Story
Weight-loss supplements debunked · what works, what's a scam
Last updated
In one pass Among weight-loss supplements, even the best performers only nudge the number on the scale. Not this — Meal replacement shakes have weight-loss magic — Meal replacements are portion-control tools, nothing more. Stop using them and the weight returns; a pooled analysis (Heymsfield 2003) found no advantage over an ordinary diet with the same calories.
Educational content, not medical advice — consult a clinician.
Story path
Chapter 1
Why supplements needn't prove they work
Among weight-loss supplements, even the best performers only nudge the number on the scale. A serious change in diet plus exercise can take off several times as much, and the prescription drugs that really work go another order of magnitude further.
So why do these products keep selling? In the United States, supplements sit in the food column of the law, not the drug column: a maker does not have to show that a product works, or that it is safe, before putting it on the shelf, and regulators can only pull it after someone has been harmed. So the shelf proves only one thing — that someone was willing to make it. That is how the trade grew to more than $30B a year worldwide.
This story does not take them down one by one. It gives you a ruler: where exactly a supplement's claim breaks.
Some weight-loss supplements have been linked to liver and heart harm. If, while taking one, your skin or the whites of your eyes turn yellow, your urine turns the color of strong tea, or you get chest pain or a racing heart, stop it and see a doctor. For chest pain that does not ease, or sudden weakness on one side or slurred speech, call your local emergency number now.
So why do these products keep selling? In the United States, supplements sit in the food column of the law, not the drug column: a maker does not have to show that a product works, or that it is safe, before putting it on the shelf, and regulators can only pull it after someone has been harmed. So the shelf proves only one thing — that someone was willing to make it. That is how the trade grew to more than $30B a year worldwide.
This story does not take them down one by one. It gives you a ruler: where exactly a supplement's claim breaks.
Some weight-loss supplements have been linked to liver and heart harm. If, while taking one, your skin or the whites of your eyes turn yellow, your urine turns the color of strong tea, or you get chest pain or a racing heart, stop it and see a doctor. For chest pain that does not ease, or sudden weakness on one side or slurred speech, call your local emergency number now.
Background · Why supplements skip proving they work
This trade stands on a law, not on a study.The 1994 US DSHEA law put dietary supplements in the food column: the FDA does not review them beforehand, a maker can sell without proving the product works or is safe, and a recall becomes possible only after harm has happened. Ephedra supplements walked exactly that path — years on the shelf, until enough sudden deaths had piled up, and a ban in 2004.
Hold on to what this rule reverses. Before a prescription drug goes on sale, the maker has to show it beats placebo. Before a supplement goes on sale, the maker only has to show it can be manufactured. The burden of proof points the other way.
So many of the fat-loss miracles on the shelf have never been tested in a (an RCT, where people are randomly split between the real product and a placebo and then compared) — not failed one, but nobody ran the trial at all. Under this rule, skipping the trial is entirely legal, and cheaper: run one, get a negative result, and you have created a fact you cannot mention.
That is also why the sales pitch has not changed in decades and looks the same everywhere in the world:
Natural means safe, and safe means effectiveNo dieting, no exerciseBurns stubborn fat, blocks the absorption of carbs and fatA celebrity lost 10 kg in 30 days
The pitch works because it goes around the only hard step. Losing weight is hard because it means changing how you eat and move, for a long time — the hard part is not knowing what to do, but making that choice again every day. A bottle of capsules promises exactly that: to skip the step. In the United States, using supplements to lose weight is common. That is not because people are gullible; it is because the skipped step really is hard.
Numbers · Every option on one ruler
Put everything on one ruler and the gap needs no argument.| Intervention | Weight lost in trials | Certainty of evidence |
|---|---|---|
| Any supplement | 0-2 kg | low to very low (mostly small, short trials) |
| Diet plus exercise (lifestyle) | 5-10 kg | high (many randomized trials) |
| Semaglutide (the STEP-1 trial) | about 15 kg | high (a large randomized trial) |
| Tirzepatide (the SURMOUNT-1 trial) | about 21 kg | high (a large randomized trial) |
| Bariatric surgery | 25-35% of body weight | high |
The time spans are not identical: the supplement and lifestyle figures mostly cover six months to a year, while both drug trials ran for more than a year.
What to keep from this table is not the numbers but the distance between them.
The supplement row is several times smaller than the lifestyle row, and an order of magnitude smaller than the prescription rows. And an order of magnitude feels like this: a change at the supplement level is drowned by your ordinary day-to-day swings — the same person morning and evening, before a drink of water and after, and the scale already jumps. A change at the prescription level is the kind a friend who has not seen you for six months notices at a glance.
So this story's conclusion is not every supplement is a scam. The more accurate version: even if their own best-case claims came true in full, they would not be worth the price. What you pay for is a change you cannot feel yourself.
One more thing to notice: the only row on the table that costs nothing (diet plus exercise) sits in the middle, not at the bottom.
Mechanism · Five steps every claim must finish
This page is what the story really wants to hand you: a ruler you can use again.For a supplement to actually make you lighter, it has to get through five steps in a row. Break any one and the earlier ones count for nothing — and marketing only ever talks about the first.
① A reaction in a dish — in a culture plate, the molecule really does meet an enzyme or a receptor and speed it up or slow it down. This is where every pitch starts, and it is the easiest step: in a dish you can use any concentration you like, with no stomach, no liver and no dilution in the blood.② It reaches the blood after you swallow it — it has to survive stomach acid and then cross the cells of the gut wall. Many plant molecules are stopped right here; most of the dose leaves with the stool and never enters your blood.③ It gets past the liver — what the gut absorbs does not go straight into general circulation. The portal vein first carries the whole batch to the liver. The liver is the body's customs post: its enzymes tag foreign molecules, cut them up and send them out. Some molecules go through once and leave only a trace in the blood.④ It reaches enough concentration at the target — what is left still has to get to where it is supposed to work: inside a fat cell, in the hypothalamus, on the gut wall. The dish concentration and the concentration reachable there often differ by several orders of magnitude.⑤ The body does not cancel it out — even if the first four steps succeed, your body is not a passive target. Absorb a little less energy and it can move a little less and make a little less heat; free a little more fat and it can ask for another mouthful. Body weight is a quantity that is regulated again and again, not an addition problem. And keep two things apart: many fat burners do make the body burn a little more fat for a while, but burning more for a while is not net fat loss, because you keep storing the fat you eat (Jeukendrup 2011).
