Dietary strategies to maintain adequacy of circulating 25-Hydroxyvitamin D concentrations

Abstract

The importance of vitamin D intake to nutritional status is a corollary of sunshine deficit. There is a dose-response of serum 25-hydroxyvitamin D (25(OH)D) concentrations to total vitamin D intake in persons who do not receive UVB exposure. This updated summary of vitamin D intakes and sources in adults and children focuses on data from North America and Europe. We explore the evidence that intakes of vitamin D are inadequate with reference to the Institute of Medicine (IOM) Dietary Reference Intakes. Due to mandatory fortification, usual vitamin D intakes are higher in the US and Canada than most of Europe, with the exception of the Nordic countries. Intakes of vitamin D in national surveys are typically below 5 μg/d in most European countries and vary according to country-specific fortification practices, sex and age. The main source of variation is the contribution from nutritional supplements. Usual vitamin D intake estimates need to capture data on the contributions from fortified and supplemental sources as well as the base diet. The current dietary supply of vitamin D makes it unfeasible for most adults to meet the IOM Estimated Average Requirement of 10 μg/d. While supplements are an effective method for individuals to increase their intake, food fortification represents the best opportunity to increase the vitamin D supply to the population. Well-designed sustainable fortification strategies, which use a range of foods to accommodate diversity, have potential to increase vitamin D intakes across the population distribution and minimize the prevalence of low 25(OH)D concentrations.

Key Words::

• Vitamin D intake
• national dietary surveys
• food fortification
• dietary supplements
• food composition data
• public health nutrition
• Estimated Average Requirement

Background and scope

The major source of vitamin D in humans is cutaneous synthesis of cholecalciferol in the presence of UVB radiation (290–315 nm). However, there are several environmental factors that impede year-round synthesis, such as latitude and prevailing weather conditions, which determine availability of UVB of sufficient intensity to stimulate the conversion of 7-dehydrocholesterol in the skin to pre-cholecalciferol. Personal attributes, such as skin pigmentation, age, attire, sunscreen, working environment, physical activity and sun exposure behavior can also prevent or impede vitamin D synthesis. Thus, substantial portions of the world's population, including all who reside at latitudes greater than ∼40^o, rely on body stores and dietary sources to maintain nutritional adequacy of vitamin D all year round. Given that body stores are dependent on sun exposure, the importance of vitamin D intake to nutritional status is a corollary of sunshine deficit (see review [1]).

Calvo et al. [2] provided a comprehensive description of vitamin D intakes and food sources around the world and examined the relative contributions of diet, fortification and supplementation to the vitamin D supply. They concluded that reliance on supplements was variable (6–47 % of average intakes at that time) and that despite various fortification policies, the overall capacity of the diet to supply sufficient vitamin D to maintain healthy serum 25-hydroxyvitamin D concentration (S-25(OH)D) (the biomarker of vitamin D exposure) was inadequate. The purpose of the current review is to provide an updated summary of vitamin D intakes and sources in adults and children, focusing on data from North America and Europe, to evaluate the evidence that intakes of vitamin D are inadequate and to explore the impact of various strategies to increase intakes. A brief overview of some methodological challenges in evaluating the role of diet in vitamin D nutrition, i.e. assessment of total habitual vitamin D intake and adequacy of food composition data for vitamin D, is also provided. While we acknowledge that the metabolic interactions between calcium, vitamin D and phosphate have implications for the nutritional status of each nutrient in the context of inadequate/excessive intakes of another, this summary focuses solely on vitamin D.

The framework for describing adequacy of vitamin D intakes is the recently updated Dietary Reference Intakes (DRI) for vitamin D from the US Institute of Medicine (IOM) [3]. Using a Risk Assessment Framework [4] supported by systematic evidence-based reviews [5,6], the IOM expert committee defined adequacy of S-25(OH)D in terms of the concentrations associated with various indicators of bone health. Data from randomised placebo-controlled studies that reported the dose-response of S-25(OH)D to total intake of vitamin D in northern latitudes in Europe and Antarctica during their respective winter seasons and in conditions of minimal UVB radiation enabled the committee to specify Dietary Reference Intakes (DRI) for vitamin D. The Estimated Average Requirement (EAR) for all persons above 12 months is 10 μg/d (400 IU) and the Recommended Dietary Allowance (RDA) is 15 μg/d (600 IU) up to the age of 70 y, and 20 μg/d (800 IU) thereafter. In the absence of sufficient data to define reference intervals, an Adequate Intake (AI) value for infants up to 12 months of 10 μg/d was indicated. The Tolerable Upper Intake Level (UL) was revised upwards to 100 μg/d (4,000 IU) in persons over the age of 9 y, with lower levels for younger children down to 25 μg/d (1,000 IU) in infants up to 6 months. Cashman has provided an overview of the US, UK and EU dietary reference intervals in this issue of the Proceedings of the 13^th Bergmeyer Conference [7].

