Bean salad: half a cup a day, and a small move in LDL

A bean salad is a dish, and nobody has studied the dish. What has been tested is the beans in it. Across 38 randomized trials in 2,095 people, about 4.6 oz (130 g) of whole pulses a day, roughly half a cup, lowered LDL cholesterol by 5.4 mg/dL (0.14 mmol/L), 95 percent confidence interval 0.19 to 0.08 lower. The certainty of that was rated moderate to high, and for beans the line kept falling up to one cup a day, at 0.25 mmol/L per extra half cup. Those trials ran a median of six weeks and fed whole pulses, with no bowl of dressing over them. In 21 trials covering 940 people, a median of 4.7 oz (132 g) of pulses a day took 0.75 lb (0.34 kg) off body weight over a median of six weeks, even where the diets were never meant to cut calories. That is a small shift over a short run, and it says nothing about a year. We have no USDA entry for bean salad in this cohort, so this page carries no figure for what a serving of the dish contains. If you cook dry kidney beans for it yourself, the FDA says to soak them for at least 5 hours, pour the water away and boil them in fresh water for at least 30 minutes. It warns against slow cookers, which often reached only 167 °F (75°C) inside. As few as four or five raw beans can cause symptoms.

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

What your body absorbs

• Phytic acid in beans is what holds their iron back. In 22 women with low iron status, 9.2% of the iron was absorbed from ordinary bean meals, and removing more than 95% of the phytate raised the iron absorbed from high-iron beans to 746 micrograms. — 22 Rwandese women with low iron status eating composite bean meals with potatoes or rice in a crossover design (n = 22) Evidence B source
  Fractional iron absorption 9.2% from control bean meals, 30% higher than from biofortified beans with native phytate; iron absorbed rose from 406 to 599 and 746 micrograms with 50% and over 95% dephytinization

What cooking and storage change

• Pooling 61 laboratory studies, extrusion cut legume allergenicity measured by antibody binding by a standardized 20.19 and fermentation by 20.8, with fermentation followed by hydrolysis the strongest at 53.34, none of it tested by feeding allergic people. — 61 in vitro studies of nuts and legumes with allergenicity measured by ELISA (n = 61 studies) Evidence D source
  Extrusion SMD -20.19 (95% CI -22.22 to -18.17), fermentation SMD -20.8 (95% CI -24.10 to -17.50), fermentation plus proteolytic hydrolysis SMD -53.34 (95% CI -70.18 to -36.5)
• The FDA says dry kidney beans should be soaked in water for at least 5 hours, the water poured away, and the beans boiled in fresh water for at least 30 minutes. It adds that as few as four or five raw beans can trigger symptoms. — consumers cooking dry red kidney beans (Phaseolus vulgaris) at home Evidence E source
  Soak at least 5 hours, discard the water, boil in fresh water at least 30 minutes; as few as four or five raw beans can cause symptoms
• The FDA says slow cookers should not be used for kidney beans, because slow-cooked casseroles often reached only 167 °F (75°C) inside, too low to destroy the lectin. Raw kidney beans carry 20,000 to 70,000 hemagglutinating units and fully cooked beans 200 to 400. — consumers cooking kidney beans or dishes that contain them in slow cookers Evidence E source
  Slow-cooked casseroles often reached only 167 °F (75°C) or less inside; boiling 10 minutes at 212 °F (100°C) destroyed the toxin; raw beans 20,000 to 70,000 hau, fully cooked 200 to 400 hau

What it does to your health

• Across 38 randomized trials in 2095 people, eating about 4.6 oz (130 g) of whole pulses a day, roughly half a cup, lowered LDL cholesterol by 5.4 mg/dL (0.14 mmol/L), and for beans the effect kept growing up to one cup a day. — 2095 participants in 38 randomized controlled trials lasting at least 3 weeks, median 6 weeks (n = 2095) Evidence A source
  LDL cholesterol mean difference -5.4 mg/dL (95% CI -7.3 to -3.1; metric: -0.14 mmol/L, -0.19 to -0.08); dose-response for beans -9.7 mg/dL (-0.25 mmol/L) per 0.5 cup (95% CI -0.48 to -0.02) up to 1 cup/d
• Adding about one serving of pulses a day, a median of 4.7 oz (132 g), produced a small weight loss of 0.75 lb (0.34 kg) over a median of six weeks in 21 randomized trials, even when the diets were not meant to cut calories. — 940 participants in 21 randomized controlled trials of at least 3 weeks (n = 940) Evidence A source
  Body weight -0.75 lb (95% CI -1.4 to -0.09, P = 0.03; metric: -0.34 kg, -0.63 to -0.04) at a median intake of 132 g/d
• For blood pressure the honest answer is no overall effect: across 16 randomized trials in 1092 people systolic pressure fell by only 1.06 mmHg, which was not significant, and a drop of 2.79 mmHg appeared only in participants with overweight or obesity. — 1092 participants in 16 randomized controlled trials, aged 17 to 75, 4 to 52 weeks, 16 to 111 oz (450 to 3150 g) of legumes a week (n = 1092) Evidence A source
  Systolic BP -1.06 mmHg (95% CI -2.57 to 0.44), diastolic BP -0.48 mmHg (95% CI -1.06 to 0.10); subgroup with overweight or obese BMI -2.79 mmHg (95% CI -4.68 to -0.90)

