Iron Fortification and the Kenyan Infant Gut Microbiome

What the study examined

The work combined two linked, double-blind, randomized, placebo-controlled trials in Msambweni County, a malaria-endemic area of coastal Kenya where maize porridge (uji) is the staple weaning food. Infants weaned at about six months of age consumed home-fortified porridge every day for four months. In each trial they received either an iron-containing micronutrient powder (MNP) or an otherwise identical MNP without iron, so the effect of iron itself could be isolated from the rest of the vitamin-and-mineral mix.

Two iron forms and doses were tested: 2.5 mg of iron as sodium iron EDTA (NaFeEDTA), a chelated ferric form, and 12.5 mg of iron as ferrous fumarate. Of 160 infants screened, 115 were enrolled and 101 completed the per-protocol analysis (28 on +2.5 mg, 21 on −2.5 mg, 26 on +12.5 mg, 26 on −12.5 mg); antibiotic use during the trial was unexpectedly high and accounts for much of the attrition. Compliance with the daily sachets was excellent (99.4% and 96.2%).

At baseline the cohort carried a heavy burden of deficiency and infection: 67.3% were anemic, 25.5% iron-deficient and 29.7% had systemic inflammation; 99% were still breastfed. The maize flour itself supplied little bioavailable iron (1.15 mg per 100 g) but was rich in the absorption inhibitor phytic acid (310 mg per 100 g), so more than 80% of any added iron passes unabsorbed into the colon. The primary outcome was gut microbiome composition, profiled by both 16S rRNA pyrosequencing and targeted quantitative PCR (qPCR); secondary outcomes were fecal calprotectin (a marker of intestinal inflammation) and diarrhea incidence.

A bifidobacteria-rich gut already carrying pathogens

At baseline the infants had the bifidobacteria-dominated microbiome typical of breastfed babies — Bifidobacteriaceae made up about 63% of all 16S rRNA reads — a composition usually considered protective against enteric infection. But beneath that dominance, the same guts already harbored a striking load of opportunistic pathogens.

Across the analyzed stool samples the team detected enteropathogenic E. coli in 65% of infants, and enterotoxigenic and enterohemorrhagic E. coli in smaller fractions; Clostridioides difficile in 56.5%, the Clostridium perfringens group in 89.7%, Staphylococcus aureus in 65.4%, Bacillus cereus in 39.5% and Salmonella in 22.4%. In other words, the community had abundant fuel for a pathogenic bloom if the ecological balance tipped — which is exactly what added iron did.

What iron did to the microbiome

By 16S pyrosequencing, iron-containing MNPs significantly increased enterobacteria — particularly the potentially pathogenic Escherichia/Shigella group (p=0.048) — and Clostridium (p=0.030), while trending toward lower Bifidobacterium. The enterobacteria-to-bifidobacteria ratio, a summary index of the shift away from a protective community, rose significantly with iron (change over time p=0.020; and was higher in the iron groups at endpoint, p=0.004).

Targeted qPCR confirmed the pattern with hard counts. Iron significantly raised the summed abundance of pathogenic E. coli (6.0 vs 4.5 log copies per gram of feces; p=0.029) and the enterobacteria-to-bifidobacteria ratio (p=0.008), while significantly lowering the beneficial butyrate-producing Roseburia / Eubacterium rectale group (p=0.020). The higher 12.5 mg ferrous fumarate dose tended to cause the larger disturbance, including a significantly greater fall in Bifidobacterium (p=0.049), though the lower 2.5 mg NaFeEDTA dose also shifted the community.

Interestingly, iron did not change overall phylogenetic diversity or fecal short-chain fatty acid concentrations — but the authors note that 95–99% of colonic SCFAs are absorbed, so fecal levels may not reflect what happens in the lumen.

Inflammation, diarrhea — and who was affected

The microbial shift was mirrored in the host. Fecal calprotectin, a neutrophil-derived marker of intestinal inflammation, was significantly higher in infants receiving iron (229 vs 123 µg/g; p=0.002), with the strongest effect at the 12.5 mg ferrous fumarate dose (248.9 vs 102.5 µg/g; p=0.008). Diarrhea requiring treatment was more than three times as common on the higher iron dose — 27.3% (6 of 22) versus 8.3% (2 of 24) in the matched no-iron group — a difference that trended but did not reach significance in this modest sample (p=0.092).

A revealing subtlety is who bore the harm. The rise in pathogenic E. coli was significant in infants who were iron-deficient at baseline (p=0.012) but not in iron-sufficient infants (p=0.327). The rise in intestinal inflammation, by contrast, was significant in infants who were already iron-sufficient (p=0.0002) but not in those who were iron-deficient (p=0.912). Together these point to the same clinical message from two directions: blanket iron given to a population that includes iron-replete infants provokes gut inflammation in exactly the children who do not need the iron.

For balance, iron was not all cost: it produced a small but significant gain in linear growth (70.2 vs 68.5 cm at endpoint; p=0.017), and there were no study-related serious adverse events. The paper is careful not to argue against treating genuine iron-deficiency anemia.

The mechanism: unabsorbed iron feeds the wrong microbes

Iron absorption in the small intestine is limited, and in the phytate-rich, frequently inflamed guts of this cohort — where the iron-regulatory hormone hepcidin further suppresses uptake — the great majority of fortificant iron reaches the colon. There it becomes a contested resource. This is nutritional immunity run in reverse: the healthy gut normally withholds free iron to restrain pathogens, and pouring bioavailable iron into the lumen relaxes that restraint.

The winners are the microbes best equipped to grab iron. Enteric pathogens such as E. coli, Salmonella and Shigella deploy high-affinity siderophores and ferrous-iron transporters (for example FeoB) whose acquisition machinery is directly tied to virulence and gut colonization. The 'barrier' commensals are comparatively iron-frugal: lactobacilli famously do not require iron at all (relying on manganese), and bifidobacteria take up iron through a divalent permease but make no siderophores. So added luminal iron systematically tilts the competition toward the opportunists and away from the protective flora — the enterobacterial bloom, bifidobacterial loss and rise in inflammation the trial recorded.

The authors also raise the possibility that host inflammatory responses generate electron acceptors that favor facultative anaerobes such as enterobacteria, and that iron drives inflammation through neutrophil infiltration, lipid peroxidation and NF-κB activation — a self-reinforcing loop between iron, pathogens and gut inflammation.

How it fits the metal-microbiome-disease axis

This trial is one of the clearest human demonstrations of the first two links of the metal-microbiome-disease axis operating together inside a single controlled experiment: a defined change in metal exposure (iron intake) drives a defined change in the microbiome (an enterobacterial bloom and loss of bifidobacteria), and that dysbiosis is coupled to a disease-relevant readout (intestinal inflammation, with a signal toward diarrhea).

It is instructive precisely because iron is an essential nutrient, not a toxic heavy metal. The axis here is about metal availability and microbial competition for it, not about poisoning — the same siderophore-mediated scramble a fortificant dose exploits is the mechanism pathogens use to overcome host metal-withholding during infection. That places this study alongside the site's work on how dietary metals reshape the infant gut, from iron to nickel.

The practical legacy has been to shift iron delivery from 'give it to everyone' toward targeting infants with true iron-deficiency anemia, using lower doses and gentler compounds such as NaFeEDTA, and pairing iron with prebiotics (as later Kenyan trials tested) to blunt the microbiome cost — while still protecting against malaria and diarrhea. For the broader thesis of this site, it is a rigorous proof of concept that what reaches the colon in metal form can reorganize microbial ecology and inflame the gut.