The Supplements That Do Nothing: A Research-Based Blacklist

How This List Was Built

Most supplement takedowns are written from vibes. This one is written from a rule, and the rule is worth stating up front because it decides every grade below.

A compound earns a failing grade here when randomized, placebo-controlled human trials in the relevant population have been run and the compound did not do the thing the label says it does. That last clause matters. Plenty of supplements work for something. The question is whether they work for the thing you are buying them for, in a body like yours.

Three distinctions do most of the work across this article, and once you internalize them the supplement aisle gets a lot smaller:

• Rodent evidence is not human evidence. Nearly every entry below has a beautiful mouse study behind it. Mice given raspberry ketones at doses equivalent to hundreds of milligrams per kilogram of bodyweight do interesting things to their fat cells. A human taking 100 mg in a capsule is not running the same experiment.
• Deficiency correction is not enhancement. Restoring a nutrient in someone who is genuinely short of it produces real, sometimes dramatic improvement. Adding more of that nutrient to someone who already has enough produces nothing. This is why zinc and vitamin D get an A for deficiency and a much lower grade as boosters, and it is the single most abused distinction in supplement marketing.
• Training status changes the answer. Untrained subjects improve on almost any intervention because they are improving on everything. A compound that shows an effect only in beginners has probably shown you the effect of beginning, not the effect of the compound.

Statistical significance also gets separated from practical significance throughout. Several compounds below produced results that are technically real and functionally worthless. A supplement that reliably removes 90 grams of fat a week is a supplement that loses to skipping one beer.

The Blacklist at a Glance

Grades reflect the strength and independence of human evidence for the marketed claim, not the enthusiasm of the label. Where a compound has a legitimate use outside the lifting context, the table says so.

The rest of this article is the evidence behind each row, in roughly the order you are likely to encounter these compounds in a store.

CLA: The 90 Grams a Week Problem

Conjugated linoleic acid is the most respectable compound on this list, which makes it the most instructive failure. It is a naturally occurring family of fatty acid isomers found in beef and dairy, it has a genuine mechanism involving fat cell metabolism, and it has been studied for decades. The rodent literature is genuinely impressive. Mice given CLA lose substantial fat mass.

The human literature is where it comes apart, and the definitive number comes from Whigham, Watras, and Schoeller, whose 2007 meta-analysis in the American Journal of Clinical Nutrition pooled 18 randomized, double-blind, placebo-controlled human trials. After adjusting to the median dose of 3.2 grams per day, CLA produced a fat mass reduction of 0.09 plus or minus 0.08 kilograms per week compared with placebo.

Sit with that figure for a second. Ninety grams of fat per week, with a standard deviation of 80 grams, meaning a substantial share of individuals in those trials got somewhere between nothing and a rounding error. Over a full year of continuous supplementation, assuming perfect linearity, which the authors explicitly note does not hold past about six months, you would be looking at under five kilograms. A modest calorie deficit does that in two months without a subscription.

The dose-response works out to roughly 0.024 kilograms of fat per gram of CLA per week, which is the kind of number that sounds meaningful right up until you multiply it by a realistic dose and a realistic adherence period.

The Isomer Problem

CLA is not one molecule. The two isomers that matter commercially are cis-9 trans-11 and trans-10 cis-12, and most fat-loss products lean on the second one because it is the one that drives the body composition signal.

It is also the one that caused problems. Riserus and colleagues published two trials in 2002 from the same subject pool of 60 abdominally obese men with signs of metabolic syndrome, randomized to 3.4 grams per day of a CLA isomer mixture, purified trans-10 cis-12 CLA, or placebo for 12 weeks. In Diabetes Care, the purified trans-10 cis-12 arm showed isomer-specific insulin resistance measured by hyperinsulinemic euglycemic clamp, the gold standard method. In Circulation, the same group reported isomer-dependent oxidative stress measured by urinary F2-isoprostanes along with elevated C-reactive protein. A follow-up in Diabetologia found hyperproinsulinemia closely associated with the impaired insulin sensitivity.

CLA is not a scam in the sense that raspberry ketones are a scam. The research is real, the compound does something, and the people who study it are serious. It just does something too small to care about and carries a tradeoff nobody prints on the tub.

Raspberry Ketones: Zero Human Trials

Raspberry ketone is the aromatic compound that gives red raspberries their smell. It became a fat-loss product almost entirely on the strength of a single daytime television segment in 2012 and a set of rodent studies showing effects on lipolysis and adiponectin.

