Are beans good for you, and how do you make them easy to digest?

Beans, lentils, chickpeas and dried peas are called pulses. They deliver protein, slowly digested starch and fiber together, which is why they lower the glycemic index of a whole diet².

What the trials found. In 26 trials, a median of 130 g a day (about three-quarters of a cup cooked) lowered LDL cholesterol by 0.17 mmol/L, roughly 7 mg/dL¹. A pulse-rich diet also led to a small weight loss of about 0.34 kg over a median of six weeks, even when calories were not restricted⁹. In observational data, each extra 50 g a day went with 6% lower all-cause mortality³, and eating four 100 g servings a week went with 14% lower coronary heart disease¹⁰. Elderly people in Japan, Sweden, Greece and Australia had lower death rates the more legumes they ate¹¹; our Blue Zones article covers the wider pattern.

Why beans cause gas. They contain raffinose-family sugars that your small intestine cannot digest, so gut bacteria ferment them. Soaking for 16 hours, discarding the leftover soaking water and then cooking cut raffinose by 25%, stachyose by 25% and verbascose by 42% in common beans⁶. Starting small also helps. In three feeding studies of half a cup a day, fewer than half of people noticed more gas in week one, and 70% or more of those who did said it eased by the second or third week⁴.

Lectins and phytic acid. Raw red kidney beans contain a lectin that causes acute poisoning, but it is inactivated by thorough cooking of well-soaked beans⁵. Soaking alone barely changed lectin levels in Canadian pulses, while cooking was far more effective⁷. Sprouting does not solve it either: in white beans, soaking and four days of sprouting slightly raised measured lectin, and only cooking presoaked beans made it disappear¹². Soaking also lowered phytic acid, which binds minerals, by 85% in one bean study⁶. Our lectins and phytic acid articles explain the debate.

Canned or dried. Canned beans arrive fully cooked, so raw-bean poisoning is not a concern. The question is the can lining. In a randomized crossover trial, volunteers who ate a daily serving of canned soup for five days had urinary bisphenol A (BPA, a plastic chemical that acts like a hormone) more than ten times higher than after fresh soup. That trial used one soup brand in 2010, so check for BPA-free cans⁸.

Cholesterol. Ha and colleagues pooled 26 randomized trials (1,037 people, each at least three weeks) comparing pulse-rich diets with similar-calorie diets. At a median of 130 g a day, LDL fell by 0.17 mmol/L (95% CI 0.09 to 0.25), but apolipoprotein B and non-HDL cholesterol did not change, and the authors asked for longer, better trials¹.

Blood sugar and weight. A meta-analysis of 41 trials found pulses alone lowered fasting glucose (standardized mean difference -0.82) and insulin, and pulses in low-glycemic diets lowered glycated proteins such as HbA1c, but heterogeneity was high and mostly unexplained². For body weight, 21 trials (940 people) showed a loss of 0.34 kg (95% CI 0.04 to 0.63) at about 132 g a day. That is modest, and long-term effects are unknown⁹.

Heart disease and death. A meta-analysis of 25 observational studies and 2 trial reports (501,791 people) found legumes went with lower coronary heart disease (risk ratio 0.86 per four weekly 100 g servings), but not significantly lower stroke or diabetes¹⁰. A later dose-response analysis of 32 cohorts (1,141,793 people, 93,373 deaths) found all-cause mortality 6% lower at higher intakes (hazard ratio 0.94) and stroke mortality 9% lower, but no clear association with cardiovascular, coronary or cancer mortality, with low to moderate certainty³. The earlier Food Habits in Later Life study followed 785 people aged 70 and over for up to seven years and reported a 7% to 8% lower mortality hazard per 20 g a day of legumes; it is the smallest of these studies and adjusted only for age, sex, smoking and, in one model, ethnicity¹¹. People who eat more beans also tend to eat better overall, so confounding is likely.

Gas. Winham and Hutchins analyzed three feeding studies (17, 29 and 80 adults) of half a cup of pinto beans, baked beans or black-eyed peas daily for 8 to 12 weeks. In week one, fewer than half reported more gas with pinto or baked beans and 19% with black-eyed peas. By week two of the crossover trials this had fallen to 6% for pinto beans, 24% for baked beans and 12% for black-eyed peas. In the 12-week parallel trial, 45% reported more gas in week one, 30% in week three and 15% to 23% in weeks 6 to 12. Between 3% and 11% reported more gas even on control foods⁴.

Lectins and antinutrients. Between 1976 and 1989, the UK recorded 50 suspected red kidney bean poisonings. Nine were confirmed by finding the lectin in the beans, and others were diagnosed from symptoms appearing 1 to 7 hours after eating; in many cases the beans had been eaten raw or heated inadequately⁵. In peas, lentils, chickpeas, faba and common beans, soaking cut lectins by only 0.11% to 5.18% and had no effect on phytic acid, whereas cooking worked better on most antinutrients⁷. In three Tunisian white bean varieties, soaking or germination did not remove the lectin (measured by immunoassay), while cooking or autoclaving presoaked beans eliminated it and abolished its blood-clumping activity¹². The "no effect of soaking on phytic acid" result conflicts with the 85% phytate drop after soaking common beans⁶, which may reflect different beans and soaking conditions.

Cans. In 75 volunteers (a randomized, single-blinded crossover with five-day phases), urinary BPA averaged 20.8 micrograms per liter after daily canned soup versus 1.1 after fresh soup, a 1,221% increase⁸. It tested one brand of vegetarian soup, not beans, in healthy adults with a median age of 27.

Null and mixed results. Pulse diets did not change apolipoprotein B or non-HDL cholesterol¹, and blood-sugar benefits varied widely between trials². In cohorts, legumes were not clearly tied to stroke or diabetes in one analysis¹⁰, or to cardiovascular, coronary or cancer mortality in another³. Soaking results for phytic acid conflict between studies^6,7.

Bottom line (confidence 4 of 5). Beans likely lower LDL cholesterol and improve blood-sugar markers a little, and they are associated with modestly lower mortality. Effects are modest, most trials are short, and cohort data may be confounded. Digestive effects are real but usually temporary.