes and vitamin D status
• the relative efficacy of vitamin D2 and vitamin D3

6.1 Experiences from countries with existing vitamin D fortification policies

Six countries with detailed information on their vitamin D fortification policies were identified and considered. Policies were mandatory in Australia, Canada and Sweden.

The most common vehicles used for vitamin D fortification were breakfast cereals, fluid milks, yogurts, butter, margarines, fat spreads and edible oils.

Three countries - Canada, Finland and the US - specify the form of vitamin D to be used for fortification. In Canada, either vitamin D2 or vitamin D3 may be used as the source of vitamin D for food fortification; in the US, vitamins D2 and D3 are used to fortify milk, yogurts, margarines and breakfast cereals, and D2 is used to fortify plant-based alternatives. In Finland, vitamin D3 is used as the source of vitamin D for food fortification.

Vitamin D fortification levels in the different products vary between countries. For example: fortification levels in margarine and fat spreads range between 5.5µg (220 IU) per 100g (Australia) and 19.5 to 21μg (780 to 840 IU) per 100g (Sweden); fortification levels in milk range from 0.4μg (16 IU) per 100g (Norway) to 2.1μg (84 IU) per 100g (USA).

Finland is the only country that has assessed the impact of vitamin D fortification on vitamin D intakes and vitamin D status of the population. The findings from this assessment suggest that vitamin D fortification has improved the vitamin D status of the Finnish population. Although voluntary, its implementation has been successful because it was adopted widely by the food industry.

The impact of vitamin D fortification policies in other countries is less clear. Evidence from countries that have collected data on intakes of vitamin D-fortified foods and serum 25(OH)D concentrations suggests that consumers have higher vitamin D intakes and higher vitamin D status compared with non-consumers.

A potential barrier to consumption of fortified foods for some population groups is choice of food for fortification. In the USA, for example, consumption of foods that are major contributors to vitamin D intake (such as fluid milk and ready-to-eat cereals) differed across various population groups and was lowest for black American adults. Another barrier to consumption of fortified foods in the USA was cost.

Evidence was limited on the impact of national fortification policies on population groups at greater risk of vitamin D deficiency, such as those with dark skin and those with minimal sunshine exposure.

In general, studies did not consider the proportion of the population exceeding upper levels of vitamin D intakes, but those that did reported that exceedance was rare. Studies also did not consider any potential adverse effects of vitamin D fortification.

The national surveys considered in this review were not designed to assess other factors that influence serum 25(OH)D concentrations including:

• factors that affect skin synthesis (such as time of year the blood sample was taken)
• the type of assay used for measurement

It was therefore not possible to consider how these factors might have affected serum 25(OH)D concentrations reported in the surveys.

6.2 Relative efficacy of vitamin D2 and vitamin D3

One eligible systematic review with meta-analysis, which evaluated the relative efficacy of vitamins D2 and D3 for raising serum concentrations of total 25(OH)D, 25(OH)D2 and 25(OH)D3 was identified (Balachandar and others, 2021).

Results of the meta-analysis suggest vitamin D3 is more efficacious than vitamin D2 for raising total serum 25(OH)D concentrations (MD: 15.69nmol per litre; 95% CI: 9.46 to 21.93; p<0.00001). However, the primary studies included in the meta-analysis were diverse in terms of vitamin D2 and D3 dosing frequency, amounts and duration. Heterogeneity between studies (I²=94%) suggests that the pooled estimate of the mean difference between vitamins D2 and D3 in raising serum 25(OH)D concentrations may not be robust across different doses and frequency.

Greater consistency (I²=66%) was observed in studies that administered daily doses between 5 and 25µg (200 and 1,000 IU). Pooled estimates from these studies suggest an advantage of vitamin D3 over D2 in raising total serum 25(OH)D concentration of approximately 8nmol per litre (95% CI: 4 to 13nmol per litre).

Studies that administered daily doses between 5 and 25µg (200 and 1,000 IU) of vitamins D2 and D3 are more representative of the vitamin D intakes likely to be achieved through food fortification. Although serum 25(OH)D concentrations were generally higher in the vitamin D3 supplemented groups, both forms were effective in raising total serum 25(OH)D concentrations.

