Running is the most elemental sport. The equipment is simple: shoes, socks, shorts, shirt. The activity is natural. We once ran after antelopes on the savannah, and we now run around playgrounds as kids. For the most part, we compete against ourselves. And because it’s so personal, and so elemental, the inevitable decline that comes with age can be wrenching.
Aging reduces our performance at everything athletic, but sometimes it’s hard to make out what’s happening. The ball doesn’t seem to go quite as far; the racket or the bat doesn’t swing quite as fast. The back starts to ache a bit. But the more complex a sport is, the more confounding factors—or excuses—there are.
In running, the evidence is right there: We ran the same distance and we went slower. What took 10 minutes now takes 11; what took three hours now takes three and a half. The evidence of the damage that time does to our bodies shows up implacably on our watches. I’ve been running for decades, and after almost every slow race I worry that I’ve stepped onto an escalator headed inexorably downward.
There isn’t, of course, a magic age at which runners start to get worse. Most of the men’s and women’s distance-running records were set by people in their mid-twenties. But those are the elites; they use their bodies hard, and their speed, when it goes, tends to go quickly.
Recreational runners, on the other hand, may slow more slowly. Science suggests that the best age to run may be 27 for men and 29 for women. The men’s world record in the marathon was just set by Eliud Kipchoge, age 33. A study of marathoners in Stockholm pegged the decline to age 34. Whatever the exact age, by the time you’re eligible to run for president, you’ve likely started to fade.
I’m quick but nowhere near elite. I ran well in high school and then badly in college. I bailed at 18 and didn’t really start again for a decade. I ran a 2:43 marathon at age 30, a moment when I was old enough to start declining but inexperienced enough to keep improving. For the next nine years these two forces—experience propelling me forward and age pushing me backward—stayed in balance. I ran marathon after marathon in roughly the same time, sometimes a bit faster, sometimes a bit slower.
But when I turned 39, I finally started to slow: Each marathon got worse. Last year, at age 42, I did a little better than the year before, but the trend line did not look good.
This year, I set a slightly fuzzy goal of two hours plus my age in minutes, a barrier I had never quite broken. I would turn 43 in the summer, which meant I wanted to run a fall marathon in 2:43. As I was beginning to plan, late in the spring, I got an unexpected note from Nike saying that they often connect non-elite runners to elite coaches. Did I want to run the Chicago Marathon under their tutelage in October? Of course I did. And that’s how I set out on a quest to understand the deeper science of running, aging, and performance.
I got on a conference call with three experts—Brett Kirby, a sports scientist with the calm demeanor and seeming wisdom of Obi-Wan; Stephen Finley, a big-hearted coach who nearly made the Olympic trials in the steeplechase; and Joe Holder, a physical trainer and former University of Pennsylvania wide receiver whose other clients include Naomi Campbell.
I told them how I’d worked out in the past and what my goals were. I wanted to get faster, but I didn’t want to spend any more time on the sport. I sent them a link to my old online logs and told them that I now tracked every run in Strava. They assured me that there was a way to beat the ravages of the years with science and a little bit of math.
The physiology of running can be broken down into three parts. There’s the body’s fitness: how fast you can get oxygen to the muscles and how fast you can go before lactate accumulates in the blood. Then there’s running economy: the efficiency with which you move. And then there’s mass: how much you weigh. Multiply fitness by running economy and divide by mass. That’s how fast you’ll go.
As a person ages, those variables don’t have to ineluctably get worse. We often gain weight, but we can lose it again. We pick up bad habits from injuries that change our form—say, a tweaked right ankle that makes us land too hard on our left. But a habit caused by an unconscious choice can usually be reversed by a conscious one.
Most important, as Kirby explained, our muscles change in ways that are both good and bad. As we train, over time, the mitochondria inside our muscle cells become more efficient at converting energy. New blood vessels develop. Tendons strengthen. On the other hand, our lean muscle mass declines with age, which is bad for marathoners and even worse for sprinters. Still, the decline doesn’t have to be steep.