Dietary Intakes of Vitamin D

Adults

Foods containing naturally occurring vitamin D are limited and many are not consumed on a regular basis. Natural sources include oily fish, meat, dairy, egg yolk and mushrooms. Depending on legislation, some foods are fortified with vitamin D, including milk, yoghurt, spreads, cheese, juices, breads and breakfast cereal. In addition, vitamin D is available as a dietary supplement, either as vitamin D2 or vitamin D3.

Bailey and colleagues at the NIH Office of Dietary Supplements described vitamin D intake from foods and supplements from the National Health and Nutrition Examination Survey (NHANES 2005–2006) [8], summarized in Table I. The rate of use of D-containing supplements varied from 21 % in young women to >–40 % in older adults. Mean daily vitamin D intake from food alone ranged from 3.6 μg in women aged 19–30 y to 5.6 μg in men over 70 y; intake from all sources ranged from ∼6 to 11 μg/d. Less than 7 % of adults over the age of 51 y had an intake greater than 10 μg/d (the AI for vitamin D in 51–70 y olds at the time) from food; this was <–1 % in older women. However, when dietary supplements were included, the prevalence of meeting the AI increased in all age groups. Notwithstanding that 49 % of them took supplemental vitamin D, adults ≥–71 y were at highest risk of not meeting their AI and 77 % had average intakes <–15 μg/d. Further analysis in the United States population examined contributions of micronutrients to usual intakes derived from all sources (naturally occurring, fortified and dietary supplements) based on NHANES 2003–2006 data [9]. The median vitamin D intake among adults was 1.8 μg/d from naturally occurring sources (base diet). When fortified foods were included, the median increased to 3.9 μg, and to 5.8 μg when dietary supplements were included. Mean intakes were higher than the median values, indicating the skewed distribution of vitamin D intakes, particularly when contributions from fortified and supplemental sources are considered. The prevalence of inadequate intakes, determined by the % of the EAR of <–10 μg/d, was 68 % overall when intakes were calculated from all sources including supplements. Exclusion of supplemental vitamin D and fortified foods increased the prevalence of not meeting the EAR to 100 %.

Table I. Vitamin D intakes (μg/d) in adults from selected national nutrition surveys and cohort studies.

Reference	Country	Survey	Year	Sex	Age group (y)	n	Mean (SD) (μg)	Median (IQR) (μg)	Sources	National fortification policy
Bailey et al. 2010 [8]	USA	NHANES	2003–2006	Men	19–30	549	6.6 (9.4) 5.1 (7.0)	NR NR	All Food	Milk, mandatory; other foods voluntary
					31–50	758	7.9 (8.3) 5.4 (8.3)	NR NR	All Food
					51–70	614	8.8 (9.9) 5.1 (7.4)	NR NR	All Food
					>–71	368	10.7 (13.4) 5.6 (7.7)	NR NR	All Food
				Women	19–30	481	5.8 (6.6) 3.6 (6.6)	NR NR	All Food
					31–50	693	7.7 (13.2) 4.4 (7.9)	NR NR	All Food
					51–70	610	10.1 (24.7) 3.9 (9.9)	NR NR	All Food
					>–71	332	10.0 (9.1) 4.5 (3.6)	NR NR	All Food
Fulgoni et al. 2011[9]	USA	NHANES	2003–2006	Both	>–19	8860	8.1 (18.9)	5.8 (3.1, 12.1)	All	Milk, mandatory; other foods voluntary
				Both	>–19		4.5 (9.4)	3.9 (2.5, 5.8)	Food
				Both	>–19		2.0 (3.8)	1.8 (1.2, 2.6)	Base diet
Vatanparast et al. 2010 [10]	Canada	CCHS 2.2	2004	Men	19–50	4650	5.8 (13.6)	5.7 (4.0, 7.8)	Food	Milk and margarine mandatory
					51–70	2730	7.0 (20.9)	5.6 (3.7, 8.3)	Food
					>–70	1605	6.7 (16.0)	5.3 (3.9, 8.0)	Food
				Women	19–50	5018	5.1 (14.2)	3.5 (1.7, 6.0)	Food
					51–70	3412	5.1 (17.5)	4.5 (3.2, 6.6)	Food
					>–70	2777	6.1 (36.9)	4.4 (3.2, 6.4)	Food
Whitton et al. 2011 [12]	UK	NDNS	2008–09	Men	19–64	181	NR	2.8 (1.9–3.6)	Food	Margarine mandatory; other foods voluntary
				Women	19–64	253	NR	2.3 (1.3–3.4)	Food
Dept of Health 2011 [11]	UK	NDNS	2008/9–‘09/10	Men	19–64	346	3.7 (3.3) 3.1 (2.3)	2.8 2.5	All Food
					>–65	96	5.0 (4.0) 3.9 (2.9)	3.9 3.2	All Food
				Women	19–64	461	3.7 (3.1) 2.6 (1.9)	2.6 2.1	All Food
					>–65	128	4.5 (3.8) 2.9 (1.7)	3.1 2.4	All Food
Hill et al. 2004 [13]	Ireland	NSIFCS	1997–1999	Men	18–64	662	4.4 (3.5)	3.3 (NR)	All	Margarine mandatory; other foods voluntary
				Men	18–64	662	3.7 (2.2)	3.1 (NR)	Food
				Women	18–64	717	4.0 (3.8)	2.6 (NR)	All
				Women	18–64	717	2.8 (1.9)	2.3 (NR)	Food
Jenab et al. 2009 [18]	Europe	EPIC	1995–2000	Men	35–74	13025	5.5 (11.4)	NR	Food	Various
	(10 countries)			Women	35–74	23009	3.6 (NR)	NR	Food
	South			Men	35–74	4530	4.2 (6.7)	NR	Food
				Women	35–74	7372	2.6 (8.6)	NR	Food
	Central			Men	35–74	3807	4.7 (6.2)	NR	Food
				Women	35–74	8561	3.4 (9.3)	NR	Food
	North			Men	35–74	4688	7.4 (6.8)	NR	Food
				Women	35–74	7076	5.0 (8.4)	NR	Food

