Summary

The article discusses the efficacy of modern genetic testing for predicting future obesity compared to traditional family history assessments, concluding that while obesity is a polygenic trait influenced by multiple genes, current genetic tests offer limited predictability, and family history remains a more reliable indicator.

Abstract

The article examines the scientific opinion on the predictability of future obesity through genetic testing, contrasting it with the traditional method of assessing family history. It explains that obesity is a polygenic characteristic influenced by numerous genes, with the FTO gene being a well-known example that accounts for only a small fraction of BMI variation. Genetic testing for obesity involves screening the entire genome to provide a genetic risk score, but the accuracy of these tests is modest, with AUC values slightly better than chance. The predictive power of genetics is somewhat higher for extreme obesity. In contrast, family history, such as parental obesity and one's own BMI history, shows stronger predictability for future obesity. The article questions the utility of genetic testing for obesity, given that the prevention strategies do not differ based on susceptibility and that environmental factors play a significant role in modifying genetic expression. It suggests that the future of body weight management may involve personalized dietary approaches based on genetic metabolism differences.

Opinions

The author suggests that genetic testing for obesity risk, while advancing, is not significantly more predictive than a simple coin toss, with AUC values ranging from 0.52 to 0.61.
The article posits that family history, particularly the body weight status of parents, is a better predictor of future obesity than current genetic tests, with an AUC of 0.79 when considering BMI at 11 years old for predicting obesity at 33 years old.
There is a concern about the lack of transparency and standardization in how genetic testing companies calculate genetic risk scores, which may lead to varying results from different companies.
The author questions the practical value of predicting obesity susceptibility since the prevention strategies are the same for all individuals, regardless of genetic predisposition.

SCIENTIFIC OPINION

Can Today’s Genetic Test Predict Future Obesity?

Modern inexpensive genetic approach vs traditional approach based on family history.

The idea of obesity-related gene became popular since the association between the FTO gene and obesity was being reported in 2007. A slight variation in the coding of the FTO gene occurs between individuals has contributed to the variable susceptibility to weight gain. Since then, a number of other genes have been proposed to be associated with weight gain, even the CDC offers some common examples here.

Recently a group of researchers published their findings on 20 April 2020 after the researchers successfully prevent weight gain in mice after artificially prevent the expression of Asxl2 gene. How cool it is if humans can be treated to eat without gaining weight?

Nowadays, many independent companies offer inexpensive genetic screenings to screen for the risk of future obesity. Upon reviewing a few scientific articles, I aim to write this article to address the following:

Is obesity a monogenic or a polygenic characteristic?
How predictable is genetic testing for future obesity?
How predictable is asking someone’s family history for future obesity?

Lastly, I will be concluding by questioning who can benefit from genetic testing for obesity and other diseases.

Obesity: A monogenic or a polygenic characteristic?

The meaning of monogenic vs polygenic

Monogenic means a biological characteristic is controlled by one gene only. For example, the cystic fibrosis and sickle cell anaemia are rare diseases due to mutations in one gene. It is often expressed as an “all-or-none” characteristic.

Polygenic means a biological characteristic is controlled by more than one gene. The polygenic characteristics are like skin colour, body height, body weight, etc, expressed in a continuum.

Obesity is polygenic and its implication on genetic testing.

Using the example of FTO gene, the “worst” coding you can get only explains a maximum of 1.45% variation in the BMI difference when compared to those who have the “ideal” FTO gene coding. The remaining 98.55% variation in BMI remains unexplained. Hence, testing a single gene for obesity is not informative.

Testing a single gene for obesity is not informative.

Since multiple genes contribute to our body weight (polygenic), genetic testing involves the screening of our entire genome to look for the genetic codings that are associated with obesity. Then, we will receive a genetic risk score that tells us whether we are at risk of future obesity.

With that being said, the number of genes associated with obesity is still increasing. Scientists make predictions from the genes that are currently known. Understandably, it is unlikely that scientists have fully discovered all genes associated with obesity, even so, scientists are still on the way to study whether a gene variant is actually causing obesity or just merely present in those with obesity.

Genetic testing for obesity: How predictable it is?

Prof Ruth Loos who leads the Genetics of Obesity and Related Metabolic Traits Program evaluated the specificity and sensitivity of the genetic test for predicting future obesity using AUC.

AUC ranges from 0.50, signifying a prediction equals to tossing a coin, to 1.00, signifying a perfect prediction.

Prof Loos reviewed a number of different genetic tests and found that their predictions ranged from 0.52–0.61. For me, it seems like the prediction is just slightly better than tossing a coin.

What is the point to predict?

Adding other predictors such as education level, employment, dietary intake, medication, smoking habits, and physical activity, may improve the prediction to 0.69. Furthermore, Prof Loos also showed us that genetics performs better in predicting extreme obesity (0.70) than anywhere between normal weight and less severe obesity.

Some researchers also argue the lack of transparency on how certain companies calculate our genetic risk score. For sure, genetic testing companies claim their algorithm to be proprietary. For this reason, it may be possible that the results you receive from one company can be different from another company.

“Are companies protecting their scientific innovation or hiding the lack of scientific methods?” — Prof Janssens from Emory University.

Can a family history predict future risk of obesity any better than genetic testing?

By using the self BMI history, our BMI at 11 years old can predict future obesity at 33 years old with an AUC of 0.79, far better than any genetic testing.

Furthermore, by asking a simple question, “Do you have an obese parent?”, often lead to good predictability. Having at least an obese parent at childhood can predict future obesity at 59–78% specificity (true positive) and 52–73% sensitivity (true negative). This is likely because assessing body weight status of parents may be a good representation of both genetics and environmental influence.

Unfortunately, the AUC statistic is not available for this testing.

Final thoughts

As we know obesity is commonly related to various diseases, it is certainly important to prevent extreme weight gain, but does it make it necessary to predict it?

Does it make any difference whether to know if we are susceptible to obesity?

We can’t change our genes. Does it make any difference whether to know if we are susceptible to obesity? Importantly, environment factor is a strong modifier of gene expression, with the ability to turn on, turn off, augment or diminish it.

If we know we are less susceptible to obesity, we may tend to increase risky behaviour that may lead to weight gain, which becomes dangerous. In contrast, if we know we are more susceptible to obesity, the current obesity prevention strategy is not different for those who are susceptible to those who aren’t. Then, what is the point to predict?

Some genetic tests have been conducted in clinics under medical supervision to predict genetic disease of a fetus. However, a genetic disease such as cystic fibrosis can be predicted at 100% specificity and 97.5% sensitivity. This makes sense to predict and such genetic disease has life-long implications on newborns.

Nevertheless, the future direction of body weight management is likely to be genetic-related. We are very interested to know how our body metabolizes carbohydrate, protein and fat differently, as we know some people may perform better on one diet than the other.

In short, the current science’s verdict is that if you want to predict your future risk of obesity, look at your parents’ and your childhood’s body size.