Should you skip period days in your TDEE? We tested it

Leaving your period days out of your TDEE sounds like it should help. The scale often jumps a pound or two around a period, and that's water, not fat. So why let it confuse your estimate?

We tested it, and for most people it does the opposite. Across 306 test runs, skipping the food you log on period days made the average estimate about four times less accurate. It also reads low: it tells you that you burn less than you do. Zolt's "Pause TDEE on period days" switch is now off by default.

These are averages across 51 user profiles, not a rule for your body. The switch is still there, and for a small group it can still help.

The short version

Average error with the switch off, then on

47 → 184cals a dayAverage error with the switch off, then on

How far low the estimate reads with it on

170cals a dayHow far low the estimate reads with it on

How long your TDEE is paused

11 to 19days a cycleHow long your TDEE is paused

How often the switch helped

2 in 100test runsHow often the switch helped

• Your weight often goes up a little around your period. It's water, and it comes back off.
• Zolt's switch skips the food you log on period days. Across 306 test runs on 51 user profiles, that raised the average error from 47 to 184 cals a day.
• It doesn't only pause your estimate. It pulls it low, by about 170 cals a day on average, because it keeps the weeks where water goes on and throws out the weeks where it comes off.
• It also pauses your TDEE for 11 to 19 days of every cycle, depending on how often you log.
• It helped about 2 in 100 test runs. If your estimate has tracked well with it on for a few cycles, you can keep it on.

Why period days look like a problem

Weight really does move with the cycle. In a study of 42 women weighed twice a week, body weight was about 0.45 kg (1 lb) higher during menstruation than in the first week of the cycle, and almost all of that was extracellular water. Read the study. In a year-long diary study of 765 cycles, self-reported fluid retention peaked on the first day of flow. Read the study.

So the worry makes sense. Zolt learns your TDEE by comparing what you eat with how your weight changes. A pound of water looks like a pound of change. If you can watch your adaptive TDEE move with your own numbers, you've probably seen it dip around your period and wondered why.

What the switch actually does

When "Pause TDEE on period days" is on, every day you log flow is marked as a period day. Zolt then leaves out that day's food. Your weigh-in for that day still counts.

Zolt needs at least 4 logged food days in a week before that week can count toward your TDEE. A 5-day period can take a week below that line, even if you logged every single day.

Same person, same food log

A normal week

1. Day 1Logged
2. Day 2Logged
3. Day 3Logged
4. Day 4Logged
5. Day 5Logged
6. Day 6Logged
7. Day 7Logged

7 food days. This week counts.

Period week, switch on

1. Day 1Logged
2. Day 2Skipped
3. Day 3Skipped
4. Day 4Skipped
5. Day 5Skipped
6. Day 6Skipped
7. Day 7Logged

2 food days, and a week needs 4. This week does not count.

Food loggedPeriod day, food skipped

When a week doesn't count, your TDEE holds at its last value until there's enough data again. The app tells you when this happens. With a period every month, it happens every month.

How we tested it

To measure how far off an estimate is, you need to know the right answer, and nobody knows a real person's true burn to the calorie. So we built user profiles with a known burn, gave them realistic day-to-day swings in food and weight, and ran Zolt's real TDEE engine on their logs every day for 100 days.

We built 17 kinds of user and ran each with three logging habits, for 51 profiles:

• Burns of 2,000, 2,400 and 3,000 cals a day.
• Different water patterns: a short bump around the period start, water that builds over the week before and clears in the first days of flow, a smooth monthly swing, and no swing at all.
• Periods of 3, 5 or 7 days. Regular cycles and irregular ones from 25 to 40 days.
• Maintaining, cutting, bulking, or getting more active partway through.
• Eating about 250 cals more in the 10 days before a period. A meta-analysis of 15 datasets found people eat about 168 cals a day more in the second half of the cycle. Read the review.
• Logging every day, logging food 6 days a week, or logging food 5 days a week.

Each case ran with 6 random seeds, and then again with 6 different seeds to check the results held up. We compared five options: keep every day, the app's switch (skip period food), skip period food and weigh-ins, skip only period weigh-ins, and skip weigh-ins from a week before each period to 3 days after.

What we found

Average error across all 51 test cases

Keeping every day was the most accurate option by a wide margin. The switch was close to the worst. The second set of seeds gave nearly the same numbers.

Here's what that looks like for one profile: someone who burns 2,000 cals a day and logs food most days.