SD, Standard deviation; IQR, interquartile range; NR, not reported; NHANES, National Health and Nutrition Examination Survey; CCHS, Canadian Community Health Survey; NDNS, National Diet and Nutrition Survey; NSIFCS, North-South Ireland Food Consumption Survey, Epic, European Prospective Investigation into Cancer and Nutrition.

In Canada, vitamin D intakes from food sources were calculated from data reported in the 2004 Canadian Community Health Survey Cycle 2.2 (CCHS 2.2) [10]. Median intakes ranged from 3.5 to 4.5 μg in women and 5.3 to 5.7 μg in men. In adults aged 51–70 y, 80 % of men and 92 % of women had intakes below the EAR and 95 % of adults over 70 y failed to meet the EAR for vitamin D of 15 μg/d from foods. Fortified milk products and meat contributed 49 and 31 % of dietary vitamin D, respectively. Ultimately, despite mandatory fortification of milk, the majority of Canadians consumed less than the EAR for vitamin D. The authors encouraged increasing the range of foods to which vitamin D can be lawfully added and wider practice of voluntary fortification.

Food consumption surveys throughout Europe have consistently reported low vitamin D intakes in the UK [11,12], Ireland [13], Denmark [14], France [15] and Finland [16]. Data from the first year of the rolling program of the UK National Diet and Nutrition Survey (NDNS 2008–2009) showed a median daily vitamin D intake of 2.8 μg in men and 2.3 μg in women from food sources (including fortified foods) [12]. Intakes were slightly lower among males than previous NDNS data and similar in females. Intakes from supplements were not included in the paper by Whitton et al. [12] but are included in the summary report by the UK Department of Health [11], which merged data from the 2008–09 and 2009–10 rolling surveys: median intakes ranged from 2.1 μg in women below 65 y to 3.2 μg in men over 65 y.

The mean daily intake of vitamin D from food sources only in Irish adults from the North/South Ireland Food Consumption Survey was estimated at 3.2 μg [13]. Despite a low prevalence of consumption of D-containing supplements at 15 %, the contribution from supplements increased the overall mean intake to 4.2 μg/d. Vitamin D-containing supplement users had intakes of 7.1 μg/d. Similar intakes were reported among Danish men and women aged 18–49 y who used supplements, at 7.8 and 7.6 μd/d, respectively, compared with 2.7 and 2.0 μg/d in non-users [14]. Intakes from food only in supplement users were 2.8 and 2.1 μg/d in men and women respectively, indicating that differences in vitamin D intakes between users and non-users were attributable to the contribution from supplements and not to any differences in dietary habits. This point is worth noting: intakes of vitamin D, unlike many other nutrients, are not necessarily higher in those who have a healthier diet per se, but are likely to be higher in those who make a conscious effort to eat healthily due to the use of dietary supplements, which is seen as a health-promoting behavior. The report by Tetens and colleagues [14] showed that Danish adults who reported a stronger “intention to eat healthy” were more likely to be users of dietary supplements.

In an analysis of the second French Individual and National Study on Food Consumption (2006–2007), Dufour et al. [15] assessed the safety of various mathematical models developed to propose maximum safe levels of nutrient addition to fortified foods and dietary supplements. They reported median vitamin D intakes from food sources (excluding fortification) of 2.1 μg/d among adults aged 18–79 y. In Finland, fortification of butter/spreads and milk products has been widespread since 2003. Using FINDIET 2002, Hirvonen et al. [16] reported median intakes of almost 5 μg/d in men and women from all sources, and recommended a refined fortification model to increase intakes. In the meantime, the updated FINDIET 2007 reported average intakes of 5.2 and 7.1 μg/d among men and women, respectively, suggesting that the fortification policy implemented in 2003 produced a slight increase in mean vitamin D intakes, at least in men [17].

