How Altitude Affects Running Pace

One of the all-time great topics for discussion on the Sunday long run is “What was so-and-so’s race performance really worth?” Andreas Almgren, for example, won this year’s New York Marathon by running the first half in 1:04:04 and the second half, including the hills of Central Park, in 1:01:57. What could he have run with a flatter course and more even pacing? Similarly, you can ponder the effects of factors like heat, wind, drafting, shoe design… and altitude.

Everyone understands that the thin air at high altitudes matters when you’re climbing Mount Everest, or even playing football in Denver. But the effects can also show up at much lower altitudes. I once spent a few months in Australia’s capital city, Canberra, which is only around 1,900 feet above sea level. I had a streak of disappointing road races—which only made sense once I realized that even at that altitude VO2 max is reduced by about five percent. Australian sports scientists had figured that out when they were setting up the Australian Institute of Sport in Canberra and noticed that their fancy new testing equipment seemed to produce abnormally low values.

This intrinsic worth of high-altitude performances is currently a hot topic because the NCAA voted over the summer to eliminate its altitude conversions for track and field qualifying standards. Until now, any performance at more than 3,000 feet above sea level was adjusted to account for the effects of less air resistance and lower oxygen levels. Sprint performances had time added, because thin air makes them faster; middle- and long-distance performances had time deducted, because the lack of oxygen made them slower.

The conversions were eliminated in part because, as the rules committee wrote, “there is no true science being applied” in formulating them. That makes a new study in the International Journal of Sports Physiology and Performance slightly late to the party—but it’s interesting nonetheless. Atle Guttormsen and Simen Guttormsen of the Norwegian University of Life Sciences analyzed more than 900,000 individual performances from track meets at 382 NCAA indoor track venues over a 14-year period to determine exactly how much difference altitude makes.

What the New Data Reveals

The statistical analysis attempted to account for as many sources of variation as possible: the design of the track (e.g. banked or unbanked), how a given performance compared to the athlete’s other performances that season, how early or late in the season it was, and so on. (Conveniently, there’s no wind indoors.) Overall, they came up with an estimate for how much altitude affected various events per 1,000 meters of altitude (which is 3,281 feet).

Here’s the resulting data for six running events:

Up to 400 meters, athletes are faster at altitude. The effect at 60 meters is slightly smaller because (I’m assuming) they spend a large fraction of the race accelerating to top speed. At 200 meters, they’re about 0.4 to 0.5 percent faster than they would be at sea level. In the long jump, triple jump, and pole vault, all of which are speed-based (but not shown above), there’s a similar advantage.

The crossover, where lack of oxygen begins to outweigh lower aerodynamic drag, happens at race distances just beyond 400 meters. The negative effects of altitude climb steeply at distances of up to 3,000 meters, then level off. That’s because in these longer events, aerodynamic drag is mostly irrelevant, so the effects of altitude are purely negative. For distances of 3,000 meters and beyond, the negative effects of altitude in this dataset seem to be about 2.7 percent for every 1,000 meters of altitude. Note that 1,000 meters is just slightly more than 3,000 feet, where the NCAA’s altitude conversions used to kick in. That equates to a roughly 30-second disadvantage for 5,000-meter runners.

The most interesting point in this graph is the 800-meter data. Women experience a greater disadvantage than men at altitude: 0.92 percent versus 0.50 percent. Top NCAA men are running about 1:45, while top women are running about 2:00. Since they’re so close to the point where oxygen and aerodynamics balance out, that 15-second difference matters. The men are going faster, so they benefit more from the reduction in drag; the women are running for longer, so they suffer more from the lack of oxygen. The Guttormsens argue that the NCAA needs sex-specific conversion factors—but the argument is moot since the conversion factors have been scrapped, at least for now.

What It Means for the Rest of Us

Assuming the oxygen-related penalty stays roughly constant at 2.7 percent beyond 3,000 meters, here’s what kind of slowdown you can expect in endurance events at various elevations:

The upshot: I can now tell myself that I was about 1.6 percent fitter than my half-marathon times in Canberra suggested. (I’ve been saying that for years anyway… but now I’ve got science!)

The problem is that altitude affects everyone differently. A few years ago, I wrote about a neat study showing that altitude-born South American cyclists perform relatively better in high-altitude race stages than European cyclists. Even though the Europeans were doing lots of altitude training camps, they couldn’t overcome the disadvantage of being born near sea level. Altitude response depends on a mix of genetics (where your ancestors lived), environment (where you live and/or train), and developmental factors (where you were born and grew up). This is another reason the NCAA cited in scrapping altitude conversion factors: the “right” conversion is different for everyone.

There’s even some evidence that trained endurance athletes are more susceptible to the negative effects of altitude than untrained people. That may be because their hearts pump blood so powerfully that it rushes past the lungs too quickly to pick up a full load of oxygen. This affects some athletes even at sea level, but becomes increasingly common as you climb in elevation.

Personally, Canberra isn’t the only history I’d like to revise: some of the most disappointing races of my life took place in Calgary (3,428 feet) and Edmonton (2,116 feet). These aren’t cities that anyone thinks of as “high-altitude,” but it’s clear that their elevation does have an effect—and that effect is bigger if you don’t realize that it exists, because in trying to stick with your expected pace you can burn yourself out early in the race. Knowing about these effects matters, and the new data gives you a starting point to make adjustments next time you head to the mountains—or even just the foothills.

For more Sweat Science, sign up for the email newsletter and check out my new book The Explorer’s Gene: Why We Seek Big Challenges, New Flavors, and the Blank Spots on the Map.