Here is the entire human evidence base, stated plainly: there are no published randomized controlled trials of raspberry ketone as a single ingredient in humans for weight loss. Not weak trials. Not trials with mixed results. None.

The one randomized, placebo-controlled, double-blind human trial that gets cited in marketing gave 45 obese adults a multi-ingredient formula containing raspberry ketone alongside five other compounds already associated with weight loss, on top of a controlled diet and exercise program, for eight weeks. The test group lost more weight. Attributing any of that to the raspberry ketone specifically is not possible from that design, and the authors did not claim otherwise.

There is a second problem that applies to the whole rodent-derived category. The rodent studies that generated the interest used doses that, scaled to human bodyweight, land in the range of grams to tens of grams per day. Commercial capsules typically contain 100 to 500 mg. Even if the mouse mechanism translated perfectly, which mechanisms almost never do, the dose in the bottle is one to two orders of magnitude short.

A compound with a plausible mechanism and no human trials is not "promising" or "early stage." In a market where running a small human trial costs less than a single influencer campaign, the absence of trials after fourteen years of sales is itself the finding.

Garcinia Cambogia: Two Pounds and a Transplant Risk

Garcinia gummi-gutta, sold as garcinia cambogia, contains hydroxycitric acid, which inhibits ATP-citrate lyase, an enzyme involved in converting carbohydrate to fat. That mechanism is real and it is why the compound got taken seriously. It is also close to irrelevant, because de novo lipogenesis contributes very little to fat gain in humans eating a normal mixed diet. Blocking a pathway you barely use does not accomplish much.

The efficacy question was settled in 2011 by Onakpoya and colleagues in the Journal of Obesity. They identified 23 eligible trials, included 12, and pooled 9 that provided suitable data. The result was a mean difference of 0.88 kilograms favoring hydroxycitric acid over placebo, with a 95 percent confidence interval running from 1.75 kilograms down to essentially zero. The authors' own conclusion was that garcinia extracts can cause short-term weight loss but the magnitude is small and the clinical relevance is uncertain.

Roughly two pounds, in trials that were mostly short and mostly of modest quality, with the confidence interval touching nothing.

The Part That Actually Matters

Efficacy is the less interesting half of the garcinia story. The safety record is where it becomes a genuine warning rather than a shrug.

Garcinia was a component of Hydroxycut after its 2004 reformulation. In 2009 the FDA reported 23 cases of severe hepatotoxicity associated with the product, including one death and one liver transplant, and the manufacturer recalled it in May of that year. Hydroxycut was later reissued without garcinia extract.

The case reports did not stop with the recall. A 2025 review in Pharmaceutical Biology examining supplements containing Garcinia gummi-gutta identified more than 200 adverse liver events, including 34 published case reports of garcinia-associated hepatotoxicity, one death, and nine liver transplants. Individual case reports of fulminant hepatic failure requiring transplantation have appeared in the World Journal of Gastroenterology and elsewhere. Causality in any single case is difficult to establish because supplement users often take multiple products, which is precisely the argument the industry makes, and it does not survive 34 reports pointing the same direction.

Chromium Picolinate: A Meta-Analysis of One Study

Chromium is an essential trace mineral involved in insulin signaling, and the pitch is that supplementing it improves glucose handling, reduces cravings, and drives fat loss. Frank chromium deficiency is rare in people eating normal food.

Pittler, Stevinson, and Ernst published the relevant meta-analysis in the International Journal of Obesity in 2003. Ten randomized, double-blind, placebo-controlled trials with body weight data were pooled, covering 489 subjects. The headline result was a weighted mean difference of 1.1 kilograms favoring chromium picolinate, with a confidence interval from 1.8 down to 0.4 kilograms.

That looks like a win until you read the next sentence of the paper. Sensitivity analysis showed that the effect was largely dependent on the results of a single trial. Remove that one study and the pooled result falls apart. The authors' conclusion was that the effect is relatively small and of doubtful clinical significance.

This is worth pausing on as a general skill, because it applies well beyond chromium. When a meta-analysis reports a positive pooled result, the sensitivity analysis tells you whether that result is a consensus across the literature or one enthusiastic study dragging nine null ones behind it. Marketing copy quotes the pooled number. It never quotes the sensitivity analysis.

Chromium picolinate gets a D minus rather than an F because a kilogram of weight over several months is not literally nothing, and because chromium is cheap and generally well tolerated. It is still a rounding error against the effect of tracking your intake for two weeks.

