Tuesday, January 28, 2014

What's the best way to predict heart attack risk?

By Alex Hutchinson

There's power in numbers. Swedish researchers just published a pretty neat study in which they followed a staggering 743,498 men for an average of 34 years. The basic analysis, published in the European Heart Journal, was very simple: the men were assessed at the age of 18 when they were conscripted into the military, with data starting between 1969 and 1984. Then their ultimate fate was assessed by checking national registers to see which of them had heart attacks. Here's what they found:

It's worth taking a minute to look carefully at this graph to figure out what it's saying. The vertical axis shows relative risk of heart attack. The shading of each bar indicates the BMI range, from less than 18 to above 30 (which is the clinical definition of "obese"). And each set of four bars is clustered according to the quartile of aerobic fitness, as measured at 18: the four bars on the left of the graph were those with the highest fitness, and those on the right had the lowest.
The major trends aren't surprising: as far as heart attack risk goes, it's better to be more fit rather than less fit; and it's better to have a lower BMI than a higher BMI. The interesting comparison comes when you compare across groups. Is it better to be fit but overweight, or "normal" weight but not very fit? This question has been the topic of debate for the last few decades, with interesting results on both sides but a bit more momentum (as far as I can tell) on the side of those arguing that fitness is more important than fatness. In this case, though, the data say exactly the opposite. It's much better to have terrible fitness and low BMI (relative risk ~2.7) than to have great fitness and BMI above 30 (relative risk ~4.6).
Well, let's be a little more specific: it was better, for long-term heart attack risk, for these Swedish men to be lean when they were 18 years old in the 70s and early 80s. Does the same balance hold true for a currently middle-aged woman in 2014? Very hard to know, and we should be careful not to overgeneralize. In a sense, the fat-versus-fit debate reminds me a bit of the nature-versus-nurture debate, in that both elements are clearly important: you can't ignore either, and the relative importance of each one will depend on the specific context.
As a final note, check out this other graph from the same study:

The set-up of this graph is exactly the same, except instead of aerobic fitness it uses knee extension strength as the measure of physical health. In this case, BMI is way more important that knee strength, which barely seems to matter at all. This is a good check to show that the aerobic fitness results weren't just showing that generally healthy people live longer, because the pattern doesn't show up in knee strength and several other fitness measures. There really is something special about aerobic fitness.

No comments: