Americans obsess over weight loss. In 2013, Americans spent over $60 billion trying to lose weight. Nearly every gossip magazine suggests a new method to slim down or tone muscle. Among all the sensational headlines, it may seem difficult to find scientific information on what really determines our weight. A growing body of research, however, may provide some answers as to why it is so difficult to maintain weight loss.

One popular mantra in the world of nutrition is that a “calorie is a calorie.” This implies that weight gain or loss is simple. If you eat more calories than you burn, you will gain weight. If you eat less than you burn, you will lose weight.

This “black box” version of nutrition has led to several rules of thumb regarding weight loss. For instance, it is frequently claimed that 3500 excess calories are required in order to gain one pound. As an energy store, one pound of fat translates into 4086 calories.  But as Marion Nestle, professor of nutrition at NYU, stresses, “Body fat is not pure fat.  It is mixed with proteins and water, thereby diluting the theoretical number [of calories].” The 3500 number is an approximation based on this dilution. By this logic, dieters know roughly how many extra calories they need to burn to lose a certain amount of weight.

The 3500 calorie rule, however, oversimplifies. Dieters often lose lean mass in addition to fat. Since calorie densities vary for different types of tissues, the 3500 approximation can be dramatically off for some individuals. Additionally, changes in body weight, body composition, and caloric intake will actually alter the “calories out” portion of the equation over time.

Recent research in nutrition has focused on how the body adapts to a change in diet and exercise over the long-term. Dr. Kevin D. Hall and his colleagues at the National Institute of Diabetes and Digestive and Kidney Diseases have studied the effects of calorie changes on individuals with varying body compositions. In an interview with the New York Times, Hall said that the longstanding assumption that cutting 3,500 calories will indefinitely produce a one-pound weight loss is inaccurate. It can discourage dieters and lead to misinformed policy like the tax on sugar-sweetened beverages proposed by Mayor Bloomberg of New York.

Dr. Hall’s research group has developed an alternative model to simulate body weight changes. In Hall’s model, body weight tends towards an equilibrium where “calories in” equals “calories out” over time.

“Calories in” is largely a result of calories consumed, with some variance due to differing calorie absorption between individuals.

“Calories out,” however, varies with weight and other factors. Lighter individuals tend to burn fewer calories while at rest and during physical activity. In addition, as individuals consume less, the body tends to expend less energy in order to digest food. The result is that a typical dieter will see an initial weight loss, but the rate of weight loss will decrease over time.

In practice, this may partly explain why it is so difficult to lose weight and keep it off in the long-term. Many weight loss programs rely on a large, up-front reduction in “calories in.”  According to Dr. Hall’s model, this will result in a large initial weight loss, but as weight loss begins to plateau, individuals may lose faith in their weight loss program or unconsciously begin to eat more.

A better approach is to start with a more modest calorie reduction. By this model, cutting out 250 calories a day — the amount in a small bar of chocolate or half a cup of premium ice cream — would lead to a weight loss of about 25 pounds over three years, with half that loss occurring the first year. With a more gradual calorie reduction, changes in eating habits and lifestyle are more likely to take hold for the long-term.

The weight loss simulator at the top of this article is based on the model developed by Dr. Hall. It estimates initial body fat composition based on age, height, and weight, assuming a stable initial weight. It then models future weight gain or loss based on a fixed increase or decrease in caloric intake, assuming a constant level of physical activity. In most scenarios, weight gain or loss is initially rapid, but decreases with time while approaching a new steady state. For more information on the model, check the NIH website.

This post was written by contributor Alex Medearis. Follow him on Twitter here. .