Five out of the 8 studies reported that vitamin D2 supplementation decreased serum 25(OH)D3 concentrations. Although the studies were generally conducted during the winter months when serum 25(OH)D3 concentrations would be expected to decrease, 3 out of the 5 studies reported that the serum 25(OH)D3 concentrations were below those in the placebo group. However, 2 studies did not include a placebo group for comparison against the expected seasonal decrease in serum 25(OH)D3 concentrations.

It was not possible to assess whether vitamin D3 supplementation has any impact on serum 25(OH)D2 concentrations because baseline serum 25(OH)D2 concentrations were below limits of detection in most studies.

6.3 Overall conclusions

In its 2016 report Vitamin D and health, SACN concluded that the risk of poor musculoskeletal health was increased at serum 25(OH)D concentrations below 25nmol per litre. SACN set the RNI for vitamin D at 10µg (400 IU) per day for the UK population. This is the average amount needed by 97.5% of the population to maintain a serum 25(OH)D concentration of 25nmol per litre or above when UVB sunlight exposure is minimal.

Since it is difficult to achieve the RNI from natural food sources alone, SACN advised the government to consider strategies for the UK population to achieve the recommended intakes of vitamin D. Following this advice, the government recommended that everyone should consider taking a daily supplement of vitamin D (10µg or 400 IU) during the autumn and winter months when sunlight exposure is not effective for vitamin D synthesis and serum 25(OH)D concentrations are at their lowest. However, this advice has had limited impact since substantial proportions of the UK population still have poor vitamin D status (serum 25(OH)D concentrations < 25nmol per litre). This suggests that other strategies may be necessary for the UK population to achieve recommended intakes of vitamin D.

One potential public health strategy to increase vitamin D intakes in the UK is vitamin D food fortification. Evidence from countries with existing vitamin D fortification policies suggest overall that an appropriately designed and well-implemented vitamin D food fortification policy has the potential to improve the vitamin D status of the UK population.

To be effective, a vitamin D fortification policy should ensure that most of the population meets recommendations for vitamin D intake, with few or no individuals exceeding upper limits. Such a policy would require determination of suitable levels of vitamin D for fortification, and identification of appropriate categories of foods and drinks that would reach all populations groups in the UK, including those consumed by groups at risk of vitamin D deficiency. To reach diverse population groups, foods fortified with vitamin D would need to be affordable and widely available.

The choice of fortificant (vitamin D2 or D3) would need to take account of different food consumption patterns across the UK (for example, vitamin D3 derived from animal sources would not be a suitable fortificant for those following a vegan diet). Although the evidence suggests vitamin D3 may be more efficacious than vitamin D2, both forms are effective in raising total serum 25(OH)D concentrations and therefore preventing risk of vitamin D deficiency (serum 25(OH)D concentration < 25nmol per litre).

Further consideration of a potential vitamin D fortification policy in the UK would require a modelling exercise to identify suitable fortification vehicles that would reach all population groups in the UK and to assess safe levels of fortification (taking account of current vitamin D levels in fortified foods and supplement use).

The impact of any fortification policy on population serum 25(OH)D concentrations would need to be carefully monitored and evaluated.

7. References

Ahn J, Yu K, Stolzenberg-Solomon R, Simon KC, McCullough ML, Gallicchio L and others. Genome-wide association study of circulating vitamin D levels. Human Molelcular Genetics 2010: volume 19, issue 13, pages 2,739 to 2,745.

Australian Government, 2019. Australian pregnancy care guidelines: vitamin D status. Department of Health and Aged Care.

Autier P, Gandini S and Mullie P. A systematic review: influence of vitamin D supplementation on serum 25-hydroxyvitamin D concentration. Journal of Clinical Endocrinology and Metabolism 2012: volume 97, issue 8, pages 2,606 to 2,613.

Balachandar R, Pullakhandam R, Kulkarni B and Sachdev HS. Relative efficacy of vitamin D(2) and vitamin D(3) in improving vitamin D status: systematic review and meta-analysis. Nutrients 2021: volume 13, issue 10, article number 3,328.

Biancuzzo RM, Clarke N, Reitz RE, Travison TG and Holick MF. Serum concentrations of 1,25-dihydroxyvitamin D2 and 1,25-dihydroxyvitamin D3 in response to vitamin D2 and vitamin D3 supplementation. Journal of Clinical Endocrinology and Metabolism 2013: volume 98, issue 3, pages 973 to 979.

