Six milligrams raised ferritin; microbiome harm was not detected
NICU360 Evidence Desk·8 min read·00
In 151 VLBW infants, 6 versus 2 mg/kg/day produced higher 4-week ferritin while prespecified microbiome measures and short-term clinical outcomes did not differ. A two-week, single-center biologic-endpoint trial is not a routine dosing rule.
THE HOT TAKE
Six milligrams per kilogram per day raised ferritin, and the trial did not detect the microbiome harm it was designed to find. That is a useful two-week result, not a routine dosing verdict.
At one level III NICU in Tampa, investigators enrolled 162 very-low-birth-weight infants. Eight failed screening, 154 were randomized, and 151 received study iron and entered the modified intention-to-treat analysis: 77 assigned to 2 mg/kg/day and 74 to 6 mg/kg/day of total enteral iron. The total included iron from feeds plus the blinded liquid ferrous sulfate intervention.
How the blinded trial moved from dose to outcomes
1
VLBW infant enrolled
Birth weight under 1500 g; key exclusions included prior NEC, intestinal perforation, existing iron treatment, and an erythropoiesis-stimulating agent.
2
Randomized around 2 weeks
Infants were at least 14 days old and receiving full feeds of about 150-160 mL/kg/day.
3
2 or 6 mg/kg/day
Total enteral iron from feeds plus once-daily blinded ferrous sulfate.
4
About 2 weeks after starting iron
Primary stool metagenomics and intestinal biomarkers; serum ferritin was measured at age 4 weeks.
5
Was this worth carrying into your next shift?
Your reaction stays connected to your NICU360 profile.
Comments
0
Join the conversation
Sign in to reply, vote, and keep your participation with your NICU360 profile.
Auditory brainstem response latency and short-term clinical safety were assessed.
THE PRIMARY ENDPOINT DID NOT SHOW A DOSE SIGNAL
The primary endpoint was change in stool metagenomic measures from before iron to about two weeks after starting it. Across bacterial diversity, individual taxa, virulence potential, bacterial overgrowth, and iron-related functional genes, the analyses found no statistically significant between-group differences. This was a failure to detect a difference, not an equivalence design and not proof of no later effect.
What the trial detected after the dose split
Measure
Finding
Interpretation
Prespecified microbiome measures
No statistically significant between-group differences
Primary result; does not establish equivalence
Urinary claudin-3
Higher after treatment in the 2 mg/kg/day arm
Possible barrier signal; no validated clinical threshold
Urinary iFABP and fecal calprotectin
No statistically significant treatment-arm effect
Secondary biomarkers
Clinical outcomes and auditory latency
No statistically significant between-group differences
Short follow-up and limited power for rare outcomes
A prespecified singleton-only exploratory analysis used 171 samples from 99 infants and found a difference in overall community trajectory over time (p=0.038). Because this was a subgroup trajectory analysis rather than a patient-level clinical endpoint, it is context, not a reversal of the primary result.
FERRITIN MOVED MORE CLEARLY
Median serum ferritin at age 4 weeks (mcg/L)
2 mg/kg/day
6 mg/kg/day
Median ferritin
In an adjusted model, lower-dose assignment was associated with 17.7% lower ferritin than higher-dose assignment (95% CI, 9.3%-25.4% lower). Each additional red-cell transfusion was associated with 13.7% higher ferritin (95% CI, 9.3%-18.7%). The transfusion effect is a reminder that the assigned supplement dose was not the only driver of iron stores.
SHORT-TERM SAFETY DID NOT SEPARATE
Investigators counted 423 adverse events: 210 in the low-dose arm and 213 in the high-dose arm. Ten serious events occurred in 7 infants, 5 low-dose and 2 high-dose. Severity, attribution, and major clinical adverse events did not differ by group. These are event counts, not infant-level risks; an infant could contribute more than one event.
Auditory brainstem responses at 36 weeks were available for 72 low-dose and 66 high-dose infants, with no difference in auditory latency. Weight gain, transfusions, phlebotomy volume, days without feeds, ventilation days, time to full feeds, and short-term clinical outcomes including NEC also did not differ. Rare harms and long-term neurodevelopment require a larger, longer trial.
THE REST OF THE EVIDENCE IS NOT A DOSE ANSWER
ESPGHAN's 2022 position paper generally places enteral iron for VLBW infants at 2-3 mg/kg/day from about 2 weeks, with doses up to 6 mg/kg/day in selected situations such as erythropoietin therapy or rapid ferritin repletion. This trial excluded infants already receiving an erythropoiesis-stimulating agent and tested 6 mg/kg/day as a fixed comparison, so it should not be read as a blanket guideline replacement.
A post hoc PENUT analysis of 692 extremely preterm infants found that each additional 50 mg/kg of cumulative enteral iron by day 60 was associated with a 0.77-point higher cognitive score at 2 years (95% CI, 0.06-1.48). Iron dose was not randomized in that analysis, and the day-90 association was not statistically significant. It makes long-term neurodevelopment relevant; it does not set an optimal dose.
THE PRACTICAL READ
The defensible shift is narrow: concern about two weeks of microbiome disruption alone is less persuasive as a reason to reject 6 mg/kg/day in every VLBW infant. The trial does not identify who needs more iron, which biomarker target best protects brain development, how transfusions or inflammation should alter interpretation, or how long higher dosing should continue.
This was a single-center trial, and its modified intention-to-treat analysis excluded 3 randomized infants who never received study iron. The main exposure window was about two weeks, ferritin had no pretreatment baseline, and clinically validated neonatal thresholds for claudin-3 and iFABP remain incomplete. ClinicalTrials.gov still lists the status as unknown, an estimated enrollment of 183, and no posted results, so the registry cannot reconcile the paper's categorical ferritin mismatch.
Six milligrams raised ferritin; microbiome harm was not detected · NICU360
