Science behind why dieters regain weight

Calorie deprivation alters body and mind, overwhelming willpower.  It is well established that dieters are able to lose weight in the short run, but tend to gain it back over time. In 2007, the graduate students in my Psychology of Eating seminar and I did a painstaking review of every randomized controlled trial of diets we could find that included a follow-up of at least two years (Mann et al., 2007). Janet Tomiyama, Britt Ahlstrom, and I updated it in 2013 with studies we had missed, as well as newer ones (Tomiyama, Ahlstrom, & Mann, 2013). The results were clear. Although dieters in the studies had lost weight in the first nine to 12 months, over the next two to five years, they had gained back all but an average of 2.1 of those pounds. Participants in the non-dieting waitlist control groups gained weight during those same years, but an average of just 1.2 pounds. The dieters had little benefit to show for their efforts, and the non-dieters did not seem harmed by their lack of effort. In sum, it appears that weight regain is the typical long-term response to dieting, rather than the exception.

The real question is why do dieters regain the weight they lose? They are motivated to maintain their lost weight, and they have the skills to enact the necessary eating behaviors, as they did so when they lost the weight. When dieters regain weight, is it because they are too weak-willed to continue restricting their food intake, or is it because their body physiologically defends a genetically-based set weight? The answer is somewhere in between those explanations and is easily misunderstood: In brief, calorie deprivation leads to changes in hormones, metabolism, and cognitive/attentional functions that make it difficult to enact the behaviors needed to keep weight off.

What does calorie deprivation do?
In response to calorie deprivation, levels of leptin, considered the satiety hormone, decrease, and levels of ghrelin, thought of as the hunger hormone, increase (reviewed in Greenway, 2015; Maclean, Bergouignan, Cornier, & Jackman, 2011). Therefore individuals in a deprived state experience more hunger, and feelings of hunger remain increased for deprived individuals even after eating a regular meal (Doucet, St-Pierre, Alméras, & Tremblay, 2003).

The most fundamental physiological adaptation to calorie deprivation is a reduction in energy expenditure (that is, in the amount of calories burned; reviewed in Greenway, 2015; Maclean et al., 2011). This reduction is due to two factors. First, after sufficient calorie deprivation, weight is lost, and therefore less energy is needed to “run” the smaller body, and less energy is expended to move the smaller body during exercise (Leibel, Rosenbaum, & Hirsch, 1995). Second, and beyond the changes resulting from the smaller body, metabolism becomes more efficient, allowing the body to survive on less energy than similar sized bodies that were not calorie deprived (e.g., Camps, Verhoef, & Westerterp, 2013; DeLany, Kelley, Hames, Jakicic, & Goodpaster, 2014). Therefore, to continue losing weight, the individual must consume even fewer calories than during the initial stages of the diet. Dieters who are unaware of this metabolic adaptation and don’t alter their diet accordingly will stop losing weight, may start to regain, and will likely be puzzled and discouraged. From their perspective they are still doing everything right – everything that already led to weight loss – and it is no longer working. Their friends and family members may erroneously conclude that they have been cheating on their diet.

In theory, exercise should prevent or overcome these effects of calorie deprivation, and studies with rodent models do find that adding physical activity to a diet improves weight loss maintenance (e.g., MacLean et al., 2009). Human studies generally find that exercise alone leads to less weight loss than diets alone, and that adding exercise to diets leads to slightly more weight loss initially, but does no better in preventing weight regain over time (reviewed in Washburn et al., 2014). This is likely due to low levels of adherence to the assigned physical activity over the long term (MacLean et al., 2015), as when these studies ignore the study conditions participants were randomly assigned to, they find that the amount of actual exercise participants maintain does correlate with weight loss maintenance (e.g., DeLany et al., 2014; Jakicic, Marcus, Lang, & Janney, 2008).

Calorie deprivation also leads to changes in a variety of cognitive and attentional functions such that dieters become preferentially focused on food. A preoccupation with thoughts of food was one of the most pronounced responses to calorie deprivation in a classic study of conscientious objectors to WWII who volunteered to be semi-starved for six months (Keys, Brozek, Henschel, Mickelsen, & Taylor, 1950). The volunteers spent much of their time talking about foods, planning future meals, reading cookbooks and even considering new careers in food-related fields.

Studies that assess attentional focus and attentional capture with eye tracking methods (Castellanos et al., 2009), the attentional blink paradigm (Piech, Pastorino, & Zald, 2010), or dot probe tasks (Placanica, Faunce, & Soames Job, 2002) find that people’s attention is biased toward food stimuli when they are calorie deprived. Brain imaging studies find increased activity in areas relevant for attention when calorie deprived individuals are shown images of palatable foods (compared to images of water or non-palatable foods; Stice, Burger, & Yokum, 2013).

It’s not the case that preferential attention to food leads to negative reactions to food. On the contrary, brain imaging studies also find increased activity in reward-relevant areas (Stice et al., 2013). Other studies find that when calorie deprived, people have improved smell functioning (Cameron, Goldfield, & Doucet, 2012), report that food tastes more palatable and are willing to work harder to earn it (Cameron, Goldfield, Finlayson, Blundell, & Doucet, 2014).

Research shows that the effects of calorie deprivation last a long time – at least through the last measurement points studies have included. For example, the conscientious objectors who agreed to go on starvation diets still had overly efficient metabolic rates and preoccupations with food thoughts a year after their starvation period ended (Keys et al., 1950). Contestants on the television show “The Biggest Loser,” who lost an average of 128 pounds in 30 weeks, still maintained the metabolic effects of calorie deprivation even six years later (Fothergill et al., 2016). Are these lengthy consequences due to the extreme nature of these diets, and if not, is everyone who has dieted in at least the last six years already suffering the effects of calorie deprivation?

