Some diseases, such as H.I.V. and prostate cancer (the PSA
test) have long been diagnosed by detecting a single protein. Pregnancy is also
detectable by a single protein. In this case, it’s a do-it-yourself pee stick
that measures a hormonal protein produced by the placenta. But the newer
research is focusing on diseases with protein “fingerprints” that can involve
thousands of proteins, as is the case with diabetes. Such diagnostic ability
has only recently become possible, thanks to powerful and sophisticated
computers.
Using enormous banks of computers and thousands of blood
samples associated with thousands of health records, a couple of companies are
developing algorithms that are beginning to identify some of these disease-related
protein fingerprints, such as for lung and pancreatic cancer and heart disease.
The fingerprints convey your odds of getting sick, the current state of your
disease, and its trajectory.
The journalist, whose article inspired this blog post, and
whose super-fit 71-year-old mother had died suddenly of a heart attack, had his
protein fingerprint analyzed. The results told him that he had an 11 percent
chance of having a heart attack within five years. He immediately started
taking cholesterol-lowering drugs and avoiding red meat. After one year, his
blood was tested again. He did not drop a single percentage point. (As an
anti-cholesterol-lowering drug person, I love this result.)
For an introduction to this blog, see I Just Say No; for a list of blog topics, click the Topics tab.
For an introduction to this blog, see I Just Say No; for a list of blog topics, click the Topics tab.
No comments:
Post a Comment