A 500,000-person study?

[iansmith]

Gene-environment interactions would be focus of NIH-led effort | By Maria W Anderson

The National Institutes of Health (NIH) is considering undertaking the largest population-based study ever done in the United States. NIH issued a request for information (RFI) from researchers earlier this month about the questions a large cohort study on the gene–environment interactions involved in common human diseases might ask, and how the study might be constructed.

A project of this kind is "the logical next step beyond the mapping of the human genome and doing case studies," said Terri Manolio, director of the National Heart, Lung, and Blood Institute's epidemiology and biometry program.

Such a project would try to survey a representative sample of the US population, explained Manolio, and may include as many as 500,000 participants from all geographic, racial, ethnic, and socioeconomic groups defined in the most recent US census. No funds have been appropriated for the project yet, and NIH officials are hesitant to speculate on how much it might cost.

Manolio said that NIH officials hope to find a way to incorporate data from previously conducted studies of individual diseases. "We want to include existing cohorts," she said, "but we have to decide, how feasible is it to add on to these disease studies?"

http://www.biomedcentral.com/news/20040526/04

[Monique]

This is very different from the usual genetic studies, where participants are recruited from high risk geno-/phenotypes (special populations isolated by religion, geography or other reasons). It's logical that they'll need such a huge sample if they want to have any statistical power to compare the different groups.. I wonder how they plan on going about this..

[arildno]

It would be great if they find backers for such a massive project!
The only comparable study in scope that I know of, is the compilation of genetic data for practically the whole Icelandic population.

[Monique]

A major difference (besides the fact that the Icelandic population is very homogeneous) is that the attitude of the US population towards biomedical research is very suspicious and defensive if it comes to collecting blood samples and predicting disease risk based on genotype. The reason I say this is that in the US people are afraid they might loose their insurance if information about their genetic makeup would be disclosed.

[Pattielli]

How do you know so much about them like that ? What do you mean by insurance ?

Thank Monique

[Monique]

Because I have talked to people who do this kind of studies and worked on a study like that myself. Researchers who performed the studies noted that the readiness of people to participate in studies is influenced by the way insurance is taken care of in a country. Compare public insurance that is taken care of by a government who treats everyone equally to private insurance companies who run a business.

[russ_watters]

I wonder how they plan on going about this.. My knee-jerk reaction is to be very wary of bias in their selection group. To get so many people in a study seems like it would require connection to a common event - say for example, blood donation. That wouldn't necessarily yeild a cross-section of the country, but rather a cross section of those in the country who would donate blood.

edit: whoa - 4,000 posts. I need a girlfriend bad...

[Monique]

My knee-jerk reaction is to be very wary of bias in their selection group. To get so many people in a study seems like it would require connection to a common event - say for example, blood donation. That wouldn't necessarily yeild a cross-section of the country, but rather a cross section of those in the country who would donate blood.
But they won't be treating the group as one, rather they want a lot of small groups that together represent the population. Two large studies have been the UK biobank and the mentioned Iceland DeCode project, the US has populations that are very different from these (the extensively studied Mormons for instance). In the article it says they want to include participants from all geographic, racial, ethnic, and socioeconomic groups defined in the most recent US census. They will be comparing probably whether people living in the mountains are at higher risk of getting stroke than people living at sea-level for instance.

edit: whoa - 4,000 posts. I need a girlfriend bad... :biggrin: