African Mango

Also sold as: Irvingia gabonensis, Dika Nut, Ogbono, IGOB131

A West and Central African food seed sold as a 150 mg twice-daily extract. Its founding trial reports 12.8 kg of weight loss in 10 weeks with no prescribed diet and no prescribed exercise — roughly a 1,400 kcal daily deficit produced by 300 mg of powder. The size of that number is the reason this page exists.

How African Mango actually works

The seed of *Irvingia gabonensis*, traditionally ground as a soup thickener, and the thickening is the clue: what is measurable in the kernel is soluble fiber, which slows gastric emptying and glucose appearance the way any viscous fiber does. The published claims go much further — leptin down, adiponectin up, adipogenesis suppressed through PPAR-gamma and C/EBP — but no molecule in the extract has ever been isolated, named, or measured in plasma in any species, so there is no compound to attach that mechanism to.

Where to get African Mango

The evidence for African Mango

Graded by what exists behind each claim.

✅ Clinically validated

• The 2009 trial randomized 102 overweight adults to 150 mg of IGOB131 or placebo twice daily for 10 weeks and reported significance on nine endpoints out of nine: body weight down 12.8 kg against 0.7 kg, waist down 16.19 cm against 5.3 cm, leptin down 48.6%, adiponectin up 159.8%, and LDL cholesterol down 27.3% to a group mean of 59.8 mg/dL.
• A systematic review found three randomized trials, all with flaws in the reporting of their methodology, all reporting significant weight and waist reductions — and concluded that the effect is unproven and that Irvingia gabonensis cannot be recommended as a weight-loss aid.
• An independent group gave the same 300 mg daily dose for 12 weeks after a 12-week run-in and found no difference between groups in metabolism, inflammation, leukocyte telomere length or aerobic capacity. What separated was plasma vitamin C and a 3-week adiponectin rise.

📊 Correlative data

• For scale: pooling 27 randomized trials and 1,428 participants, isolated soluble fiber for at least 12 weeks produced -1.25 kg with high certainty, and psyllium at a mean 10.8 g/day produced -2.1 kg. Pooled soluble fiber did not significantly change serum leptin or adiponectin at all.

🧪 Theoretical / extrapolated benefits

• The plausible mechanism is viscosity — 100% dika kernel powder has the highest water absorption and swelling capacity of the ogbono ingredients — and the dose excludes it. 0.3 g cannot thicken a stomach when the fiber trials use 10 g.
• The mechanism actually claimed is transcriptional, through PPAR gamma, leptin, adiponectin and glycerol-3-phosphate dehydrogenase. No molecule in this extract has ever been isolated, named, synthesized or measured in plasma in any species.

How to read these tiers: they say how much human evidence exists, not how well something works — and ✗ flags harm, never a disappointing trial. How the evidence tiers work →

What African Mango actually does

There are 2 candidate mechanisms for this seed, they make opposite predictions about dose, and the product is sold at a dose that fits only one of them. Getting that straight is the whole mechanistic content of this page.

Start with what the seed is. Irvingia gabonensis kernel — dika nut, ogbono — is a West and Central African food, not a botanical extract by origin. Its dominant macronutrient is fat: a hard, largely saturated kernel fat used as a cooking fat and as a lipid matrix in formulation work. Fed to young Wistar rats at 0, 5.1, 7.34 and 13.48% of the diet for 3 weeks, it raised serum cholesterol and triglycerides significantly at the highest level, at p below 0.01, though the cholesterol rise sat in the high-density lipoprotein fraction and low-density lipoprotein receptor abundance did not change Nangue 2011. The second property is rheological: ground dika kernel is the thickener in ogbono soup, and in a formulation study the 100% dika kernel powder had the highest water absorption, swelling capacity and bulk density of every sample tested, which the authors attributed to its crude fiber content and low moisture Bamidele 2015.

So the plausible mechanism is on the record and it is viscosity. A ground seed that absorbs water and swells behaves as a viscous soluble fiber. In the small intestine a viscous gel raises chyme viscosity and slows the degradation and absorption of nutrients, which is exactly how psyllium produces its effects on glycemia, serum cholesterol and body weight Gibb 2023. This is a real mechanism with a measurable size, and the measurable size is what creates the problem below.

