This is general safety information, not medical advice. Consult a doctor before travelling to altitude, before starting a new training programme, or before taking any medication. In an emergency, call the local emergency number — 112 anywhere in the EU, 911 in the US and Canada.

Most mountain training advice is confident and unsourced. This article reports what peer-reviewed work and alpine-club guidance establish, and names the popular claims with no evidence behind them — an unsupported claim is a finding, not a gap. It is not a programme, and deliberately does not become one.

A trail grade describes the ground, not the load on you

On an officially "easy" trail, measured oxygen uptake averaged "1339 ± 479 ml/min (17.9 ± 5.5 ml/kg/min)" — "41.50 ± 14.28%" of peak on average, with a maximum of "74.05 ± 16.21%". The spread is the finding: "82.3 ± 16.5% of the trail was conducted at low intensity", yet "17% completed the entire trail at low intensity, while two subjects spent more time at high intensity than at moderate or low intensity." The same paper puts a figure on the work of going up: 300 m per hour to 3,500 m "could require an altitude-adjusted relative VO2-hike of 18–22 ml/min/kg."

The grade on the signpost is a property of the path; the intensity is a property of you.

Sources: PMC — Br Med Bull 2025.

The aerobic base, and what it is actually protecting

The UIAA Medical Commission is unambiguous that fitness is the entry condition: "A relatively high level of fitness is a prerequisite for safe mountaineering", and endurance training beforehand "may be advisable for physically inactive men over 40."

The protection is large, and what earns it matters. Austrian case-control data found hikers "who performed mountain sports activities for more than 2 weeks per year benefited from a 77% risk reduction" in sudden cardiac death (0.23; 0.1–0.4); among skiers, high-intensity exercise more than once a week cut risk by "83%" (0.17; 0.04–0.74). That is habitual exposure across a year, not a training block bought in the weeks before a trip.

Sources: UIAA MedCom Consensus No. 21 — Cardiovascular Diseases (PDF); PMC — Burtscher et al., Int J Environ Res Public Health 2021.

Load carriage: where the weight sits, and the rule with no source

Knapik's review gives the principle and the number that explains footwear: "locating the load center of mass as close as possible to the body center of mass results in the lowest energy cost… each kilogram added to the foot increases energy expenditure 7% to 10%; each kilogram added to the thigh, 4%." The injuries are named — "foot blisters, stress fractures, back strains, metatarsalgia, rucksack palsy, and knee pain" — and hip belts "should be used whenever possible."

A 2022 physiology review frames the limit as intensity rather than mass: "load and speed should be carefully managed to maintain an exercise intensity ~45% V̇O2 max", with "47% V̇O2 max as a threshold above which intensity increases at a greater rate", and "gradient has a larger effect than load mass in determining the cardiovascular response". Cold compounds it — at −10 °C, "a 20% increase in V̇O2… compared to measurements taken at 20°C."

A rule that failed verification. The familiar pack-weight rule stated as a percentage of body weight has no admissible source we could find — not an alpine club, not a standards body, not the literature — so we will not print a figure nobody stands behind. The sourced framing is the one above: manage load, pace and gradient so the effort stays sustainable, and weigh your boots as carefully as your pack.

Sources: Europe PMC — Knapik et al., Mil Med 2004; PMC — Physiol Rep 2022.

Training the descent is a safety intervention, not a performance one

This is the section that matters most. In the Austrian Alps, "About 75% of all falls happened during the descent" — the statistic our turnaround guide is built on, and the reason the second half of a day deserves preparation.

A 2023 study supplies the mechanism. After 30 minutes of downhill walking on a 20° decline at natural pace and low perceived effort, "Leg dexterity (maximal force score) decreased significantly… (t(18) = 5.5, p < 0.001, d = 1.3)", and anterior–posterior balance fell too (d = 0.6) — while "Knee extensor isometric strength did not change post-exercise (p = 0.511)." Strength intact, foot placement degraded: eccentric contractions "disrupt the muscle proprioception pathways necessary for… dynamic foot–ground interactions." You feel fine, and are measurably less able to put your foot where you intended. The authors conclude that hiking safety "even on easy paths—might be improved by… stepping more carefully during descents."

The second finding is a surprise about dose. A 40-minute downhill bout at "-28%" gradient and "5 km/h" carrying "10% body mass" cost controls "-19.2 ± 6.9%" of maximal voluntary contraction torque against "-9.9 ± 9.6%" in a group that had walked downhill for five minutes a week earlier; peak creatine kinase "635.5 ± 306.0 U/L" against "339.3 ± 148.4 U/L"; soreness "81.4 ± 14.8 mm" against "27.8 ± 16.8 mm". The authors: "performing small volume of downhill walking is crucial in preparation for trekking."

That is one study, reported as one. But it argues what most training advice omits: descent exposure belongs in preparation at all, at a dose measured in minutes rather than weekends.

