HOW DID THE PROS STACK UP?

In addition to author Bill Gifford, we had three professional riders tested for fitness-related genes: Taylor Phinney, Jens Voigt, and Phil Gaimon. The testing was done by Helix and the results analysed by DNAFit, a UK-based fitness and diet-oriented genetics company. The results were interesting. Our author, for one, looks forward to out-sprinting Jensie…

I PROBABLY DIDN’T need to take a DNA test to know that I’m a middle-of-the-road athlete, but there it was: On a sliding scale titled ‘Your Aerobic Potential’, an arrow pointed decisively at the word ‘low’.

Really? Low? I was hoping for at least ‘above average’. I like to think that I’m better than at least some people, like those flatlanders I flew past on my local climb yesterday. There had to have been some mistake. Or at least missing information.

As direct-to-consumer genetic testing grows in popularity – more than 14 million Americans have had their DNA analysed by the two largest players, AncestryDNA and 23andMe – new tests are targeting amateur athletes. At last count, close to 40 companies offer some kind of gene testing for athletes, often packaged with training and nutrition advice based on the results. In the US, prices start around R750 a pop.

I had signed up with one of the best-known providers: a UK-based start-up called DNAFit, which offers a range of services, including an analysis of fitness-related genes, and one that looks at genes related to metabolism and diet. “We help you be the best you can be with unique genetic insights,” its website promised. A package had arrived in the mail, containing a clear plastic tube and instructions to fill it with my saliva. A few weeks after I mailed my spit in, the report from DNAFit landed in my inbox.

“We’d say you kind of have a low aerobic potential,” said Craig Pickering, a former British Olympic sprinter who is now DNAFit’s chief scientific officer. “You’re more likely to see small improvements and you’ll probably have to work much harder.”

I’m not the only one who learned something new about myself via genetic testing. Another rider I know, Phil, took a similar test, and it had reached a similar conclusion to mine. “Your genetic profile indicates you may not be as genetically well-suited to endurance sports as elite endurance athletes,” declared his report from coaching company Carmichael Training Systems.

Unlike me, Phil knew very well the results were laughably wrong. Phil – last name, Gaimon – was a pro for US domestic and international road teams, including Cannondale-Drapac (now EF Education First-Drapac p/b Cannondale), for a decade before leaving the pro peloton in 2016.

I started to wonder: are genetic tests for athletes worthwhile? Can they help us train more intelligently? And how much of a role do genes play in determining talent? Is there a ‘bike gene’?

Seeking answers, I subjected myself to tests by four different companies. I also persuaded

three current or former pro riders to be tested: Gaimon, retired German hardman Jens Voigt, and Taylor Phinney, an Olympian with Olympian parents. IF ANYONE HAS the ‘bike gene’, it’s probably Gaimon. For his first 19 years on Earth he showed no signs of anything elite, much less athletic. But when he started riding a bike, he soared from the entry-level Category 5 up to near-elite Category 2 in six months. Within two years, he had a US domestic pro contract with Jelly Belly, and was soon riding in breakaways in the Tour of California. Not too long after that, he was racing in Europe. “I didn’t use the word ‘genes’, but I sort of compared it to discovering a superpower,” he says.

A good number of us, it turns out, might be secret superheroes. In the 1990s, researchers at Louisiana’s Pennington Biomedical Research Centre and four other institutions recruited about 750 sedentary people to ride exercise bikes three times a week for 20 weeks. When they measured the subjects’ VO2 max (maximal oxygen uptake) at the end of the study, they found something striking: average oxygen-using capacity for the group had gone up with training, but there was an immense variation. Some had more than doubled their VO2 max, while others barely moved the needle.

Further investigation revealed that the key factor seemed to be heredity: siblings and immediate relatives were much more likely to have improved by roughly the same amount. The HERITAGE Family study, as it was called, pointed to a major role of genetics in determining one’s VO2 max, as well as what sports scientists call ‘trainability’, or how strongly individuals respond to training. The authors estimated that up to 47 per cent of an individual’s response

to their training programme was determined by their genes.

The next question was obvious: Which genes?

The human genome consists of between 20 000 and 25 000 genes. The genetic tests I took do not sequence the entire genome, an elaborate endeavour that can cost many thousands. Instead, they look for single nucleotide polymorphisms (SNPs), which are variations in a single DNA building block at a specific location on a given chromosome.

The variants we would look for are those that promote endurance and give us a high VO2 max, plus muscular power to let us wind up in a sprint, and the ability to recover from hard rides. But it turns out the genetics that underlie those traits are not so simple to identify.

