G6PD Genotype

Some medications destroy red blood cells in a few people and leave everyone else alone, and the difference is usually one gene. If you carry a deficiency variant, a course of the antimalarial primaquine, a dose of rasburicase during cancer treatment, or the antibiotic dapsone can break apart enough red cells to cause acute anemia within a day or two. Many people find out this way, unless newborn or family testing finds it first.

This test reads the gene itself rather than measuring how hard the enzyme is working on the morning of the blood draw. That difference matters most for women, who can carry a faulty copy and still look completely normal on the standard enzyme test. You get the answer once and keep it for life.

Inside the Red Cell

The gene sits on the X chromosome and makes an enzyme called glucose-6-phosphate dehydrogenase, usually shortened to G6PD. Every cell in your body makes it, but mature red blood cells depend on it completely. They have no backup system, so this one enzyme helps keep their protective chemistry recharged while they carry oxygen all day.

Take that defense away and the sequence is fast. Hemoglobin clumps inside the cell, the membrane gives out, and the cell is destroyed, either inside the vessels or by the spleen. A completely absent enzyme is not compatible with human development at all, which is why every variant that shows up in living people leaves at least some function behind.

Drugs That Set It Off

The most usable thing this test gives you is a prescribing rule. The Clinical Pharmacogenetics Implementation Consortium writes gene-based prescribing guidance. Its G6PD guideline sorts more than 30 medications by risk in people with a deficiency genotype: avoid the high-risk ones, use the medium-risk ones cautiously, treat the rest normally. Primaquine, tafenoquine, rasburicase, and dapsone are at the sharp end of that list.

For the two antimalarials that clear the dormant liver stage of vivax malaria, primaquine and tafenoquine, testing before the first dose is standard practice rather than a precaution. Nobody wants to discover a deficiency by watching hemoglobin fall on day two of treatment. If you carry a variant, the result belongs in your pharmacy record and your chart, not in a folder at home.

Fava Beans and Infections

Fava beans are the food trigger with real evidence behind them, and they are where the older name for the condition comes from. In a real-world Israeli cohort of about 32,000 people with deficiency, fava beans accounted for roughly 72% of the hospitalizations for major hemolysis and infection for about 9%, both ahead of drugs. Infections are the trigger people hit more often than a risky prescription, because an immune response generates the same kind of chemical stress a drug does. In a study of adult dengue in Thailand, the hemoglobin trajectory through the illness differed depending on which variant a person carried.

Jaundice in the First Days of Life

In newborns, the danger is not usually anemia but bilirubin. Red cells break down faster than a newborn liver can clear the pigment they release, and severe jaundice left untreated can injure the brain permanently. Phototherapy and, rarely, exchange transfusion prevent that, which is the reason so many countries screen at birth. If you carry a variant, this is the newborn risk your family needs to know before a baby is born.

Why Carrier Women Get Missed

Women have two X chromosomes, and each red cell silences one of them at random early in development. A carrier ends up with a mosaic: some cells with a working enzyme, some without. The enzyme test averages across that mixture and often comes back reassuringly normal.

How often is not a rounding error. Among 4,067 Italian newborns sequenced alongside routine screening, more than 60% of the girls carrying a pathogenic variant had normal enzyme results. Work in Thailand puts the share of carrier women misclassified as normal by enzyme testing alone at 40% to 44%. Those women can still have red cells break apart when the trigger is strong enough.

Sources: Bombaci et al. (2026); Tan et al. (2024); Martin et al. (2025).

A normal enzyme result is not a clean bill of health, and a newborn screen that came back fine decades ago does not settle the question either. Clinical-grade DNA testing is the part enzyme testing lacks, particularly for women and before a high-risk drug.

Why These Variants Spread

These variants are common in exactly the places malaria has been common for centuries, and that is not a coincidence. Carrying the Mediterranean variant cut symptomatic vivax malaria by 76% in men with one copy and women with two copies. A meta-analysis of African populations found protection against uncomplicated malaria, concentrated in carrier women, with no protection against the severe form.

A large case-control analysis then reported that deficiency lowered the risk of cerebral malaria while raising the risk of severe malarial anemia. That opposing-effects result is contested rather than settled: a later reanalysis argued that the apparent protection against cerebral malaria can be manufactured by how severe cases are selected and classified, not by biology. The simpler point survives the argument. A fragile red cell is a poor home for a parasite and a poor home for you, so the same variant can help in one setting and hurt in another, depending on what you are exposed to.

