Summary: Whether dieting makes people fatter has been a subject of considerable controversy over the past 30 years. More recent analysis of several prospective studies suggest, however, that it is dieting to lose weight in people who are in the healthy normal range of body weight, rather than in those who are overweight or obese, that most strongly and consistently predict future weight gain. This paper analyses the ongoing arguments in the debate about whether repeated dieting to lose weight in normal-weight people represents unsuccessful attempts to counter genetic and familial predispositions to obesity, a psychosocial reaction to the fear of fatness or that dieting per se confers risks for fatness and hence a contributing factor to the obesity epidemic. In addressing the biological plausibility that dieting predisposes the lean (rather than the overweight or obese) to regaining more body fat than what had been lost (i.e. fat overshooting), it integrates the results derived from the re-analysis of body composition data on fat mass and fat-free mass (FFM) losses and recoveries from human studies of experimental energy restriction and refeeding. These suggest that feedback signals from the depletion of both fat mass (i.e. adipostats) and FFM (i.e. proteinstats) contribute to weight regain through the modulation of energy intake and adaptive thermogenesis, and that a faster rate of fat recovery relative to FFM recovery (i.e. preferential catch-up fat) is a central outcome of body composition autoregulation in lean individuals. Such a temporal desynchronization in the restoration of the body's fat vs. FFM results in a state of hyperphagia that persists beyond complete recovery of fat mass and interestingly until FFM is fully recovered. However, as this completion of FFM recovery is also accompanied by fat deposition, excess fat accumulates. In other words, fat overshooting is a prerequisite to allow complete recovery of FFM. This confers biological plausibility for post-dieting fat overshooting – which through repeated dieting and weight cycling would increase the risks for trajectories from leanness to fatness. Given the increasing prevalence of dieting in normal-weight female and male among young adults, adolescents and even children who perceive themselves as too fat (due to media, family and societal pressures), together with the high prevalence of dieting for optimizing performance among athletes in weight-sensitive sports, the notion that dieting and weight cycling may be predisposing a substantial proportion of the population to weight gain and obesity deserves greater scientific scrutiny.
Alex’s Notes: It seems paradoxal that dieting would make someone fatter, and yet 15 prospective studies since the 1990s have suggested that dieting to lose weight predicts future weight gain and obesity in adults, adolescence, and even children. For instance, a study of high school females showed that those who were on a diet in 9th grade (baseline) had a 3-fold increased risk of becoming obese by the end of high school (12th grade). In a separate study of younger girls, those who dieted with parental encouragement before the age of 11 had greater increases in BMI from 9-15 years compared to non-dieters.
Of particular interest is a 6-15 year follow-up study of adults that showed the risk of major weight gain (>20 lbs) was consistently greater in initially normal weight men and women, but that results were less consistent in overweight persons of the same cohort. These findings suggest that leaner individuals may be more susceptible to future weight gain than overweight persons. Moreover, a study of over 4,000 twins shows that those with the lowest baseline BMI gain more weight than those with higher BMI and suggests that there is a dose-response relationship between the number of lifetime intentional weight cycling and future weight gain. Being twins, this strongly suggests that dieting itself may induce subsequent weight gain independently of genetic factors.
So does dieting, per se, cause weight gain?
The answer to this question, and the central concept of the review at hand, can be answered from classic food deprivation and refeeding studies. In these studies where various methods of starvation and weight loss were used in otherwise normal weight and healthy adults, there was a profound hyperphagic response that persisted for several weeks after body weight had been restored during the refeeding phase, leading to weight and fat gain greater than baseline.
So why would weight and fat gain continue to increase after reaching baseline levels? In the Minnesota starvation experiment, it was shown that when body fat had been completely recovered (i.e. 100% of baseline values), lean body mass (LBM) was still below baseline and hyperphagia did not cease until LBM was restored. This observation raises fundamental questions about the weight-recovery control systems, the answers of which are explained by revisiting the Minnesota experiment with a new mindset in the current review.
What is the relative importance of fat vs. LBM depletion during post-starvation hyperphagia?
Hunger and appetite are driven by the extent of body mass depletion, with fat mass being a stronger determinant than LBM. Nonetheless, LBM still plays a role and thus drives hyperphagia despite recovered fat mass stores owed to the disproportionate body mass regain (see below).
Why is fat-mass recovery faster than LBM recovery?