Glutamine: Right Molecule, Wrong Person

Glutamine is the most abundant free amino acid in human plasma and skeletal muscle. It is a fuel source for enterocytes and immune cells, a nitrogen shuttle, and genuinely important. None of that means supplementing it builds muscle in a healthy lifter, and the trial that tested the proposition tested it hard.

Candow and colleagues randomized 31 healthy young adults aged 18 to 24 in a double-blind design to receive either glutamine at 0.9 grams per kilogram of lean tissue mass per day or placebo through six weeks of total-body resistance training. For an 80 kilogram lifter at 15 percent body fat, that dose works out to roughly 61 grams of glutamine a day, which is somewhere between 12 and 20 times what a typical tub recommends.

Both groups improved. Neither group improved more than the other on lean tissue mass, squat one-rep max, or knee extensor force production. The glutamine group's numbers were nominally slightly higher and the difference did not reach significance.

Glutamine is classified as conditionally essential, and that word is the entire explanation. Your body synthesizes plenty under normal conditions. It becomes limiting only under severe catabolic stress: major burns, sepsis, trauma, prolonged critical illness, some post-surgical states. That is where the credible glutamine literature lives, and it is real literature. Resistance training in a well-fed adult does not come close to that level of physiological stress, no matter how the session felt.

Glutamine also happens to be one of the most abundant amino acids in whey protein, so anyone drinking a shake is already getting a meaningful dose without a separate purchase. If you want the details on what is actually in your tub, the protein powder buyer's guide covers the label math.

HMB: It Works, Just Not on You

Beta-hydroxy beta-methylbutyrate is a metabolite of leucine, and it is the most scientifically defensible entry in this article. It has a plausible and specific mechanism, which is attenuation of muscle protein breakdown rather than stimulation of synthesis. It has a large body of human trials. Some of those trials are impressive.

The catch is that the trials split cleanly by training status, and the split is the whole story.

Rowlands and Thomson's 2009 meta-analysis in the Journal of Strength and Conditioning Research assessed HMB supplementation during resistance training on strength, body composition, and muscle damage in trained and untrained young men. Across eleven studies the finding was a small benefit to strength in untrained subjects and trivial gains in trained lifters, with trivial effects on muscle mass in both. Later meta-analyses that restricted inclusion to trained and competitive athletes, including a 2017 analysis in the Journal of Science and Medicine in Sport and a 2020 analysis in the Journal of Strength and Conditioning Research, arrived in the same neighborhood.

The mechanism explains the pattern perfectly. If HMB works by reducing muscle protein breakdown, it has the most to offer someone whose breakdown rate is elevated: a beginner whose muscles are being wrecked by unfamiliar loading, an older adult losing mass, a patient on bed rest, someone in a severe deficit. A trained lifter eating 1.6 to 2.2 grams of protein per kilogram has already pushed breakdown about as low as it goes. There is no headroom left for HMB to occupy.

HMB is the clearest illustration of the training-status rule. It is not fraudulent. It is simply sold to the one population it has been repeatedly shown not to help.

The Testosterone Shelf: Tribulus, DAA, ZMA, Antler

The testosterone booster category deserves a fuller treatment than a round-up can give, and the Tongkat Ali guide covers the one compound in that aisle with defensible human data plus the grading framework for the rest. This section extends that table with the four compounds that have been tested in resistance-trained men specifically and failed.

Tribulus Terrestris

Rogerson and colleagues published the cleanest test in 2007 in the Journal of Strength and Conditioning Research. Elite rugby league players took tribulus terrestris for five weeks during preseason training, a period of heavy, supervised, standardized loading. The result was no significant effect on strength, body mass, or fat-free mass, and no meaningful influence on serum testosterone concentration.

Tribulus does have documented effects on libido in some populations, likely through mechanisms unrelated to circulating testosterone. Subjective libido and serum testosterone are not the same outcome, and the label sells the second while the evidence supports at most the first.

D-Aspartic Acid

D-aspartic acid had the most promising start of any compound in this section. Early work in sedentary men showed testosterone increases, which is exactly the kind of finding that justifies a proper trial in the target population.

Willoughby and Leutholtz ran it. Resistance-trained men trained four times per week for 28 days while taking either 3 grams of D-aspartic acid or placebo. Body composition and muscle strength improved in both groups from the training, with no difference between them, and no changes in testosterone or the associated hypothalamic-pituitary-gonadal axis hormones.

