Cowpeas: iron on the label, very little in the blood

Cowpeas, the group that includes black-eyed peas, are often counted as an iron food, and the absorption measured in people is low. In 16 young women eating cowpea meals with iron added, 1.4 to 1.7 percent of it was absorbed as NaFeEDTA and 0.89 to 1.2 percent as ferrous sulfate, with no difference between red and white cowpeas. Fourteen landraces from northern Ghana held 4.9 to 8.2 mg of iron per 3.5 oz (100 g) dry weight alongside 477 to 1110 mg of phytate, and the authors infer, without measuring it, that the phytate limits absorption. In a controlled trial, 241 Ghanaian schoolchildren ate cowpea meal fortified with 10 mg of iron three days a week for about 7 months. Iron deficiency fell by 30 percent and iron deficiency anemia by 47 percent against the same meal unfortified, so the gain belongs to the added iron. The name also covers leaves, pods, fresh seeds and dried seeds, which is why USDA's 20 records run from 26 to 639 micrograms of folate per 3.5 oz (100 g).

The numbers, per 3.5 oz (100 g)

How much is actually in it

• USDA's 20 cowpea records differ more than twentyfold on folate, from 26 to 639 micrograms per 3.5 oz (100 g), because leaves, pods, fresh seeds and dried seeds are all sold as cowpeas. — 20 USDA SR Legacy records for cowpeas, reference record Cowpeas, leafy tips, raw (n = 20 records) Evidence E source
  Reference record vitamin C 36 mg (40% DV), thiamin 0.354 mg (30% DV), folate 101 micrograms (25% DV); across 20 records folate 26 to 639 micrograms, vitamin C 0.2 to 36 mg, thiamin 0.063 to 0.853 mg
• Fourteen cowpea landraces from northern Ghana held 4.9 to 8.2 mg iron and 2.7 to 4.1 mg zinc per 3.5 oz (100 g) dry weight, alongside 477 to 1110 mg of phytate that likely limits how much is absorbed. — 14 cowpea landraces from northern Ghana, biochemical analysis (n = 14 landraces) Evidence C source
  Iron 4.9 to 8.2 mg/100 g dw; zinc 2.7 to 4.1 mg/100 g dw; phytate 477 to 1110 mg/100 g dw; polyphenol 327 to 1055 mg/100 g dw

What a real portion looks like

• Adults in southern Benin eat about 2 to 2.5 oz (58 to 71 g) of cowpea seeds a day, which supplies around 42% of their recommended fiber, 37% of magnesium and 30% of folate. — 1217 adults aged 19 to 65, urban and rural southern Benin, one-week food frequency questionnaire (n = 1217) Evidence C source
  Mean intake 71 g (urban) and 58 g (rural) seeds per adult per day; contribution to RNI fiber 42%, magnesium 37%, folate 30%, protein 26%, energy 15%

What your body absorbs

• Only 0.9 to 1.7% of the iron in fortified cowpea meals was absorbed by 16 young women, with similar values for red and white cowpeas. — 16 young women, paired crossover with stable iron isotopes, red and white cowpea meals (n = 16) Evidence B source
  Iron bioavailability 1.4 and 1.7% with NaFeEDTA, 0.89 and 1.2% with FeSO4 (red and white cowpea); NaFeEDTA added 0.05 to 0.08 mg absorbed iron; no difference between varieties
• In 241 Ghanaian schoolchildren, cowpea meal fortified with 10 mg iron as NaFeEDTA three days a week for about 7 months cut iron deficiency by 30% and iron deficiency anemia by 47%. — 241 children aged 5 to 12, rural northern Ghana, malaria endemic area, double-blind controlled trial (n = 241) Evidence B source
  Prevalence of iron deficiency down 30% and iron deficiency anemia down 47% versus unfortified cowpea meal (P < 0.05)

