Beans and peas: a small effect you have to actually eat

Beans and peas do one measurable thing, and the measurement is modest. A meta-analysis of 38 randomized trials in 2095 adults, at a median dose of 4.6 oz (130 g) of cooked pulses a day for a median of six weeks, found LDL cholesterol down by 5.4 mg/dL (95% CI -7.3 to -3.1; metric: 0.14 mmol/L, -0.19 to -0.08). Non-HDL cholesterol fell by 8.5 mg/dL (-12 to -5.4; metric: 0.22 mmol/L, -0.30 to -0.14), with GRADE certainty from moderate to high for everything except apoB, and the authors call the 1.2 mg/dL (0.03 mmol/L) drop in HDL trivial. An older pool of 10 trials in 268 adults eating non-soy legumes put the fall in total cholesterol at 11.8 mg/dL (95% CI -16.1 to -7.5), in a set where 70 percent of participants were men and nothing ran longer than a few months. Blood sugar moves as well. A review pooled 41 randomized trials, and in the 11 of them that gave pulses on their own fasting glucose fell by 0.82 standardized units (95% CI -1.36 to -0.27) in people with and without diabetes, with heterogeneity between trials that the authors call high and mostly unexplained. A median dose of 130 g is half to two thirds of a cup, most days, which is the part almost nobody does.

How much is actually in it

• Eight analyzed samples of dry cannellini beans held from 5.26 to 8.57 mg of iron per 3.5 oz (100 g) and from 103 to 184.5 mg of calcium, so one label number covers a 50% spread. — USDA Foundation Foods, 8 analyzed samples of Beans, cannellini, dry (n = 8 samples) Evidence E source
  Iron 6.673 mg per 3.5 oz (100 g) (5.26-8.57, spread 50%), calcium 142.9 mg (103.0-184.5, 57%), copper 0.7419 mg (0.578-0.882, 41%), manganese 1.784 mg (1.38-2.28, 50%), phosphorus 412.3 mg (337.0-481.0, 35%), magnesium 153.6 mg (142.0-160.0, 12%)
• Eight analyzed samples of canned green peas ranged from 0.913 to 1.38 mg of iron per 3.5 oz (100 g) and from 21.4 to 36.9 mg of calcium, a spread of 41% and 55% around the mean. — USDA Foundation Foods, 8 analyzed samples of canned sweet green peas with sodium added (n = 8 samples) Evidence E source
  Iron 1.139 mg per 3.5 oz (100 g) (0.913-1.38, spread 41%), calcium 28.21 mg (21.4-36.9, 55%), copper 0.1233 mg (0.0847-0.159, 60%), phosphorus 77.74 mg (64.9-95.3, 39%), magnesium 21.91 mg (20.0-26.0, 27%), manganese 0.2849 mg (0.24-0.334, 33%)

What a real portion looks like

• Most of twenty human trials reported improvements in blood lipids, blood pressure, inflammation and body composition at a daily amount of 5.3 oz (150 g) of cooked pulses, with a range from 54 to 360 g. — twenty human dietary intervention publications longer than one day, on whole pulses, reviewed for the amount needed to see an effect (n = 20 publications) Evidence A source
  Daily amount 150 g cooked, range 54 to 360 g per day, with reported improvements in blood lipids, blood pressure, inflammation markers and body composition
• Only 1% of UK children and 2% of UK adults ate a full 2.8 oz (80 g) portion of pulses a day, even though over half of both groups ate some, with baked beans the most common. — 7656 children and 7999 adults with food diaries in the UK National Diet and Nutrition Survey 2008 to 2019 (n = 15655) Evidence C source
  Pulses eaten by 55.1% of children and 56.7% of adults, a full portion by 1% and 2%, legumes eaten by 72.5% and 79%, a full portion by 1.4% and 5.2%
• A cup of blackeyed peas cooked from dried weighs 6.5 oz (185 g) in the US food survey, and one ounce of dry peas yields 72 g after cooking. — USDA FNDDS measures for Blackeyed peas, from dried, WWEIA category Beans, peas, legumes Evidence E source
  1 cup = 185 g, 1 oz dry yields 72 g after cooking, quantity not specified = 92 g

