Does cooking change the purine content of food?

Most purine charts list one number per food. Open a database that records the preparation as well, and that turns out to be a fiction. The same fish, measured raw and measured grilled, can differ by 160 mg per 100 g — more than the entire gap between the "low" and "high" bands.

This matters because you do not eat foods, you eat preparations of them. Here is what the measurements actually show.

Drying concentrates purines, dramatically

Drying removes water. The purines stay. So 100 g of a dried food started life as several hundred grams of the fresh one, and the number per 100 g rises accordingly.

The clearest case in this database is anchovy:

Preparation	Purines per 100 g
Anchovy, fresh	303.9 mg
Anchovy, raw	411 mg
Anchovy, canned	321 mg
Anchovy, dried	1,109 mg

Dried anchovy is the third-highest value in the entire 830-food database, behind two raw fish skins. It is also, in practice, a seasoning — you might use 5 g of it, contributing about 55 mg, less than a modest serving of chicken. The headline number is alarming and the practical contribution often is not. Both facts are true, and a chart that shows only the first one is the reason people end up avoiding the wrong things.

The same pattern runs through mushrooms. Fresh tea tree mushroom measures 48 mg; dried, 293 mg. Fresh king oyster mushroom is 94 mg; dried abalone mushroom is 424 mg. If you have ever wondered why mushrooms appear on both "safe" and "avoid" lists, this is why — the lists are describing different states of the same ingredient.

Grilling and roasting raise the number too

Dry heat drives off moisture, so the effect is the same in a milder form:

• Cod — 71 mg fresh, 230 mg baked
• Snakehead fish — 169 mg raw, 214 mg cooked
• Crucian carp — 154 mg raw, 190 mg cooked
• Grass carp — 134 mg raw, 162 mg cooked

Note what happens to the labels. Under the NHC bands raw cod is fairly high and baked cod is high — the top band. Nothing was added to it; water left.

Boiling can move purines the other way

Purines are water-soluble, so boiling can leach some of them into liquid you then discard. Tofu shows this clearly, because the database happens to include timed measurements:

Preparation	Purines per 100 g
Soybean (dried)	218 mg
Tofu, nigari	100 mg
Dried tofu (soft)	94 mg
Water tofu (firm)	68 mg
Tofu, deep-fried	54.4 mg
Tofu, boiled 5 minutes	35.8 mg
Tofu, boiled 3 minutes	21.9 mg

From dried soybean to briefly boiled tofu is a tenfold drop. Some of that is processing and water content rather than boiling alone, but the direction is consistent and it is the opposite of what drying does.

There is a corollary people miss: the purines go into the broth. A stock simmered from bones or meat carries what left the solids. Beef bone broth after hotpot appears in this database as its own entry for exactly that reason. If you are being careful about a meat dish, the soup made from it is not the safer half of the meal.

What this means when you read any purine chart

Three practical habits follow from the data.

Check which preparation the number describes. If a chart does not say, it is quoting one measurement and implying it covers all of them. On this site every row names the preparation it was measured in, and each food page lists all the variants side by side.

Convert to your portion before judging. A 20 g garnish of a 220 mg food contributes 44 mg. A 250 g main course of a 105 mg food contributes 262 mg. The second one is on nobody's warning list. Portion arithmetic beats category labels almost every time.

Prefer the wet method when it is a real choice. Where you would happily eat something boiled or in a broth-discarded preparation, the number is lower — sometimes much lower. Where you would not, forcing it is not worth the misery. A diet you abandon has an effect size of zero.

The honest limit of all this

Purine content is a property of food. It is not a prediction about your uric acid, let alone about whether you will have a flare. Most uric acid is produced by the body rather than absorbed from meals, and the measured effect of diet alone on serum urate is modest next to urate-lowering medication.

What these numbers are genuinely good for is comparison: this preparation versus that one, this portion versus that portion, tonight's option A versus option B. That is a real and useful thing, and it is worth knowing that the comparison changes depending on how the food met the heat.

Search the full database to see every preparation recorded for a food, or start with the high-purine list — where dried and grilled foods dominate the top for exactly the reasons above.