Safety of vitamin D₃ addition to feedingstuffs for fish

Abstract

During 2012–2014, the EFSA Panel on Additives and Products or Substances used in Animal Feed (FEEDAP) issued three opinions on the safety and efficacy of vitamin D₃ for all animal species and concluded that no safety concern was identified for the use of vitamin D₃ for fish at the maximum authorised content of 0.075 mg/kg feed. The Norwegian Food Safety Authority made available to the Commission some studies on the safety of vitamin D₃ for fish and consumers at substantially higher levels (1.5 mg/kg feed) than those proposed by EFSA. The European Commission asked EFSA to review the information provided to estimate if it would be possible to increase the current levels of vitamin D₃ in feed for fish. The increasing use of plant‐based feed materials in aquaculture feeds could induce a decrease in vitamin D₃ content in feedingstuffs. However, there is no evidence that the current total (background + supplemented) maximum EU content of vitamin D₃ may cause any appreciable risk of deficiency in salmonids. The possible contribution of vitamin D₂ in plant‐based ingredients to the total vitamin D intake is considered to be low, although it cannot be reliably estimated. The FEEDAP Panel concludes that a total level of 1.5 mg vitamin D₃/kg compound feed is safe for salmonids with a margin of safety of at least 10. For other fish, insufficient data are available to conclude on the safety of a total level of 1.5 mg vitamin D₃/kg feed. Although the assessment of safety for the consumer is impaired by uncertainties concerning the transfer of vitamin D₃ from feed to fish flesh, it was concluded that an increase of total vitamin D content in fish feeds up to 1.5 mg/kg feed would not lead the tolerable upper intake level to be exceeded even in high consumers.

Summary

During 2012, 2013 and 2014, the Panel on Additives and Products or Substances used in Animal Feed (FEEDAP) issued three opinions on the safety and efficacy of vitamin D₃ (cholecalciferol) as an additive to feed and water for drinking for all animal species. The Panel concluded that the current maximum authorised contents were temporarily acceptable for the safety of target animals and considered a complete review of the most recent literature necessary to maintain or revise these maximum contents. Particularly, no safety concern was identified for the use of vitamin D₃ for fish, for which the maximum authorised content is 3,000 IU/kg feed (corresponding to 0.075 mg/kg feed). The FEEDAP Panel also concluded that the use of vitamin D in animal nutrition at the currently authorised maximum dietary content has not and will not cause the tolerable upper intake level for consumers to be exceeded.

The Standing Committee on Plants, Animals, Food and Feed discussed and agreed to support the revision of the maximum content of vitamin D₃ in feed for farmed fish. The Norwegian Food Safety Authority (NFSA) made available to the European Commission (EC) some studies on the safety of vitamin D₃ for fish and consumers at substantially higher levels (1.5 mg/kg feed corresponding to 60,000 IU/kg feed) than those proposed by EFSA. The EC asked EFSA to review the information provided by the NFSA and to take into account any other relevant information to estimate if it would be possible to increase the current levels of vitamin D₃ in feed for fish and the impact it may have on the safety for the consumer.

It was considered possible that the increasing use of plant‐based feed materials in aquaculture feeds could induce an appreciable decrease in vitamin D₃ content in compound feedingstuffs. However, the available evidence evaluated by the FEEDAP Panel does not indicate that an appreciable decrease in total vitamin D₃ has occurred in the EU. Furthermore, the estimated levels of vitamin D₃ in the novel fish feed formulations (with fish meal/oil replaced by plant protein sources) and the current upper legal level of vitamin D₃ in the EU do not indicate any risk of vitamin D₃ deficiency in farmed fish. The possible contribution of vitamin D₂ in plant‐based ingredients to the total vitamin D intake is likely to be low but it cannot be reliably estimated to date.

The FEEDAP Panel reiterates its previous conclusion that the currently authorised level for vitamin D₃ is sufficient to meet the requirements for salmonids. Therefore, it does not see a need to increase the current maximum authorised content based on considerations of animal nutrition.

The FEEDAP Panel concludes that a total level of 1.5 mg vitamin D₃/kg compound feed is safe for salmonids with a margin of safety of at least 10. For other fish, insufficient data are available to conclude on the safety of a total level of 1.5 mg vitamin D₃/kg feed.

The assessment of safety for the consumer is impaired by uncertainties concerning the transfer of vitamin D₃ from feed to fish flesh, including the influence of factors such as fish species, the background content and the supplementation level of vitamin D₃, and farming conditions. Even taking into account the identified uncertainties, it was concluded that an increase in total vitamin D content in fish feeds up to 1.5 mg/kg feed would not lead the upper level to be exceeded even in high consumers.

1. Introduction

1.1. Background and Terms of Reference as provided by the requestor

Regulation (EC) No 1831/20031 establishes rules governing the Community authorisation of additives for use in animal nutrition and, in particular, Article 9 thereof defines the terms of such authorisation by the Commission.

The applications for authorisation of the additive pursuant Article 10(2) and 4 (use in water for drinking for terrestrial animals) were submitted by Lohmann Animal Health GmbH, DSM and Fermenta Biotech Ltd. Three opinions were issued by the FEEDAP Panel on 13 November 2012, 20 June 2013 and 30 January 2014, respectively.2

All applications requested the additive to be classified in the category ‘nutritional additives’ (Table 1).

Table 1.

