Scientific opinion on the tolerable upper intake level for iron

Abstract

Following a request from the European Commission, the EFSA Panel on Nutrition, Novel Foods and Food Allergens (NDA) was asked to deliver a scientific opinion on the tolerable upper intake level (UL) for iron. Systematic reviews were conducted to identify evidence regarding high iron intakes and risk of chronic diseases, adverse gastrointestinal effects and adverse effects of iron supplementation in infancy, young childhood and pregnancy. It is established that systemic iron overload leads to organ toxicity, but no UL could be established. The only indicator for which a dose-response could be established was black stools, which reflect the presence of large amounts of unabsorbed iron in the gut. This is a conservative endpoint among the chain of events that may lead to systemic iron overload but is not adverse per se. Based on interventions in which black stools did not occur at supplemental iron intakes of 20-25 mg/day (added to a background intake of 15 mg/day), a safe level of intake for iron of 40 mg/day for adults (including pregnant and lactating women) was established. Using allometric scaling (body weight^0.75), this value was scaled down to children and adolescents and safe levels of intakes between 10 mg/day (1-3 years) and 35 mg/day (15-17 years) were derived. For infants 7-11 months of age who have a higher iron requirement than young children, allometric scaling was applied to the supplemental iron intakes (i.e. 25 mg/day) and resulted in a safe level of supplemental iron intake of 5 mg/day. This value was extended to 4-6 month-old infants and refers to iron intakes from fortified foods and food supplements, not from infant and follow-on formulae. The application of the safe level of intake is more limited than a UL because the intake level at which the risk of adverse effects starts to increase is not defined.

Keywords: adverse effects; iron homeostasis; iron overload; safe level of intake.

FIGURE 1

Stepwise approach for evidence integration and uncertainty analysis. BoE, body of evidence; LoE, line of evidence.

FIGURE 2

Approach applied to assign the final level of certainty in a causal relationship. Adapted from OHAT/NTP (2019). LoE, line of evidence; NCC, nested case‐control; PC, prospective cohort; RCT, randomised controlled trial; RoB, risk of bias. ^aAs an example, a 'high level of certainty' means that, based on the available evidence, experts are 75 to 100% certain that iron intake is positively and causally associated with the adverse health outcome of interest.

FIGURE 3

Distribution of iron content in food supplements as displayed on labels in EU Member States and Norway (mg/serving). Source: Mintel GNPD. Search for iron‐containing supplements available in the EU market from December 2017 to December 2022. A total of 1055 products available in 24 EU Member States and Norway were identified, of which 754 contained complete data on mg iron/serving.

FIGURE 4

Mean, median, 5th and 95th percentiles of iron intakes in infants (≥ 4 to < 12 months old), toddlers (≥ 1 year to < 3 years old), other children (≥ 3 years to < 10 years old) and adolescents (≥ 10 years to < 18 years old), by sex and country. Estimates for females in orange and for males in blue. Squares correspond to medians and stars to means. Lines represent the range between the 5th and 95th percentiles. Estimated intakes from 5th and 95th percentiles are not presented when sample size is below 60 participants. DE, Germany; FI, Finland; FR, France; IT, Italy; LV, Latvia; NL, The Netherlands. Source: (EFSA NDA Panel, 2015) except for infants.

FIGURE 5

Mean, median, 5th and 95th percentiles of iron intakes in adults (≥ 18 to < 65 years old) and elderly as well as very elderly (≥ 65 to < 75 years old, and ≥ 75 years old), by sex and country. Estimates for females in orange and for males in blue. Squares correspond to medians and stars to means. Lines represent the range between the 5th and 95th percentiles. Estimated intakes from 5th and 95th percentiles are not presented when sample size is below 60 participants. FI, Finland; FR, France; IE, Ireland; IT, Italy; LV, Latvia; NL, The Netherlands; SE, Sweden. Source: (EFSA NDA Panel, 2015).

FIGURE 6

Prospective cohort studies investigating the association between total iron intake and the risk of T2DM. *Exposure = cumulative iron intakes. Intake estimates provided by Jung et al. (2021) (KoGES) as ratio of total‐iron‐to‐total‐energy intake were recalculated by EFSA. appr., approximately; av., average; CHNS, China Health and Nutrition Survey; CI, confidence interval; CN, China; F, females; GNHS, Guangzhou Nutrition and Health Study; HR, hazard ratio; HPFS, Health Professionals' Follow‐up Study; JACC, Japan Collaborative Cohort Study for Evaluation of Cancer Risk; JP, Japan; KoGES, Korean Genome and Epidemiology Study; KR, South Korea; M, males; max, maximum; med., median; NHS, Nurses' Health Study; OR, odds ratio, PE, point estimate; Q, quantile; RR, risk ratio; US, United States; WHS, Women's Health Study; y, years.

FIGURE 7

Intervention studies investigating the effect of iron supplements on adverse GI effects using structured tools for their assessment. *For cross‐over studies, no effect estimates could be calculated. In addition, Bries et al. (2019) did not report on the percentage of individuals with at least one symptom. It is, however, reported that consuming ferrous sulfate resulted in a tendency for a higher incidence of constipation, diarrhoea, nausea and abdominal discomfort. **Based on data provided by the authors of the study to EFSA. All OR estimated by EFSA.

FIGURE 8

Intervention studies investigating the effect of iron supplementation on infant growth. (A) Weight gain per day in the study period. (B) Weight at baseline and the end of the study either as absolute weight or WAZ and length gain. Median weight and length gains of infants within the studied age range and averaged over the studied time period from the WHO growth reference standard are also given. For the study by Majumdar et al. (2003), SE for the individual age groups were imputed using the SE for the combined age groups. Weight and length gains of individual age groups were read from graphs. For Gahagan et al. (2009), length gains were read from a graph. For Ziegler et al. (2009) and Gahagan et al. (2009), the doses of iron supplements expressed as mg/kg bw were calculated based on the absolute weights of infants at the beginning as well as at the end of the studies. Base, baseline; B‐l, baseline length; B‐w, baseline weight; bw, body weight; CI, confidence interval; CL, Chile; Dur, duration; E‐w, end weight; E‐l, end length; F, females; G1, intervention group; G2, control group; ID, Indonesia; IN, India; Lg, length gain; Lv, length velocity; M, males; m, months; MD, mean difference; N, number; SE, Sweden; suppl.; supplement; US, United States of America; WHO, World Health Organization; wv, weight velocity; z‐s, z‐scores.

FIGURE 9

Intervention studies investigating the effect of iron supplementation during pregnancy on birth weight of the offspring. CI, confidence interval; G1, intervention group; G2, control group; GA, gestational age; MD, mean difference; N, number; PE, point estimate.