(FORTUNE Magazine) – Don't try telling Ira Lipman to relax. The success of Guardsmark, the private, $400-million-a-year corporate-security firm he built from scratch, hinges on the vigilance he has cultivated in himself and his employees. "I'm in the business of prevention," he says. But none of that helped four years ago when Lipman's doctors told him that the burning sensation clutching his throat whenever he picked up his pace was a sign that his heart wasn't getting enough oxygen. He was only 55, but his arteries were dangerously clogged. "I had a stress test at 1:30 in the afternoon," he says, "and by 5:30 I was having an emergency bypass."

In the lonely weeks of recovery that followed, Lipman had plenty of time to wonder what warning signs he could have missed. He has never smoked cigarettes and has no family history of heart problems. His dad, a Memphis private investigator, lived to 88, and his mother, at 91, still travels the world. Lipman's blood pressure was fine. At his physical six months earlier, his cholesterol scorecard had raised concern; his doctor had written him a prescription for a cholesterol-busting drug. But Lipman hadn't rushed to fill it, and patient and doctor had agreed it was worth trying diet changes and exercise first, to see whether that would be enough to bring the numbers in line. "My doctor said I should watch the desserts and play a little more tennis," says Lipman. Neither man had realized how perilous his condition was. Now Lipman wanted to know: How could this happen?

That question put Lipman on the trail of America's No. 1 killer, and what he discovered shocked and angered him. Though the prevention and treatment of heart disease have greatly improved in the past two decades, there are still gaping holes in the nation's defenses. The cardiac risk factors that Lipman and everyone else have been trained to pay attention to--smoking, blood pressure, obesity, and the like--still can't account for at least a quarter of all heart attacks. And although cholesterol-busting drugs like Lipitor, Pravachol, and Zocor are now swallowed by ten million Americans every day and save thousands of lives as they bring pharmaceutical companies $9 billion in annual sales, they are no panacea. Half of the nation's 1.5 million heart attacks each year happen to people with normal levels of cholesterol.

Lipman began tracking down top specialists, a hunt that led him to Ronald Krauss, an endocrinologist at the Lawrence Berkeley National Laboratories at the University of California. There are tests, Krauss told Lipman, that look beyond the usual definitions of good and bad cholesterol, that separate the bad from the really bad and the mildly good from the angelic. A small but growing body of research suggests that among people who look okay on standard blood tests, the subtler measures may be much more predictive of who's going to get a heart attack and who's not. They're also a guide to more tailored treatment.

Krauss analyzed a sample of the CEO's blood and turned up something Lipman--like most people--had never heard of. He had "LDL Pattern B," a cholesterol profile that fits one in three American men Lipman's age, and one in five postmenopausal women. Roughly translated, the diagnosis means that because of genes Lipman inherited from one or both parents, the particles of so-called bad cholesterol (low-density lipoprotein, or LDL) in his blood tend to be extra small. For reasons scientists don't entirely understand, extra small means extra bad. Though Lipman's cholesterol results at his physical had been unhealthy (total cholesterol, 276; HDL, 45; LDL, 191; triglycerides, 202), Krauss' detailed analysis gave the numbers a much more dire cast.

The arteries of LDL Pattern B patients clog faster than those whose cholesterol is Pattern A. In fact, the risk of heart disease among people with Pattern B is 300% higher. But with treatment they can improve faster too. Soon, with the help of prescription niacin, which increases LDL size, Lipman's cholesterol shifted from the risky Pattern B to a healthier Pattern A. Today, by every measure, his coronary-artery disease is in check.

Still, the man whose business is prevention knows that in 1996 he was one blood clot away from a massive heart attack, and that made him angry. "If I'd known ten years ago there were better tests than just measuring cholesterol, I'd have had them done and gotten treatment earlier," Lipman says. "I'd never have needed the bypass. We're talking about a simple blood test--why don't doctors do it routinely?" Other facts he learned from Krauss made him even madder. Medicare and other insurers will pay for the added tests in some cases. Yet some of the most influential researchers in cardiology still consider them experimental and too expensive. They have yet to recommend the tests, so most doctors and patients don't know they exist.

Ira Lipman thinks that's a scandal, and in 1998 his conviction led him to involve some of the people he worries about most in an unconventional partnership with science. Reaching into his own pocket, he funded a pilot study by Krauss among a cross-section of healthy people. The volunteers: 300 Guardsmark executives and employees.

