Decision 10 of 10

Why the meal it preselects depends on where you are

When you log a meal, the app has already picked breakfast, lunch, dinner or snack before you type anything. It used to pick from four hours somebody in Belgium made up.

Written and maintained by Mattias Geniar. Last updated .

What the app does
Takes meal times from national time use diaries, per country, based on the time zone your phone reports. It never asks where you are.
How much we believe it
The differences between countries are real and large. A national average is not a person, and it is one chip you change with one tap.

The old hours were: before 11:00 breakfast, before 15:00 lunch, before 21:00 dinner, and snack after that. That is wrong for most of the people this app is translated for. It offered a Spaniard dinner at 15:00, and told a Norwegian eating middag at 16:30 that it was lunch.

Where the hours come from now

Eurostat's Harmonised European Time Use Surveys ask people to write down what they were doing in each of the 144 ten-minute slots of a day. Eurostat publishes, per country, the share of the population eating in each slot (dataset tus_20startime for the round collected around 2020, and tus_00startime for 2000 and 2010). That gives a curve of meal times with three peaks in it, and the peaks are never in the same place twice.

Three countries, one day, as their own diaries describe it Share of the population eating in each ten-minute slot, from 04:00 to midnight.
0% 10% 20% 30% 40% 50% 06:00 09:00 12:00 15:00 18:00 21:00 24:00

France, 2010 diary — peak 50.0% at 12:30 Spain, 2010 diary — peak 39.9% at 14:30 Norway, 2020 diary — peak 13.6% at 17:30

At 12:30 on an average day, 50.0% of France is eating. Spain's largest peak is at 14:30 and reaches 39.9%. Norway's evening peak is at 17:30 and never gets above 13.6%, because Norwegians do not all eat at the same time. Drawn from the published diaries rather than from a smoothed copy, so the peaks in the key are the peaks in the data. Generated by tools/meal-windows.js, the same script that produces the table below and the meal times in the app.

Each window in the table below is one of those peaks, widened until the curve drops to 65% of the peak. It stops at the quietest slot between two meals and never goes more than 90 minutes either side. Then it is rounded to the half hour. The diaries are kept in ten-minute slots, so the unrounded answer for Belgium is 07:10 to 09:40. Printing that would claim a precision nobody has: a few thousand diaries from 2010 do not know your breakfast to within ten minutes. Anything outside all three windows is a snack, which is most of the day. Your country comes from the time zone your phone reports, so we never ask.

Every window the app ships Breakfast, lunch and dinner per country, and the survey round the diary comes from.
CountryBreakfastLunchDinnerDiary
Austria07:00–10:0012:00–13:0018:00–19:302020
Belgium07:00–09:3012:00–13:0018:00–20:002010
Bulgaria07:00–10:0012:00–13:3019:00–21:002020
Canada06:30–09:0011:30–13:3017:30–19:30peak only
Croatia07:30–09:3012:00–14:3018:30–20:302020
Denmark07:00–09:3011:30–13:3017:30–19:30peak only
Estonia07:00–10:3012:00–13:3017:30–20:002020
Finland07:00–10:0011:00–13:3015:30–19:002020
France07:00–09:3012:00–13:3019:00–21:002010
Germany07:30–10:0012:00–13:3018:00–20:002020
Greece07:30–09:3013:30–15:0020:00–22:002010
Hungary07:00–09:3012:00–13:0018:00–20:002020
Ireland07:30–10:3012:30–14:3017:30–19:30peak only
Italy07:00–09:0012:30–14:0019:30–21:002010
Latvia07:00–09:0012:00–14:0018:30–20:302000
Lithuania07:00–10:0012:00–14:3018:00–20:302000
Luxembourg07:00–09:0012:00–13:3019:00–20:302010
Netherlands07:30–10:3012:00–13:0017:30–19:002020
Norway07:30–11:0011:30–13:3016:00–19:002020
Poland07:00–10:0013:00–16:0018:00–20:302020
Romania07:00–09:3013:00–14:0019:00–20:302010
Serbia07:30–10:3014:00–17:0018:30–20:302020
Slovenia07:30–10:0012:00–14:0018:30–20:302000
Spain08:00–11:0014:00–15:3021:00–22:302010
Türkiye09:00–11:3012:00–13:3018:30–21:302010
United Kingdom07:00–10:0012:00–14:0017:30–20:002010
United States06:30–09:0011:30–13:3018:00–20:00peak only
Anywhere else07:30–10:0012:30–14:0018:30–20:30the average of the 23
Generated from Eurostat by tools/meal-windows.js and written into both this page and resources/pwa/lib/meal-windows.js by the same run. A page saying “these hours come from diaries” while showing hours that no longer match the app would be worse than having no section at all.

The four rows marked “peak only” have no Eurostat diary. Their peaks come from Martín-Olalla & Rodríguez-Sánchez (2018), Table X, which reads the same figure off national time use surveys for 19 countries. That paper is also the best evidence that these differences are real rather than noise. Dinner falls about three hours after winter sunset, in a band one hour wide. That band is 40% narrower than the spread of dinner clock times. Europe eats by the sun, and it is the clock that disagrees.

For the United States, the USDA's Economic Research Service reports from the 2022 American Time Use Survey. Eating and drinking as a main activity peaks between 12:00 and 12:59 and again between 18:00 and 18:59. About 3 in 10 people are eating in each. Our US row fits inside both.

The evidence against, and what this is not

This says nothing about whether the time you eat is good for you. It is a default in a form. Whether meal timing affects health is a real and disputed field of research. We have not read enough of it to have a position, and logging at 03:00 gets you a snack rather than a lecture.

Several diaries are old. Latvia, Lithuania and Slovenia are from the 2000 round; Belgium, Spain, France, Italy, Greece, Luxembourg, Romania, Turkey and the United Kingdom are from 2010. Habits change. Where a country appears in more than one round we use the most recent, which is why the 2020 round is used for eleven of them.

“Eating” in a diary includes drinking, so a country that spends its morning with coffee has a long, low morning rather than a peak. That is exactly why the window is cut at 65% of the peak rather than at half of it. At half, the Dutch morning ran into lunch and the Finnish evening covered six hours. The 65%, the 90-minute limit and the rounding to half hours are engineering choices we made by looking at the curves, not findings. The rounding moves an edge by up to a quarter of an hour, which is less than the difference between two survey rounds of the same country.

A national average is not a person. Shift workers, students, anybody fasting, and anybody who simply eats late are all inside these numbers and none of them is described by them. Living in Madrid does not make your habits Spanish either. All of this is one preselected chip you change with one tap. Nothing overrides your choice afterwards, not even the analyser. A plate of pasta at 13:00 once came back as “dinner” and overwrote a correct lunch.

Four countries are derived rather than measured. Table X publishes peaks for employees on a weekday, who eat earlier than everybody else. So those four rows carry a correction. It is the median gap between employees and everybody else in the four countries the paper reports both for: 38 minutes at breakfast and roughly nothing at lunch and dinner. A median of four pairs is thin evidence, and it is the weakest number on this page.

And most of the world is not in here at all. There is no diary in this dataset for Japan, Australia, Brazil, India, Mexico or anywhere else outside Europe apart from the two North American rows. Everyone there gets the average of the 23 measured countries, which is a schedule nobody actually keeps. We would rather add real rows than invent them, so if you know of a national time use survey with an eating-by-time-of-day breakdown we have missed, send it to us.