Cognition and Neuropsychological Changes at Altitude—A Systematic Review of Literature

High-altitude (HA) exposure affects cognitive functions, but studies have found inconsistent results. The aim of this systematic review was to evaluate the effects of HA exposure on cognitive functions in healthy subjects. A structural overview of the applied neuropsychological tests was provided with a classification of superordinate cognitive domains. A literature search was performed using PubMed up to October 2021 according to PRISMA guidelines. Eligibility criteria included a healthy human cohort exposed to altitude in the field (at minimum 2440 m [8000 ft]) or in a hypoxic environment in a laboratory, and an assessment of cognitive domains. The literature search identified 52 studies (29 of these were field studies; altitude range: 2440 m–8848 m [8000–29,029 ft]). Researchers applied 112 different neuropsychological tests. Attentional capacity, concentration, and executive functions were the most frequently studied. In the laboratory, the ratio of altitude-induced impairments (64.7%) was twice as high compared to results showing no change or improved results (35.3%), but altitudes studied were similar in the chamber compared to field studies. In the field, the opposite results were found (66.4 % no change or improvements, 33.6% impairments). Since better acclimatization can be assumed in the field studies, the findings support the hypothesis that sufficient acclimatization has beneficial effects on cognitive functions at HA. However, it also becomes apparent that research in this area would benefit most if a consensus could be reached on a standardized framework of freely available neurocognitive tests.


Introduction
In the last century, mountaineering has gained in popularity and is now a mass phenomenon with a continually growing number of individuals hiking, skiing, climbing, and trekking in high-altitude (HA) environments. Additionally, HA exposure can occur in workrelated contexts (aviation and astronauts or health care services and alpine rescue). However, at higher altitude the reduced partial pressure of oxygen and the decreased barometric pressure can cause acute mountain sickness (AMS) [1], or other possibly life-threatening diseases such as high-altitude pulmonary (HAPE) [2], or cerebral edema (HACE) [3]. With a ratio of about 21 percent of oxygen, the composition of the atmosphere at HA is roughly the same as at sea level. However, the total atmospheric pressure, and therefore oxygen pressure, declines exponentially with altitude [4][5][6]. Symptoms of altitude disease can occur when reaching above an elevation of about 2500 m (8202 ft). HA symptoms become more severe with increasing altitude, and with faster ascent rates, with the decisive factor for acclimatization being the sleeping altitude (of about 1500 m [4921 ft]) [4]. Altitude is  Table A1).  Table A1).
The inclusion criteria were studies with: (1) healthy adult subjects exposed to an altitude equal or above 2440 m (8000 ft) (either actual or through generated conditions corresponding to the intended altitude); (2) a cognitive function assessment in the context of altitude exposure using an experimental design; and (3) within-subject design with a control condition, where the baseline test condition may at most be measured up to 1500 m (4921 ft); or between-subjects design with a matched control group tested under normoxia.
Excluded were: (1) studies with a first test time point one month or longer after arrival at altitude and studies with subjects exposed to chronic hypoxia, such as miners, for the possibility of drawing conclusions from the direct effects of altitude; (2) studies that explicitly examined the effect of altitude by modulating another variable, such as sleep deprivation, exercise, or the intake of pharmaceutical agents; (3) studies performing only electrophysiological cognitive measures (e.g., event related potentials); (4) studies conducted with professional pilots or military personnel; and (5) case reports, reviews, expert opinions, comments, or letters to the editor, studies on animals, conference or abstract reports and articles not written in English.
The primary search provided the following: 1404 articles, 21 abstracts, 14 reviews, and two meta-analyses. Additionally, 30 sources were gathered from inspecting the references with the original search criteria. As a whole, 52 articles were included. On two occasions, there are two articles ( [31][32][33][34]) with each referring to the same study but presenting different tests; therefore, all four articles are listed in the following. The excluded articles and their reasons for being ruled out are listed in Table A1 in Appendix A.

