Brought to you by the letter U
The u- or j-shaped
curve is a feature of a lot of research studies on health-related topics. “Moderation in all
things,” is as true now as it was when Hesiod and/or Aristotle doled
out the advice thousands of years ago. It’s also a key to some of
the trickiness of health advice – you can get too little or
too much of a thing. People like to latch onto particular
foods or behaviors as being healthy or unhealthy, but none of them
exists in isolation. I mean, sure, kale is packed with all kinds
of minerals, but if you only eat kale, the results aren’t going to
be pretty.
Here’s another
maxim from the ancients: “The dose makes the poison,” attributed
to Paracelus, a Swiss alchemist (b. 1494-ish, d. 1541). That can
apply to everything from Tylenol to red wine, both of which will, in
sufficient quantities, utterly destroy your liver, but which are
pretty okay in small doses. Fish, likewise, is good for you (Omega 3 fatty acids are important), but
because it also often contains some heavy metals (not so good for you) you should go easy on it.
Then there’s exercise. None is very bad, and any amount more is better, until you start to get to the point
where you’re utterly wearing yourself out.
One of the first
places I ran across this relationship was when I was wondering about
optimal weight. According to the stock advice these days, normal
weight is defined by the 18.5
to 24.9 range of BMI (Body Mass Index, more on this in another post).
However, “Normal” does not match the range of the BMIs with the
lowest all-cause mortality. “… many epidemiological studies show
an inverse relationship between mortality and BMI inside the normal
BMI range. Other studies show that the lowest mortality in the entire
range of BMI is obtained in the overweight range (25–29.9).” This
paper has a long title: “How
Much Should We Weigh for a Long and Healthy Life Span? The Need to
Reconcile Caloric Restriction versus Longevity with Body Mass Index
versus Mortality Data.” It
also has a nice little chart which
shows lowest mortality at a BMI of 23.4, but no significant
difference from 20.8 to 27.7.
The u- or j-shaped
curve also shows up when you look at the results of exercise, eating red meat
or protein, drinking alcohol, or eating sugar, though in those later
two cases it’s unclear that any amount is really good
for people. Sleep is another problematic example, which merits a blog post of its own.
For
almost everything, there’s going to be a balance. Fortunately,
we’re adaptable creatures, and that balance point is very broad without a gigantic drop-off for most things. Don’t get too stressed out if
you’re not hitting the bullseye. Being more or less on target
really is good enough. Still, it helps to have some idea of what the targets are, and that's what I'll get into in a little while.
Comments