Estimate your maintenance calories after stopping GLP-1 medication with the Mifflin-St Jeor equation: resting energy from weight, height, age and sex, multiplied by an activity factor (Mifflin 1990; FAO/WHO/UNU 2001). It is a starting point, not a target: in two studies of Chinese adults by Singapore researchers it came within 10% of the measured value for 67% and 75% of people (Camps 2016; Song 2014).

Maintenance calories are an estimate of the energy a body uses in a day, and eating about that much keeps weight steady. The estimate has three parts: resting energy (BMR, the energy that keeps basic functions going), the energy of moving through the day, and the energy of digesting food (Westerterp 2004).

After weight loss the figure is lower: a smaller body uses less energy, and in one study people holding a lower weight used less than their size alone predicted (metabolic adaptation) (Leibel 1995). To check an estimate, follow the weight trend over several weeks, because daily weight moves in a weekly pattern (Orsama 2014). Appetite matters too: as the medication clears over weeks (semaglutide label; tirzepatide label), its appetite effect is expected to fade, so hunger and habits also decide how much is eaten. At GetLean the programme is designed to end: habits are built with daily WhatsApp coaching while the medication is working, and Dr Quek steps the dose down at monthly reviews towards the lowest effective dose, rather than stopping overnight. We call it the GetLean GLP-1 Taper Plan. Individual results vary; any change to a prescribed medicine is for the doctor who prescribed it.

Maintenance calories: the amount of food energy that matches what the body uses in a day, so that weight stays steady. It is an estimate of total daily energy expenditure (TDEE), not a target.

BMR (basal metabolic rate): the energy that keeps basic functions going, like breathing and circulation, measured at rest. Fat-free mass was the best single predictor of it in one study (Johnstone 2005).

What are maintenance calories?

Maintenance calories are the food energy that matches what the body spends in a day, so that weight neither rises nor falls. Calories here means kilocalories (kcal), the unit used for food energy.

Daily energy expenditure has three components: basal metabolic rate, diet-induced thermogenesis (the energy of digesting, absorbing and processing food) and the energy cost of physical activity (Westerterp 2004). On a mixed diet eaten at energy balance, digestion accounts for 5 to 15% of the day's energy expenditure (Westerterp 2004).

Searches for "TDEE" and "maintenance calories" are the same question asked from two sides: TDEE is what the body spends, and maintenance calories are the intake that matches it.

How are maintenance calories estimated?

With a two-step estimate: an equation for resting energy, then a multiplier for activity.

Step one, resting energy. The Mifflin-St Jeor equation was derived from 498 healthy adults aged 19 to 78, 234 of them with obesity, whose resting energy expenditure was measured by indirect calorimetry (Mifflin 1990). In its simplified form, resting energy in kcal a day is 10 × weight (kg) + 6.25 × height (cm) − 5 × age (years), then + 5 for men or − 161 for women (Mifflin 1990). It is one of four equations most used in clinical practice, and in a systematic review it put more people within 10% of their measured value than any of the others tested (Frankenfield 2005).

Step two, activity. A joint FAO, WHO and United Nations University expert consultation expresses the 24-hour energy requirement as a multiple of basal metabolic rate and groups habitual activity into three bands (FAO/WHO/UNU 2001).

Habitual activity band Multiple of basal metabolic rate
Sedentary or light activity lifestyle 1.40 to 1.69
Active or moderately active lifestyle 1.70 to 1.99
Vigorous or vigorously active lifestyle 2.00 to 2.40

Source: FAO/WHO/UNU 2001, Table 5.3. Bands, not targets.

The bands are wide. Within the first one, the top of the range (1.69) is about a fifth higher than the bottom (1.40), so the choice of band alone moves the daily figure. The equation estimates resting energy expenditure and the bands are defined against basal metabolic rate, so putting the two together is an approximation on top of an approximation.

How accurate is the estimate?

Close enough to be a starting point, and not close enough to be a target.

In a systematic review, Mifflin-St Jeor was the most reliable of the four equations most used in practice, with the narrowest error range, and the authors still note that noteworthy errors exist when it is applied to individuals (Frankenfield 2005). Two Singapore studies checked it against measured energy at rest in Chinese adults, basal metabolic rate in one and resting energy expenditure in the other:

  • In 232 Chinese adults (BMI 16 to 41), it was within 10% of measured basal metabolic rate in 67% of people, and a Singapore-derived equation reached 78% (Camps 2016).
  • In 96 healthy Singaporean Chinese men aged 21 to 40, it overestimated the group's mean resting energy by 2.4%, with 75% of predictions within 10%, and individual errors across all the equations tested were high (Song 2014).

Both studies were in Chinese adults, so they describe how the equation performed in that group. Resting energy can be measured directly with indirect calorimetry, which is the method used to build and check these equations (Mifflin 1990; Camps 2016; Song 2014).

Why is the number lower after weight loss?

Because a smaller body uses less energy, and because energy use can fall by more than body size alone explains.

The smaller body. In the Mifflin-St Jeor equation, each kilogram of body weight adds 10 kcal a day, so with height and age unchanged a body 10 kg lighter has a resting estimate 100 kcal a day lower (Mifflin 1990). Fat-free mass, which includes muscle, explained 63% of the variance in basal metabolic rate between 150 adults in one study, against 6% for fat mass and 2% for age (Johnstone 2005).