Quantitative Description of the Research Methods
Of the analyzed studies, 29 studies (31 articles) were field studies and 23 were classified as laboratory studies, including two studies that investigated both conditions [36,37]. In 21 of the field studies, participants reached the altitude partially by active climbing, and in eight studies there was a passive ascent only. Of the laboratory studies, seven were performed establishing hypoxia via inhalation of a hypoxic gas mixture, eleven took place in a normobaric hypoxia chamber (NHC), and in five studies, the subjects were tested in a hypobaric hypoxia chamber (HHC). In the field investigations, the study sites were located at altitudes between 2590 m (8497 ft) [38] and 7200 m (23,622 ft) [39], with the highest point also reached at 8848 m (29,029 ft) [40]. Among the laboratory studies, the simulated altitudes under normobaric and hypobaric conditions ranged from 2440 m (8000 ft) [41] to 8848 m (29,029 ft) [42] (see also Table 1). All studies were published between 1963 and 2021.

Field Studies (n = 29)
Passive and active ascent (n = 21) The use of familiarization sessions/test trials before the start of the actual cognitive assessments has been mentioned in several articles and was included in the tables. These familiarization trials are supposed to minimize learning effects or practice effects in the intraindividual comparisons. To take these learning effects over the course of several tests into account, control groups were involved in eight articles [25,28,31,36,42,47,50,67]. Learning effects were also discussed or considered in another 32 articles. In three articles, extensive familiarization trials were used to try to establish a performance plateau in advance [24,41,51]. In Lefferts et al. (2019) [51], additional training sessions occurred during the field study "to prevent loss of task familiarity owing to the passage of time between test days". Another method to prevent learning effects, the use of parallel versions or alternative forms of the applied tests, was mentioned in 17 articles [25,27,31,34,40,41,43,46,49,50,53,54,56,67,[72][73][74] and is marked in the following by x 1 (meaning alternate forms used). Among them, test batteries were used in nine articles: the Computerized Neurocognitive Test Battery CNS Vital Signs [46,74], the CogState Computerized Battery [25], the Cognition Test Battery [43], the PC based Multiple Attribute Task Battery [41], the FACTRETRIEVAL2 Test Battery [54], the Cambridge Neuropsychological Test Automated Battery [31], the ANAM-4th Edition [72,73], and the Defense Automated Neurobehavioral Assessment Test Battery [34].

Classification According to the Cognitive Domains Studied
As shown in Table 2, a total of 112 different neuropsychological tests were found, with the Stroop Test being the most frequently used with ten applications, eleven if adding the modified version. Overall, of the 173 test applications performed, 74 showed significant impairment (see also Table 3).

Orientation
Orientation for time and place depend on the awareness of self in relation to one's surroundings, presuppose a consistent and reliable integration of perception and attention as well as an awareness of the ongoing narrative and, thus, memory [14]. Cognitive domains with the superordinate term orientation were tested in three studies. Here, the studies conducted in the field by Davranche et al. (2016) [24] at 4350 m (14,272 ft) and Nelson (1990) [54] at over 6500 m (21,325 ft) yielded impairments.
Attentional capacity, processing speed and working memory Attention, concentration, and tracking are necessary skills for goal directed behavior. They can only be measured in the context of a cognitive activity sequence, depending on the focus. The temporary storage of information plays an essential role and a common characteristic is them having limited capacity [14]. Twelve different tests were performed to examine attentional capacity in short-term memory. Regarding the field tests, the Auditory Digit Span Test showed impairments at 4280 m (14,042 ft) [45] and at 5100 m (16,732 ft) [25], whereas other authors ( [39,56]) found no impairments at higher altitudes in the field. Furthermore, the Memory Search Task at 4330 m (14,206 ft) [28], the Picture Recognition Test at 4280 m (14,042 ft) [45], and the Verbal Free Recall Test [55] showed reductions in performance at 4500 m (14,764 ft) as well as at 5040 m (16,535 ft). Three laboratory tests examined attention, with the Corsi Block Forwards and Digit Span Test-Forward at 4500 m (14,764 ft) [66], and the latter also at 7620 m (25,000 ft) [27] showing deteriorations.
Working memory, also referred to as mental tracking, can deal with more complex cognitive operations and allows information to be maintained in temporary storage and be manipulated [14]. This includes an executive control mechanism to focus attention and block out interference [79]. In the working memory, in 14 performed experiments, seven of them showed changes due to HA, with four tests in De Aquino's (2012) [66] [72] also showed alterations in their first study of 2015. All five tests performed in the field showed no alterations.