Description of the additive as referred to in the applications for authorisation

Category of additive	Nutritional additive
Functional group of additive	Vitamins, provitamins and chemically well‐defined substances having a similar effect
Description	DSM – Cholecalciferol + precholecalciferol (67.2%) Fermenta Biotech Ltd – Vitamin D3 40 MIU/g powder and Vitamin D3 resin 22 MIU/g Lohmann Animal Health GmbH – Vitamin D3 84%
Target animal category	All species and categories. Fish 3,000 IU maximum content (equivalent to 0.075 mg/kg)
Applicant	DSM, Lohmann Animal Health GmbH and Fermenta Biotech Ltd
Type of request	Generic opinion

During the discussions at the Standing Committee on Plants, Animals, Food and Feed it was agreed to request the Authority to revise the proposed levels (3,000 IU −0.075 mg/kg) for fish feed, taking into account the current feeding practices that replace the ingredients from animal origin rich in vitamin D by ingredients from crop origin with negligible content of this vitamin.

The proposed levels seemed to be insufficient to satisfy the dietary requirements for fish.

The Commission has now received supplementary information concerning the safety of vitamin D, containing, in particular, the following items:

• Changes in the composition of feeds: animal protein/versus crop protein

• Real contents of vitamin D in feed for fish

• Correlation between vitamin D intake and deposition in edible parts

• Consequences of a low upper limit in fish production

• Tolerance of fish to higher levels of vitamin D

• Effects of an increase in vitamin D in fish on consumer health

The supplementary information has been submitted by the Norwegian Food Safety Authority (NFSA). The NFSA concluded that levels of 1.5 mg/kg vitamin D₃ supplementation are safe for fish and for the consumer.

The Commission asks EFSA to review the information provided by the NFSA and to take into account any other relevant information. The opinion should estimate if it is possible to increase the current levels of vitamin D₃ in feed for fish.

1.2. Interpretation of the Terms of Reference

The Commission asked EFSA to estimate if it is possible to increase the current levels of vitamin D₃ in feedingstuffs for fish up to 1.5 mg/kg complete feed.

The supplementary information provided by the NFSA and the scope of their assessment were mainly referring to safety for target animals and consumer: as a consequence, the FEEDAP Panel considers that the present assessment should be limited to the assessment of the safety for target animals and consumers. In addition, since the dataset did not include any new information concerning the safety for the user and the environment, this opinion will not address the potential effects of the proposed increase in the maximum authorised levels of vitamin D₃ on the safety for the user and the environment.

1.3. Additional information

In 2012, 2013 and 2014, the Panel on Additives and Products or Substances used in Animal Feed (FEEDAP) issued three opinions on the safety and efficacy of vitamin D₃ (cholecalciferol) as an additive to feed and water for drinking for all animal species (EFSA FEEDAP Panel, 2012a, 2014, 2015). The FEEDAP Panel noted that whereas for many terrestrial species (turkeys for fattening, equines, bovines, ovine and pigs) the current maximum authorised contents of vitamin D₃ in feeds do not provide any margin of safety, no safety concern was identified for the use of vitamin D₃ in fish. Additionally, notwithstanding the long history of supplementing compound feed with vitamin D and the absence of publicly reported intolerances, the FEEDAP Panel could not draw final conclusions on the safety of vitamin D in animal nutrition based on the NRC data collection, which has not been revised for 25 years. The FEEDAP Panel considered the current maximum authorised contents to be temporarily acceptable for the target animals. Nutritional surveys in 14 European countries showed that vitamin D intake by consumers is sufficiently below the upper level (UL) (EFSA NDA Panel, 2012). The FEEDAP Panel assumes that foodstuffs of animal origin monitored in these studies were produced following current production practices, including vitamin D₃ supplementation of feed. Therefore, the FEEDAP Panel concluded that the use of vitamin D in animal nutrition at the currently authorised maximum dietary content has not and will not cause the UL to be exceeded.

Vitamin D in the form of vitamin D₃ (E 671) is included in the European Union Register of Feed Additives. Maximum limits have been established for inclusion in complete feeds of different target species3: in fish the current maximum limit is 3,000 IU/kg equal to 0.075 mg/kg.4 The simultaneous use of vitamin D₂ (ergocalciferol) and vitamin D₃ is prohibited for all species.

Vitamin D in the form of cholecalciferol (D₃) and ergocalciferol (D₂) is authorised for use in food,5 and in food supplements,6 for addition for specific nutritional purposes in foods for particular nutritional uses,7 to processed cereal‐based foods for infants and young children8 and to infant formulae and follow‐on formulae when reconstituted as instructed by the manufacturer.9 Vitamin D is also listed as a pharmacologically active substance in veterinary medicinal products and is not subject to maximum residue levels when used in medicinal products for food‐producing animals.10

Cholecalciferol is described in the European Pharmacopoeia (PhEur), Monograph (MG) 0072 (European Pharmacopoeia, 2010a). Cholecalciferol concentrated, oily form, powder form and water‐dispersible forms are described in MG 0575, MG 0574 and MG 0598, respectively (European Pharmacopoeia, 2010b,c,d).

One milligram of vitamin D₃ is equal to 40,000 international units (IU). In the present opinion, the amounts of vitamin D in feed are expressed as mg/kg feed, and in food as μg/kg food.

2. Data and methodologies

2.1. Data

The present assessment is based on data submitted by the NFSA11 in support of the review of the use of vitamin D₃ as a feed additive, particularly to estimate if it is possible to increase its current levels in feed for fish.