The mechanism actually claimed for the product is a different one. IGOB131 is sold on a transcriptional story: an in vitro adipogenesis result said to run through peroxisome proliferator-activated receptor gamma, leptin, adiponectin and glycerol-3-phosphate dehydrogenase, and the human trial cites that work as its rationale Ngondi 2009. That mechanism does not need the seed to be viscous at all. It needs some molecule in a 150 mg extract to be absorbed, reach adipose tissue and change gene expression. No such molecule has ever been isolated or named, and no pharmacokinetic study of any constituent of this extract has been published in any species.

Here is where the 2 mechanisms separate, and it is arithmetic. A viscosity effect is a grams-per-meal effect: the psyllium meta-analysis pooled a mean dose of 10.8 g/day Gibb 2023, and the soluble-fiber trials pooled for weight run in the same range Huwiler 2022. A receptor or transcriptional effect can work at 150 mg. This product is dosed at 150 mg twice daily Ngondi 2009, which is 300 mg a day — about one thirty-sixth of the psyllium dose. Three hundred milligrams of powder cannot raise the viscosity of a stomach containing half a liter of fluid. Whatever this extract does, it does not do it by thickening a meal.

Which leaves the mechanistic position exactly here: the plausible mechanism is excluded by the dose, and the claimed mechanism has no identified molecule, no measured absorption and no target engagement data. That is not a hostile summary. It is the complete list of what is known about how this is supposed to work.

Cell, rodent, human — and where it stops

The trial. One hundred and two overweight or obese volunteers, body mass index above 25 kg/m2, randomized to 150 mg of IGOB131 or matching placebo twice daily, 30 to 60 minutes before lunch and dinner, double-blind, for 10 weeks. Significant improvements were reported against placebo in body weight, body fat, waist circumference, total cholesterol, low-density lipoprotein cholesterol, blood glucose, C-reactive protein, adiponectin and leptin — 9 endpoints out of 9 Ngondi 2009.

Now the effect sizes, because they are the review. Mean body weight fell from 97.9 to 85.1 kg on the extract and from 96.4 to 95.7 kg on placebo: 12.8 kg against 0.7 kg in 10 weeks. Waist circumference fell 16.19 cm against 5.3 cm. Leptin fell 48.6% against 9.3%. Adiponectin rose 159.8% against 23.4%. Total cholesterol fell 26.2% and low-density lipoprotein cholesterol fell 27.3%, to a group mean of 59.8 mg/dL Ngondi 2009.

Why an implausibly large effect size is itself evidence, which is the thing this category never explains. Do the energy arithmetic. Losing 12.8 kg of body tissue is on the order of 12.8 times 7,700, or about 98,600 kcal. Spread over 70 days that is a sustained deficit near 1,400 kcal a day — in a trial where no dietary change and no exercise were prescribed, produced by 300 mg of seed powder. Fourteen hundred kilocalories a day is more than most people achieve on a supervised diet; it is in the range of the first months after bariatric surgery. It is also 13.1% of starting body weight in 10 weeks, where the best-supported obesity pharmacology of the last decade takes about 68 weeks to reach 15%. A result that size is not a strong version of a plausible result. It requires an energy flux that has to come from somewhere countable, nothing in the trial accounts for it, and at that point the effect size stops being evidence about the product and starts being evidence about the study.

The dispersion says the same thing more quietly. Baseline leptin is reported as 32.9 plus or minus 1.6 ng/mL across 102 overweight adults Ngondi 2009. Serum leptin varies several-fold between people at the same body mass index and runs roughly twice as high in women as in men at matched fat mass, so a standard deviation of 1.6 ng/mL across a mixed cohort is narrower than the between-person biology of the analyte. The body-weight standard deviation also contracts from 9.1 to 3.1 kg over the 10 weeks, which is the opposite of what a real intervention does: real interventions produce responders and non-responders, and dispersion widens.

Replication is the next question and the answer is that there has not been any. When the randomized evidence was systematically reviewed, 3 trials existed. All 3 had flaws in the reporting of their methodology, all 3 reported statistically significant reductions in body weight and waist circumference favoring the plant, and the reviewers concluded that the effect on body weight is unproven and that Irvingia gabonensis cannot be recommended as a weight-loss aid Onakpoya 2013. Two of the 3 came from the same group: the 2009 IGOB131 trial Ngondi 2009 and a 2008 trial of a Cissus quadrangularis/Irvingia gabonensis combination in 72 participants over 10 weeks, in which both active arms separated from placebo on all 6 measured variables by week 10 Oben 2008. Six for 6, then 9 for 9. A literature in which every endpoint in every trial from one group reaches significance is a literature that has not been independently reproduced, whatever any individual trial reports.