Concentration/Focused attention
Attention in terms of concentration or focused attention is believed to be the basis for other, more complex components of cognition [14,80]. Complex attention also requires visual scanning, visuomotor coordination, motor persistence, and response speed and can be measured via symbol substitution tests [81]. Divided attention can be assessed by Trail Making Tests and it further requires scanning, visuomotor tracking, and cognitive flexibility [14]. To investigate focused attention, altitude-induced changes were examined using 20 neuropsychological tests. Limitations in performance were found in the Code Substitution Task at 4330 m (14,206 ft) [28], in the Continuous Performance Test at 5500 m (18,045 ft) [74], and in four investigations with the Digit Symbol Substitution Test, starting at 3269 m (10,725 ft) [34,49,60,74]. In contrast, improvements in the Digit Symbol Substitution Test were found at 5100 m (16,732 ft) [25] with no changes in two other examinations [43,46]. Using the Frankfurt Attention Inventory-2, subjective impaired attentional functions were reported in both field and laboratory testing [36]. In the Paced Auditory Serial Addition Test, the results were heterogenous, reaching from deterioration, improvement, to no change. The Trail Making Test A showed impaired performance in two of five studies, once in a field study at 3500 m (11,483 ft) [50] and in a laboratory study at 5334 m (17,500 ft) [27]. Two tests for response inhibition showed no changes with the Go/No-Go test [34,72].

Processing speed
Processing speed can be examined with the help of tests of response speed in regards to reaction time and accuracy data, whereas deceleration is often subject to attentional deficits [82]. Processing speed was studied in eleven experiments, approx. half of which showed no significant changes in altitude. Limitations were found in different Reaction Time Tests, e.g., in Subudhi et al. (2014) [34] in two investigations at 5260 m (17,257 ft). The results using the Psychomotor Vigilance Test showed significantly impaired performance at 3800 m (12,467 ft) [43] and at 5050 m (16,568 ft) [32]. At an altitude of 2590 m (8497 ft), Latshang et al. (2013) [38] showed no changes in the outcome of the Psychomotor Vigilance Test, as did   [49] at 3269 m (10,725 ft), although deteriorations correlated with poor sleep. In Harris et al. (2009) [25], improved reaction times were found at 5100 m (16,732 ft).

Perception
Another topic of particular interest in HA research is visual and auditory perception. Visual perception tests require little or no physical handling of the test material. However, the complexity of brain functions makes overlap inevitable, and most tests also assess other functions such as attention, spatial orientation, or memory. Auditory perception examines, in part, skills in phoneme discrimination and speech sensation [14]. In perception, five of eight trials were unaffected by altitude [34,43,53,56]; Reading of Briefly Displayed Letters improved at 3450 m (11,319 ft) in both the field and the laboratory [37], whereas the Pattern Comparison Task showed deterioration at 4330 m (14,206 ft) [28].

Memory
Efficient memory can on one hand retain information, and is examined by assessing encoding using immediate retrieval trials. Second, it can retain the information even after the application of interference during a delay period. For the measurement of memorability, it is also important to understand whether underperformance is a retention or a retrieval problem [14]. For memory, three subcategories were examined in more detail in a total of eight studies. Long-term memory was not affected at altitudes up to 7100 m (23,294 ft) [54]. For verbal memory, two positive tests were found, one using Rey's Auditory-Verbal Learning Test in the field at 5300 m (17,388 ft) [50], and the Verbal Memory Test in the laboratory at 5500 m (18,045 ft) [74]. Visual memory was also impaired at 5500 m (18,045 ft) [74] and on the Match to Sample Test at 5260 m (17,257 ft) [34].

Verbal functions and language skills
This heading includes tests for aphasia. Here, spontaneous speech, the ability to repeat words or sentences, syntax comprehension, the power to name things, and reading and writing are tested [14]. Two tests for verbal functions showed impairments beyond an altitude of 4340 m (14,239 ft) [56,57].

