Millet: the grain with real trial evidence on iron-poor diets

A meta-analysis of 13 randomized trials with 1,139 people, mostly children and adolescents, found that millet of mixed types, sorghum included, raised hemoglobin by 0.68 standardized mean difference units compared with refined staples. A second meta-analysis pooled 8 feeding trials in children in Africa and Asia. Millet diets in place of rice improved height gain by 28.2% and weight gain by 26% over 3 months to 4.5 years, though the trials disagreed so sharply (I2 of 94%) that the size is poorly pinned down. The pooled height estimate had a confidence interval running from 0.23 to 5.62. So the direction is clear for children whose staple is rice, and the amount is not. Millet is a grain, and a cup of cooked millet weighs 6 oz (170 g). Eight US samples of whole grain millet averaged 2.5 mg of iron per 3.5 oz (100 g), with samples ranging from 1.3 to 3.3 mg.

The numbers, per 3.5 oz (100 g)

As people eat it, per 3.5 oz (100 g)

How much is actually in it

• US lab samples of whole-grain millet average 2.5 mg of iron per 3.5 oz (100 g), but samples range from 1.3 to 3.3 mg. — USDA Foundation Foods, 8 samples (n = 8 samples) Evidence E source
  Iron 2.534 mg (1.34-3.28); zinc 2.258 mg (1.52-3.03); magnesium 105.5 mg (58.0-126.0)
• UK tables list foxtail millet flour at 0.68 mg thiamin per 3.5 oz (100 g), above the US whole-grain millet range of 0.19 to 0.55 mg. — CoFID 2021 vs USDA Foundation Foods (n = 1 entry + 8 samples) Evidence E source
  CoFID 0.68 mg vs USDA 0.411 mg (0.19-0.55)

What a real portion looks like

• US dietary survey coding counts a cup of cooked millet as 170 g, and 1 oz of dry millet yields 110 g cooked. — USDA FNDDS portion measures (n = 4 measures) Evidence E source
  1 cup cooked 170 g; 1 oz dry yields 110 g; 1 cup dry yields 760 g

What your body absorbs

• In 20 Beninese women, iron-biofortified pearl millet meals doubled the iron absorbed, 1125 vs 527 ug, although the absorbed fraction stayed at 7.5% in both. — Beninese women with marginal iron status, multiple meals over 5 days (n = 20) Evidence B source
  Fractional absorption 7.5% vs 7.5%; total absorbed 1125 vs 527 ug (P < 0.0001); post-harvest fortified 10.4%
• In 40 two-year-olds in India, biofortified pearl millet meals delivered 0.67 mg of absorbed iron a day versus 0.23 mg from ordinary millet. — children aged 2 y, Karnataka, India, about 84 g dry millet flour a day (n = 40) Evidence B source
  Iron absorbed 0.67 vs 0.23 mg/d (P < 0.001); zinc 0.95 vs 0.67 mg/d (P = 0.03)
• Lab studies pooled in a millet iron review show fermentation and germination raise the iron the body can take up from millet. — 8 in vitro studies within a meta-analysis of 30 studies (n = 8 in vitro studies) Evidence D source
  Bioavailable iron significantly higher after fermentation or germination (p < 0.01); size not in abstract

What cooking and storage change

• In a crossover RCT in 18 women, millet cooked and chilled for 24 h gave a 29.5% smaller glucose rise than fresh-cooked rice, while fresh-cooked millet cut the rise after the next meal by 39.1% against fresh rice. — healthy young women, 50 g carbohydrate test meals (n = 18) Evidence B source
  Cold-stored millet iAUC -29.5% vs fresh rice; fresh millet second-meal iAUC -39.1%
• In a crossover trial in 14 healthy volunteers in Mali, traditional sorghum and millet porridges and millet couscous took about 2 times as long to leave the stomach as rice, potato or pasta. — healthy volunteers in Bamako, 14 in the main study plus 6 in validation (n = 14) Evidence B source
  Gastric half-emptying time about 2x longer (p < 0.0001)

