Scientific Opinion on Dietary Reference Values for iron


Panel on Dietetic Products, Nutrition and Allergies
EFSA Journal
EFSA Journal 2015;13(10):4254 [115 pp.].
Panel members at the time of adoption
Jean Louis Bresson, Barbara Burlingame, Tara Dean, Susan Fairweather-Tait, Marina Heinonen, Karen Ildico Hirsch-Ernst, Inge Mangelsdorf, Harry McArdle, Androniki Naska, Monika Neuhäuser-Berthold, Grażyna Nowicka, Kristina Pentieva, Yolanda Sanz, Alfonso Siani, Anders Sjödin, Martin Stern, Daniel Tomé, Dominique Turck, Hendrik Van Loveren, Marco Vinceti and Peter Willatts.

The Panel wishes to thank the members of the Working Group on Dietary Referenve Values for Minerals: Peter Aggett, Carlo Agostoni, Susan Fairweather-Tait, Marianne Geleijnse, Ambroise Martin, Harry McArdle, Androniki Naska, Hildegard Przyrembel and Alfonso Siani for the preparatory work on this scientific opinion and EFSA staff: Anja Brönstrup, Sofia Ioannidou, Laura Martino and Liisa Valsta for the support provided to this scientific opinion.

Opinion of the Scientific Committee/Scientific Panel
On request from
European Commission
Question Number
23 September 2015
21 October 2015
European Food Safety Authority (EFSA), Parma, Italy
Download Article (2.6 MB)

Following a request from the European Commission, the Panel on Dietetic Products, Nutrition and Allergies derived Dietary Reference Values (DRVs) for iron. These include Average Requirement (AR) and Population Reference Intake (PRI). For adults, whole-body iron losses were modelled using data from US adults. Predicted absorption values, at a serum ferritin concentration of 30 µg/L, of 16 % for men and 18 % for women were used to convert physiological requirements to dietary iron intakes. In men, median whole-body iron losses are 0.95 mg/day, and the AR is 6 mg/day. The PRI, calculated as the dietary requirement at the 97.5th percentile, is 11 mg/day. For postmenopausal women, the same DRVs as for men are proposed. In premenopausal women, additional iron is lost through menstruation but, because losses are highly skewed, the Panel set a PRI of 16 mg/day to cover requirements of 95 % of the population. In infants and children, requirements were calculated factorially, taking into consideration the needs for growth, replacement of losses and percentage iron absorption from the diet (10 % up to 11 years and 16 % thereafter). PRIs were estimated using a coefficient of variation of 20 %. They are 11 mg/day in infants (7–11 months), 7 mg/day in children aged 1–6 years and 11 mg/day in children aged 7–11 years and boys aged 12–17 years. For girls aged 12–17 years, the PRI of 13 mg/day is the midpoint of the calculated dietary requirement of 97.5 % of girls and the PRI for premenopausal women; this approach allows for the large uncertainties in the rate and timing of pubertal growth and menarche. For pregnant and lactating women, for whom it was assumed that iron stores and enhanced absorption provide sufficient additional iron, DRVs are the same as for premenopausal women.


Following a request from the European Commission, the EFSA Panel on Dietetic Products, Nutrition and Allergies (NDA) was asked to deliver a Scientific Opinion on Dietary Reference Values (DRVs) for the European population, including iron. These include Average Requirement (AR) and Population Reference Intake (PRI).

Iron is required for oxygen transport, electron transfer, oxidase activities and energy metabolism. The main components of the body that contain iron are erythrocyte haemoglobin and muscle myoglobin, liver ferritin, and haem and non-haem enzymes.

Dietary iron consists of haem (from animal tissues) and non-haem (including ferritin) iron. Foods that contain relatively high concentrations of iron include meat, fish, cereals, beans, nuts, egg yolks, dark green vegetables, potatoes and fortified foods.

Iron is inefficiently and variably absorbed, depending on dietary and host-related factors. Iron absorption occurs primarily in the duodenum. A proportion of non-haem iron in foods is solubilised in the gastrointestinal lumen, reduced by duodenal cytochrome b reductase to Fe2+ and transported into the enterocyte by the transmembrane divalent metal transporter 1. There, iron is either stored as ferritin, some of which is subsequently lost when the cells are sloughed, is taken up by mitochondria for the synthesis of haem, or is transported across the basolateral membrane by ferroportin where it is carried in the circulation as diferric-transferrin after oxidation to Fe3+ by hephaestin. The mechanisms of absorption of haem iron and ferritin iron are uncertain, but once taken up iron is released from haem iron by haem oxygenase and then follows the same pathways as non-haem iron.

