Chromium — Body Handbook

Supplements · §491

Chromium

You picked up the bottle because it said "supports healthy glucose metabolism" or "curbs sugar cravings" — and one Chinese trial from 1997, never replicated in any Western population, is doing most of the heavy lifting for that claim. Chromium is a trace mineral the body uses in microgram quantities; whether it's even essential is a live disagreement between US and European regulators, and the biochemist who first proposed how it works inside cells has since switched sides. The pooled effect on blood sugar across twenty-five trials is real but small — a third of what metformin does — and the body-composition story is honestly weaker than the marketing. The supplement is cheap and broadly safe; that's the kindest thing the evidence supports.

Know · Once Evidence Mixed Chapter Supplements

The honest read: cheap, mostly safe, marginal where it works, oversold everywhere else. A diabetic might see a small dent in blood-sugar numbers; a craving-prone eater might feel less drawn to the mid-afternoon sugar — both signals are real and both are small. For everyone else, taking it is mostly buying the sense that you're doing something. The cheapest experiment in the supplement aisle, also one of the easiest to leave on the shelf.

The cells of your muscle and liver carry a small protein called chromodulin that binds chromium. When your insulin receptor lights up after a meal, chromodulin docks onto it and — in test-tube preparations and in obese, insulin-resistant rats — amplifies the receptor's signal, helping more glucose move from the bloodstream into the cell. That mechanism is the story sold on every chromium bottle, and it is a real story in those preparations.

The trouble is it has never been shown to actually happen this way inside a living human at the chromium levels people eat. John Vincent — the biochemist who first purified chromodulin in the 1990s — now writes that the in-vivo evidence isn't there, and that modern animal nutrition cannot reliably produce a chromium-deficiency state at all on diets that aren't contaminated with traces of other minerals Vincent 2017. The mechanism is plausibility, not proof.

One more thing worth knowing about the supplement form: picolinate is a chelating molecule the chromium rides into your cells on, and the picolinate itself is bioactive — it binds zinc, iron, and copper and crosses cell membranes more readily than free chromium would. Several of the effects sometimes attributed to "chromium picolinate" actually trace to the picolinate ligand, not the metal it's carrying.

What the trials actually show

The trial doing most of the work for chromium is from Beijing in 1997 — 180 adults with type 2 diabetes, given either placebo, 200 micrograms a day, or 1,000 micrograms a day of chromium picolinate for four months. The high-dose group's three-month blood-sugar average dropped from 8.5% to 6.6% and fasting glucose fell about 17%. Those numbers are metformin-tier, and they have shaped the supplement market ever since.

The trouble is replication. American researchers gave Western type 2 diabetics comparable doses for comparable periods and found nothing — no insulin-sensitivity gain, no blood-sugar drop Cefalu et al. 2002. A Dutch group ran the same experiment in obese, poorly-controlled diabetics, also with no effect Kleefstra et al. 2007. The most-charitable explanation is that the Beijing diet was so low in chromium to begin with that supplementing fixed a real deficit, and Western diets are not deficient enough for that fix to matter. The less-charitable one is that the Anderson trial was an outlier; the NIH's own systematic review rated nearly half of the twenty trials it included as poor-quality and concluded the overall evidence in type 2 diabetes is inconclusive Costello et al. 2016.

What does survive a meta-analysis is small. Pooling twenty-five trials, the most-cited recent review found chromium dropped HbA_1c by about half a point on average — a real effect, concentrated in the picolinate form at doses above 200 μg/day Suksomboon et al. 2014. For context, metformin lowers the same number by a full point or more; the newer GLP-1 and SGLT2 drugs by more again. The American Diabetes Association does not recommend chromium in its annual practice standards ADA 2024.

For people without diabetes — the typical reader of this entry — the visible claim is fat loss. Two independent meta-analyses with no commercial ties on either side pool roughly half a kilo to one kilo of weight loss versus placebo over twelve to twenty-four weeks Pittler et al. 2003 Onakpoya et al. 2013, and explicitly call the size "of doubtful clinical significance." A kilo over six months is below the noise of a normal fluctuation week. A separate controlled trial in young women combining the supplement with a resistance programme found no body-composition advantage at all Lukaski et al. 2007.