What it does to your health

• A meta-analysis of 13 RCTs with 1,139 people found eating millet raised blood hemoglobin by 0.68 standardized mean difference units compared with refined staples. — 590 intervention and 549 control participants, mostly children and adolescents (n = 1139) Evidence A source
  SMD +0.68 (p < 0.01); hemoglobin +13.6% vs +4.8% in controls
• A meta-analysis of 8 randomized feeding trials in children found millet-based diets in place of rice improved height gain by 28.2% and weight gain by 26% over 3 months to 4.5 years. — children in Africa and Asia, rice-based control diets (n = 8 trials) Evidence A source
  Height +28.2%, weight +26%, MUAC +39%, chest circumference +37% vs rice diets

Safety, limits and interactions

• In a Canadian market survey, millet flour not labeled gluten-free averaged 272 mg of gluten per kg, far above the 20 mg per kg gluten-free limit. — 640 naturally gluten-free flour and starch samples, 8 Canadian cities, 2010-2012 (n = 640) Evidence C source
  Millet composite mean 272 mg/kg; 9.5% of all samples above 20 mg/kg
• A 2024 systematic review of nine studies found the claim that pearl millet causes goiter does not hold up, since only three were in humans and none had a control diet. — 9 studies, 8 on pearl millet and 1 on fonio; 3 human studies from Sudan (n = 9 studies) Evidence C source
  No controlled human evidence; animal studies used raw 100% millet diets

What people believe that the data does not support

• The EU did not authorize a claim that a millet extract mix supports healthy hair growth. — EU Register of nutrition and health claims (n = 1 claim) Evidence E source
  Status Non-authorized

Who this works differently for

• In a 9-month RCT in 223 Mumbai toddlers, iron- and zinc-biofortified pearl millet raised hemoglobin slightly and cut anemia to 33.8% versus 57.6%, without changing ferritin, zinc or growth. — children 12-18 months, urban slums of Mumbai, 9 months (n = 223) Evidence B source
  Hb change +0.10 vs -0.15 g/dL (P = 0.04); anemia 33.8% vs 57.6% (P = 0.005)
• In an RCT in 75 Ghanaian mothers, a millet-soya-ginger drink raised breast milk volume by 1.7 tbsp (24.9 mL) at day 14 versus controls. — postpartum mothers, Tamale, Ghana, 14 days (n = 75) Evidence B source
  +1.7 tbsp (24.9 mL) milk (95% CI 22.1-27.7); prolactin +25.0 ng/mL (95% CI 22.0-28.0)

What is still unknown

• The pooled height effect in that meta-analysis had very high heterogeneity, I2 of 94%, with a standardized mean difference of 2.92 and a confidence interval from 0.23 to 5.62. — children in 8 randomized feeding trials (n = 8 trials) Evidence A source
  SMD 2.92 (95% CI 0.23-5.62), I2 = 94%
• A 2026 systematic review of millet and blood sugar in adults found only one parallel RCT, three crossover trials and two non-randomized studies. — adults with or without type 2 diabetes, searches to January 2025 (n = 6 studies) Evidence B source
  No pooled estimate; qualitative synthesis

