How to choose your activity level honestly

Every other input to a calorie calculator is a fact you can measure. Your height is your height. Activity level is the one field that asks you to summarise your entire life in a single word, and then multiplies your whole result by it.

On a resting rate of 1,639 kcal, sedentary gives 1,967 and extra active gives 3,114. That is a 1,147 kcal spread produced by nothing but a dropdown: larger than the effect of being twenty years older or twenty kilograms heavier. Get this wrong and no amount of precision elsewhere rescues the number.

Note what these descriptions are counting: exercise sessions. That is the convention nearly every calculator follows, and it is also the source of most of the confusion, because exercise is not where most of the variation between two people's days actually lives.

People choose their level from how their training feels rather than from how their day looks. Four hard gym sessions a week is genuinely demanding, and it is entirely reasonable to think of yourself as an active person because of it. But an hour of training is one twenty-fourth of a day. If the other twenty-three hours are spent sitting at a desk, in a car and on a sofa, the honest description of that week is closer to lightly active than to very active.

The reverse error exists too and is much rarer: someone who does not formally exercise but works a warehouse shift, walks fifteen thousand steps and lifts things all day will underestimate badly by picking sedentary on the grounds that they never go to the gym.

A useful reframing: the multiplier is not rating your effort. It is estimating how much energy your body moved through in twenty-four hours. Those are different questions, and the second one is far less flattering.

That is the boundary most people are actually stuck on, and the two levels are 0.175 apart on the multiplier, which is roughly 290 kcal a day on a resting rate of 1,639. Six real weeks, and where each one honestly lands:

Notice rows two and four. A desk worker who trains five times a week and a nurse who never trains at all land on the same multiplier, because the nurse spends eight hours upright and the desk worker spends eight hours seated. The gym is one hour; the job is a third of the day.

If your week sits between two rows, take the lower one. The scale will correct you within three weeks, and being corrected downward is a much easier problem than eating at maintenance while believing you are in a deficit.

Most people now have a number from a phone or a watch, which is a far better input than a guess. The awkward part is that the two scales were never designed to line up: the multipliers describe exercise sessions per week, while a step count describes movement across a whole day.

There is a published classification for step counts. Tudor-Locke and Bassett proposed these bands for healthy adults in 2004, and they are still the ones most commonly used:

The left two columns are from the paper. The right column is not. It is an approximation, and it is worth being explicit that no study maps step counts onto Mifflin-St Jeor multipliers, because the two were developed separately and measure different things.

Two things follow from that. Steps miss activity that does not involve walking, so cycling, swimming and a heavy lifting session all register as almost nothing. And the table assumes steps are your only activity: if you walk 12,000 steps and also train four times a week, you are above the last row, not on it.

Used sensibly, though, a step count is still the most honest input most people have, precisely because it was recorded rather than recalled.

Three rules that between them handle most cases:

1. Count the whole week, not the best week. Pick the level that matches a normal week including the flat ones, not the week you had when you were training for something.
2. Count your job, not just your workouts. A desk job and a physical job are a whole level apart before any exercise is added.
3. When two levels feel equally true, take the lower one. An underestimate produces a slightly steeper deficit than you planned, which shows up on the scale and is easy to correct. An overestimate produces a target above maintenance, and looks like nothing happening at all for a month.

The scale settles this. Eat near the number your chosen level produced, hold it steady for two to three weeks, and look at the trend rather than any single morning.

If your weight is stable when you expected to lose, your real expenditure is lower than the estimate, so drop a level and recalculate. If you are losing faster than intended, you were too conservative. This is the only way to find your actual maintenance level, and it beats any equation, because it is measured on you.

It is also worth revisiting after any real change in routine: a new job, an injury, a season off. The multiplier you chose in January is not necessarily still true in July.