The FEEDAP Panel used the data provided by the NFSA together with data from other sources, including previous risk assessments by EFSA or other expert bodies, peer‐reviewed scientific papers, other scientific reports, experts' knowledge and information received through a call for data from EU Member states and EFSA Focal Point network,12 to deliver the present output.

Following a call for data on 29 April 2016, a total of seven European countries submitted data consisting on scientific papers and grey literature (technical reports, outputs from databases, non‐published data).13 A summary of the data received is available in Appendix A.

2.2. Methodologies

The approach followed by the FEEDAP Panel to review the safety of vitamin D₃ is in line with the principles laid down in Regulation (EC) No 429/200814 and the relevant guidance documents: Guidance on nutritional additives (EFSA FEEDAP Panel, 2012a), Guidance for establishing the safety of additives for the consumer (EFSA FEEDAP Panel, 2012b).

3. Assessment

The two major natural sources of vitamin D are cholecalciferol (vitamin D₃, which occurs in animals) and ergocalciferol (vitamin D₂, which occurs predominantly in plants). Vitamin D₃ is metabolised to the active steroid hormone 1,25‐dihydroxyvitamin D₃ (1,25(OH)D₃) by successive hydroxylation in the liver and kidney (in fish both hydroxylation steps occur in the liver only). Vitamin D₂ is metabolised to 1,25‐dihydroxyvitamin D₂ by the same enzyme systems. Vitamin D₃ is also produced by endogenous synthesis in mammalian species and birds; it is also recognised that trout may synthetize vitamin D₃ in the skin when exposed to blue light (Pierens and Fraser, 2015). 7‐Dehydrocholesterol in the skin is converted by exposure to ultraviolet light and then enzymatically into vitamin D₃. Vitamin D₂ is formed by photochemical reaction from ergosterol in plants, fungi and lower life forms. Moreover, a few plants (e.g. Solanum glaucophyllum and Trisetum flavescens) are known to produce vitamin D₃‐related metabolites (EFSA FEEDAP Panel, 2015). Both forms of the vitamin can also be produced by chemical synthesis. Details of the vitamin D‐metabolism in fish have been reviewed by Lock et al. (2010).

The principal physiological role of vitamin D in all vertebrates (including teleost fish, reviewed by Lock et al., 2010) is in calcium and phosphorus homeostasis. Vitamin D is a key regulator of transcellular calcium uptake. Furthermore, vitamin D modulates the expression of Na⁺‐dependent inorganic phosphate (Pi) transporters (i.e. in the intestine and the kidneys). Bone represents the largest store of calcium phosphate, and vitamin D directly affects osteoblast activity and osteoclast formation. The vitamin D‐regulated metabolic pathways also play important roles in other biological processes not related to calcium and phosphorus homeostasis, i.e. in muscle function, immunological regulation and cardiovascular physiology. Vitamin D has been shown to affect cell proliferation and differentiation. Like other steroids and transcription factors, vitamin D is a specific ligand of a nuclear receptor, the vitamin D receptor (VDR) (EFSA NDA Panel, 2012, EFSA FEEDAP Panel 2012a, EFSA NDA Panel 2016 and references herein). However, a full vitamin D endocrine function, characterised by a specific VDR, specific vitamin D metabolizing CYP450 enzymes regulated by calciotropic hormones and a dedicated plasma transport protein, is only found in terrestrial vertebrates. The morphology of fish bone is very different from that of terrestrial vertebrates and the calcium reservoir of bone in fish does not represent a major factor in calcium homeostasis even in situations of calcium stress. The evolutionary biology of vitamin D in vertebrates is reviewed by Bouillon and Suda (2014).

Excess of vitamin D disrupts the Ca‐P homeostasis with consequences, e.g. on bone health in vertebrates. In humans, an upper tolerable intake level (UL) of 100 and 50 μg/day for adults and children, respectively, has been assessed (EFSA NDA Panel, 2012).

Vitamin D₃ (IUPAC name: (3β,5Z,7E)‐9,10‐secocholesta‐5,7,10(19)‐trien‐3‐ol; synonyms: 9,10‐ secocholesta‐5,7,10(19)‐trien‐3‐ol, cholecalciferol, calciol) is identified with the Chemical Abstracts Service (CAS) number 67‐97‐0 and the European Inventory of Existing Chemical Substances (EINECS) number 200‐673‐2. The molecular weight is 384.64 g/mol. The structural formula of cholecalciferol is shown in Figure 1.

Figure 1.

Structural formula of cholecalciferol

The FEEDAP Panel adopted three scientific opinions on different vitamin D₃ products (EFSA FEEDAP Panel 2012a, 2013, 2014). The Panel concluded that the current maximum authorised contents were temporarily acceptable for the safety of all animal species but recommended a complete review of the most recent literature. The currently maximum authorised level of vitamin D in feed for fish is 0.075 mg/kg feed (corresponding to 3,000 IU/kg feed). In the meantime, the FEEDAP Panel did not foresee any risk of animal deficiency at the current authorised levels in the EU, noting that data on requirements and allowances are easily available.

The aim of the present assessment is to evaluate the consequences for fish health and consumer safety of levels of vitamin D₃ substantially higher than the one currently authorised, as proposed by Norway (maximum total level of 1.5 mg vitamin D₃/kg feed). To this purpose, data on the tolerance of salmon to vitamin D₃ supplementation and on the correlation between vitamin D₃ in feed and in edible fish products were submitted by the NFSA.