Sources: German Journal of Sports Medicine — Faulhaber et al., 2020; PMC — Rausch et al., 2024; PMC — Int J Environ Res Public Health 2023; PMC — Maeo et al., PLoS ONE 2017.

What sea-level training cannot buy you

Fitness is worth having for every reason above. It is not altitude protection. A study measuring VO2max and hypoxic ventilatory response before travel to 4,500 m found "no correlation between AMS scores and HVR… or with VO2max". Among 38 men ascending passively to 3,450 m, endurance-trained against untrained ("V̇O2max 66 ± 6 vs 45 ± 7 mL·min·kg"), "On day 1 at altitude, ET presented with a higher AMS incidence (42% vs 11%; P < 0.05)" — with no group difference on days 2 and 3. Exertion during the ascent accelerates onset too: an active group "expended more energy (1,265 ± 351 vs 408 ± 208 kcal)", saturated lower ("75 ± 3 vs 82 ± 3%"), and peaked at "81%" AMS on day one.

State it precisely, though, because the pooled evidence is weaker than those two studies. A meta-analysis of 58 studies and 2,705 participants (pooled AMS prevalence "48.25%") lists altitude above 5,500 m (RR 1.89), rapid ascent under 6 h (RR 1.60), an SpO2 decline over 10% (RR 2.02), prior AMS (RR 1.36) and male sex (RR 1.15) — and does not list fitness, training status or VO2max. Consistent that fitness does not protect; mixed on whether it briefly raises risk on day one.

Pre-acclimatisation works, at a dose almost nobody does. "Spending ~75 hours in hypoxia… reduces the risk of AMS by roughly 50%", and "~200 hours… may reduce the subsequent AMS risk to nearly zero", exposure duration accounting for "76% of the variance". Short protocols do nothing: seven hours in total produced a "12%" reduction, and "Seven passive 1-h hypoxia exposures do not prevent AMS in susceptible individuals." For anyone with cardiovascular disease the UIAA is blunter — "nothing is better than acclimatising slowly in the mountain environment."

Sources: Europe PMC — Eur Respir J 1991; Europe PMC — Sareban et al., Med Sci Sports Exerc 2020; Europe PMC — J Appl Physiol 2025; PMC — AMS meta-analysis, PLoS ONE 2026; PMC — Burtscher et al., J Travel Med 2026; UIAA MedCom Consensus No. 21 (PDF).

Where the popular advice runs out of evidence

Static stretching is the clearest case. A systematic review found "SS and PNF stretching had no clear effect on all-cause or overuse injuries", with immediate performance costs of −3.7% and −4.4%, worse at "≥60 s (-4.6%)" than "<60 s (-1.1%)", and range-of-motion gains "typically last[ing] <30 min". The recommendation is a warm-up "that includes additional poststretching dynamic activity."

For the protocols people actually follow — stair-machine sessions, "hike with a weighted pack N weeks before" — we found no study evaluating them. The honest word is untested, not disproven: absence of evidence is a reason to hold a claim loosely, not to reverse it. The nearest evidence is useful anyway — the descent study supports pre-exposure at a far smaller dose than folk advice prescribes.

Sources: Europe PMC — Behm et al., Appl Physiol Nutr Metab 2016; PMC — Maeo et al., PLoS ONE 2017.

The preparation that happens in a doctor's office

What follows is a reason to talk to a doctor. It is not a screening you can run on yourself, and none of it is medical advice.

The Austrian data explain why official bodies raise the subject: "50% of all deaths during downhill skiing and mountain hiking are SCDs"; "More than 90% of all SCDs… affected males aged over 34"; "about 50% of all SCDs recorded occurred on the first day." Risk concentrates in identifiable history — prior myocardial infarction carried a "10.9 (3.8–30.9) times higher adjusted SCD risk" in hikers, diabetes "7.4 (1.6–34.3)", hypercholesterolaemia "3.4 (2.2–5.2)".

The screening position quoted in that literature is written for clinicians, not walkers: in subjects "35 years and older… cardiovascular evaluation is recommended and should include as a minimum family history, symptoms, physical examination, and 12-lead resting ECG". An attribution note: we have that wording as quoted in the papers we could read, not from the source guideline — so treat it as reported, and as something a doctor decides, never a checklist you work through yourself.

One recommendation in that literature sits outside the clinic: "Rest or only slight physical activity on the first one or two days in the mountains and a gradual increase in activity on the following days." If you are inactive, mid-life, or carry any condition named above, the highest-value preparation is a conversation with a doctor.

None of this is a programme and we are not going to write one. The evidence says something narrower: the grade on the sign is not your intensity, the descent degrades your foot placement before it touches your strength, sea-level fitness does not buy altitude tolerance, and for some walkers the most consequential session is an appointment they have not made.

Sources: PMC — Burtscher et al., Int J Environ Res Public Health 2021; PMC — Br Med Bull 2025.

Sources