The most defensible felt-experience claim is on carbohydrate cravings. A 113-person trial in atypical depression — a subtype where carb craving and oversleeping are core symptoms — ran 600 μg/day chromium picolinate for eight weeks; the subgroup with high cravings improved on the depression scale, driven mostly by the appetite items Docherty et al. 2005. A smaller, similar trial in overweight craving-prone women found reduced free-eating intake Anton et al. 2008. Both trials were funded by the supplement industry; neither has been replicated outside that funding source. That doesn't make them wrong, but it's the right caveat to keep in mind.

The essentiality dispute, in one screenful

You will see chromium called "an essential trace mineral" on almost every bottle. That is true in the US — the Institute of Medicine set an Adequate Intake number for it in 2001 (35 micrograms a day for adult men, 25 for women) IOM 2001. But "adequate intake" is a softer category than "recommended dietary allowance": it is set by what people happen to eat, not by what the body has been shown to require. The IOM acknowledged at the time that no chromium-deficiency syndrome had been observed in free-living adults.

In 2014, the European Food Safety Authority looked at the same evidence and went further. There is no convincing evidence chromium is essential at all; no average requirement can be defined; no dietary reference value is warranted EFSA 2014. The chromodulin biochemist behind the proposed mechanism agreed, three years later, in print Vincent 2017. The only well-documented human deficiency is in patients fed through a vein for months on chromium-free intravenous-nutrition formulas Jeejeebhoy et al. 1977 — and even those cases are now under question as possibly reflecting other mineral imbalances rather than chromium per se.

The second durable misconception is the glucose tolerance factor story — the idea that chromium, niacin, and amino acids form a defined biological complex that helps insulin work. The complex was never chemically pinned down; the framing is now obsolete Vincent 2017. When you see "GTF chromium" on a label, the chemistry behind that phrase has not held up.

Who has at least directional evidence

Three small slices of the population, in order of strongest signal:

Type 2 diabetics close to goal on metformin alone. The pooled half-point of HbA_1c reduction is real, even if small, and the supplement is cheap enough to try without much downside Suksomboon 2014. It is not a substitute for guideline therapy; it is a marginal addition on top of one.
People with atypical depression and strong carbohydrate cravings. The most-defensible felt-experience signal in the literature, though it rests on one industry-funded trial Docherty 2005.
Habitual overeaters with intrusive carb cravings of any cause. A smaller, similar signal Anton 2008, same caveats about funding and replication.

For healthy adults eating a varied diet, with no diabetes and no craving problem, the evidence does not support a benefit EFSA 2014 Vincent 2017. "Just in case" is not what the data supports.

What you are actually missing if you skip it

For most healthy adults: essentially nothing. No documented functional deficit shows up in free-living people eating mixed diets, and the regulators willing to set a number for habitual intake all set it within range of what people already eat EFSA 2014 Vincent 2017. The bottle you didn't buy isn't a bottle you needed. The friend who's been on it for two years and swears by it isn't lying — they may genuinely feel sharper or leaner — but the most-likely cause is the rest of what they changed when they decided to start "doing something about their health," not the capsule.

For a type 2 diabetic, though, the stake of treating chromium as a serious alternative to guideline therapy is real. Half a point of HbA_1c, in the best-case pooled estimate, is roughly a third of what metformin does and an order of magnitude below the modern injectables your doctor is choosing among. The cost of a "natural" alternative that does not carry the load shows up five and ten years out — in retinas, in kidneys, in the toes that stop healing — not in the dollars you saved by skipping the prescription.

If you are going to try it

The trials that show anything use chromium picolinate specifically, at 200 to 1,000 micrograms a day, for at least eight to twelve weeks. Below 200 μg there is no detectable blood-sugar effect in any meta-analysis Suksomboon 2014. Other chromium forms (polynicotinate, chloride) do not carry even the modest signal that picolinate does in pooled data.

The discipline is the target. The supplement is cheap enough that the failure mode is drift — keeping it in the daily pill organiser out of inertia, five years on, because nothing got worse when you started and you never set up a way to notice nothing got better. Decide what you're measuring, measure it, drop the bottle if nothing budges.