One of the first sports-related genes to be discovered was called ACTN3, which popped up in a study of elite Australian sprinters circa 2003. It contributes to the development of fast-twitch muscle fibres. This ‘sprint gene’ is now probably the most-famous and most-studied gene in sports.

Another early example was a gene called ACE, which helps regulate blood pressure. People with one variation of the ACE gene tend to have a higher maximum heart rate and greater oxygen-carrying capacity, and tend to be better at endurance sports. A different version of ACE, studies suggested, supports strength and power.

The discovery of these ‘sports genes’ launched the direct-to-consumer genetic testing business for athletic talent. When they first appeared in the mid-2000s, many of these tests targeted parents who wanted to know if their children would grow up playing soccer, winning medals as track sprinters, or running marathons.

Since then, the field has become more sophisticated – as have the products. My report from DNAFit was a sleek 15-page PDF that began with ‘Understanding Genetics’ and continued on to the main dish: my ‘Power/Endurance’ score, based on a complex algorithm weighting my results from 15 SNPs on genes relevant to athletic performance.

According to DNAFit, my physiology was 70.6 per cent favourable to power-oriented sports such as weightlifting, sprinting, and track cycling, but only 29.4 per cent geared towards endurance pursuits such as long-distance running, mountain climbing, or road cycling. Hmm. I did win a Cat 5 track sprint that one time. But I hate weightlifting, and have the skinny arms to prove it. I love long, steady climbs on the bike, and hiking up mountains. I have gravitated toward these kinds of endurance activities since university – and now I find out I’ve been a secret sprinter all along?

The point of the test was not to determine my athletic destiny, said DNAFit’s Pickering. Rather, it was to help me train more effectively. “What we’re interested in is matching people to the type of training that works best for them,” he said. “In the gym, you’re probably going to respond better to high-intensity resistance training.” Indeed, I was soon receiving emails prodding me to sign up for one of DNAFit’s training programmes, like the 100-Day Muscle Builder, at about R900 per month.

But I still had one overarching question: Was any of this actually true? MY QUASI-SCIENTIFIC experiment – to get actual pro cyclists tested, and to compare my results from different services for consistency – aimed to answer this question. It didn’t take much persuading to get Voigt, Phinney and Gaimon on board. “I’m curious, because I’ve been told I’m like a freak of genetics my whole life,” said Phinney when I reached him at his training base in Girona, Spain. The son of Olympic road race gold medallist Connie Carpenter Phinney and Tour de France stage winner Davis Phinney, he seemed to have been bred for cycling glory. And indeed, he competed in the 2008 Olympics in Beijing at the age of just 18, finishing 7th in the individual pursuit.

Not surprisingly, DNAFit rated Phinney’s aerobic potential as ‘high’, and found that he was more or less equally balanced between power and endurance genes. Jens Voigt also scored well on endurance, which makes sense for a rider who seemed to be in every Tour de France breakaway ever, and also set a world hour record (since broken) in 2014, at the age of 42.

But there was still Gaimon’s mystifying result from Carmichael Training Systems, which said he had ‘typical’ endurance ability. This conflicts not only with his real-life result of making the ProTour, but also with his recorded VO2 max of 88.7, which is near the upper end of the testing range even for pro cyclists. (A fit amateur cyclist is likely to have a VO2 max in the 40s or 50s.) “I think the [DNA] test is missing something,” Gaimon noted drily.

Things got even stranger when I submitted my own DNA to the three other services: Athletigen, Orig3n, and Carmichael Training Systems.

DNAFit and Carmichael both use a third-party provider called Helix to do the gene sequencing, then apply their own interpretation to the raw data. Orig3n does its own sequencing. And Athletigen accepts data from third-party platforms such as 23andMe and AncestryDNA (I uploaded my data from 23andMe), and offers its own sequencing.

Not surprisingly, CTS and DNAFit agreed that my aerobic potential is nothing special – DNAFit rated it ‘low’, and CTS said it’s ‘normal’. Athletigen offered a little bit more hope. Things got brighter when I opened my report from Orig3n. It concluded that – contrary to my other results – I really was a genetically gifted endurance god. Strange. Were these companies looking at entirely different genes to draw their conclusions?

“They absolutely could be using different variants to test the same phenotype,” says Emily Spencer, a genetics researcher at the Scripps Research Institute in La Jolla, California, who has written critically about direct-to-consumer fitness genetic testing. “It doesn’t necessarily mean either is ‘wrong’; just that neither is the complete story.” SO WHAT DOES the research community think of genetic tests? In 2015 an international group of two dozen leading sports scientists and geneticists published a joint paper that began with this sentence: “The general consensus among sport and exercise genetics researchers is that genetic tests have no role to play in talent identification or the individualised prescription of training to maximise performance.”

