Discussion

The present study used a follow-up method to investigate the change in perceptual organization during HA adaptation and deadaptation. For the adaptation process, compared with the sea level baseline, increased HR and DBP and decreased SpO2 and VC were found when entering HA. Shorter latencies of P1 and N170 showed a faster process of perceptual organization, as well as a larger amplitude of complete face N170. For the deadaptation process, all physiological parameters recovered to baseline levels one week after returning to sea level. Compared with HA, the amplitude of the P1 component of the incomplete face decreased after one week, while that of the complete face decreased after one month. The right hemisphere N170 amplitude increased after entering HA and one week after returning to sea level but returned to baseline one month later. The changes in physiological data and ERP components were out of sync; the perceptual organization function recovery takes a relatively long time. These findings offer novel insights into the neural mechanism of HA adaptation and deadaptation on cognitive processes, especially in the nonprefrontal domain.
After twenty-five days of HA exposure, shorter P1 and N170 latencies were observed, consistent with previous chronic (two-year) HA exposure study . The results suggested that higher levels of cortical excitation within the visual cortex leads to a faster perceptual process when entering HA. After staying at HA for approximately four weeks, the organism is in a state of hypoxia stress. The reduced SpO2 and VC and increased HR and DBP revealed a decrease in the partial pressure of oxygen under a hypoxic environment at HA, causing sympathetic nerve excitement, which may lead to faster perceptual processing. This acceleration is compatible with previous studies that reported that the resting EEG of people who first arrived at HA had enhanced β wave activity, which is a manifestation of increased excitability and excessive alertness in the cerebral cortex . According to a previous study, hypoxia results in the facilitation of perception tasks . The perception task involves activating sensory cortices, which are greatly influenced by neurotransmitter activity, including dopamine (DA), 5-hydroxytryptamine (5-HT), and peptide corticotropin releasing factor (CRF). HA exposure induces high concentrations of neurotransmitters, leading to massive anoxic depolarization of neurons in cortical tissues and causing brain hyperexcitability , which could positively affect perceptual process speed. Existing studies have also suggested that cognitive performance decreased obviously when SpO2 was 80%, but activation and increased performance were observed when SpO2 was 90%, and this activation could have a general stimulating effect on central nervous system (CNS) functional adaptation to the impacts of hypoxia , which may reflect a mechanism attempting to restore/protect neuronal homeostasis . Therefore, it might be interpreted as an adaptive trait rather than a deficit, perhaps attenuating hypoxemia effects under mild hypoxia. Further studies are needed to provide definitive conclusions on the mechanisms relating to the acceleration of perceptual organization processing under HA adaptation.
The N170 amplitude was larger than that at sea level after entering HA for twenty-five days, which was also similar to the result of two years of exposure . The increased N170 amplitude under HA adaptation may be related to maintaining the facial configural encoding process efficiency during hypoxic stress, as certain cortical areas were activated to a greater extent to compensate for decreased processing efficiency due to HA exposure . The increased N170 amplitude may indicate a higher level of neural activity at HA . Interestingly, similar observations in terms of the N170 enhancement phenomenon were also found among elderly adults, which may be due to an increase in the sensitivity to visual stimuli or a decline in brain adaptability with aging . A recent study revealed that cognitive resources were insufficient under HA exposure during a visual attention paradigm, which is similar to normal aging . Moreover, hypoxia was found to accompany inflammatory brain states, such as neurodegeneration . Evidence underscores that hypoxia triggers both the accumulation of cerebral amyloidogenesis and tau phosphorylation, the pathological markers of Alzheimer’s disease, which are featured cognitive aging . Chronic exposure to an HA environment may have a similar neural mechanism to cognitive aging, or chronic exposure to HA may accelerate brain aging, but this inference needs more evidence to be confirmed.