When not to take it

If you are on diabetes medication — insulin, sulphonylureas, or any drug that can drop blood sugar on its own — adding chromium can stack additively, and case reports describe hypoglycaemia in well-controlled patients who added it without telling their clinician NIH ODS 2022. The fix is communication, not abstention: tell whoever manages your diabetes if you start it, and watch your numbers in the first month.

The chromium picolinate form has a thirty-year-old test-tube safety footnote: at high concentrations in hamster cell cultures, the picolinate ligand (not the chromium) caused chromosome damage Stearns et al. 1995. Subsequent assays have been mixed; no human-cancer signal has emerged after three decades of widespread sales. It is a footnote, not a settled question — worth knowing about, not enough to forbid the supplement. Pregnancy and breastfeeding intake at supplement doses are not well characterised; food-source chromium is the conservative choice during those windows. Rare case reports of kidney and liver problems at sustained very high doses exist and are scattered NIH ODS 2022 — a reason not to chase 1,500+ μg a day for months without a target.

What actually moves the metrics chromium claims to

For blood sugar: weight loss, walking right after meals, resistance training, and dietary carb-quality changes — each on its own does more than the supplement's best-case pooled effect. Pharmacologically, metformin and the newer GLP-1 and SGLT2 drugs lower HbA_1c by a full point or more ADA 2024. For body fat: a calorie deficit and resistance training, the boringly-effective levers. For carbohydrate cravings specifically: protein-forward meals, sleep adequacy, and — at the pharmacological end of the same craving signal — the GLP-1 drugs again, which collapse cravings at a magnitude chromium has never approached.

Chromium-rich foods are also a real option: broccoli, grape juice, brewer's yeast, whole grains, lean meat. A diet that already includes them is almost certainly past the EFSA's "no evidence of deficiency" threshold without any supplement at all.

What changes if it works for you

Honest about magnitude. If the diabetes signal lands for you, the change shows up on a blood test three months out, not in how you feel walking around. If the craving signal lands, the most-likely experience is the mid-afternoon sugar reach feeling slightly easier to skip — a small lift, not a personality change. If the body-composition story lands at the meta-analytic mean, the scale moves about a kilo over six months. None of these are wrong things to want; all of them are smaller than the bottle advertises.

If nothing has moved by week twelve, that is the answer. The discipline of this entry is the willingness to read that answer and not keep the bottle in the organiser anyway. The supplement is so cheap and so easy that the failure mode is not poisoning — it is paying twenty dollars a year for a small ritual that's doing nothing, indefinitely, because nothing told you to stop.

Worth knowing next

If the blood-sugar story brought you here, the real leverage lives in metformin and the modern oral and injectable diabetes agents — read those before reading any other supplement. If the craving story brought you here, the GLP-1 receptor agonists collapse the same craving signal at a different order of magnitude, and a protein-forward breakfast does more for an afternoon than a capsule does. If body composition brought you here, resistance training and a calorie deficit are the entries the magazine pages keep pretending are interchangeable with chromium. And if you want to see what your own blood sugar is actually doing between visits, a continuous glucose monitor will tell you more in two weeks than a year of any supplement ever could.

Substance + claimed effects

Chromium is a trace metal taken in trivalent form (Cr³⁺) from food and from supplements — most often as chromium picolinate (chromium chelated to picolinic acid), the form invented by USDA chemist Gary Anderson and patented for diabetic blood-sugar support in the early 1990s. The hexavalent form (Cr⁶⁺, industrial) is a known carcinogen and is not what supplements contain; this dossier is exclusively about Cr³⁺. Marketing has long claimed the supplement improves insulin sensitivity, lowers fasting glucose and HbA_1c, reduces body fat, blunts carbohydrate cravings, and supports lean-mass gains in trainees. The entry covers each of those claims plus the foundational dispute over whether chromium is an essential nutrient at all — a question the US and EU regulators answer differently EFSA 2014 IOM 2001 — and the cellular-toxicity flag that has followed picolinate since the mid-1990s Stearns et al. 1995.