Jenab et al. [18] described vitamin D intakes among 35–74 y old adults in 10 countries in the European Prospective Investigation into Cancer and Nutrition (EPIC) study. Mean daily vitamin D intakes from food sources only were 5.5 and 3.6 μg/d among men and women with a pronounced North-South gradient, shown in Table I. In a further investigation of the geographical variation in nutrient intakes, Freisling et al. [19] observed lowest vitamin D intakes in Italy and highest in Sweden and Norway, with most other countries within ∼25 % of the EPIC mean (reported here as 4.8 and 3.3 μg/d in men and women, respectively). As part of the EC-funded EURRECA (EURopean micronutrient RECommendations Aligned Network of Excellence; www.eurreca.org), nutrient intake data were obtained from the European Nutrition and Health Report II, and summarised according to participating countries [20]. Intakes of vitamin D among men were lowest in Spain (<–2 μg), between 3 and 4 μg in most other countries (UK, Germany, Denmark, Ireland, Netherlands, Italy, Portugal) and 6, 7 and 11 μg/d in Sweden, Finland and Norway, respectively. Intakes among women were lower in almost all countries. Similarly with most micronutrients, reported intakes of vitamin D in national surveys and large cohort studies (notably EPIC, which uses a standardised dietary assessment method and food composition database) vary according to country-specific fortification practices, sex and age.

Children

The report by Bailey et al. [8] also described vitamin D intakes from food and supplements in boys and girls in different age groups using NHANES 2003–2006 data, summarised in Table II. Mean vitamin D intakes in boys aged 4–18 y were 5.7–6.4 μg from food and 6.9–9.3 μg from all sources. In girls of the same age, intakes ranged from 3.8–5.5 μg from food and 5–7.9 μg/d from all sources. The highest prevalence (80 %) of meeting the AI of 5 μg at that time was among boys aged 4–8 y and the lowest prevalence (32 %) was among girls aged 14–18 y. Prevalence of D-containing supplement use ranged from 16 % in 14–18 y old boys to 43 % in 4–8 y olds and from 27 % of 14–18 y old girls to 34 % of 1–8 y olds. Further analysis of the same survey [9], which disaggregated the contribution from the base diet, fortification and supplements, showed that in all children and adolescents aged 2–18 y (n =–7,250), mean intake of vitamin D from naturally occurring foods was 1.7 μg/d, which increased to 6.1 μg when fortification was included and to 8.3 μg when supplements were added. The percentage of 2–18 y olds with usual intakes below the EAR of 10 μg was 100 % from the base diet, decreasing to 87 % when fortified foods were included and further to 73 % with the addition of the contribution from supplements.

Table II. Vitamin D intakes (μg/d) in children and adolescents from national nutrition surveys.