Marketing's standard move is to say ① and let you hear ⑤. The ad line science shows it inhibits a fat-making enzyme is step one; what you hear, so I will get thinner, is step five. The three steps in between are skipped, and that is exactly where most ingredients die.
Read on with this ruler, and each ingredient will be marked with the step where it breaks: raspberry ketone breaks at ④, short not by a little but by tens of times; bitter orange is one of the few that clearly gets into the whole body and does something there, but what it does lands first on the heart and blood vessels, while the weight-loss evidence is contradictory and weak; Garcinia and white kidney bean get stuck between ① and ⑤; green coffee is the extreme case, where the human trial propping it up was itself retracted.
Chapter 2
Some effect, but too small
This group was not made up out of nothing: human trials have measured at least something for each of them. Some reached statistical significance in a (CLA, for example); some did not clear even that bar once the trials were pooled (green-tea extract trials outside Japan came out at −0.04 kg). But statistically significant and something you can feel are two entirely different things.
Statistical significance tells you one thing only: the difference is probably not chance. It says nothing at all about how big the difference is. Losing 5% of body weight is the usual bar for clinical value — for a 70 kg person, more than 3.5 kg. All four ingredients in this group fall below that line, and most fall below half of it.
In other words: even where they did move the scale, the movement is so small that on the day you weigh yourself you cannot tell whether it was the capsule or one glass of water fewer the day before.
Statistical significance tells you one thing only: the difference is probably not chance. It says nothing at all about how big the difference is. Losing 5% of body weight is the usual bar for clinical value — for a 70 kg person, more than 3.5 kg. All four ingredients in this group fall below that line, and most fall below half of it.
In other words: even where they did move the scale, the movement is so small that on the day you weigh yourself you cannot tell whether it was the capsule or one glass of water fewer the day before.
Safety · Why the liver takes the hit
Green-tea extract (, the most abundant catechin in green tea) — breaks at ④, and pays a price at ③.The Cochrane systematic review (Jurgens 2012) pooled 6 randomized trials from outside Japan (n=532) and got −0.04 kg, with a 95% (CI, the range the true effect very likely falls in) of −0.5 to 0.4 — in the authors' own words, statistically non-significant. Not a small effect but essentially zero, and a zero whose interval crosses zero at that.
(Japan has 8 more trials, too heterogeneous for Cochrane to pool, with single results ranging from −0.2 to −3.5 kg, reported separately. Treating that end of the range as green tea's effect is the most common way this topic gets cherry-picked.)
The half really worth learning is the other one: why tea is fine and capsules cause trouble.
Keep one general test in mind: an extract is not the food. It is a small part of the food, pulled out, dried, concentrated, and then often swallowed in one go on an empty stomach.
When you drink tea, EGCG is spread out — diluted in a lot of water, sipped across the day, entering the gut slowly along with the rest of the tea. In a capsule, the same molecule becomes one concentrated lump that floods the gut in a short window.
Then comes the key link, step ③ of the five: what the gut absorbs goes first not to the whole body but to the liver. The portal vein carries everything the gut took up to the liver as one batch, and liver enzymes decide whether to let it through, modify it or get rid of it. Usually that is protection — the liver stands between you and everything you swallow. But protection has a cost: it means liver cells face a higher concentration than anywhere else in your body.
So you get a pairing that looks odd at first: a weight effect small enough to ignore, and liver injury that is real. At high doses, more than 80 cases of acute liver injury have been recorded, a few needing a liver transplant. EFSA 2018 placed the risk signal at ≥ 800 mg a day (in supplement form), but the same opinion carries an equally important sentence: it could not identify any clearly safe supplement dose, and there are case reports below 800 mg too. So 800 is where the evidence is strongest, not a safety floor. It also states that brewed tea and concentrated extracts are different things. Following the mechanism, the reason the liver is the organ that gets hurt lies in that portal route.
If, while taking an extract like this, your skin or the whites of your eyes turn yellow, your urine turns the color of strong tea, your upper right belly hurts or you feel unusually exhausted, stop it and see a doctor promptly.
So green tea burns fat is at most a slight side benefit of drinking tea, not something worth buying as a supplement. The same molecule in a different form brought no bigger gain but a risk of a different order — a question worth asking of any other plant extract too.
efsa-2018-green-tea-catechins
Evidence · CLA acts, in the wrong direction
CLA (conjugated linoleic acid) — the textbook case of acting without helping.The Onakpoya 2012 pooled 7 randomized trials lasting at least 6 months: body weight −0.70 kg (95% −1.09 to −0.32), fat mass −1.33 kg (−1.79 to −0.86). Those are two different endpoints — calling about 1.3 kg over half a year "weight loss" quietly passes off the fat-mass line as body weight. The authors' own conclusion: the effect is small and its clinical relevance uncertain. The accompanying changes deserve more attention: some studies saw the "good" cholesterol carried in high-density lipoprotein () fall and the "bad" cholesterol carried in low-density lipoprotein () rise, meaning blood lipids moved the wrong way.
Here the mechanism lesson matters more than the numbers.
CLA is not an inert molecule; it really does something in the body. In animal models it changes where fat is stored, but one of the costs is fat starting to build up in the liver — the opposite of the story the marketing tells. That is an animal result; whether the same happens in people is not settled.
Put the two together and you have a second ruler: a molecule acting and a molecule acting in the direction you want are two separate questions. Supplement copy only ever answers the first. And pathways in the body are rarely one-way valves: block one destination for fat and it may go to another; change how fat is distributed in the body and blood lipids and the liver get redistributed along with it.
So when you read a sentence like it has been shown to affect fat metabolism, the question to ask is not is it true but in which direction, and who pays the cost.
Evidence · Was it the vinegar, or the study?
Apple cider vinegar — the break is not in the body but in the study design.A few trials found that drinking vinegar every day brought only a very small weight loss. Most of those trials were small and many were not blinded, so it is hard to separate how much of that little effect came from the vinegar and how much from diet changes happening at the same time.
Why does not blinded take so much weight out of a weight-loss study's conclusion?
Because weight loss is the result of behavior, and behavior is extremely sensitive to being watched. Imagine you signed up for a vinegar weight-loss study: you have to remember to drink it every day, weigh yourself on schedule, come back for visits, and you know someone is looking at your numbers. That alone is enough to make you think for two extra seconds at every meal. So whether the bit of weight that left the scale was the vinegar's doing, or taking part in a study, nobody can tell.
