Eat More, or Starve Yourself

According to a recent L.A. Times Science article, there’s new evidence that a calorie-deprived diet, with 10% to 30% fewer calories than the usual recommended intake, leads to better health and longer life. The evidence comes from a study of rhesus monkeys, though researchers believe (based on accumulating evidence) that the effect is probably common to all primates, including humans. The conclusion? “Even those who maintain a healthy weight probably should be eating less.”

But wait! Turns out they buried the lead, waiting until page 2 to reveal this:

Are they happy? Are they hungry? Can they think as fast?

When UCLA evolutionary biologist Jay Phelan put mice on caloric restriction, he got the distinct impression that they didn't appreciate it.

"They bit people and were more agitated," he said. In contrast, the mice who ate a normal diet "would just sit around and let you pick them up."

Okay, maybe that’s not the lead, because Phelan only studied mice; the article does not report on the monkeys’ happiness.

Still, this is the important issue, is it not? If eating less increases your lifespan but decreases your happiness, then there’s a very real trade-off between quantity and quality of life. Which means the author’s conclusion that even people with a healthy weight “should probably be eating less” is simply unjustified -- unless we make some fairly heroic assumptions about what’s loaded into that key word “should.”

Setting aside the quality-quantity trade-off and assuming health and longevity are the only worthwhile goals, how much should you be eating? It’s beginning to look like the relationship between calorie intake and health is aggressively non-linear. And by “aggressively,” I mean it’s not even a simple quadratic (U-shaped) relationship.

Clearly, starvation is not optimal. Nor is morbid obesity. Between these suboptimal endpoints, the usual assumption is that you’re best off with a BMI in the middle of the recommended 18.5-24.9 range. But two factors argue against this conclusion. First, the monkey study adds to a growing body of evidence that people with a lower BMI live longer. (This isn’t immediately clear, since the article doesn’t say anything about the BMI under caloric restriction, but I assume it’s lower than what you get with regular calorie intake.) Second, there’s also good evidence that people with a BMI higher than the recommended range – between 25 and 30 – also live longer.

Putting it all together, the relationship between BMI and longevity (setting aside a variety of other factors, of course) seems to be something like this: From starvation up to a BMI somewhere on the low side of normal, there’s a positive relationship. From there into the middle of the normal range, there’s a negative relationship. Then the relationship goes positive into what is typically regarded as the overweight range. And then it goes negative again, all the way to morbid obesity. Thus, there seem to be two peaks on either side of the alleged normal range of BMI, and a local minimum in the middle of the normal range. Yet that’s the range regularly recommended to the public!

Anyway, I’m far from an expert on this. Maybe the weirdness goes away when you separate the effects of different kinds of weight, such as fat vs. muscle. BMI is a very blunt instrument for measuring health, after all. I just wish that articles about health and weight would spend more time discussing questions like these, instead of continuing to push the simplistic “Americans are too fat and need to slim down” line.