Evidence by addressing question

mechanism

The standing biochemical model is that ingested Cr³⁺ binds to a low-molecular-weight oligopeptide named chromodulin (also called LMWCr), and the chromium-saturated complex then docks onto insulin-stimulated insulin receptors, amplifying their intrinsic tyrosine-kinase activity and downstream GLUT4 translocation in muscle and adipose tissue Vincent 2017. The model is built from in-vitro insulin-receptor activation assays and from rodent work in obese, insulin-resistant strains (e.g., JCR:LA-cp rats). It has never been confirmed in human tissue at physiological doses; Vincent 2017 — the biochemist who isolated chromodulin — now writes that the proposed autoamplification of the insulin receptor has not been demonstrated in vivo and that animal-deficiency studies have failed to produce a consistent chromium-deprivation phenotype on modern, contamination-controlled diets. The mechanism is plausibility, not established physiology.

Picolinic acid, the chelating ligand in the dominant supplement form, is itself bioactive: it binds zinc, iron, and copper and crosses cell membranes more readily than free Cr³⁺. Several of picolinate's reported in-vitro effects (and most of its mutagenicity signal — see contraindications) trace to the ligand, not the metal Stearns et al. 1995.

evidence

The diabetes literature is the largest and most-cited body of human work. The signal RCT is Anderson et al. 1997 — 180 Chinese adults with type 2 diabetes, randomised to placebo, 200 μg/day, or 1,000 μg/day chromium picolinate for four months. At 1,000 μg/day, HbA_1c fell from a baseline near 8.5% to 6.6% and fasting glucose dropped 15–19% versus placebo. The effect size is dramatic — closer to metformin than to a supplement — and the trial drove a decade of marketing.

Western replications have not matched it. Cefalu et al. 2002 and Kleefstra et al. 2007 found no insulin-sensitivity or glycaemic benefit in US and Dutch type 2 cohorts at comparable or higher doses. The plausibility-favouring interpretation is that the Chinese cohort was chromium-deficient at baseline — habitual intake in rural China was much lower than in Western diets — and that the supplement only "works" where dietary intake is already short. The skeptic interpretation is that Anderson 1997 is a quality outlier: Costello et al. 2016, the NIH Office of Dietary Supplements' systematic review, rated almost half of the 20 included trials as poor quality and concluded that overall results in type 2 diabetes are inconclusive.

The two pooled estimates that survive published meta-analysis are modest. Suksomboon et al. 2014 (25 RCTs) found a mean HbA_1c reduction of −0.55% (95% CI −0.88 to −0.22) with chromium versus placebo in diabetic patients; the effect was concentrated in the picolinate form and at doses above 200 μg/day. Yin and Phung 2015 reported similar magnitudes. For context, metformin lowers HbA_1c by roughly 1.0–1.5% — chromium's pooled effect, if real, is about a third of that and an order of magnitude below the next-generation oral agents (GLP-1s, SGLT2s). The American Diabetes Association's annual Standards of Care consistently declines to recommend chromium for glycaemic management on grounds of insufficient evidence ADA 2024.

In non-diabetic readers — the typical catalogue audience — the body-composition story is the visible claim. Pittler, Stevinson and Ernst 2003 pooled ten RCTs (n=489) and reported −1.1 kg weight loss versus placebo at typical supplement doses; Onakpoya, Posadzki and Ernst 2013, updating with 11 RCTs (n=866), pooled −0.5 kg and a small body-fat reduction with no BMI or waist-circumference change. Both groups (independent meta-analysts, no commercial ties) concluded the effect is statistically detectable but of doubtful clinical significance. A double-blind trial in young women supplementing through a 12-week resistance programme found no body-composition advantage over placebo Lukaski, Siders and Penland 2007.