Biancuzzo RM, Young A, Bibuld D, Cai MH, Winter MR, Klein EK and others. Fortification of orange juice with vitamin D(2) or vitamin D(3) is as effective as an oral supplement in maintaining vitamin D status in adults. American Journal of Clinical Nutrition 2010: volume 91, issue 6, pages 1,621 to 1,626.

Bikle DD. Vitamin D metabolism, mechanism of action, and clinical applications. Chemistry and Biology 2014: volume 21, issue 3, pages 319 to 329.

Binkley N, Gemar D, Engelke J, Gangnon R, Ramamurthy R, Krueger D and others. Evaluation of ergocalciferol or cholecalciferol dosing, 1,600 IU daily or 50,000 IU monthly in older adults. Journal of Clinical Endocrinology and Metabolism 2011: volume 96, issue 4, pages 981 to 988.

Bouillon R, Verlinden L and Verstuyf A. Is vitamin D2 really bioequivalent to vitamin D3? Endocrinology 2016: volume 157, issue 9, pages 3,384 to 3,387.

Calvo MS and Whiting SJ. Survey of current vitamin D food fortification practices in the United States and Canada. Journal of Steroid Biochemistry and Molecular Biology 2013: volume 136, pages 211 to 213.

Calvo MS, Whiting SJ and Barton CN. Vitamin D fortification in the United States and Canada: current status and data needs. American Journal of Clinical Nutrition 2004: volume 80, issue 6, pages 1710s to 1716s.

Cashman KD, Kiely ME, Andersen R, Grønborg IM, Madsen KH, Nissen J and others. Individual participant data (IPD)-level meta-analysis of randomised controlled trials with vitamin D-fortified foods to estimate dietary reference values for vitamin D. European Journal of Nutrition 2020: volume 60, issue 2, pages 939 to 959.

Chile Ministry of Health, 2022. Modified Supreme Decree No. 977 of 1996, Sanitary Regulation.

Clemens TL, Adams JS, Henderson SL and Holick MF. Increased skin pigment reduces the capacity of skin to synthesise vitamin D3. The Lancet 1982: volume 319, issue 8263, pages 74 to 76.

Committee on Toxicity of Chemicals in Food, Consumer Products and the Environment (COT), 2014. Statement on adverse effects of high levels of vitamin D (accessed January 2024).

Dunlop E, James AP, Cunningham J, Rangan A, Daly A, Kiely M and others. Vitamin D fortification of milk would increase vitamin D intakes in the Australian population, but a more comprehensive strategy is required. Foods 2022: volume 11, issue 9, article number 1,369.

Dunlop E, Kiely ME, James AP, Singh T, Pham NM and Black LJ. Vitamin D food fortification and biofortification increases serum 25-hydroxyvitamin D concentrations in adults and children: an updated and extended systematic review and meta-analysis of randomized controlled trials. Journal of Nutrition 2021: volume 151, issue 9, pages 2,622 to 2,635.

Durrant LR, Bucca G, Hesketh A, Möller-Levet C, Tripkovic L, Wu H and others. Vitamins D(2) and D(3) have overlapping but different effects on the human immune system revealed through analysis of the blood transcriptome. Frontiers in Immunology 2022: volume 13, article number 790,444.

European Food Safety Authority (EFSA), 2012. Scientific opinion on the tolerable upper intake level of vitamin D.. EFSA Panel on Dietetic Products, Nutrition and Allergies (accessed January 2024).

EFSA) Scientific opinion on the tolerable upper intake level for vitamin D, including the derivation of a conversion factor for calcidiol monohydrate. EFSA Journal 2023: volume 21, issue 8, article number e08145.

Ethiopian Standards Agency, 2022. Fortified edible oils - with vitamin A and D - specification.

Food and Drug Administration (FDA), 2016. Food additives permitted for direct addition to food for human consumption; vitamin D2. Federal Register, pages 46,578 to 46,582.

FDA, 2022. Code of Federal Regulations - Title 21, Chapter 1, Subchapter B: Food for Human Consumption.

Finnish National Nutrition Council, 2010. Report of vitamin D working group (Valtion ravitsemusneuvottelukunta. D-vitamiinityöryhmän raportti) (in Finnish).