Safety, limits and interactions

• The safety question for a bean salad is whether the beans were properly cooked, because the lectins in common beans act as toxic hemagglutinins and can cause intestinal disorders when they survive. — laboratory and animal studies of Phaseolus vulgaris lectins reviewed systematically Evidence D source
  Lectins described as anti-nutritional factors causing intestinal disorders, and as toxic allergens and hemagglutinins
• Beans and lentils are themselves allergens for some people, both in their own right and by cross-reaction with other legumes, and a 2025 review found no conclusive data on how safe these proteins are once processed. — literature on faba bean, mung bean, lentil, black gram, cowpea and other alternative protein sources Evidence D source
  Allergenic properties through intrinsic allergens or cross-reaction, plus antinutritional factors, phytoestrogens and oligosaccharides in all sources reviewed
• Between 1976 and 1989 the UK recorded 50 suspected incidents of red kidney bean poisoning, 9 confirmed by lectin found in the beans and 23 more on symptoms. In many of them the beans were eaten raw or after inadequate heating. — 50 suspected incidents of red kidney bean poisoning reported in the UK between July 1976 and February 1989 (n = 50) Evidence C source
  9 incidents confirmed by hemagglutinin in the beans, 23 more diagnosed on symptoms and preparation; nausea, vomiting and diarrhea within 1 to 7 h

What people believe that the data does not support

• The gas people expect from beans may be smaller than feared. Eating half a cup daily for 8 to 12 weeks, fewer than half reported more flatulence in the first week, and only 19% did on black-eyed peas. On the bean-free control diets 3 to 11% reported it too. — 86 adults across three feeding studies eating half a cup of pinto beans, black-eyed peas or baked beans daily for 8 or 12 weeks (n = 86) Evidence B source
  Under 50% reported increased flatulence in week one on pinto or baked beans, 19% on black-eyed peas, and 3 to 11% on control diets without flatulence-producing components

What is still unknown

• What nobody can yet quantify is the safety of legume proteins after they have been cooked, canned or processed, since the 2025 review of 10 alternative protein sources found no conclusive data on it. — literature on alternative plant protein sources including beans, lentils and cowpea Evidence D source
  No conclusive data available to quantitatively assess safety after processing

The bottom line

Blood pressure is where the honest answer is no. Across 16 randomized trials in 1,092 people eating 16 to 111 oz (450 to 3,150 g) of legumes a week for 4 to 52 weeks, systolic pressure fell by 1.06 mmHg, confidence interval 2.57 lower to 0.44 higher. That does not reach significance. Diastolic pressure moved 0.48 mmHg, interval 1.06 lower to 0.10 higher. A fall of 2.79 mmHg did appear in participants with overweight or obesity, interval 4.68 to 0.90 lower, and that was an exploratory subgroup inside a body of evidence the authors rated low certainty. The gas may be smaller than its reputation. In three feeding studies covering 86 adults who ate half a cup of beans a day for 8 or 12 weeks, fewer than half reported more flatulence in the first week on pinto or baked beans, and 19 percent did on black-eyed peas. On the control diets, which had nothing gas-forming in them, 3 to 11 percent reported it anyway. Those were symptoms people wrote down rather than gas anybody measured, and the studies were built around heart disease markers. If you eat this for the iron, the phytate in the beans works against you. In 22 Rwandese women with low iron status, 9.2 percent of the iron in ordinary bean meals was absorbed, measured with stable isotopes. Stripping out more than 95 percent of the phytate raised the iron absorbed from high-iron beans from 406 to 746 micrograms. That stripping was done in a laboratory, and nothing in our sources says what soaking beans or rinsing a tin at home does to it. The caution that matters for this dish is whether the beans were cooked through. Lectins in common beans act as hemagglutinins and are called toxic in a systematic review of laboratory and animal work, causing intestinal disorders when they survive cooking. In the UK, 50 suspected incidents of red kidney bean poisoning were reported between 1976 and 1989, 9 confirmed by lectin found in the beans and 23 more on symptoms, with nausea, vomiting and diarrhea within 1 to 7 hours. In many of them the beans had been eaten raw or after inadequate heating. The FDA advises soaking dry kidney beans for at least 5 hours, discarding the water and boiling them in fresh water for at least 30 minutes. It warns against slow cookers for kidney beans, because slow-cooked casseroles often reached only 167 °F (75°C) inside, too low to destroy the lectin, while boiling for 10 minutes at 212 °F (100°C) destroyed it. Raw kidney beans carried 20,000 to 70,000 hemagglutinating units and fully cooked beans 200 to 400. Beans and lentils are also allergens in their own right, and by cross-reaction with other legumes. Pooling 61 laboratory studies, extrusion cut measured allergenicity by a standardized 20.19 and fermentation by 20.8, with fermentation followed by hydrolysis strongest at 53.34. Those are antibody readings on a plate rather than allergic people eating anything, and none of those treatments is what happens when you open a tin and make a salad. A 2025 review of ten alternative protein sources found no conclusive data on how safe these proteins are once cooked, canned or processed, which is the open question under this whole page.