Construction and motor performance
This includes copying abilities, assembling and building, and, further, the assessment of motor skills and manual dexterity functions [14]. With regard to construction and motor performance, seven of twelve different tests showed impairments, although these were only examined in one study each. Two tests with multiple occurrences showed different results, such as Finger Tapping, with a significant deterioration in a laboratory test at 5500 m (18,045 ft) [42]. Deficits in the Pegboard-Psychomotor Test were shown at an altitude of 5300 m (17,388 ft) [50] and again at 5500 m (18,045 ft) [42]; however, Merz et al. (2013) [53] found unimpaired performance at 6265 m (20,555 ft) using the Pegboard-Psychomotor Test.

Concept formatting and reasoning
This category contains investigations by means of verbal procedures and visual formats such as mathematical, i.e., arithmetic reasoning problems, or, for example, by sorting and shifting [14]. For problems requiring concept formatting and reasoning, three of the six studies with six different tests were significantly altered at HA, two of them with impairment, namely in the Category Search Task performed at 4330 m (14,206 ft) [28] and the Number Ordination-Rey's Test at 5500 m (18,045 ft) [42]. However, in the Robinson's Numbers Test [57], the performance improved at HA at 4340 m (14,239 ft).

Executive functions
Executive functions, such as planning and decision making, are the most complex behaviors and are solid cognitive abilities with sufficient accountability necessary to respond appropriately to novel situations. At HA, the assessment of potential dangers is essential for survival. They further form the basis of many skills in the areas of cognition, emotion, and social skills. Willpower, planning and decision-making, goal-directed action, and effective performance are listed as four major components in the concept of executive functions [14]. With 22 tests and 43 experiments, executive functions were the most extensively studied cognitive domain in the collected studies. Approximately half of the experiments were associated with impairment in executive functions or increased risk behavior. In contrast, three studies found improvements or reduced risk behavior at HA. Regarding the three most common tests, the N-Back Number Task, Stroop Test, and Trail Making Test B, there are still mixed results. For the N-Back Number Task, impaired performance was found at the highest altitude tested, 5160 m (16,929 ft) [51], and in the only laboratory test at 4500 m (14, [75]. The Stroop Test was used ten times [23,27,46,50,52,59,62,64,66,74]. In the field examinations, the results were very heterogeneous; twice there was no disturbance [23,46], twice there was a worsening [50,59], but on the follow-up examination there was no further deviation, and once there was an improvement [52]. In the laboratory studies, four out of five tests showed impaired functions in the subjects [27,62,66,74]; the altitudes investigated ranged from 3500 m (11,483 ft) [62] to 7620 m (25,000 ft) [27]. The study by Ochi et al. (2018) [64], which investigated three altitudes from 2000 to 5000 m (6562-16,404 ft), merely found improvements in reaction time. Trail Making Test B revealed impairments in two of five tests, in the field above 3500 m (11,483 ft) [50] and the laboratory above 5334 m (17,500 ft) in [27]. The field study by Harris et al. (2009) [25] elicited improved results at 5100 m (16,732 ft).

Further and mixed domains
Lastly, the remaining domain of affective flexibility yielded mixed results in two examinations, with a significant deterioration at 4300 m (14,108 ft) [73] and no changes at 4500 m (14,764 ft) [78]. A Mini-Mental State Examination, screening multiple cognitive domains (concentration or working memory, language and praxis, orientation, memory, and attention span [83]), failed to detect any deficits at 4500 m (14,764 ft) [77].                         Table 3, more than twice as many took place in the field than in the laboratory. Table 3 below provides an overview of the frequency of results depending on the research conditions, with test results divided into no alteration, impairment, and improvement. In the literature, the improvements listed in the tables are most likely attributed to learning effects from the repeated measurements and are not further considered here for simplification reasons. The numbers from Table 3 suggest that the measured impairments might have something to do with the rate of acclimatization. Mean values were calculated in relation to the altitudes showing impairment in field (1) versus laboratory studies (2). Regarding the results showing impairment on several levels of altitude, the lowest significant level was chosen. Due to different variances, a t-Test assuming unequal variances was calculated, and the mean altitudes did not differ (3).

