Heme iron: de et al., 2026
An 18-week crossover RCT in 27 older adults (60+) found that a pulse-protein diet providing only non-heme iron maintained ferritin and hemoglobin as well as a meat-protein diet, despite lower methionine intake. This is a small but well-controlled trial showing non-heme iron can support iron status when absorption factors are optimized.
Key takeaways
- Both diets maintained hemoglobin and ferritin over 18 weeks — no clinically meaningful difference between meat-based (heme + non-heme iron) and pulse-based (non-heme only) protein sources.
- The pulse diet provided 14.6 mg/day iron (all non-heme) vs. 17.4 mg/day in the meat diet (mixed heme/non-heme). Serum ferritin dropped slightly on pulses (−3.5 µg/L) but remained within normal range.
- Methionine intake was 40% lower on pulses. Lower methionine tracked with lower ferritin and higher FGF21, suggesting a metabolic signal linking amino acid intake to iron regulation.
- This is a controlled-feeding crossover trial — participants ate only what researchers provided. That removes real-world variability but makes the findings clean and interpretable.
- Sample size of 27 completers is small for a crossover study. The study was powered for protein outcomes, not iron endpoints — these iron findings are exploratory, not pre-specified.
The study
de Vargas et al., Frontiers in Nutrition, 2026. PubMed link. This was a secondary analysis of PRODMED1, an 18-week randomized crossover controlled-feeding trial in 27 adults aged 60+ (mean 67 years, 63% women, BMI 20-35). Participants ate two isocaloric diets for 8 weeks each, separated by a 2-week washout: a meat-protein diet (MPD) with beef, pork, poultry, and a pulse-protein diet (PPD) with beans, lentils, chickpeas. Both diets aligned with DGA macros: 15% protein, 30% fat, 55% carbs. Researchers measured hemoglobin, serum ferritin, iron intake, methionine, homocysteine, cystine, and FGF21 (a metabolic hormone) at baseline and end of each diet period. All food was provided; participants ate nothing else.
How to read this study
What this paper does well: The crossover design is a strength — each person serves as their own control, which removes between-person variability in iron absorption genetics (some people absorb non-heme iron well, others don't). Controlled feeding is the gold standard for diet studies: no self-report error, no compliance guessing. The 8-week diet periods are long enough to see ferritin changes (ferritin reflects 2-3 months of iron status). They measured multiple iron markers — hemoglobin (functional iron), ferritin (stored iron), and dietary iron intake — not just one surrogate.
What this paper is missing or weak on: The sample size of 27 is small, and the study was originally powered for protein metabolism, not iron outcomes. The authors call these findings "exploratory," which means they didn't pre-register iron as a primary endpoint — that raises the risk of p-hacking (testing many outcomes, reporting the interesting one). They didn't measure hepcidin, the master iron regulator, or soluble transferrin receptor, which would clarify whether iron absorption was upregulated on the pulse diet. The methionine-FGF21-ferritin link is associational, not causal — they saw a correlation but didn't test whether methionine restriction *causes* the ferritin drop.
How I'd weight this paper: I treat this as hypothesis-generating, not definitive. The crossover design and controlled feeding give me confidence the effect is real within this sample, but the small size and exploratory nature mean I wouldn't base a recommendation on this alone. If future pre-registered trials replicate the methionine-ferritin link, that becomes interesting. For now, it's a well-executed pilot that shows non-heme iron *can* maintain status in older adults when intake is adequate and diet quality is high.