Blinding and a control group exist to fix exactly this: the control group goes through the identical routine and drinks an identical-looking cup every day, so the only difference left is the molecule. The gap between the two groups is what the molecule gets credit for.
Three questions to take with you. Next time a weight-loss pitch says a study proved it works, ask them in order:
Was there a control group? If not, you may only be seeing the effect of taking part in something.Did participants know which one they were taking? If they did, expectation alone changes behavior.Who did the weighing? Self-reported weight and weight measured by a researcher on the same scale at the same time are two different kinds of data.
These three questions need no nutrition knowledge, and they screen out most of the backed by research claims on the market.
Myth · Berberine borrowed a name
Berberine — it has real effects, just not on weight.It is mainly used to lower blood glucose and improve blood lipids; its effect on weight is weak (about 1-2 kg in ). Calling it nature's Ozempic is misleading — its weight effect trails that of real semaglutide (the ingredient in Ozempic) by 5-10 times. The full breakdown is in the berberine story (see berberine).
The marketing move itself is worth learning on its own.
In the name nature's Ozempic, the part doing the work is not berberine; it is Ozempic. It borrows the trust you already place in that other product: you do not have to be persuaded that berberine works, only to accept that it belongs in the same class as the thing that works.
And that same class is false. The two do not even act in the same place. One helps along the glucose-metabolism line. The other imitates — a fullness hormone your own gut releases when you eat, which at the same moment tells the pancreas to release insulin, the stomach to empty more slowly and the brain that you have had enough. The multiple between them is not the same job done a bit worse; it is not the same job at all.
The test: whenever a product is named after another product, assume it is borrowing credit, then look up what evidence exists under its own name. Treat anything sold as the natural version of X or the budget version of X the same way.
In practice · Same money: why walking wins
Why even this group is not worth buying.Spend the same money on 1,500 extra steps a day and 20 g more protein a day, and by a rough estimate the expected loss over 6 months is 3-5 kg — 2-3 times everything above added together, with no risk to your liver.
A little more walking and a little more protein are not magic at all. They win because they act in a different place.
A supplement acts at one binding site on one enzyme — small, specific, and liable to break at any of the five steps. Walking and protein act on behaviors you repeat many times a day: you take thousands of steps every day, and decide three times a day what to eat. A small change multiplied by that many repeats adds up to a sizable total.
That is also why a supplement's effect is measurable in a trial but invisible in daily life, while a change of habit is slow to show up in a trial but snowballs in daily life. The first is a one-off, over once it is swallowed; the second happens again every day.
So this story is not handing you the lecture do not buy supplements. It is handing you a conversion: the money you would pay for a bottle of capsules, turned into a pair of shoes that makes you want to go out, or into the extra protein in your kitchen, buys several times the expected weight loss — and buys muscle, heart and lung fitness and sleep along with it, which a capsule cannot even claim.
Chapter 3
Never shown to work in people
The problem with this group runs deeper than a small effect: some have barely been tried in people (raspberry ketone), some have been tried and landed near zero (Garcinia, chromium), and one has its effect claimed on its behalf by another ingredient in the formula (bitter orange). What the marketing does in every case is take a reaction seen in a dish or a mouse and present it as the effect you will get.
Raspberry ketone shows at once how this whole group works. The amount that worked in rodent experiments, scaled up to a person, means swallowing several grams a day; the bottle you buy holds a few percent of that. That is not a weak effect; that is never taking a dose that could work — and not even the doses on sale have ever been checked for safety in people.
The others each break in their own way: bitter orange does get into the whole body, but its effects land first on the heart and blood vessels; Garcinia and white kidney bean are stuck between inhibiting an enzyme in a dish and a person actually getting lighter; chromium keeps hovering around zero in clinical terms.
Raspberry ketone shows at once how this whole group works. The amount that worked in rodent experiments, scaled up to a person, means swallowing several grams a day; the bottle you buy holds a few percent of that. That is not a weak effect; that is never taking a dose that could work — and not even the doses on sale have ever been checked for safety in people.
The others each break in their own way: bitter orange does get into the whole body, but its effects land first on the heart and blood vessels; Garcinia and white kidney bean are stuck between inhibiting an enzyme in a dish and a person actually getting lighter; chromium keeps hovering around zero in clinical terms.
Evidence · Raspberry ketone fails on dose
Raspberry ketone — breaks at step ④: it never reaches a working concentration.One line on a US TV health show (Dr. Oz) in 2012, a miracle fat burner, made it famous overnight and sales explodedHuman trials: according to the summary by the US National Institutes of Health Office of Dietary Supplements (NIH ODS), there is only one randomized trial, and it used a multi-ingredient product, so raspberry ketone's own effect cannot be separated outThe rest of the supporting data come from rats and isolated fat cellsThe effective rat dose scaled to a person: about 90 mg/kg a day, which for a 70 kg adult means 6.3 g a dayCommercial supplements contain 100-200 mg, which is 1/30 to 1/60 of thatSafety: the safety of the doses on sale has never been evaluated in people
How that order-of-magnitude gap arises is worth spelling out — it turns up with almost every ingredient sold as proven in the lab.
In a dish, researchers soak fat cells directly in a liquid containing the molecule. They can use any concentration they like, with no stomach, no liver and no dilution in the blood. A cell twitching under those conditions is a low bar — at a high enough concentration, many molecules will make a cell do something.
In a living animal, rebuilding that concentration around the cells means feeding a much larger amount by mouth, because you have to subtract what is never absorbed, what the liver breaks down and what is diluted in the whole body's blood. That is why the rat dose looks so large per kilogram.
Scale the same ratio to a person and you get several grams of powder a day. No maker will sell it that way — a bottle would not last a few days, the price would look ugly, and nobody has ever tested what happens when a person swallows that much, so the maker would carry the liability if something went wrong. So the label carries a dose that is unlikely to cause harm and bound to fail: too small to do much harm, and too small to work.
This is the line to take away: that dose is not cautious. It was chosen to stay legal and sellable, and it never had anything to do with working.
A question to take with you: next time you see research confirms it works, first ask whether the amount used in that research is the same order of magnitude as the amount in the capsule in your hand. That one question screens out this whole group.
Safety · Why bitter orange pushes the heart
Bitter orange (synephrine, from Citrus aurantium) — one of the few ingredients in this story that clearly gets into the whole body and does something there. That is not good news.After ephedra was banned in 2004, it became the legal stand-inA similar mechanism: it stimulates adrenaline receptors, bringing central stimulation, a faster heart rate and higher blood pressureWeight effect: one gives about 1.3 kg over 6-12 weeks, but most of the trials added caffeine, so that small effect probably comes mainly from the caffeine rather than from synephrine itself; NIH ODS judges the evidence that it works to be contradictory and weakAdverse events: NIH ODS lists reports of chest pain, headache, anxiety and a faster heart rate; there are also case reports of hypertensive crisis, arrhythmia, stroke and heart attack (especially with caffeine), and case reports cannot tell you how often these happenSome sports bodies list it as a stimulant that athletes may not use or that is being monitoredThe risk outweighs the benefit
First, put stimulating adrenaline receptors into plain words.
When something startles you, your adrenal glands release a signal into the blood. That signal docks on receivers on many kinds of cells, and each kind of receiver does something different with it: heart muscle beats faster and harder; the smooth muscle in blood-vessel walls tightens and blood pressure rises; fat cells let go of stored fat, releasing fatty acids into the blood as a reserve.
That last one is the line the fat-burning pitch grabs, and it is not wrong — the signal really does loosen fat. Synephrine in bitter orange does something similar: it looks enough like that signal to dock on the same receivers.
The problem: you cannot send this signal to fat cells alone.
It travels in the blood, and the blood passes through every organ. There are receivers on the heart, on the blood vessels and in your brain, and all of them hear the same message. What you wanted was the fat around your waist letting go; what you actually sent was the whole body going into stress mode. Following the mechanism, those case reports of hypertensive crisis, arrhythmia, stroke and heart attack look less like this molecule's side effect and more like the same effect landing on other organs.
If you get chest pain or a heartbeat that is irregular or very fast after taking a weight-loss product containing bitter orange or caffeine, stop it and see a doctor. For chest pain that does not ease, or sudden weakness on one side or slurred speech, call your local emergency number now.
The second lesson is about who gets the credit. Almost every combination product has this shape: one cheap ingredient that everyone agrees does a little (caffeine is the favorite), plus an expensive star, and then the whole bottle's measured effect is credited to the star.
The test: when you see a combination label, look first for caffeine or anything else known to work. If it is there, the star gets no credit until someone shows the effect remains once that ingredient is taken out.
Evidence · Garcinia names an enzyme, not a pathway
Garcinia cambogia (its active ingredient is hydroxycitric acid, HCA) — it breaks at every link between ① and ⑤.The pitch: blocks a fat-making enzyme (-citrate lyase)The Onakpoya 2011 (*Journal of Obesity*; 12 randomized trials included, 9 with poolable data): a mean body-weight difference of −0.88 kg against placebo (95% −1.75 to −0.00). The authors' own conclusion: the effect is small and its clinical relevance uncertainThe most rigorous randomized trial was negative (Heymsfield 1998, *JAMA*: 135 overweight adults, both groups on a high-fiber, low-energy diet, 12 weeks, 1500 mg of HCA a day): the HCA group lost 3.2 kg and the placebo group 4.1 kg, a non-significant difference (p = 0.14) — the direction even ran the wrong wayLiver injury: of the 1,418 cases registered in the US Drug-Induced Liver Injury Network (DILIN) from 2004-2018, 22 were attributed to Garcinia (5 taken alone, 16 with green tea, 1 with ashwagandha); 91% needed hospital care, including 1 death and 1 liver transplant. A registry has no denominator, so this is a safety signal, not a rateThe FDA's 2017 notice is not what you might think: it targeted a product labeled as Garcinia that turned out to contain hidden sibutramine — a prescription weight-loss drug withdrawn in 2010 over cardiovascular risk. In other words, the truly dangerous ingredient in that batch was not on the label at all. That is not a one-off: in the FDA's database of adulterated products, 269 of 317 weight-loss products (84.9%) were spiked with sibutramine. Those 317 come from the 776 adulterated supplements, from 146 companies, that drew FDA warnings in 2007–2016: products already caught, not a random sample from the shelf, so 84.9% tells you what the caught products were spiked with, not how much of the market is spiked. A useful rule: if a weight-loss supplement works startlingly well, suspect first that it contains a drug
The shape of this pitch is worth studying: it gives you the name of a real enzyme.
ATP-citrate lyase exists, and it really is part of the route by which the body turns surplus sugar into fat. Naming it makes the whole sentence sound technical — and that is its job: a specific molecular name makes people assume the later steps have already been covered.
But between naming an enzyme and showing that blocking it makes you lighter lie those same five steps. Take them one at a time:
Can it reach the enzyme? The enzyme is inside cells. Swallowed HCA has to cross the gut wall, get past the liver, and then get inside fat cells and liver cells, losing some at every gate.Is the inhibition strong enough? A dish can be soaked at a high concentration; a living person cannot.How big a share does this route carry? The key point: the fat on your body comes by more than one route. One is the body turning surplus sugar into fat — that is the route this enzyme controls. The other is the fat you eat, which is already fat and can be stored directly after absorption without passing through this enzyme at all. Close one door and the other stays open.Will the body compensate? Sugar that does not become fat still has to go somewhere — burned, stored as glycogen, or making you slightly less hungry. The body does not stop just because one door has been shut.
Put the four questions together and you can see why the most rigorous randomized trial was negative: not because the enzyme does not exist, but because an enzyme existing is a long way from losing half a kilo of flesh.
As for liver injury — it is that portal route again: what you swallow passes the liver first, at high concentration. Something whose weight effect hovers around zero has still left 22 cases of liver injury in one registry, one of them fatal. There is no need to run that risk-benefit calculation twice.
If, while taking a product like this, your skin or the whites of your eyes turn yellow, your urine turns the color of strong tea, your upper right belly hurts or you feel unusually exhausted, stop it and see a doctor promptly.
onakpoya-2011-garcinia-metaheymsfield-1998-garcinia-rctvuppalanchi-2022-dilin-garciniatucker-2018-weight-loss-supplements
Mechanism · Where the blocked starch goes
White kidney bean extract (Phaseolus vulgaris, an α-amylase inhibitor) — breaks at ⑤: your body takes the energy back.The pitch: a carb blocker, so rice and bread will not make you gainIn a dish: it inhibits α-amylase, so less starch is digestedRandomized trials in people: a few small ones (n = 60-100) show 1-3 kg more lost than placebo over 12 weeks, with high heterogeneity; NIH ODS judges that it may have a modest effect, that the trials vary in quality, and that there are few safety concerns at up to 3000 mg a day for 12 weeksClinical relevance is marginal, long-term data are missing, and bloating is a common side effect (undigested starch reaches the colon)
First, where α-amylase works and what it does.