Classification According to the Numerical Frequency of Results
Of the 173 tests applied (sum of field and laboratory tests), 70.5% were conducted in the field. Of the field tests, slightly more than half of the total 122 test applications were performed under active ascent. Of all tests performed in the field, 60.7% showed no impairment, 5.7% showed improvement, and 33.6% found significant differences at HA. In active ascent, 60.3% showed no alteration, and 32.4% showed deterioration. In passive ascent, 63.2 % of the trials had no alterations, and 34.2% had deteriorations. Of the 51 laboratory tests, 31.4% showed no change, in 3.9% there was an improvement, and 64.7% showed deterioration in percentage terms. Numerically, in hypoxic gas mixture testing and hypobaric chamber studies, no impairment occurred about as often as impairment. Of the studies with normobaric hypoxia, 82.1 % showed hypoxia-related deterioration. improvement. In the literature, the improvements listed in the tables are most likely attributed to learning effects from the repeated measurements and are not further considered here for simplification reasons. The numbers from Of the 173 tests applied (sum of field and laboratory tests), 70.5% were conducted in the field. Of the field tests, slightly more than half of the total 122 test applications were performed under active ascent. Of all tests performed in the field, 60.7% showed no impairment, 5.7% showed improvement, and 33.6% found significant differences at HA. In active ascent, 60.3% showed no alteration, and 32.4% showed deterioration. In passive ascent, 63.2 % of the trials had no alterations, and 34.2% had deteriorations. Of the 51 laboratory tests, 31.4% showed no change, in 3.9% there was an improvement, and 64.7 % showed deterioration in percentage terms. Numerically, in hypoxic gas mixture testing and hypobaric chamber studies, no impairment occurred about as often as impairment. Of the studies with normobaric hypoxia, 82.1 % showed hypoxia-related deterioration.

Discussion
The aim of the review was not only to report an overview of the study results an summarize the evidence for and against cognitive impairment at moderate, high and ex treme altitudes, but also to provide a closer look at the neuropsychological tests used. Fo this purpose, they were grouped under the respective cognitive domains and the resul of each neuropsychological test were listed in detail.

Major Findings
The major findings of the current analysis were that 112 different tests have bee used. With 74 of 173 test applications per subject, less than half of the tests resulted i impairment. A novel approach in this review was to assign each neuropsychological te to its cognitive domain. Since some brain regions appear to be more dependent on oxygen it would be interesting to see whether this also manifests itself in differences in the ind vidual cognitive domains. Attentional capacity, concentration, and executive function were the most frequently studied cognitive domains. However, it is not possible to con clude that they were more sensitive than others, because we found impairments in a cognitive domains.
In the most extensively studied cognitive domain "Attentional capacity, processin speed and working memory", 5 out of 21 field studies yielded impaired results. However, is noticeable that out of the thirteen laboratory tests, nine found deteriorations above th simulated altitude of 4300 m (14,108 ft). It seems possible that some cognitive domains ar more likely to be distracted by sudden hypoxic conditions than others, as might be th case in the laboratory studies.

Discussion of Threshold Altitude
An obvious conclusion would be that there are preconditions, such as a threshol altitude, under which significant differences collected with a particular test can be repro duced. A good example that this is partially true can be found in the Psychomotor Vigilanc

Discussion
The aim of the review was not only to report an overview of the study results and summarize the evidence for and against cognitive impairment at moderate, high and extreme altitudes, but also to provide a closer look at the neuropsychological tests used. For this purpose, they were grouped under the respective cognitive domains and the results of each neuropsychological test were listed in detail.

Major Findings
The major findings of the current analysis were that 112 different tests have been used. With 74 of 173 test applications per subject, less than half of the tests resulted in impairment. A novel approach in this review was to assign each neuropsychological test to its cognitive domain. Since some brain regions appear to be more dependent on oxygen, it would be interesting to see whether this also manifests itself in differences in the individual cognitive domains. Attentional capacity, concentration, and executive functions were the most frequently studied cognitive domains. However, it is not possible to conclude that they were more sensitive than others, because we found impairments in all cognitive domains.
In the most extensively studied cognitive domain "Attentional capacity, processing speed and working memory", 5 out of 21 field studies yielded impaired results. However, it is noticeable that out of the thirteen laboratory tests, nine found deteriorations above the simulated altitude of 4300 m (14,108 ft). It seems possible that some cognitive domains are more likely to be distracted by sudden hypoxic conditions than others, as might be the case in the laboratory studies.