3.1. Safety

3.1.1. Safety for the target species

The information submitted by the NFSA suggests that the increasing use of plant‐based feed materials could induce a decrease in vitamin D₃ content in fish feed. Consequently, the NFSA proposed to increase the total level of vitamin D₃ up to a maximum of 1.5 mg/kg complete fish feed.

The following sections aim (i) to assess the potential impact of the change in the composition of feed on fish nutrition and (ii) to establish whether an increase in the current levels of supplemental vitamin D₃ in feed for fish up to 1.5 mg/kg feed is safe for fish.

3.1.1.1. Changes in the composition of feeds for salmonids and the impact on fish nutrition

The composition of fish feed has changed considerably in the last 20 years as a result of a continuing trend to lower the content of fish‐derived feed materials, especially for Atlantic salmon. Previously, feed for Atlantic salmon comprised mainly fish‐derived materials, about 45% fish meal, 30% fish oil, plus wheat, vitamins and minerals; currently, the Atlantic salmon feed contains only about 20% of marine ingredients (10% fish meal and 10% fish oil).15 The rapidly growing farming industry and the limited global marine resources to be used for fish feed have increased the use of alternative feed ingredients, mainly of plant origin; moreover, the shift toward plant‐based ingredients can effectively reduce the burden of bioaccumulating contaminants in farmed fish, such as methylmercury (fish meals) and dioxins and polychlorinated biphenyls (PCB; fish oil) (reviewed by Mantovani et al., 2015). A 2011 report of Norden Nordic Innovation (an inter‐government foundation to boost innovation and competitiveness in the Nordic Countries) lists a number of main plant‐based ingredients of the new aquaculture feeds: soy (meal, protein concentrate, cake), peas and pea protein, horse beans, grains (wheat, corn gluten) as well as rapeseed oil (Norden Nordic Innovation, 2011).

Fish meals and oils can provide a good, albeit highly variable, source of vitamin D₃. According to the NFSA, fish meal and fish oil show highly variable content of vitamin D₃: according to the available data (Table 2), fish meal may contain from 0.01 to 0.18 mg/kg and fish oil from 0.03 to 3.9 mg/kg, depending on the species and the season.16 Therefore, based on these figures, in traditional complete feeds for farmed fish, fish meal (45%) and fish oil (30%) are estimated to provide vitamin D₃ in the range of 0.01–1.25 mg/kg. Nowadays, with the new feed formulation containing more vegetable based ingredients, vitamin D₃ provided by fish meal (10%) and fish oil (10%) is estimated to range from 0.004 to 0.408 mg/kg complete feed.17

Table 2.

Available data on vitamin D₃ content in fish meal and fish oil

Fish meal	Fish oil
0.01–0.18 mg/kg (Horvli and Lie, 1994)	0.2–3.9 mg/kg (Opstvedt et al., 1997)
0.02–0.15 mg/kg (Biomar database)a	0.03–2.4 mg/kg (Biomar database)

^aIn‐house database cited by the applicant in the technical dossier Annex II Vitamin D supplementing note to the Commission.

The FEEDAP also considered the available information on vitamin D₂ content in feed materials. Vitamin D₂ has a markedly lower efficiency in fishes compared to vitamin D₃ (NRC, 2011). The USDA nutrients database18 indicates that cereals and soybeans do not contain vitamin D. Rao and Raghuramulu (1996) studied the concentration of provitamin D and vitamin D in fresh water plankton. Vitamin D₂ was markedly more abundant in zooplankton than in phytoplankton, but concentrations were lower than vitamin D₃ in both commodities: Vitamin D₂ concentrations in phytoplankton and in zooplankton were 0.053 and 0.724 mg/kg dry matter (DM), whereas concentrations of vitamin D₃ were 0.804 and 2.717 mg/kg DM in phytoplankton and in zooplankton, respectively. To the best knowledge of the FEEDAP Panel, no studies have yet measured vitamin D₂ in currently used aquaculture feeds. Data on vitamin D₂ content in plant‐derived ingredients normally used in fish feed were requested in the call for data launched by EFSA, but no information was received. Therefore, the FEEDAP Panel cannot assess the contribution of plant‐based materials to vitamin D intake of farmed fish. However, considering the available data on vitamin D₂ in a range of feed materials, including phytoplankton, and the lower efficiency of vitamin D₂, its contribution to the overall vitamin D content of aquaculture feeds is expected to be a minor one.

Although the data obtained in the Norwegian fish feed surveillance program appear to indicate that total vitamin D₃ content in commercial feed samples has decreased from 2002 to 2008, a clear trend was not apparent (Sissener et al., 2012; see Appendix B). Across all years, the mean value of vitamin D₃ was always equal to or higher than 0.2 mg/kg complete feed. However, the data provided do not clarify whether and to what extent these fish feeds were supplemented with vitamin D₃ (from either fish meal/oil or chemically synthesised). The FEEDAP Panel notes that the average values are always at least fivefold above the vitamin D requirements for salmonids of 0.04 mg/kg (rainbow trout; NRC, 2011) as well as always at least 2.5‐fold above the maximum authorised content in the EU of vitamin D in complete feedingstuffs for fish of 0.075 mg/kg.

