In “Association of Body Mass Index With Lifetime Risk of Cardiovascular Disease and Compression of Morbidity,” by Sadiya Khan, Hongyan Ning, John T. Wilkins, Norrina Allen, Mercedes Carnethon, Jarett D. Berry, Ranya N. Sweis, Donald M. Lloyd-Jones argue that the claim moderately overweight people have greater longevity gives the wrong idea. For one thing, people who are already sick can lose weight for two different reasons: the direct effect of their illness, and doctors telling them they need to lose weight. When people talk about the effect of weight on illness, what they would normally be thinking of is the effect of one’s weight when seemingly healthy on later disease and mortality risk.

Also, some studies don’t look at survival to a given age, instead looking at survival from time of diagnosis. If risk of disease goes up for any reason, one is likely to get the disease at a younger age. Conditional on being diagnosed with a disease, one will probably survive more years beyond that time of diagnosis if one is younger at the time of diagnosis.

One way to address both of these problems is to find a set of apparently healthy people of the same age, some of normal weight and others in various overweight categories, and track what happens to them.

The simple answer from this study is that obese folks (body mass index 30 or above) tend to die quicker and get more of all kinds of diseases. Folks who are merely overweight (25.0-29.9 body mass index) die at about the same rate as those of normal weight (18.5 to 24.9 body mass index), but die a lot more from cardiovascular and heart disease and a lot more from other causes. Of course, we are all mortal, so if we don’t die of one thing we will die of another. But overall death rates at given ages are pretty similar for the overweight and those of normal weight.

I take this to mean that, leaving out the officially “obese,” the typical variation that makes some people of normal weight is helping those normal weight folks reduce their cardiovascular and heart disease risk a lot more than it is helping them reduce their cancer risk. The authors argue that reducing cardiovascular disease and heart disease risk is especially valuable, so there is a great benefit to whatever it is that makes people of normal weight. But I would argue that we need to understand where the extra cancer risk is coming from.

We don’t even have conclusive science on what makes some people of normal weight while others are overweight; there is a big debate about whether dietary sugar or dietary fat is a bigger culprit. (I blame dietary sugar.) And we don’t know why whatever makes people of normal weight instead of overweight contributes to cancer. But here is one hypothesis that needs to be closely examined: what if those who consume less sugar and other refined carbohydrates consume more protein? My reading suggests that both sugar and too much protein (especially animal protein) are cancer risks. If you reduce sugar and increase protein, what happens depends on the relative strength of the effect of sugar and the effect of protein. Among the health conscious, I think protein has much too positive a reputation at this point in history.

Another possibility is that the combination of having a metabolism that allows one to eat a lot without gaining weight and actually eating that food is a cancer risk. This seems a real possibility to me. A cancer cell with a high metabolism and plenty of available food might grow fast.

In all of this, it is crucial to notice my emphasis on the variation in the population that causes most of the difference between being (moderately) overweight and of “normal” weight. People who lose weight by regular fasting are probably not well represented in population data sets because not many people in the population have been serious fasters in the past. (I hope more people use fasting as a weight-loss and health tool in the future.) So losing weight by fasting could have a dramatically different effect on cancer mortality at each age than, say, cutting back on sugar while increasing animal protein consumption.

Don’t forget that normal weight seems protective against cardiovascular disease and heart disease. That is a benefit. But unless you are obese to begin with, you probably need to lose weight the right way to reduce cancer risk.

Posts on Anti-Cancer Eating:

• How Fasting Can Starve Cancer Cells, While Leaving Normal Cells Unharmed

• Why You Should Worry about Cancer Promotion by Diet as Much as You Worry about Cancer Initiation by Carcinogens

• Good News! Cancer Cells are Metabolically Handicapped

• How Sugar, Too Much Protein, Inflammation and Injury Could Drive Epigenetic Cellular Evolution Toward Cancer

• Meat Is Amazingly Nutritious—But Is It Amazingly Nutritious for Cancer Cells, Too?

• My Annual Anti-Cancer Fast

• Cancer Cells Love Sugar; That’s How PET Scans for Cancer Work

For annotated links to other posts on diet and health, see:

• Miles Kimball on Diet and Health: A Reader's Guide