Genes Have "No Clinical Relevance" in Predicting Disease, Scientists Admit

Monday, November 09, 2009 by: David Gutierrez, staff writer
Tags: genetic diseases, health news, Natural News

(NaturalNews) Genetic analysis is essentially useless in predicting a person's risk of cancer, heart attack or other common diseases, according to a set of commentaries published in the New England Journal of Medicine.

The decoding of the human genome in 2003 led to a flood of research into the contributions that genetic variation might make to the risk of various chronic diseases that tend to develop late in life, such as diabetes, heart disease or cancer. Since then, a number of for-profit companies have begun offering genetic screenings and disease risk assessments.

"With only a few exceptions, what the genomics companies are doing right now is recreational genomics," said David B. Goldstein of Duke University, author of one of the commentaries. "The information has little or in many cases no clinical relevance."

A few diseases, such as sickle cell anemia or Tay Sachs disease, can be caused by a mutation on a single gene. Most diseases, however, are much more complex, and develop due to an interaction between environmental factors and the contributions of a variety of genes.

In order to calculate the genetic contribution to these diseases, geneticists developed the technique known as genomewide association study, in which the genomes of groups of healthy people are compared with the genomes of those with a certain disease. The hope among researchers has been that this analysis could point out genetic differences between the groups that might be linked to disease risk.

While genomewide analysis has successfully identified some differences between the genetic codes of healthy and diseased patients, these differences provide little information about disease risk. Researchers expected to find a small number of common genetic variations that were responsible for each disease. Instead, common gene variants appear to contribute to disease risk only very marginally.

Goldstein suggests that the genetic component of chronic disease risk might arise from a large number of rare genetic variants, making genetic screening far less helpful in predicting risk. If only a small number of variants are involved, they might provide information on the biological pathways that lead to the development of the disease. If many different variations are implicated, however, this would tell researchers nothing except that an error in the functioning of nearly any bodily system can contribute to disease risk.

"In pointing at everything," Goldstein wrote, "genetics would point at nothing."

Based on the failure of more than 100 genomewide studies -- carried out on thousands of patients in a number of different countries -- to deliver useful results, Goldstein suggests that the approach should be abandoned as a waste of resources. Genomewide analysis is only effective at uncovering common variations with large effects, Goldstein says. Discovering rare variants with smaller effects is "beyond the grasp of the genomewide association studies."

A single genomewide analysis study costs several million dollars to carry out.

In addition, as more such studies are carried out, each one will deliver less new information on gene-related disease risk -- a problem of diminishing returns.

Also writing in the New England Journal of Medicine, researchers Peter Kraft and David J. Hunter of the Harvard School of Public Health disagree with Goldstein that the studies are incapable of providing useful data. They agree, however, that they have provided no clinically useful information to date.

Kraft and Hunter warned that any risk estimate based on genetic analysis is likely either too low or too high, because researchers simply do not have enough information on genetic contribution to disease. Risks that are calculated as lower than they actually are will likely never be reported to patients. This means that any genetic risk a patient actually gets told about is probably an overestimate, and only likely to cause unnecessary alarm.

Sources for this story include: www.nytimes.com.