The main reason that runners slow isn’t our bodies. It’s our lives. We get married, we have children, we work longer, our parents get sick. We have more important things to do with our time. Running is a sport that rewards consistent effort, and once you step away it’s hard to come back. Your body frays, which makes running less enjoyable, which accelerates the decline. We go slower as we age, but we also age when we start to go slower.
For about 13 years, my training routine has been roughly the same. I live four miles from work and, on weekdays, I usually run to the office and back home. (Yes, there’s a shower.) When I don’t have a marathon on the horizon, I’ll end up covering 30 to 40 miles a week. In the three months leading up to a marathon, I’ll do 20-mile runs on the weekend and speed up some of my commutes. Those weeks, I run closer to 50 or 60 miles.
According to Michael Joyner, a sports physiologist at the Mayo Clinic and a historian of running, there’s been an evolution in the way elite runners train. A century ago, the world’s fastest distance runner, Alfred Shrubb, just ran at a steady pace for less than an hour a day, three to five times a week. Gradually, people realized they could get faster by running longer and varying the pace. By the 1950s, the world’s best marathoner, Emil Zápotek, was running more than two hours a day and adding in interval training: workouts where you run a set distance (say a mile) at a faster-than-usual pace and then recover for a set amount of time (say, two minutes).
Now, the classical training program followed by elites includes interval runs, fast steady runs, long runs, and recovery runs. They run twice or sometimes three times a day, pushing their bodies right up to the red line where injury occurs. In all, they typically cover around 120 miles a week, which sounds like a lot but doesn’t actually take much time. An elite marathoner might spend 15 hours a week running, while a cyclist, swimmer, or cross-country skier might spend twice as much time training because they can do it without getting hurt. The easier a sport is on your joints, the more time elite training takes.
My new coaches listened to my description of my training regimen and told me it was fine—but far from optimal. The long runs I was doing were good. The total volume of training was OK. Ideally, I’d run many more miles a week, but that’s not a variable you can easily change without risking injury. The one variable I could truly improve was time spent running fast.
I wasn’t doing remotely enough work to improve one of the key metrics of running: VO2 max, a measure of the body’s ability to bring oxygen to the blood cells during intense exercise. Nor was I doing enough to improve my lactate threshold, a measure of the body’s ability to clear lactate from the blood. As Joyner put it, a runner’s VO2 max is equivalent to a car’s engine size, and his or her lactate threshold is the red line on the tachometer. I needed to improve both.
VO2 max improves mostly through speed workouts—running quarter miles, or miles, to the point of near exhaustion, resting briefly, and then running them again. Lactate threshold improves through what are called threshold runs: running hard at a pace that’s tiring but that doesn’t bring you to your knees. So, starting in early July, I began a new routine. I still commuted in by foot, but on Tuesdays I added focused runs to tax my VO2 max, and on Fridays I added threshold runs.
My new coaches wrote out plans for every workout in a Google Doc, and I reported how I did. Soon I was running mile repeats and 9-mile threshold runs before or after work. Within a month, something had started to change. I likely hadn’t run a mile faster than five minutes in a quarter century.
But in a track workout after coming home one day, I somehow churned out a 4:59. I felt bliss, convinced I had stumbled on a code that had magically advanced me to the next athletic level. Four days later, I botched a similar workout and remembered that getting faster is hard.
I wasn’t precisely following the prescribed program because life always intervened. I often had to run with a fanny pack to transport my wallet and keys, or a backpack to carry my clothes. My three sons supervised one track workout, which ended early when the 4-year-old justifiably got bored. Other runs involved stops at the dentist, the dry cleaner, and soccer practice. Red-eyes scuttled planned runs, as did sudden conference calls.
The beauty of running, though, is that it’s the simplest sport to fit into a day. Keep your sneakers nearby, and, when the opportunity arises, just get up and go.
The second big change in my routine came from data. I’d long believed that any data beyond the most basic was a distraction. The quantified self is often a neurotic soul. So for years I kept track of my runs with my stopwatch and counted up the number of miles I ran each week. But that was it. Kirby persuaded me that this was a habit to break.