What is the role of willpower?
Self-control (which I use synonymously with willpower here) is “the capacity to alter or override dominant response tendencies and to regulate behavior, thoughts, and emotions” (de Ridder, Lensvelt-Mulders, Finkenauer, Stok, & Baumeister, 2012, p. 77). It tends to be a statistically significant predictor of body mass index (BMI) in longitudinal studies that measure self-control in children and then measure weight from three (Duckworth, Tsukayama, & Geier, 2010) to 30 years later (Schlam, Wilson, Shoda, Mischel, & Ayduk, 2013).

It is clear that self-control plays some role in weight, but it is also important to appreciate that this role is quite small, explaining from 1 to 4 percent of the variance in BMI in these studies. It’s not the case that self-control is simply an unreliable or poor measure in general. The same measures of self-control in these studies do powerfully predict large proportions of variance in other outcomes, including, for example, 32 percent of the variance in scores on the quantitative SAT, and 28 percent of the variance in scores on a measure of coping resiliency (Mischel, Shoda, & Rodriguez, 1989), as well as 45 percent of the variance in eighth grade GPA (Duckworth & Seligman, 2005). Similarly, a meta-analysis found that a commonly used self-report measure of self-control was more powerful in predicting outcomes in the domains of school achievement, well-being (which included self-esteem, happiness and depressed mood), and even interpersonal functioning (which included relationship commitment, loyalty tendencies and perceived support), than in eating and weight (de Ridder et al., 2012).

In sum, self-control matters for weight, but not as much as one might expect, and not as much as it matters for other types of activities. This limited relationship between self-control and BMI may be partly due to a unique feature of controlling eating (or other consumption behaviors), compared to controlling other behaviors (e.g., retaining focus on one’s schoolwork). With eating, failures of self-control erase prior successes. For example, foods in the immediate environment (say, cookies on one’s kitchen counter) may need to be resisted repeatedly, each time one notices them. If an individual resists them ten times during one evening, but succumbs and eats one on the eleventh encounter, there is nothing to show for the ten successes. The relationship between willpower and consumption may be lower than expected because people with weaker willpower (who presumably succumb on an earlier encounter) and people with stronger willpower (who presumably succumb on a later encounter) have similar outcomes: they ate the tempting food. Note that this is not the case with controlling one’s attention to schoolwork – a single failure of focus (such as a few minutes of goofing off) does not erase the effects of many previous minutes of successful focus – the work accomplished does not disappear due to the brief lapse.

Why is weight regain misunderstood?
Calorie deprivation leads to physiological, neural, and attentional changes, and those changes make it difficult to engage in the behaviors necessary to keep weight off. But since those changes do not directly cause weight to return, it is still possible to keep weight off, which a minority of dieters do. This possibility allows people to discount the powerful role of these changes, and instead to argue that if people regain the weight, it must be due to their poor self-control. And because the changes ultimately do operate through eating behavior, the weight regain does seem to be the fault of the apparently weak-willed dieter. As many people have said to me about failed dieters, “they’re still the ones holding the fork.” The key misunderstanding here is the different physical and cognitive context in which dieters hold the fork compared to non-dieters: they feel hungrier, their attention is biased toward food, they find food tastier, and they get more reinforcement from it. Plus they need to consume an even smaller quantity of food than earlier in the diet (as well as less than a non-dieter of the same size), because their more efficient metabolism is burning fewer calories. So dieters don’t necessarily have worse willpower than non-dieters, but calorie deprivation has put them in a situation that requires much more willpower in order to successfully limit consumption.

What should people do?
Because I do research in this area, people inevitably ask me for diet advice. I generally recommend avoiding restrictive dieting, since for the majority of dieters, the weight loss is unlikely to be maintained, and people can be healthy at most weights if they engage in healthy behaviors (Loef & Walach, 2012). Studies show that mortality rates for individuals who are physically fit do not differ by weight category (i.e., normal weight, overweight, or obese; Gaesser, Tucker, Jarrett, & Angadi, 2015), and that engaging in four healthy habits (eating five or more servings fruits/vegetables per day, limiting alcohol consumption, not smoking, and exercising over 12 times per month) minimizes weight-based differences in mortality (Matheson, King, & Everett, 2012).

Therefore, instead of dieting, I recommend engaging in the other healthy behaviors, and in recent years, my lab has focused on strategies to promote vegetable eating without worrying about reducing calories. Our most successful strategy aims to minimize the “competition” vegetables get from less healthy and more liked foods. Instead of having these foods side by side on a plate, we moved the vegetables to their own separate course, before the meal. If vegetables are eaten before other foods are visible or available, they may not be passed over for tastier options. This strategy was successful when used with adults who were watching videos in a lab setting (Redden et al., 2015), and it led to dramatic increases in vegetable consumption among elementary school children who were served vegetables before they entered the cafeteria for lunch (Elsbernd et al., 2016).

In addition to increasing vegetable consumption, there are certain unhealthy foods that should be minimized (although it is not necessary to forbid them entirely), including added sugars, refined grains, and highly processed foods. We have found that we can help people reduce their sugar consumption in coffee with a simple mindfulness intervention, and this intervention was more effective than the typically recommended strategy of gradually reducing sugar (Lenne & Mann, 2017). A new study supports this approach to healthy eating (increasing vegetable consumption and decreasing certain unhealthy foods without monitoring calories), but the long-term effects are not yet known (Gardner et al., 2018).