Skip to content
NatureHQ

Ecologyadaptation

High-altitude adaptation

The air at altitude is still 21 per cent oxygen. It is the pressure that has changed — and Tibetans and Andeans solved that in two different ways.

Air at altitude is not short of oxygen as a fraction — it is thinner, so each breath delivers less. Human highland populations have solved this in more than one way, and none of their solutions is what a visitor’s body does over three weeks.

A common misunderstanding is worth clearing first: the air at 4,000 metres is still 21 per cent oxygen. What has changed is the pressure, so each breath contains fewer molecules of everything, oxygen included, and the partial pressure driving oxygen into the blood is lower. That is why the problem cannot be solved by breathing a different mixture of the same thin air. A lowlander who arrives at altitude acclimatises, and the process is well described: breathing rate rises within hours, and over two or three weeks the body produces more red blood cells, so more oxygen is carried per unit of blood. This works, up to a point, and it reverses within weeks of descending. It is also, importantly, not free — more red cells means thicker blood, and thicker blood is harder to pump, which over years produces the chronic mountain sickness seen in some long-term highland residents. Now the interesting part. Tibetan highlanders do not do this. Their haemoglobin concentrations are close to lowland values, and they cope by breathing faster at rest, maintaining much higher nitric oxide levels and correspondingly greater blood flow. Andean highlanders do the opposite: markedly elevated haemoglobin, closer to an extreme version of the acclimatisation response. Two populations, the same altitude, two different answers — which is a useful demonstration that adaptation does not converge on one inevitable solution. The Tibetan case has been traced to a variant near EPAS1, a gene governing the low-oxygen response, and that variant turns out to have come from Denisovans: inherited from another human lineage and then strongly selected for. The direction of its effect is the memorable part. Selection favoured damping the standard response, because over a lifetime the standard response costs more than it delivers.

Developed coverage · 50% complete · reviewed 2026-09-03

What this page covers

Documented in humans on the Tibetan plateau, in the Andes and in Ethiopia; in birds including several waterfowl and passerines; and in mammals such as deer mice, yaks and vicuñas. The human cases are the best resolved, because the genomes are.

Often confused with: Acclimatisation, which is what a visitor’s body does over weeks and is reversible; Fitness or training, which does not change the oxygen available; Cold adaptation, which is a separate problem that happens to occur in the same places

Quick facts

What actually changes
Pressure, not the percentage of oxygen
Two populations, two answers
Andeans raise haemoglobin; Tibetans do not, and breathe faster instead
Where the Tibetan variant came from
Denisovans — inherited from another human lineage, then selected
Which way it works
It damps the usual response, because extra red cells cost more than they give

Where this appears

Every organism below has been linked to this page because the evidence links them. Each one carries its own evidence, and its own limits.

What a visitor does, and what a resident population has done

The clearest place on this site to see the difference between the two words.

Tibetan and Andean highlanders solved the same problem differently. Andeans carry much more haemoglobin; Tibetans do not, and instead breathe faster with far higher nitric oxide and greater blood flow.

Well supported

Good evidence backs this, though some details remain open.

Long-resident Tibetan and Andean high-altitude populations display distinct oxygen-transport phenotypes: Andean highlanders show elevated haemoglobin concentration with normal ventilation, while Tibetans maintain lowland-typical haemoglobin with elevated resting ventilation, elevated exhaled nitric oxide and increased peripheral blood flow.

Who this applies to
Two long-resident human highland populations, compared at similar altitudes.Do not extend this beyond the taxa listed — the popular version over-reaches.
Studied in
Homo sapiens
Why we rate it this way, and what the caveats are
Well supportedModerate confidence

The physiological differences are measured and consistent. Attributing them to genetic adaptation rather than to lifelong developmental exposure is well supported for the Tibetan case by genomic evidence, and less directly for the Andean.

How far it can be extended

The point of the comparison is that the two populations differ; generalising either phenotype to highlanders as a group is precisely the error it corrects.

Caveats

  • Population comparisons cannot fully separate inherited difference from a lifetime of developmental exposure to altitude.
  • Both populations vary internally, and the contrast describes averages rather than individuals.

Still unanswered

  • Whether the difference reflects the length of time each population has been at altitude, or different starting variation, or both.

Last reviewed 2026-09-03

The evidence (2 studies)
Three responses to the same thin air
AspectVisiting lowlanderAndean highlanderTibetan highlander
TimescaleHours to weeksGenerationsGenerations, and more of them
HaemoglobinRises substantiallyHighClose to lowland values
Resting ventilationRisesNear lowland levelsElevated
Blood flow and nitric oxideLittle changeLittle changeMarkedly elevated
ReversibleYes, on descendingNoNo
Long-term costNot applicableRisk of chronic mountain sicknessLower

The bottom two rows contain the point. The visitor’s response reverses, because it is something their body is doing rather than something their population has become. And the visitor’s response is the same one the Andean population has, in a heritable and stronger form — while the Tibetan population has gone a different way entirely, and appears to pay less for it.

A variant inherited from another kind of human

Adaptation in the strict sense, traced to its source.

The Tibetan altitude variant came from Denisovans, inherited from another human lineage and then selected for. It damps the usual low-oxygen response rather than amplifying it.

Well supported

Good evidence backs this, though some details remain open.

The EPAS1 haplotype at high frequency in Tibetans closely matches Denisovan sequence and is essentially absent elsewhere, indicating archaic introgression followed by strong positive selection. It is associated with attenuated erythropoietic response to hypoxia rather than an enhanced one.