Carbohydrate craving is the smallest but most consistent felt-experience claim. Docherty et al. 2005 randomised 113 adults with atypical depression (a subtype characterised by carb craving and oversleeping) to 600 μg/day chromium picolinate or placebo for eight weeks; in the high-craving subgroup, total HAM-D-29 scores improved significantly versus placebo, driven largely by the appetite-and-cravings items. Anton et al. 2008 ran a smaller laboratory-meal study in overweight women with reported food cravings: 1,000 μg/day chromium picolinate reduced free-eating intake and self-rated hunger over eight weeks. Neither result has been replicated at scale, and both were funded by the supplement manufacturer.

protocol

The trials that show any effect cluster around 200–1,000 μg/day elemental chromium, with the picolinate form carrying the bulk of positive results in the Suksomboon meta-analysis Suksomboon et al. 2014. Below 200 μg/day, no glycaemic effect is detectable in pooled data. The US adequate intake is far lower — 35 μg/day for adult men, 25 μg/day for adult women, dropping to 30 and 20 after age 50 IOM 2001 — which is what a mixed diet supplies; supplement protocols are pharmacological doses an order of magnitude above habitual intake. Onset of detectable effects in positive trials runs 8–16 weeks; carb-craving effects appear sooner (Docherty's signal emerged by week 2 in his subgroup). No tolerable upper intake level was set by the IOM for Cr³⁺ from food; supplement labels typically cap at 500–1,000 μg.

contraindications

The most cited safety flag is in-vitro: Stearns et al. 1995 reported chromosomal aberrations in Chinese hamster ovary cells exposed to chromium picolinate at concentrations of 0.05–1.0 mM, with damage attributable to the picolinate ligand (chromium nicotinate and CrCl₃ did not produce the effect, and free picolinic acid alone was clastogenic). Subsequent assays in CHO and Salmonella systems have been mixed and no human carcinogenicity signal has emerged after three decades of supplement sales, but the regulatory verdict remains "insufficient evidence of harm at typical doses" rather than "shown to be safe."

Co-administration with oral antidiabetic drugs and insulin can stack — case reports describe additive hypoglycaemia when chromium is added on top of sulphonylureas or insulin in well-controlled patients NIH ODS 2022. Rare case reports of rhabdomyolysis, nephritis, and hepatic dysfunction at high cumulative doses (1,200–2,400 μg/day for months) exist but are scattered NIH ODS 2022. Picolinate competes with iron for transferrin binding; iron-deficiency states warrant caution. Pregnancy and breastfeeding data are limited; the AI in pregnancy is set higher (30 μg/day) but supplement-level intakes are not characterised.

misconceptions

The most durable misconception is that chromium is an essential micronutrient, on the level of iron or zinc. The US IOM set an Adequate Intake in 2001 — a level set by observed habitual intake, not by a demonstrated requirement IOM 2001. The European Food Safety Authority's 2014 reassessment looked at the same evidence and reached the opposite conclusion: there is no convincing evidence that chromium has an essential physiological role in healthy humans, no average requirement could be defined, and dietary reference values are not warranted EFSA 2014. Vincent 2017 — who first purified chromodulin in the 1990s — endorses the EFSA position based on the failure of modern, contamination-controlled animal studies to produce a chromium-deficiency phenotype.

The "glucose tolerance factor" idea — that chromium plus niacin and amino acids forms a discrete biological complex that potentiates insulin — has not held up to chemical characterisation since the 1990s and is now obsolete Vincent 2017. The only well-documented human deficiency is in patients on long-term unsupplemented total parenteral nutrition Jeejeebhoy et al. 1977; modern TPN formulas have included trace chromium for decades, and even those cases have come under doubt as possibly reflecting contaminating zinc or iron rather than chromium itself.

audience

Sub-populations with at least directional evidence of benefit:

Type 2 diabetics with poor glycaemic control on metformin alone. The largest signal comes from Anderson 1997; pooled effect −0.55% HbA_1c in Suksomboon 2014. ADA does not recommend it on the strength of available evidence ADA 2024.
People with atypical depression and high carbohydrate craving. Docherty 2005. Industry-funded, unreplicated at scale.
Habitual overeaters with intrusive carb cravings. Anton 2008, small, industry-funded.

