Using mathematics to predict who will lose weight

Background

Losing weight is perhaps the most popular ongoing health-promoting activity attempted by people living in industrialized nations. As most people are painfully aware, however, there is no one-size-fits-all solution to weight loss, and many people attempting to lose weight do not succeed. There are many approaches we can take to lose weight, such as decreasing calories, altering food intake patterns, increasing energy expenditure, and a combination of these factors. But people often need to try several tactics before finding one that works for them. Wouldn’t it be nice to have evidence that a certain diet and/or exercise program is likely to work early on so that another can be substituted if needed? Dr. Diana Thomas (Montclair State University) and colleagues recently set out to determine whether they could develop mathematical models that could predict the efficacy of a diet in its early stages. Results from this study are published in the March 2015 issue of The American Journal of Clinical Nutrition. Accompanying this manuscript is an editorial penned by Dr. Nicholas Finer from the ULC Institute of Cardiovascular Science (London).

Study Design

Data used for this study were collected as part of the Preventing Obesity Using Novel Dietary Strategies (POUNDS Lost) study in which 683 overweight adults were randomly assigned to one of four weight-loss diets varying in fat, protein, and carbohydrates. The researchers statistically evaluated and compared a series of equations to determine which one best predicted loss of at least 5% of body weight one year after the study was initiated.

Results

The analysis showed that including variables such as a person’s baseline weight, sex, age, target calorie intake, and weight-loss goal in addition to early weight loss success increased the predictive value of the equations.

Conclusions

The scientists concluded that these mathematical models could be used by healthcare providers to modify treatment strategies in a timely manner. Finer concurs but also asks whether identifying genetic markers related to successful weight loss might be more prognostic in the long run. Clearly, there is much left to learn in terms of establishing and maintaining a healthy body weight – including the personalization of weight-loss programs.

Reference Thomas DM, Ivanescu AE, Martin CK, Heymsfield SB, Marshall K, Bodrato VE, Williamson DA, Anton SD, Sacks FM, Ryan D, Bray GA. Predicting successful long-term weight loss from short-term weight-loss outcomes: new insights from a dynamic energy balance model (the POUNDS Lost study). American Journal of Clinical Nutrition. 2015;101:449-54.Finer N. Predicting therapeutic weight loss. American Journal of Clinical Nutrition. 2015;101:419-20.