The first significant finding of this study was an unexpected difference in PCr recovery halftime (and hence mitochondrial function) between climbers and trekkers at baseline, with the climbers having significantly shorter halftimes (and hence better mitochondrial function) than their altitude-naïve counterparts. The climbers also had significantly higher inorganic phosphate concentrations than trekkers (the possible significance of this difference will be discussed below). These results were particularly surprising considering the fact that none of the climbers had been to high altitude for at least 5 months. There was a significant difference in age between the groups, but this would be expected to result in the opposite effect, at least on mitochondrial function (the climbers were older, and mitochondrial function generally declines with age 
). Nor did the climbers engage in any structured physical training prior to the expedition, being aware that differences between individuals in baseline cardiorespiratory fitness are unrelated to hypoxia tolerance (see 
and references therein).
Several possible explanations suggest themselves. First, the climbers might have been different to the trekkers because years of climbing had selected against particular phenotypes (for example, those genetically ill-suited to high altitude hypoxia are unlikely to continually expose themselves to it). Second, altitude exposure might induce stable changes in phenotype, perhaps through epigenetic modifications. Finally, it may be that repeated hypoxic exposure causes more conventional physiological adaptations that are unusually persistent. There are no published data on the timecourse of ‘de-acclimatisation’ after hypoxic exposure. The most relevant information therefore probably comes from experiments examining the timecourse of detraining after stopping prolonged exercise training. Coyle et al. observed that subjects who had previously been well-trained but who had stopped training altogether for 84 days still had, on average, a 17% higher
max than controls who had never trained at all 
. This was due to persistent peripheral adaptations (larger mixed-arterial/venous O2
difference, and therefore better O2
extraction by skeletal muscle). So despite the fact that none of the climbers had climbed for at least 5 months prior to this study, it seems possible that the difference in mitochondrial function observed was a long-term (or even stable and transmissible) effect of exposure to high-altitude hypoxia. More work investigating these differences is clearly warranted, as are comparisons with high-altitude natives and the offspring of successful climbers.
Exposure to hypobaric hypoxia is associated with an involuntary loss of body mass 
, whether under laboratory conditions 
or in the field 
. In women, nitrogen balance is negative soon after exposure to 4300 m altitude 
and remained negative in men throughout a 7,102 m ascent 
. In keeping with these observations, we observed a significant reduction in muscle cross-sectional area after hypoxic exposure (). There was no difference in the degree of atrophy between the groups (no statistically significant effect of trekker/climber grouping).
Such weight loss does not seem to relate simply to excessive metabolic demands related to exertion. Indeed, physical activity levels (PAL, assessed as maximal exertional metabolic rate as a multiple of basal metabolic rate (BMR)) are normally 2.2–2.5 at sea level, and twice that in trained athletes. However, near Everest's summit, PAL is limited to 2.0–2.7, meaning that exertional energy loss is minimized 
. Thus, although perceived exertion is great, actual energy expenditure is much less 
. The notion that weight loss is not purely due to excessive metabolic demands is given further credence by the observation that obese subjects exercising three times each week in 15% oxygen lose more weight than those exercising in air 
A variety of factors are thought to contribute to altitude-induced weight loss. First, energy expenditure may rise at altitude, due in part to an increase in BMR 
. This effect may be altitude-dependent: BMR has risen by 6% in men at 3,650 m 
, by 10% at 3,800 m 
, and by 27–28% in men and women by day 2–3 at 4,300 m 
. In the absence of calorie supplementation, such absolute increases in BMR do not seem sustained 
, although BMR per unit mass may actually remain elevated 
Second, pro-inflammatory cytokines may play a role. In eight sea-level residents, the interleukin-6 (IL-6) response to 60 min of bicycle ergometer exercise was found to be greater during exposure to acute hypoxia (4100 m altitude) than that seen in normoxia 
. After 6 weeks of exposure to 4100 m, IL6 levels remain elevated 
, a finding in keeping with similar observations over four days of exposure to 4350 m in males 
and with 12 days of exposure to 4300 m in women 
. Interleukin-6 is implicated in the pathogenesis of cancer-associated weight loss, driving lipid catabolism and muscle protein catabolism 
, perhaps through both lysosomal (cathepsin) and non-lysosomal (proteasome) pathways 
There was no decline in volume-scaled mitochondrial function (PCrt1/2
) after hypoxic exposure (). Taken in the context of significant atrophy, this means that whole muscle aerobic capacity was reduced. Yet, rather surprisingly in the face of significant muscle atrophy and a loss of aerobic capacity, the expedition did not have any adverse effects on muscle function during exercise. Subjects were able to complete the same exercise tasks pre and post exposure, and exercising metabolites were unchanged (). Although at variance with previous reports that muscle mitochondrial enzyme activities (per unit of cross sectional area) are decreased by hypoxic exposure 
, our results suggest that in vivo
function might somehow be maintained. Further experiments specifically targeted at illuminating changes in muscle mitochondrial function in vivo
in response to hypoxic exposure are required.