For healthy adults eating a typical mixed diet, there is no documented baseline deficiency and no functional outcome the supplement reliably moves EFSA 2014 Vincent 2017.

alternatives

For the glycaemic-control claim, the proven interventions are weight loss, resistance and aerobic exercise, dietary carbohydrate-quality change, and — pharmacologically — metformin and the newer GLP-1 receptor agonists and SGLT2 inhibitors, each with HbA_1c reductions an order of magnitude above chromium's pooled estimate ADA 2024. For body composition, calorie deficit and resistance training dominate any supplement's contribution. For carbohydrate cravings specifically, the strongest non-pharmacological lever is protein-forward meals and sleep adequacy; on the pharmacological side, GLP-1 agonists collapse the same craving signal at very different magnitudes.

failure-modes

The most common reason "I tried chromium and nothing happened" is that the reader was not chromium-insufficient at baseline. The Anderson Chinese cohort had habitual intakes far below modern Western diets; positive trials cluster in populations where baseline status is plausibly low. A second failure mode is dose: under 200 μg/day, no glycaemic effect is detectable in any meta-analysis. A third is form: chromium polynicotinate and CrCl₃ do not carry the (admittedly small) signal that picolinate does in pooled data Suksomboon 2014.

practicalities

Chromium picolinate is one of the cheapest supplements on the shelf — generic bottles deliver a year's supply at typical dose for under $20. It is regulated as a dietary supplement in the US (no FDA pre-market efficacy review) and as a novel-food-grade additive in the EU, where chromium-rich yeast and chromium chloride are also permitted in supplements. Half-life in plasma is short; excretion is renal. No fasting or timing-with-meals discipline meaningfully changes pharmacokinetics.

stakes

For a healthy reader, the stake of not taking chromium is essentially nothing — no reproducible functional deficit has been documented in free-living adults eating a mixed diet EFSA 2014 Vincent 2017. For a type 2 diabetic, the stake of relying on chromium instead of guideline therapy is real — the absolute glycaemic gap between chromium (best pooled estimate −0.55%) and metformin (~1.0–1.5%) plus the next-line agents is the difference between organ-damaging and well-controlled disease.

payoff

If the carbohydrate-craving signal holds for a given reader (the cleanest positive result in the literature is the Docherty subgroup), the payoff is felt as smaller portion sizes and less mid-afternoon sugar reach. If the glycaemic signal holds (best-case −0.55% HbA_1c), it is invisible to felt experience and only registers on a blood test. The body-composition payoff at the meta-analytic mean (−0.5 to −1.1 kg over months) is below the noise floor of a normal weight-fluctuation week.

The credibility range

The optimist case

Chromium has a real biochemical interaction with insulin signalling — Vincent's own chromodulin isolation, multiple rodent insulin-receptor activation studies, and the Anderson Chinese trial together describe a mechanism in search of the right population. Western trials have under-detected the effect because Western diets are not chromium-poor enough to expose the deficiency the supplement corrects. In the populations where deficiency is real (rural Chinese diabetics in Anderson, chromium-poor TPN patients pre-supplementation, atypical depressives with craving phenotypes), the supplement produces a clean signal. Pooled HbA_1c reduction of −0.55% in Suksomboon's 25-trial meta-analysis is not nothing — it is comparable to a third-line oral agent, at $0.05 a day, with a safety profile that thirty years of widespread use has not flagged in humans. The Docherty craving signal is biologically coherent (insulin and central serotonin both modulate carb appetite) and clinically meaningful for the subgroup it identifies.

The skeptic case

Modern, contamination-controlled animal nutrition cannot produce a chromium-deficiency syndrome at all EFSA 2014 Vincent 2017. The proposed in-vivo autoamplification mechanism has never been demonstrated outside cell culture and rodent receptor preparations Vincent 2017. The Anderson trial is an outlier on quality-rated review Costello 2016 and has not replicated in two independent Western RCTs Cefalu 2002 Kleefstra 2007. The body-composition meta-analyses — independently performed, no commercial ties — find statistically detectable but clinically inconsequential effects and explicitly say so Pittler 2003 Onakpoya 2013. The craving trials are industry-funded with small samples. The picolinate ligand is bioactive in its own right and the Stearns clastogenicity signal has never been fully extinguished Stearns 1995. The most authoritative regulatory voice in the field has concluded chromium is not an essential nutrient EFSA 2014; the ADA does not recommend it for diabetes management ADA 2024. The most parsimonious explanation for the persistent supplement market is the legacy of the picolinate patent and decades of unchecked marketing claims rather than a real, generalisable physiological effect.