Melville and colleagues then tested whether the dose was the problem, comparing 3 grams and 6 grams in resistance-trained men. The 3 gram arm did nothing. The 6 gram arm decreased total and free testosterone, with no compensating change in the other hormones measured. More was worse.

ZMA

ZMA is zinc monomethionine aspartate plus magnesium aspartate and vitamin B6. The original industry-affiliated study reported large testosterone increases in football players and launched the product.

Wilborn and colleagues at the Baylor University Exercise and Sport Nutrition Lab ran the independent replication and published it in the Journal of the International Society of Sports Nutrition in 2004. Forty-two resistance-trained males took either ZMA or a dextrose placebo 30 to 60 minutes before sleep across eight weeks of standardized resistance training. The result was no significant effect on total or free testosterone, IGF-1, growth hormone, cortisol, the cortisol to testosterone ratio, muscle or liver enzymes, one-rep max strength, muscular endurance, or body composition.

The important nuance is that the individual ingredients are not useless. If you are genuinely zinc deficient or magnesium deficient, correcting that is worth doing and will show up in how you feel and possibly in your bloodwork. The magnesium guide covers who actually falls short and by how much. Buying a proprietary blend of two minerals at boosted doses in the hope of a hormonal effect is a different proposition, and it is the one that failed.

Deer Antler Velvet

Deer antler velvet is sold on its IGF-1 content, and the pitch collapses on basic physiology before you get to the trials. IGF-1 is a 70 amino acid polypeptide. Swallowing a polypeptide subjects it to gastric acid and proteases, which digest it into amino acids like any other protein. Oral IGF-1 does not arrive in your bloodstream as IGF-1, which is why the injectable version exists and is a controlled substance.

The trials match the physiology. Sleivert and colleagues in 2003 randomized 38 active males to deer antler velvet extract, powder, or placebo through a ten-week strength program and found no significant endocrine changes in any group. Syrotuik and colleagues later gave 560 mg of antler velvet to rowers and found no alteration in serum growth hormone, cortisol, or any measured hormonal or physiologic response versus placebo.

Oral Arginine and the Pump Aisle

L-arginine is the direct substrate for nitric oxide synthase, which makes it the most intuitive "pump" ingredient imaginable. More substrate, more nitric oxide, more vasodilation, better pump and blood flow. The logic is clean and the biochemistry is correct.

It fails on delivery, for two reasons that compound each other.

First, oral arginine takes a heavy first-pass hit. Bioavailability sits around 60 percent, and a meaningful share of what survives absorption is metabolized in the intestine and liver before it reaches systemic circulation. Second, the enzyme arginase competes with nitric oxide synthase for the same substrate, and in healthy young people with already-adequate arginine levels, adding more mostly feeds the competing pathway.

The measured outcome bears this out. Reviews of the literature note that plasma nitrate and nitrite concentrations, the standard markers of nitric oxide production, are generally insensitive to L-arginine supplementation across durations from 1 to 28 days and doses from 6 to 12 grams. An acute supplementation trial published in Nutrition and Metabolism found no increase in nitric oxide production in healthy subjects.

Arginine draws a D rather than an F because there are populations where it does something, including people with endothelial dysfunction, and because high enough doses can nudge markers in some contexts. For a healthy lifter chasing a pump, the compound loses to its own precursor. L-citrulline bypasses first-pass metabolism, converts to arginine in the kidneys, and raises plasma arginine more effectively than arginine itself does. That mechanism and the supporting trials are covered in the L-citrulline guide, and the broader pre-workout ingredient list is graded in pre-workout supplements decoded.

The Ones We Already Took Apart

Three more compounds belong on this blacklist and have already been given full treatments elsewhere on the site. Short version and a link, rather than a rerun:

• BCAAs for muscle growth. Branched-chain amino acids in isolation cannot build muscle because muscle protein synthesis requires all nine essential amino acids, and providing three of them stimulates a process that then stalls for lack of the others. Anyone hitting a reasonable protein target is already consuming far more BCAAs from food than a scoop provides. The full teardown is in The Truth About BCAAs.
• Turkesterone and ecdysteroids. Marketed on rodent work and a single much-criticized human trial, with essentially no controlled human data supporting the anabolic claims, and independent testing has repeatedly found little or none of the labeled compound in commercial products. See Turkesterone and Ecdysteroids.
• Collagen for joint pain. This one is more nuanced than a flat F. There is a legitimate signal for collagen peptides plus vitamin C on tendon and ligament adaptation when timed before loading, and a much weaker case for the broad joint-health claims on most bottles. The distinction is drawn in Collagen Supplements: Joint Health Claims vs. the Actual Research.