Reference	Country	Survey	Year	Sex	Age group (y)	n	Mean (SD) (μg)	Median (IQR) (μg)	Sources	National fortification policy
Bailey et al. 2010 [8]	USA	NHANES	2003–2006	Boys	4–8	431	9.3 (8.3) 6.4 (6.2)	NR NR	All Food	Milk, mandatory; other foods voluntary
				Boys	9–13	522	7.5 (16.0) 5.7 (4.6)	NR NR	All Food
				Boys	14–18	654	6.9 (12.8) 6.1 (10.2)	NR NR	All Food
				Girls	4–8	468	7.9 (13.0) 5.5 (6.5)	NR NR	All Food
				Girls	9–13	525	7.7 (22.9) 5.3 (13.7)	NR NR	All Food
				Girls	14–18	643	5.0 (12.7) 3.8 (5.1)	NR NR	All Food
Fulgoni et al. 2011 [9]	USA	NHANES	2003–2006	Both	2–18	7250	8.3 (25.5) 6.1 (8.5) 1.7 (3.4)	6.3 (3.9, 0.5) 5.4 (3.5, 7.9) 1.5 (1.0, 2.1)	All Food Base diet	Milk, mandatory; other foods voluntary
Vatanparast et al. 2010 [10]	Canada	CCHS 2.2	2004	Both	1–8	5655	6.2 (7.5)	5.6 (4.1, 7.5)	Food	Milk and margarine mandatory
				Boys	9–18	4536	7.3 (6.7)	6.9 (4.9, 9.4)	Food
				Girls	9–18	4406	5.4 (6.6)	5.0 (3.7, 6.8)	Food
Dept of Health 2011[11]	UK	NDNS	2008/9–2009/10	Boys	4–10	210	2.7 (2.1) 1.9 (0.9)	2.2 1.8	All Food	Margarine mandatory; other foods voluntary
					11–18	238	2.5 (1.5) 2.4 (1.3)	2.2 2.1	All Food
				Girls	4–10	213	2.4 (1.7)	2.0	All
					11–18	215	1.9 (1.0) 2.2 (1.8) 1.9 (1.2)	1.8 1.7 1.6	Food All Food
Whitton et al. 2011 [12]	UK	NDNS	2008–09	Boys	4–10	119	NR	1.8 (1.2–2.5)	Food
					11–18	114	NR	2.0 (1.3–2.9)	Food
				Girls	4–10	119	NR	1.9 (1.3, 2.4)	Food
					11–18	110	NR	1.8 (1.3–2.4)	Food
Verkaik-Kloosterman et al. 2011 [21]	Netherlands	DNFCS young children	2005–06	Both	2–3	640	4.4 1.8	3.8 (2.1, 6.3) 1.8 (1.4, 2.2)	All Food	Voluntary
					4–6	639	2.7 2.0	2.3 (1.7, 3.2) 2.0 (1.5, 2.4)	All Food
Elmadfa et al. 2009 [20]	Europe	ENHR	2009	Boys	4–6	2541	1.8–5.8	NR	Mixed	Various
				Girls	4–6	2295	1.5–6.5	NR	Mixed
				Boys	7–9	2736	1.5–6.4	NR	Mixed
				Girls	7–9	2597	1.5–5.1	NR	Mixed
				Boys	10–14	4069	1.5–4.8	NR	Mixed
				Girls	10–14	4056	1.2–4.5	NR	Mixed
CSIRO 2008 [24]	Australia	Children's Survey	2007	Boys	4–8	NR	3.0	NR	Food	Margarine and fat spreads mandatory; dairy, voluntary
				Boys	9–13	NR	3.4	NR	Food
				Boys	14–16	NR	4.0	NR	Food
				Girls	4–8	NR	2.7	NR	Food
				Girls	9–13	NR	2.7	NR	Food
				Girls	14–16	NR	2.8	NR	Food
Irish Universities Nutrition Alliance [25]	Ireland	NCFS	2003–2004	Boys	5–12	293	2.2 (2.2) 1.4 (1.1)	1.4 1.1	All Food	Milk, mandatory; other foods voluntary
				Girls	5–12	301	2.3 (2.3) 1.6 (1.2)	1.5 1.2	All Food
Irish Universities Nutrition Alliance [26]	Ireland	NTFS	2005–2006	Boys	13–17	224	3.0 (2.6) 2.4 (1.9)	2 1.9	All Food	Milk, mandatory; other foods voluntary
				Girls	13–17	217	2.3 (2.2) 1.8 (1.6)	1.6 1.3	All Food

SD, standard deviation; IQR, interquartile range; NR, not reported; NHANES, National Health and Nutrition Examination Survey; CCHS, Canadian Community Health Survey; NDNS, National Diet and Nutrition Survey; DNFCS, Dutch National Food Consumption Survey; ENHR, European Nutrition and Health Report 2009; Australian National Children's Nutrition and Physical Activity Survey; NCFS, National Children's Food Survey; NTFS, National Teens’ Food Survey.

The Canadian CCHS 2.2 reported vitamin D intakes from food sources only in children [10] that were very similar to adults. Boys and girls aged 1–8 y had mean intakes of 6.2 μg/d and more than 60 % were achieving >–5 μg. Boys aged 9–18 y had higher intakes (7.3 μg/d) than all other sex and age groups. Reflecting the vitamin D food fortification policy in Canada, where milk and margarine are fortified on a mandatory basis, milk products contributed 75 % of dietary vitamin D intake in children aged 1–8 y. Based on food records from four consecutive days, data from the NDNS rolling programme [11,12] showed that the mean daily vitamin D intake from food sources in the UK was between 1.9 and 2.4 μg in boys and girls aged 4–18 y. When supplements were included, intakes in boys increased to 2.7 and 2.5 μg in ages 4–10 y and 11–18 y, respectively, and intakes in girls increased to 2.4 and 2.2 μg in ages 4–10 y and 11–18, respectively.

Verkaik-Kloosterman et al. 2011 [21] estimated habitual vitamin D intake from food and supplements among young children using data from the Dutch food consumption survey. From food sources only, mean vitamin D intake was 1.8 and 2 μg/d in 2–3 and 4–6 y olds, respectively. Vitamin D intakes in users of dietary supplements were 5.6 μg/d in 2–3 y olds and 3.7 μg/d in 4–6 y olds. Similarly with the trends observed in adults in Denmark [14], vitamin D intakes from food sources only were comparable between users and non-users of dietary supplements. Total vitamin D intakes were below the EAR of 10 μg/d in almost everyone, with the exception of ∼5 % of supplement users. Similar data were reported in Flemish preschoolers [22]: mean (95 % CI) intakes of vitamin D from food only were 2.0 (1.8, 2.2) μg/d in 2.5 to 6.5 y olds. Intakes of vitamin D decreased with age and only 4 % of 2–3 y olds and 1 % of children 4–6 y had intakes of vitamin D at the EAR of 10 μg/d from food.

