The American Heart Association (AHA) is used to damage control, but its focus has generally been on protecting hearts, not its reputation. Ever since the organization of heart professionals, along with the American College of Cardiology (ACC), issued new guidelines for measuring and treating cholesterol levels, it’s been on the defensive.
The radical new way it’s suggesting that primary care doctors evaluate patients for heart disease risk has physicians, not to mention the public, confused and conflicted. Now a calculator that it developed to factor in things such as blood pressure, diabetes and cholesterol levels to assess a person’s risk of heart disease – and need for taking cholesterol-lowering drugs — is also coming under fire for over-estimating risk. That excess could translate to as many as 31 million people who might have to start taking drugs called statins.
The risk calculator, available on the AHA’s website, is intended for doctors to use when deciding which of their otherwise healthy patients might benefit from statins. This group is the hardest to treat, since it has no history of heart disease, but may be at higher risk of heart events or a stroke because of a number of factors, including age, weight, family history of heart disease, smoking history, and blood pressure — all in addition to cholesterol levels. Using the new model, the AHA and ACC recommend that everyone who plugs their information into the calculator and has a risk of having a heart event or stroke that is 7.5% or higher in the next 10 years, should consider taking a statin.
Here’s what you need to know about how the calculator was developed and what it means for your health.
Why did the guidelines for cholesterol treatment change?
The previous guidelines were based on risk factors established by the Framingham Heart Study, a long-term trial of heart disease that began in 1948 and tracked participants’ heart health until their deaths. Most of the volunteers were white and male, and more recent studies suggest that women and African-Americans have higher risks of heart disease and stroke than the study’s participants and may need to be treated differently.
Also, the target cholesterol levels that doctors established – below 200 mg/dL for most adults, and below 100 mg/dL for LDL, or bad cholesterol, were not based on solid scientific studies. There was nothing magic about those numbers; they came from studies of people with heart attacks and were extrapolated to healthy populations to help them prevent first heart events.
Is the calculator flawed?
“None of the risk prediction models is perfect,” says Dr. Sidney Smith, chair of the AHA’s guidelines executive committee, who participated in writing the cholesterol guidelines and is professor of medicine at the University of North Carolina at Chapel Hill and past president of AHA. No algorithm that tries to predict disease is 100% accurate, and the AHA and ACC are standing by the formula it developed to predict risk – acknowledging that as new data becomes available, it may change.
But Dr. Paul Ridker, director of the center for cardiovascular prevention at Brigham and Women’s Hospital, and Dr. Nancy Cook at Harvard Medical School, raised questions about the calculator in an article in the journal Lancet, noting that it overestimates risk by as much as two-fold. They based their conclusion on running the model through three sets of patients from large trails where the participants’ blood pressure, cholesterol and other readings were available, as well as their heart health outcomes.
Ridker says he reviewed the proposed calculator when it was sent to some experts about a year ago for comments. At the time, he raised the potential for over-estimating risk. Dr. Neil Stone, chair of the committee that wrote the cholesterol guidelines and a professor at Northwestern Feinberg School of Medicine, says Ridker suggested using LDL, HDL and C-reactive protein (CRP) levels to assess risk instead. CRP is an indicator of inflammation, which recent studies show can be just as important as cholesterol in contributing to heart attacks. Ridker led those landmark studies and shares a patent on the blood test for CRP, which the AHA recommends as an additional factor for doctors to consider when prescribing statins for patients who show a heart disease risk of 7.5% or higher.
The AHA says that its formula isn’t flawed, but that it will take a closer look at the data Ridker and Cook used, which hasn’t been published in study form, to see if changes need to be made to the calculator. “I will sit down, we will look at that [data] and we will see if there is any information there that could allow us to maybe alter the tool,” says Smith.
Ridker says the data he and Cook considered involved more recent populations than the published, validated trials the AHA relied upon. Since those trials were conducted, many changes – from significant drops in smoking rates, as well as a general trend toward healthier diets and increased physical activity, have altered risk profiles for heart disease. So the assumptions upon which the risk calculator were built – that 35% of the population smoked, for example – may no longer be relevant. “Their data is accurate,” says Ridker. “But what happens over time is an improvement in secular trends, all of which drive heart events down. You need to make sure your score is built on the most recently available data; you can’t build on outdated data because things have changed.”
Smith points out, however, that in order to tally heart events properly in a scientific study, you have to follow people for 10 years or more, since heart disease develops over years, not days. “You will always be a little bit behind the breaking wave in terms of accurately predicting risk,” he says.
It’s also not clear whether the populations in Ridker’s and Cook’s analysis were at particularly low risk – they evaluated health professionals in several large studies, who may be more likely to have their blood pressure and cholesterol levels under control. It’s also not clear whether some of these volunteers started on statin therapy to lower their cholesterol once they entered the study, and therefore actually had lower rates of heart disease than their initial cholesterol levels might suggest. While Ridker admits that’s a possibility, he says that even if all of the volunteers were taking statins, the drugs could not explain the dramatic over-estimation of risk.
Do I have to start taking a statin if the calculator says I’m at high risk of heart disease?
Not necessarily. “If your risk is 7.5%, some people are saying that means you should automatically be on a statin. That’s patently wrong,” says Stone.
The guidelines are just that – guidelines. The AHA always intended for the calculator to be something that doctors could use as one additional piece of information in deciding which of their patients could benefit from statins. The new advice does potentially broaden the group of people who might consider statins, and that reflects the latest studies showing that some people who currently don’t fall into the cholesterol-only criteria for statins won’t be asking their doctors about them. These include those with high CRP levels, or a family history of heart problems, but who currently have normal cholesterol levels. “Do we walk away from the opportunity to prevent first heart events, when we looked at figures from the CDC [Centers for Disease Control] and one in three Americans dies of a heart attack, and 60% of people will have a heart attack in their lifetime?” says Stone. “It’s a big problem.”
So for otherwise healthy people with no history of heart problems, the calculator takes the following into account – age, sex, race, total cholesterol, good cholesterol (HDL), blood pressure, diabetes and smoking history.
If the formula shows that you have a higher than 7.5% risk of having a heart event in the next 10 years, then that’s an indication to your doctor that he should discuss whether you might benefit from a statin – not an automatic prescription. Other factors he might consider before sending you to the pharmacy include things such as your family history of heart disease or stroke, and whether you eat a healthy diet and exercise. He may also fold in additional tests for things such as C-reactive protein, A CT scan of the heart can also detect bits of calcium in the heart arteries, which are the seeds of plaques that can block vessels and cause heart disease.
“This is not computer medicine and we’re not suggesting that all you need to do is go to a computer, put in your numbers, and if it comes out at a certain level, the computer gives you a prescription and your next stop is to pick up a statin,” says Smith. “That’s wrong.” The risk assessment is more like a talking point for the dialogue that a doctor and his patient should have about that particular patient’s risk of having a heart disease, and the best way to help avoid that outcome.