What your body absorbs

• Phytic acid, tannins, saponins, lectins and enzyme inhibitors in beans and peas bind minerals and hold back their absorption, which is why soaking, germination and fermentation exist. — food systems review of anti-nutritional factors and the processing methods that reduce them, laboratory and technological studies rather than human feeding trials Evidence D source
  Direction only, no absorption figures for any single bean or pea are given

What cooking and storage change

• Boiling legumes for 15 to 20 minutes and draining them keeps 60% of the folate, while boiling the same group for 2 to 2.5 hours keeps only 30%, and thiamin falls from 65% to 40%. — USDA retention factors for the whole group Legumes and Legume Products, boiled and drained, at two cooking durations Evidence E source
  15 to 20 minutes boiled and drained: folate 60%, thiamin 65%, vitamin C 65%, copper 70%, niacin 70%, vitamin B6 70%, magnesium 80%, iron 85%, zinc 85% · 2 to 2.5 hours: folate 30%, thiamin 40%, vitamin B6 50%, copper 55%, niacin 55%, magnesium 70%, iron 75%, zinc 85%

What it does to your health

• Across 38 randomized trials with 2095 people eating a median 4.6 oz (130 g) of pulses a day for a median of 6 weeks, LDL cholesterol fell by 5.4 mg/dL (0.14 mmol/L) and non-HDL cholesterol by 8.5 mg/dL (0.22 mmol/L). — 2095 adults in 38 randomized trials (52 comparisons) of at least 3 weeks, median dose 130 g per day, which is about half to two thirds of a cup (n = 2095) Evidence A source
  LDL -5.4 mg/dL (95% CI -7.3 to -3.1; metric: -0.14 mmol/L, -0.19 to -0.08), non-HDL -8.5 mg/dL (-12 to -5.4; metric: -0.22 mmol/L, -0.30 to -0.14), apoB -0.08 g/L (-0.13 to -0.03), HDL -1.2 mg/dL (-1.9 to -0.39; metric: -0.03 mmol/L, -0.05 to -0.01), linear dose response for beans up to 1 cup a day
• Pooling 10 randomized trials with 268 adults who ate non-soy legumes for at least 3 weeks, total cholesterol fell by 11.8 mg/dL and LDL cholesterol by 8.0 mg/dL against control diets. — 268 adults in 10 randomized controlled trials of at least 3 weeks comparing a non-soy legume diet with a control diet of similar energy and macronutrients (n = 268) Evidence A source
  Total cholesterol -11.8 mg/dL (95% CI -16.1 to -7.5), LDL cholesterol -8.0 mg/dL (95% CI -11.4 to -4.6)
• In the 11 trials of a review of 41 randomized trials that gave pulses on their own, fasting blood glucose fell by 0.82 standardized units and fasting insulin by 0.49 in people with and without diabetes. — participants of 41 randomized controlled trials of at least 7 days, with and without diabetes, taking pulses alone, in low glycemic index diets or in high fiber diets (n = 41 trials) Evidence A source
  Pulses alone (11 trials) fasting glucose SMD -0.82 (95% CI -1.36 to -0.27) and insulin -0.49 (-0.93 to -0.04), pulses in low GI diets (19 trials) glycated proteins -0.28 (-0.42 to -0.14), pulses in high fiber diets (11 trials) fasting glucose -0.32 (-0.49 to -0.15)
• Pooling 9 acute feeding trials, a meal with pulses raised the satiety curve by 31% against an equal calorie control, and people still ate the same amount at the next meal. — adults in 9 acute controlled feeding trials of beans, peas, chickpeas and lentils against isocaloric control meals (n = 9 trials) Evidence A source
  Satiety incremental area under the curve ratio of means 1.31 (95% CI 1.09 to 1.58, P = 0.004, I2 = 0%), second meal intake mean difference -19.94 calories (95% CI -75 to 35, P = 0.48)
• In a 2026 review of 40 human intervention studies, pulses gave a lower glucose response than the starchy control in 28 of them and a lower insulin response in 16 of 24. — 40 human intervention studies of non-oil-seed pulses against starchy controls, mostly wheat flour products or white rice, measuring postprandial events (n = 40 studies) Evidence B source
  Lower glucose in 28 of 40 studies, lower insulin in 16 of 24, reduced hunger in 3 of 5, significantly greater satiety in only 1 of 6 and greater fullness in 3 of 6