Who this applies to
Tibetan highland populations specifically.
Studied in
Homo sapiens
Why we rate it this way, and what the caveats are
Well supportedHigh confidence

The frequency difference is among the largest recorded between closely related human populations, and the archaic match is unusually clear. The causal path from variant to phenotype is association rather than experiment, which is the remaining gap.

Caveats

  • The available Denisovan genome is a single individual from a site far from Tibet, so the archaic source population is inferred.
  • How the variant produces the physiological effect is not established mechanistically.

Still unanswered

  • Why the Andean populations did not arrive at a comparable damping solution, given the same problem and a long residence.

Last reviewed 2026-09-03

The evidence (3 studies)

This is the co-option theme of the whole sprint, appearing at the level of populations rather than molecules. The antifreeze protein of Antarctic fish was a digestive enzyme first. The desiccation tolerance of resurrection plants was a seed programme first. And the Tibetan altitude variant was Denisovan first — it did not arise in the population that now carries it, but was acquired from another human lineage and then favoured. Useful things are more often recruited than built, and this is as clear a demonstration of that as the sprint contains.

  • Why did Andean populations not arrive at a comparable damping solution?

    Why it matters: The same problem, thousands of years, and a different answer. Whether the difference reflects time at altitude, the variation each population started with, or the availability of an archaic variant to one and not the other, would say a good deal about how repeatable adaptation is.

    What would settle it: Comparable genomic scans in Andean populations with the same resolution, plus better dating of each population’s residence at altitude.

Other animals at altitude

The same problem, and mostly the same answer: improve every step a little.

There is no single thing that lets a bar-headed goose fly in thin air. Every link in the chain that moves oxygen from air to muscle is improved a little, and the effect is cumulative.

Well supported

Good evidence backs this, though some details remain open.

High-altitude flight capacity in Anser indicus derives from coordinated modification along the entire oxygen cascade: higher haemoglobin–oxygen affinity, elevated ventilatory and cardiac capacity, greater lung diffusing capacity, and flight muscle with increased capillary density and mitochondria redistributed closer to the sarcolemma.

Who this applies to
The bar-headed goose in detail, with comparable patterns in other high-altitude birds.
Studied in
Anser indicus, Aves
Why we rate it this way, and what the caveats are
Well supportedModerate confidence

Each component is measured; that the whole cascade rather than any one link is what matters is a well-argued synthesis rather than a single experimental result.

How far it can be extended

Several of the same modifications have been found independently in unrelated high-altitude bird species.

Caveats

  • Much of the comparative physiology comes from captive birds in hypoxic chambers rather than from birds in flight.
  • Which links limit performance in free flight is not resolved.

Still unanswered

  • Whether any single link is limiting, or whether the cascade is balanced so that no one step dominates.

Last reviewed 2026-09-03

The evidence (2 studies)

Related

The research behind this page

6 studies, newest first. Each one has a page explaining what it found and what it could not show.

2014Nature

Altitude adaptation in Tibetans caused by introgression of Denisovan-like DNA

The Tibetan EPAS1 haplotype closely matches the Denisovan sequence and is essentially absent from other modern populations, indicating it entered modern humans by interbreeding with an archaic group and was then favoured by selection.

2011Proceedings of the National Academy of Sciences

The trans-Himalayan flights of bar-headed geese

The geese flew mostly at night and in the early morning, tracked the terrain closely rather than cruising at a constant high altitude, used passes and valleys, and spent very little time above 6,000 metres.

2011Journal of Experimental Biology

Elevated performance: the unique physiology of birds that fly at high altitudes

High-altitude flight rests on a chain of modest improvements at every step of the oxygen pathway rather than on one dramatic feature — a lung that already outperforms a mammal’s, haemoglobin with higher oxygen affinity, greater cardiac and ventilatory capacity, and muscle with denser capillaries and mitochondria positioned closer to them.

2010Science

Sequencing of 50 human exomes reveals adaptation to high altitude

A variant near EPAS1, a gene regulating the response to low oxygen, showed one of the largest frequency differences ever recorded between closely related human populations, and is associated with the lower haemoglobin characteristic of Tibetans.

2007Proceedings of the National Academy of Sciences

Two routes to functional adaptation: Tibetan and Andean high-altitude natives

The two populations solve the same problem differently.

1982Paleobiology

Exaptation — a missing term in the science of form

Current utility and evolutionary origin are separate questions, and a large fraction of useful traits were co-opted rather than built for the job they now do.

This page is a stop on a longer route

A guided journey reads several subjects in a deliberate order, with an argument for why one follows another. You can join in the middle.

Where to go from here

Each of these follows from something on this page — a relationship in the evidence, a claim people ask about, or the next mechanism along.

How complete this page is, and what it is still missing

NatureHQ publishes its own gaps. This page is at 50% completeness against what we would call a finished subject, and was last reviewed on 2026-09-03. It carries 4 claims and answers 7 mapped search questions.

  • no research from the last few years is attached — check for newer work
  • more experiments could be explained in plain English
  • no popular claim about this subject has been checked yet
  • Ethiopian highlanders, who appear to have a third solution again, are mentioned only in passing.
  • Non-human mammals at altitude — deer mice especially, where the experimental work is strongest — are not covered.
  • How the EPAS1 variant produces its physiological effect is not established and is stated as such.