The Sports Gene: Inside the Science of Extraordinary Athletic Performance

by David Epstein

  In 1997, Lee’s group named GDF-8, a gene on chromosome two, and its protein “myostatin.” The Latin myo-, meaning muscle, and -statin, to halt. Something that myostatin does signals muscles to cease growing. They had discovered the genetic version of a muscle stop sign. In the absence of myostatin, muscle growth explodes. At least it did in the lab mice.

  Lee wondered whether the gene might have the same effect in other species. He contacted Dee Garrels, owner of the Lakeview Belgian Blue Ranch in Stockton, Missouri. Belgian Blue cattle are the result of post–World War II breeding that sought more meat to accommodate the increased demand of Europe’s postbellum economy. Breeders in Belgium crossed Friesian dairy cows with stocky Durham shorthorns and got cattle with heaps of muscle. Double muscle, to be precise. Belgian Blues look as if somebody unzipped them and tucked bowling balls inside their skin. “Hotline,” Garrels’s 2,500-pound prize Belgian Blue bull, once ripped a steel restraining gate off its hinges and flicked it aside en route to a cow in heat.

  Lee asked Garrels for blood samples from her double-muscled cattle. Sure enough, the Belgian Blues were missing eleven of the DNA base pairs—out of more than six thousand—from the myostatin gene. It left them without a stop sign for their muscles. Another breed of double-muscled cattle, Piedmontese, also had a genetic mutation that resulted in no functional myostatin.

  So Lee went hunting for human subjects. First stop: the grocery store, where he loaded his cart with muscle mags, the kind with cover photos of bulging-veined men in itty-bitty skivvies. Lee has jokingly been called “the skinniest man in the world” by a colleague, and he still remembers the sideways look from the cashier. Nonetheless, he placed an ad in Muscle and Fitness and was immediately swamped with willing volunteers, many of whom mailed him photos of themselves flexing and scantily clad, or not clad at all. He took samples from 150 muscular men, but found no myostatin mutants.

  He put the work aside until 2003, when Markus Schuelke called to talk about the bulging baby boy who was born at Charité hospital three years earlier and whose development he was monitoring. The following year, Schuelke, Lee, and a group of scientists published a paper that would introduce the world to the “Superbaby,” as the media would name him. The German boy, whose identity has been carefully guarded, was the human version of a Belgian Blue. Mutations on both of his myostatin genes left him with no detectable myostatin in his blood. Even more provocatively, Superbaby’s mother had one typical myostatin gene and one mutant myostatin gene, leaving her with more myostatin than her son but less than the average person. She was the only adult with a documented myostatin mutation, and she was a professional sprinter.

  •

  Double muscle might seem like an unconditional blessing, but myostatin exists for a reason. It is, in evolutionary terms, “highly conserved.” The gene serves the same function in mice, rats, pigs, fish, turkeys, chickens, cows, sheep, and people. This is probably because muscle is costly. Muscle requires calories and specifically protein to sustain it, and having massive muscles can be a massive problem for organisms—like ancestral humans—that don’t have steady access to the protein necessary to feed the organs. But that is a diminishing concern in modern society.

  In Superbaby’s case, doctors initially worried that the boy’s lack of myostatin might cause his heart to grow out of control. So far, though, no major health concerns have been reported in him or his mother.* Thus, it seems unlikely that an individual with a myostatin mutation would ever even think to get tested. The result is that nobody has any idea just how rare the myostatin mutation is, other than that most people (and animals) don’t have it. But the facts that the one boy with two of the rare myostatin gene variants has exceptional strength, and that his mother had exceptional speed, are no coincidence. Superbaby and his mother fall precisely in line with racing whippets.

  Since the late nineteenth century, speed-seeking whippet breeders unknowingly created dogs that, like Superbaby’s mother, have single myostatin mutations and that are lightning fast. In the highest level of whippet competition—racing grade A—where dogs hit a top speed of thirty-five miles per hour, more than 40 percent of the dogs have what is normally an exceedingly rare myostatin mutation. In racing grade B only about 14 percent have it. In grade C, it all but disappears.