It’s not that they disagree with the premise that athletic ability is largely influenced by heredity. The problem is that the science behind the genes being tested for is far from settled. “Scientists who work in this area will tell you very clearly that you cannot use any of this genomic stuff as predictive,” says Timothy Lightfoot, a professor of genetics at Texas A&M University.

It all boils down to the complexity of genetics – and the complexity of athletic talent. There are a few obvious traits that depend on single genes, such as the ‘coffee gene’, which impacts how quickly one metabolises caffeine. For more complex traits, however, the picture is less clear. “A phenotype like exercise endurance is incredibly complicated,” says Lightfoot. “As is trainability, as is strength.”

Even if you did get your entire genome sequenced, for R15 000 or so, we simply have not uncovered all of the genes that play a role in athletic ability, much less worked out how they might interact. The handful of sportsrelated genes that have been identified to date all have fairly small individual effects on performance. Even ACTN3, the ‘sprinter gene’, has been found to confer just 1 to 1.5 per cent of sprint speed performance among elite male athletes.

In the joint paper mentioned above, the authors wrote that both ACTN3 and ACE – the best-studied sports genes – have “zero” predictive value in determining athleticism. More importantly, genetics researchers say, we don’t even know which or how many genes are involved in the traits that contribute to sports performance. “They [genetic testing providers] are trying to tell you what the puzzle depicts, based on one or two pieces,” David Epstein, author of tells me. “They don’t even know how many pieces there are in the puzzle.”

Even fairly simple traits turn out to be mindbogglingly complex. Take height, which is easy to measure, and highly inheritable – scientists believe that about 80 per cent is genetically determined, if childhood malnutrition is not a factor. Still, a study of more than 250 000 subjects,

published in 2014, found that at least 697 gene variants determined the eventual height of a growing child. And even those genes, the authors estimated, accounted for only about 20 per cent of the variability in height.

In this light, the formula used by DNAFit – which it calls its ‘Peak Performance Algorithm’, and is based on just 15 genes that are each given a weighted score – does seem rather flimsy.

SPEAKING OF HEIGHT, let’s go back to Taylor Phinney. Even if he did inherit a monster engine from his parents, he also grew to be 1.96m. He could train for years and never be able to haul his 82kg-plus bulk over the Alps and the Pyrenees to win the Tour de France – although his size and power profile probably did help him finish eighth in last year’s Paris-Roubaix. So according to his test results, Phinney has a handful of favourable endurance genes, but he also has at least 697 other genes that made him grow so tall and muscular.

Either way, Phinney doesn’t spend much time dwelling on his cycling pedigree. He has learned that natural talent only takes you so far. “I quickly found out when I turned professional that everybody in the World Tour is extremely talented,” he says, “but whether that talent is a physical gift or more of a mental gift depends on the person. In my experience, the mind trumps the body, every time.”

He cites fellow rider and friend Mark Cavendish, who is widely known for winning sprint stages almost at will. Cavendish, he points out, has been open about the fact that he wins races in spite of past test results that indicated a mediocre (for an elite athlete) physiology. “They were almost taking him off the team because his [VO2 max and lactate threshold] test scores were so low,” Phinney says. “But mentally he just that he’s the best, and he’s the fastest, and he’s going to win.”

He sees the same characteristic in his father, Davis Phinney, who was also a fast sprinter who won two Tour de France stages and over 300 other bike races in his career: “I don’t think that my dad – physically – was that talented either, at least as a cyclist. I think he had this really deep, strong, innate confidence.”

That confidence just doesn’t register on genetic tests; nevertheless, it’s an important part of his athletic toolkit.

Clearly, Phinney wouldn’t have got anywhere close to where he did without a pretty stellar set of genes. We just don’t know what those are yet, exactly. And we may not, any time soon – as Lightfoot explained, finding the genetic recipe for elite-level performance has been a relatively low priority in genetics research.

But maybe that’s okay. Circling back to my own results, when I first started riding more than 30 years ago, little did I know my aerobic potential was ‘low’. Maybe it was. I’m not ever going to be a pro – that’s been sorted. But that didn’t stop me from riding and racing all these years, and it won’t stop me from heading out onto my favourite climb tomorrow, a 45-minute, switchbacking piece of singletrack called Jenni’s. And when I do, I will once again revel in the feeling that ride by ride, week by week, I’m getting better, stronger, and faster – regardless of my genetic make-up.