What they found
Hemoglobin stayed stable on both diets: 14.0 g/dL at baseline, 14.1 g/dL after meat, 14.0 g/dL after pulses (p = 0.60). Serum ferritin dropped 3.5 µg/L on pulses vs. rising 1.4 µg/L on meat (p = 0.03 for the difference), but both remained in normal range (meat: 89 µg/L, pulses: 84 µg/L). Iron intake was 17.4 mg/day on meat (mixed heme/non-heme) and 14.6 mg/day on pulses (all non-heme). Methionine intake was 1.8 g/day on meat vs. 1.1 g/day on pulses (p < 0.001). Serum methionine, cystine, and homocysteine all rose significantly on meat (p < 0.001 for all three) and dropped on pulses. FGF21, a hormone that rises during methionine restriction, increased 46% on pulses (p = 0.02). In regression models, lower methionine and higher FGF21 both associated with lower ferritin (p < 0.05), independent of dietary iron and vitamin C intake.
What it means for the average man
If you're eating plant-based protein or shifting away from meat, your body can maintain iron status on non-heme iron alone — but you need to get the rest of the diet right. The pulse diet in this study provided 14.6 mg/day iron (above the RDA of 8 mg for men) and included vitamin C-rich foods at meals, which boosts non-heme absorption 3-4×. The small ferritin drop on pulses is statistically significant but not clinically worrying — it stayed well above deficiency. The methionine finding is intriguing: lower methionine intake tracked with slightly lower ferritin and higher FGF21, a longevity-associated hormone in animal models. We don't know yet if that's good, bad, or neutral for long-term health. Practical takeaway: if you're eating beans instead of beef, pair them with citrus, tomatoes, or peppers, and track your ferritin annually.
The caveats
This is a secondary analysis of a trial designed to study protein, not iron — the authors didn't pre-register iron as an outcome, so there's selection bias risk. The sample of 27 is small; larger trials are needed to confirm. Controlled feeding means participants ate researcher-provided meals for 18 weeks — real-world adherence to a pulse-based diet is different. The study excluded people with anemia or chronic disease, so we don't know if these findings hold in iron-deficient individuals or those with inflammation (which raises ferritin). The methionine-ferritin link is associational; we can't say methionine restriction *caused* the ferritin change. Finally, they didn't measure iron absorption directly (e.g., via isotope tracing), so we're inferring mechanisms from blood markers.
Frequently asked questions
Should I trust an exploratory analysis?
Exploratory means the iron outcomes weren't pre-specified — the authors ran the trial for protein metabolism, then looked at iron data afterward. That increases the risk of finding a "significant" result by chance (p-hacking). But crossover controlled-feeding trials are high-quality designs, and the effect is biologically plausible. I treat this as a strong signal worth replicating, not as proof to act on alone.
What's the difference between heme and non-heme iron absorption?
Heme iron (from meat) is absorbed at 15-35% regardless of other foods. Non-heme iron (from plants) is absorbed at 2-20% and is highly sensitive to enhancers (vitamin C, meat) and inhibitors (phytates, tannins). This study shows non-heme iron can maintain status when intake is adequate and the diet includes absorption enhancers — the pulse diet had both.
Is lower methionine intake a good thing?
In rodents, methionine restriction extends lifespan and raises FGF21, which improves metabolic health. In humans, we don't know yet. This study found lower methionine tracked with slightly lower ferritin and higher FGF21 — whether that trade-off benefits long-term health is speculative. Methionine is essential; severe restriction is harmful. The pulse diet here provided 1.1 g/day, which is above the RDA of 0.9 g/day.
Does this mean I should switch to plant-based protein?
Not necessarily. This study shows plant protein *can* support iron homeostasis when intake and absorption factors are optimized. It doesn't show it's superior to meat. If you digest and tolerate meat well, there's no iron-based reason to switch. If you're reducing meat for other reasons, this study suggests you can maintain iron status on pulses with attention to total intake and meal composition.
Sources
- de Vargas B., et al. A DGA-aligned pulse-protein diet may support iron homeostasis in older adults: an exploratory analysis of the PRODMED1 trial. Front Nutr. 2026. PubMed.
- Hurrell R., Egli I. Iron bioavailability and dietary reference values. Am J Clin Nutr. 2010;91(5):1461S-1467S.
These statements have not been evaluated by the Food and Drug Administration. This product is not intended to diagnose, treat, cure, or prevent any disease.
