Weight swings aren’t a factor of just appetite and exercise, or even metabolism. Relapsing obesity after dieting, otherwise known as unsuccessful dieting, turns out to rely not only on our intake of ice cream but to the type of bacteria we harbor in our intestines.
Not only that: scientists say they can predict how much weight a mouse will gain after dieting, based on the germ population in its little guts. (Or rather, a machine-learning algorithm can.) And one day the machine-learning algorithm may be able to do the same for the roughly 90% of dieters who, after a year, find themselves back at their original weight or worse.
The discovery was made through a multidisciplinary collaboration by scientists working in immunology and computer science, of all things, at the Weizmann Institute of Science.
Certainly, obesity is an issue. The Institute for Health Metrics and Evaluation at the University of Washington reported in 2014 that almost a third of the world’s population is overweight or obese, and that 2.8 million adults die each year from obesity-related problems.
Another study published last year concluded that the probability of an obese person (BMI 30-35) being normal weight after a year of dieting (not including through surgical intervention) is 1 in 210 for man, and 1 in 124 for women.
At BMI levels of 40-45, the probability is even lower: 1 in 1,290 for men and 1 in 677 for women.
The blame game
It has long been a frustrating mystery why some people can gorge on croissants and remain slim, while others glance at a potato and gain weight. Many people assume overweight people are simply eating in secret.
But last year a correlation was discovered by the Weizmann team between one’s personal physiological response to foods, and one’s personal collection of intestinal bacteria. It really is true that one man’s meat is another man’s misery. Now the Weizmann people have taken their groundbreaking study to the next level and correlated our pet bacteria with another problem: weight gain after dieting.
True, the study is in mice, not men, Eran Elinav and colleagues report in Nature this week, but it should hold water in humans as well.
What exactly leads to the so-called yo-yo effect among dieters – losing weight, gaining back, losing, gaining – has never been tied down. It also probably differs widely between people. People may assume that the hapless dieter simply lacks self-control. But one element that clearly matters is the composition of the microbiome, which means “the bacteria that live inside us”.
“Nobody looked at the possible role of intestinal bacteria in relapsing obesity, though the dynamic of swinging weight characterizes about 80% of overweight people in the world,” Elinav told Haaretz. “When we began to read the literature, we understood that [science is] just scratching the surface, and that we know nothing about weight gain.”
It’s typical in a state of ignorance to blame everything on the patient, Elinav adds, “But the psychological explanations could not explain the numbers. When you look at the vast body of literature on dieting, it’s pretty surprising to see that almost all diets work in the short run, and people do lose weight, but in the longer run, 12 months or more, 95% of them regain their original weight, or more.”
It bears saying that we each have our own microbiome, meaning, the precise composition of our bacterial populations will differ from person to person. And from mouse to mouse.
Elinav, Dr Eran Segal and the team took mice and made the rodents’ weight swing by feeding them high fat, normal mouse food, high fat, normal mouse food, etc. Actually they created three different models of relapsing obesity (following successful weight loss) and discovered a common factor. After a successful diet, all physiological and metabolic parameters in the mice returned to normal – except for the gut bacteria, Elinav says.
The team then identified changes to the bacterial population in the murine gut that persisted after successful weight loss, that also correlated to accelerated weight regain when the mice were fed the obesity-promoting diet.
To further test their theory, the scientists did poo transplants to mice that had not been subjected to the weight-cycling diet. These mice also gained weight faster.
The bottom line: Helped by computer science, the Weizmann team created a machine-learning algorithm that factors in the gut bacteria and can accurately predict how much weight an animal will regain after dieting, they claim.
The authors also find that the altered microbiome contributed to reduced energy expenditure. And meanwhile, their conclusion is that in the future, one day, analyzing out personally harbored bacteria could help of lose weight and keep it off without forgoing the potato forever more.
Meanwhile, the Weizmann team suggests that fecal transplants from bacteria with no obesity-related “memory” might be useful. Or flavinoids that affect the bacterial population usefully could be added to the diet. They hope to start human clinical trials of their methods soon.
Earlier this year, a different team discovered that a high-fat diet during pregnancy can have a lasting impact on the bacteria living in her baby’s gut. In that case, researchers at Baylor College of Medicine in the U.S. found an association between the mothers’ diets and distinct changes in their child’s microbiome – changes that lasted some six weeks. The bottom line, they reported in Genome Medicine, is that a high-fat diet hurts bacteria that help baby digest certain carbohydrates. That could thus have significant consequences for the baby’s ability to extract energy from food and for its immune development.