Few topics in endurance sport get used as an explanation as often as genetics. One person "just has it", another "isn't built for it". That is convenient, because it settles every uncomfortable question about volume, intensity distribution or sleep. It is also only half true. Genes do explain a substantial share of the differences between people — but they explain almost nothing about what happens to your training over the next two years.

What genes actually explain

Genetics sets the range, not the point within that range. Concretely: muscle fibre distribution is largely innate (roughly 45 to 50 percent heritable), maximal oxygen uptake has a heritability of about 40 to 50 percent, and even the response to an identical training programme varies genetically. What genes do not do: they do not dictate whether you run 60 or 90 kilometres this week, whether your easy sessions are genuinely easy, or whether you sleep eight hours.

The thinking error usually comes from mixing two questions. Question one: why is untrained person A faster than untrained person B? Genetics answers a large part of that — heart size, capillarisation, fibre types, body structure. Question two: how much better can I get in two years? Genetics answers that only at the margins. Training history, dosing and recovery decide here.

The number worth knowing: heritability of VO2max

Twin studies comparing untrained identical and fraternal pairs consistently arrive at a heritability share of VO2max of around 40 to 50 percent. Identical twins resemble each other more closely when untrained than fraternal twins — the classic finding. The remaining 50 to 60 percent is environment: daily movement, childhood sport, body composition, altitude, eating habits.

Two caveats belong to that number, because otherwise it gets misread. First, it applies to untrained people: once someone has trained for years, the share explained by baseline talent shrinks, because training shifts absolute capacity. Second, group heritability says nothing about an individual. From "50 percent heritable" it does not follow that "half of mine is fixed" — only that within the group of untrained people, heritable differences exist.

Trainability varies — but it is not sealed

The best-known study on this is the HERITAGE Family Study: more than 700 untrained adults completed a standardised programme five days a week for 20 weeks. At the end, the change in VO2max ranged from practically zero to over 40 percent. That finding is often cited as proof of "non-responders" — people for whom training supposedly does nothing.

The more accurate reading is less comfortable and more useful. Part of the spread is genuinely heritable, but a larger part is methodologically explainable: different baselines, imprecise measurement, different recovery, underestimated energy intake, sleep debt, stress. Someone who does not respond to a standard programme often responds to a different one — more volume, more intensity, different distribution, more recovery. In practice, "non-responder" usually means the wrong stimulus, not missing talent.

The famous genes: ACTN3, ACE and the rest

Anyone reading about sports genetics inevitably encounters three abbreviations. They are real, but their significance is dramatically overstated.

  • ACTN3 (the "speed gene"): The gene encodes alpha-actinin-3, a structural protein in fast-twitch muscle fibres. One variant (R577X) means roughly 18 percent of people of European descent produce no alpha-actinin-3 at all — with no apparent everyday disadvantage. In studies, sprinters are somewhat more often carriers of the functional variant, endurance athletes somewhat more often without it. The explained variance sits in the low single-digit percent range. There are world-class marathon runners without functional alpha-actinin-3.
  • ACE (I/D polymorphism): The enzyme is involved in blood pressure and fluid balance, so a link to endurance performance is biologically plausible. The evidence, however, is contradictory: meta-analyses find small effects that do not reproduce consistently across populations and sports. That is not enough for a training recommendation.
  • PPARGC1A and HIFA/EPAS1: Both are involved in mitochondrial biogenesis and oxygen transport. The best-known examples come from altitude research: Tibetan populations carry EPAS1 variants that favour oxygen handling at high altitude. That is a genuine, well-documented evolutionary finding — but one at population level after millennia, not information for your next training block.

The common denominator: none of these variants allows a prediction useful for individual training decisions. Effect sizes are small, studies are often small, and most findings have never been independently confirmed.

Why DNA tests are of little use to athletes

Direct-to-consumer sports DNA tests promise to derive a "training profile" from a saliva sample: power type or endurance type, optimal intensity, injury risk. The business model is understandable; the science behind it is thin. Two problems weigh heavily.

First: predictive power. Even if you add up all known performance-associated variants, the explained share of performance variance stays in the single-digit percent range. For a training plan that responds with weekly adjustments, that is not information — it is noise with marketing attached.

Second: the missing instruction. Suppose a test tells you "you are an endurance type". What follows from that for whether your interval session should be 4 × 1,000 m or 6 × 800 m, whether the long run belongs on Saturday or Sunday, or how much volume you tolerate in week six? Nothing. Training decisions need feedback from your body, not from a nucleotide sequence.

What is actually predictive can be measured and costs less: the trend in resting heart rate and heart rate variability, the ratio of load to recovery, time spent at target pace, the number of sessions missed. These are all signals from practice — and exactly the quantities an adaptive training engine uses.

What genetics really does influence

Genetics is not a myth; it just acts in different places than most people assume. Four areas where innate disposition genuinely shows up:

  • Body structure and levers: Height, leg length and the torso-to-leg ratio determine mechanical efficiency. Very long legs help in the marathon and tend to be a hindrance in uphill running and climbing on the bike. That is race-relevant information — but not information you can change. What you can do: choose the discipline where your build pays.
  • Muscle fibre distribution: The share of slow-twitch (type I) fibres sits between 40 and 60 percent in most people and above 80 percent in elite marathon runners. That distribution is largely innate and barely trainable in either direction — what is trainable is the capacity of each fibre. A 50-percent distribution can still produce world-class performance after ten years of training, if the other factors line up.
  • Heat tolerance and fluid balance: Sweat rate, sweat sodium concentration and thirst perception are partly inherited. That is why one person pours sweat at 25 °C and another barely glistens — and why not everyone needs the same fluid intake. Hydration for running explains how to determine your personal rate.
  • Tendon and ligament injury susceptibility: Variants in collagen genes (such as COL5A1) appear statistically more often in people with tendon problems. The effects are small, and they change nothing about the most important measure: controlled load progression with strength training as the foundation.

Notably, what genetics does not explain is the ability to train consistently for years. That trait is the strongest single predictor of endurance performance at amateur level — and it is almost entirely behaviour, not inheritance.

The role of age — and why it gets overstated

The second popular explanation next to genetics is age. The two are often confused because they justify the same thing: "at my age that doesn't work anymore." The data is clearer than the feeling. VO2max declines from around age 30 by roughly 0.5 to 1 percent per year — in untrained people. In people who maintain their training, the decline is markedly slower, in some longitudinal studies below 0.5 percent per year into the seventh decade.

Most of the typical age-related decline is therefore not biological fate but a consequence of falling training intensity. Anyone who stops running intervals at 55 loses exactly the capacity that slows the decline. Training over 50 shows how the intensity distribution shifts without the quality disappearing.

The practical conclusion: what you can control

If innate disposition sets the frame and training sets the position within the frame, the decisive question is not "what am I?" but "what do I do with what I have?" Four levers that make the difference between two runners with identical genetics larger than genetics itself:

  • Consistency over years: Ten years of five sessions a week beat two years of ten sessions and four injury breaks. At amateur level that effect is bigger than any genetic difference.
  • Correct intensity distribution: 80 percent easy, 20 percent hard — measured against your threshold, not against a generic heart rate. Anyone running every session "moderate" gives away adaptation at both ends. Details in Polarized training.
  • Load progression at a tolerable rate: Weekly increases of 5 to 10 percent are sustainable; jumps of 30 percent produce injuries. Increasing training volume lays out the system.
  • Recovery as part of training: Adaptation does not happen during the session but after it. Sleep, energy intake and rest days are not softeners but preconditions. Sleep and endurance performance gives the concrete numbers.

If you want the terms around load, adaptation and recovery clearly separated, the definitions are in the Peakora training glossary. And to see how these levers distribute across a whole season, the system is laid out in the marathon training overview.

Common mistakes around this topic

  • Talent as an excuse: "I'm just not a runner's build" is a self-description that devalues every training decision. It may hold for the world elite and not for your next half marathon.
  • DNA test instead of a training plan: A DNA test delivers no weekly schedule, no intensity zones and no adaptation logic. A training plan delivers all three.
  • Confusing heritability with determinism: 50 percent heritable means 50 percent not heritable — and the 50 percent you can influence is precisely what decides at amateur level.
  • Choosing the wrong role models: Professionals have a genetic setup that occurs once in several million people. Copying their plans one-to-one ignores exactly the part that differs in you: recovery time.
  • Not documenting the trend: Anyone who does not know how they trained a year ago cannot judge whether they are improving. A training log is the cheapest form of performance testing.

How Peakora handles individual variation

Adaptive training control does not replace a DNA test — it makes one superfluous, because it measures what matters: your response. Peakora works from your actual sessions, your load history and your feedback. The training stress score sums the load, CTL and ATL represent fitness and fatigue, and the weekly structure adapts when you miss sessions, run too fast or are too tired.

That is the fundamental difference to a static plan from the internet: the plan does not know you, your data does. Someone genetically mid-pack who trains consistently and correctly dosed for ten years will pass almost everyone who trained irregularly for three years with better genes.

Frequently asked questions about genetics in endurance sport

How much of endurance performance is genetic?

Roughly 40 to 50 percent of the VO2max differences between untrained people can be attributed to genetics in twin studies. The rest comes from training, daily activity, sleep and body weight. Important: that figure describes differences within an untrained group, not the limit of one individual after years of systematic training.

Can I do endurance sport with bad genes?

Yes. A genetic disadvantage shifts the upper limit, not the training response. Someone starting at 40 can double their endurance performance within two years regardless of starting point. Trainability itself is only about half heritable; the rest is dosing, consistency and recovery.

What does a DNA test tell me about athletic talent?

For training planning, practically nothing. Known variants such as ACTN3 or ACE together explain only a few percent of performance differences, and studies show no reproducible prediction of training success. A DNA test tells you neither which intensity helps you nor how much volume you tolerate.

Is trainability genetically fixed?

Partly. In the HERITAGE study, VO2max change after 20 weeks of identical training ranged from almost zero to over 40 percent. Some of that spread is heritable, but much of it is measurement error, sleep, nutrition, energy availability and training history. Someone labelled a low responder often simply had the wrong stimulus.

Is injury susceptibility inherited?

Some variants in collagen genes are statistically linked to tendon and ligament injuries. Predictive power for an individual is low, though. Training build-up, strength level, load spikes and recovery matter far more — all factors you can control.

Training control instead of the talent debate

Peakora adapts your weekly structure to your actual load and recovery — instead of to an innate disposition you cannot change anyway. Start free, no credit card required.

Start free