The discovery was made by two international research teams working independently, decoding the DNA of about 30,000 people.
The gene mutations were rare in people under 40, but found in about 10 per cent of those over 65 and in nearly 20 per cent of folks over 90.
Having one of the mutations does not destine someone to develop a blood cancer, but it raises the risk of that more than tenfold. It also increases the chance of a heart attack or stroke, and of dying from any cause over the next four to eight years.
"We are hopeful that someday we would be able to use this as a screening test and identify individuals who are at risk," said one study leader, Dr. Benjamin Ebert of Brigham and Women's Hospital in Boston.
However, "nobody should go out tomorrow and look for these mutations," because there's no treatment for having one or way to prevent cancer from developing, he said.
The studies are to be presented next month at an American Society of Hematology conference but were published ahead of time, online Wednesday by the New England Journal of Medicine, because of their importance. They were led by the Broad Institute of MIT and Harvard-affiliated groups, and mostly funded by the National Institutes of Health.
Each year in the United States alone, about 140,000 people are diagnosed with a blood cancer. All cancers are caused by bad genes, but most of them are not inherited — they pick up flaws during someone's lifetime for a variety of reasons. It usually takes several mutations for a cell to become so abnormal that cancer results.
Researchers wanted to see if they could find the first step in that process, the initial mutation, and detect a pre-malignant state long before cancer develops and symptoms appear.
One group looked at more than 17,000 people in a study of diabetes and heart disease risks. The other looked at more than 12,000 people in a study of mental disorders. All of their genes had been sequenced, which takes about a week and now costs less than $1,000. That produced a trove of "big data" that can be mined to study new things, Ebert said.
The teams independently found three genes whose mutations accounted for most of the blood cancer risk they saw. About 1 per cent of people with one of the mutations went on to develop a blood cancer within a year, or 10 per cent within a decade.
"These are exciting" findings, because they reveal the origins of some of these cancers, said Dr. Janis Abkowitz, blood diseases chief at the University of Washington in Seattle and past president of the American Society of Hematology. But people who have their genes analyzed for other reasons and accidentally discover one of these mutations should not panic, because "the odds are they won't develop cancer at all," she said.
The studies also found that people with one of the blood gene mutations had more than twice the normal risk of heart and strokes.
"That is a little bit of a surprise," said Dr. Parameswaran Hari, a blood cancer specialist at Froedtert & the Medical College of Wisconsin. "We don't know why but I think we'll figure it out soon."
Dr. Wyndham Wilson, a senior scientist at the National Cancer Institute, said it is way too soon to consider screening certain age groups because there is no evidence it would do any good or prolong survival.
"The number of people being harmed by all the tests, the anxiety, would far outweigh the benefits," he said.
It's not like screening for pre-cancers like polyps, which enables surgery that can prevent or cure colon cancer, he added.
"With blood cancers, you can't take them out."
Marilynn Marchione can be followed at http://twitter.com/MMarchioneAP