Some people who are working on losing weight like the approach of a calorie prescription, which is the daily amount of calories in food/drink that they should not exceed in order to lose weight.

This calorie prescription starts with an estimation of Resting Energy Expenditure (REE), which equals the number of calories we burn at rest over 24 hours, and can be estimated by any of a number of equations that have been developed for this purpose.  These equations incorporate various factors that influence REE, including age, gender, height, and weight.  From there, we typically multiply the REE by an activity factor to calculate the number of calories a person needs in a day, and then usually subtract 500 calories per day in order to achieve an initial rate of weight loss of around 1lb (0.5kg) per week. 
Most of these equations were generated using normal weight individuals, including very few people who carry excess weight.  But are these equations accurate in people with obesity?
A study, published in the International Journal of Obesity, evaluated a number of these formulae in 1,851 people with obesity, comparing the calculations to actual measures of Resting Energy Expenditure (using a technique called indirect calorimetry). 
They found that the accuracy of the equations to predict Resting Energy Expenditure was very low in people with obesity, and were even less accurate in people with a higher degree of obesity (BMI >40), especially females.   Even the Mifflin St-Jeor equation, commonly cited as the most accurate equation, performed poorly. These equations generally underestimated the calorie needs of participants by several hundred calories, with the degree of underestimation increasing with increasing BMI.
As these equations come in low, the calorie prescription ends up being too low.  This could mean that a person with obesity leaves their health care provider’s office with a calorie prescription that is too
restricted – for example, that patient may be told that she should take in 1500 kcal per day in order to lose 1 lb per week, when actually her prescription should be 1900 kcal per day to lose 1lb per week.  For her, sticking to 1500 kcal per day would be very difficult – it may cause more rapid weight loss at the beginning but would be very tough to stick with.  
So why would these equations be less accurate in people with obesity? Fat tissue is less metabolically active than lean tissue (eg muscle), so having a higher proportion of fat can reduce accuracy of estimation using equations that were developed in a lean population.  It is also not clear which weight to use in these equations – actual weight, ideal weight, or adjusted weight.
Indirect calorimetry is a much better way to estimate calorie needs than equations, but has a price tag and limited availability. Clearly, we are in need of equations that are validated in people with obesity to estimate
resting energy needs. 

Follow me on twitter! @drsuepedersen © 2017