Benefits
Body weight and fat mass
Reviews show mixed results. An independent meta-analysis of 8 placebo-controlled trials (543 adults with overweight or obesity) found about 1.6 kg more weight loss and 1.2 kg more fat loss with the extract. An earlier review of 6 flawed trials found the weight difference not statistically significant, and EFSA rejected a weight claim for one extract as unreplicated.
Smaller blood sugar and insulin rise after a starchy meal
In 12 healthy volunteers, 100 mg of a purified extract taken with a mixed meal blunted the 30-minute glucose rise and the insulin response. In a 13-person bread test, only a 3,000 mg powder mixed into the food lowered the glycemic index; capsules at 1,500 to 3,000 mg did not. An older study of a purified inhibitor also found smaller rises after 50 g of starch.
How much starch the extract actually blocks
Measured calorie blocking is the weak link. Early 1980s starch-blocker tablets did not change stool calories after a 100 g starch meal. A purified inhibitor infused straight into the small intestine stopped over 94% of amylase activity, but no study has measured how many starch calories a modern capsule keeps from being absorbed.
HbA1c, a 3-month blood sugar marker
In a placebo-controlled trial in China in adults with type 2 diabetes on insulin or sulfonylureas, 1.5 g of extract before each meal lowered HbA1c (a 3-month blood sugar marker) by 0.66 points versus 0.22 on placebo after 2 months. It is a single trial, and blood sugar changes on these medicines need a doctor's oversight.
Appetite and desire to eat after a meal
In the same 12-volunteer meal study, the extract kept the hunger hormone ghrelin from rebounding in the third hour after eating and lowered rated desire to eat compared with placebo. This was one small study of a single meal, so effects on daily food intake are unknown.
Mechanism of action
Alpha-amylase inhibition in the gut
Phaseolamin binds alpha-amylase, the salivary and pancreatic enzyme that splits starch into sugars. A purified bean inhibitor stayed active in human stomach and duodenal juices in the lab and, infused into the human duodenum, switched off most amylase activity there. How much a capsule achieves depends on its potency.
Slower starch digestion and a lower glucose load
With less amylase activity, starch is broken down more slowly, so less glucose reaches the blood soon after a meal. This is the proposed route to smaller post-meal glucose and insulin rises; the share of starch calories that escapes absorption altogether has not been measured for modern extracts.
Undigested starch reaches the colon
Starch that escapes digestion passes to the large bowel, where gut bacteria ferment it. In one trial, stool levels of Bifidobacterium and other fermenting bacteria rose with the extract. Fermentation is also the likely source of the gas some users notice.
Clinical trials
Double-blind, randomized, placebo-controlled trial of a branded white kidney bean extract at 1000 mg or 700 mg three times daily, 30 minutes before meals, during a calorie-restricted diet; sponsored and funded by supplement companies including the extract supplier, and two authors consult for the maker (Jäger et al. 2024, Sci Rep)
81 completers with overweight or moderate obesity, 12 weeks.
Body weight, fat mass, BMI and waist fell more than with placebo, with a larger effect at the higher dose. Tolerability was rated good; one case of gas was judged possibly related to the extract.
Randomized, double-blind, placebo-controlled trial of an extract from beans grown in southwestern China, taken before each meal on the usual diet (Wang et al. 2020, Food Sci Nutr)
120 adults with obesity enrolled (60 women, 60 men); 114 analysed.
Weight fell 2.24 kg with the extract versus 0.29 kg with placebo (p < 0.01). BMI, body fat, skinfolds, waist and hip also fell in the extract group. No adverse effects were observed. The trial lasted only 5 weeks.
Randomized, double-blind, placebo-controlled trial of a branded white bean extract taken with meals (Udani et al. 2004, Altern Med Rev)
50 adults with obesity screened, 39 completed screening and 27 completed the study.
Weight loss was 3.79 lb with the extract and 1.65 lb with placebo, a difference that was not statistically significant (p = 0.35). Triglycerides fell more with the extract, but this also missed significance (p = 0.07). No adverse events were attributed to the extract.
Randomized, double-blind, placebo-controlled meal study of a standardized, purified extract added to a balanced meal with 60% of energy from carbohydrate (Spadafranca et al. 2013, Br J Nutr)
12 healthy volunteers, with blood sampled for 3 hours after the meal.
Glucose rose 15.4% at 30 minutes with the extract versus 26.1% with placebo, and insulin and C-peptide rises were smaller. Ghrelin did not rebound as it did on placebo, and desire to eat was lower. A single-meal study in a small group.
Open-label, randomized crossover glycemic index test of a branded extract at 1500, 2000 and 3000 mg, supported by the manufacturer (Udani et al. 2009, Nutr J)
13 randomized subjects.
No capsule dose changed the glycemic index of white bread significantly, and the 1500 mg capsule had no effect at all. Only the 3000 mg powder mixed into butter on the bread lowered it significantly, by about 34%. Comparisons were not adjusted for multiple testing.
Randomized, double-blind trial of 1.5 g extract versus maltodextrin placebo 30 minutes before meals, with a 2-month intense phase and a 2-month maintenance phase (Feng et al. 2022, Front Endocrinol (Lausanne))
Adults aged 35 to 75 with type 2 diabetes treated with sulfonylureas or insulin in Wuxi, China; 90 randomized 2 to 1 (60 extract, 30 placebo); 83 analysed at 2 months (57 extract, 26 placebo) and 52 at 4 months.
HbA1c, the primary outcome, fell 0.66 points with the extract versus 0.22 with placebo at 2 months, and 0.72 versus essentially no change at 4 months. Stool Bifidobacterium increased. No adverse events were reported. Only volunteers willing to continue entered the 4-month phase, so that group was small and self-selected.
Placebo-controlled one-day calorie-balance study of commercial bean amylase-inhibitor tablets with a high-starch meal (Bo-Linn et al. 1982, N Engl J Med)
Healthy volunteers eating a 100 g starch meal of spaghetti, tomato sauce and bread.
Stool calories were the same on tablet and placebo days (80 versus 78 kcal), although full starch blocking should have added about 400 kcal. These older tablets did not block starch absorption; later work traced the failure to too little anti-amylase activity.