Junkfood Science: Questions the media didn’t ask

November 13, 2008

Questions the media didn’t ask

You’ve no doubt heard about a study presented this week at the American Heart Association conference in New Orleans that's being reported as finding “striking evidence” that children who are fat or with ‘high’ cholesterol have early signs of heart disease. The thickness of the inner walls of their carotid arteries (“carotid intima-media thickness” measurements or CIMT) was reported as resembling those of a 45-year old, someone 30 years older than their actual age. This has made news headlines around the world as “alarming” proof that fat children are at risk for early heart attacks, strokes and death.

No it isn’t.

As has become routine, the strength of the science is inversely related to the amount of media coverage. When hundreds of news outlets around the world report on a single study, of the hundreds released each day, all saying exactly the same thing, you can be sure someone issued a press release. Sure enough, this paper came with a press release, too. Marketing departments issue press releases.

The AHA press release added a special note to the media about its partnership with the William J. Clinton Foundation in the creation of the Alliance for a Healthier Generation and it linked to their anti-childhood obesity initiatives surrounding healthy lifestyle choices. The media appears to have complied and, without any evidence at all, linked CIMT measurements to diet and exercise. This abstract is being reported as a wake up call for the urgent need to take steps to prevent obesity and high cholesterol levels before children’s arteries grow older than their years. By implementing lifestyle changes, like diet and exercise and maybe cholesterol-lowering medications in children as young as eight, perhaps early artherosclerosis in fat kids can be reversed, we’ve been hearing. Yes, press releases are marketing.

This media coverage has been based on an abstract (#6077) presented at a meeting. No study has been published, which means the study hasn’t undergone any peer-review, nor is there even a study available for healthcare professionals to examine its data, methodology or interpretations. Studies presented as abstracts at meetings have this troubling habit of being so flawed they never make it through the peer-review process to get published.

Still, this abstract, alone, provides all the information needed to know this was not a fair test to support any of the claims in the news.


This abstract was a collaborative student project that was first presented on April 25th at the University of Kansas City School of Medicine Student Research Day by statistics graduate student Danna Zhang (whose faculty mentor was Jie Chen, Ph.D.). Her co-authors were medical students Joseph Le and Spencer Menees (whose mentor is Geeth Raghuveer, M.D. at Children’s Mercy Hospital) and fourth-year pediatric resident at Children’s Mercy Hospital, David McCrary, M.D.

They calculated the CIMT from carotid artery ultrasounds they did on 70 high-risk children (ages 6-19 years, evenly boys and girls) at the cardiology department of Children’s Mercy Hospital. Most of the kids were white (89%), and just over half had BMIs that were above the 95th percentile on growth curves (BMIs 25.6±6.0kg/m2).

The children’s mean CIMT values were 0.45mm±0.03, with a wide range (maximum 0.75mm).

Because there is no published data on CIMT values in young people, the authors compared their CIMT calculations to CIMT percentile tables published for race and sex-matched adults. They reported that the CIMTs of the children were evenly divided between those who were above the 25th percentile on the charts for 45-year olds, with 38 children with triglyceride levels over 100mg/dL falling above the 25th percentile for 45-year olds. That’s it.

Questions we didn’t hear asked

Journalists failed to ask the right questions or provide any help to parents as to what this study means for their children, if anything.

What are normal CIMT values among all children? Since CIMT values have never been studied in children, what if all kids have similar ranges and always have? How do the ranges seen among these high-risk children being evaluated for heart problems compare to other children?

Without a control group for comparison, no conclusions can be made that the CIMT ranges among these children suggests that fat children have higher or abnormal values.

Do CIMT measurements even matter? Is the thickness of the inside wall of the carotid artery even a measurement of atherosclerosis and plaque build-up? Or, is it just another risk factor (correlation) with no clinical value in predicting heart disease? Does a single measurement in time predict which children will go on to have heart disease as adults or do values normally fluctuate from year to year or even week to week?

No children were followed to see which ones went on to get heart disease. There was no evidence presented to suggest their CIMT measurements predict those who will.

And the most obvious question: Is all of this just a high-tech way to pile on scares about childhood obesity?

Balancing information we didn’t hear

No journalist reported of the controversy surrounding CIMT. None balanced their stories with reassuring information that many researchers have shown that CIMT isn’t a good measure of atherosclerosis or predictive of clinical outcomes and premature death.

Role among healthy people. Japanese researchers at Kugayama Hospital in Tokyo, for example, evaluated the relationship between plaque formation, age and CIMT measurements of 319 healthy adults who had no cardiac risk factors, no atherosclerotic disease, no diabetes, etc. These subjects ranged in age from 21 to 105. They found that CIMT naturally rose with age at every decade of life, including among the centenarians, with normal values of 0.009 X age +0.116. [Note, these values would have put the children this week’s abstract right in line with these healthy adults.] Most importantly, these Japanese researchers found the thickness of the inner arterial wall was not related to plaque formation and was distinct from any pathological plaque formation process.

Role among unhealthy people. The Rotterdam Study, also published in a 2001 issue of Stroke, the journal of the American Heart Association, specifically investigated the role of CIMT in predicting future coronary heart disease and cerebrovascular disease. They compared 374 older adults who had had a stroke or heart attack to nearly 1,500 healthy controls and followed them an average of 4.2 years. They found no value in adding CIMT as a risk factor and that CIMT added little predictive value as a screening tool for heart or cerebrovascular disease. Swedish researchers with the Angina Prognosis Study in Stockholm followed CIMT measures, lumen diameter and plaques in the carotid and femoral arteries of 809 older patients (about 60 years old) with stable angina for 3 years. They, too, found CIMT to be a weak predictor of those who went on to have cardiovascular events, heart attacks or die.

CIMT correlates with other risk factors. Given the single most important risk for death is age, and CIMT is most related to age, may help explain why CIMT is often found to add limited value to risk assessments among asymptomatic people.

Population weight studies. Nor has CIMT been shown to be related to body weight when examined among the general population in adults. As reported earlier this year in the August issue of Archives of Internal Medicine, German researchers at the University of Tubingen examining NHANES data from 1999-2004 found that even obese people with insulin sensitivity had CIMT values no different from normal weight people.

Role in treatment. Without exhaustively dissecting the science, what’s valuable for the public to know is that CIMT as a measure of clinical value, even when treating people with familial hypercholesterolemia and traditional coronary risk factors, has been questioned among medical professionals for years. In the current issue of Mayo Clinic Proceedings, Dr. J. David Spence, MBA, MD, FRCPC with the Stroke Prevention & Atherosclerosis Research Centre London in Ontario, Canada, wrote that “CIMT is not a good way to assess effects of therapy on atherosclerosis.” CIMT, plaque and stenosis are all biologically distinct. CIMT is a measure most closely related to age and is also genetically distinct, he wrote, whereas traditional coronary risk factors are more associated with plaque. Even aggressive treatment for atherosclerosis doesn’t show up with changes in CIMT, and the annual change (about 0.015mm/year) is about 20 times less than the resolution of the carotid ultrasonography (0.3mm) equipment.

Remember that all a risk factor means is that someone has found something associated with a disease. That doesn’t mean it means anything. There is nothing in this abstract that should leave young persons afraid or fearing that if they don’t have their CIMTs monitored and diet and lose weight, they’re going to drop dead of a heart attack 30 years before their time.

© 2008 Sandy Szwarc

Bookmark and Share