Safety, limits and interactions

• Lectins in common beans act as toxic allergens and hemagglutinins and can cause intestinal disorders, which is the reason undercooked beans make people ill. — laboratory and animal studies of Phaseolus vulgaris lectins, their purification and the effect of heating, proteolysis, irradiation and high pressure on them Evidence D source
  Direction only, with heating among the treatments that eliminate the activity of the lectins

What people believe that the data does not support

• US labeling rules put beans, peas and lentils in the same protein group as lean meat, seafood, eggs, nuts and seeds, while no health claim for pulses and cholesterol exists in either the US or the EU register. — 21 CFR 101.65, implied nutrient content claims, paragraph (i), and the EU Register on nutrition and health claims, snapshot 2026-09-21 Evidence E source
  Beans, peas and lentils named in the protein group of the US rule, zero claims under the name beans or peas in the EU register of 2330 claims

Who this works differently for

• In the UK national survey children ate 10.6 g of pulses a day and adults 15.0 g, so both groups sit far below the 80 g portion that dietary advice counts. — 7656 children aged 1.5 to 18 years and 7999 adults aged 19 to 96 years in the UK National Diet and Nutrition Survey 2008 to 2019, with 3 or 4 day food diaries (n = 15655) Evidence C source
  Pulses 10.6 ± 27.0 g per day in children and 15.0 ± 21.0 g in adults, legumes 16.7 ± 32.5 g and 27.3 ± 26.0 g, and a portion of 2.8 oz (80 g) linked to a 3.7 point higher EAT-Lancet index

What is still unknown

• A 2026 UK mapping review found 50 ongoing legume research projects clustered at the two ends of the chain, with four on product development and none at all on processing, food service or retail. — 50 ongoing UK research projects on beans, peas and other pulses identified in three databases for 2019 to 2023 (n = 50 projects) Evidence C source
  21 projects on production, 21 on consumption, 4 spanning the whole chain, 4 on product development, 0 on processing, food service or retail

The bottom line

Most of twenty human trials reported improvements in lipids, blood pressure and inflammation at a daily amount of 5.3 oz (150 g) of cooked pulses, with a range from 54 to 360 g, in a review whose authors flag methodological problems in the included work and no settled mechanism. Set that against what people eat. In the UK national survey of 7656 children and 7999 adults, more than half of both groups ate some pulses. Only 1 percent of the children and 2 percent of the adults reached a full 2.8 oz (80 g) portion on an average day, with baked beans the most common entry. Mean intakes were 10.6 g a day in children and 15.0 g in adults, from self-reported food diaries rather than weighed plates. The distance between 15 g and 130 g is the whole story of this food. Pulses also make a meal feel longer. Pooling 9 acute feeding trials, a meal with pulses raised the satiety curve by 31 percent (95% CI 1.09 to 1.58) against an equal calorie control. Intake at the next meal was unchanged. That is the effect of one serving rather than of a habit. Cooking time costs more than most people expect. USDA retention factors for the legume group, set on the whole group rather than on any one bean and dated 1975, leave 60 percent of the folate and 65 percent of the thiamin after 15 to 20 minutes of boiling and draining. After two to two and a half hours the same factors fall to 30 percent and 40 percent. Iron and zinc hold up at 75 and 85 percent even after the long boil. Much of what is lost leaves in the water you pour away. Absorption is a separate question, and the answer we have is thin. A review of food processing studies lists phytic acid, tannins, saponins, lectins and enzyme inhibitors as compounds that hinder mineral absorption, and it reports direction only, with no absorption figure for any single bean or pea. That is the reason soaking, germination and fermentation exist, and it is a laboratory review rather than a feeding trial in people. Beans need real cooking, not warming. A review of laboratory and animal work on Phaseolus vulgaris lectins describes them as toxic allergens and hemagglutinins that can cause intestinal disorders, with heating among the treatments that destroy the activity, and it gives no dose at which a person falls ill. That is a mechanism and a warning about raw or undercooked beans, not a measured risk. We have no card on pulses in pregnancy or in kidney disease and none on any drug interaction, so this page says nothing about them. For scale, a cup of blackeyed peas cooked from dried weighs 6.5 oz (185 g) in the US food survey, more than the 130 g that produced the cholesterol result, and one ounce of dry peas yields 72 g cooked. US labeling rules already count beans, peas and lentils in the same protein group as lean meat, seafood, eggs, nuts and seeds, while no cholesterol claim for pulses exists in the US register or in the EU register of 2330 claims. A 2026 mapping review of 50 ongoing UK research projects found 21 on production and 21 on consumption, four on product development and none at all on processing, food service or retail, which maps what is being studied rather than what is known.

Sources

1. bnpe-r4-11 · food composition dataset
   Beans, cannellini, dry | Iron, Fe 6.673 MG range 5.26-8.57 (median 6.565), 8 samples, spread 50% | Calcium, Ca 142.9 MG range 103.0-184.5 (median 136.0), 8 samples, spread 57% | Copper, Cu 0.7419 MG range 0.578-0.882 (median 0.7685), 8 samples, spread 41% | Manganese, Mn 1.784 MG range 1.38-2.28 (median 1.675), 8 samples, spread 50% | Phosphorus, P 412.3 MG range 337.0-481.0 (median 426.5), 8 samples, spread 35% | Magnesium, Mg 153.6 MG range 142.0-160.0 (median 154.0), 8 samples, spread 12%
   USDA FoodData Central, Foundation Foods, release 2026-04-30 · checked 2026-09-25
2. bnpe-r4-12 · food composition dataset
   Peas, green, sweet, canned, sodium added | Iron, Fe 1.139 MG range 0.913-1.38 (median 1.16), 8 samples, spread 41% | Calcium, Ca 28.21 MG range 21.4-36.9 (median 26.75), 8 samples, spread 55% | Copper, Cu 0.1233 MG range 0.0847-0.159 (median 0.1295), 8 samples, spread 60% | Phosphorus, P 77.74 MG range 64.9-95.3 (median 78.6), 8 samples, spread 39% | Magnesium, Mg 21.91 MG range 20.0-26.0 (median 21.1), 8 samples, spread 27% | Manganese, Mn 0.2849 MG range 0.24-0.334 (median 0.283), 8 samples, spread 33%
   USDA FoodData Central, Foundation Foods, release 2026-04-30 · checked 2026-09-25
3. bnpe-r4-05 · systematic review of human trials
   Most of the twenty eligible publications reported improvements in blood lipid profile, blood pressure, inflammation biomarkers, as well as, in body composition, resulting from pulse daily amounts of 150 g (minimum-maximum: 54-360 g/day; cooked).
   Crit Rev Food Sci Nutr, 2021 · checked 2026-09-25
4. bnpe-r4-10 · cross-sectional survey analysis
   Pulses were consumed by 55.1% (n = 4221/7656) and 56.7% (n = 4537/7999) of the UK children (1.5–18 years) and adults (19–96 years), which represented mean intakes of 10.6 ± 27.0 g/d and 15.0 ± 21.0 g/d respectively. Only 1% (n = 60/7656) of the children and 2% (n = 140/7999) of the adults included in the analysis consumed on average one portion (80 g) or more of pulses/day.
   Eur J Nutr, 2025 · checked 2026-09-25
5. bnpe-r4-14 · USDA FNDDS, dataset
   FNDDS 41301010 Blackeyed peas, from dried (fdc_id 2707413, WWEIA Beans, peas, legumes): 1 cup 10205 = 185 g · 1 oz dry, yield after cooking 40014 = 72 g · Quantity not specified 90000 = 92 g
   USDA FNDDS, food code 41301010 · checked 2026-09-25
6. bnpe-r4-07 · systematic review of laboratory and processing studies
   Naturally occurring compounds in foods, such as tannins, saponins, phytic acid, gossypol, lectins, and various enzyme inhibitors (protease inhibitors (PIs) and amylase inhibitors), function as anti-nutritional factors (ANFs) that hinder the absorption of essential vitamins and minerals, negatively impacting nutritional bioavailability and human health.
   J Food Sci, 2025 · checked 2026-09-25
7. bnpe-r4-13 · USDA nutrient retention factors, dataset
   USDA-RETN-R6:501 LEGUMES,CKD 15/20MIN,BOILED,DRAINED (group 16 Legumes and Legume Products): folate_total_ug 60%, vitamin_c_mg 65%, thiamin_mg 65%, copper_mg 70%, niacin_mg 70%, vitamin_b6_mg 70%, potassium_mg 75%, riboflavin_mg 75%, magnesium_mg 80%, calcium_mg 85%, iron_mg 85%, zinc_mg 85% || USDA-RETN-R6:541 LEGUMES,CKD 2/2.5HRS,BOILED,DRAINED: folate_total_ug 30%, thiamin_mg 40%, vitamin_b6_mg 50%, copper_mg 55%, niacin_mg 55%, potassium_mg 65%, vitamin_c_mg 65%, magnesium_mg 70%, iron_mg 75%, riboflavin_mg 75%, calcium_mg 85%, zinc_mg 85%