Starch is a long string of glucose units linked together, and a chain that long cannot be absorbed through the gut wall. The α-amylase in saliva, and the α-amylase the pancreas releases into the small intestine, has one job: cutting the long chain into short pieces, which the small intestine can then recognize and take into the blood. Inhibiting the enzyme means cutting a bit less.
Note that this does not make starch disappear; it leaves starch uncut. Uncut starch does not vanish. It moves on down and arrives in the colon.
The colon is home to huge numbers of bacteria that are glad to take over this undigested starch — they ferment it and make gas. So the common bloating side effect is itself a sign that the mechanism is working: the starch really was not absorbed in the small intestine; it was just handled somewhere else.
That is why the effect is so small. Following the mechanism, part of the energy you thought you blocked is taken over by bacteria in the colon, part of what they produce by fermentation can still be absorbed, and the body has no intention of throwing away what it can still collect.
There is one more ceiling: the inhibitor comes in a fixed amount, and your meal does not. Each capsule holds a set dose of inhibitor, and it does not grow because you ate two bowls of rice today instead of one. The more you eat, the smaller the share it can stop — so in the I ate too much moment, when you would most want it, it works least.
The test: for any product that claims to block absorption, ask where the blocked material goes. If the answer is to the colon, it has mostly just moved somewhere else, and handed you a belly full of gas on the way.
tucker-2018-weight-loss-supplements
Evidence · Chromium, and the script they share
Chromium picolinate:The NIH ODS summary: 0.5-1.1 kg more lost than on placebo over 8-26 weeks, of debatable clinical relevanceCase reports of liver and kidney toxicity at high dosesChromium itself has a story of its own
Put this whole group side by side and the same script plays out again and again:
① An effect in a dish or an animal is presented as an effect in people. The three steps in between (reaching the blood, getting past the liver, reaching a working concentration) are never mentioned, because mentioning them would give the game away.② One or two small studies are quoted over and over. The same paper appearing on dozens of product pages creates the impression of lots of research — when it is one piece of evidence photocopied many times.③ Large, independent, long-term randomized trials are mostly negative, or show effects below clinical relevance. Note the word independent: when manufacturer-funded and independent trials give different results, that split is itself information.④ Side effects are played down. Put a natural label on it and liver injury becomes individual variation, while palpitations become your metabolism kicking in.
You do not actually need to remember these ingredient names. A new batch comes along every few years — this year's star is next year's clearance item on the bottom shelf. What is worth remembering is the script, because the next ingredient will play it out in exactly the same order.
When you judge one, it is enough to hold it up against the five steps in Why supplements needn't prove they work: at which step does this claim run out of evidence? Almost always, at the first.
nih-ods-chromium
Chapter 4
Natural doesn't mean effective or safe
The word natural printed on a bottle carries no information.
It sounds as if it is saying two things — this works, and it is safe — but it says neither. Natural only tells you the molecule was extracted from something that was once alive, and that has nothing to do with what it will do once inside your body.
The cleanest counterexample is ephedra (which contains ephedrine): entirely natural, used for more than a thousand years, and genuinely effective for weight loss. It was banned precisely because it works — a signal that makes fat let go is also a signal that drives your heart.
At the other extreme, most products sold as natural fat burners have none of ephedra's effect and so, conveniently, none of its danger either, until safe is the only selling point left.
It sounds as if it is saying two things — this works, and it is safe — but it says neither. Natural only tells you the molecule was extracted from something that was once alive, and that has nothing to do with what it will do once inside your body.
The cleanest counterexample is ephedra (which contains ephedrine): entirely natural, used for more than a thousand years, and genuinely effective for weight loss. It was banned precisely because it works — a signal that makes fat let go is also a signal that drives your heart.
At the other extreme, most products sold as natural fat burners have none of ephedra's effect and so, conveniently, none of its danger either, until safe is the only selling point left.
Myth · Natural does not mean effective
Trap 1: natural does not mean effectiveAshwagandha (also called Indian ginseng): moderate data for stress relief and better sleep (1-2 randomized trials showing lower cortisol); the weight-loss evidence is very weak, with only 1-2 small studies (n < 50) showing 1-2 kg lost over 8 weeks, which is not conclusive evidenceIts fat-loss pitch rests on a syllogism: stress goes down, so cortisol goes down, so belly fat goes down. Not one of the middle steps has been tested in a randomized trialGreen coffee bean extract (mainly chlorogenic acid): a 2012 randomized trial by Vinson (n = 16) showed 8 kg lost over 22 weeks, but the study was retracted in 2014, and the marketing built on it also drew a fraud lawsuit from the US Federal Trade Commission (FTC); later independent trials did not reproduce the result
The syllogism claim is the shape supplement marketing loves most, and it is worth learning to spot it.
Look at the chain: take it and stress falls; stress falls and cortisol falls; cortisol falls and belly fat falls. Each link sounds reasonable on its own, and each can find its own paper to lean on. The trouble is in joining them up.
Think of each link as a probably. One probably, you might still bet on; three probablys strung together leave very little certainty, because a chain's strength is not an average but a product. Worse still, the last link (a small drop in cortisol taking waist fat down with it) has never been tested on its own. It is the weakest link in the whole argument, and the only one that actually matches your goal.
So the test is this: when a product's pitch reads A causes B, B causes C, and C causes the result you want, jump straight to the last link and ask — has anyone measured A against the result you want, directly? If someone has and found nothing, the middle links do not matter however elegantly they are told.
Green coffee is a problem of a different order. The ingredients above have weak evidence; for this one the evidence itself went wrong: the study was retracted. Yet notice that it is still doing fine on e-commerce sites today — when a paper is retracted, the ads that cite it are not retracted with it. Science's self-correction does not reach the shelf, and that time lag is the space this whole industry lives in.
Safety · Natural does not mean safe
Trap 2: natural does not mean safeEphedra: entirely natural, used in Chinese medicine for more than a thousand years. After the US FDA had received reports of more than 155 deaths (from arrhythmia, stroke and heart attack), it banned supplements containing ephedrine alkaloids in 2004Kava: a traditional Pacific-island drink, once banned or pulled from sale in Germany, France and Canada over liver toxicityBitter orange (synephrine): see Never shown to work in peopleGreen-tea extract (high-dose ): see Some effect, but too smallGarcinia: see Never shown to work in peopleThe shared lesson: plant does not mean the molecule is safe. Strychnine, tetrodotoxin and ricin are natural too.