The European Nutrition and Health Report [23] summarised vitamin D intake in children and teens from a variety of sources. Data showed an interval of intakes from 1.2–6.5 μg in children and teenagers aged 4–14 y, depending on country, sex and age group. Highest intakes of vitamin D in children were reported in the Nordic countries, where children between 10 and 14 y had intakes of 5.1–6.8 μg/d. In adolescents aged 15–18 y in Denmark, Germany, Italy, Norway, Poland, Slovenia, Spain and the Netherlands, mean intakes ranged from 1.5 μg/d in Spanish girls to 7.5 μg/d in Norwegian boys, respectively. Intakes were <–5 μg/d among adolescents of all ages in every participating country except for Norwegian boys and girls and Polish boys. Vitamin D intakes from food sources reported by the CSIRO in Australian children [24] were 2.8–4 μg/d, with the lowest intakes in girls and the highest in adolescent boys. In Ireland, the national food surveys in children (NCFS) [25] and teens (NTFS) [26] reported mean vitamin D intakes from food sources of 1.5 μg in children aged 5–12 y and 2.4 and 1.8 μg in boys and girls aged 13–17 y, respectively. Almost all (98 %) children and teens had intakes below the EAR of 10 μg/d, and although supplement users had higher intakes of vitamin D, at 6–7 μg/d, 90 % of users had inadequate intakes.

Data from European national dietary and food consumption surveys use various methods of data collection, analysis and reporting, making meaningful comparison problematic. Nonetheless, with regard to vitamin D, it appears from these data that the main source of variation in intake estimates across the population >–2 y is the contribution from nutritional supplements. Surveys that do not include supplemental vitamin D omit the largest source of vitamin D in supplement users, which accounts for up to 50 % of the adult population and a variable percentage of children depending on age; the prevalence of supplement use decreases in older children [8]. This has implications for assessing the public health significance of low vitamin D intakes if intake estimates are derived from food sources only, as higher total intakes increase circulating S-25(OH)D. In an analysis of the role of D-containing supplements in increasing the prevalence of Canadians meeting DRI thresholds of circulating S-25(OH)D of 30, 40 and 50 nmol/L [3], Whiting et al. [27] reported that supplement users aged 6–79 y, who represented 31 % of the population surveyed during the 2007–2009 Canadian Health Measures Survey-Cycle I, had not only higher S-25(OH)D than non-users but also a much lower prevalence of S-25(OH)D <50 nmol/L (19 vs. 37 %). The data summarized here show that the current availability of vitamin D in the food supply (excluding the contribution from supplements, which is a matter of personal choice) is insufficient to reduce the prevalence of the population with intakes of vitamin D from food sources below the EAR of 10 μg/d.

Food Fortification with vitamin D

During the First and Second World Wars, fortification of food became widespread in order to prevent micronutrient deficiencies and to address nutrient losses during food processing. During this time, it became common practice to add a number of nutrients to foods, including iodine to salt; vitamins A and D to margarine; vitamin D to milk; and thiamine, riboflavin, niacin, and iron to flour [28]. Fortification of food with vitamin D began in response to the development of rickets in children in industrialized countries. In the 1920s, an estimated 75 % of children in New York City were afflicted with rickets [29] and 339 related deaths were reported nationwide in 1933 [30]. In 1932, soon after the structure of vitamin D2 was determined, vitamin D fortification of milk began in the United States at a level of 400 IU/quart (1 quart = 0.95 L) [31]. This was followed by the fortification of many products in the 1930:s, including peanut butter, hot dogs, soft drinks, bread and beer, leading to the virtual eradication of rickets [30]. In the UK, fortification of margarine with vitamin D first became mandatory at the start of World War II, at a level of 8 μg/100 g, to match the nutritional profile of butter [32]. However, the fortification process was poorly monitored and, when over-fortification of some milk products shortly after World War II led to vitamin D intoxication in young children, the practice largely ceased [33].