After returning to sea level, the body makes adjustments to adapt to the sea level, and all the physiological parameters (SpO2, VC, HR, SBP and DBP) recover to baseline. Compared with HA, SBP decreased one week after returning to sea level, even though it had not been significantly altered by HA adaptation. These findings imply that the autonomic nervous system outcomes induced by HA exposure are reversible within one week after returning to low altitudes. P1 and N170 latency also returned to baseline less than one week after descending to lowlands, suggesting that the change in the perceptual process was also reversible after returning to sea level. However, hypoxia was found to have a longer lasting effect on brain activity than on blood, which might be because the brain consumes the most oxygen and is more sensitive to hypoxia . Cortical activation recovery under the HA deadaptation process always took more than one week in previous studies . A neuroimaging study showed that entering the HA or returning to sea level in one week were both associated with a sustained greater level of excitement in the posterior parietal cortex and occipital cortex, which are the foundation of the perceptual processes . A longitudinal EEG study also conforms to this view; increased beta power was found after one month of HA exposure (3,800 m), and reoxygenation occurred seven days after returning to sea level, which revealed a sustained higher level of cortical excitation within the visual cortex during HA deadaptation, which may be explained by the rise in capillaries induced by chronic hypoxia . The latency recovered to baseline sooner in our study, which indicates that the change in cortical activation may not be the direct cause of the change in latency. The change in latency may be modulated by a more basic physiological process, which may also regulate the activation of the cortex. The subtle changes in perceptual organization observed here may be a result of synergistic interactions between physiological deadaptation and the aftereffect of hypoxia on neurons . The physiological mechanism of latency changes and its relationship with cortical activation need to be further studied.
During the deadaptation process, compared with HA, the incomplete P1 amplitude decreased one week after returning to sea level, while the complete P1 amplitude decreased after one month. As the P1 amplitude was associated with attention arousal , the decrease in the P1 amplitude may reflect lower arousal to the low-level visual features during the HA deadaptation process. An MRI study was conducted on participants who carried out the same task following the same schedule and found that after one week of reoxygenation exposure, cell activity was inhibited in the visual cortex . This implied that the low-level face perception process was inhibited for sojourners who went through the deadaptation process because of the shortage of sufficient arousal. Although this initial visual processing of faces under both the incomplete condition and complete condition in the deadaptation process showed no difference compared to the baseline at sea level, the low-level face perception process of incomplete faces and complete faces was different after HA deadaptation. This may be because compared with the complete face, the processing of incomplete faces involves not only parietal lobe activity but also frontal lobe activity to compensate for incomplete visual information . After returning to sea level, this compensation mechanism of incomplete faces soon disappeared, so the P1 amplitude decreased in one week. For complete faces, attention arousal decreased gradually, leading to progressively smaller P1 amplitudes.
Seven days after descending to the lowlands, the N170 amplitude remained larger in the right hemisphere, but it was no longer detectable after one month, indicating that it takes more than one week but less than one month for configural face-encoding processing to recover to baseline. This face processing configuration stage is also reversible, although slower than autonomic nervous system recovery. Previous studies found physiological retention (arterial blood gases and hemoglobin) of adaptation after returning to sea level after seven days. During hypoxia in rats for three weeks, microvessels in the brain significantly increased, and remained elevated after reoxygenation . Therefore, the found N170 retention might be relevant to physiological responses that we had not measure or molecular and cellular responses in the brain . Only a tentative interpretation is possibly based on the present data, and further studies should be conducted to elaborate the neural mechanisms.
Both P1 and N170 tend to exhibit a larger amplitude in the right hemisphere than in the left hemisphere , and this dominance of face perception in the right hemisphere did not change under HA exposure. Our results also revealed that the complete face still evoked larger P1 and N170 amplitudes than the incomplete face in the HA exposure condition. Moreover, P1 latency for incomplete faces is shorter than that for complete faces, while N170 latency is longer in incomplete faces than in complete faces . This pattern represents the basic law of face perception organization processing, which was not altered by HA adaptation or deadaptation.
Notably, there are several limitations regarding the current study. First, it lacks a sea level control cohort. Moreover, we did not collect ERP data when participants arrived at HA immediately because hair washing is not recommended within three days of arrival at HA to prevent cold. Further study could assess the acute effects of HA exposure on perceptual process and compare it with chronic HA exposure condition. One more limitation is that we examined healthy, young, and prepared university students during an incremental ascent. Before entering HA, they had regular exercise training and took nutrition products in advance to prevent any possible sickness. As such, some generalization to a broader population may be limited.
Following the limitations outlined above, it would be advisable to replicate our findings in different altitude exposure circumstances (e.g., different HA durations, different altitudes, and different rates of ascent). Another useful addition to the design would be to elucidate the impact of HA adaptation and HA deadaptation on other cognitive function types. Finally, future studies could broaden existing research by investigating the specific relationships between physiological parameters (such as cerebrovascular and cardiovascular parameters) and cognitive alterations, especially how physiological and biochemical activities regulate neuronal activities.