The saga of cholesterol testing began in the late 1940s, with two guys putting a new lab toy through its paces. John Gofman, a chemist and Manhattan Project alum, had turned his attention to public health after the war, graduating from medical school and taking a research job at the University of California at Berkeley. One day Gofman and grad student Frank Lindgren discovered that when they used a newly invented gadget, the analytic ultracentrifuge, to spin test tubes of blood faster than anyone had before, they got unexpected results. Most of the blood proteins settled to the bottom in layers of sediment according to weight--no surprise there. But several layers ran in the wrong direction: A strange, waxy goop had floated to the top.

Closer inspection showed the goop to be a mix of protein, fat (a.k.a. lipids), and a waxy steroid called cholesterol--the same sort of sludge that researchers had earlier noted in the arteries of animals force-fed a high-fat diet. Gofman and Lindgren isolated several types of lipoprotein from the fatty layers, and classified the particles into four main groups by density (high-density lipoprotein, or HDL; low, LDL; intermediate, IDL; and very low, VLDL). They split the groups further into subclasses that differed according to particle size.

Their work suggested that the link between a bacon cheeseburger and a heart attack is a lot more complicated than just a gullet full of grease clogging the pipes. For one thing, as research soon revealed, cholesterol is far from useless; it's a crucial building block in the construction of cell membranes and sex hormones, among other handy substances. And the fatty triglycerides packed in every VLDL particle are concentrated fuel. Everybody needs cholesterol, even bad cholesterol. People get into trouble, apparently, only when they have too much of a bad thing.

The problems start when, instead of ferrying a load of cholesterol and fat to cells, LDL particles squeeze beneath the lining of arteries and dump their load within the blood-vessel wall. Good HDL particles, the recycling trucks of the fleet, reclaim excess cholesterol from the blood and artery walls before it mixes with other substances to form scarlike plaque. But, especially among people who are genetically vulnerable, diets high in saturated fat and cholesterol can overwhelm the recycling fleet by pumping too much LDL into the blood. The result is bulging plaque that narrows the blood vessel and, under the right conditions, can rupture the lining to trigger a clot--and a heart attack.

Doctors ignored Gofman's taxonomy for decades because the method of measurement he used, while extremely accurate, was also expensive and too difficult for the common clinic. Even researchers stuck mostly to simpler checks of total cholesterol, as few universities had analytic ultracentrifuges. When cheap and reliable methods of measuring HDL and LDL came along in the late 1970s, most doctors didn't pay much attention because of the lack of evidence that knowing the numbers could help treatment.

All that changed when a family of drugs known as statins burst onto the scene. Older drugs had lowered cholesterol by 10% to 20%, about the same reduction achieved by changes in diet and lifestyle. In contrast, statins can safely more than double that drop. The drugs reduce LDL by slowing the production of cholesterol in the liver. That forces the organ to get the cholesterol building blocks it needs by skimming LDL from the blood.

Even the usually conservative American Heart Association started gushing that statins were a breakthrough in treatment when large, long-term studies published in the 1990s showed that the drugs could reduce the number of fatal heart attacks by 25% to 40%. The advent of statins is a big reason that today's routine blood tests measure not only total cholesterol but also HDL, LDL, and triglycerides, which intensify risk to the arteries, though not as much as LDL.

Still, Krauss and other researchers were bothered by the fact that large numbers of people were not being helped enough by statins. For many people, he says, focusing just on LDL "misses where the action is. At high LDL levels--160mg/dl and over--then, no question, it's a very good marker. But as you get into the garden-variety heart-disease patient whose LDL is only borderline high, you run into trouble if that's all you're paying attention to." Every study of people taking statins, he points out, turns up a large number of folks who lower their LDL and still have heart attacks. For them, these cardiovascular wonder drugs are a failure.

Hoping to protect such people, Krauss and colleagues at LBNL developed a better technique for probing the subtle variations of good and bad cholesterol. The method, which is cheaper and easier to scale up to commercial levels than Gofman's approach, also measures a few more subspecies--seven types of LDL and five kinds of HDL, plus other particles and associated proteins. Called gradient gel electrophoresis, the technique sifts and identifies various blood elements according to particle size by propelling the blood through a gel using an electric current.