A further fall. In a study of people held at a weight at least 10% below their usual weight, total energy expenditure was lower than the change in body composition alone predicted, in people who had never been obese as well as in people who had (Leibel 1995). This is metabolic adaptation, and its size varies between people. An equation that uses only weight, height, age and sex has no input for it. What was measured, in whom, and what can be shaped is in metabolic adaptation explained.

Why does appetite matter alongside the arithmetic?

An equation estimates what the body uses. It cannot estimate how hungry someone will be.

Semaglutide has an elimination half-life of about one week and is present in the circulation for roughly 5 to 7 weeks after a last 2.4 mg dose (DailyMed); tirzepatide's is about 5 to 6 days (DailyMed). As the medication clears, the appetite effect it was producing is expected to fade, so hunger, habit and the food around also decide how much is eaten. Appetite after stopping GLP-1 sets out what is known.

In the STEP 1 extension, where the medication stopped overnight with no taper and the study's lifestyle support ended the same day, about two-thirds of the lost weight returned within a year (Wilding 2022). That is a group average, and individual results vary widely.

At GetLean, the programme is designed to end: our philosophy is that GLP-1 medication should act as a catalyst — not something to depend on indefinitely. The habits are built while the medication is still quieting appetite, with daily WhatsApp coaching. At the monthly reviews, Dr Quek looks at how each patient is doing, and the dose comes down gradually towards the lowest effective dose rather than stopping overnight. Any dose change is a decision for the doctor who prescribed it. The full picture is in what carries a loss after the medication stops and how a planned taper works, and the first answer in the series is in the coming-off guide.

How do you check the estimate in real life?

Follow the weight trend over several weeks, not a single day.

Weight moves in a weekly pattern. In a study of 80 adults and 4,657 daily weights, weight was higher on Sunday and Monday and decreased during the week, and the authors concluded that weekend-to-weekday variation is normal rather than a sign of weight gain (Orsama 2014). Comparing the same day each week, or a weekly average, avoids reading that pattern as a change.

Weight also responds slowly to a change in intake: in a mathematical model of adult metabolism, the half time of the bodyweight response was about a year (Hall 2011). So the trend over weeks shows direction, and it is the direction that tests the estimate.

  • A level trend means intake is matching expenditure for now, and the estimate is working as a maintenance figure.
  • A trend that keeps rising or falling means intake and expenditure are not matched, so the estimate was off.

What to change, and by how much, is a question for a doctor or dietitian at a review, not for guesswork. Regular weighing itself is associated with better maintenance: in a registry cohort, people who cut back on how often they weighed themselves gained 4.0 kg over a year, against 1.1 kg for those who began weighing more often, an observational finding (Butryn 2007). During weight loss the scale misleads about composition (why the scale misleads); in maintenance it does a different job, as an early signal of drift.

Where do protein and strength fit in?

They are the muscle side of the same number.

Fat-free mass was the best single predictor of resting energy in the study that built the equation (Mifflin 1990) and explained 63% of the between-person variance in another (Johnstone 2005). Lean mass was statistically unchanged in the trials where resistance training accompanied caloric restriction (Lopez 2022). That result is from the trials that combined resistance training with calorie restriction, so it describes the weight-loss phase. What protein intake suits a steady weight is a question for the doctor or dietitian.

The protein and strength pages cover the practical side:

How much energy is in a Singapore meal?

It depends on the dish and, above all, on the portion.

A laboratory analysis of 45 commonly consumed Singapore meals measured portion weight and energy density (Yeo 2021). A 412 g portion of steamed chicken rice measured 789 kJ per 100 g, which works out to about 780 kcal for the portion; a 285 g portion of roasted chicken rice measured 768 kJ per 100 g, about 520 kcal for the portion (Yeo 2021; kcal derived by dividing kJ by 4.184 and multiplying by portion weight). Per 100 g the two are almost the same, about 189 and 184 kcal, so most of the gap is the portion. These are measured examples from one study, not averages for every stall. The wider table is in hawker dishes ranked by protein per calorie.

Individual results vary, and clinical-trial figures describe the populations studied. Never stop or change a prescribed medicine on your own.

Common questions

How many calories should I eat after stopping GLP-1 medication?

There is no single number that is right for everyone. Maintenance calories are an estimate of resting energy, movement and digestion, worked out with a published equation such as Mifflin-St Jeor (Mifflin 1990) and an activity multiplier, then checked against your weight trend over several weeks. Any change to a prescribed medicine is a decision for the doctor who prescribed it.

What is the difference between BMR and maintenance calories?

BMR is the energy the body uses at rest. Maintenance calories also include the energy of moving through the day and of digesting food, which together are the three parts of daily energy expenditure (Westerterp 2004).

Why are my maintenance calories lower after losing weight?

A smaller body uses less energy: in the Mifflin-St Jeor equation each kilogram of body weight adds 10 kcal a day (Mifflin 1990). Energy expenditure can also fall by more than the change in body size alone predicts (Leibel 1995), which is called metabolic adaptation.

How accurate is a TDEE or BMR calculator?

It is an estimate. In two studies of Chinese adults by Singapore researchers, the Mifflin-St Jeor equation came within 10% of the measured value (basal metabolic rate in one study, resting energy expenditure in the other) for 67% and 75% of people (Camps 2016; Song 2014). The activity multiplier adds its own uncertainty.

How do I check whether my maintenance calories estimate is right?

Follow your weight trend over several weeks rather than a single day, because weight moves in a weekly pattern, higher on Sunday and Monday and lower later in the week (Orsama 2014). A level trend means intake is matching expenditure for now; a trend that keeps rising or falling means the estimate was off.