Homeostasis is mediated via the regulation of iron absorption, as there are no active pathways for excreting iron. In healthy individuals, the mucosal uptake and transfer of iron is inversely related to systemic serum ferritin concentrations, and control is exerted via the expression of the hepatic hormone hepcidin.

If the supply of iron is insufficient to meet physiological requirements, iron stores will be mobilised and iron deficiency will develop once the stores are exhausted. Iron deficiency anaemia (a microcytic anaemia with haemoglobin concentrations below normal) is the most common nutritional deficiency disorder, being found in all countries of the world. Subjects at greatest risk are those with high iron requirements owing to growth (infants, children, pregnant women) or high losses (women with high menstrual losses), or those with impaired absorption, e.g. in the presence of infection/inflammation.

The risk of systemic iron overload from dietary sources is negligible with normal intestinal function. Chronic iron overload may occur as a result of specific clinical conditions and genetic mutations, but there is no evidence that heterozygotes for haemochromatosis are at an increased risk of iron overload.

The Panel considers that health outcomes cannot be used to derive DRVs for iron because of the uncertainties in intake measurements, the poor correlation between intake and iron status, and the presence of confounders that prevent the determination of dose–response relationships and the assessment of risks associated with deficiency or excess.

A factorial approach was used to derive dietary iron requirements. Data on iron turnover and total obligatory iron losses from the body (including skin, sweat, urine and faeces) obtained from radioisotope dilution measurements were used to determine iron requirements in men and premenopausal women. Although these data were collected from a North American population group, the Panel agreed to use them as a basis for the estimation and probability modelling of the mean and approximate variability of distribution percentiles for the iron losses of adult men and premenopausal women in the European Union (EU) population. Summary statistics were estimated for the main variables related to iron losses for men and premenopausal women and for associations among the variables which were considered to be explanatory for iron losses. From these, a regression model equation for iron losses (as mg/day) was fitted to the data using a set of potentially relevant variables. This stage included an assessment of outliers and goodness of fit. The regression model was then used to derive a distribution for iron losses, combining the model equation with parametric distributions fitted to the sampling observations of each of the explanatory variables.

Dietary (haem and non-haem) iron absorption was estimated from a probability model, based on measures of iron intake and status in a representative group of men and women from the UK National Diet and Nutrition Survey. This provides estimates of total iron absorption from a mixed Western-style diet at any level of iron status. The Panel selected a target value of 30 µg/L for serum ferritin concentration. At this level, the predicted iron absorption is 16 % in men and 18 % in premenopausal women. The Panel decided to use 16 % for adults (except premenopausal women) and children aged 12–17 years when converting physiological requirements into dietary intakes, based on the assumption that the relationship between serum ferritin concentration and efficiency of absorption holds for all age groups, as there are no indications that age will affect the relationship.

In men, the 50th percentile of the model-based distribution of obligatory iron losses is 0.95 mg/day. The 90th, 95th and 97.5th percentiles are, respectively, equal to iron losses of 1.48, 1.61 and 1.72 mg/day. Using 16 % iron absorption to convert the physiological requirement into the dietary requirement results in a calculated dietary requirement at the 50th percentile of 5.9 mg/day and of 10.8 mg/day at the 97.5th percentile. After rounding, an AR of 6 mg/day and a PRI of 11 mg/day were set. In the absence of information on the iron requirement for postmenopausal women and despite their lower body weight, the Panel decided to set the same DRVs for postmenopausal women as those set for adult men.

In premenopausal women, the 50th percentile of the model-based distribution of obligatory iron losses is 1.34 mg/day. The 90th, 95th and 97.5th percentiles are, respectively, equal to iron losses of 2.44, 2.80 and 3.13 mg/day. Using 18 % absorption to convert the physiological iron requirement into the dietary requirement results in a calculated dietary requirement at the 50th percentile of 7.4 mg/day. Intakes meeting the dietary iron requirement of approximately 90, 95 and 97.5 % of the premenopausal women are calculated as 13.6, 15.6 and 17.4 mg/day, respectively. After rounding, the Panel derived an AR of 7 mg/day and a PRI of 16 mg/day for premenopausal women. The Panel considers that the PRI meets the dietary requirement of 95 % of women in their reproductive years and is derived from a group of premenopausal women, some of whom used oral contraceptives, as is the case in the EU. The Panel decided that women with very high iron losses should not be included in the premenopausal group, as this would result in unrealistically high DRVs for the majority of this population group.