Why They All Fail the Same Way

Eleven compounds, one recurring failure mode. Recognizing the shape of it is more useful than memorizing any individual verdict, because the next product will be a compound none of us have heard of yet and it will follow the same script.

Stage One: A Real Finding in the Wrong Species

Almost every entry above started with a legitimate result. Mice on CLA lose fat. Rodents on raspberry ketones show altered lipolysis. Hydroxycitric acid really does inhibit ATP-citrate lyase. The finding is not fabricated. It is simply produced in an organism with different metabolism, at a dose per kilogram that no human capsule approaches, over a lifespan that compresses years into weeks.

Stage Two: A Mechanism Becomes a Claim

The mechanism gets written into label copy, where "supports healthy metabolism" or "supports healthy testosterone levels" is legally permissible without any human outcome data. The phrasing is engineered. "Supports" is not "increases," and the disclaimer at the bottom of the bottle says the statements have not been evaluated by the FDA.

Stage Three: The Human Trial Runs and Nothing Happens

Eventually someone runs a proper randomized controlled trial in the population being sold to. The result comes back null, or so small that it is meaningless, or in the case of 6 grams of D-aspartic acid, backwards. This paper gets cited a few dozen times in the academic literature and zero times in marketing.

Stage Four: The Product Keeps Selling

Nothing in the US regulatory framework requires the product to change or come off the shelf. Sales continue on the strength of stage two. If a safety signal accumulates, as it did for garcinia, action arrives years and dozens of case reports later.

There is a fifth stage worth naming for people who buy supplements based on how they feel afterward: the effect you notice is usually not the compound. Starting any supplement typically coincides with training harder, tracking food, sleeping more deliberately, and paying attention. Placebo effects in exercise research are large and well documented. If you change five things and improve, the capsule gets the credit it did not earn. The broader pattern-matching toolkit for this is in Fitness Influencer Red Flags.

Where the Money Should Go Instead

The honest version of a blacklist has to say what passes, otherwise it is just cynicism. The list of supplements with strong, independent, replicated human evidence for a healthy lifter is genuinely short, and that is the point.

Creatine alone has more high-quality independent human evidence behind it than all eleven blacklisted compounds combined, and it costs less than most of them individually.

The larger point is that supplements are the last five percent of the picture and they cannot substitute for the first ninety-five. Total protein intake, energy balance, sleep, and consistent progressive training account for essentially all of the variance in outcomes between two lifters. A CLA subscription against a chronic four hours of sleep is not a close contest.

How to Vet the Next One in 90 Seconds

New compounds arrive constantly. Rather than waiting for someone to write an article about each one, run this sequence. It resolves most products before you finish the label.

1. Search for a human RCT in your population. Type the compound name plus "randomized controlled trial" plus "resistance trained" into PubMed. If nothing comes back in humans, you have your answer regardless of how good the mechanism sounds.
2. Check the subjects. Untrained beginners, sedentary older adults, and clinical populations improve on nearly everything. If the only positive trials are in those groups, the compound has not been shown to work for a trained lifter.
3. Find the effect size, not the p-value. "Statistically significant" tells you the effect probably was not chance. It says nothing about whether it matters. Convert the finding into kilograms, pounds on the bar, or percent, then ask whether you would notice it.
4. Read the sensitivity analysis in any meta-analysis. Chromium's entire pooled result rested on one trial. This is common and it is never in the marketing.
5. Check who funded it. Industry-funded studies are not automatically wrong, and they do skew positive. Look for independent replication, which is exactly what killed ZMA.
6. Compare the trial dose to the label dose. Rodent-derived products routinely under-dose by an order of magnitude relative to the studies used to sell them.
7. Look for a proprietary blend. A blend hides the dose of every ingredient in it, which means no dose can be compared to any trial. That is the function of the blend.
8. Require third-party testing. NSF, NSF Certified for Sport, USP Verified, or Informed Sport. Without it you do not know what is in the capsule, which matters more than efficacy for anyone who gets drug tested.

Common Mistakes

Treating a mechanism as a result. Every compound above has a mechanism. Mechanisms tell you what to test and nothing about what happened when someone tested it.