Why does a natural molecule that really works almost always carry side effects? It is not coincidence; it is how the body is built.
A molecule that can affect weight does so by saying something the body understands — that is, by docking on a receptor. And receptors do not grow only on the patch of flesh you want to shrink. The same kind of receptor sits on the heart, the blood vessels, the gut and the brain, because the body already uses one message to direct several organs at once.
Once the molecule is in the blood, the blood carries it everywhere. The blood does not know an address called waist. You want it to speak only to the fat cells in your belly, but it says the same thing, equally, to every cell carrying that receptor.
So there is a trade-off you can hardly get around: the more forcefully a molecule speaks, the more strongly the other organs that hear it respond. Ephedra could make people thinner because it shouted loud enough; it could kill people because it shouted loud enough — the same thing.
Run that backward and you get a useful instinct: a natural product that claims a big effect with no side effects at all has probably not found some clever target; more likely it is doing nothing. No action, and so no side effect. Many of the gentle and safe fat-loss products on the shelf are selling exactly that kind of safety.
The last reminder is worth copying down: natural was never a safety grade. It is a note about where something came from.
Numbers · The order-of-magnitude gap
Trap 3: between natural and drug, the effect differs by an order of magnitude| Intervention | Weight lost in trials (kg) | Class |
|---|---|---|
| Ashwagandha (limited data) | ~1-2 | natural |
| Green-tea EGCG | −0.04 (not significant) | natural |
| CLA | −0.70 (body weight, not fat mass) | natural |
| Berberine | ~1-2 | natural |
| Semaglutide 2.4 mg | ~15 | prescription |
| Tirzepatide 15 mg | ~21 | prescription |
The rows come from trials of different lengths: the natural-supplement rows mostly cover a few weeks to half a year, while both drugs were tested in large randomized trials lasting more than a year.
The gap: the weight effect of the prescription drugs and of the natural supplements differs by 10-20 times. That is not natural and therefore gentle; it is natural supplements doing almost nothing for weight.
Watch how the word gentle gets slipped in.
What marketing wants you to hear is: drugs hit hard, natural is soft, so you can go at your own pace. That hides a premise — that the two travel the same road and differ only in speed. But the gap in the table is not a bit slower; it is barely moving.
Feel the gap another way. The numbers in the top half of the table sit inside a person's ordinary weight swings — the same person on different days, at different hours, before and after a drink of water, and the scale already moves by that much. So even if those effects were entirely real, you would have no way to confirm they happened to you. The bottom half of the table is the size your own trousers will tell you about.
That is why it worked for me is almost never believable for a supplement: at that effect size, a person has no way of telling. In any change you noticed, the season, your mood or a few extra walks may all have played a bigger part.
Mechanism · How a natural molecule becomes a drug
A more accurate picture:If a natural product really had a strong weight-loss effect, it would either carry obvious side effects (ephedra) or deserve to be developed into a prescription drug (paclitaxel and metformin both came from plants)There is currently no natural weight-loss agent that is both highly effective and free of side effects — if there were, it would long since have been developed into a drug itself is a natural hormone — scientists simply modified its structure so that it lasts about a week in the body, which allows one injection a week (the STEP-1 trial used once-weekly semaglutide). That is the proper route: a natural molecule plus rigorous development
That last point is worth opening up, because it states the whole story's position: the problem was never natural; the problem is untested.
That hormone was yours to begin with. When you eat, your gut senses food passing and releases it; it travels once around the bloodstream and speaks to three places at once — telling the pancreas it is time to release insulin, telling the stomach to empty more slowly, telling the brain that's enough. That is why, halfway through an ordinary meal, you slow down without trying.
The natural version has one peculiarity: as soon as it enters the blood an enzyme cuts it up, and within minutes it is gone. That is a sensible design for the body — the message should be sent after a meal, not all the time. But if you want to borrow that message to help someone lose weight, said and gone becomes the obstacle.
What scientists did was not invent a new signal but rework its structure so the scissors cannot cut it. The same message can then carry on for a long time. There is no unnatural magic here: it is the same message, docking on the same receptor, and the main difference is how long it keeps talking.
So both halves of that route are essential: a molecule taken from the body gives it a sound mechanism, and rigorous development and trials give it a believable effect size. The products on the shelf usually have only the first half, and the first half on its own produces nothing.
Chapter 5
What's actually worth your money
If you want to spend money on supplements, four things are worth buying: protein powder, caffeine, creatine and fish oil. But first be clear about the job they do: none of them burns fat.
During a fat-loss phase, when you take in less energy than you use each day, the body breaks down two things at once — fat and muscle. It does not pick only the part you dislike. And muscle is one of the body's more power-hungry tissues: the more you lose, the fewer calories you burn doing nothing; following that mechanism, the later stages get harder, and regaining weight after you stop may get easier.
So the real job of the first three is to make the body take apart more fat and less muscle, and to give you the strength to keep training. Fish oil is for the heart and blood vessels and has nothing to do with weight.
During a fat-loss phase, when you take in less energy than you use each day, the body breaks down two things at once — fat and muscle. It does not pick only the part you dislike. And muscle is one of the body's more power-hungry tissues: the more you lose, the fewer calories you burn doing nothing; following that mechanism, the later stages get harder, and regaining weight after you stop may get easier.
So the real job of the first three is to make the body take apart more fat and less muscle, and to give you the strength to keep training. Fish oil is for the heart and blood vessels and has nothing to do with weight.
In practice · Why protein matters in a deficit
① Protein powder (whey, casein or plant protein)Its core job: helping you reach a total protein intake of 1.6-2.4 g/kg a day, to hold on to muscle during a fat-loss phaseThe Morton 2018 (49 randomized trials, n = 1863, in healthy adults who were all doing resistance training, mostly not in a calorie deficit): resistance training plus protein supplements added 0.3 kg more lean mass (fat-free mass) and about 9% more strength gainLongland 2016 (a 4-week proof-of-principle trial in 40 young men with an energy deficit of about 40%): the group eating 2.4 g/kg of protein a day plus hard training gained 1.2 kg of lean mass while losing 4.8 kg of fatIt is not burning fat; it is keeping muscle while you cutHow to handle protein in a deficit has a story of its own (see Protein During a Deficit)
Why protein in particular? Because it is the only macronutrient the body has no dedicated store for.