The bottom line

Iron-biofortified pearl millet delivers more iron to the body. In 20 Beninese women with marginal iron status, meals made from it doubled the iron absorbed, 1125 against 527 ug, while the absorbed fraction stayed at 7.5% for both. In 40 two-year-olds in India eating about 3 oz (84 g) of millet flour a day, it gave 0.67 mg of absorbed iron a day against 0.23 mg from ordinary millet, measured over one or two test days. In a 9-month trial in 223 toddlers in Mumbai slums, iron and zinc biofortified millet moved hemoglobin by +0.10 g/dL against -0.15 in the control group, without changing ferritin, zinc or growth. Anemia at the end was 33.8% against 57.6%, a difference the trial was not designed to detect. That paper is a 2026 republication after the authors retracted their 2022 version to reanalyse hemoglobin. Lab studies suggest fermentation and germination raise the iron available from millet, and the review gives no size. In a crossover trial in 18 healthy young women, millet cooked and chilled for 24 hours gave a 29.5% smaller glucose rise than fresh rice. Fresh millet cut the rise after the next meal by 39.1% against fresh rice. Both were short tests in one group. In Mali, traditional sorghum and millet porridges and millet couscous took about twice as long to leave the stomach as rice, potato or pasta in 14 healthy volunteers, and how full people felt did not line up with it. For adults and blood sugar in general, a 2026 systematic review found only one parallel randomized trial, three crossover trials and two non-randomized studies, and gave no pooled estimate. In 75 Ghanaian mothers, an unblinded trial of a drink made of 74% millet with soya and ginger raised breast milk volume by 1.7 tbsp (95% CI 1.5 to 1.9; metric: 24.9 mL, 22.1 to 27.7) at day 14 against controls. Millet has no gluten of its own, and that is exactly where people with celiac disease get caught. In a Canadian survey of 640 naturally gluten-free flours, millet flour not labeled gluten-free averaged 272 mg of gluten per kg, far above the 20 mg per kg limit. The old warning that pearl millet causes goiter rests on weak ground. A 2024 systematic review found nine studies, only three of them in humans and none with a control diet, and the animal studies fed raw millet as the whole diet. A controlled human test does not exist yet. We found no study of millet in pregnancy or alongside medicines. The EU refused a hair growth claim for a mix of millet extract, L-cystine and vitamin B5.

