Glutathione: One Infusion Trial in FDA's Search and in Mine, and Nine of Twenty Five Stopped Early

01What it is

Glutathione is a tripeptide, three amino acids joined end to end, glutamate, cysteine and glycine, and every cell in the body makes it. Formula C10H17N3O6S, PubChem CID 124886. My Drugs@FDA search by active ingredient found no approved United States drug containing it, only an eye irrigation solution and a methotrexate record, neither of them a glutathione product.

That eye solution matters because compounders have pointed at it. It contains the oxidised form, which FDA answers in a footnote: oxidised glutathione is a dimer, two molecules joined, not a salt and not the same substance.

Injectable glutathione is approved elsewhere, in the Philippines as an add on to cisplatin chemotherapy and, Sonthalia and colleagues report, in India for liver disorders. What a United States clinic puts in the bag is neither: Empower Pharmacy sells a compounded 200 mg per mL vial under the line that safety and efficacy have not been evaluated by FDA.

02What the trials found

For skin lightening by infusion, FDA's literature search identified one trial, and my own PubMed and registry searches returned the same one.

Zubair and colleagues, 2016, read through FDA's summary, not the original. Fifty women randomised into a vein twice weekly for six weeks, to saline or to a proprietary preparation carrying glutathione 1200 mg alongside ascorbic acid, which is vitamin C, and 35 mg of hydrolysed collagen. That trial of the injection tested a mixture, not the molecule on its own. Six of the sixteen who finished on the drug were judged lighter, against three of sixteen on saline, at a p value of 0.054, which does not clear the conventional threshold. The improvement then faded, and at six months one patient had kept it.

Sonthalia and colleagues took it apart in 2018: nine of twenty five dropped out of the treatment arm, and the colour card scale used to grade the skin is in their words unreliable for changes this subtle. (PMID 29445569)

By mouth the picture is thin rather than empty: 60 Thai medical students on capsules 250 mg twice daily for four weeks had pigment fall at all six measured sites and beat placebo at two. (PMID 20524875) The tightest skin trial in the file is a lotion and not this molecule: 30 women, split face, 2 percent oxidised glutathione for 10 weeks, pigment below placebo, three of its authors at a glutathione manufacturer. (PMID 25378941)

Away from skin, FDA assessed glutathione against 24 nominated uses at once and found no information or insufficient evidence for any, adding that some evidence may support skin lightening but that no published study supported the other uses. The oncologists' guideline says it should not be offered for chemotherapy nerve damage, where a randomised trial in 185 patients found no difference and two measures favouring placebo.

03Whether swallowing it works, which decides the rest

The market rests on a prior claim: swallowed glutathione does not get in, so it has to go in a vein.

Witschi and colleagues, 1992. Seven healthy volunteers, one oral dose of 0.15 mmol per kilogram, as much as 3.0 grams in a 65 kg adult, watched for 270 minutes. Plasma glutathione started at 6.2 micromoles per litre and did not move measurably. (PMID 1362956) One dose, seven people, four and a half hours.

Richie and colleagues ran it for six months instead: randomised and placebo controlled, 250 mg or 1,000 mg a day by mouth, 54 adults analysed. The higher dose group finished up 30 to 35 percent in red cells, plasma and lymphocytes. (PMID 24791752) Allen and Bradley gave 1,000 mg a day for four weeks to 40 adults and nothing moved. (PMID 21875351)

Having read all of it, FDA still concluded that the amount of swallowed glutathione that reaches the bloodstream is minimal. The trade quotes the four and a half hour study, and I found no clinic page mentioning the six month one.

Now the other half. Aebi and colleagues, read through FDA's evaluation rather than the original, infused 10 healthy volunteers at 2 grams per square metre of body surface: plasma glutathione rose about 64 fold, to 444 micromoles per litre, then cleared with a half life of roughly 15 minutes, the time for half of it to go, as cysteine, its breakdown product, rose in blood and urine. FDA adds that it does not easily cross the blood brain barrier.

It is like watering a houseplant down the outside of the pot. You can measure the water going in, and I could find nobody who has dug down to see whether any of it reached the roots.

04How it compares

Four things get sold and cited as one substance.

Every controlled skin lightening result I found came from a capsule or a lotion. The injection trial missed.