Extra sugar is stored as glycogen and fat; extra fat goes straight into fat cells. Both have a warehouse and a stock. Protein is different: the body has no protein reserve tank. The amino acids circulating in the blood last only a short while, and once they run short, the only place the body can draw from is tissue already on duty — your muscles and organs.
So in a fat-loss phase the situation is this: you have opened a gap, and the body has to make it up from its stock. Fat has stock; muscle has no stock, only tissue in use. Eating enough protein every day keeps giving the body a reason not to break down muscle.
Resistance training is the other half, and you need both. Protein supplies the material; training supplies the signal that this muscle is still in use. The body's logic is simple: tissue that goes unused for a long time is expensive to maintain and gets cut first. Doing a few sets each week tells it again and again: keep this one.
That is why Morton's meta-analysis tested protein supplements alongside resistance training — eat without training and the material piles up unused; train without eating and you have a signal but no material. In Longland's trial, fat loss and lean-mass gain happened together because both were done together; it was a short trial in young men under a harsh deficit, showing that this can happen, not that everyone will get the same numbers.
One last time: protein powder has no magic over eggs, tofu, fish or lean meat. Its only advantage is convenience — on days when you have no time to cook and would otherwise fall short, it lets you make up the amount. That is all, but it is often enough.
In practice · Caffeine and creatine buy training volume
② CaffeineRaises basal metabolic rate (BMR, the energy you burn each day lying still) a little (about 3-11%), and improves exercise performance (endurance, strength and high-intensity work)As a source, coffee, tea or dark chocolate (which bring polyphenols) beat a plain caffeine pillYou do not need a supplement dose; 200-400 mg a day (1-3 cups of coffee) is enough
Caffeine does not give you energy; what it blocks is sleepiness. The longer you stay awake, the more a rest-signaling molecule builds up in your brain — adenosine — and as it docks everywhere you grow sleepier. Caffeine is shaped enough like it to take those docking sites first: the signal keeps building, but for a while it cannot be delivered. So what caffeine buys is not new strength but pushing tiredness back for a stretch.
That has two practical consequences. First, following the mechanism, the postponed signals do not vanish; when the caffeine wears off they arrive together, which is the familiar caffeine crash. Second, its direct contribution to weight loss is small enough to ignore. What is really valuable is that it helps you finish the workout.
③ Creatine (creatine monohydrate)
One of the best-supported nutrition supplements, with > 1000 randomized trialsImproves training performance and muscle mass and helps preserve muscle during fat loss, improving body composition indirectly3-5 g a day, no loading phase needed, and cheap (about $0.10 a day)
Creatine is stored in muscle as phosphocreatine, a kind of small-change purse for energy. In the first seconds of an all-out effort, muscle cannot draw energy slowly from sugar and fat, so it relies on this instantly available little pool. A slightly bigger pool means one or two more reps per set and a slightly faster recovery for the next set.
Notice the shape of this causal chain — it takes a detour, but every link is solid: creatine adds one or two reps per set, a week of that adds up to more training volume, more muscle is kept, and basal metabolism drops less during the fat-loss phase. It never touches fat from start to finish, yet it really does help body composition.
That is the biggest difference from the ingredients in the earlier chapters: they were sold as acting directly on fat, and the chain broke in the middle; the four items here honestly take the long way round, and every link can be checked. A long road that gets there is worth far more than a shortcut that does not.
Evidence · Worth buying is not the same as weight loss
④ Fish oil (omega-3 fatty acids, and )Its place is heart and blood vessels, inflammation and brain function, not fat lossThe REDUCE-IT trial: in people already on a statin, with raised and high cardiovascular risk, 4 g a day of highly purified EPA lowered major adverse cardiovascular events () by 25% in relative termsFor most people, a common supplement amount is 1-2 g of EPA plus DHA a day
It is in this chapter to sort out something that is easily confused: worth buying and helps you lose weight are two separate questions.
Fish oil has a solid place, but that place is not weight loss. Its strongest evidence comes from cardiovascular endpoints, and note the population in that trial — raised triglycerides and high cardiovascular risk. Swap in a healthy young adult and the same conclusion does not automatically hold.
That leads to a general way of reading that applies to all nutrition evidence: a study's conclusion is tied to its population and its endpoint. The same bottle can bring a clear benefit in one group and measure nothing at all in another. Ads drop the population and keep only the percentage.
So buy the four recommendations in this story for their own reasons: protein powder to keep muscle, creatine for training volume, caffeine to finish the workout, fish oil for the heart and blood vessels. None of them is bought for weight loss — and precisely because they do not pretend to cause weight loss, you can trust what else they say.
bhatt-2019-reduce-it
In practice · Where the same money pays off most
A real value-for-money ranking for weight loss (same money spent):1. Good food (quality protein, vegetables and whole grains, replacing ultra-processed food) — far ahead of any supplement
2. A gym membership, a pair of running shoes — strength plus aerobic work, two kinds of metabolic gain
3. Sleep (blackout curtains, earplugs, a cool bedroom) — people who sleep under 6 hours a night have about 1.5 times the risk of obesity (, RR 1.5; this is an observed association and does not show that short sleep is the cause)
4. A bathroom scale and a kitchen scale — self-tracking, a core tool for changing behavior
5. (When medically indicated) a drug — assessed by a doctor; $300-1300 a month, but an effect of 5-15 kg, so the cost per kilogram lost is far below that of supplements
Why does food come first, and not any bottle?
Because it works in a different place. A supplement acts at one binding site on one enzyme — small, specific, and it has to clear five gates before it even gets a turn. What you eat today is a decision you make several times a day and tens of thousands of times in a lifetime. A small change to that decision, multiplied by all those repeats, ends up bigger than any capsule.
The items further down are variations on the same idea. A pair of shoes that makes you want to go out changes how much you move each day. Earplugs and a blackout blind change how you sleep each night — and short sleep tends to push appetite up and the will to train down at the same time. A scale turns I think I've put on weight lately into an actual number, so you can correct course earlier and by less, instead of waiting until your trousers feel tight.
What these have in common: they do not enter your body; they change what you do every day. And weight loss is, from start to finish, mainly a matter of behavior — the body part was never the hard bit.