The author's call

The skeptic case is the heavier one — EFSA's reassessment and the chromodulin biochemist's own retraction of the essentiality claim are load-bearing. The honest read is: chromium is probably not essential for free-living adults eating a varied diet; the supplement effects on glycaemic control are real but small and concentrated in populations with low baseline intake (most readers are not those populations); the body-composition story is over-marketed against its own meta-analytic data; the carbohydrate-craving signal is the most defensible felt-experience claim but rests on one industry-funded subgroup analysis. The supplement is cheap and broadly safe, so it can be tried — for a clear, defined target (HbA1c at the third-decimal-place, or carb cravings the reader can rate before and after) — and dropped without ceremony when nothing budges in 12 weeks. It is not a foundation; it is a marginal lever, contested at the foundations. evidence: 2, controversy: 3.

Stakeholder + incentive map

Picolinate patent holders and supplement makers. The chromium picolinate compound was patented by USDA researcher Gary W. Evans in the late 1980s, then licensed and marketed aggressively for blood-sugar and body-composition claims through the 1990s and 2000s. Most positive carb-craving and body-composition trials (Docherty, Anton) carry industry funding or sponsorship.
FTC and regulators. The US FTC settled with multiple chromium picolinate marketers in the late 1990s over weight-loss and disease-prevention claims; the legacy is a paragraph of regulatory scepticism in any modern NIH ODS or EFSA opinion NIH ODS 2022 EFSA 2014.
Diabetes guideline bodies. The ADA has consistently declined to recommend chromium in the annual Standards of Care; it is mentioned as a supplement that "has not shown clear benefit" ADA 2024.
Independent academic biochemists. John Vincent (the original chromodulin isolator) has been the most public voice arguing against essentiality; the EFSA Panel adopted that framing EFSA 2014 Vincent 2017.
Functional-medicine and "metabolic health" practitioner subcultures. Continue to recommend chromium for insulin sensitivity, PCOS, and craving control, citing the Anderson signal and the chromodulin mechanism. Largely impervious to the EFSA reassessment.

Population variability

The signal is concentrated in populations with low baseline chromium status — the Anderson Chinese cohort being the clearest example. Western adults consuming mixed diets average 25–50 μg/day from food, near or above the AI; signal-to-supplementation in this group is weak. Type 2 diabetics show plasma chromium concentrations lower than non-diabetic controls in some cohorts, but causality (does diabetes deplete chromium, or does low chromium contribute to diabetes?) is unsettled. The atypical-depression carb-craving subgroup in Docherty 2005 is a phenotypic minority; the general-depression population in the same trial did not benefit. Older adults absorb less chromium and have lower plasma levels, but no functional deficit in the absorption decline has been documented. Vegetarian and high-refined-carbohydrate diets are lower in chromium than mixed omnivorous diets, but the practical relevance is unclear.

Knowledge gaps

Whether chromium has any essential physiological role in healthy humans remains unresolved — the two leading regulatory bodies disagree, and no chromium-deficiency syndrome can be reliably produced in modern animal models EFSA 2014. The Anderson Chinese trial has never been independently replicated at scale in a non-Western, low-baseline-intake population — a single well-powered repeat in (say) rural India or sub-Saharan Africa would settle whether the signal generalises beyond one cohort. The carb-craving claim rests on one industry-funded subgroup analysis (Docherty 2005) and one small follow-up (Anton 2008); a non-industry replication in the same atypical-depression-with-craving phenotype would change the credibility range materially. Long-term safety data at supplement doses (1–10 years at 200–1,000 μg/day) is observational rather than trial-grade; the Stearns clastogenicity signal remains a footnote rather than a resolved question Stearns 1995. The mechanism — whether chromodulin amplification is operative at all in vivo at physiological chromium concentrations — would be settled by a tracer study in humans that the field has not yet run Vincent 2017.