Soon I was running with a Garmin Forerunner 935 on my wrist, an external heart-rate monitor attached to my arm, and little pods attached to my shoes and waistband that measured how balanced I was as I ran. My wife told me I looked like Voltron.
After a run, I’d sync the Garmin and study the data. I now had precise information on how much of my energy had gone to going forward and how much I’d wasted swaying side to side. I could see how my feet had pronated and whether my heart had been beating as fast while going the same speed as two weeks before. I learned about RunScribe power, which is a measure pioneered by cyclists of the amount of power generated by a runner at a given moment. Running a six-minute mile on a clear day and a flat course may take less power than running a 6:20 in the wind and over hills.
I also became fascinated with something that Garmin calls TRIMP, a metric that’s hard to learn about because Google autocomplete assumes you’re searching for the man in the Oval Office. The acronym stands for training impulse, a measurement determined by multiplying the length of an exercise by exertion, as measured by heart rate. I never changed a workout because of President TRIMP, but I felt good when the numbers were high.
In mid-August, I traveled to Portland, Oregon, where I finally met Kirby in person. I walked on a mat that measures pressure and impact and learned that I land heavily on my forefoot, transition heavily to my heel, and then off again from my forefoot. The middle of my foot, it seems, does nothing. I did a test of ankle flexibility and learned about computational apparel design. And then Kirby persuaded me to do something I’d both wanted and feared to do since high school: take a VO2 max test.
A runner’s VO2 max is determined in large part by genetics, and it’s frustratingly hard to change; hard workouts improve it, but within a limited band. The average man scores about 40; a good athlete will score 50. Lance Armstrong’s was famously 81.2. Kilian Jornet, the greatest mountain runner in the world, is said to hit 92.
Some great runners have relatively low scores: Frank Shorter, who won a gold medal in the marathon, tested only 71.3. But, for the most part, VO2 max is destiny. Test everyone in a high school and you’ll get a very good sense of who should run track. I feared that my score would be high, suggesting that I’d never lived up to my potential. And I feared, too, that my score would be low, suggesting that there was nowhere left to go.
The actual metric is the volume (V, measured in milliliters per minute) of oxygen (O2) that a person of a given mass can transmit through the bloodstream every minute. And the way to measure it is to put a runner on a treadmill, strap an oxygen mask to his face, and grind him into exhaustion.
I dutifully submitted and raced as quickly as I could, from a beginning pace of 7 minutes a mile to one of 5:10. Every three minutes, I’d stop and Kirby would prick my fingers and take some blood. When the test was over, he ran the numbers through his computer and gave me my score: 60, pretty much a Goldilocks number. Training could make that number go up, and I could get better. But I had also probably made a good choice when I quit the college cross-country team as a freshman.
Running is a simple sport, but in a city, at least, you really can’t run naked. You need a shirt that keeps you warm and dry, shorts with little pockets for your keys, and socks that keep your feet from blistering. Most important, you need shoes. Your feet constantly pound the pavement. With each stride, your shoe absorbs force from the crash, returns energy to your legs, and helps to shape your stride and the distribution of pressure through your body.
It didn’t escape me during this process that Nike might not have selected me for this experiment entirely randomly or even on the merits. I do happen to edit a magazine that covers athletic performance and gear. I’ve written about shoes. Executives from Nike have come to our conferences.
And so I tried to stay honest by remaining ecumenical: My main running club is sponsored by New Balance; in the middle of the summer I ran a mountain marathon sponsored by Adidas; I’ve long trained mainly in Asics Gel-Kayanos; I did some of my best early runs in On Cloudventures trainers. (While we’re at it, I’d like to give a shout-out to these Rothy sneakers, too.)
But Nike was running this particular project, and so I was going to run this particular race in their gear. In Portland, I sat down with the head of Nike’s sports research lab to talk about their philosophy of shoes and to learn more about the VaporFly 4%, a shoe they introduced to great fanfare last year, and in which I had run my last marathon faster than expected.