Following the call for data launched by EFSA, more recent data on the content of vitamin D₃ in fish feed were available. Data obtained from the Norwegian feed surveillance program of NIFES showed average contents of vitamin D₃ in commercial samples of 0.14 mg/kg (range 0.05–0.46, n = 73) in 2014 (Sanden et al., 2015) and 0.11 mg/kg (range 0.05–0.21, n = 63) in 2015 (Sanden et al., 2016), still approximately threefold the requirements in salmonids and above the maximum vitamin D content authorised in the EU. Experiments with feeds based on formulations with low level of fish products (replaced by plant protein sources) indicate that vitamin D levels from the feed ingredients can be reduced to 0.07 mg/kg (Waagbø et al., 2013, unpublished data).19 Data from the Danish control of nutrients in fish feed of 2015 (unpublished) showed an average vitamin D₃ content of 0.03 mg/kg (0.02–0.04, n = 8).20 Feed monitoring data collected by federal inspection agencies of the German States from 2006 to 2016 showed no appreciable decrease in vitamin D₃ (Appendix C).21 In these two last cases, the vitamin D₃ contents in feeds were compliant with the maximum level authorised in the EU. However, uncertainties remain on how representative are these two data sets for the whole EU.

The vitamin D requirements of fish are in a broad range, 0.006–0.06 mg/kg feed, depending on species as well as farming conditions: salmonids are close to the upper range, i.e. 0.04–0.06 mg/kg (NRC, 2011; Woodward, 1994). There is no risk of deficiency at levels of vitamin D compliant with the upper limit in the EU of 0.075 mg/kg feed.

In conventional fish feed (containing fish meal/oil) the requirements are met, at least to a large extent, by the background concentration of vitamin D naturally present in the marine resources used (Ostermeyer and Schmidt, 2006). The new feed formulations that largely use plant‐based ingredients are expected to contain only 25–33% of the vitamin D₃ content in the conventional formulations, owing to the greatly reduced content of fish meal and fish oil (see above for calculations). Although the contribution by vitamin D₂ cannot be reliably estimated, it is likely to be a minor one.

Conclusions on changing feed composition

The increasing use of plant‐based feed materials could result in a decrease in vitamin D₃ content in feedingstuffs. However, there is no evidence that the current total (background + supplemented) maximum EU content of vitamin D₃ may cause any appreciable risk of deficiency in salmonids.

3.1.1.2. Tolerance of fish to vitamin D

In its previous opinions on vitamin D₃, the FEEDAP Panel did not identify any safety concern for the use of vitamin D₃ in fish (EFSA FEEDAP Panel, 2012a, 2013, 2014). The current maximum authorised feed content of vitamin D₃ (0.075 mg/kg) is well below (about 300‐fold) the upper safe level estimated for fish (rainbow trout) for a period of more than 60 days (NRC, 1987).

The existing literature on elevated feed levels of vitamin D has been reviewed by Lock et al. (2010), Darias et al. (2011) and NRC (2011). The NFSA submitted some publications which show that carnivorous species, such as Atlantic salmon, have evolved having a normal high intake of vitamin D and may tolerate overdosing of two magnitude orders (Horvli and Lie, 1998; Graff et al., 2002a,b). When Atlantic salmon were given diets containing 0.04, 2.21 and 28.68 mg vitamin D₃/kg feed, no effects were observed on weight gain, survival, plasma level of calcium, red blood cell count or haematocrit (Horvli and Lie, 1998). The effects on growth, mortality, calcium content and bone formation were investigated in Atlantic salmon fry fed diets containing 0.2, 5 and 57 mg vitamin D₃/kg feed from first feeding for 14 weeks. No differences were observed on growth parameters, kidney calcium concentration, skeletal malformations or histopathological changes between the different feeding groups, suggesting that salmon can tolerate up to 57 mg/kg feed (Graff et al., 2002a). Comparable data are not available for other fish species.

Conclusions on the tolerance of fish to vitamin D

As regards tolerance to high levels of vitamin D₃, the FEEDAP Panel reiterates its previous conclusion that salmonid fish are highly tolerant (EFSA FEEDAP Panel, 2012a, 2013, 2014). Based on the available literature, the FEEDAP Panel estimated a safe upper dietary level of about 25 mg/kg in fish feed. A total level of 1.5 mg/kg as proposed by the NFSA will remain safe for salmonids with a more than 10‐fold margin of safety.

For other fish species no sufficient data are available. An interspecies uncertainty factor of 10 would not suffice to protect some fish species as e.g. channel cat fish (NRC, 1987). Therefore, the FEEDAP Panel cannot conclude on the safety of a total level of 1.5 mg vitamin D₃/kg feed in fish species other than salmonids.

3.1.1.3. Conclusions on safety for the target species

The FEEDAP Panel reiterates its previous conclusion that (i) the currently authorised levels for vitamin D₃ are sufficient to meet the requirements for salmonids. (ii) Vitamin D₃ is highly tolerated by salmonids. Consequently, a total level of 1.5 mg vitamin D₃/kg compound feed as proposed by the NFSA would still afford a more than 10‐fold margin of safety in salmonids.

For other fish, no sufficient data are available to conclude on the safety of a total level of 1.5 mg vitamin D₃/kg feed.