Fisk CM, Theobald HE and Sanders TA. Fortified malted milk drinks containing low-dose ergocalciferol and cholecalciferol do not differ in their capacity to raise serum 25-hydroxyvitamin D concentrations in healthy men and women not exposed to UV-B. Journal of Nutrition 2012: volume 142, issue 7, pages 1,286 to 1,290.

Food Standards Australia New Zealand. Food Standards Code legislation - Chapter 2: food standards. Australian Government Federal Register of Legislation.

Fulgoni VL, 3rd, Keast DR, Bailey RL and Dwyer J. Foods, fortificants, and supplements: where do Americans get their nutrients? Journal of Nutrition 2011: volume 141, issue 10, pages 1,847 to 1,854.

Glendenning P, Chew GT, Inderjeeth CA, Taranto M and Fraser WD. Calculated free and bioavailable vitamin D metabolite concentrations in vitamin D-deficient hip fracture patients after supplementation with cholecalciferol and ergocalciferol. Bone 2013: volume 56, issue 2, pages 271 to 275.

Glendenning P, Chew GT, Seymour HM, Gillett MJ, Goldswain PR, Inderjeeth CA and others. Serum 25-hydroxyvitamin D levels in vitamin D-insufficient hip fracture patients after supplementation with ergocalciferol and cholecalciferol. Bone 2009: volume 45, issue 5, pages 870 to 875.

Government of Canada, 1997. Interim marketing authorization to permit the optional addition of vitamin and mineral nutrients to plant-based beverages.

Government of Canada, 2022a. Interim policy on the use of expired interim marketing authorizations related to food fortification.

Government of Canada, 2022b. Regulations amending the Food and Drug Regulations (nutrition symbols, other labelling provisions, vitamin D and hydrogenated fats or oils): SOR/2022-168. Canada Gazette: part 2, volume 156, number 15.

Government of Canada, 2023. Food and Drug Regulations (Consolidated Regulations of Canada, c.870) - Part D.003 - Addition of vitamins, mineral nutrients or amino acids to foods. Document number: D.01.003 (1) (b).

Hammami MM and Yusuf A. Differential effects of vitamin D2 and D3 supplements on 25-hydroxyvitamin D level are dose, sex, and time dependent: a randomized controlled trial. BMC Endocrine Disorders 2017: volume 17, issue 1, article number 12.

Hartwell D, Hassager C and Christiansen C. Effect of vitamin D2 and vitamin D3 on the serum concentrations of 1,25(OH)2D2, and 1,25(OH)2D3 in normal subjects. Acta Endocrinologica (Copenhagen) 1987: volume 115, issue 3, pages 378 to 384.

Hayes A and Cashman KD. Food-based solutions for vitamin D deficiency: putting policy into practice and the key role for research. Proceedings of the Nutrition Society 2017: volume 76, issue 1, pages 54 to 63.

Higgins J, Thomas J, Chandler J, Cumpston M, Li T, Page M and others. Cochrane Handbook for Systematic Reviews of Interventions. John Wiley and Sons, 2019: second edition. Chichester, UK.

Holick MF, Biancuzzo RM, Chen TC, Klein EK, Young A, Bibuld D and others. Vitamin D2 is as effective as vitamin D3 in maintaining circulating concentrations of 25-hydroxyvitamin D. Journal of Clinical Endocrinology and Metabolism 2008: volume 93, issue 3, pages 677 to 681.

Institute of Medicine, 2011. Dietary reference intakes for calcium and vitamin D. Ross AC, Taylor CL, Yaktine AL and others, editors. Washington (DC): National Academies Press (US), National Academy of Sciences.

Itkonen ST, Andersen R, Björk AK, Brugård Konde Å, Eneroth H, Erkkola M and others. Vitamin D status and current policies to achieve adequate vitamin D intake in the Nordic countries. Scandinavian Journal of Public Health 2021: volume 49, issue 6, pages 616 to 627.

Itkonen ST, Skaffari E, Saaristo P, Saarnio EM, Erkkola M, Jakobsen J and others. Effects of vitamin D2-fortified bread v. supplementation with vitamin D2 or D3 on serum 25-hydroxyvitamin D metabolites: an 8-week randomised-controlled trial in young adult Finnish women. British Journal of Nutrition 2016: volume 115, issue 7, pages 1,232 to 1,239.