What cooking and storage change

• Boiled cowpea tested as a low glycemic index food, with a GI of 49 relative to white bread, in a Sri Lankan study of carbohydrate-rich foods. — healthy volunteers, glycemic index testing against white bread Evidence B source
  GI of boiled cowpea 49 plus or minus 8 (white bread reference); boiled chickpea 29, mungbean 57
• Sun-drying and cooking cowpea leaves kept over 60% of their beta-carotene, and Kenyan preschool children eating the dried leaf recipe raised serum beta-carotene and retinol. — preschool children, Machakos District, Kenya; cowpea and amaranth leaf recipe Evidence C source
  All-trans beta-carotene 806 micrograms/g dry matter in fresh cowpea leaves; over 60% retained after dehydration and cooking; hemoglobin up 4.6% from baseline
• A school Salmonella outbreak with 401 cases in Sichuan was traced to cowpea salad kept 3 hours at room temperature in an unsanitised bucket (odds ratio 3.1). — 81 case and 104 control students, boarding school, Sichuan, China, 2013 (n = 185) Evidence C source
  Cowpea salad eaten by 85% of cases vs 60% of controls, OR 3.1 (95% CI 1.3 to 7.8)

What it does to your health

• In 38 adults with high cholesterol, 25 g a day of cowpea protein isolate for 6 weeks cut LDL cholesterol by 18.9% compared with casein, without changing inflammation markers. — 38 adults with moderate hypercholesterolemia, randomized double-blind crossover, 6 weeks per diet (n = 38) Evidence B source
  Total cholesterol -12%, LDL -18.9%, non-HDL -16%, apoB -14%, HDL +2.7%; no change in CRP or adhesion molecules
• Daily cowpea or common bean for Malawian infants aged 6 to 12 months did not change gut microbial diversity or overall community structure in 236 infants. — 236 breastfed rural Malawian infants, 1249 stool samples, within a randomized double-blind trial (n = 236) Evidence B source
  Alpha-diversity unchanged; neither cowpea nor common bean altered overall 16S configuration at any age

Safety, limits and interactions

• In Senegalese markets, cowpea was the only one of eight staple foods contaminated with ochratoxin A, in 16% of samples at an average of 34 micrograms per kg. — 166 market samples of eight food and feed types, Dakar and Kaolack, Senegal (n = 166 samples) Evidence C source
  Ochratoxin A in 16% of cowpea samples, mean 34 micrograms/kg; aflatoxin B1 in almost all samples
• A systematic review of emerging protein sources lists cowpea among legumes with allergenic properties, either through their own allergens or cross-reaction, and with antinutritional factors. — literature on faba bean, mung bean, lentil, black gram, cowpea, quinoa, hemp and other protein sources Evidence D source
  Allergenic potential and antinutritional factors identified; no quantitative safety data after processing

Who this works differently for

• Adding cowpea flour to the food of Malawian infants from 6 to 9 months slowed their loss of length for age (LAZ change -0.14 vs -0.27 with corn-soy blend) in a trial of 355 infants. — 355 rural Malawian infants aged 6 months, 288 breastfed throughout, daily feeding for 6 months (n = 355) Evidence B source
  LAZ change 6 to 9 months: cowpea -0.14 (95% CI -0.24 to -0.04), control -0.27 (-0.35 to -0.19), P = 0.048; no difference in lactulose excretion
• In Malawian children aged 12 to 23 months, 48 weeks of roasted cowpea flour did not improve linear growth compared with corn-soy blend in 331 children who finished the trial. — Malawian children aged 12 to 23 months at entry, 48 weeks, 331 completed (n = 331) Evidence B source
  No significant effect on LAZ, change in LAZ or weight-for-length z score for either legume versus control
• In 69 Nigerian boys hospitalized with acute watery diarrhea, a maize-cowpea-palm oil diet shortened liquid stools to a median 42 hours versus 140 hours on soy formula, with similar treatment failure. — 69 Nigerian boys aged 6 to 24 months hospitalized for acute watery diarrhea (n = 69) Evidence B source
  Median duration of liquid stool 42 h vs 140 h (p < 0.001); treatment failure 5.7% vs 8.8% (p = 0.67); soy formula group ate more and gained more weight