3.1.2. Safety for the consumer

3.1.2.1. Deposition of vitamin D and vitamin D₃ in fish

Limited studies are available on the feed‐fillet transfer ratio for vitamin D₃ in Atlantic salmon. The available data indicate that the feed‐fillet transfer ratio for vitamin D₃ at high supplementation levels, i.e. above 2 mg/kg feed, can range from 0.1 (Horvli et al., 1998) to 0.13 (Graff et al., 2016); when supplementation levels are lower by one order of magnitude (i.e. in the 0.2 mg/kg feed range), the transfer ratio may be higher, i.e. 0.4 (Graff et al., 2016). However, an experimental study in rainbow trout (450 g body weight, fed for 4 months) performed by Mattila et al. (1999) showed no correlation between increasing dietary concentrations of vitamin D₃ (89, 174 and 539 μg/kg feed) and vitamin D₃ concentration in fillet muscle (which ranged from 5.7 to 15.6 μg/100 g fillet). Upon a request of data to EFSA focal points, two sets of unpublished experimental data in finfish species were received on the transfer of vitamin D₃ from fish feed to fish flesh following high‐dose feed supplementation with vitamin D₃ (e.g. 10, 100 or 1,000 times the maximum authorised content of 0.075 mg/kg feed). According to unpublished data from Denmark, a supplementation of 1.4 mg vitamin D₃/kg feed results in 100 μg/kg fillet in salmon, suggesting a feed‐flesh transfer ratio of 0.07. Unpublished data from Portugal indicate that 0.75 mg/kg feed in rainbow trout results in approximately 120 μg/kg in fillet, with a feed‐flesh transfer ratio of 0.16. The FEEDAP Panel notes that in most of the above studies, no information was provided on the background content of vitamin D₃ in feed; considering this uncertainty, the feed‐flesh transfer is expected to be somewhat lower than the estimates given. However, taking into account the highly variable vitamin D₃ background in aquaculture feeds, no reliable assumption can be made to account for the uncertainty.

Overall, the transfer of vitamin D₃ from feed to salmonid flesh seems variable, possibly in relation to the species and the farming conditions. Even taking into account the uncertainty outlined above, for high supplementation levels the transfer ratio is estimated to be in the range of 0.1–0.15. Accordingly, the proposed increase in the total authorised content of vitamin D₃ in feed to 1.5 mg/kg would lead to vitamin D₃ content in fish flesh ranging from 150 up to 220 μg/kg.

3.1.2.2. Assessment of consumer safety

A UL of 100 and 50 μg/day for adults and children, respectively, has been indicated by the EFSA NDA Panel (2012).

According to the EFSA NDA Panel, data from European populations indicate that vitamin D intakes from all sources in high consumers are about 25%, 75%, 30% and 8% of the UL for adults, infants (up to 1 year), children (1–10 years) and adolescents (11–17 years), respectively. This assessment is consistent with that of the FEEDAP Panel, where the highest intake of vitamin D₃ by adults would result from the consumption of fish (salmon) and milk amounting to 23% of the UL. In particular, the consumption of salmon flesh containing 160 μg/kg of vitamin D₃ (figures from Mattila et al., 1999; Souci et al., 2008 and upper range of NIFES data in 2006) gives the following intakes: high‐consumer toddlers (65 g/day) 10.4 μg/day, high‐consumer adults (125 g/day) 20.0 μg/day, both figures corresponding to approximately 20% of the respective UL (EFSA FEEDAP Panel, 2014, see also Table 3).

Table 3.

Daily exposure of consumers to vitamin D₃ resulting from the consumption of food of animal origin (milk and fish fillet) calculated at the current authorised vitamin D levels (EFSA FEEDAP Panel, 2014) and at the proposed supplementation levels

		Toddlersa		Adultsb
	Concentration in food (μg/kg)	Chronic intakec (g)	Exposure (μg)	Chronic intake (g)	Exposure (μg)
Currently authorised vitamin D levels (0.075 mg/kg)
Milkd	2.1	1,050	2.21	1,500	3.15
Fish (salmon)	160	65	10.4	125	20.0
Proposed supplementation levels (1.5 mg/kg, transfer rate 0.15)
Milkd	2.1	1,050	2.21	1,500	3.15
Fish (salmon)	220	65	14.3	125	27.5

^aToddlers: 1–3 years of age, 12 kg body weight.

^bAdults: 18–65 years of age, 60 kg body weight.

^cChronic intake is the 95th percentile of the distribution of average individual consumption levels (over the survey period) for consumers only from all available EU national surveys.

^dMilk including dairy products.

The FEEDAP Panel notes that a concentration of 220 μg/kg in salmon flesh, resulting from the highest estimate of feed‐flesh transfer with 1.5 mg/kg feed, will increase the intake of vitamin D from salmon to 28% of the UL in high consumers (Table 3). Whereas uncertainties exist regarding the influence of factors such as fish species, the background content of vitamin D₃ and the farming conditions, as well as the limitations of the available data set, the FEEDAP Panel considers unlikely that the proposed increase in vitamin D₃ supplementation in aquaculture feeds would pose any concern that the UL might be exceeded. In addition, considering that, based on FEEDAP Panel calculations (EFSA FEEDAP Panel, 2014), milk and other foods of animal origin would provide no more than 5 μg/day (5% of the UL) in high‐consumer adults, the increase in vitamin D₃ in fish flesh would not pose safety concerns for the consumer. This is consistent with the conclusion of the EFSA NDA Panel that the UL for vitamin D is unlikely to be exceeded in the EU even for high consumers (EFSA NDA Panel, 2012).

3.1.2.3. Conclusions on consumer safety

The FEEDAP Panel recognizes several uncertainties associated with the conditions influencing the transfer of vitamin D₃ from feed to fish flesh, including (but not limited to) the fish species, the background content and the supplementation level of vitamin D_3, and environmental conditions.