Scope match against the brief. The brief named insulin signalling, fasting glucose, body composition, and carbohydrate cravings, plus the "modest and contested evidence" framing. Each consequence is covered in the article: insulin/glucose in mechanism and evidence; body composition in evidence (Pittler 2003, Onakpoya 2013, Lukaski 2007); cravings in evidence (Docherty 2005, Anton 2008); the contested-evidence frame structures the dek, the highlights, and the misconceptions section. No silent narrowing.

Hard scoping calls.

Picolinate vs other chromium forms. Treated chromium picolinate as the load-bearing supplement form throughout. The pooled glycaemic signal in Suksomboon 2014 is concentrated in picolinate; the body-composition meta-analyses are picolinate-only. Polynicotinate and CrCl3 mentioned briefly in protocol but not given their own beats — there is no positive trial signal to write about.
Hexavalent chromium (industrial Cr⁶⁺). Deliberately not covered in the reader-facing article — it is a different substance with a different regulatory and toxicological story, not a supplement. Flagged in the research dossier to keep the distinction in the reviewer's mind.
PCOS + chromium. Skipped. There is a small literature suggesting modest glycaemic and ovulation effects in PCOS at high doses, but the body of evidence is too thin and too noisy to land cleanly in a general-audience entry. A future PCOS entry is the better home — see future-link candidates below.
Action verb. Chose know over do. The entry's value is the calibrated read, not a recommendation to take the pill — most readers should not. The protocol section gives concrete guidance for the minority who want to try, but the dominant action is "be informed."
Dream narrative at score <40. Wrote one in relief / debunking mode (overall ≈ 13). The honest hook here is the reader getting their hope and their twenty dollars back, not aspiration. The tagline reflects that lever without the dream-tier marketing-words licence.

Rating difficulties.

health_short_term = 1 was close to 0. Voted 1 because the diabetic-glycaemic signal, even at -0.55% pooled, is large enough to register on a lab test for the narrow subgroup; if the entry were scoped only to non-diabetics the honest score would be 0.
mood = 1 rests almost entirely on Docherty 2005's atypical-depression-with-craving subgroup, which is industry-funded and unreplicated at scale. A defensible alternative call is 0; I leaned to 1 because the felt-experience signal is the cleanest in the literature and the subgroup is real even if narrow. If the catalogue gets a future "atypical depression" entry, that may be a better home for the same evidence.
evidence = 2 vs 3 was the hardest call. Twenty-five-trial meta-analyses with a statistically significant effect would normally land at 3, but the Anderson-replication-failure pattern (Cefalu, Kleefstra), the Costello "inconclusive" verdict, and the EFSA non-essentiality finding pulled it down. Going with 2 to avoid inflating evidence for a substance the most-authoritative regulator says may not even be needed.
controversy = 3 reflects the IOM-vs-EFSA split and Vincent's mid-career switch, both of which are real expert disagreements with stakes for the entry's framing.
applicability = 3 leans on the broad decision-audience rule from meta.md §6 — the entry's value is to the much wider set of supplement-curious adults weighing whether to add chromium, not just the ~10% with diagnosed type 2 diabetes. A narrow "type 2 diabetics only" read would score this 2.

Future-link candidates (entries that don't yet exist).

PCOS — chromium has a thin but real literature here that deserves its own treatment.
Atypical depression — the Docherty subgroup result would land more meaningfully inside a clinical-subtype entry than as a footnote in a supplement entry.
Glucose tolerance factor / chromodulin biochemistry — only if the catalogue ever takes on substance-of-historical-interest entries; not a priority.
Hexavalent chromium / industrial exposure — distinct from this entry; would belong under environmental or occupational health if the catalogue grows that way.

Separate-entry candidates surfaced during research. Metformin, GLP-1 receptor agonists, and SGLT2 inhibitors are each name-checked in alternatives and out-of-scope; each warrants its own catalogue entry if the diabetes-pharmacology side of the catalogue gets built out. Resistance training and continuous glucose monitoring are already implied catalogue neighbours.

Evidence-handling note for reviewers. The industry-funding flag on Docherty 2005 and Anton 2008 is named in-line in the article rather than buried — the craving signal is the most-defensible felt-experience claim in the literature but is also the one most exposed to publication and funding bias. The article gives the reader both pieces at once rather than choosing which to omit.