  Even in racing grade A, the myostatin mutation is not a prerequisite, but it is clearly beneficial. The drawback in the whippet breeding system is that some dogs end up with too much muscle.

  Every whippet puppy inherits one copy of its myostatin gene from each parent. If two sprinter whippets—dogs that each have one copy of the myostatin mutation—have four puppies, this is the likely scenario: one puppy will have zero copies of the mutation and be normal; two puppies will have one copy of the mutation, like Superbaby’s mother, and be sprinters; the fourth puppy will have two copies of the mutation, like Superbaby, which make for a double-muscled “bully” whippet. Bully whippets are cartoonishly buff. A bully whippet looks like a shrink-wrapped rock pile stuck to a cuddly face. Bully whippets are too bulky to sprint, so breeders often put them down.

  The more scientists look, the more species they find that hold with the pattern of myostatin gene mutations and speed. In early 2010, two separate studies independently found that variations in the myostatin genes of Thoroughbred racehorses were powerful predictors of whether the horses were sprinters or distance runners. And horses with a so-called C version of the myostatin gene—a variant that results in less myostatin and more muscle—earned five and a half times more money in purses than did their counterparts that carried two T versions and were flush with myostatin.

  The scientists who discovered this have, not surprisingly, started genetic testing companies for Thoroughbred breeders.

  •

  As soon as Lee published his first mighty mice results in 1997, he was inundated with messages from parents of children with muscular dystrophy (no surprise), and also from athletes (surprise!) ready and willing to offer themselves for genetic experimentation. Some of the athletes hardly knew what they were talking about. They asked Lee where they could purchase myostatin, not realizing that it is the absence of myostatin that leads to muscle growth.

  Lee himself is a huge sports fan. He can recite the last forty-five NCAA basketball champions and he reverse-engineers the memory of a two-decades-old date with his wife by first thinking of who the St. Louis Cardinals’ pitcher was that day. But he has been reticent to talk with sportswriters about his work. He is troubled by the apparent willingness of athletes to abuse technology that isn’t even technology yet, and that is meant for patients with no other options. He hopes that any future myostatin-based treatments won’t be stigmatized the way steroids have been because of their role in sports scandals.

  The occasional keyhole view into the genetic cutting edge has proven understandably tantalizing for athletes. After myostatin, Lee moved on to mice in which he both blocked myostatin and altered another protein involved in muscle growth, follistatin. The result: quadruple muscle. In collaboration with researchers at the pharmaceutical company Wyeth, Lee then developed a molecule that was shown to bind to and inhibit myostatin and with just two injections increased mouse muscle 60 percent in two weeks. A subsequent trial by the pharmaceutical company Acceleron reported in 2012 that a single dose of that same molecule boosted muscle mass in postmenopausal women. Several companies now have myostatin inhibitor drugs in clinical trials.

  For pharmaceutical companies, this is the search not simply for a cure for muscle-wasting diseases, but for the all-time pharma pot of gold: a cure for the normal muscular decline of aging. And myostatin is not the only gene that has emerged in the quest for explosive muscle growth.

  The year after Lee’s mighty mice made headlines, H. Lee Sweeney, a physiology professor at the University of Pennsylvania, introduced the world to his own ripped rodents, made by injecting them with a transgene—a gene engineered in a lab—to produ
ce the muscle-building insulin-like growth factor, or IGF-1. Like Lee, Sweeney was flooded with calls. A high school wrestling coach and a high school football coach both offered up their teams as genetic guinea pigs. (Of course, the offers were rejected.)

  The gene-doping era may even already be here. In 2006, during the trial of German track coach Thomas Springstein on charges of providing performance-enhancing drugs to minors, evidence emerged that the coach had been seeking Repoxygen, an anemia drug that delivers a transgene that prompts the body to produce red blood cells.

  Before I traveled to the Beijing Olympics in 2008, a former world powerlifting champion gave me the name of a Chinese company that he said bodybuilders were using for gene therapy techniques. Once in China, I contacted the company, and a representative did respond to discuss potential genetic technologies. But I suspect it was just a strategy to tantalize patients, and that the company was not actually performing gene therapy.