Sources

1. bsal-r4-04 · randomized crossover trial
   With native PA concentrations, fractional iron absorption from the control bean meals was 9.2%, 30% higher than that from the biofortified bean meals (P < 0.001)...PA strongly decreases iron bioavailability from iron-biofortified beans, and a high PA concentration is an important impediment to the optimal effectiveness of bean iron biofortification.
   J Nutr, 2014 · checked 2026-09-24
2. bsal-r4-08 · meta-analysis of in vitro studies
   In single processing treatment, the extrusion and fermentation had the largest reduction in allergenicity, considering their SMD value, that is, -20.19 (95% CI: -22.22 to -18.17; the certainty of evidence: moderate) and -20.8 (95% CI: -24.10 to -17.50; the certainty of evidence: moderate), respectively.
   J Food Sci, 2023 · checked 2026-09-24
3. hf2-bsal-1 · agency guidance
   Studies done by British scientists suggest that beans should be soaked in water for at least 5 hours, the water poured away, and the beans boiled in fresh water for at least 30 minutes. ... Toxic dose: As few as four or five raw beans can trigger symptoms.
   US FDA, Bad Bug Book, 2nd edition (2012), chapter Phytohaemagglutinin (kidney bean lectin) · checked 2026-10-08
4. hf2-bsal-2 · agency guidance
   Raw kidney beans contain from 20,000 to 70,000 hau, while fully cooked beans contain from 200 to 400 hau. ... Slow cookers should not be used to cook these beans or dishes that contain them. Studies of casseroles cooked in slow cookers revealed that the food often reached internal temperatures of only 75°C or less, which is inadequate for destruction of the toxin.
   US FDA, Bad Bug Book, 2nd edition (2012), chapter Phytohaemagglutinin (kidney bean lectin) · checked 2026-10-08
5. bsal-r4-01 · systematic review and meta-analysis
   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])...A linear inverse relationship was shown for beans up to 1 cup/d for low-density lipoprotein-cholesterol (coefficient, -0.25 mmol/L/0.5 cup [95% CI, -0.48 to -0.02])
   J Am Heart Assoc, 2026 · checked 2026-09-24
6. bsal-r4-02 · systematic review and meta-analysis
   The pooled analysis showed an overall significant weight reduction of -0.34 kg (95% CI: -0.63, -0.04 kg; P = 0.03) in diets containing dietary pulses
   Am J Clin Nutr, 2016 · checked 2026-09-24
7. bsal-r4-03 · systematic review and meta-analysis
   No significant overall effect between legume consumption and BP amelioration was observed in the meta-analysis (SBP-MD: -1.06 mm Hg; 95% CI: -2.57, 0.4410 mm Hg; I2 = 45%; DBP-MD: -0.48 mm Hg; 95% CI: -1.06, 0.10 mm Hg; I2 = 0%)...Significant subgroup differences in SBP were found when studies were grouped according to participant BMI, with SBP reduction found for participants with overweight/obese BMI (MD -2.79 mm Hg, 95% CI: -4.68, -0.90 mm Hg)
   Adv Nutr, 2023 · checked 2026-09-24
8. bsal-r4-06 · systematic review of laboratory evidence
   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 2026-09-24
9. bsal-r4-07 · systematic review of laboratory evidence
   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 2026-09-24
10. hf2-bsal-3 · analysis of reported outbreaks
    Between July 1976 and February 1989, 50 incidents of suspected red kidney bean poisoning were reported in the UK. Nine incidents in which nausea, vomiting and diarrhoea developed within 1-7 h of ingestion, were confirmed by the detection of haemagglutinin in the beans. The diagnosis was made on a further 23 incidents on the basis of symptoms, incubation time and the description of preparation of beans prior to consumption. ... In many of the outbreaks reported the implicated beans were consumed raw or following an inadequate heat process.
    Epidemiol Infect, 1990 · checked 2026-10-08
11. bsal-r4-05 · randomized feeding trials
    Less than 50% reported increased flatulence from eating pinto or baked beans during the first week of each trial, but only 19% had a flatulence increase with black-eyed peas...People's concerns about excessive flatulence from eating beans may be exaggerated.
    Nutr J, 2011 · checked 2026-09-24
12. bsal-r4-09 · systematic review of laboratory evidence
    Further, the effects of food processing on these proteins remain poorly understood and no conclusive data are available to quantitatively assess their safety after processing.
    Compr Rev Food Sci Food Saf, 2025 · checked 2026-09-24