Confusing statistical significance with practical significance. CLA's fat loss is statistically real and worth 90 grams a week. Both facts are true and only one is relevant to a buying decision.

Generalizing from the wrong population. HMB in beginners, D-aspartic acid in sedentary men, and glutamine in burn patients all produced positive results. None of those subjects is you.

Assuming "natural" means safe. Garcinia is a fruit rind extract with a documented liver transplant record. Plant origin is not a safety argument.

Stacking blacklisted compounds and expecting the sum to work. Multiplying eleven null effects still yields nothing, but it does multiply cost, pill burden, and the number of ways a contaminated product can reach you.

Buying deficiency-correction supplements without checking for deficiency. Zinc, magnesium, iron, and vitamin D can all produce meaningful improvement in someone genuinely short of them and nothing at all in someone who is not. A blood panel costs less than a year of guessing, and in the case of iron and the fat-soluble vitamins, guessing carries real toxicity risk.

Judging a supplement by how you feel in week one. Starting a supplement almost never happens in isolation. Motivation, sleep, food, and training effort usually improve at the same time, and the capsule collects credit for all of it.

Frequently Asked Questions

Does CLA actually burn fat?

Technically yes, and the number is the problem. The 2007 meta-analysis by Whigham, Watras, and Schoeller in the American Journal of Clinical Nutrition pooled 18 randomized, double-blind, placebo-controlled trials and found that at the median dose of 3.2 grams per day, CLA produced a fat mass reduction of 0.09 plus or minus 0.08 kilograms per week compared with placebo. That is about 90 grams of fat a week with an error bar that nearly reaches zero. Separately, the trans-10 cis-12 isomer that drives most of the fat-loss signal is the same isomer that produced isomer-specific insulin resistance, elevated C-reactive protein, and increased oxidative stress in obese men with metabolic syndrome in the 2002 Riserus trials. You are paying real money and accepting a real metabolic tradeoff for an effect smaller than the daily noise in your own bodyweight.

Is glutamine a waste of money for lifters?

For building muscle in a healthy person, yes. Candow and colleagues gave 31 young adults 0.9 grams of glutamine per kilogram of lean tissue mass every day through six weeks of total-body resistance training, a dose far above what any tub recommends, and found no significant advantage over placebo on lean tissue mass, squat one-rep max, or knee extensor force. Glutamine is conditionally essential, which means healthy bodies make plenty of it and it only becomes limiting under severe physiological stress such as burns, sepsis, or major surgery. Those clinical settings, plus some gut-barrier and immune contexts, are where the credible glutamine literature lives. Bench pressing is not one of them.

Does HMB work if I already lift?

The training status of the subject decides the answer, which is why HMB generates such contradictory headlines. The 2009 meta-analysis by Rowlands and Thomson in the Journal of Strength and Conditioning Research found a small benefit to strength in untrained young men and trivial effects in trained lifters. Later meta-analyses restricted to trained and competitive athletes reached the same place. HMB is a metabolite of leucine that appears to work mainly by blunting muscle protein breakdown, so it has the most to offer someone whose muscle is being torn up by unfamiliar training, illness, or bed rest. A trained lifter eating adequate protein has already minimized that breakdown, so there is nothing left for HMB to rescue.

Are testosterone boosters like tribulus, D-aspartic acid, and ZMA worth trying?

No, and these three are the clearest failures in the category because they were tested properly and lost. Rogerson and colleagues gave elite rugby league players tribulus terrestris for five weeks during preseason and found no effect on strength, body mass, or fat-free mass, with no meaningful influence on serum testosterone. Willoughby and Leutholtz gave resistance-trained men 3 grams of D-aspartic acid daily through 28 days of heavy training and found no change in testosterone or performance beyond placebo, and Melville and colleagues later found that 6 grams per day actually lowered total and free testosterone. Wilborn and colleagues gave 42 resistance-trained men ZMA before sleep for eight weeks and found no effect on total or free testosterone, IGF-1, growth hormone, cortisol, strength, or body composition. Correcting a documented zinc or vitamin D deficiency raises testosterone. Adding these compounds on top of a normal panel does not.

Is garcinia cambogia dangerous or just useless?