Discussion of Threshold Altitude
An obvious conclusion would be that there are preconditions, such as a threshold altitude, under which significant differences collected with a particular test can be reproduced. A good example that this is partially true can be found in the Psychomotor Vigilance Test, which was used in four field studies. In an altitude-dependent order of the results, two, namely the studies of Frost et al. (2021) [43] at 3800 m (12,467 ft) and, further, by   [32] at 5050 m (16,568 ft), show impairments. The investigations of reaction time using the Psychomotor Vigilance Test at 2590 m (8497 ft) [38] and at 3269 m (10,725 ft) [43] show no impairments. Therefore, a logical conclusion would be to assume that the subjects' ability to respond at sea level are limited with an increase in the altitude of the study. Regarding the Psychomotor Vigilance Test, in particular, this would be between an altitude of 3269 m (10,725 ft) and 3800 m (12,467 ft). The picture is similar for the N-Back Number Task, which was used in four tests and twice showed an impairment at HA. The field test of Frost et al. (2021) [43] at 3800 m (12,467 ft) was without changes, whereas the laboratory test of Williams et al. (2019) [75] at 4500 m (14,764 ft) shows significant differences.
That this "threshold altitude" cannot be generalized to other tests appears evident, with reference to the most extensively studied cognitive domain of executive functions. In this domain, a large number of different tests was used, some of which showed no impairment up to very high altitudes, such as the Ruff Figural Fluency Test of Merz et al. (2013) [53] at 6265 m (20,555 ft) in the field.
The Stroop Task is the most represented with ten applications and an overview of the results of the studies concerned raises new questions. Namely, of the five field tests, the study conducted at the lowest point, 3109 m (10,200 ft), by Weigle et al. (2007) [59], deviates from the norm, whereas the study conducted at the highest point, 5500 m (18,045 ft), by Issa et al. (2016) [23] is found to be unaffected.

Effects of the Ascent Mode
The mode of ascent could influence the results of neurocognitive tests at HA. Roughly broken down, the passive mode of ascent using cars or mountain cable cars shows higher ascent rates than for an active ascent on foot. As described before, a slower ascent contributes to better acclimatization [10,84]. With reference to Davrache et al. (2016) [24], better acclimatization would result in fewer negative effects caused by HA. In laboratory tests, in most cases, hypoxic conditions were established much faster than they would be via active ascent in field studies. Looking again at the Stroop Test, it is noticeable that impairments were found in four out of five laboratory tests, starting at 3500 m (11,483 ft) [62], up to 7620 m (25,000 ft) [74].
In addition, there seem to be other issues such as test-retest reliability, which can be seen unprecedentedly in the two studies by Seo et al. (2015 and. For example, the first study in 2015 [72] showed significant changes in the Running Memory Continuous Performance Test, but the re-run in 2017 [73] could not replicate this result. If mathematical significance calculations can exclude accidental events, there must be other influencing mediators that have not yet been conclusively clarified. The obvious and already extensively researched ones would be, for example, physical strain, cold, or the quality of sleep.