05Where the dose came from

Empower Pharmacy states the convention plainly, and it is what circulates, not a recommendation: 600 mg to 1,400 mg per session, two to three times per week.

Both ends are Parkinson's disease studies, and the 600 mg travelled without its schedule. That trace is mine.

600 mg is Sechi 1996: nine patients with early, untreated Parkinson's disease, none of them yet on levodopa, the standard drug for it, into a vein twice a day for 30 days, open label, no control group, all nine improving. (PMID 8938817) That is about 8,400 mg a week, against the 1,200 to 1,800 a clinic schedule comes to.

1,400 mg is Hauser 2009: randomised and placebo controlled, 21 patients with Parkinson's disease, three times a week for four weeks, improving 2.8 units more than placebo on daily living and motor scores, a gap well inside what chance produces at that sample size, P = 0.32, and then worsening over the eight weeks after, which at P = 0.54 is also inside what chance produces. (PMID 19230029) I found no study behind the twice weekly end.

For skin lightening the number came from the sellers. Sonthalia and colleagues report manufacturers recommending 600 to 1200 mg weekly or twice weekly with no stated duration, citing a 2005 review that contains no dosing trial. (PMID 18492181) Zubair used the top of that range.

Empower's page carries eight references: consumer drug sites, supplement and skincare blogs, marketing pages of other clinics, WebMD and the National Institutes of Health. Not one is a study.

06What could go wrong

The skin lightening trial both searches found is also the harm record. Of the 25 healthy women who received glutathione, nine stopped for severe adverse effects, eight for deranged liver tests and one for anaphylaxis, a whole body allergic reaction that can close the airway. All 25 reported some side effect. Nobody on saline did.

FDA's adverse event database held 17 reports across twenty two years, seven of them anaphylaxis or hypersensitivity, five given into a vein. One woman had anaphylaxis, was given the drug again ten days later, and needed resuscitation.

Then endotoxin, the fragment of dead bacterial cell wall that survives sterilisation. No infection, but a violent reaction within minutes. Two clusters of seven patients each turned up in my reading and they are not the same event. Johnstone and colleagues investigated seven cases of probable endotoxin poisoning in New South Wales after compounded glutathione infusions. In January 2019, seven patients at an American clinic were each given 1,400 mg at 200 mg per mL and were ill within minutes, and FDA's laboratory found endotoxin in that powder as high as five times the appropriate limit. The label on that powder read both "Caution: Dietary Supplement" and "For Manufacturing, Processing, Repackaging and Pharmaceutical Compounding". A third seven patient cluster sits in the same FDA document and is a different failure again: infusions pushed over two minutes instead of the intended ten.

That 1,400 mg is Hauser's Parkinson's dose and also the top of the range a clinic session runs to. A pharmacy in Austin, Texas withdrew three lots of 200 mg per mL vials on 5 August 2026 for elevated endotoxin, seven years on, with no published link between them.

The Philippine regulator's 2019 advisory says it plainly: no published clinical trials of injectable glutathione for skin lightening, no published dosing guidance, no product approved there for it, and reported side effects on the liver, kidneys and nervous system. Zubair was published three years earlier, and the two sit unreconciled.

FDA's reviewers recommended against adding glutathione to the substances compounders may lawfully use, naming the route: injections give rapid and irreversible exposure. On 8 June 2022 the advisory committee voted eight to five, one abstention, against its own agency. Four years on FDA has neither added it nor refused.

07What nobody knows

Whether repeated infusions raise glutathione inside any human tissue. It floods the space outside cells and clears in a quarter of an hour, and I found no study that biopsied one afterwards.

Whether the six month oral result holds. Richie stands alone at that duration, and I found no independent replication.

Whether the liver signal is real. Eight of 25 healthy women with deranged liver tests over six weeks would matter if confirmed, and the trial reported it without enzyme levels.

Whether the modified versions are the story. In cell culture, as Sonthalia and colleagues report Chung's work, glutathione itself did not inhibit the pigment enzyme inside the cell and two chemically altered versions did. Nobody in this chain opened Chung's paper, and I found no trial of those versions in people.

08Where to look next

FDA's briefing document for the Pharmacy Compounding Advisory Committee meeting of 8 June 2022 is public. Read the glutathione evaluation inside it and see whether the trial the practice rests on reads the way clinic pages describe it.

If you have had one of these drips, I want to know what the clinic told you was in the bag besides glutathione, and whether anybody checked your liver enzymes before or after.

Frequently asked

Is injectable glutathione approved by the FDA?+

There is no drug approved in the United States whose active ingredient is glutathione. A search of the Drugs@FDA database by active ingredient returns an eye irrigation solution and a methotrexate record, and neither is a glutathione product. What American clinics use is a compounded preparation, mixed to order by a pharmacy rather than manufactured as an approved drug, and the compounder's own page states that safety and efficacy for that formulation have not been evaluated by FDA. Injectable glutathione is approved in some other countries, including the Philippines, where the regulator says it is approved as an add on to cisplatin chemotherapy and for nothing else.

Does an intravenous glutathione drip lighten skin?+

One clinical trial of the infusion for skin lightening exists in FDA's literature search and in mine. Fifty women were randomised twice weekly for six weeks to saline or to a preparation carrying a trial dose of glutathione 1200 mg alongside vitamin C and hydrolysed collagen, given into a vein, and 37.5 percent on the drug were judged lighter against 18.7 percent on saline, at a p value of 0.054, which does not clear the conventional threshold. The lightening faded after treatment stopped and at six months one woman out of sixteen still had it. Dermatologists reviewing the trial called the colour card scale it used unreliable for changes this subtle.

Does swallowed glutathione work, or does it have to be injected?+

It depends entirely on how long you look. A single oral dose given to seven people who were then watched for four and a half hours did not move plasma glutathione measurably, and that 1992 result is the basis for the whole injection argument. A six month randomised trial of capsules, at trial doses of 250 mg or 1,000 mg a day and not as a recommendation, with 54 adults analysed, found body stores up 30 to 35 percent at the higher dose, returning to baseline a month after stopping. A four week trial at the same higher trial dose in 40 adults found nothing move. FDA, having read all of it, still concluded that the amount of swallowed glutathione that reaches the bloodstream is minimal.

Where does the 600 to 1,400 mg clinic session come from?+

Both ends of that circulating range come from Parkinson's disease research, and this is reported as convention rather than as a recommendation. 600 mg is an open label study with no control group in nine patients with early Parkinson's disease who were not yet on levodopa, where it was given into a vein twice a day for thirty days, about 8,400 mg a week rather than the 1,200 to 1,800 a clinic schedule comes to. 1,400 mg three times a week is a randomised placebo controlled trial in 21 patients with Parkinson's disease that missed its endpoint at P = 0.32. I found no study behind the twice weekly end of the range, and no clinic page that cited a study for its dose.

What harms have been reported with glutathione infusions?+

In the only skin lightening trial of the infusion that FDA's literature search turned up, and the only one mine turned up, all 25 healthy women who received glutathione reported side effects and nobody on saline did, and nine of the 25 stopped early for severe adverse effects, eight of them for deranged liver function tests and one for anaphylaxis, a whole body allergic reaction that can close the airway. FDA's adverse event database held 17 reports over 22 years, seven of them allergic reactions and five of those given into a vein, including a woman who had anaphylaxis, was given the drug again ten days later and needed resuscitation. Separately, bacterial endotoxin, the fragment of dead bacterial cell wall that survives sterilisation, is the probable cause of two clusters of seven patients each, one in New South Wales and one at an American clinic, and a compounding pharmacy recalled three lots of glutathione vials on 5 August 2026 for the same failure, with no published link between them.

Is oxidised glutathione the same thing as glutathione?+

It is not, and FDA states so directly: oxidised glutathione is a dimer, two glutathione molecules joined, rather than a salt, and the two carry different ingredient identifiers. This matters because the most tightly controlled skin study I found, a split face trial of a 2 percent lotion in 30 women over 10 weeks, used the oxidised form on the skin rather than glutathione in a vein, and three of its four authors work for a glutathione manufacturer.

---

This content is for educational and informational purposes only and is not medical advice. Peptides discussed are research compounds and may not be approved for human use. Nothing here should be used to diagnose, treat, cure, or prevent any disease. Always consult a qualified healthcare professional before starting any peptide, supplement, or protocol. Individual responses vary. Do not self-administer compounds without proper medical supervision.