My colleague Ed Caesar has written thousands of words about the shoe, but the general idea is that it’s based on three innovations. The first involves weight. When a runner lifts her leg after hitting the ground, the motion resembles that of a pendulum, meaning that the mass of a shoe matters more than mass anywhere else on the body. An extra pound in your midsection will be less of a hindrance than 50 extra grams on your shoe. The VaporFlys weighed 190 grams each, which is roughly equivalent to wearing six pairs of socks.
The second innovation was the introduction of a new form of cushioning called Zoom foam, which is based on aircraft insulation. When your foot strikes the ground, the ideal shoe would absorb most of the pressure as it travels up your legs. Adding more foam increases cushioning, but it also increases weight.
This foam, according to Nike, is the most efficient one they’ve found, a claim backed up by Runner’s World lab tests. It’s not the prettiest: It smudges, for example, faster than you’d like. But if you’re just using the shoe in a race, who cares? When your face is contorted and drooling, it doesn’t much matter if there’s a mud streak on the side of your shoe.
The last innovation was the introduction of a carbon-fiber plate inside the sole that looks something like a spoon. The idea is that it bends on impact and offers back a slight propelling force. Critics call it illegal, which surely just makes demand for the shoes go up. Other shoe companies have introduced competitive marathon shoes this year that may perform just as well. But I was going to wear the VaporFlys again, whether coached by Nike or not.
I also ran in a special lightweight Aero Swift singlet designed for Kipchoge. A shirt seems simple, but it needs to perform a variety of functions. It has to help the body cool itself, by letting it sweat, and then letting the sweat evaporate. It needs to minimize the torso’s wind resistance. It needs to be light, and it really needs to not stick to your skin.
One of Kipchoge’s singlets had little spikes around it that, in wind tunnel tests, reduced the impact of wind. Mine had a big opening in the back to help with cooling. Relatively few people had worn this particular singlet, and it gave me a perverse thrill to know that I would be running the slowest marathon it had ever seen.
Meanwhile, a couple of times a week, I would fire up an app on my phone and actually do some non-running exercise and stretching. When the process started, I had headed into a gym to see Holder and had almost immediately started to feel like a man trying to eat a bowl of soup with a fork.
Holder was baffled that I didn’t know how to do a lunge and that I could barely do a squat. He laughed when I tried to touch my toes and reached just past my knees. I had spent 13 years training my cardiovascular system but had cared little for my joints, tendons, or ligaments.
Running breaks systems down and, over the years, I’d slightly injured almost every muscle group from the bottom of my feet (plantar fasciitis) to my belly button (lower abdomen strain). Stronger, more flexible muscles can absorb more stress. This training cycle, by stretching and planking more than ever, I stayed reasonably healthy. When my quads did start to ache after a day running fast down mountain trails, Holder showed me how to stretch muscles in my upper quads that I don’t think I’d ever stretched before.
His last instruction was to drink beet juice every day. Beets contain nitrates, which your body converts into nitric oxide. That seems to possibly increase blood flow and stamina. I ordered a case, stashed the little bottles in the fridge, and started popping one every morning. It’s unnerving to pee red, but at least I knew it was for the sake of science.
In a marathon, everything comes down to the day. Really the most important variable in running a fast one is the weather, especially the temperature. If it seems cold for walking around, say, 45 or 50, it’s probably perfect for running. Above 60, you’re in trouble. Above 70, you need to start seriously adjusting goals. Above 80, you should probably just stay home.
Wind can also be a huge factor, particularly in a race like New York, where most of the course travels north, and the wind seems to always blow to the south. Or Boston, where the whole course goes to the east; sometimes the wind is at your back, and sometimes it blasts in your face. In 2007, running into a headwind, the men’s winner clocked 2:14. In 2011, with the wind at his back, the winner ran 2:03. Even when the wind is light, according to Kirby, tucking in behind another runner can reduce the energy you expend at a given pace by about 1 percent.
The Chicago course is blessedly flat, but the forecast didn’t look particularly good as race day approached. Neurotically, I plugged my race goal, along with the expected temperature, humidity, and wind, into an online calculator a few days before; it told me to adjust my goals downward by four minutes. That doesn’t sound like much, but it meant that I’d miss the target I’d held in my head for months.
