Every week, it seems, doctors at Harvard and around the country find something new for the health-minded consumer to worry about.

If you don't eat vegetables, you'll get a stroke.

If you take certain prescription drugs, you'll become infertile.

If you don't exercise at least 20 minutes per day, you'll have a heart attack.

These are just a few of the often controversial risk studies produced over the past few years by epidemiologists, or doctors who study statistical relationships. But should anyone listen to the constant barrage, each study with more dire consequences than the last?

One indicator of the tests' importance, it would seem, is their financial impact. And apparently, insurance companies basically ignore the dozens of disease risk studies released by the Medical School each year.

Insurance companies unanimously say that their premiums are based on real past data and proven medical techniques, rather than epidemiologic studies from medical institutions. And even when premiums are reduced or increased, it's hardly a nervous reaction to a single study.

"You may find that your rates could confirm a study or two, but you look at claim experience to determine the rates, so those studies are moot," says Diana Lipps, public relations consultant for the Prudential.

The nature of most of the studies, in which participants fill out questionnaires on their habits and on their incidence of a given illness or condition, means that doctors never actually produce a basis for the increased risk. They can simply associate two groups of data, one of say, eating vegetables, and one of say, suffering strokes.

Nancy M. Kane, assistant professor of management at the School of Public Health, says she doesn't believe insurance companies base premiums on medical study results, but rather on past statistics.

"It has to be a real experience," says Kane. "Actuaries and mathematicians are not so good to say what the future might bring. It's harder to work an actuarial analysis."

Disease risk studies themselves may not affect insurance premiums, but John A. Sanders, general manager of Aetna Health Care Plans for the New England area, says changes in medical technique resulting from medical research results are definitely a part of the actuarial process.

"The key determinant will be if those changes in medical practice would have an effect on medical pricing," he said.

Nancy Peskin, account executive for MetLife, says MetLife has an underwriting policy committee with several medical directors that meets once a month to determine premium changes.

"[The doctors] stay current with the literature, read all the journals, keep up the date with medical developments, and they all have a medical specialty," says Peskin. "They bring to the committee matters which they feel relate to underwriting."

Peskin says MetLife adjusted rates two years ago for policyholders suffering from coronary heart disease and undergoing modern treatments, such as angioplasty--topics of much recent epidemiology research.

"The literature showed that these treatments were very effective and so we did reduce hundreds of thousands of people's health insurance premiums by as much as 50 percent," says Peskin. "It was an acknowledgement that some these treatments were in fact working."

Sanders says changes in response to new medical techniques were not immediate, and that any changes would affect the amount of benefits, not the specific illnesses covered.

"Most of the coverage that we provide is mandated by the state, so there's very little in a contract that pertains to specific diseases," says Sanders. "It might affect the level of benefits for instance under mental nervous benefits, or restrictions on total benefits paid out, but under group insurance contracts there's typically not a line-item exclusion for certain diseases."

For instance, acupuncture is covered in California, says Sanders, because of a state mandate, but not in Massachusetts.

According to Kane, companies generally base group health insurance premiums on industry classes, rather than epidemiologic studies. She says insurance companies compile tables of statistics for different industries, and determine premiums according to the risks peculiar to each industry.

For example, health insurance rates for bars and restaurants may be high because smoking may be more prevalent there.

"The classes can get pretty specific," says Kane. She says that beauty salons are a separate class, because there is a high HIV infection rate there. Construction workers are another group that would fall under a different set of premiums.

But risk factors are still taken account for individuals seeking personal health insurance. According to Peskin, individual policyholders are asked if they smoke or have high blood pressure.

Kane says premiums based on small group or individual experiences were more easily skewed because a single member who used a significant amount of health insurance in the past could greatly distort the group's health rating.

"They're trying to get rid of that [system], and that's what a lot of insurance reform right now is about," says Kane. "They're trying to price people according to utilization, and getting to a community rating."

But is money being wasted in discovering the same relationships twice? In the end, says Kane, epidemiologic studies merely confirm the statistics amassed by insurance companies.

"The medical establishment figures out the why, the insurance industry is projecting on what is," she says. "Medical research is just confirming the scientific and systematic causes."

Medical researchers produce hundreds of studies per year, warning us that our habits are unhealthy and may lead to any number of life-threatening conditions. But should we take all of them seriously?