The independent test, at the same dose. An unrelated group monitored body weight for 12 weeks, then randomized overweight and obese adults to 300 mg/day of Irvingia gabonensis kernel extract or placebo for a further 12 weeks — the same 300 mg total daily dose. What separated from placebo was plasma vitamin C at 6 and 12 weeks and serum adiponectin at 3 weeks. There were no differences between groups in metabolism, inflammation, leukocyte relative telomere length or aerobic capacity, and no between-group body-weight result is reported Nonsa-Ard 2022. Same dose, longer study, and a 12.8 kg signal does not appear.

And here is what a real, mechanism-appropriate effect looks like. Pooling 27 randomized controlled trials and 1,428 participants, isolated soluble dietary fiber supplementation for at least 12 weeks in overweight and obese patients produced a mean difference in body weight of -1.25 kg, 95% confidence interval -2.24 to -0.25, with high certainty of evidence Huwiler 2022. Psyllium at a mean 10.8 g/day for a mean 4.8 months produced -2.1 kg, 95% confidence interval -2.6 to -1.6 Gibb 2023. One to 2 kg, from grams of fiber, with tight intervals. The claim under discussion is roughly 10 times that from 300 mg.

The adipokine claim has its own benchmark and fails it the same way. Pooled across randomized trials, soluble fiber supplementation did not significantly change serum adiponectin, standardized mean difference -0.49, 95% confidence interval -1.20 to 0.21, or leptin, standardized mean difference -0.8, 95% confidence interval -1.70 to 0.08. The only significant subgroup was a small leptin reduction in overweight and obese participants, standardized mean difference -0.22 Zeinabi 2023. A 159.8% rise in adiponectin is not a larger version of that. It is off the scale the proposed mechanism operates on.

African Mango — which form, and does it matter

Ask this product the question that decides every other botanical on the shelf — standardized to what? — and there is no answer. A Coleus product declares 10% forskolin. A milk thistle product declares 80% silymarin. An African mango product declares 'Irvingia gabonensis seed extract, 150 mg' and stops: no extraction ratio, no solvent, no marker compound, no fiber content. There is no accepted marker compound for this plant because the active molecule has never been identified Ngondi 2009.

IGOB131 is a supplier code, not a specification. It names who made the material for the 2009 trial. It does not state a constituent, a ratio or an assay, which means a second manufacturer cannot reproduce it and a third-party laboratory cannot verify it. Reviews of the obesity supplement category list this ingredient alongside the others without being able to say what it is standardized to either Bonetti 2022.

That matters more here than usual, because the 2 candidate mechanisms live in different fractions of the same seed. The swelling and water-absorbing behavior is in the crude-fiber fraction of the ground kernel Bamidele 2015; the kernel's mass is largely a saturated fat Nangue 2011; and an extract concentrated for either of those would look nothing like an extract concentrated for the other. Two capsules both reading 'Irvingia gabonensis seed extract 150 mg' can be a defatted mucilage powder and an ethanolic extract of the same seed, and nothing on either label distinguishes them.

The pharmacokinetics, stated as the negative that it is. No constituent has been named, so no constituent has a measured oral bioavailability. No first-pass metabolism has been characterized, no gut-wall or hepatic clearance step has been described, no cytochrome P450 or glucuronidation pathway has been proposed, and no half-life exists to quote. That is not a gap in one study; it is the state of the entire literature. What can be said with arithmetic is the negative one: 300 mg a day is 0.3 g, a viscous-fiber effect on chyme requires a dose 2 orders of magnitude larger — 10.8 g/day in the psyllium pooling Gibb 2023 — and a gel capable of slowing nutrient absorption cannot be formed by 0.3 g dispersed in 500 mL of gastric fluid. Any mechanism that survives that arithmetic has to be a small-molecule one, and there is no small molecule.

The one form with a plausible mechanism at a plausible dose is the food. Ogbono is prepared with grams of ground kernel per bowl, which is in the right range for viscosity, and it arrives with the kernel fat Nangue 2011 Bamidele 2015. Nobody has run a trial on it. There is no injectable form, no isolated molecule to inject, and no reason to expect one — which is itself the honest summary of where this pharmacology stands.

