---

Digest news about food
with caution

People often complain that nutrition recommendations change from one week to the next. Much of this confusion is caused by reporters and readers not understanding a few key terms.

These are "correlation," "associated with," "risk" and "cause and effect."

"Correlation" implies that as one thing goes up, another thing simultaneously goes up (positive correlation) or goes down (negative correlation). This is not "cause and effect."

For example: In the 1970s it was shown that as the number of TV sets in a house increased, the incidence of dying from heart attack increased for men living in the house.

The correlation between TV sets and heart disease was positive. But it would be naive to suggest that TV sets cause heart attacks. Having many TV sets was likely associated with having a high-stress job, smoking, working long hours, being sedentary and not eating well. It is fairly obvious that these factors were more likely causing the heart attacks, not the TVs.

Not all correlations are so obviously unrelated to "cause and effect." For example, eating lots of fruits and vegetables high in beta-carotene is correlated with a decreased risk of certain cancers.

Giving people beta-carotene supplements, however, does not predictably decrease cancer incidence and can even increase some cancers. Therefore, the cancer protection of fruits and vegetables might be due to other factors.

It could be that people who eat more fruits and vegetables eat less high-fat foods and/or less total calories. Obviously, a great number of studies are needed to sort out these types of associations.

Most of the reports about foods and health are from correlation studies. Rather than showing cause and effect, they do demonstrate the concept of "risk."

The risk of heart disease increases with increased blood cholesterol. However, some people with low blood cholesterol develop heart disease, and some with high blood cholesterol do not. Still, when cholesterol is high, the overall statistical chance (risk) of dying from heart disease is greater.

THE RESULTS of these correlation studies frequently lead to research with animals and humans designed to sort out "cause-and-effect" relationships. Cause-and-effect studies on humans are very costly, often starting at $500,000 for a single study. But, even with these carefully controlled studies, it is virtually impossible to control every aspect of the research participants' lives. Therefore, it can be extremely difficult to sort out absolute cause-and-effect relationships.

So how does a consumer of health information figure out what to believe? Here are a few tips we use:

1. Don't put too much stock in any single study or any single group of researchers.

2. Wait for a body of other studies to confirm or refute results.

3. Question everything you read.

4. Pick your experts carefully.

Health professionals avoid basing their recommendations on only one study, even if it appears to be well designed. However, someone selling a product or a book based on a single study often will throw all their eggs in that one basket to support their belief and promote their moneymaking scheme. Buyer beware!

---