Differences of Field and Chamber Test Applications and Consequent Assumptions
Far more than twice as many test applications took place in the field and there is a difference in the number of significant results. Proportionally, the tests that found impairments in the field at HA amounted to about one third of all tests conducted in the field and slightly less than two thirds showed no changes. In contrast, for laboratory tests, the ratio of tests without altitude-induced changes and those with impairments seemed to be almost reversed. Since there was no difference between the calculated mean altitude of impairments in field versus laboratory test settings, it is yet unclear what exactly is the reason for cognitive performance being proportionally more often impaired in laboratory studies than in field studies. The more rapid onset of compromising circumstances and insufficient acclimatization could be held accountable. As mentioned previously, Martin et al. [10] had concluded that the duration of the altitude exposure and altitude level had the greatest impact on cognition, and that altitude acclimatization appeared to have a positive effect on cognitive performance. In field studies, the ascent rate is decisively lower compared to that of passive field tests. In addition, the overall duration is longer than in most laboratory studies, which means that acclimatization can take place over a longer period. In conclusion, this review also suggests that sufficient acclimatization has a beneficial effect on cognitive functions at HA.
Alongside this, however, it must also be mentioned that in different experimental conditions, different stressors are in the foreground as well. Besides hypoxia, supplemental stressors in the field may include physical exhaustion from the climb, cold, sleep disturbance, and psychological factors, such as stress and anxiety, over the climb. Participants in passive ascents are exposed to hypoxic conditions more rapidly, which is why the problems of inadequate acclimatization are more likely to occur, as measured by the Lake Louise Scale, which is designed for the clinical diagnosis of AMS and assessment of its severity [85]. In laboratory studies, it is probably also the latter. Additionally, in chamber studies, difficulties were caused by unfamiliar situations, such as the confined space that allows only limited freedom of movement and isolation from the other group members, not to neglect psychological factors, such as claustrophobia. The extent to which these specific factors affect the study results is unclear at this time and needs to be considered.

Limitations
Throughout this review, an attempt has been made to give space to the complexity of the studies analyzed and to list the research modalities and results as completely as possible. On the other hand, the aim was to approach the object of investigation, the effects of hypoxia on cognitive abilities, as straightforwardly and comprehensibly as possible. These intentions are, in view of modern research with consideration of numerous cofactors in the statistical models, contrary to a simple approach. It is important to mention the risk of potential bias in this work. For one thing, this may be the case due to the so-called "publication bias", i.e., the potentially selective reporting of complete studies and nonpublication of study results without significant findings. However, "outcome reporting bias" within individual studies also has to be discussed as a reason for sources of error.
Despite the broad sample of studies, it must also be mentioned that a large number of studies, some of them highly qualitative, were excluded because they were conducted with military personnel or aircrew. The decision against including studies with professional pilots or military personnel is based on the fact that, on the one hand, highly specialized individuals, e.g., air combat pilots, seem difficult to compare with mountaineers. Second, the studies were often computer-based multi-tasking assessments (such as those used in the study of Kourdidou-Papadeli et al. [41]) that examined occupation-specific scenarios and made it difficult to draw direct conclusions about underlying cognitive functions due to the multifactorial and comprehensive nature of the assignments.
The bundling of the test procedures into cognitive domains was conducted by drawing artificial dividing lines that are not entirely objectively reproducible. As already noted in the methods section, this artificial simplification is a source of error. One possibility to at least contain it or to standardize it across the studies would also be the creation of a standardized framework of tests. One difficulty and possible bias in classifying the tests was that some of the same neuropsychological tests were used, but their names differed somewhat. In places where this was clearly evident, summaries were made under the same test name. However, this was probably not successful everywhere, which is one more reason to use standardized test batteries.

Conclusions
This review explicitly reports on applied neuropsychological tests, classifies them into cognitive domains, and correlates their results to the investigated altitudes and the intervention methods. Altitude and duration of altitude exposure seem to have the greatest impact on cognition, and sufficient acclimatization has beneficial effects on cognitive functions at HA.
Future studies should try to find a common consensus and complement each other. One possibility would be the creation of a standardized framework of tests, for instance an open access test battery. If used frequently, a standardized test battery could make the different experiments more comparable and help to identify fundamental influencing factors. To date, there have been studies with computerized test batteries; however, these were provided by commercial service companies and were therefore only used in occasional studies.
The "STAR" data reporting guidelines provide investigators with a suggested pathway of quality aspects to examine. Inter-study comparisons can be made more easily. The STAR guidelines for data reporting have proved useful in assessing study quality, and their utilization in future studies may contribute to a further standardization of methods.
Supplementary Materials: The following supporting information can be downloaded at: https:// www.mdpi.com/article/10.3390/brainsci12121736/s1, Table S1. STAR data reporting guidelines [35] assessed in the studies included in the current analysis; Table S2. List of articles and their findings of altitude impact on cognitive performance.