A cardiologist interested in genetic epidemiology, Christopher O’Donnell is a natural fit as the scientific director of the Framingham Heart Study’s SHARe project. GT’s Ciara Curtin recently caught up with O’Donnell to discuss how the study keeps up with new advances and where it’s all going. What follows are excerpts of their conversation, edited for space.
Genome Technology: What is SHARe and why was it launched?
Christopher O’Donnell: Well, SHARe stands for SNP Health Association Resource. The first SHARe program is in the Framingham Heart Study, where we have conducted a genome-wide association scan in about 9,500 subjects, spanning three generations of participants and, included in the database, literally over 10,000 different clinical variables.
Why would you want to do this? It has been well-established in Framingham and other studies that many of these common diseases — like hypertension, diabetes, heart attack, stroke, and congestive heart failure — have a genetic predisposition, but until very recently there’s been very little in the way of information about the specific genes and variants in those genes that underlie these common conditions. The notion for conducting a genome-wide association study in a large, well-characterized population like Framingham is that one could obtain information on the genetic determinants of not just one or even three or four diseases, but literally hundreds of different disease measures — both clinically apparent disease and quantitative traits that underlie the disease.
GT: Framingham began 60 years ago. How has it modernized?
CO: What has happened since its inception in 1948 are essentially three things. One is that a second, and then a third, generation of participants were enrolled, so it really became a family study. The second thing is that as new technologies have occurred, either new blood measures of risk markers such as C-reactive protein or particularly new imaging measures — such as computed tomography or ultrasound or magnetic resonance imaging — these kinds of measures have been introduced. The third thing is that we collected DNA beginning in the early 1990s. We obtained lymphocyte samples [that were] subsequently transformed to allow for perpetual production of DNA.
All along the way, we’ve collected environmental measures such as diet information and lifestyle factors, such as smoking and exercise and the like, to allow us to potentially study modifying effects as we uncover new genetic findings.
We have made available a database that can be accessed through dbGaP. Investigators need to provide appropriate documentation of protections, particularly protections of privacy and confidentiality, which I will underscore is incredibly important to the investigators on the study. Once that application has been approved for a qualified investigator, they can access the database.
GT: How do you see SNP data from the Framingham study being used?
CO: The long-term promise of such studies is to focus on what we like to call the three Ps: prediction, prevention, and pre-emption of established disease. In the immediate term, I see a large number of new discoveries — discoveries of new genes and loci implicated in variation, in important disease traits, and in clinically apparent disease as well as confirmation of existing findings. We conducted a smaller, 100,000 SNP association study in about 1,200 of our participants. We saw many confirmations already in that much smaller scan. The hope with such findings is they will have a role in contributing to a new type of medicine: genomic medicine.
As investigators at the Framingham Study, [we] recognize that we are creating an incredibly valuable resource but also that this is an experiment of sorts that we are conducting in terms of how research is going to be conducted in the future.