  Still, Sweeney says that one method of delivering transgenes, simply pouring them into the bloodstream, is not necessarily safe but is simple enough that it could be accomplished by a sharp undergrad studying molecular biology. Sweeney has helped World Anti-Doping Agency officials prepare to fight gene doping, but if gene therapy is proven completely safe, he says, his impetus for keeping it out of sports disappears.*

  But perhaps the most interesting question is whether common DNA sequence variations in genes like IGF-1 and myostatin, as opposed to rare mutations, help to determine whether one gym-goer will pack on muscle faster than her spotting buddy. Comparisons of common variants of the human myostatin gene in both weight lifting and sedentary subjects have had less than eye-popping results. Some studies have found slight differences and some none at all. Other genes involved in the process of muscle building, though, are emerging as critically important to understanding why some people get sculpted when they pump iron while others struggle toward buffness in vain.

  •

  Muscles are pieces of meat made of millions of tightly packed threads, or fibers, each a few millimeters long and so thin as to be barely visible on the end of a needle. Along each fiber are a number of command centers, or myonuclei, that control muscle function in the area. Each command center presides over its fiber fiefdom.

  Outside of the fibers hover satellite cells. These are stem cells that wait quietly, until muscle is damaged—as happens when one lifts weights—and then they swoop in to patch and build the muscle, bigger and better.

  For the most part, as we gain strength we do not gain new muscle fibers but simply enlarge the ones we already have. As a fiber grows, each myonuclei command center governs a larger area, until the point when the fiber gets big enough that the command center needs backup. Satellite cells then form new command centers so the muscle can continue to grow. A series of studies in 2007 and 2008 at the University of Alabama–Birmingham’s Core Muscle Research Laboratory and the Veterans Affairs Medical Center in Birmingham showed that individual differences in gene and satellite cell activity are critical to differentiating how people respond to weight training.

  Sixty-six people of varying ages were put on a four-month strength training plan—squats, leg press, and leg lifts—all matched for effort level as a percentage of the maximum they could lift. (A typical set was eleven reps at 75 percent of the maximum that could be lifted for a single rep.) At the end of the training, the subjects fell rather neatly into three groups: those whose thigh muscle fibers grew 50 percent in size; those whose fibers grew 25 percent; and those who had no increase in muscle size at all.

  A range from 0 percent to 50 percent improvement, despite identical training. Sound familiar? Just like the HERITAGE Family Study, differences in trainability were immense, only this was strength as opposed to endurance training. Seventeen weight lifters were “extreme responders,” who added muscle furiously; thirty-two were moderate responders, who had decent gains; and seventeen were nonresponders, whose muscle fibers did not grow.*

  Even before the strength workouts began, the subjects who would ultimately make up the extreme muscle growth group had the most satellite cells in their quadriceps, waiting to be activated and build the muscle. Their default body settings were better primed to profit from weight lifting. (Incidentally, one possible reason steroids help athletes gain muscle rapidly is because the drugs prompt the body to make more satellite cells available for muscle growth.)

  Every similar strength-training study has reported a broad spectrum of responsiveness to iron pumping. In Miami’s GEAR study, the strength gains of 442 subjects in leg press and chest press ranged from under 50 percent to over 200 percent. A twelve-week study of 585 men and women, run by an international consortium of hospitals and universities, found that upper-arm strength gains ranged from zero to over 250 percent.

  The results evoke the American College of Sports Medicine’s new motto: “Exercise Is Medicine.” Just as areas of the genome have been identified that influence how well different people respond to coffee, Tylenol, or cholesterol drugs, so does every individual seem to have a physiologically personalized response to the medicine of any particular variety of training.

  The Birmingham researchers took a HERITAGE-like approach in their search for genes that might predict the high satellite cell folk, or high responders, from the low responders to a program of strength training. Just as the HERITAGE and GEAR studies found for endurance, the extreme responders to strength training stood out by the expression levels of certain genes.

  Muscle biopsies were taken from all subjects before the training started, after the first session, and after the last session. Certain genes were turned up or down similarly in all of the subjects who lifted weights, but others were turned up only in the responders. One of the genes that displayed much more activity in the extreme responders when they trained was IGF-IEa, which is related to the gene that H. Lee Sweeney used to make his Schwarzenegger mice. The other standouts were the MGF and myogenin genes, both involved in muscle function and growth.