What is still unknown

• Cowpeas are named in definitions of the traditional African diet, yet few of the 45 studies behind it reported how much of any food people ate or how often. — 45 studies defining a traditional African diet (n = 45 studies) Evidence C source
  Cowpeas among lesser-cited food items; few studies reported quantities or frequency of consumption

The bottom line

Boiled cowpea scored a glycemic index of 49 in a Sri Lankan test. That value is set against white bread, so it reads higher than a glucose-based table would show, and the abstract does not say how many volunteers took part. In 38 adults with moderately high cholesterol, 25 g a day of cowpea protein isolate for 6 weeks lowered LDL by 18.9 percent compared with casein. That was an extracted protein powder, and 25 g of protein is more than a normal serving of beans provides, so the trial does not tell you what a bowl of black-eyed peas does. Cowpea leaves kept over 60 percent of their beta-carotene after sun-drying and cooking. Kenyan preschool children eating a leaf recipe mixed with amaranth raised serum beta-carotene, in a study with no control group reported in the abstract. Most of the trials were in young children in Africa. Malawian infants given cowpea flour from 6 to 9 months lost less length for age than those on corn-soy blend, -0.14 against -0.27 (P = 0.048) in 355 infants, and every group still fell behind. In the second year of life, 48 weeks of roasted cowpea flour did not improve growth in 331 children who finished. Daily cowpea did not change gut microbial diversity in 236 infants. In 69 Nigerian boys in hospital with acute watery diarrhea, a maize, cowpea and palm oil diet cut liquid stools to a median 42 hours against 140 on soy formula, with similar treatment failure, and the soy group ate more and gained more weight. Adults in southern Benin, where cowpea is a staple, report eating 2 to 2.5 oz (58 to 71 g) of the seeds a day, which covers about 42 percent of their recommended fiber by one week of self-report. The cautions come from storage. A Salmonella outbreak at a Sichuan boarding school was traced to cowpea salad held 3 hours at room temperature in a bucket that had held raw ingredients (odds ratio 3.1, 95% CI 1.3 to 7.8). In late 1980s Senegalese markets, cowpea was the only one of eight staples carrying ochratoxin A, in 16 percent of samples at an average of 34 micrograms per kg. A systematic review lists cowpea among legumes with allergenic properties, from its own allergens or cross-reaction, on mostly laboratory data that give no figure for how common cowpea allergy is. We found no study on cowpeas in kidney disease or alongside medicines.