Jääskeläinen T, Itkonen ST, Lundqvist A, Erkkola M, Koskela T, Lakkala K and others. The positive impact of general vitamin D food fortification policy on vitamin D status in a representative adult Finnish population: evidence from an 11-y follow-up based on standardized 25-hydroxyvitamin D data. American Journal of Clinical Nutrition 2017: volume 105, issue 6, pages 1,512 to 1,520.

Jayaratne N, Hughes MC, Ibiebele TI, van den Akker S and van der Pols JC. Vitamin D intake in Australian adults and the modeled effects of milk and breakfast cereal fortification. Nutrition 2013: volume 29, issue 7 to 8, pages 1,048 to 1,053.

Lamberg-Allardt C, Kärkkäinen M, Outila T and Natri A-M. Suomalaisten D-vitamiinitilanteen kohentaminen—auringosta, ruoasta vai purkista? [Improving vitamin D status in Finland—sun, food or supplements?]. Suomen lääkärilehti 2003: volume 58, pages 1,055 to 1,056.

Langlois K, Greene-Finestone L, Little J, Hidiroglou N and Whiting S. Vitamin D status of Canadians as measured in the 2007 to 2009 Canadian Health Measures Survey. Health Reports 2010: volume 21, issue 1, pages 47 to 55.

Lehmann U, Hirche F, Stangl GI, Hinz K, Westphal S and Dierkes J. Bioavailability of vitamin D(2) and D(3) in healthy volunteers, a randomized placebo-controlled trial. Journal of Clinical Endocrinology and Metabolism 2013: volume 98, issue 11, pages 4,339 to 4,345.

Lips P, Cashman KD, Lamberg-Allardt C, Bischoff-Ferrari HA, Obermayer-Pietsch B, Bianchi ML and others. Current vitamin D status in European and Middle East countries and strategies to prevent vitamin D deficiency: a position statement of the European Calcified Tissue Society. European Journal of Endocrinology 2019: volume 180, issue 4, pages p23 to p54.

Logan VF, Gray AR, Peddie MC, Harper MJ and Houghton LA. Long-term vitamin D3 supplementation is more effective than vitamin D2 in maintaining serum 25-hydroxyvitamin D status over the winter months. British Journal of Nutrition 2013: volume 109, issue 6, pages 1,082 to 1,088.

Ministry of Trade and Industry of Finland, 2002. Regulation 917/2002 about adding of vitamins and some other substances to foodstuffs (in Finnish).

Moore CE, Radcliffe JD and Liu Y. Vitamin D intakes of adults differ by income, gender and race/ethnicity in the USA, 2007 to 2010. Public Health Nutrition 2014: volume 17, issue 4, pages 756 to 763.

Moyersoen I, Devleesschauwer B, Dekkers A, Verkaik-Kloosterman J, De Ridder K, Vandevijvere S and others. A novel approach to optimize vitamin D intake in Belgium through fortification based on representative food consumption data. Journal of Nutrition 2019: volume 149, issue 10, pages 1,852 to 1,862.

Nälsén C, Becker W, Pearson M, Ridefelt P, Lindroos AK, Kotova N and others. Vitamin D status in children and adults in Sweden: dietary intake and 25-hydroxyvitamin D concentrations in children aged 10-12 years and adults aged 18-80 years. Journal of Nutritional Science 2020: volume 9, article number e47.

Nikooyeh B and Neyestani TR. The effects of vitamin D-fortified foods on circulating 25(OH)D concentrations in adults: a systematic review and meta-analysis. British Journal of Nutrition 2021: volume 127, issue 12, pages 1 to 18.

Nimitphong H, Saetung S, Chanprasertyotin S, Chailurkit LO and Ongphiphadhanakul B. Changes in circulating 25-hydroxyvitamin D according to vitamin D binding protein genotypes after vitamin D3 or D2 supplementation. Journal of Nutrition 2013: volume 12, article number 39.

Norwegian National Nutrition Council, 2018. Vitamin D i Norge: Behov for tiltak for å sikre god vitamin D-status? [Vitamin D in Norway: need for measures for to ensure good vitamin D status?] (PDF, 1.4MB).

