Study Questions Definition of Hypertension
Millions of people classified with high blood pressure may not really have the problem after all, according to a provocative UCLA study that counters the prevailing wisdom and questions government health guidelines for treating hypertension.
The study, which appears today in the Lancet medical journal, drew sharp criticism from hypertension researchers, who said it ignored extensive clinical evidence on the benefits of controlling hypertension.
The UCLA team suggests that a standard definition of hypertension--a systolic pressure of 140 or above--is inappropriate for many older people because of flawed statistics that doctors and researchers have embraced for decades. Systolic pressure is the force exerted on blood vessels while the heart contracts.
Instead, the UCLA scientists say, blood pressure may exceed that level in many people before it increases the risk of dying. The most dramatic proposed revision involves women 65 to 74, whose systolic blood pressure may not be dangerously high until it reaches 167, the study found. Thus 40% of women in that age group currently defined as having hypertension would be reclassified as having normal blood pressure.
Similarly, the study sets the cutoff for high systolic blood pressure at 158 for women ages 55 to 64, 159 for men ages 65 to 74, and 148 for men 55 to 64 years old.
Current standards apply to men and women under 45, according to the UCLA research, which was based on a reanalysis of 30-year-old data from the Framingham Heart Study.
Government and other outside health authorities who previewed the study expressed fear that it might cause many people to eschew measures to treat hypertension, which afflicts about 50 million Americans and is associated with the nation’s No. 1 and No. 3 killers, heart disease and strokes.
The findings “reach beyond the data and draw conclusions that are inappropriate,” said Dr. Daniel Levy of the National Heart, Lung and Blood Institute and current director of the Framingham Heart Study. “Sending the message that current national blood pressure guidelines are mistaken . . . flies in the face of enormous advances that have occurred in public health.”
Although blood pressure is a fairly easy thing to measure, it nonetheless reflects body functions that experts do not entirely understand. Researchers do not know the primary cause of as much as 95% of high blood pressure, the American Heart Assn. says. Still, quitting smoking, losing excess weight, exercising and other measures have been found to lower blood pressure somewhat.
Guidelines issued in 1997 by the National Heart, Lung and Blood Institute define optimal blood pressure as 120 / 80 millimeters of mercury. The top number is the systolic pressure, and the lower number is the diastolic pressure, the force between contractions.
The guidelines urge lifestyle changes or medical treatment for all people with a blood pressure 140 / 90 or above. Chronic high blood pressure is associated with strokes, kidney failure, heart disease and heart failure. The UCLA analysis did not analyze diastolic measurements.
The new research fits into a long-running scientific effort to pin down the precise relationship between blood pressure and disease risk.
The prevailing view, reflected in the national treatment guidelines and based on scores of studies with hundreds of thousands of people, is that disease risk goes up in lock step with each blood pressure point above the optimum. Accordingly, even moderately elevated blood pressure should be addressed, and the seemingly natural tendency of blood pressure to rise with age should be curbed.
Arguing against that view is the UCLA study, which suggests there is a wide range of “normal” blood pressure within which the risk of harm is stable--followed by a sudden steep risk increase.
The two mathematical views represent different conceptions of reality. In the prevailing view, every little bit of excess blood pressure adds to the damage. In the opposing view, it’s as though the body can withstand a certain amount of blood pressure-related stress before being damaged, rather like a bridge supporting great weight before it collapses.
The researchers, led by mathematician Sidney Port and cardiologist Linda Demer, say they arrived at this “threshold” model after subjecting blood pressure and death data from the Framingham study to a statistical analysis that biomedical researchers seldom use.
That method, a more complex analysis that Port called a spline model, is more appropriate for probing such data, the researchers say, than is the commonly employed form of so-called linear regression.
Port said that other biomedical researchers had not thought to use this newer approach because they have long been accustomed to the other one. “ ‘Monkey see, monkey do’ is the word in medicine,” he said.
The researchers chose to reanalyze the first 18 years of the Framingham study involving more than 5,000 people, because it has been so influential and was so carefully done. The study is named after the Massachusetts city whose residents first volunteered to participate 52 years ago.
“The main consequence of this study is that a large portion of the population currently considered at increased risk, due to systolic blood pressure over 140 . . . are not,” wrote the researchers.
Demer, UCLA’s cardiology chief, said she was “very confident” of the finding and that she is expecting “negative reaction to it.”
Levy, the Framingham study’s current director, said he did not dispute the UCLA analysis as far as it goes. But he said it ignores mountains of other evidence for the direct relationship between blood pressure and disease risk, including recent data from the continuing Framingham study. Further, he said the researchers were remiss in not correlating blood pressure readings with hypertension-related outcomes other than deaths, such as strokes, kidney failure and heart attack.
Dr. Daniel Jones, a hypertension expert at the University of Mississippi Medical Center and an American Heart Assn. research official, said: “I think the conclusions are wrong. It’s a valid research argument . . . but these authors have basically just taken a different statistical approach to analyzing someone else’s data and it’s pretty bold of them to suggest everyone else is wrong.”
He expressed concern that, because of the publicity about the report, people with borderline high blood pressure might be dissuaded from taking steps to reduce it or that physicians might stop treating high blood pressure in the elderly: “This is not new data that patients and physicians need to react to.”
(BEGIN TEXT OF INFOBOX / INFOGRAPHIC)
Pressure Change?
A UCLA study challenges conventional thinking about high blood pressure, defined as 140/90 or above. Here are the study’s proposed threshold values for dangerous systolic pressure, the fraction’s top number:
*
