"The formula of diseases"
The most extensive and detailed scheme of the interrelations of diseases has been compiled
Dmitry Safin, "Kompyulenta"
A group of researchers from Northeastern University and Harvard University (both in the USA) studied the data of 32 million claims for insurance compensation filed by Americans over 35 years of age from 1990 to 1993, and compiled a diagram of the interrelationships of diseases, which included more than ten thousand names.
In their work, the authors used the standard International Classification of Diseases. All statistical data and final versions of the schemes have been made freely available by scientists on this site; there is also a special form in which you can enter the name of the disease to view the corresponding part of the scheme.
On the graphs obtained by the researchers, diseases are located in nodes, and the edges represent the connections between them, and the "distance" (the number of intermediate nodes) from one disease to another reflects the degree of their interdependence. To establish these ratios, the criteria of relative risk RRij and phi-correlation φij were applied (see Figure). The relative risk indicator is focused on rare diseases, while the phi-correlation value is more suitable for assessing the links between common diseases.
After analyzing the constructed schemes, the scientists found several interesting patterns. For example, the appearance in patients of a disease associated with many others (representing a great danger to life) usually occurs against the background of the development of a less "common" disease; therefore, according to the position of each individual in the scheme, it is possible to predict how many years he will live.
The researchers also noted a pronounced dependence of the type of personal scheme of diseases on the gender and nationality of a person.
"Every disease can have many causes, so the discrepancies in the relationships between diseases observed in different populations will allow us to identify the causes most characteristic of specific populations," said Cesar Hidalgo from Harvard University in an interview with the PhysOrg website. In the future, the authors believe, the schemes they have created will be used in everyday medical practice: for example, during an examination, a doctor will be able to obtain information about all the patient's diseases, related diseases and – in the future – a set of genes responsible for their development. In addition, scientists hope to interest representatives of pharmaceutical companies in the results of their research.
The full version of the report is published in the journal PLoS Computational Biology.
Prepared based on the materials of PhysOrg.
Portal "Eternal youth" www.vechnayamolodost.ru20.04.2009