Oberg J, Jorde R, Almås B, Emaus N and Grimnes G. Vitamin D deficiency and lifestyle risk factors in a Norwegian adolescent population. Scandinavian Journal of Public Health 2014: volume 42, issue 7, pages 593 to 602.

Pilz S, März W, Cashman KD, Kiely ME, Whiting SJ, Holick MF and others. Rationale and plan for vitamin D food fortification: a review and guidance paper. Front Endocrinol (Lausanne) 2018: volume 9, article number 373.

Provincial Assembly of Khyber Pakhtunkhwa, 2022. The Khyber Pakhtunkhwa Food Fortification Bill.

Public Health England, 2020. National Diet and Nutrition Survey rolling programme: results from years 9 to 11 (2016 to 2017 and 2018 to 2019).

SACN, 2016. Vitamin D and health report.

SACN, 2020a. Rapid review: vitamin D and acute respiratory tract infections.

SACN, 2020b. Update of rapid review: Vitamin D and acute respiratory tract infections.

SACN, 2023. Framework and methods for the evaluation of evidence that relates food and nutrients to health.

Schleicher RL, Sternberg MR, Lacher DA, Sempos CT, Looker AC, Durazo-Arvizu RA and others. The vitamin D status of the US population from 1988 to 2010 using standardized serum concentrations of 25-hydroxyvitamin D shows recent modest increases. American Journal of Clinical Nutrition 2016: volume 104, issue 2, pages 454 to 461.

Shea BJ, Reeves BC, Wells G, Thuku M, Hamel C, Moran J and others. AMSTAR 2: a critical appraisal tool for systematic reviews that include randomised or non-randomised studies of healthcare interventions, or both. BMJ 2017: volume 358, article number j4008.

Statistics Canada, 2015. Vitamin D blood levels of Canadians.

Summerhays E, Eliasson M, Lundqvist R, Söderberg S, Zeller T and Oskarsson V. Time trends of vitamin D concentrations in northern Sweden between 1986 and 2014: a population-based cross-sectional study. European Journal of Nutrition 2020: volume 59, issue 7, pages 3,037 to 3,044.

Swedish Food Agency, 2007. Regulations on changes in the Swedish Food Agency’s regulations (SLVFS 1983:2) on fortification of certain foods (PDF, 45KB, in Swedish).

Swedish Food Agency, 2018. The Swedish Food Agency’s regulations on fortification of certain foods (LIVSFS 2018:5) (PDF, 58.2KB, in Swedish).

Thacher TD, Fischer PR, Obadofin MO, Levine MA, Singh RJ and Pettifor JM. Comparison of metabolism of vitamins D2 and D3 in children with nutritional rickets. Journal of Bone and Mineral Research 2010: volume 25, issue 9, pages 1,988 to 1,995.

Trang HM, Cole DE, Rubin LA, Pierratos A, Siu S and Vieth R. Evidence that vitamin D3 increases serum 25-hydroxyvitamin D more efficiently than does vitamin D2. American Journal of Clinical Nutrition 1998: volume 68, issue 4, pages 854 to 858.

Tripkovic L, Lambert H, Hart K, Smith CP, Bucca G, Penson S and others. Comparison of vitamin D2 and vitamin D3 supplementation in raising serum 25-hydroxyvitamin D status: a systematic review and meta-analysis. American Journal of Clinical Nutrition 2012: volume 95, issue 6, pages 1,357 to 1,364.

Tripkovic L, Wilson LR, Hart K, Johnsen S, de Lusignan S, Smith CP and others. Daily supplementation with 15 μg vitamin D(2) compared with vitamin D(3) to increase wintertime 25-hydroxyvitamin D status in healthy South Asian and white European women: a 12-wk randomized, placebo-controlled food-fortification trial. American Journal of Clinical Nutrition 2017: volume 106, issue 2, pages 481 to 490.

Tylavsky FA, Cheng S, Lyytikäinen A, Viljakainen H and Lamberg-Allardt C. Strategies to improve vitamin D status in northern European children: exploring the merits of vitamin D fortification and supplementation. Journal of Nutrition 2006: volume 136, issue 4, pages 1,130 to 1,134.

Wang TJ, Zhang F, Richards JB, Kestenbaum B, van Meurs JB, Berry D and others. Common genetic determinants of vitamin D insufficiency: a genome-wide association study. The Lancet 2010: volume 376, issue 9736, pages 180 to 188.