Throughout the '80s, Krauss and his team used the technique in studies, soliciting patients through the university clinic. They not only correlated various markers with disease progression but also, more importantly, began to track the ability of different treatments to push those markers in the right direction. Though far from conclusive, the science was promising, and as word of the tests spread at medical meetings, cardiologists began calling to ask if Krauss would analyze blood samples from their patients. By 1995 he was getting more requests than his small lab could handle, so he helped a colleague, cardiologist Robert Superko, start a larger clinic. Superko's clinic eventually became Berkeley HeartLab in San Mateo, Calif. (Lipman now has $140,000 invested in the privately held company.) Any doctor can send blood samples to BHL, which charges from $150 to $650 per patient, depending on the number of tests run.

Using Krauss' test or their own equivalent, an increasing number of university scientists have begun looking beyond good and bad cholesterol to the subfractions within. In the past few years, they've run the tests on frozen blood samples from some of the largest and best-known studies of heart disease. Most seem impressed with what they've found. "At least 95% of everybody who studies this now agrees that small, dense LDL are the bad guys," says John Brunzell, a leading lipid researcher at the University of Washington in Seattle. The scientific story is still just suggestive, not conclusive, he notes; most research has demonstrated only an association between LDL size and heart attacks or clogged arteries. The causal link, if one exists, is harder to establish. Just one study has shown that plaque accumulates less quickly among people who have been nudged with drugs from Pattern B to Pattern A. Nobody has shown that people who manage to enlarge their LDL particles live longer--very expensive research that would take years to complete. Yet even without the scientific i's dotted, Brunzell says, "the world is shifting toward using this measure as an independent predictor of heart disease."

For anyone willing to bet that the correlation is actually a causal link, the findings have important implications for treatment. Though statins reduce the total amount of LDL in the blood--a crucial first step for someone, Pattern B or not, who has very high cholesterol--they generally don't affect the size of LDL particles. On the other hand, weight loss of as little as five pounds can boost particle size. For some people, daily exercise is enough to drop the pounds needed to shift them from Pattern B to A. Diets low in saturated fat work too--though Krauss warns that extremely low-fat fad diets may backfire dangerously. In some people, they pull down the size of LDL particles to such an extent that they can throw the dieter from Pattern A into B. For other Pattern B sufferers, like Lipman, losing weight isn't enough; they must rely on a couple of old-time heart-disease drugs--fibrates or prescription-strength doses of niacin--to achieve the shift.

Beyond looking at LDL size, scientists are beginning to think that other particles among the lipid subfractions may be worth measuring as potential markers for heart disease (see box). That could mean big opportunity for the Berkeley HeartLab and other specialty labs springing up to offer advanced lipid testing, like Lipomed of Raleigh, N.C., and Atherotec of Birmingham, Ala. BHL, which expects to top $10 million in sales this year, will follow the cholesterol story wherever it leads, says Superko, adding tests for the most promising, cutting-edge markers for anyone whose doctor orders them. Its primary market: the 19 million Americans who already have obvious symptoms of heart disease and whose tests Medicare and insurance companies will generally cover. Says BHL Chief Executive Frank Ruderman: "Nineteen million people times $300 per person is enough to support this corporation for a very long time."

Brunzell thinks the test for small LDL might prove useful for an even larger group. Any middle-aged adult with high triglycerides who has a sibling or parent with heart disease would be a good candidate, he says. Even people without such a family history might consider getting the tests if their good cholesterol is a little low and their triglycerides borderline high.

The commercialization of any of these tests before they've been fully vetted scientifically makes the guardians of diagnostic standards extremely nervous. Among the most influential is a panel of experts convened by the federally sponsored National Cholesterol Education Program. The panel, which meets every few years to overhaul diagnostic and treatment guidelines, is set to release its latest revision in May. Panel members refuse to say what exactly is in the new guidelines; word on the street has it that tests for small LDL will be mentioned but downplayed. Instead doctors will likely be encouraged to prescribe cholesterol-busting drugs more aggressively and to badger patients more about overeating and underexercising.

"They simply decided that LDL heterogeneity was too much for the general practitioner to be able to handle," Brunzell says. "I personally disagree. The average doctor can handle it, and there are some patients for whom it could be really important."

In the end, the difference between new- and old-guard cardiologists seems to be as much about how they view the world as about the strength of the scientific evidence for any particular risk factor. Traditionalists draw confidence from being able to explain half of all heart disease cases, and like to keep the public health message simple. It's hard enough, they argue, to get people to stop smoking, get off the couch, and quit stuffing themselves--measures that everyone agrees would be great for the heart and would lengthen many lives. How many people would tune out completely if that message were more complicated?