Related stories:
Weight Management — the underlying framework for weight lossGLP-1 agonists — the real drugs, as a comparisonProtein During a Deficit — protein strategy during a fat-loss phaseBerberine (see berberine) and chromium — each has a detailed breakdown of its own
Cut the $50-100 a month supplement budget and swap it for 2 kg more protein, a monthly gym pass and a good pillow. Six months later you will have lost more weight, and you will understand why.
References · 9
- Wilding, J. P. H., Batterham, R. L., Calanna, S., Davies, M., Van Gaal, L. F., Lingvay, I., et al. (2021). Once-weekly semaglutide in adults with overweight or obesity (STEP-1). New England Journal of Medicine, 384(11), 989-1002. 1961 adults with BMI >= 30 (or >= 27 with a weight-related condition) and no diabetes, randomized 2:1 to semaglutide 2.4 mg weekly or placebo, plus lifestyle intervention, for 68 weeks. Mean body-weight change -14.9% vs -2.4% (difference -12.4 percentage points); >= 15% loss in 50.5% vs 4.9%; -15.3 kg vs -2.6 kg; discontinuation for gastrointestinal events 4.5% vs 0.8%. Funded by Novo Nordisk (abstract, PMID 33567185). 10.1056/NEJMoa2032183
- Jastreboff, A. M., Aronne, L. J., Ahmad, N. N., Wharton, S., Connery, L., Alves, B., Kiyosue, A., Zhang, S., Liu, B., Bunck, M. C., & Stefanski, A. (2022). Tirzepatide once weekly for the treatment of obesity. The New England Journal of Medicine, 387(3), 205–216. 2539 adults with BMI >= 30, or >= 27 with a weight-related complication, people with diabetes excluded; randomized 1:1:1:1 to tirzepatide 5, 10 or 15 mg weekly or placebo for 72 weeks (20-week dose escalation). Mean weight change (treatment-regimen estimand) -15.0%, -19.5% and -20.9% vs -3.1%; >= 20% loss in 50% and 57% on 10 and 15 mg vs 3%. Adverse events led to discontinuation in 4.3%, 7.1% and 6.2% vs 2.6%. Funded by Eli Lilly (abstract, PMID 35658024). 10.1056/NEJMoa2206038
- Jeukendrup, A. E., & Randell, R. (2011). Fat burners: nutrition supplements that increase fat metabolism. Obesity Reviews, 12(10), 841-851. Reviews caffeine, carnitine, green tea, conjugated linoleic acid, forskolin, chromium, kelp and fucoxanthin; an acute rise in fat oxidation does not necessarily translate into fat loss. 10.1111/j.1467-789X.2011.00908.x
- Jurgens, T. M., Whelan, A. M., Killian, L., Doucette, S., Kirk, S., & Foy, E. (2012). Green tea for weight loss and weight maintenance in overweight or obese adults. Cochrane Database of Systematic Reviews, (12), CD008650. Pooled non-Japan studies (6 RCTs, n=532): mean difference -0.04 kg (95% CI -0.5 to 0.4), statistically non-significant. 10.1002/14651858.CD008650.pub2
- Onakpoya, I. J., Posadzki, P. P., Watson, L. K., Davies, L. A., & Ernst, E. (2012). The efficacy of long-term conjugated linoleic acid (CLA) supplementation on body composition in overweight and obese individuals: A systematic review and meta-analysis of randomized clinical trials. European Journal of Nutrition, 51(2), 127–134. 10.1007/s00394-011-0253-9
- National Institutes of Health, Office of Dietary Supplements. (2022). Dietary supplements for weight loss: health professional fact sheet (updated 18 May 2022). Ingredient-by-ingredient evidence review. Raspberry ketone: only one RCT of a multi-ingredient product, so its effect cannot be isolated, and the safety of the doses sold has never been evaluated in humans. Bitter orange/synephrine: efficacy evidence contradictory and weak, with reported chest pain, headache, anxiety and raised heart rate. Garcinia cambogia: effect on body weight remains uncertain; 10 published cases of liver toxicity. White kidney bean (Phaseolus vulgaris): possible modest effect, trials of varying quality, few safety concerns up to 3000 mg/day for 12 weeks. Chromium picolinate: 0.5-1.1 kg more than placebo over 8-26 weeks, of debatable clinical relevance. Green tea extract: any weight effect is small and unlikely to be clinically relevant, with growing human evidence of liver damage from extracts. ods.od.nih.gov/factsheets/WeightLoss-HealthProfessional
- Onakpoya, I. J., Wider, B., Pittler, M. H., & Ernst, E. (2011). Food supplements for body weight reduction: a systematic review of systematic reviews. Obesity, 19(2), 239-244. Overview of systematic reviews of food supplements marketed for body-weight reduction. 10.1038/oby.2010.185
- Morton, R. W., et al. (2018). A systematic review, meta-analysis and meta-regression of the effect of protein supplementation on resistance training-induced gains in muscle mass and strength in healthy adults. British Journal of Sports Medicine, 52(6), 376–384. 49 RCTs, 1,863 participants, resistance training of 6 weeks or more. Protein supplementation added 2.49 kg to 1RM and 0.30 kg to fat-free mass; the effect fell with age and was larger in trained people. Break point for FFM gains at 1.62 g/kg/day (95% CI 1.03-2.20; 42 study arms, 723 participants; the biphasic model was not statistically significant, p = 0.079); given the CI, the authors say ~2.2 g/kg/day may be prudent for those maximising gains; timing, post-exercise dose and source play a minor if any role; they cite per-dose MPS break points of 0.24 (younger) and 0.40 g/kg (older). One author reports grant support from the US National Dairy Council (abstract and full text, PMC5867436). 10.1136/bjsports-2017-097608
- Longland, T. M., Oikawa, S. Y., Mitchell, C. J., Devries, M. C., & Phillips, S. M. (2016). Higher compared with lower dietary protein during an energy deficit combined with intense exercise promotes greater lean mass gain and fat mass loss: A randomized trial. American Journal of Clinical Nutrition, 103(3), 738–746. 40 young men (20 per group), 4 weeks at a ~40% energy deficit with resistance training plus high-intensity intervals 6 days a week; 2.4 vs 1.2 g protein/kg/day. Lean body mass +1.2 ± 1.0 kg vs +0.1 ± 1.0 kg; fat mass -4.8 vs -3.5 kg; exercise performance improved similarly in both groups. The authors call it a proof-of-principle trial (abstract, PMID 26817506). 10.3945/ajcn.115.119339