Many genes contribute to obesity, so it's difficult to design a DNA test: Shots

Scientists who recently announced an experimental genetic test that can help predict obesity were immediately repelled by other researchers, who wonder if it's really helpful.

The story behind this back and forth is essentially knowing when it is worthwhile to immerse yourself in the DNA databanks when there is no obvious way to put this information into practice.

The basic facts are not disputed. Human behavior and our environment favoring obesity have led to a skyrocketing of this disease in recent decades. Today, about 40% of American adults are obese and even more overweight.

But genetics also plays an important role. People inherit genes that make them more or less likely to be overweight.

Although some diseases (such as Huntington's and Tay-Sachs's) are caused by a single gene that went wrong, this is certainly not the case with common diseases, including obesity. Instead, thousands of genes apparently play a role in increasing the risk of obesity.

Many of these gene variants pose minimal risk. Sekar Kathiresan, Harvard cardiologist and geneticist at the Broad Institute, asked if he and his team could find a set of these genetic variants and add the effects. The goal was to identify genetic models that expose people to the highest risk.

This genetic information "could explain why a person is so fat, why it's so hard to keep weight," said Kathiresan.

His team has identified more than 2 million potential DNA variants. He thinks most of these variants are irrelevant, but his intuition is hidden somewhere in the change. A few thousand changes each contribute at least to the risk of developing obesity.

No gene can do much to move the needle. But he says that the composite result, called the polygenic risk score, is still potentially useful. Individuals with the highest risk scores were more likely to be severely obese (with a body mass index greater than 40). In fact, 43% of people with the highest genetic scores were obese.

But the score is far from perfect. For example, 17% of people with the highest scores had a normal body weight.

The team, with lead author Amit Khera, published his findings in the journal Cell.

"The impact of genetics – and this has been a huge surprise for us too – begins very early in life, in preschool age, around the age of 3," said Kathiresan. .

This finding suggests that prevention efforts are more likely to succeed if they also start in childhood. Kathiresan also has a more philosophical idea of ​​her work.

"I hope this work will de-stigmatize obesity and make it very similar to all other diseases, which combine both lifestyle and genetics."

Numerous elaborate genetic analyzes are at the origin of this study, which involved more than 300,000 people. But the general conclusions are not new.

Scientists already knew that genetic risk factors could contribute significantly to obesity. And other studies show that obese children are at high risk of becoming obese adults.

Emory University epidemiologist Cecile Janssens, who is a professor at Emory University, does not like this strategy of adding the minimal risks of millions of genetic variants to a cumulative risk score.

"In all fairness, we do not know if all these variants really matter," she says. When asked if it was worthwhile to conduct a study like this, she replied, "I have no idea."

"It does not really answer a question that is very relevant from a biological point of view, nor really to a question that is very relevant from a clinical point of view," she says.

This type of analysis does not reveal anything about the individual genes that contribute to obesity, which means you can not use it to understand the underlying biology. If obesity was a rare disease, a test of this type could be helpful in identifying high-risk individuals. However, Janssens estimates that, since 40% of Americans suffer, prevention efforts should concern everyone.

She is part of a group of scientists who are informally rebelling against the gene-centric way of looking at the disease. It is frustrating for them to see so much money invested in this type of genetic work, rather than in the efforts to change the environment and behaviors that contribute to diseases such as cancer. ;obesity.

Janssens also states that, despite the tremendous efforts made to study 2 million genetic variants, the score obtained still does not account for even 10% of the variation observed by scientists in terms of body mass index. (Kathiresan, who formulates her findings differently, says the score accounts for about a quarter of the genetic risk.)

Scientists doing this kind of work hope that such data, once presented to individuals, will induce them to change their behavior.

Unfortunately, this is not based on scientific journals.

"This type of personalized risk information has little [or] no impact on people's actual behavior, "said Theresa Marteau, director of the Behavior and Health Research Unit at the University of Cambridge.

In fact, researchers are concerned that when they learn that they have a high genetic risk for diseases such as obesity, they become fatalistic and stop getting sick. try to change their behavior. Fortunately, Marteau said "in a review, we found no evidence of this". It seems that they simply ignore the information.

Ewan Birney, who heads the European Institute of Bioinformatics, has been following this debate over the years. Birney agrees with critics who argue that obesity is not the ideal disease for this type of analysis.

"You have to do more than just show a strong statistical association," he says. "You really have to show that you can then use that to do an intervention."

Birney also fears doing too much of this information because it relies mainly on data from the British Biobank, as well as American samples, in which racial minorities are not well represented.

There are other cases where these polygenic risk scores may be useful, he says. For example, a score that identifies people at high risk for heart disease identifies the people who benefit most from cholesterol-lowering drugs called statins. (But we do not know if it would be beneficial to give statins to people who score high on this test but who would not normally be identified as candidates for this drug).

Using a different assay method, called a genome-wide association study, scientists have identified more than 140 genes that can slightly increase the risk of obesity. Although their individual influence is weak, they provide clues to the biology of the disease.

For example, one of the relatively powerful variants "actually concerns calorie-seeking behaviors," says Ali Torkamani, director of genome computing at Scripps Research Translational Institute. Unsurprisingly, other variants are related to the function of fat cells.

It is also possible that a careful analysis of the genes – rather than the abstract risk score – can identify genetic variants that actually reduce a person's risk of obesity. An article in the same issue of Cell as the one who introduced the work of Kathiresan's group points in this direction.

While genes influence a person's risk of obesity, the epidemic in this country is obviously much larger than the single individuals at high risk. And Torkamani notes that the risk score is not destiny. "It's just a probability," he says. "And you know, when you throw a coin, sometimes it jumps to the head and sometimes to the tail too."

You can contact NPR Scientific Correspondent Richard Harris at [email protected].