A Blood Sugar Test That Reads Low

Here is the consequence almost nobody anticipates. A1c, the test that estimates your average blood sugar over three months, works by measuring sugar stuck to hemoglobin inside red cells. If your red cells are being replaced faster than usual, they spend less time collecting sugar, and the number comes back lower than your actual glucose warrants.

In a large UK cohort study, undiagnosed deficiency among Black and Asian men biased A1c downward by an average of 0.9%, enough to keep people under the diagnostic line for years. In those men, diabetes was diagnosed 4.1 years later on average, and small-vessel complications were about 37% more likely. Separate work in African-born adults in the United States and in East Asian men points the same direction: A1c underperforms as a screening test when this enzyme is deficient.

So if you carry a variant, stop letting A1c alone rule out diabetes. Use fasting glucose, or an oral glucose tolerance test if there is any doubt. The evidence here is strongest in men, who have a single X chromosome and are therefore more often fully deficient, but the same logic applies to any carrier whose red cells are turning over faster than usual. For many carriers, this may change more everyday decisions than the drug list does.

What the Genotype Can and Cannot Tell You

DNA testing tells you whether you carry a variant, including in the people enzyme testing overlooks. It does not tell you how much working enzyme you have circulating right now, and in carrier women the gap between those two things is wide. Two women with the identical variant can land in very different places depending on how their cells divided the X chromosomes years ago.

Variant classification is still catching up. A systematic cataloging effort identified 1,341 variants in this gene and could supply functional interpretations for 268 of them. Even position is not destiny: different DNA changes affecting the same protein spot have produced clearly different clinical pictures. The genotype and a quantitative enzyme measurement answer different questions, and the pair matters most when the result will drive a drug decision.

When a Result Can Mislead

• Panel coverage: a negative result clears the variants the assay was built to read, not every possible change in the gene. Rare and locally common variants such as Orissa, Coimbra, and Union have repeatedly surfaced only when researchers sequenced past the standard panel.
• Ancestry: which variants matter depends on where your family comes from. The A- combination dominates in African populations, Mediterranean in the Middle East and southern Europe, and Canton, Kaiping, Gaohe, and Viangchan across East and Southeast Asia.
• Unclassified findings: a report may name a variant whose clinical meaning nobody has established yet. Reports may call this a variant of uncertain significance. Add enzyme testing and specialist input instead of assuming it is harmless.
• Consumer chips: a consumer test may check only a few common variants. Confirm a positive or negative consumer result with clinical-grade testing before a high-risk drug, a pregnancy decision, or family testing.
• Somatic versus inherited findings: a variant found on tumor sequencing, or a blood DNA test after a bone marrow transplant, may not be your inherited result. Confirm with germline testing before using it for family decisions.
• A normal enzyme test afterward: recent hemolysis, a high reticulocyte count, or a transfusion all push enzyme readings up temporarily. A normal activity result under those conditions does not overrule a confirmed variant.

One Result, Decades of Use

Your sequence does not change, so there is no trend to follow and no reason to repeat this test. The only situation that warrants a second look is a call you do not trust: a result from a limited panel or a consumer chip is worth confirming with a clinical-grade DNA test before you build decisions on it.

What does need ongoing attention is everything downstream. If you carry a variant, get a quantitative enzyme activity measurement to pair with it, especially if you are a woman. Check hemoglobin and a reticulocyte count during or right after any acute illness or trigger exposure. Use glucose-based testing whenever diabetes is being screened, rather than leaning on A1c.

Acting on an Unexpected Result

A confirmed variant should trigger four moves. Add quantitative enzyme activity so you know where you sit functionally. Get the result into your pharmacy system and primary care chart, since the whole value of knowing is that it fires before a prescription, not after. Before travel to a malaria region, tell the travel clinic rather than waiting to be asked.

If you have ever had unexplained anemia, yellowing eyes, or dark urine after an infection or a new drug, order a complete blood count, reticulocyte count, bilirubin, lactate dehydrogenase, and haptoglobin during or shortly after the next episode. Together those show whether red cells were actually destroyed, which turns a suspicion into a documented event.

Bring in a hematologist if you carry one of the rare severe variants that cause hemolysis without any trigger, or if the report names a variant of unclear significance. Bring in a genetic counselor if you are planning a pregnancy or if there is a history of severe newborn jaundice in the family. Then tell your relatives: a mother of an affected son is usually a carrier, an affected father passes the variant to every daughter and no son, and a carrier mother has a 50% chance of passing the variant to each child.