Pensum
Adaptive TDEE, explained
A formula gives you a first guess at your maintenance calories. An adaptive TDEE measures the real number from your own weigh-ins and food diary, and keeps measuring it.
TDEE stands for total daily energy expenditure, the calories your body uses in a full day. Eat at it and your weight holds steady over time. Many trackers work it out once from a formula and never revisit it. An adaptive TDEE is measured from two things you already record, what you eat and what the scale says, and it's re-measured as new days come in. Pensum does this every day and can move your calorie target with it.
What adaptive TDEE means
Energy you eat and don't burn gets stored, mostly as fat, some as glycogen and the water that comes with it. Run that backwards and expenditure follows from intake and weight change: over any stretch of days, it's your average daily intake minus the change in stored energy divided by the number of days.
The stored-energy part comes from your weight trend. Say you averaged 2,500 kcal a day for two weeks and your trend fell by half a kilogram. Pensum counts each kilogram the trend moves as 5,500 kcal, so that half kilogram is about 2,750 kcal taken out of storage, near 196 a day, and your expenditure comes out close to 2,700.
The textbook figure is about 7,700 kcal per kilogram, derived from the fat content of adipose tissue. Part of what the trend moves is lean tissue and glycogen with its bound water, which hold far less, and reading every kilogram as fat makes the estimate overreact to water. I use 5,500, fitted on real data and explained in the full expenditure algorithm.
Reading energy balance backwards from weight change is an established method, validated in 2015 against doubly labelled water, the gold standard for measuring free-living energy use.
Why a TDEE formula is only a starting estimate
A formula reads your height, weight, age and sex, estimates your resting metabolism, and multiplies it by an activity level you pick from a list. Two people with identical stats can burn a couple of hundred calories a day apart, and the formula gives both of them the same number.
That gap can decide whether a cut works. Losing half a kilogram a week takes a deficit of about 550 kcal a day. If your real maintenance sits a couple of hundred below the calculated one, a large part of that deficit is gone before you start, and the scale only shows it weeks later.
Pensum still starts from a formula, because on day one there's nothing else to go on. The TDEE calculator runs the same equations in your browser. With less than a full week of your own data, the app shows the starting value untouched. After a few weeks it no longer affects the number. I go into the formula's error in how accurate a TDEE calculator is.
How a tracker measures expenditure from weigh-ins and the diary
The raw scale is too noisy to read directly. It can move a kilogram between two mornings with no change in fat, most of it water.
Pensum first smooths the weigh-ins into a trend. Each new weigh-in moves the trend 10% of the way toward itself, which gives it a half-life of about a week. Missed days are filled with a straight line between the weigh-ins on either side.
Each day, the estimator then looks at the last two weeks. It takes the average calories of the days you tracked and subtracts the slope of the trend converted to energy. That daily reading is still noisy, so it feeds a second, slower average that puts 3% weight on each new reading. The data behind the number you see is on average five to six weeks old, and the number moves by a few kcal a day.
I made it slow on purpose. A real change in your expenditure, a new job on your feet for example, takes six to nine weeks to show halfway, and in return a salty dinner or a heavy weekend barely registers. Three limits sit on top of the averaging:
- The number moves at most 100 kcal from one day to the next.
- It stays within 40% of the formula value. A reading outside that band means the input data is broken.
- A tracked day whose calories fall below half the two-week median is dropped, because it looks like breakfast logged and dinner forgotten.
For the first month the average runs faster, so a wrong starting value gets corrected quickly. Each recompute replays up to a year of your diary, so the average starts from your own history.
What adaptive TDEE needs from you
It needs regular weigh-ins. The trend fills gaps, but each real weigh-in makes its slope more reliable. Weighing in most mornings, at the same point in your routine, gives the estimator the most to work with.
It also needs a consistent diary. People under-report what they eat by around 20% on average in studies against doubly labelled water. If your diary reads 15% light, the measured expenditure reads about 15% low as well. Your target is computed in the same under-logged units, so the deficit between the two, which is what drives weight change, still lands where it should.
That holds as long as your habits stay the same. If you change how you log partway through, for example moving from photo estimates to a kitchen scale, the number shifts for reasons unrelated to your metabolism, and it takes a couple of months to settle again.
Three situations that break a naive adaptive TDEE
A naive version runs the equation on whatever the scale and the diary say. Each of these feeds it a false signal.
Switching between cutting and bulking
When you end a cut and start eating in a surplus, your muscles refill glycogen, and each gram of it binds roughly three to four grams of water. That puts one to three kilograms on the scale over a few days. The equation reads that as stored energy and pushes a naive expenditure down exactly when the real one is steady or rising. For 28 days after a change of direction, Pensum counts each daily reading at a fifth of its usual weight and nudges the estimate the way physiology moves when you switch diet phase.
Missing and under-logged days
If you log four days out of seven, the estimator fills the gaps with the median of the same weekday's recent intake, because skipped days tend not to be typical ones. If you log almost nothing for a stretch, it pauses and holds the last value.
Period water
Around a period the scale typically rises by about half a kilogram of water, highest on the first day of bleeding and gone within the following week. A naive estimate dips and recovers every cycle. If you log your period, Pensum subtracts the expected water from each weigh-in before the trend is computed, and learns your own pattern once two logged periods have weigh-ins around them. The research behind it is on the period tracking page.
Adaptive TDEE without a subscription
MacroFactor is the app known for adaptive expenditure, and it has no free version: after a 7-day trial, logging requires a subscription from $71.99 a year. In Pensum the measured expenditure and the targets that follow it are part of the free app, with no ads and no account.
A MacroFactor user sent me their export, covering a cut, maintenance and a lean bulk. It includes the expenditure MacroFactor showed each day, so I fed the same weigh-ins and calories into Pensum and compared. From January to October 2026 the two numbers were 15 kcal a day apart on average, and never more than 63. It's one person's data, and I keep checking as more exports come in. The day-by-day chart is in MacroFactor's expenditure number, free in Pensum, and prices and switching are in Pensum vs MacroFactor.
If you already have months of weigh-ins and diary days in another tracker, you can import that history and the estimator runs on it right away. Switching calorie trackers goes through the exports app by app. Starting fresh, the number begins at the formula value and becomes yours over the first few weeks. After a month, compare your weight trend with what the measured expenditure implies at your logged intake.