Recovery is acutely hampered with changes in altitude, after which it improves through acclimatisation
Similar to sleep HR, we found that average sleep HRV also changed dramatically during transitions. We observed that average sleep HRV decreased from ~45 ms on August 24th to ~31 ms on August 27th. Sleep HRV continued to decrease to ~24 ms on August 28th and then increased sharply over the next 2-3 days (Figure 1b). Before the first transition, sleep HRV at sea-level ranged from ~34 to ~48 ms, whereas sleep HRV in the mountains ranged from ~23 to ~36 ms during end-August. Unlike sleep HR, sleep HRV did not show much of an acclimatisation effect upon returning to sea-level from altitude in the period monitored.
Resting HR and HRV are well-established markers that indirectly assess physiological recovery, with higher HR and lower HRV likely suggesting physiological impairment or heightened stress6,7. We measured these metrics along with sleep movements for four days before and after altitude transitions in both directions. We found that average sleep HR increased substantially from ~57 bpm on August 24th (the day before leaving sea-level) to ~65 bpm on August 27th (the day after arriving at the mountains). Average sleep HR continued to increase to ~68 bpm until August 29th, after which we observed a sharp decrease to ~60 bpm on August 30th (Figure 1a). The athlete’s usual HR at sea-level ranged from ~52 to ~59 bpm, which increased to ~60 to ~68 bpm during the high altitude stay. We also observed an opposite effect during the second transition, where sleep HR remained high, around 59 bpm on September 9th (the first day after returning from the mountains) but quickly dropped to ~56 bpm the next day, post acclimatisation (Figure 1a). This trend suggests that the transition from lower to higher altitude places greater strain on the athlete’s physiology than the reverse, primarily due to the high altitude adaptation mechanisms in play.
Finally, we found that sleep movements, a proxy for sleep disturbances, were impacted acutely but also improved with acclimatisation. From August 24th to August 27th, we observed that sleep movements increased from ~1 to ~3, suggesting more disturbed sleep. This was corroborated by the athlete as settling down was uneven at the base camp on the first day, but may have been due to carbohydrate loading on the next two days. However, by August 28th, sleep movements declined to ~1 and remained constant thereafter (Figure 1c). Upon return to sea-level, we found that sleep movements increased substantially from ~4 to ~12 until September 10th, after which they decreased sharply to ~1 by September 12th (Figure 1c).
Figure 1: Graphical representation of mean sleep HR, HRV, and movements for four nights before and after both altitude transitions.
Long-distance running at higher altitude is accompanied by elevated HR and lower glucose levels
To evaluate the impact of high altitude on a comparable physical exercise stint, we compared running HR over time across two 18 km runs - one done at sea-level and the other in the mountains. We found that HR at high altitude was substantially higher than at sea-level for the first 60 minutes of the session. This might indicate higher cardiovascular strain to compensate for reduced oxygen density and is supported by previous studies8. We also found greater variability in running HR in the mountains, which ranged from ~70 to ~170 bpm, compared to running HR at sea-level, which ranged from ~85 to ~125 bpm (Figure 2a). Despite this, we observed no statistically significant differences in overall average running HR between both runs across the entire 18 km span.
We also observed that during the mountain run, median glucose levels, as measured by the Ultrahuman M1 sensor, were lower than those of the sea-level run. As shown in Figure 2b, glucose levels were lower than those of the sea-level run for the entire duration, with a diminished difference towards the end of the two runs. This is perhaps due to increased glucose disposal with high altitude exercise, as demonstrated in previous research9,10. Here, we found a trend-level statistical difference in overall glucose levels (p~0.1).
Figure 2: Graphical representation of granular running HR and glucose level over time across both 18 km runs. Running HR and glucose levels have been resampled from original rates of 5 minutes and 15 minutes respectively.
Tracking HR across steep changes in altitude within the ultramarathon period
We found that the athlete’s running HR during their 72 km ultra-marathon showed marked variability and was acutely influenced by both altitude changes and rest periods. As altitude increased from ~4000 to ~5370 m (summit) during the 0 to 35 km stretch, we observed a general decrease in average running HR from ~110 to ~60 bpm. During this period, we also observed two key rest points: a relatively long rest between 25 and 30 km and a food break around 30 to 35 km. Both rest periods had a stabilising effecton running HR, with the food break substantially decreasing HR before the athlete began his descent (Figure 3).
Post-summit, between 30 and 40 km, we found a sharp increase in running HR, despite decreasing altitude - likely reflecting an intensified effort after the break, combined with the digestive load of food. Glucose in this stretch spiked simultaneously to ~160 mg/dL - its highest level during the ultra-marathon; this was substantially higher than average glucose levels during the race (~119 mg/dL). We observed another rest period around 40 to 45 km, during which running HR again stabilised. As the marathon progressed and altitude continued to decrease, HR fluctuations diminished, displaying smaller peaks and troughs. In the final kilometres of the race (70 to 72), running HR exhibited a slight increase, possibly due to cumulative fatigue and increased effort moving towards the finishing line within the prescribed time span (Figure 3).
Figure 3: Graphical representation of average running HR over the duration of the ultra-marathon, grouped into 5 km increments and with respect to the mountain’s changing altitude. Labelled points on the blue HR trace represent rest periods, and the blue shaded region represents variation in running HR across each 5 km increment during the ultra-marathon.
Conclusions, Limitations and Future Directions
This
case study offers valuable insights into how altitude affects recovery, glucose disposal, and performance in an ultramarathon athlete. Our findings show that transitions between sea level and high altitude have an acute impact on sleep quality and recovery metrics, with sleep HR, HRV, and movements showing noticeable changes before gradually improving through acclimatisation. Notably, elite athletes may experience disrupted sleep during altitude transitions, but as observed in this case, they tend to acclimatise and recover quickly, with rapid reductions in sleep movements within just two days of either transition. This is in contrast to a previous
study, where we demonstrated how a single episode of sleep deprivation in the general population impacts sleep, and recovery extends over days afterwards. Running at higher altitudes involved increased cardiovascular strain, and elevated glucose processing was easily visible when comparing the two 18 km runs. During the ultra-marathon itself, HR varied considerably, influenced by both altitude changes and rest periods, underscoring the critical role of pacing and recovery strategies in high-altitude environments. Glucose utilisation during the ultra-marathon was available with gaps, likely due to connectivity issues in areas with sparse network coverage. This data will be analysed in another study.
Case studies with singular athletes limit their generalisability, and the relatively short observation period may not fully capture long-term adaptation effects. Additionally, factors such as nutrition, hydration, and psychological stress were not controlled for, which could have influenced the physiological responses observed
11,12,13.
This investigation offers an initial foundation on which future research can be built, expanding to larger groups of athletes across various endurance disciplines and skill levels.
The Ultrahuman Ring AIR, with its extended battery life (>14 h with high sampling rates) and robust data capture capabilities, proves to be a reliable tool for monitoring athletes in endurance sports, even in areas with inconsistent network coverage. Our results hold substantial implications for athletes, coaches, and sports scientists preparing for high-altitude endurance competitions. They highlight the importance of allowing adequate time for acclimatisation, closely monitoring recovery metrics, and employing tailored nutrition and pacing strategies to optimise performance at altitude.
Reach out to support@ultrahuman.com for commercial queries and science@ultrahuman.com for scientific queries.