In infants aged 7–11 months, the requirement for absorbed iron is 0.79 mg/day to replace obligatory losses (0.19 mg/day) and increase haemoglobin mass, tissue iron and storage iron (0.6 mg/day). Assuming 10 % absorption, this gives an AR of 8 mg/day and, based on a coefficient of variation (CV) of 20 %, which allows for high individual variation relating to growth rate, iron losses, absorption and dietary patterns, the PRI is 11 mg/day. In children aged 1–6 years, the AR is 5 mg/day, calculated from the sum of the requirements for growth (0.25 mg/day for ages 1–3 years and 0.27 mg/day for ages 4–6 years) and obligatory losses of 0.022 (1–3 years) and 0.012 (4–6 years) mg/kg body weight per day, and absorption of 10 %. Based on a CV of 20 %, the PRI is 7 mg/day. In children aged 7–11 years, requirements for growth increase to 0.39 mg/day, but losses per kilogram of body weight do not change. Assuming 10 % absorption, the AR (after rounding) is 8 mg/day and, based on a CV of 20 %, the PRI is 11 mg/day.

In boys and girls aged 12–17 years, the requirements for absorbed iron are 1.27 and 1.13 mg/day, respectively, calculated from losses of 0.012 mg/kg body weight per day and menstrual blood losses of 0.25 mg/day in girls, and growth needs of 0.61 mg/day for boys and 0.26 mg/day for girls. Assuming 16 % absorption, the AR (after rounding) is 8 mg/day for boys and 7 mg/day for girls. The PRI for boys is 11 mg/day based on a CV of 20 %. In girls, because of the uncertainties related to the rate and timing of physiological development and the onset of menarche, and because of the skewed distribution of menstrual losses, the Panel decided to set the PRI as the mean of the calculated dietary requirement of 97.5 % of girls aged 12–17 years (9.9 mg/day) and the PRI for premenopausal women (16 mg/day). After rounding, the PRI is 13 mg/day for girls.

In pregnancy, iron intake should cover basal losses during the first trimester, taking into account the cessation of menstruation. The requirements then increase exponentially, and this is associated with a dramatic increase in the efficiency of iron absorption. The total quantity of iron required for a singleton pregnancy is 835 mg. If the serum ferritin concentration is 30 µg/L at conception, around 120 mg of stored iron can be mobilised to support the pregnancy, which means that the total dietary requirement of iron is 715 mg. If the relevant percentage absorption figures determined from a study in pregnant women are applied to the entire pregnancy (7.2 % during weeks 0–23, 36.3 % during weeks 24–35 and 66.1 % during weeks 36–40 for non-haem iron, plus 25 % absorption for haem iron throughout the whole pregnancy), the total quantity of iron absorbed from a diet providing 13 mg iron/day is 866 mg. The Panel notes that using the absorption figures from single-meal studies in fasting mothers may be an overestimate, but, nevertheless, the quantity of iron absorbed is well in excess of the estimated 715 mg calculated by a factorial approach, and the progressive fall in serum ferritin concentration will be accompanied by an increased efficiency of absorption, irrespective of other homeostatic mechanisms. The Panel therefore considers that no additional iron is required in pregnancy.

During lactation, the quantity of iron secreted in breast milk is approximately 0.24 mg/day. When this is added to basal losses of 1.08 mg/day (obtained from data in postmenopausal women), the requirement for absorbed iron during the first months of lactation is calculated to be 1.3 mg/day, assuming that menstruation has not yet resumed. This requirement is slightly less than in non-pregnant, non-lactating women, but, for depleted iron stores to be replenished and to cover losses of iron when menstruation is re-established, the Panel considers that the AR and PRI for lactating women are the same as for non-pregnant women of childbearing age.

iron, Average Requirement, Dietary Reference Value, probabilistic modelling, factorial approach
Print on demand
Number of Pages