Both, in a bad ratio. The 2011 systematic review and meta-analysis by Onakpoya and colleagues in the Journal of Obesity pooled nine trials and found a mean weight difference of 0.88 kilograms favoring hydroxycitric acid over placebo, which the authors themselves described as small and of uncertain clinical relevance. On the other side of the ledger, a 2025 review in Pharmaceutical Biology identified more than 200 adverse liver events linked to garcinia, including 34 published case reports of hepatotoxicity, one death, and nine liver transplants. Garcinia was also a component of the Hydroxycut formulation that the FDA linked to 23 cases of severe liver injury, including one death and one transplant, before its 2009 recall. Two pounds of theoretical weight loss is not worth a nonzero transplant risk.

Why do these supplements keep getting sold if the research says they do not work?

Because under the 1994 Dietary Supplement Health and Education Act, a supplement in the United States does not need to prove it works before it is sold. The manufacturer needs only to avoid claiming it treats a disease, add a disclaimer, and stay off the FDA's radar. Enforcement happens after harm is documented, which is why garcinia stayed on shelves through years of case reports. On top of that, most of these compounds have a genuine rodent result or a plausible enzyme mechanism behind them, and a mechanism is enough to write a label. The rodent-to-human failure rate in this space is close to total, but the rodent study is what gets quoted.

What should I buy instead?

The list of supplements with strong, independent, repeatedly replicated human evidence for lifters is short enough to memorize: creatine monohydrate at 3 to 5 grams daily, caffeine at 3 to 6 milligrams per kilogram before hard sessions, protein powder as a convenience food if you struggle to hit 1.6 to 2.2 grams per kilogram from whole food, vitamin D3 if you are actually deficient, and a small number of targeted deficiency corrections confirmed by bloodwork. Everything on the blacklist in this article combined has less evidence behind it than creatine alone. The money saved by skipping the blacklist covers a year of creatine several times over.

The Bottom Line

The supplement industry does not primarily sell you fraud. It sells you rodent studies wearing a human label.

Every compound on this blacklist has a real finding somewhere in its history. CLA works in mice. Hydroxycitric acid inhibits the enzyme it claims to inhibit. HMB reduces muscle protein breakdown, and in the right person that produces measurable strength gains. D-aspartic acid raised testosterone in sedentary men before anyone tested it in lifters. The science was never entirely made up. It was carried across a species boundary, a dose gap, or a training-status gap that it could not survive, and then printed on a label.

What separates the blacklist from the short list of things that work is not the presence of a mechanism. It is whether the compound was tested in a person like you, at a dose you can buy, against a placebo, by someone with no financial stake in the answer, and whether it produced a difference large enough to notice. Creatine clears every one of those bars. Caffeine clears them. Nothing in this article does.

The practical version is short. Stop buying the eleven compounds above. Take the roughly $60 to $150 a month that frees up and spend $6 of it on creatine. Put the rest toward food, a gym membership you will actually use, or nothing at all. Then go make the ninety-five percent of the picture better, because no capsule on any shelf competes with adequate protein, a controlled calorie target, seven to nine hours of sleep, and four years of showing up.

