For many humans affected by obesity, the challenge of initial weight loss is less than the challenge of maintaining the weight loss over the long term. Studies in large populations have found that, while long-term maintenance of weight loss is possible, its prevalence in real life remains below satisfactory levels. Individuals may experience recurrent weight gain and re-emergence of metabolic complications, a phenomenon sometimes called the “yo-yo effect”. But growing evidence of the role of gut microbiota in obesity and metabolic health may bring the potential to address this problem in new ways.

Recently, researchers from the Weizmann Institute of Science in Rehovot (Israel), led by Eran Segal and Eran Elinav, undertook a detailed exploration of mechanisms underlying post-dieting weight regain in mice. In the initial investigation, researchers exposed mice to two diets—a high-fat diet (HFD) and a normal chow diet (NCD)—in cycles. The experimental mice were exposed to a cycle of HFD followed by NCD, with a further cycle of HFD. Control groups were either continuously fed a HFD, continuously fed a NCD, or fed a single cycle of HFD followed by NCD.

The mice in the experimental group responded according to the composition of the diet in each phase: weight gain and metabolic syndrome developed during exposure to the HFD, and a return to baseline weight occurred when mice consumed a NCD. After the return to baseline weight on the NCD, the body fat, serum cholesterol, glucose tolerance, and serum insulin levels of these mice were not significantly different from control (non-cycling) mice. Upon re-exposure to the HFD, the mice once again experienced the weight gain and metabolic disturbances, but with this second HFD exposure they experienced a greater increase in total body fat, enhanced glucose tolerance, and elevated serum levels of both leptin and low-density lipoprotein (LDL).

Authors noted the same pattern of greater weight gain with secondary HFD exposure even when weight loss was enhanced by the pharmacological intervention celastrol. It also occurred when obesity was induced not by a HFD but by hyperphagia (when leptin signaling was pharmacologically inhibited in mice on a NCD, causing them to overeat and become obese). These mice returned to a normal weight when the intervention was ceased, but reintroduction of the leptin antagonist induced more pronounced weight gain compared to mice who had received the intervention for the first time.

Studying the gut microbiota of mice in these groups revealed, as expected, alterations induced by primary obesity; yet the mice that had successfully returned to a normal weight and metabolic homeostasis while consuming a NCD exhibited gut microbiota compositions that did not return to baseline. The post-dieting intestinal microbiome signature appeared to be an intermediate configuration between the obesity-associated state and the baseline state. Functionally, the researchers saw a similar pattern: the results of shotgun metagenomic sequencing showed the abundance of 773 bacterial genes was altered by the HFD and did not return to control levels after weight reduction. That is, reversal of obesity led to only a partial recovery of microbial functions.

When the microbiota was transferred to germ-free mice on a HFD, the recipients of post-dieting microbiota gained weight faster than those receiving microbiota from mice of identical weight who had been maintained on a NCD, showing that the post-dieting microbiome signature played a causal role in accelerated weight regain. If the germ-free recipient mice were fed a NCD, however, their weight and glucose tolerance changed little. This showed the post-dieting gut microbiome was not itself obesogenic, but that its interaction with diet (HFD vs. NCD) was crucial for its impact on phenotype.

In the next group of experiments, researchers developed a machine-learning algorithm that enabled them to predict the extent of weight regain after obesity and weight loss, using each mouse’s individual microbiome data. A 2-step algorithm, which first predicted obesity history and then predicted weight regain based on predicted history, accurately forecasted the extent of weight gain in individual mice.

Researchers then addressed the question of how to ameliorate the greater weight regain after obesity and weight loss. One strategy involved fecal microbiota transplantation: after obesity and dieting, the mice receiving gut microbiota from healthy mice on a NCD showed reduced secondary weight gain, reduced glucose tolerance, decreased body fat, and increased lean mass compared to those receiving the dysbiotic post-dieting gut microbiota. The second strategy was a ‘post-biotic’ intervention: the researchers found gut dysbiosis after obesity and weight loss contributed to a reduction in levels of two intestinal flavonoids (apigenin and naringenin) in the mice, which appeared to impact host energy expenditure. Administering these flavonoids daily to the post-dieting mice ameliorated the larger secondary weight gain and normalized host energy expenditure, with no effects on microbiome composition.

The main idea of this paper is that gut microbiota dysbiosis induced by obesity is not immediately reversed after a return to baseline weight—a caution against the assumption that gut microbiota will always ‘match’ phenotype. The gut microbiota dysbiosis appeared to accelerate weight gain under obesity-promoting conditions. Authors say the work also shows a personalized approach may help diagnose and treat exaggerated post-dieting weight regain; in the mice, either fecal microbiota transplantation or administration of selected flavonoids successfully treated mice whose secondary obesity was induced by an altered gut microbiota in combination with a high-fat diet.

Yet to be determined is the extent to which these findings apply to humans: most individuals affected by obesity already know the idea of “dieting” (a temporary change in eating patterns, often involving food restrictions) is not sufficient for long-term weight loss and are counselled instead by medical professionals to adopt long-term strategies for dietary change and regular exercise.




Thaiss CA, Itav S, Rothschild D, et al. Persistent microbiome alterations modulate the rate of post-dieting weight regain. Nature. 2016; 540:544-551.