What would have to be true, and how you would know it was not

1. The replication, and it is the decisive one. One hundred adults, 150 mg twice daily before meals, 10 weeks, no dietary prescription, an independent group, weight on a calibrated scale and fat by dual-energy X-ray absorptiometry. Predict a between-group body weight difference of 0 to 1.5 kg — the range the high-certainty soluble-fiber pooling reports Huwiler 2022. If a 12.8 kg difference reappears Ngondi 2009, this page is wrong and the result is one of the most important findings in obesity medicine.

2. Leptin and adiponectin at 4 and 10 weeks, with fasting insulin and HOMA-IR alongside. Predict adiponectin moves by tens of percent at most, if at all, in line with the pooled fiber data Zeinabi 2023, and predict fasting insulin is the endpoint most likely to move if anything does. Leptin tracks adipose mass, so predict any fall in leptin is proportional to the fall in fat mass — a 48.6% leptin fall alongside a 6.3-point fall in body fat percentage Ngondi 2009 is a relationship to check, not a result to accept.

3. The lipid panel at 10 weeks, and this is the cheapest falsification available. Predict total cholesterol and low-density lipoprotein cholesterol fall by the single-digit percentages a few grams of soluble fiber would produce, not by 26.2% and 27.3% Ngondi 2009. A group mean low-density lipoprotein of 59.8 mg/dL in an untreated overweight cohort is a statin result. If 300 mg of seed extract reproduces it, the finding belongs in a cardiology journal and not in a weight-loss aisle.

4. hs-CRP at baseline and 10 weeks. Predict it tracks fat mass and does not fall independently of it. C-reactive protein is largely produced by hepatocytes under interleukin-6 drive and interleukin-6 comes substantially from adipose tissue, so an independent fall in hs-CRP without a matching fall in fat mass would imply a direct anti-inflammatory action nobody has proposed a mechanism for Ngondi 2009.

5. The bench test that costs almost nothing and settles the mechanism. Disperse a full day's dose, 300 mg, in 250 mL of simulated gastric fluid at 37 degrees and measure apparent viscosity against 10 g of psyllium in the same volume Gibb 2023. Prediction: the difference from fluid alone is not detectable. If it is, the viscous-fiber mechanism is live at the retail dose and the central argument of this page collapses in an afternoon.

What nobody has tested yet

Nobody has named the active molecule. Two decades of trials, a proprietary extract code, a transcriptional mechanism cited as the rationale Ngondi 2009, and no compound has been isolated, identified, synthesized or dosed on its own. Every other ingredient in this category can at least name the thing it is standardized to.

Nobody has measured the absorption of anything in the extract. There is no plasma concentration of any constituent, after any dose, in any species, in the published literature. That single absence makes every dose recommendation for this ingredient arbitrary.

Nobody has run the dose-response that separates the 2 mechanisms. Three arms — 300 mg of extract, 3 g of defatted kernel powder and 10 g of defatted kernel powder — one body-weight endpoint, 12 weeks. If the effect is a fiber effect it must rise with dose Huwiler 2022; if it is a small-molecule effect it should already be maximal at 300 mg Ngondi 2009. One study distinguishes them and it has never been done.

Nobody has replicated the trial at its own primary endpoint. The closest independent attempt used the same 300 mg daily dose for 12 weeks and reported no between-group difference in metabolism, inflammation, telomere length or aerobic capacity Nonsa-Ard 2022, and the systematic review found the 3 existing trials too poorly reported to pool Onakpoya 2013.

Nobody has published the participant-level data from the 2009 trial. The contracting standard deviations are the reason to want them Ngondi 2009, and a dataset would answer in a week what another decade of commentary will not.

And nobody has tested the food. The traditional preparation delivers grams of the swelling fraction per meal Bamidele 2015, which is the only version of this ingredient whose mechanism and dose agree with each other, and it has never been the intervention in a trial.

African Mango — its own safety story, not its category's

The reported harms are mild and the reporting is thin. Across the randomized trials that have been systematically reviewed, the adverse events noted were headache and difficulty sleeping Onakpoya 2013. Flatulence, bloating and loose stool belong to any fiber-containing seed material and are dose-related.