Even taking into account the identified uncertainties, an increase in total vitamin D content in fish feeds up to 1.5 mg/kg feed would not lead the UL to be exceeded even in high consumers. Therefore, the proposed increase is not of concern for consumer safety.

4. Conclusions

The FEEDAP Panel reiterates its previous conclusion that the currently authorised level for vitamin D₃ is sufficient to meet the requirements for salmonids. Therefore, it does not see a need to increase the current maximum authorised content based on considerations of animal nutrition.

The FEEDAP Panel concludes that a total level of 1.5 mg vitamin D₃/kg compound feed is safe for salmonids with a margin of safety of at least 10. For other fish, insufficient data are available to conclude on the safety of a total level of 1.5 mg vitamin D₃/kg feed.

The proposed increase is not of concern for consumer safety.

5. General remarks

The FEEDAP Panel recommends that targeted studies are carried out in order to reduce the identified uncertainties concerning nutritional and safety aspects of vitamin D in feeds of farmed fish species other than salmonids. Particular consideration should be given to the tolerance and the influence on vitamin deposition in fish flesh of such factors as: fish species, the capacity for endogenous synthesis, the background content of vitamin D₃ and D₂ of different feed formulations, the supplementation level of vitamin D₃ and farming conditions.

If the total maximum level of vitamin D₃ in fish feed is increased as proposed by the NFSA, this may have implications for the safety for the user and for the aquatic compartment of the environment, not considered here.

Documentation provided to EFSA

1. Vitamin D₃ addition to feedingstuffs for fish. August 2014. Submitted by Norwegian Food Safety Authority.

2. Vitamin D₃ addition to feedingstuffs for fish. September 2016. Submitted by EU Member states and EFSA Focal Points network.

Abbreviations

ANS

EFSA Scientific Panel on Additives and Nutrient Sources added to Food

CAS

Chemical Abstracts Service

DM

dry matter

EC

European Commission

FEEDAP

EFSA Scientific Panel on Additives and Products or Substances used in Animal Feed

IU

International Unit

IUPAC

International Union of Pure and Applied Chemistry

IENECS

European Inventory of Existing Chemical Substances

NFSA

Norwegian Food Safety Authority

NIFES

Norwegian National Institute of Nutrition and Seafood Research

NRC

National Research Council

PCB

polyclhorinated biphenyl

PhEur

European pharmacopoeia

UL

upper level

VDR

vitamin D receptor

Vitamin D

vitamin D₂ and vitamin D₃

VKM

Norwegian Scientific Committee for Food Safety

Appendix A – Data obtained from Member states and EFSA focal points through a call for data

1.

Denmark:

• Data from Danish food composition database (Technical University of Denmark): vitamin D₃ content in raw Atlantic salmon from aquaculture was 6.74 μg/kg (n = 2, http://www.fooddata.dk).

• Unpublished data of part of EU project F 7 ODIN. Trial in salmon fed up to 57,000 IU (1.425 mg) vitamin D₃/kg. Salmon contain up to 10 μg vitamin D₃/100 g flesh.

• Unpublished data of Danish Veterinary and Food Administration. Results from Danish control of nutrients in fish feed sampled and analysed in 2015. Eight samples had an average of 1,152 IU/kg, ranging from 889 to 1,692 IU/kg (average 0.029 mg/kg, range 0.022–0.042 mg/kg).

United Kingdom:

• Department of Health. Institute of Food Research. Nutrient Analysis of Fish and Fish Products. Sampling Report. 2013. http://www.dh.gov.uk/documents. Vitamin D content of canned salmon (red or pink) and smoked salmon (hot or cold smoked).

Germany:

• Federal Office of Consumer Protection and Food Safety (BVL) provided data on analytical results of vitamin D in food and feed from the inspection agencies of the 16 Federal States in Germany. Analytical values of 44 samples (2006–2016) for vitamin D₃ in compound feedingstuffs of fish were provided.

• Federal Research Institute of Nutrition and Food (Max Rubner‐Institute, MRI) provided information on vitamin D content in different fish species from a database of nutritional values of about 15,000 foods (Bundeslebensmittelschlussel, BLS): 87 Analytical values of vitamin D‐cholecalciferol (μg/100 g) of different fish species (period 1989–2010). Salmon raw 3.8 μg/100 g.

• MRI, department of safety and quality of milk and fish products. List of publications on vitamin D contents in fishery products including fish meal and on analytical methods (68 references). In addition, data of:

  • –European nutrition values databases
  • –USDA nutrients database
  • –Souci Fachmann Kraut online database
  • –Swedish food composition database
  • –McCance and Widdowson's composition of foods integrated dataset (Public Health England)
  • –Extract from literature, vitamin D₃ content in rainbow trout, raw:
    • –Ostermeyer and Schmidt (2006) Vitamin D and provitamin D in fish, n = 110, 8.14 μg/100 g
    • –Health D.o. 2013. Nutrient analysis of fish and fish products, n = 9, 6.99 μg/100 g

Greece:

• Greek Ministry of Rural Development and Food: two literature references and vitamin D₃ content in feed for sea bass and sea bream.

Iceland:

• Icelandic Food and Veterinary Authority: published report of the Matís laboratory – Nutrient value of seafoods (http://www.matis.is/media/matis/utgafa/33-11-Naeringargildi-sjavarafurda.pdf), containing data on Vitamin D₃ in farmed salmon (pooled sample containing up to 10 subsamples).

