Keeping a gluten and allergen log

A gluten-free week is not one decision. It is forty or fifty small ones, most of them made in a shop aisle in under ten seconds, and almost none of them remembered by Thursday. A log exists to fix the remembering, not the deciding.

Ask anyone who has been asked “what did you eat this week?” in a consultation. The honest answer is a shrug and a partial reconstruction, weighted heavily towards whatever went wrong. Human memory for routine food is bad in a specific way: it keeps the exceptions and quietly discards the ordinary, which is the opposite of what a useful record would do.

That is the gap a scan history fills. Every check you already make — a barcode in the supermarket, a photograph of a plate, a look at an unfamiliar ingredient — is a data point you generated and then threw away. Keeping them costs nothing extra, because you were going to make the check anyway.

What the daily view records

The app's home screen keeps a running tally for the day: the total allergen load recorded so far, how many products you have checked, how many came back clear, and the trend across the week. It is a count of decisions, not a physiological measurement, and reading it that way keeps it honest.

The useful pattern is usually not the number itself but its shape. Three checks on Monday and nineteen on Saturday says something about where the effort is going. A run of days with no entries at all is not evidence of a clean week; it is evidence that the log went quiet, which is a different thing and worth knowing.

Log the reason, not just the verdict

A result that says only contains gluten is nearly useless a week later. Every scan result names the ingredient that triggered it, lists the other allergens found, and carries a confidence score, and those three things are what make the entry legible after the fact.

The confidence score matters most. A product whose ingredient list names wheat flour outright is a different kind of answer from one where the catalogue entry is sparse or the wording is ambiguous, and a log that flattens both into a red mark will mislead you later. Looking back at an entry and seeing this was a strong match on barley malt extract rather than this was a guess is the difference between a record and a mood.

A log is not a tolerance test. Controlled trials have repeatedly found that symptoms track gluten exposure poorly, in both directions — a quiet day is not proof of a clean one, and a bad day is not proof that something got through. A history of scans records what you checked and what the labels said. It does not establish what you react to.

The meals with no barcode belong in it too

A log that only contains packaged food describes a supermarket, not a week. Restaurant plates, canteen lunches and anything from a counter are exactly the entries most likely to matter, and they are the ones with nothing to scan.

For those, the app analyses a photograph: it identifies the dishes in the picture and flags the allergens each one is likely to carry. This is genuinely weaker evidence than an ingredient list — it reads a plate, not a recipe, and it cannot see a shared fryer — so it is best read as a range rather than a verdict. It is still far better than an empty row. Checking food with no barcode goes through that in more detail.

Comparing beats judging

Foods can also be ranked by allergen content score, high to low, which turns an isolated yes-or-no into a choice between two candidates. Standing in front of four brands of stock cubes, the question is rarely “is this one acceptable” and almost always “which of these four is the least bad”.

A history makes that comparison cumulative. Once a brand has come back clear four times, it stops being a decision and becomes a default, and the ten seconds it used to cost go somewhere else. Most of the value of logging is in the decisions it lets you stop making.

Bringing it to someone who can read it

The one use of a food log that is unambiguously worth the effort is handing it to a clinician or dietitian. They are trained to see patterns in it, they will ask about things you would not have recorded on your own, and they can tell the difference between a signal and a coincidence. What they cannot do is work from a reconstruction produced in the waiting room.

Interpreting the log yourself is where it goes wrong. The temptation to draw a line from Tuesday's rough afternoon to Monday's unfamiliar sauce is strong and the reasoning is unreliable, because a week contains far too many variables and the body offers no confirmation either way. Collect carefully; conclude cautiously; leave the interpretation to a person who knows your history.

What stays on your phone

A food log is one of the more revealing records a person can keep — where you shop, what you can afford, when you eat, when you stopped. GlutenScanning requires no account, keeps your data on your device, collects no personal information and treats analytics as optional and anonymous. That is a claim worth checking in the privacy policy, and worth checking against any other app you would trust with the same history.

This page is information, not medical advice. GlutenScanning is a reference tool. It does not diagnose any condition, does not establish that a food is safe for you, and a log it produces is not evidence about what you can tolerate. Decisions about a gluten-free diet, and any reading of a symptom pattern, belong with a clinician or dietitian who knows your history.

Common questions

What does the daily tally actually count? +

The home screen shows the allergen load recorded for today, how many products you have checked and how many came back clear, with the trend across the week alongside it. It is a count of the checks you made, not a measurement of what your body absorbed.

Can a food log tell me what I react to? +

No, and it is worth being firm about that. Research on low-dose gluten exposure has repeatedly found that symptoms track exposure poorly in both directions. A log is a record of what you checked and what the labels said. Interpreting a pattern in it is work for a clinician or dietitian.

Why does each result carry a confidence score? +

Because the underlying evidence varies. A product whose ingredient list names wheat flour outright is a different kind of answer from one where the catalogue entry is thin or the wording is ambiguous. The score is there so a weak answer does not get logged as though it were a strong one.

Does food with no barcode go into the log? +

Yes. Meals can be photographed rather than scanned; the analysis identifies the dishes in the photo and flags the allergens each is likely to carry, and that goes into the same daily picture as a packaged scan. It is a weaker signal than an ingredient list and worth reading as a range rather than a verdict.

Where is the log stored? +

On your device. GlutenScanning requires no account, collects no personal information and treats analytics as optional and anonymous. A food log is an intimate record, and that claim is worth checking against the privacy policy of any app you let keep one.

Start the record with the next thing you buy

GlutenScanning checks a barcode against gluten and 60+ other allergens, names the ingredient behind every verdict, and keeps the day's tally on your phone. Free, no account required.

Download on the App Store: GlutenScanning for iPhone

Keep reading

“Gluten-free” versus “may contain”
657 products tested, and only one of the two claims predicted failure.
Food that has no barcode
What to do with a plate, a counter or a loose ingredient.
Checking a whole cart
Where the risk actually sits in a supermarket shop.
The research library
The evidence behind the numbers this site quotes.