The risk specific to this product is not toxicity. It is unverifiability. A material with no named active constituent and no marker compound cannot be assayed by anyone — not by a manufacturer, not by a third-party laboratory, not by a regulator. Identity testing can confirm plant species; it cannot confirm that the extract is the extract that was studied Ngondi 2009. On the shelf next door, when somebody did apply chromatography and mass spectrometry to a botanical supplement category with a defined marker, they found substitution with entirely different plants. An ingredient that cannot be assayed at all is a harder problem than one that fails an assay, and that is stated here as reasoning rather than as a measurement, because the measurement does not exist.

The interaction to know, and its real size. Anything viscous taken with a medicine can delay or reduce that medicine's absorption, which is the standard advice to separate fiber from levothyroxine, from antibiotics and from narrow-therapeutic-index drugs by 2 hours. At 300 mg a day that advice is close to moot for the reason given in the form section — 0.3 g cannot thicken a stomach Gibb 2023 — and it becomes real if somebody eats the food rather than the capsule Bamidele 2015.

The lipid claim cuts the other way, and this is the practical risk. If a 300 mg capsule genuinely lowered low-density lipoprotein cholesterol by 27.3% Ngondi 2009, that is a drug-sized effect that would need to be managed alongside statin therapy. Because it is unreplicated Onakpoya 2013 Nonsa-Ard 2022, the real hazard runs the opposite direction: substituting an unverified seed extract for a treatment with demonstrated cardiovascular outcomes.

Two specifics, briefly. Irvingia gabonensis is a tree seed and anyone with tree-nut allergy should treat it as one, since no cross-reactivity data exist either way. And the kernel fat is a saturated fat that raised serum cholesterol and triglycerides in rats at 13.48% of the diet over 3 weeks, with the rise falling in the high-density lipoprotein fraction Nangue 2011 — which is irrelevant to a 300 mg capsule and relevant to anyone eating the traditional preparation daily, the one form this page says has a plausible mechanism. There are no controlled human data in pregnancy. Nothing here is medical advice, and none of these statements has been evaluated by the Food and Drug Administration.

Sources read for this page

• Ngondi JL. IGOB131, a novel seed extract of the West African plant Irvingia gabonensis, significantly reduces body weight and improves metabolic parameters in overweight humans in a randomized double-blind placebo controlled investigation.. Lipids Health Dis 2009 · PMID 19254366
• Oben JE. The use of a Cissus quadrangularis/Irvingia gabonensis combination in the management of weight loss: a double-blind placebo-controlled study.. Lipids Health Dis 2008 · PMID 18377661
• Onakpoya I. The efficacy of Irvingia gabonensis supplementation in the management of overweight and obesity: a systematic review of randomized controlled trials.. J Diet Suppl 2013 · PMID 23419021
• Nonsa-Ard R. Effects of Irvingia gabonensis Extract on Metabolism, Antioxidants, Adipocytokines, Telomere Length, and Aerobic Capacity in Overweight/Obese Individuals.. Nutrients 2022 · PMID 36364907
• Huwiler VV. Prolonged Isolated Soluble Dietary Fibre Supplementation in Overweight and Obese Patients: A Systematic Review with Meta-Analysis of Randomised Controlled Trials.. Nutrients 2022 · PMID 35807808
• Gibb RD. Psyllium is a natural nonfermented gel-forming fiber that is effective for weight loss: A comprehensive review and meta-analysis. J Am Assoc Nurse Pract 2023 · PMID 37163454
• Zeinabi A. Soluble Fiber Effect on Human Serum Leptin and Adiponectin: A Systematic Review and Dose-Response Meta-Analysis.. Clin Nutr Res 2023 · PMID 37969941
• Nangue TJ. Irvingia gabonensis fat: nutritional properties and effect of increasing amounts on the growth and lipid metabolism of young rats wistar sp.. Lipids Health Dis 2011 · PMID 21375740
• Bamidele OP. Physico-chemical properties of instant ogbono (Irvingia gabonensis) mix powder.. Food Sci Nutr 2015 · PMID 26288723
• Bonetti G. Dietary supplements for obesity. J Prev Med Hyg 2022 · PMID 36479472

How you would know if it worked

African Mango — safety & side effects

Bloodwork to run alongside African Mango

Baseline first, then again at 8–12 weeks.

The Metabolic Health & Prediabetes panel covers these in one order — 8 markers, $81.45 with the discount applied.

Check results you already have → · All 103 markers A–Z

African Mango — frequently asked questions

What African Mango is used for

African Mango appears under 1 goal in the goal router.

Related Metabolic & Weight supplements

← Browse the full Supplement Vault

↑ Back to on this page