References

1. Whigham, L.D., Watras, A.C., & Schoeller, D.A. (2007). Efficacy of conjugated linoleic acid for reducing fat mass: a meta-analysis in humans. American Journal of Clinical Nutrition, 85(5), 1203-1211.
2. Riserus, U., Arner, P., Brismar, K., & Vessby, B. (2002). Treatment with dietary trans10cis12 conjugated linoleic acid causes isomer-specific insulin resistance in obese men with the metabolic syndrome. Diabetes Care, 25(9), 1516-1521.
3. Riserus, U., Basu, S., Jovinge, S., Fredrikson, G.N., Arnlov, J., & Vessby, B. (2002). Supplementation with conjugated linoleic acid causes isomer-dependent oxidative stress and elevated C-reactive protein: a potential link to fatty acid-induced insulin resistance. Circulation, 106(15), 1925-1929.
4. Riserus, U., Vessby, B., Arner, P., & Zethelius, B. (2004). Supplementation with trans10cis12-conjugated linoleic acid induces hyperproinsulinaemia in obese men: close association with impaired insulin sensitivity. Diabetologia, 47(6), 1016-1019.
5. Onakpoya, I., Hung, S.K., Perry, R., Wider, B., & Ernst, E. (2011). The use of Garcinia extract (hydroxycitric acid) as a weight loss supplement: a systematic review and meta-analysis of randomised clinical trials. Journal of Obesity, 2011, 509038.
6. Hepatotoxicity of dietary supplements containing Garcinia gummi-gutta (L.) N. Robson (2025). Pharmaceutical Biology.
7. Fong, T.L., Klontz, K.C., Canas-Coto, A., Casper, S.J., Durazo, F.A., Davern, T.J., Hayashi, P., Lee, W.M., Seeff, L.B., & Stickel, F. (2010). Hepatotoxicity due to Hydroxycut: a case series. American Journal of Gastroenterology, 105(7), 1561-1566.
8. Pittler, M.H., Stevinson, C., & Ernst, E. (2003). Chromium picolinate for reducing body weight: meta-analysis of randomized trials. International Journal of Obesity, 27(4), 522-529.
9. Candow, D.G., Chilibeck, P.D., Burke, D.G., Davison, K.S., & Smith-Palmer, T. (2001). Effect of glutamine supplementation combined with resistance training in young adults. European Journal of Applied Physiology, 86(2), 142-149.
10. Rowlands, D.S., & Thomson, J.S. (2009). Effects of beta-hydroxy-beta-methylbutyrate supplementation during resistance training on strength, body composition, and muscle damage in trained and untrained young men: a meta-analysis. Journal of Strength and Conditioning Research, 23(3), 836-846.
11. Sanchez-Martinez, J., Santos-Lozano, A., Garcia-Hermoso, A., Sadarangani, K.P., & Cristi-Montero, C. (2018). Effects of beta-hydroxy-beta-methylbutyrate supplementation on strength and body composition in trained and competitive athletes: a meta-analysis of randomized controlled trials. Journal of Science and Medicine in Sport, 21(7), 727-735.
12. Rogerson, S., Riches, C.J., Jennings, C., Weatherby, R.P., Meir, R.A., & Marshall-Gradisnik, S.M. (2007). The effect of five weeks of Tribulus terrestris supplementation on muscle strength and body composition during preseason training in elite rugby league players. Journal of Strength and Conditioning Research, 21(2), 348-353.
13. Willoughby, D.S., & Leutholtz, B. (2013). D-aspartic acid supplementation combined with 28 days of heavy resistance training has no effect on body composition, muscle strength, and serum hormones associated with the hypothalamo-pituitary-gonadal axis in resistance-trained men. Nutrition Research, 33(10), 803-810.
14. Melville, G.W., Siegler, J.C., & Marshall, P.W.M. (2015). Three and six grams supplementation of d-aspartic acid in resistance trained men. Journal of the International Society of Sports Nutrition, 12, 15.
15. Wilborn, C.D., Kerksick, C.M., Campbell, B.I., Taylor, L.W., Marcello, B.M., Rasmussen, C.J., Greenwood, M.C., Almada, A., & Kreider, R.B. (2004). Effects of zinc magnesium aspartate (ZMA) supplementation on training adaptations and markers of anabolism and catabolism. Journal of the International Society of Sports Nutrition, 1(2), 12-20.
16. Alvares, T.S., Meirelles, C.M., Bhambhani, Y.N., Paschoalin, V.M.F., & Gomes, P.S.C. (2011). L-Arginine as a potential ergogenic aid in healthy subjects. Sports Medicine, 41(3), 233-248.
17. Alvares, T.S., Conte, C.A., Paschoalin, V.M.F., Silva, J.T., Meirelles, C.M., Bhambhani, Y.N., & Gomes, P.S.C. (2012). Acute L-arginine supplementation increases muscle blood volume but not strength performance. Applied Physiology, Nutrition, and Metabolism, 37(1), 115-126.
18. Sleivert, G., Burke, V., Palmer, C., Walmsley, A., Gerrard, D., Haines, S., & Littlejohn, R. (2003). The effects of deer antler velvet extract or powder supplementation on aerobic power, erythropoiesis, and muscular strength and endurance characteristics. International Journal of Sport Nutrition and Exercise Metabolism, 13(3), 251-265.
19. Syrotuik, D.G., MacFadyen, K.L., Harber, V.J., & Bell, G.J. (2005). Effect of elk velvet antler supplementation on the hormonal response to acute and chronic exercise in male and female rowers. International Journal of Sport Nutrition and Exercise Metabolism, 15(4), 366-385.
20. Jagim, A.R., Harty, P.S., & Camic, C.L. (2019). Common ingredient profiles of multi-ingredient pre-workout supplements. Nutrients, 11(2), 254.
21. Dietary Supplement Health and Education Act of 1994, Pub. L. No. 103-417, 108 Stat. 4325.