Nutrient intakes, unlike substances such as drugs or chemical toxicants where there is zero to minimal background exposure, can pose a dual risk, due either to consumption at a level too low to deliver benefit (deficient), or sufficiently high to pose the threat of an adverse effect (toxic) [34]. To minimize the risk of excess micronutrient intakes, safe upper intake levels (UL:s) for micronutrients have been established by authoritative agencies internationally. The UL for vitamin D was recently increased to 100 μg/d by the IOM [3] and while several analyses of existing data have shown that much higher doses are not associated with adverse effects [35], dedicated studies are needed to assess adverse health effects related to long-term, high dose (although not necessarily “toxic”) intakes of vitamin D [3]. IARC pointed out that there are no data on the health hazards of maintaining high S-25(OH)D in healthy persons over long periods, and urged caution to be mindful of past experiences with other compounds (e.g. some anti-oxidants and hormone replacement therapies) that showed serious adverse effects [36]. U- and reverse J-shaped distributions have been described for S-25(OH)D and adverse consequences, including all-cause mortality, PTH suppression and intrauterine growth restriction [3], which deserve serious consideration by researchers investigating health effects of vitamin D in randomised controlled trials as well as those investigating adverse effects.

In the modern food industry and regulatory context, the likelihood of excessive intakes due to food fortification is relatively low. Liberal fortification practices have been in place in a number of countries for many years, with no reported adverse health effects [37]. Currently, fortification practices vary between countries and may be applied voluntarily by manufacturers or implemented by national legislation. In order to harmonize voluntary food fortification practices throughout EU member states, and to provide consumer protection, a regulation governing the addition of vitamins and minerals was proposed by the European Commission [38], which does not affect existing national rules regarding compulsory fortification. Micronutrients may be added to foods for the purpose of restoring the nutrients lost during processing; for producing a substitute food that resembles a common food in appearance and nutritive value; or for the purpose of enriching a food in order to provide a nutritional or physiological effect.

In the UK and Ireland, a significant number of foods are fortified on a voluntary basis. Hannon et al. [39] reported that fortified foods contributed 11 % of total vitamin D intake in Irish women aged 18–64 y, increasing the median daily intake by 23 %, with no concomitant risk of excessive intakes of vitamin D. The Finnish experience over the past 10 years is exemplary with respect to implementation and evaluation of vitamin D fortification policies. In 2003, the government initiated regulations for the optional vitamin D3 fortification of milks and yoghurt (0.5 μg/100 g) and margarine and spreads (10 μg/100 g) [40]. The resulting increase in daily vitamin D intakes in 100 adolescents (12–17 y) was 3.3 in 12–14 y olds and 1.7 in 15–17 y olds [40]. Circulating S-25(OH)D were measured immediately prior to implementation of the fortification policy and one year later 196 Finnish men in the Defence Forces (18–28 y) [41]. After fortification, mean S-25(OH)D during the winter months increased by 50 % and the prevalence of S-25(OH)D <–40 nmol/L decreased from 78 % in January 2003 to 35 % in January 2004. The prevalence of S-25(OH)D <–25 nmol/L decreased from 19 to 5 % in the same year. After fortification, 5 % of participants had S-25(OH)D >–100 nmol/L (101–111 nmol/L); however, no clinical signs of toxicity were seen. Vitamin D intakes increased from 2.1 to 4.5 μg/d and S-25(OH)D increased from 54.7 to 64.9 nmol/L in 4 y old children during winter months, before (n =82) and after (n =36) fortification [42]. Lehtonen-Veromaa et al. [43] found a modest increase in intakes (from 4 to 5.4 μg/d) adolescent girls due to the fortification policy, but no change in S-25(OH)D, probably due to low dairy consumption among adolescent girls.

The problem of fortifying a single staple, for example milk, or focusing on a commodity sector such as dairy, is that it does not increase the vitamin D supply in non-consumers. For example, Babu & Calvo [44] suggested that fortification of wheat flour may be more efficacious in alleviating vitamin D deficiency in countries such as India and Jordan, where pasteurized milk is not widely consumed. Van Horn et al. [45] showed that African-American girls relied more heavily on meat and beans as a source of vitamin D than white girls, emphasising the need to account for diversity in food consumption patterns when developing fortification strategies.

O'Donnell et al. [46] carried out a systematic review to assess the efficacy of food fortification on S-25(OH)D from randomised controlled trials using vitamin D fortified foods and found evidence for benefit of fortification. We recently updated the systematic review by O'Donnell [46] and carried out a meta-analysis to re-evaluate the evidence for efficacy of vitamin D fortification. On the basis of 16 separate randomized controlled studies from around the world, our analysis showed that foods fortified with vitamin D increase circulating S-25(OH)D in a dose-dependent manner [47]. However, there is a need for stronger data on the effect of vitamin D-fortified food on circulating S-25(OH)D, deficiency prevention and potential health benefits. Careful consideration must be given to the range of products used for fortification and amount of vitamin D used in each, to optimize effectiveness and minimize risk of excessive intakes. This can only be achieved by modeling usual food consumption intakes in representative populations and evaluating potential fortification initiatives by carrying out high quality food-based randomized controlled studies in the community that measure the impact on S-25(OH)D in the population to achieve efficacy without compromising safety.