Portugal:

• Data from a Portuguese feed producer on vitamin D₃ content in fish feed during initiation and pre‐fattening (all species). Additional data on vitamin D₃ content in fish feed during fattening of trout 1,250 IU/kg (0.031 mg/kg) but not of salmon.

• Data (Poster of Ramallho‐Ribeiro et al., year not provided) on transference of vitamin D₃ from feed to trout flesh originating from the European Project FP 7 ODIN. Feeding rainbow trout with a control diet (0.031 mg vitamin D₃/kg feed) or an enriched diet (maximum legal limit of 0.075 mg vitamin D₃/kg feed), the content of vitamin D₃ in trout flesh is 18% higher in the trout fed enriched diet (10 μg/100 g fillet in the control vs 11.8 μg/100 g fillet in the enriched).

Norway:

• Data from the Norwegian feed surveillance program, years 2014 (n = 73, average 0.14 mg vitamin D₃/kg feed, range 0.05–0.46) and 2015 (n = 63, average 0.11 mg vitamin D₃/kg feed, range 0.05–0.21). Sanden et al. (2015, 2016)

• Unpublished data on declining levels of vitamin D₃ in fish (Lock E.J. et al., in preparation)

• Unpublished data on vitamin D₃ content in feed for rainbow trout (Waagbo et al., 2013)

• Experimental data of transference of vitamin D₃ from fish feed to fish flesh (Graff et al., 2016), including an exposure assessment for the consumer.

Appendix B – Data submitted by the Norwegian Food Safety Authority on the changes of vitamin D₃ in commercial fish feed

1.

Figure B.1.

Available data on vitamin D in commercial Norwegian fish feed in the period 2002–2008. All data are given as mg/kg, represented by mean ± standard error. The line shows the linear regression analysis and the letters above the bars that do not share the same letters are significantly different (p < 0.05) (Reprinted from Sissener et al., 2012. Copyright John Wiley and Sons)

Figure B.2.

Vitamin D₃ analyses fish feed, grower feed size (fish feed company: EWOS) (copyright Norwegian Food Safety Authority)

Appendix C – Raw data on the content of vitamin D3 in feedingstuffs for fish provided by the German Federal Office of Consumer Protection and Food Safety (BVL)

1.

The German Federal Office of Consumer Protection and Food Safety (BVL) provided the following raw data on vitamin D₃ content in compound feedingstuffs for fish. All the analyses were performed following the VDLUFA Bd. III, Chapter 13.8.1 method. It was obtained through the Federal Inspection Agencies of German States within the period 2006–2016.

No	Year of sample collection	Year of analysis	Water content (%)	Vitamin D3 (mg/kg) on the base of 88% DM of mixed feed
1	2013	2013	4.7	0.03
2	2014	2014	7	0.10
3	2013	2013	8.5	0.08
4	2013	2013	6.8	0.06
5	2011	2011	6.4	0.06
6	2012	2012	6.7	0.07
7	2013	2013	4.7	0.36
8	2013	2013	6.1	0.13
9	2013	2013	7.9	0.04
10	2013	2013	6.6	0.16
11	2013	2014	7.6	0.06
12	2013	2014	7.1	0.06
13	2013	2013	9.19	0.07
14	2013	2013	7.9	0.15
15	2013	2013	4.08	0.04
16	2013	2013	4.08	0.05
17	2013	2014	7.27	0.05
18	2013	2013	8.74	n.b.
19	2013	2013	8.5	0.02
20	2013	2014	5.8	0.03
21	2014	2014	8.5	0.06
22	2014	2014	6	0.10
23	2014	2014	8.4	0.16
24	2014	2014	9.3	0.06
25	2014	2015	8.28	0.04
26	2014	2015	7.49	0.03
27	2009	2009	9.2	0.04
28	2013	2013	1.7	0.08
29	2013	2013	6.4	0.04
30	2013	2013	5.9	0.06
31	2014	2014	10.7	0.06
32	2016	2016	8.1	0.08
33	2014	2014	7.8	0.05
34	2011	2011	8.1	0.05
35	2010	2010	7.9	0.06
36	2010	2010	7.8	0.04
37	2009	2010	8	0.03
38	2006	2006	7.9	0.07
39	2006	2007	7.8	0.09
40	2013	2013	6.8	0.04
41	2015	2015	6.2	0.06
42	2014	2014	7	0.01
43	2015	2015	6.75	0.02
44	2013	2013	5.4	0.05

Figure C.1.

Graphic representation (scattered plot) of the data obtained by the German Federal Office of Consumer Protection and Food Safety (BVL) through the Federal Inspection Agencies of German States on vitamin D₃ content in compound feedingstuffs for fish in the period 2006–2016. The red line indicates the maximum level of vitamin D₃ allowed in fish feed. The black line indicates absence of a lineal trend (copyright BVL)

Suggested citation: EFSA FEEDAP Panel (EFSA Panel on Additives and Products or Substances used in Animal Feed) , Rychen G, Aquilina G, Azimonti G, Bampidis V, Bastos ML, Bories G, Chesson A, Cocconcelli PS, Flachowsky G, Gropp J, Kolar B, Kouba M, López‐Alonso M, López Puente S, Mantovani A, Mayo B, Ramos F, Saarela M, Villa RE, Wester P, Costa LG, Dierick N, Manini P, Tarrés‐Call J and Wallace RJ, 2017. Scientific opinion on the safety of vitamin D₃ addition to feedingstuffs for fish. EFSA Journal 2017;15(3):4713, 19 pp. doi: 10.2903/j.efsa.2017.4713