Kirby and Finley had a much more sophisticated model. The day before the race, I met them in a hotel in Chicago. Kirby pulled out a chart showing my overall fitness and level of fatigue, both of which he had extracted from my Garmin data using a formula that took into account my speed, my heart rate, and all the other factors I was logging on every run.
My fitness level had been climbing since early July, and it had leveled off the past 10 days while I had started to ease off my training in the blissful part of marathon prep called the Taper. My fatigue was starting to decline. Performance, he pointed out, is fitness minus fatigue. That number was at its peak.
Next, Kirby plugged my training data into a map of Chicago overlaid with the weather. It seemed to them that I should finish perhaps a bit under my six-year-old personal best of 2:39. They told me to tuck in behind runners when going north, to limit the wind, and to make sure to take in lots of carbohydrates and water as I ran. They would hand me bottles filled with a mixture of water and energy gels on the course; my kids would be on the side to do the same at mile 9.
Finley advised that I run the first half at around 1:19. Sometimes, the best thing about external validation is that it allows for internal validation. I could feel that the training had gone well. But without the data, and the experts examining the data, I would have been full of doubts about how I could do. I often sleep poorly the night before a race; that night, I drifted off with ease.
The morning was cold and windy, the start crowded as always, and I began much too slowly. I’m always cautious early in a marathon, and I went through the first mile in 6:35, a pace that would put me across the finish line in 2:52. By mile 4, I had picked it up and was clicking off miles at between 5:57 and 6:04. I ate when I needed to; I tucked in behind groups when I could. I went through the half in 1:19:30, which put me on pace for a 2:39.
Mile 20 is the spot at which marathoners traditionally start to fade. The body can hold roughly 2,000 calories of carbohydrates, and you use around 100 for every mile you run. Sure enough, at just about that point, I started to slow. I ran a 6:08 and worried that I was about to hit the wall. But I also knew that I had taken in carbohydrates along the way and that the reason many marathoners bonk at 20 is insufficient training, not glycogen depletion.
I ran mile 21 in 5:57, and then held that pace the rest of the way. My final time was 2:38:25, much faster than two hours plus my age in minutes. I celebrated in my usual way: bending over and shivering in pain. I high-fived my coaches, and Kirby took a sonogram of my thigh for a study on post-race muscle inflammation. I took the L to meet my kids, and then we zipped off to the airport.
Running faster always feels good, but running faster when you’re getting old feels good in a different way. When we’re young we wake up in the morning knowing that we’re probably stronger, taller, and faster than the day before. We want to be older because we want to be at our peak.
At some point, that desire fades, and then it flips. We wake up knowing we’re weaker and slower, and we have to work hard to stay in the same place as the day before. It’s all partly an illusion: We get a day older, whether we spend it running up a mountain or lying on a couch. Still, going faster means we’re doing something right.
I know, though, that it all has to end at some point. My joints will give out, my back will start to ache, or maybe reason will prevail. There are advantages to finding hobbies that don’t depend on exhaustion and obsession. Still, I don’t think I’m quite ready. One of the perversities of the sport is that, like gambling, it’s very hard to quit after you’ve done well.
Things went smoothly, and there’s always a way to think about how they could have gone better. I know that Father Time has an undefeated record. But for now I just need a new goal, a new plan, and new data. Once that’s set, I’ll make another spreadsheet, put on my shoes, socks, shorts, and shirt, and head out again.
Update: After publication of this article, Nicholas Thompson continued racing marathons. You can read about the rest of his journey here.
When you buy something using the retail links in our articles, we may earn a small affiliate commission. Read more about how this works.
- This gear helped me run faster marathons in my 40s
- Your poop is probably full of plastic
- We've been talking about self-driving car safety all wrong
- It was as an online gaming prank. Then it turned deadly
- PHOTOS: The badass motorbike taxi drivers of Nairobi
- Algorithms can be a tool for justice—if used the right way
- Get even more of our inside scoops with our weekly Backchannel newsletter
