The most common enzyme deficiency in humans, and for almost everyone who has it nothing ever happens. It matters on the days it does: a short list of medicines, and a bean.
Solid lines are connections this site curates. Dashed lines mean the two ends share a research paper — worth knowing, and not a claim that one explains the other.
Glucose-6-phosphate dehydrogenase is the first enzyme in a pathway whose only real job, in a red blood cell, is to keep a supply of something that mops up oxidative damage. A red cell has no nucleus and cannot make new proteins, so it lives its four months on the enzymes it started with. When that supply runs short and something oxidising arrives, the cell breaks.
Roughly 400 million people carry a deficient version. It is the most common enzymopathy there is, and it is common for the reason most common things in human genetics are common: it offers some protection against malaria, so it kept being passed on where malaria was.
Most people with G6PD deficiency go through life without an episode. The enzyme is not absent, it is reduced, and ordinary life does not put enough oxidative stress on a red cell to matter.
What causes trouble is a specific and knowable list: certain medicines, infection, and fava beans. The last one is not folklore — the condition's other name is favism, and eating broad beans can trigger haemolysis severe enough to need transfusion in some people who carry certain variants.
This is the part a genotype actually changes. Several ordinary drugs are safe for nearly everyone and are not safe here — and in each case there is usually something else to prescribe.
The full list and what to do about each is on the drugs pages.
Two reasons, and both matter.
The first is that the gene sits on the X chromosome. A male has one copy, so a deficient allele means a deficient enzyme. A female has two, and which one is active differs from cell to cell — so a heterozygous woman can have anywhere from near-normal to clearly reduced activity, and her genotype does not tell you which.
The second is coverage. More than two hundred deficiency-causing variants are known. A consumer array carries a handful of the most common ones. Not finding a variant is not the same as not having the condition.
The test that answers the question is a measurement of enzyme activity — and even that reads falsely normal during an episode, because the cells with the least enzyme have already broken and the young ones replacing them have more.
If you are told you have it, the useful thing is not a diet or a supplement. It is that the information should be in your medical record before somebody prescribes for you in a hurry — and that is a conversation with a doctor, not an entry in a genotype file.
Source. Gammal et al., CPIC Guideline for G6PD Genotype and Medication Use (Clin Pharmacol Ther 2023, PMID 36049896). The guideline assigns G6PD phenotype from genotype and lists medications for which the phenotype alters prescribing.
Genetics. X-linked. Hemizygous males and homozygous females with two deficient alleles are deficient. Heterozygous females are the difficulty: lyonisation produces a mosaic of cells expressing either allele, so enzyme activity spans the range and phenotype cannot be assigned from genotype. CPIC classifies such individuals as indeterminate rather than assigning a phenotype.
Alleles listed here. rs1050828 (c.202G>A, p.Val68Met) and rs1050829 (c.376A>G, p.Asn126Asp). The A- allele carries both; the A+ allele carries rs1050829 alone and has near-normal activity. This is why the two positions are not independently interpretable, and why an array reporting one without the other does not settle the question. The Mediterranean allele rs5030868 and many others are not represented on typical consumer arrays.
Testing. Quantitative enzyme activity is the diagnostic standard. Activity may be spuriously normal during or shortly after an acute haemolytic episode because the surviving and newly produced erythrocytes have higher activity; retesting after recovery is standard practice.
Clinical use. Rasburicase is contraindicated. Several other agents require avoidance or altered monitoring; the CPIC guideline is the reference. The relevance of this genotype is almost entirely prescriptive — there is no treatment for the deficiency itself, and no restriction on ordinary life other than the avoidance of known triggers.
What a 23andMe/AncestryDNA export or raw VCF can and can't tell you about G6PD Deficiency comes down to these specific, well-studied positions — not a diagnosis.
Research-derived gene–drug associations only — not a prescription, dosing guide, or medical advice. Always follow your prescriber's guidance.
| Gene | Drug | What the research shows |
|---|---|---|
| G6PD | Rasburicase | Contraindicated. Rasburicase carries a boxed warning against use in G6PD deficiency: it generates hydrogen peroxide as it breaks uric acid down, which is precisely the oxidative stress a G6PD-deficient red cell cannot handle. Severe haemolysis and methaemoglobinaemia have been reported. (CPIC Guideline for G6PD Genotype and Medication Use (Clin Pharmacol Ther 2023, PMID 36049896).) |
| G6PD | Primaquine and tafenoquine | These clear the dormant liver stage of malaria and both cause dose-dependent haemolysis in G6PD deficiency. Testing before prescribing is standard where they are used, which makes this one of the few places in the world where a pharmacogenetic test is genuinely routine. (CPIC Guideline for G6PD Genotype and Medication Use (Clin Pharmacol Ther 2023, PMID 36049896).) |
| G6PD | Dapsone, nitrofurantoin, methylene blue, toluidine blue and pegloticase | All are oxidising agents that can precipitate acute haemolysis in a G6PD-deficient person at ordinary doses. In most cases an alternative exists, which is why knowing the phenotype before prescribing is the whole point — there is nothing to do about the deficiency itself. Toluidine blue, a dye closely related to methylene blue and used to stain tissue during surgery, is on the same list for the same reason, and is the kind of exposure nobody thinks of as a prescription. (CPIC Guideline for G6PD Genotype and Medication Use (Clin Pharmacol Ther 2023, PMID 36049896).) |
Not necessarily. More than two hundred deficiency-causing variants are known and a consumer array carries only a few of the most common. The question is answered by measuring enzyme activity, not by reading a file.
Genuinely uncertain, and that is the correct answer rather than a hedge. The gene is on the X chromosome and which copy is active differs from cell to cell, so activity in heterozygous women ranges from near-normal to clearly reduced. CPIC classes this as indeterminate rather than assigning a phenotype. An enzyme measurement settles it.
If you have confirmed G6PD deficiency, yes — favism is real and can be severe. If you have not confirmed it, the useful step is the enzyme test rather than a change of diet based on a genotype file.
Free to reuse. This page's text is original writing from freely-available research, licensed CC BY 4.0 — reuse it, including commercially, with attribution to MyGeneLog. It's general research-derived information, not medical advice or a diagnosis — see Terms of Use.