I have a unique vantage point when it comes to the issue universal cholesterol screening in children, when compared to most pediatricians.  My unique view stems from the fact that I am also an internist who deals with those children after they grow up on KFC Double Downs.  The AAP's recommendations are supposed to be forward-thinking, addressing the increased rates of obesity in our country (in case you hadn't heard) and responding in kind.

The study that raised this issue recently appeared in the Journal Pediatrics, and it found the following (from a report on the study by TIME:

Among the 5,798 children who would not have been screened, nearly 10% had elevated LDL cholesterol levels (above 130 mg/dL), according to the study, published in the August issue of Pediatrics. And 1.7% had levels high enough (above 160 mg/dL) to warrant cholesterol-lowering medications. Indeed, of all the children in the study whose LDL levels were above 160, one-third were in the group who would not have been screened.

The current guidelines were put in place in the 1990s, and at the time, experts predicted that high cholesterol would be missed in as many as 25% of children, says Neal. But doctors assumed that in most cases, these children would have only slightly elevated cholesterol levels that would eventually be detected when they became adults and would be lowered with diet and exercise.

Neal's data show that may not be the case. Further, Neal says, the findings apply not only to heart-disease risk but also to the risk of diabetes, since high cholesterol at a young age is a strong predictor of prediabetes and diabetes. "We worry about that because if these children don't change their ways, then they are going to have Type 2 diabetes," he says. "It's something we would become more aware of if more children were screened."

So it seems obvious that we should be screening these children, right?  I am fairly certain that many of my pediatric colleagues will become more aggressive on cholesterol screening because of this article.  I think this is a mistake.

In my post, 10 Rules For Good Medicine, I stated in rule #1: ?Order as few tests as possible.  No test should be ordered for informational purposes only; the question, “What will I do with these results?” should always be answerable.  If it is not, the test should not be done.  The validity of this comment was debated by Dr. Centor on Kevin, MD, and I stand by that statement.  We don't order tests to heap on guilt.  We don't order tests to "just know" things.  The reason to order a test is to make a decision: should I treat this patient, or should further testing be done?

In this case, I think nearly all cholesterol testing by a pediatrician is on very thin ice from an evidence standpoint, and for that reason, I don't check cholesterol in children or in teens unless there is a family history of premature coronary heart disease (under age 40).  Furthermore, I think that any doctors who are doing so should stop until there is more evidence.  But why shouldn't we find out if kids have high cholesterol?  What's the harm in finding out?  Here's the harm:

  1. We don't have a clear understanding of what high cholesterol numbers mean in children.
  2. We don't know if treating cholesterol in children reduces heart disease risk.
  3. We don't know the long-term harm in children and teens of using cholesterol-lowering medications.

Isn't it clear that high cholesterol causes heart disease?  Isn't it clear that lowering cholesterol prevents heart disease?


High cholesterol is associated with a higher risk of heart disease.  Lowering cholesterol in certain specific situations with certain specific medications is associated with a lower rate of heart disease.  We need to be very careful in this.  Why?  Because much of the claims of alternative medicine are based on the inference that association implies causality. It may be that cholesterol itself (LDL and HDL in particular) is one of the agents that causes heart disease, but that hasn't been proven.  Why did Vytorin, which lowers cholesterol more than Zocor, not also lower the rate of heart attacks or fatal cardiac events?  (Vytorin contains Zocor plus one other medicine that lowers cholesterol more.)  The fact that it doesn't, suggests that cholesterol doesn't tell the whole story.

The bottom line?  I am not testing kids for cholesterol because I don't want information I can't use.  Sure, I can use it to scare parents and kids, but I don't think scare tactics really work.  It doesn't change smokers' behavior to tell them how deadly cigarettes are.  Besides, since there is no proof that giving a kid Lipitor will help him more than it harms him (you don't give it to girls because of birth defect risk).  So you end up giving the same "eat right and exercise" talk you give if you hadn't checked cholesterol in the first place.

I'm sure the drug companies wouldn't mind if kids were tested, though.

(See Gary Schwitzer's Coverage on this study for more perspective on it.)