Cholesterol, Statins, Disease & Journalism

A couple of questions posted to the listserv of the Association of Health Care Journalists got me thinking.

One asked about side effects of statins. That post was followed by a question about the science linking cholesterol (which statins can lower) and heart disease. The topics raise a number of important points that journalists should keep in mind. I posted some thoughts that are relevant also to the readers, viewers and listeners of the stories we journalists tell.

Great discussions on statins and cholesterol. There are a lot of fundamental questions about both topics that deserve more insightful and nuanced reporting. I think some points are worth emphasizing or clarifying.

As one post noted, most people with high cholesterol will die of something other than a heart attack or related cardiovascular ailment. (Meanwhile, most people who have heart attacks were not identified as high risk in advance.) Keep in mind that this fact neither supports nor undercuts the evidence associating cholesterol with heart disease or the use of statins to reduce risk. It does underscore the importance of distinguishing anecdotes from evidence... and clarifying the difference between what choices and perspectives an individual might have given the gray areas in knowledge... and what policies may gain support on a population basis.

In other words, there is really no conflict between a population-based policy of trying to lower cholesterol levels (based on the consistent correlation of cholesterol levels and heart disease rates) and an individual's choice to not swallow a statin (based on the fact that you have to treat many people in order to change the outcome of one, so most people do not actually receive a measureable personal benefit). I remember being in a group of docs in a workshop discussing how to get a man at high risk of a heart attack to take a statin. In the scenario, the number needed to treat was 8, meaning that for every 8 people treated there probably would be one less heart attack in a given time period... but the other 7 would not have any different outcome (heart attack or not) than if they had not been treated. The population basis rationale for prescribing a statin to that individual makes sense, for every 100 such people treated you'd expect to see about a dozen fewer heart attacks... but it doesn't mean that he as an individual will be necessarily better off, in fact the odds are 7 to 1 against changing the ultimate outcome for that specific individual.

It's also important not to confuse the questions about cholesterol and those about statins. Some people suspect that the effect of statins seen in clinical trials may be related to suppressing inflammation in arteries, reducing the risk of plaque ruptures, or other mechanisms beyond reducing cholesterol.

Always be clear about whether the topic is primary prevention (risk factors) or secondary prevention (preventing a second heart attack, etc.). Almost all of the trials that show giving someone a statin reduces the odds of bad stuff happening have been done in people who have already had a heart attack or are otherwise considered to have heart disease (as opposed to being at risk for developing heart disease). There are relatively few big trials that included people who seem healthy, but have high cholesterol or other risk factors. There is a lot of debate about where to set the parameters for recommending statins to people who haven't had a heart attack or other event, but much less argument about treating people who have already had something bad happen.

I'm wary of those who have bad things to say about statin pills, but recommend foods (or supplements) containing similar substances that are supposed to lower cholesterol or offer other benefits. I'd want to hear a good explanation for how the food would be better than the pill... and what the evidence is. Maybe there is something in some foods that is lost when you distill one ingredient into a pill. We've certainly seen many examples of how people who eat diets high in certain vitamins do well, but then supplements with those same vitamins fail in trials. And if the food is said to be less likely to cause side effects... is that just because the dose of the active ingredient is lower... which suggests any benefit would be less as well?

And be careful to check whether the evidence offered is from experimental trials or just observational studies. Diet studies are notorious for tripping over confounding factors and other flaws... that is, people who eat a more 'healthful' diet often do a lot of other things that are associated with good health. They also may have more education, higher incomes, live in better neighborhoods, and have other characteristics associated with health and longevity. I am very skeptical of diet studies that don't measure socioeconomic status, education and other personal characteristics.

I'll close this tome with a reminder to always ask about relative risk, absolute risk and number needed to treat. It's particularly important in the matter of using statins for primary prevention. Reducing the risk of a heart attack by a third sounds wonderful. But that statistic means nothing without knowing the absolute risk. The effect on 100 individuals with a 20% risk of a heart attack in a decade would be about six or seven fewer heart attacks. But in a group with a 3% risk of a heart attack you would be treating 100 people in order to maybe prevent one heart attack. What's more, the risk of side effects is probably about the same in both groups, so the risk-benefit ratio may be dramatically different.

Sorry to go on at such length. We need more solid coverage of cholesterol, statins, and so on... but like most of the topics on the health beat, there are important details that have to be considered in the editorial process, even if you can't include everything in the final story.

Cheers,

Andrew