A challenge encountered with fortifying several foods, is that the food composition databases are hard-pressed to keep step with innovations. Access to current and accurate food composition data is a requirement for the estimation of vitamin D intakes. More comprehensive coverage of the vitamin D content, including D3, D2 and 25(OH)D, of staple foods is needed. Although there are a number of accessible food composition databases, values are often acquired from recipe calculation procedures, borrowed from other databases, or from unknown sources. In addition, vitamin D values for many potentially vitamin D-containing foods, such as certain milk, meat and fish products, remain unreported. There is currently no authoritative source of vitamin D food composition data. McCance & Widdowson's The Composition of Foods integrated dataset (CoF IDS) [48] contains vitamin D data for 3,423 foods, primarily from HPLC analysis and calculated from cholecalciferol and 25-hydroxycholecalciferol. These data date back to 1995 and are currently being updated. The USDA National Nutrient Database for Standard Reference (USDA SR23, September 2010) [49] is the most current source of analysed vitamin D composition data and contains vitamin D values for approximately 600 foods, including meat, fish and mushrooms; however, analyses do not currently account for 25(OH)D, as the analytical methodology used to determine this metabolite of vitamin D is considered to be insufficiently validated. Further problems arise when estimating vitamin D intakes from manufactured products and nutritional supplements: new food products are continually introduced to the market, and are constantly reformulated, and nutritional supplements are often omitted from food composition databases. The most significant advance in standardisation and harmonization of food composition data to date has been the EC-funded EuroFIR [European Food Information Resource] Network of Excellence, www.eurofir.net, which included 49 partners from 27 countries, most of them national food composition database compilers, including the USDA. Through the EuroFIR eSearch Prototype, the project succeeded in linking 25 European food composition databases compiled using the standardised EuroFIR approach. Investment in the provision of quality food composition data for vitamin D is necessary to support assessment of vitamin D intakes in national surveys and research in nutrition and health.

In conclusion, usual vitamin D intakes are higher in the US and Canada than most of Europe, with the exception of the Nordic countries. Intakes of vitamin D in national surveys are typically below 5 μg/d in most European countries and vary according to country-specific fortification practices, sex and age. As the main source of variation is the contribution from nutritional supplements, usual vitamin D intake estimates need to capture data on the contributions from fortified and supplemental sources as well as the base diet. The current dietary supply of vitamin D makes it unfeasible for most adults to meet the IOM Estimated Average Requirement of 10 μg/d. While supplements are an effective method for individuals to increase their intake, food fortification represents the best opportunity to increase the vitamin D supply to the population. Well-designed sustainable fortification strategies, which use a range of foods to accommodate diversity, have potential to increase vitamin D intakes across the population distribution and minimize the prevalence of low S-25(OH)D. Success will depend on international collaboration and strategic investment.

Questions and Answers

R Vieth, Canada

The only published paper comparing vitamin D2 and vitamin D3 is that of Holick. It is forgotten that the method he used was the Quest Laboratories MS 25(OH)D method, which was subsequently withdrawn. The amount of vitamin D being given was small, generating a very small signal, difficult to detect. The assay results must be suspect.

Furthermore, never once did the vitamin average vitamin D2 group result in a S-25(OH)D concentration greater than or equal to that in the vitamin D3 group.

M Kiely

My point was that administering a bolus dose is not the same as taking a nutritionally relevant small quantity on a regular basis. In the Holick study, 25 μg was the daily dose. This does create a significant signal and leads to an increase in S-25(OH)D of 30–40 nmol/L which produces a measurable signal.

G Jones, Canada

There are at least three other publications which support the findings of Holick. I believe it is the dose and being given frequently, rather than once a month. The difference between the two sets of findings also relates to the differences in the catabolic rates of the vitamin D2 metabolites. So, if you give a lot, you end up with this more obvious difference.

M Kiely

This is relevant, because it is being used as a clinical, therapeutic measure to cure deficiency. It is important that the vitamin D2, vitamin D3 issue and the duration of therapy is agreed.

E Delvin, Canada

Taken that many people are vitamin D deficient I am curious about other issues related to clinical manifestations of the deficiency, e.g. rickets. Have you looked at issues relating to diseases in which vitamin D might be involved?

M Kiely

We have reported data in relation to bone heath in small groups. The best example of a large group is the Young Hearts Cohort study, in Northern Ireland in which S-25(OH)D3 concentration was looked at in relation to CVD markers and bone markers. There was not much of evidence in relation to CVD markers, not surprisingly because of the age group. In relation to bone markers there were significant adverse outcomes in having low S-25(OH)D3.

Acknowledgement

We acknowledge the support of the EURRECA (EURopean micronutrient RECommendations Aligned) and EuroFIR [European Food Information Resource] Networks of Excellence, funded by the European Commission, for supporting L Black.

Declaration of interest: The authors have no conflicts of interest.