Sources

1. cowp-r8-01 · USDA FoodData Central SR Legacy, dataset
   Cowpeas, leafy tips, raw (FDC 169224, SR Legacy, Vegetables and Vegetable Products): vitamin C 36 MG = 40% DV; thiamin 0.354 MG = 30% DV; folate 101 UG = 25% DV; across 20 cowpea records vitamin C 0.2-36, thiamin 0.063-0.853, folate 26-639; none of the 20 is a Foundation record, so there is no sample spread
   USDA FoodData Central, SR Legacy, FDC 169224 · checked 2026-09-29
2. cowp-r8-02 · composition analysis
   Cowpea landraces contained iron and zinc in the range of 4.9-8.2 mg/100 g d.w and 2.7-4.1 mg/100 g d.w respectively. The landraces also contained high amounts of phytate (477-1110 mg/100 g d.w) and polyphenol (327-1055 mg/100 g d.w).
   PLoS One, 2013 · checked 2026-09-29
3. cowp-r8-09 · dietary survey
   The mean amount consumed was 71 g and 58 g of seeds/adult/day in urban and rural areas respectively. The mean daily contribution of cowpea-based dishes to RNI was 15% for energy, 42% for fibre, 37% for magnesium, 30% for folate, 26% for protein, and just above 15% for zinc and potassium.
   Nutrients, 2023 · checked 2026-09-29
4. cowp-r8-03 · randomized crossover trial
   Iron bioavailabilities from red and white cowpeas were 1.4 and 1.7%, respectively, in NaFeEDTA-fortified meals and 0.89 and 1.2%, respectively, in FeSO(4)-fortified meals.
   J Nutr, 2012 · checked 2026-09-29
5. cowp-r8-04 · randomized controlled trial
   Fortification resulted in a 30 and 47% reduction in the prevalence of iron deficiency (ID) and iron deficiency anemia (IDA) (P < 0.05), respectively.
   J Nutr, 2012 · checked 2026-09-29
6. cowp-r8-08 · glycemic index trial
   Kurakkan (Eleusine coracana) flour roti, rice flour roti, atta flour roti, boiled breadfruit (Artocarpus altilis/Artocarpus communis) and boiled legumes (mungbean, cowpea and chickpea) were categorized as low-GI foods (relative to white bread; Prima Crust Top), and the corresponding GI (+/- standard error of the mean) values were 70+/-8, 69+/-7, 67+/-9, 64+/-7, 57+/-6, 49+/-8 and 29+/-5, respectively.
   Int J Food Sci Nutr, 2009 · checked 2026-09-29
7. cowp-r8-10 · feeding intervention
   All-trans-BC levels in uncooked fresh cowpea and amaranth leaves were 806.0 μg/g and 599.0 μg/g dry matter, respectively, while the dehydration and cooking processes retained the β-carotene levels at over 60 %.
   Eur J Nutr, 2013 · checked 2026-09-29
8. cowp-r8-11 · case-control
   Eighty-five percent of case students, consumed cowpea salad compared to 60% of control students at lunch on 12 June (odds ratio 3·1, 95% confidence interval 1·3-7·8). The cowpeas were stored at room temperature for 3 h in a bucket previously used to store raw ingredients.
   Epidemiol Infect, 2016 · checked 2026-09-29
9. cowp-r8-07 · randomized crossover trial
   Consumption of cowpea protein significantly reduced total cholesterol ( 12 %), LDL cholesterol ( 18.9 %), non HDL-cholesterol ( 16 %) and apoB ( 14 %), and increased HDL cholesterol (+2.7 %).
   Nutr Hosp, 2015 · checked 2026-09-29
10. cowp-r8-15 · randomized controlled trial
    Neither cowpea nor common bean altered the overall 16S configuration at any age.
    Am J Clin Nutr, 2020 · checked 2026-09-29
11. cowp-r8-12 · contamination survey
    Only cowpea (Vigna unguiculata), from a legume widely distributed in Senegal, was contaminated (16% of samples), at an average level of 34 micrograms/kg.
    IARC Sci Publ, 1991 · checked 2026-09-29
12. cowp-r8-13 · systematic review
    Also, these protein sources have been shown to display allergenic properties, either through the intrinsic presence of allergens or through cross-reaction.
    Compr Rev Food Sci Food Saf, 2025 · checked 2026-09-29
13. cowp-r8-05 · randomized controlled trial
    The changes in LAZ [mean (95% CI)] for the cowpea, common bean, and control groups from 6 to 9 mo were -0.14 (-0.24, -0.04), -0.27 (-0.38, -0.16), and -0.27 (-0.35, -0.19), respectively. LAZ was reduced less in infants receiving cowpea than in those receiving control food from 6 to 9 mo (P = 0.048).
    Am J Clin Nutr, 2017 · checked 2026-09-29
14. cowp-r8-06 · randomized controlled trial
    No significant effects on LAZ, change in LAZ, or weight-for-length z score were observed due to either intervention legume, compared to the control.
    J Nutr, 2018 · checked 2026-09-29
15. cowp-r8-14 · randomized controlled trial
    Children in group MCP had a substantially reduced duration of liquid stool excretion (estimated median duration 42 h versus 140 h; p < 0.001).
    Acta Paediatr, 1994 · checked 2026-09-29
16. cowp-r8-16 · scoping review
    Other, lesser-cited food items include cassava, cowpeas, fish, fruit, legumes, millet, and sorghum. ... Few studies reported the quantities of foods consumed and the frequency of consumption.
    Front Nutr, 2025 · checked 2026-09-29