  The activity levels of the MGF and myogenin genes were turned up in the high responders by 126 percent and 65 percent, respectively; in the moderate responders by 73 percent and 41 percent; and not at all in the people who had no muscle growth.

  •

  The network of genes that regulates muscle growth is only beginning to be delineated, but one biological cause of individual differences in strength building is already well-known. Some athletes have greater muscle growth potential than others because they start with a different allotment of muscle fibers.

  Coarsely speaking, muscle fibers come in two major types: slow-twitch (type I) and fast-twitch (type II). Fast-twitch fibers contract at least twice as quickly as slow-twitch fibers for explosive movements—the contraction speed of muscles has been shown to be a limiting factor of sprinting speed in humans—but they tire out very quickly.* Fast-twitch fibers also grow twice as much as slow-twitch fibers when exposed to weight training. So the more fast-twitch fibers in a muscle, the greater its growth potential.

  Most people have muscles comprising slightly more than half slow-twitch fibers. But the fiber type mixes of athletes fit their sport. The calf muscles of sprinters are 75 percent or more fast-twitch fibers. Athletes who race the half-mile, as I did, tend to have a mix in their calves closer to 50 percent slow-twitch and 50 percent fast-twitch, with higher fast-twitch proportions at the higher levels of competition. Long-distance runners are skewed toward the slow-twitch muscle fibers that can’t produce explosive force as quickly, but which tire very slowly. Frank Shorter, the last American man to win the Olympic marathon, was found to have 80 percent slow-twitch muscle fibers in a leg muscle that was sampled. It begs the question of whether the athletes get their unique muscle fiber combinations via training or whether they gravitate to and succeed in their sports because of how they’re already built.

  A vast body of evidence suggests that it is more of the latter. No train
ing study ever conducted has been able to produce a substantial switch of slow-twitch to fast-twitch fibers in humans, nor has eight hours a day of electrical stimulus to the muscle. (That caused a fiber type switch in mice, but failed to do so in people.) A 2010 review of muscle fiber type studies in the Scandinavian Journal of Medicine & Science in Sports had this answer in response to the question of whether significant fiber type switches can occur through training: “The short (disappointing) answer is, ‘Not really.’ The long answer has some uplifting nuances.”* Meaning that aerobic training can make fast-twitch fibers more endurant and strength training can make slow-twitch fibers stronger, but they don’t completely flip. (Save for extreme circumstances, like if one’s spinal cord is severed, in which case all fibers revert to fast-twitch.)

  Both gene and fiber type data suggest that innate qualities of each individual ensure that there is no one-size-fits-all sport or method of training. Some sports scientists have already put that notion to practical use.

  •

  With just 5.5 million residents, Denmark cannot afford to squander its top athletes. So Jesper Andersen makes sure that Danish athletes and coaches are thinking about muscle fiber type.

  Andersen was a national-level 400-meter runner and later coached the Danish national team sprinters. Now he is a physiologist at the world-renowned Institute of Sports Medicine Copenhagen. He works with elite athletes ranging from Olympic runners to soccer players on Denmark’s best team, F.C. Copenhagen, which competes in Europe’s Champions League. And he sees individualized responses to training programs every day.

  When Andersen took muscle biopsies of Danish shot-putters in 2003, he found that Joachim Olsen had a much higher proportion of fast-twitch fibers in his shoulders, quads, and triceps than the other top throwers. Andersen became convinced that Olsen had not nearly reached his muscle growth potential, given his high proportion of fast-twitch fibers. So he urged Olsen to stop weight training throughout the year and instead to focus on shorter periods of extremely heavy weight lifting, followed by periods of total rest with no weight lifting at all. Over the course of one season, Olsen’s muscle fibers ballooned—as confirmed by another biopsy—and the following summer he won the bronze medal at the 2004 Olympics in Athens. The feat propelled him to celebrity status in Denmark, and he subsequently won the Danish version of Dancing with the Stars and was elected to parliament.