When fully recovered after a period of exercise, PCr concentrations are often higher than pre-exercise values, a phenomenon known as PCr ‘overshoot’ 
. There is very little published literature regarding the mechanisms underlying PCr overshoot in skeletal muscle. One hypothesis states that PCr overshoot is the result of a slow decay in one of the signals that directly activates oxidative phosphorylation 
. If this were the case, then the data here suggest a tightening of off-exercise oxygen kinetics, perhaps to prevent unnecessary oxygen consumption. An alternative explanation would be that inorganic phosphate is being lost during exercise (perhaps due to calcium-phosphate precipitation 
There were several changes in resting muscle high-energy phosphates after hypoxic exposure. While the changes in estimated free [ADP] will be discussed below, the increase in muscle phosphate is noteworthy because it mirrors a difference that was observed between the climbers and trekkers at baseline. There are two possible mechanisms for an increased steady-state cell [Pi]. The first is an increased Na+
-dependent Pi uptake. This mechanism is poorly understood, but can be driven by an increase in insulin, possibly indirectly via effects on the Na+
. The second theoretical possibility is reduced permeability to Pi efflux, although convincing examples are currently lacking.
Our calculations of [ADP] rest on a number of assumptions regarding muscle metabolite (ATP and creatine) concentrations. Although these assumptions are generally sound 
, one cannot discount the possibility that the extreme conditions experienced by our subjects invalidated them. Thus our findings need to be interpreted with caution. PCr concentration was calculated from the PCr/ATP ratio (assuming an intramuscular [ATP] of 8.2 mM L−1
) and was unchanged by the expedition, strongly suggesting that [ATP] was also unchanged. However, we did not directly measure creatine and the observed increase in calculated [ADP] could be accounted for by an increase in total creatine. Despite these reservations, it seems reasonable that [ADP] might have decreased in response to hypoxia. It is now widely accepted that mitochondrial oxidative rate is matched to ATP demand by feedback mediated by intracellular phosphorylation potential or some function of it. Therefore a reduction in the resting ADP concentration would indicate either a change in the control parameters linking oxidative rate to [ADP] or a reduction in resting muscle oxidative rate.
Resting (but not exercising) pH was significantly lower following exposure. This was not a result of systemic ketoacidosis, as there was no correlation between pH and β-hydroxybutyrate (correlation not shown). We therefore suggest that it was the result of either increased metabolic proton production or decreased capacity for cellular proton extrusion (for example, a reduced activity or sensitivity of Na+/H+-ATPase).
There are a number of limitations to this study. First, the trekkers and climbers were different from each other before the study began and we have chosen to treat the subjects as a single group (as well as separately). We justify this based on an absence of any statistical evidence that the groups responded differently. Second, several of the climbers were very slow descending from altitude before revisiting Oxford. However, this has provided an unexpected benefit: because there were no differences between the responses of trekkers (who returned to Oxford immediately) and climbers, it is unlikely that the observed changes were acute responses. For example, this shows that the post-exposure reduction in muscle pH (equally present in both groups) was not due to an acute disturbance in acid-base status.
We used magnetic resonance spectroscopy and imaging to study the effects of a trip to high altitude (Mount Everest) on a mixed cohort of altitude-naïve trekkers and experienced climbers. The climbers had unexpectedly better mitochondrial function than the trekkers at baseline. Both groups responded similarly to the hypoxic insult. Climbers had higher resting [Pi] than trekkers before the expedition and resting [Pi] was raised across both groups on their return. There was significant muscle atrophy post-CXE, yet exercising metabolites were unchanged. These results suggest that, in response to high altitude hypoxia, skeletal muscle function is maintained in humans, despite significant atrophy.