   USDA Table of Nutrient Retention Factors, Release 6, 2007, codes 501 and 541 · checked 2026-09-25
8. bnpe-r4-01 · meta-analysis of randomized trials
   Thirty-eight trials (52 trial comparisons, n=2095) with a median of 6 weeks and dose of 130 g/d (0.5-0.67 cup/d) showed that whole dietary pulses decreased low-density lipoprotein-cholesterol (mean difference, -0.14 mmol/L [95% CI, -0.19 to -0.08]), non-high-density lipoprotein-cholesterol (-0.22 mmol/L [95% CI, -0.30 to -0.14]), apoB (apolipoprotein B) (-0.08 g/L [95% CI, -0.13 to -0.03]) and high-density lipoprotein-cholesterol (-0.03 mmol/L [95% CI, -0.05 to -0.01]) with no effects on other lipids.
   J Am Heart Assoc, 2026 · checked 2026-09-25
9. bnpe-r4-02 · meta-analysis of randomized trials
   Pooled mean net change in total cholesterol for those treated with a legume diet compared to control was -11.8 mg/dL (95% confidence interval [CI], -16.1 to -7.5); mean net change in low-density lipoprotein cholesterol was -8.0mg/dL (95% CI, -11.4 to -4.6).
   Nutr Metab Cardiovasc Dis, 2011 · checked 2026-09-25
10. bnpe-r4-03 · meta-analysis of randomized trials
    A total of 41 trials (39 reports) were included. Pulses alone (11 trials) lowered fasting blood glucose (FBG) (-0.82, 95% CI -1.36 to -0.27) and insulin (-0.49, 95% CI -0.93 to -0.04).
    Diabetologia, 2009 · checked 2026-09-25
11. bnpe-r4-04 · meta-analysis of acute feeding trials
    Nine trials met the eligibility criteria. ... Pooled analyses show that dietary pulses contribute to acute satiety but not second meal intake.
    Obesity (Silver Spring), 2014 · checked 2026-09-25
12. bnpe-r4-06 · systematic review of human trials
    Most of the articles in this review reported that non-oil-seed pulses cause lower glucose (28 of 40 studies) and insulin (16 of 24 studies) responses than other starchy foods in the control groups (mainly wheat-flour-based products or white rice).
    Nutr Rev, 2026 · checked 2026-09-25
13. bnpe-r4-08 · systematic review of mechanistic studies
    Studies have shown that lectins are anti-nutritional factors that can cause intestinal disorders. Owing to their ability to act as toxic allergens and hemagglutinins, the Phaseolus vulgaris lectins are of grave concern for human health and safety.
    Crit Rev Food Sci Nutr, 2018 · checked 2026-09-25
14. bnpe-r4-15 · US federal regulation
    (E) Lean meat, seafood, eggs, beans, peas, lentils, nuts, or seeds.
    21 CFR 101.65, Implied nutrient content claims and related label statements · checked 2026-09-25
15. bnpe-r4-09 · cross-sectional survey analysis
    Children and adults consumed mean ± SD 10.6 ± 27.0 g/day and 15.0 ± 21.0 g/day of pulses, and 16.7 ± 32.5 g/day and 27.3 ± 26.0 g/day of legumes, respectively.
    Eur J Nutr, 2025 · checked 2026-09-25
16. bnpe-r4-16 · mapping review of research activity
    Fifty ongoing research projects were identified, revealing a focus on the two ends of the value chain: production (twenty-one projects) and consumption (twenty-one projects). ... Few projects focused on the middle of the value chain (four projects on product development) with none focused on processing, food service or retail.
    Public Health Nutr, 2026 · checked 2026-09-25