Testing advocates, on the other hand, look at America's heart-disease problem and see large numbers of people like themselves--intelligent, highly motivated individuals who know how to use information and are hungry for more. There will always be debate, they agree, about how best to spend prevention dollars--not every cutting-edge test that can be done should be. But even if you don't count the health risks posed by a heart attack, bypass surgery, or angioplasty, the economic burden they represent is huge--the American Heart Association estimates that coronary-artery disease this year will cost the U.S. $101 billion. In that light, a noninvasive, $300 blood test that's likely to identify people in danger seems to advocates a no-brainer bargain.

Not surprisingly, Ira Lipman is an evangelist for detailed cholesterol screening. He has no patience with the traditionalist approach, and he decided to jump into the middle of the debate himself in hopes of helping the science along. Guardsmark has a tradition of encouraging employees to take care of themselves; the Memphis company offers annual physicals and monthly blood-pressure checks to senior employees, for instance, and invites medical specialists to sales meetings to give lectures on the latest in prostate cancer, breast cancer, cardiology, and even gene mapping. It didn't really seem so much of a stretch, Lipman says, to offer tests by Krauss' lab free to the Guardsmark management team--more than 300 employees. "It's just a pilot study," Krauss cautions. "But it gave us a chance to see what sorts of risk factors would turn up if we offered the test to everybody."

Individual results were kept confidential, but the aggregate findings were eye-opening. Some 30% turned out to have blood abnormalities that put them at risk for heart disease. Half of those cases, Krauss says, would probably not have raised a red flag among doctors relying on conventional tests.

Just as Lipman hoped, the new information is already making a difference in some people's lives. Clara Johnson, a slim 55-year-old accountant for Guardsmark, always considered herself the healthy sibling--it was her older brother who was plagued by heart attacks. She has always watched what she eats, and was quick to start on medication in 1997 when a screening at the company showed she had high blood pressure. The doctor also checked her cholesterol. "It wasn't enormously elevated," Johnson says, "but they put me on a low dose of Lipitor." Despite the drugs, she started having difficulty breathing a year later and, like Lipman, wound up needing a heart bypass. Today, because Krauss' test showed that Johnson has inherited at least two risk factors for heart disease--LDL Pattern B and elevated Lp(a)--she's now added prescription niacin to her drug regimen. She also heads to the gym three times a week and is hoping those things will do the trick.

Not all of the Guardsmark employees got with the program: A number opted not to take Krauss' test. "Some people are in denial about this stuff," Lipman says. "But that's their right." And some younger employees who took the test have had a particularly hard time, as vice president Don Pettus puts it, "getting religion."

Tony Howard is typical. A 45-year-old facilities manager, Howard says he always knew his cholesterol was sky-high; he just didn't take it seriously until the Krauss test. A doctor put him on a stiff dose of Lipitor, and now his number is down from 315 to 200. But he is LDL Pattern B, and his ratio of good to bad cholesterol has his doctor (and Lipman) worried.

Does Howard exercise or avoid fatty foods? Nah. Not really. "My wife's potato-and-bacon soup is ungodly good," he says. "Sausage, donuts, bacon--I love that stuff! I know the medicine's done all it's going to do, and the next step is up to me. But until something really scares me, I don't think I'm going to do anything about it."

That's the cardiovascular dilemma, served in a bowl of delicious potato-and-bacon soup. In the end, many public-health gurus still believe that the sort of testing the Berkeley HeartLab does is unnecessary. Many people who are Pattern B, the gurus rightly point out, also have high levels of triglycerides and most are, like more than half of all Americans, overweight. Wouldn't money be better spent on a prescription to pull down those people's overall cholesterol, or on a gym membership to help them drop the pounds? Surely the Guardsmark results prove, above all, that simply providing information is rarely enough to get people to change their unhealthy ways.

Lipman sees it differently. For him the $90,000 or so he put into the tests was well spent, if only to turn up the small number whose lives, everyone agrees, it may have helped save. Testing led those people to a drug, diet, and exercise regimen that may make all the difference. That's needless for many, perhaps, but of precious benefit to the few.

FEEDBACK: dfranklin@fortunemail.com