Data from a private insurance company gave scientists a new way of looking at whether nature or diet is more important for staying healthy despite illness. Although the answer is neither definitive nor accurate – it varies according to each of the 560 diseases studied – the technique seems promising to bring more prospects in the future.

The traditional way of studying nature in relation to food is based on twins. Since identical twins share the same genetic code, comparing their health can help determine whether genetic or environmental factors play a larger role in their health. The problem is that it can be difficult to find multiple pairs of twins. Most studies of twins therefore use small sets of data and examine one disease at a time. The new study, published this week in Nature Genetics, uses a database of 45 million people, including more than 56,000 pairs of twins.

It is hoped that the results will help guide future research into the causes of various conditions

Certain conditions, such as Huntington, are 100% genetically influenced, which means that if you inherit a genetic mutation, you have a 100% chance of contracting the disease, regardless of your wealth, where you live or what you are doing. that you eat. The risks of contracting other diseases, such as asthma, are much more influenced by environmental factors, such as climate and wealth, than by their genetic code. the Nature Genetics One study showed that genes affected at least 40% of the 560 diseases, with cognitive impairment being most influenced by genetics. About a quarter of the diseases were at least partly caused by the environment, with eye diseases having the greatest influence on the environment.

The data comes from the private insurer, Aetna, who shared them (with no credentials) with the Harvard University Biomedical Informatics Department. The researchers had the idea of ​​this study when they found that the data included dependents of the principal underwriter. Basically, they could see the information on children's insurance that was covered by the Aetna insurance of their parents, says the co-author of the study. Chirag Patel, professor of biomedical informatics at Harvard University. Next, the researchers looked at birth dates to determine if they were twins. Then, they used a statistical technique to determine the probability that the twins are identical or brothers. (The fraternal twins share only half of their DNA.)

The database included postal codes (which scientists used to extrapolate such factors as socio-economic status and air pollution), a register of doctor visits and diagnostic codes from the International Classification of Diseases, explains the first author Chirag Lakhani, biomedical computer scientist at Harvard. By combining and analyzing all these data, scientists were able to determine the relative contribution of genetic factors to those related to the environment for these 560 conditions, from heart disease to connective tissue diseases to blood diseases.

According to Dr. Lakhani, the hope is that the results will help guide future research into the causes of various diseases. "For example, if you're interested in lead poisoning, genetics plays a very small role and we have to think about the environment," he says. "But in other cases, like ADHD, likely to have a harder genetic component, we can consider other ways to question the disease," he said.

Patel and Lakhani emphasize that their study has limitations. Firstly, they did not examine ultra-rare diseases and, as they were looking for twins young enough to continue to benefit from their parents' health insurance, the analysis excludes diseases such as Parkinson's disease or of Alzheimer's disease that develop during old age.

Dan Belsky, professor of population health at Duke University who did not participate in the study, said the study method solved a major problem in medical research: people who sign up to participate in studies could be fundamentally different from those that are not. This makes the results incomplete and unrepresentative. "Nobody will collect data from as many people, but you can partner with the data holders to exploit the extraordinary data entry into our lives to advance science," he said. "This is a very careful study and I find it very exciting to see this scale of data put to this question."

That said, today's study does not solve this problem perfectly either. As Patel points out, people with private insurance like Aetna have different situations than those with, for example, Medicare. The next step is to try using the method on many different databases. "I think it's extremely possible to apply these same methods to these populations to better identify other important factors that did not appear as strong," he says.