Sources

1. mlt-daily-r4-15 · composition dataset
   Millet, whole grain: Iron, Fe 2.534 mg, range 1.34-3.28, 8 samples; Zinc, Zn 2.258 mg, range 1.52-3.03, 8 samples; Magnesium, Mg 105.5 mg, range 58.0-126.0, 8 samples
   USDA FoodData Central, Foundation Foods, release 2026-04-30, Millet, whole grain (FDC 2512379) · checked 2026-09-29
2. mlt-daily-r4-16 · composition dataset
   Flour, millet, foxtail, CoFID 11-016: Thiamin 0.68 mg, Protein 5.8 g, Energy 354 kcal; Millet, whole grain (USDA Foundation): Thiamin 0.4113 mg, range 0.19-0.55, 8 samples
   CoFID 2021 (Public Health England), code 11-016; USDA FoodData Central Foundation 2026-04-30 · checked 2026-09-29
3. mlt-daily-r4-17 · portion dataset
   Millet, FNDDS code 56202000, fdc_id 2708377: 1 cup, cooked, measure 10043, 170 g; 1 cup, dry, yields, measure 10074, 760 g; 1 oz, dry, yields, measure 40049, 110 g; Quantity not specified, measure 90000, 170 g
   USDA FNDDS, food code 56202000, FDC 2708377 · checked 2026-09-29
4. mlt-daily-r4-04 · randomized crossover trial
   Fractional iron absorption from test meals based on regular-iron millet (7.5%) did not differ from iron-biofortified millet meals (7.5%; P = 1.0), resulting in a higher quantity of total iron absorbed from the meals based on iron-biofortified millet (1125 vs. 527 μg; P < 0.0001).
   J Nutr, 2013 · checked 2026-09-29
5. mlt-daily-r4-05 · randomized controlled trial
   The mean (± SD) quantities of iron absorbed from test and control groups were 0.67 ± 0.48 and 0.23 ± 0.15 mg/d, respectively (P < 0.001). The quantities of zinc absorbed were 0.95 ± 0.47 and 0.67 ± 0.24 mg/d, respectively (P = 0.03).
   J Nutr, 2013 · checked 2026-09-29
6. mlt-daily-r4-10 · in vitro
   Eight in-vitro studies showed that traditional processing methods such as fermentation and germination can improve bioavailable iron significantly (p < 0.01)
   Front Nutr, 2021 · checked 2026-09-29
7. mlt-daily-r4-06 · randomized crossover trial
   Cold-stored non-glutinous rice (CR) and cold-stored non-glutinous millet (CM) had a 24.4% and 29.5% lower incremental area under the curve (iAUCglu) of glucose within 240 min compared to the control (fresh-cooked rice non-glutinous, FR), respectively (p < 0.05). ... After a standard lunch, the fresh-cooked non-glutinous millet (FM) achieved a 39.1% lower iAUCglu0-180 compared to the FR (p < 0.05).
   Nutrients, 2024 · checked 2026-09-29
8. mlt-daily-r4-07 · randomized crossover trial
   In both trials, traditional sorghum and millet foods (thick porridges and millet couscous) had gastric half-emptying times about twice as long as rice, potato, or pasta (p < 0.0001).
   Nutrients, 2018 · checked 2026-09-29
9. mlt-daily-r4-01 · meta-analysis
   Thirteen published studies featuring randomized control trials involving 590 individuals in the intervention group and 549 control individuals were eligible for the meta-analysis. The difference-in-differences analysis revealed highly significant (p < 0.01) positive effects of millet consumption on hemoglobin concentration, with an effect size of +0.68 standardized mean difference units.
   Front Nutr, 2024 · checked 2026-09-29
10. mlt-daily-r4-02 · meta-analysis
    Eight eligible randomized feeding trials were included in the meta-analysis. Results from the randomized effect model showed a significant effect (p < 0.05) of millet-based diets on mean height (+28.2%) (n = 8), weight (n = 9) (+26%), mid upper arm circumference (n = 5) (+39%) and chest circumference (n = 5) (+37%) in comparison to regular rice-based diets over for the period of 3 months to 4.5 years, which was based on largely substituting rice with millets.
    Nutrients, 2022 · checked 2026-09-29
11. mlt-daily-r4-11 · contamination survey
    Looking at the total mean (composite) contamination for specific ingredients the largest and most consistent contaminations come from higher fibre ingredients such as soy (902 mg kg⁻¹), millet (272 mg kg⁻¹) and buckwheat (153 mg kg⁻¹).
    Food Addit Contam Part A Chem Anal Control Expo Risk Assess, 2013 · checked 2026-09-29
12. mlt-daily-r4-12 · systematic review
    Existing evidence to support the claim that pearl millet consumption causes goiter does not stand up to scientific rigor. This is so for several reasons: no control groups and intervention groups were fed with the treatment diet (pearl millet or non-millet) in any of the human studies
    Front Nutr, 2024 · checked 2026-09-29
13. mlt-daily-r4-14 · regulatory decision
    POL-HC-6967 Combination of millet extract + L-cystine + vitamin B5 (calcium panthothenate) Supports healthy hair growth from within. Supports the growth of healthy hair & nails. Supports metabolism of the hair roots. Non-authorised
    EU Register on nutrition and health claims, claim POL-HC-6967, snapshot 2026-09-21 · checked 2026-09-29
14. mlt-daily-r4-08 · randomized controlled trial
    Consumption of FeZnPM increased hemoglobin concentration compared to CPM [change over time: 0.10 g/dL (-0.60, 0.90) vs. CPM: -0.15 g/dL (-1.00, 0.40) P = 0.04]. ... Though the trial was not designed to detect this difference, the prevalence of anemia at endpoint was 33.8 % among those consuming FeZnPM, compared to 57.6 % in the CPM group (P = 0.005).
    Clin Nutr, 2026 · checked 2026-09-29
15. mlt-daily-r4-09 · randomized controlled trial
    The Zim-So drink group demonstrated the largest effect size, with mean breast milk volume increasing by 24.9 mL (95% CI, 22.1-27.7) and prolactin levels by 25.0 ng/mL (95% CI, 22.0-28.0) at Day 14 compared to the control (p < 0.001), indicating a stronger and sustained impact on lactation outcomes.
    Int Breastfeed J, 2025 · checked 2026-09-29
16. mlt-daily-r4-03 · meta-analysis
    The DID was positive and significant (p < 0.05) with high heterogeneity (I2 = 94%) among studies and a pooled SMD of 2.92 (95% CI 0.23–5.62).
    Nutrients, 2022 · checked 2026-09-29
17. mlt-daily-r4-13 · systematic review
    One randomized-controlled study, three randomized-crossover studies, and two non-randomized studies were included.
    Diabetes Metab Syndr, 2026 · checked 2026-09-29