One test profile over 100 days

Switch offSwitch onEstimate paused (switch on)Period days

Skipping weigh-ins also made things worse, just less so. In an earlier round of tests, a user with a steady 28-day water swing went from 14 cals a day of error to 163 when the weigh-ins around each period were hidden. The cycle swings up and down, and removing one part of it leaves a lopsided picture.

The pattern held for every kind of user we built. A few of them, averaged over the three logging habits:

Average error in cals a day. Longer periods and bigger water swings made the switch worse. Users with no water swing were hurt least, but still hurt, because skipping days only removes data.

Why it reads low

The water you gain before a period and the water you lose during it should cancel out. The switch breaks that.

Why the estimate drifts low

The week before your period

Water comes on and the scale goes up. No period days, so the week counts. It says you burned a bit less than you ate.

Kept

The week of your period

Water comes off and the scale goes down. This week would say you burned a bit more. But most of its food days are skipped, so it does not count.

Thrown out

The days before your period have no flow logged, so their food counts and that week stays in. The scale is creeping up, so that week says you burned a little less than you ate. The week of your period is when the water comes off. That week would say you burned a little more, but most of its food days are skipped, so it gets dropped.

Over and over, Zolt only sees the half of your cycle that says "you burn less." In our tests, the switch read low by about 170 cals a day on average. For the 7-day, 3.5 lb case it was about 500 low.

That matters if you use your TDEE to set a calorie target. A TDEE that reads 170 low means a target 170 lower than it needs to be, every day.

Appetite can push it the other way. When we gave a profile extra hunger before each period but no water swing, the switch read high instead, by 33 to 84 cals a day across the two seed sets. Either way, the estimate moves for reasons that have nothing to do with your metabolism.

How long it pauses

Days the estimate is paused, per 28-day cycle

Switch offSwitch on

Even with perfect logging, the switch paused the estimate for about 11 days of every 28. If you miss a day or two a week, each missed day brings that week closer to the 4-day line. At 5 logged days a week, the estimate was paused for about two thirds of each cycle. A paused estimate can't react if something real changes, like starting a cut or a new training block.

Who it can still help

These numbers are averages across many profiles, and your own results can land well away from them.

Who each option helped, out of 306 test runs

HelpedAbout the sameHurt

The switch helped 7 of 306 test runs, about 2 in 100. Skipping weigh-ins around each period, rather than food, helped more: about 1 in 7 runs. Those were mostly people who weigh in about 4 times a week and have a clear spike around their period, where one heavy weigh-in can drag a whole week. It still hurt most runs, so we don't recommend it as a default either.

If you've used the switch for a few cycles and your TDEE has matched your real results, keep it. You have evidence about your own body, and that beats an average.

What we changed in Zolt

"Pause TDEE on period days" is now off when you set up cycle tracking. If you already had it on, it stays on until you change it. You'll find it in Cycle settings and on the TDEE page.

With it off, period days count like any other day, and your period is still tracked.

What to do instead

• Keep logging your food on period days. It's real food, and your TDEE needs it.
• Weigh in as usual. A single heavy day is less of a problem than a missing week.
• Read the trend over a few weeks, not a single week. The water comes back off.
• Compare the same point in each cycle when you check progress. Reading your cycle pattern walks through how.
• If your estimate looks off for more than a cycle, check these common causes before changing settings.

What this test can't tell you

Knowing each profile's true burn is what let us measure the error, but your water pattern may not match any of the shapes we tested. Our next step is checking anonymized real logs from people who track their cycle in Zolt, and we'll update this post if the result changes.

The results are also specific to how Zolt estimates TDEE. Another app that handles missing days differently could get a different answer. For the basics of how an adaptive estimate learns, read understanding adaptive TDEE.

This is about calorie estimates, not medical advice. Talk to a doctor about severe or new swelling, or bleeding that changes a lot from cycle to cycle.

Common questions

Should I turn the switch off?

For most people, yes. In our tests it made the estimate less accurate for about 9 in 10 test runs. If it's been working for you over several cycles, keep it on.

Does my period change how many calories I burn?

We didn't test that: every profile burned the same amount all month. The swing you see on the scale around a period is mostly water, as the studies above found. Keeping every day in lets Zolt pick up whatever real change in burn there is.

Should I skip weighing myself on my period?

Not as a habit. In our tests, skipping period weigh-ins nearly doubled the average error, from 47 to 85 cals a day. Weigh in as usual and let the trend do its job.

Why did the switch exist in the first place?

Hiding noisy days sounds reasonable, and we assumed it would help. It took testing to find out it mostly didn't.