Should We Believe Anything We Read?

There ought to be an addendum to the old adage, “Don’t believe everything you read.” When reading mainstream media coverage of scientific studies, change “everything” to “anything.”

Case in point: ‘Cholesterol levels linked to early signs of Alzheimer’s in brain’, an article recently published by an NBC senior health reporter, heralds a new study that found a correlation between certain cholesterol levels and evidence of certain proteins in the brain, which are known to help form the plaques found in the brains of Alzheimer’s patients.

While these findings may serve as the foundation for future inquiries into the possible causes of one of our most prolific killers, this is not how the NBC reporter chose to introduce her busy readers to the story (to be fair, “potential foundation for future inquiry found” does not make for a very splashy lede). Rather than honing her words to convey the subject of her report accurately, she led with the most common mistake a science reporter can make, inferring causation from mere correlation:

“High levels of “good” HDL cholesterol and low levels of the “bad” LDL kind are not just helpful for your heart, they’re better for your brain as well, a new study finds.”

Do you see it? If not, try it without the distracting (and misleading) “not just …” phrase:

“High levels of “good” HDL cholesterol and low levels of the “bad” LDL kind are … better for your brain …, a new study finds.”

Instead of explaining the correlation found by the study (i.e., that subjects with certain levels of cholesterol tended to present with more evidence of certain proteins in their brains), the lede implied a causal relationship between cholesterol levels and brain health (i.e., certain levels of cholesterol are “better for your brain”) that was neither found, nor posited, by the scientists conducting the research .

ATG Rule #2 for Evaluating Truth in Science Reporting:

Correlation never proves causation. Ever.

(see our Facebook feed from this morning for Rule #1)

Look at it this way. If a Martian arrived on Earth and happened upon an NBA basketball game, he may note the correlation between the relatively large size of the humans on the court and their participation in the game, but he will not have proven that playing basketball makes a human being taller (nor that great height makes for better ballers). He might infer that height is a beneficial trait in the pursuit of a pro basketball career, but this inference would be a mere hypothesis, which, according to the scientific method, would require rigorous testing before a causal link between height and likely success in the NBA could be properly established.

Unfortunately, the NBC report’s transgressions do not end with the correlation vs. causation debacle. Indeed, the opening sentence itself tweaks yet another pet peeve by perpetuating a popular (READ: false) assumption–that reducing LDL cholesterol prevents heart disease.

Repeat A LieThe reporter can almost be forgiven the misapprehension, as the sorry state of the science on this point has left most doctors with the same impression. For decades, “leading authorities” have admonished us to mind our cholesterol and aim for high levels of “good” (HDL) cholesterol and low levels of “bad” (LDL) cholesterol, under the pretense (among others) that the bad kind clogs arteries and leads to heart disease. The “HDL good, LDL bad” mantra has long outlived whatever overly simplistic usefulness it may once have had, but the medical establishment and those who claim to “report” on its research go on parroting it at every turn.

The great shame of such lazy reporting, especially when it comes to the science of health, is that it misleads and confuses us, causing many of us to give up even attempting understand potentially useful implications for own health. This, in turn, allows food and drug companies to make claims that we often feel powerless to parse or evaluate, and what is lost in the process is much more precious than an appreciation of promising scientific research into a tragic condition well worth understanding.

With neither the time nor space to properly clean up such a mess as this, I will at least try to leave you with a few bits of information to remember the next time you come across a news story about cholesterol.

● Cholesterol is a lipid (fat) produced by the liver and other cells. It exists in the outer layer of every cell in the human body and is essential for proper functioning. It also can be ingested from food.

● As the NBC article notes, HDL stands for high density lipoprotein and LDL stands for low density lipoprotein. What the piece (and most others) fail to explain is that these are not types of cholesterol. Rather, they are proteins, just as their names indicate. They are called “lipo-” proteins because they happen to be proteins that carry cholesterol (and other lipids) through the bloodstream.

● “HDL cholesterol” and “LDL cholesterol” (sometimes noted as HDL-C and LDL-C) refer only to the amount of cholesterol being carried by HDL and LDL particles, respectively. Furthermore, the amount of HDL-C in a person’s blood can only be estimated by reference to other cholesterol measurements, including LDL-C (HDL-C cannot be measured directly by the standard blood panels typically ordered by physicians).

● The body produces roughly three times more cholesterol than is ingested from food, and it would take at least four weeks to eat as much cholesterol as the body has stored at any given moment. Furthermore, very little of the cholesterol we eat is absorbed and retained–most is excreted with other waste.

● While there is ample good science correlating higher levels of HDL particles with lower risk of heart disease, there is virtually none indicating that heart disease is caused by lower levels of HDL-C or higher levels of LDL-C, alone. Nor that lowering one’s LDL-C will prevent a first heart attack. What matters are the numbers of HDL and LDL particles, among other factors, not the volume of the cholesterol payload the particles carry.

For a far more detailed explanation of these points, see Part I of Dr. Peter Attia’s comprehensive blog series entitled, “The Straight Dope on Cholesterol.”

Share Button

1 thought on “Should We Believe Anything We Read?

  1. Hey Brian!

    So many great points. As a whole, we tend to take presented information at face value, and accept it as fact. Not enough people are taking the time to ask “Why?” when presented with new information. Even worse, people take action based on that same information they never questioned in the first place!

    Great reminder to always verify what you are told! Such actions will serve us all in every aspect of life.

    Kudos again!
    Jes @

Leave a Reply

Your email address will not be published. Required fields are marked *