The starvation experiments demonstrated that thermogenesis, which was suppressed during weight loss, remained suppressed during weight recovery, leading to the idea of “fat-stores memory” referred to as adipose-specific control of adaptive thermogenesis. This is distinct from the non-specific thermogenesis that functions to attenuate energy imbalances created by the amount of calories consumed. The adipose-specific control functions to accelerate fat-store replenishment, thereby driving the disproportionate gain of body mass, while adaptive thermogenesis is the slow-down seen with dieting that operates under the influences of the leptin-sympathetic-thyroid neurohormonal axis.
What determines the large inter-individual differences during weight recovery and fat-overshooting?
A fundamental concept that all persons interested in dieting, bulking, and body weight manipulation is the “p-ratio” (partitioning-ratio). This is what determines what is lost when dieting and gained when refeeding, and the differences between persons can be explained mostly by the initial body fat percentage. The more fat you have at baseline, the more fat you will lose or gain from dieting and overfeeding, respectively.
I’ll come back to the p-ratio in a moment, but first I want to put together our three answers above to draw an image explaining their implications. The greater the severity of the fat and LBM loss, the greater the adipose-specific suppression of thermogenesis that preferentially accelerates fat mass recovery and explains why fat stores are replenished before LBM. However, since depleted LBM also drives hyperphagia, the consequences of this unequal repletion of body mass is that overeating and weight gain continue until LBM is also restored. Since LBM can only be gained in addition to fat mass, with the proportions of each determined by the individual’s p-ratio, fat gain also continues until LBM is back to normal.
As mentioned, the p-ratio dictates how much fat and LBM is lost during dieting. The lower the initial adiposity (body fat %), the greater the proportion of energy that is derived from body protein (muscle) during weight loss. Thus, during the refeeding phase there is more LBM that must be recovered and thus a greater overshoot of fat. This was indeed shown in the Minnesota starvation participants for whom initial body fat percentage was shown to be negatively associated with the amount of fat-overshoot. This also explains why lean persons are more susceptible to post-dieting weight gain, and why many prospective studies have not found consistent associations between future weight gain in overweight and obese persons, because they lose primarily fat with little LBM and thus yo-yo between their starting and post-dieting weights without fat overshooting.
The sensitivity of the p-ratio can be overlooked either. Amazingly, the steepest curve lies between 8-20% body fat, with a shift from the lower to the upper ends of this range reflecting a 2.5- to 3-fold change in the p-ratio.
The fatal flaw
So how then can natural bodybuilders drop from 14.8% body fat to 4.5% body fat in six months, and return to baseline values following the competition without issue? This is the central flaw with the p-ratio and fat overshooting hypothesis. The studies it was based on were done with, for lack of better phrasing, completely asinine dieting and exercise protocols. This isn’t to say the p-ratio is useless; quite the opposite, it provides hope that we can take advantage of it for body composition goals.
To restate for the third time (it’s important), the p-ratio determines the proportion of LBM and fat mass lost or gained depending on energy availability. This makes sense when dieting, as an obese person could probably just eat less and sit around all day without risking the loss of muscle. However, those at the lower ranges of body fat need to be more diligent with their dieting and exercise strategies in order to manipulate the p-ratio and preserve muscle mass as much as possible while continuing to promote fat gain.
In the bodybuilding case study linked above, energy intake was far above starvation levels, with daily exercise in the form of four days of resistance training (5 hours total per week), two days of high intensity interval training (HIIT, 40 minutes total per week), and one day of low-intensity, steady-state, aerobic exercise (30 minutes per week).
On the opposite side of the spectrum, consider those who want to bulk up and build muscle. A simultaneous goal would be to minimize fat gain, while appreciating that it is a necessary evil of building muscle. Again, the leaner you are, the more muscle you will build when overeating as dictated by the p-ratio. If you are fat to begin with, then you will predominantly store fat when overeating. Throw in some smartly programmed resistance training, and you have a recipe for success.
If you leave this article with anything, make it the knowledge of the p-ratio. The too-long-didn’t-read bottom line is that you lose more muscle when dieting if you are already lean, but this can be offset with proper exercise and dieting methods. For those looking to bulk, it really pays in the long run to diet down to lean levels before trying to gain weight. As long as you exercise to increase partitioning towards LBM and control your food intake so as not to significantly overeat, then there won’t be any risk of fat-overshoot.
My personal preference and advice is to diet down to 8-10% body fat before trying to bulk up as this level of leanness is achievable by anyone who is willing to put in a little extra effort. From there, bulk up to 12-15% body fat depending on how you feel, as anecdotally, 15% appears to be the threshold where insulin sensitivity takes a dive and fat gain accelerates.