Benefits
Post-meal blood sugar support
In a crossover study in 19 adults over 40 with or without pre-diabetes, drinking about 400 mg of leaf extract as a tea with 200 g of cooked rice lowered the blood glucose rise at 30, 90 and 120 minutes and cut the glucose area under the curve by about 20% versus hot water. In hospitalized diabetes patients it lowered the 2-hour reading, more mildly than the drug voglibose. These were small maker-run trials.
Fasting blood sugar and insulin measures with daily tea
In a 12-week study in 15 men with pre-diabetes or mild type 2 diabetes who drank the tea with each meal, fasting blood glucose edged down, insulin fell from about 9 to 7 microunits/mL and an insulin-resistance index (HOMA-IR) dropped. A separate 8-week study in medicated patients saw lower HbA1c in those whose starting value was raised. Both were small, maker-run and had no placebo group.
Blood cholesterol and triglyceride support
In 8-week studies of the leaf tea taken with meals, people who started with raised readings saw total cholesterol and triglycerides come down, and a blood fat-linked hormone, adiponectin, rose in those who began low. HDL cholesterol did not change much. They were seen in small, maker-run Japanese trials, some in people also on cholesterol or diabetes medicines, and have not been confirmed in independent placebo-controlled trials.
Carbohydrate-enzyme and glucose-transport effects in lab and animal studies
In test-tube work the leaf extract slowed the carbohydrate-digesting enzymes alpha-amylase, maltase and sucrase, and in cell and mouse studies it also reduced glucose transport across the gut lining. In several rodent models the extract lowered blood glucose after a sugar load and over weeks of feeding. These are mechanism and animal findings, not measures of outcomes in people.
Guava fruit is a separate vitamin C-rich food
The fruit, not the leaf, is a well-known source of vitamin C: in trials in Indian children, adding fresh guava to a meal raised the meal's vitamin C and improved hemoglobin and iron markers. Older trials of eating large amounts of guava fruit in a high-fiber diet reported lower blood pressure and blood fats, but those effects came from the whole fruit and diet, not from the leaf extract that supplies the blood-sugar evidence.
Mechanism of action
Slows carbohydrate-digesting enzymes
The water-soluble leaf polyphenol competitively slows alpha-amylase, maltase and sucrase, the gut enzymes that break starch and table sugar into absorbable glucose. In test-tube studies the extract inhibited these enzymes dose-dependently, with the strongest effect on alpha-amylase. This resembles how alpha-glucosidase drugs such as acarbose and voglibose work.
Limits glucose uptake across the gut lining
In human intestinal (Caco-2) cells and in mice, guava leaf and fruit extracts cut the transport of glucose from the gut into the body, acting more on the GLUT2 route than the SGLT1 route, and the effect held up after simulated digestion. In mice this was followed by a marked drop in blood glucose. These are laboratory and animal findings.
Flattens the after-meal glucose peak
By slowing both digestion and uptake of carbohydrate, the extract lowers the peak and the overall rise in blood glucose after a meal, which in turn asks less of the pancreas and can lower the matching insulin response. In a head-to-head human test its effect on the 2-hour reading was real but milder than the drug voglibose.
Low signal for cytochrome P450 drug interaction in the laboratory
Because the tea is taken by people on diabetes and cholesterol medicines, its maker tested it against drug-clearing liver enzymes. In test-tube work the leaf extract and tea were weaker inhibitors of CYP2C8, CYP2C9 and CYP3A4 than quercetin or grapefruit juice, and repeated dosing did not induce these enzymes in rats. Formal human interaction studies are still lacking.
Clinical trials
Review from the tea's manufacturer summarizing in vitro, animal and several small Japanese human trials of an aqueous guava leaf extract tea (Bansoureicha), an approved Food for Specified Health Uses in Japan; the underlying human trials are Japanese-language papers not indexed in PubMed (Deguchi and Miyazaki 2010, Nutr Metab (Lond)).
Human trials described include 19 adults with or without pre-diabetes (single-serving crossover), 20 hospitalized type 2 diabetes patients (tea versus voglibose), 15 men with pre-diabetes or mild diabetes over 12 weeks, 22 medicated diabetes patients over 8 weeks, and 23 people with borderline or mild high blood fats over 8 weeks.
A single serving with a rice meal cut the post-meal glucose area under the curve by about 20%; versus the drug voglibose the tea lowered the 2-hour reading more mildly. Over 8 to 12 weeks of daily use, fasting glucose, insulin, HOMA-IR and HbA1c (in those with raised starting values) fell, as did total cholesterol and triglycerides in people with high readings, with adiponectin rising in those who began low. No hypoglycemia occurred alongside diabetes or cholesterol medicines. The trials were small, maker-run and mostly lacked a placebo group.
Laboratory study in human intestinal (Caco-2) cells and in mice of guava leaf and fruit extracts on glucose transport and post-meal glucose (Müller et al. 2018, Mol Nutr Food Res).
Caco-2 cell cultures and C57BL/6N mice; no human participants.
Both leaf and fruit extracts inhibited glucose transport by up to about 80% in cells, with a larger effect on GLUT2 than SGLT1, and the activity survived treatment with digestive juices. In mice the extracts significantly lowered blood glucose in a time-dependent way (about a fourfold reduction). A mechanism and animal study, not a human trial.
Animal study of aqueous and ethanol guava leaf soluble-solid extracts in a streptozotocin-plus-nicotinamide rat model of type 2 diabetes (Shen et al. 2008, Phytother Res).
Sprague-Dawley rats with chemically induced type 2 diabetes.
Both acute and several-week feeding of the leaf extracts lowered blood sugar, and long-term feeding raised plasma insulin, glucose use and the activity of liver enzymes of glucose metabolism (hexokinase, phosphofructokinase, glucose-6-phosphate dehydrogenase). An animal study, cited here for mechanism; it does not measure outcomes in people.
Animal study of a guava leaf aqueous extract (50 to 800 mg/kg) in normal and diabetic rats and in salt-sensitive hypertensive rats, with chlorpropamide as the reference drug (Ojewole 2005, Methods Find Exp Clin Pharmacol).
Normal and streptozotocin-diabetic rats, and Dahl salt-sensitive hypertensive rats; no human participants.
The extract lowered blood glucose in a dose-related way in normal and diabetic rats, and intravenous doses lowered blood pressure and heart rate in hypertensive rats; the blood-pressure effect was not blocked by atropine. These are animal findings only, and the human leaf evidence does not cover blood pressure.
Randomized, single-blind controlled trial of eating guava fruit before meals, in a foods-to-eat high-fiber approach, for 12 weeks in people with high blood pressure (Singh et al. 1992, Am J Cardiol).
120 patients with essential hypertension (61 guava group, 59 control).
Adding guava fruit shifted the diet toward more fiber and less saturated fat and was followed by net falls in total cholesterol (about 10%), triglycerides (about 8%) and blood pressure (about 9/8 mmHg) and a rise in HDL cholesterol. The effects are attributed to the whole fruit and the fiber-rich diet, not to the leaf extract; cited here to keep the fruit separate from the leaf.
Seven-month randomized controlled trial of adding fresh guava fruit (about 170 mg vitamin C) to a mungbean meal in schoolchildren (Rani et al. 2024, J Nutr).
200 children aged 6 to 10 years in rural North India, 46% anemic at the start.
Adding guava raised hemoglobin and lowered the rate of anemia compared with the meal alone, though body iron stores did not rise significantly. This trial uses the fruit for its vitamin C, which helps iron absorption; it is included to show the fruit is a separate vitamin C food, not to support a blood-sugar claim for the leaf.