Visuospatial attention at high altitude in two ethnically distinct groups of soldiers

Subjects and study protocol

The minimum sample size identified for the statistical analysis was 15 from each group with an effect size of f = 0.25 (this indicates a medium effect) and power, b = 0.95. However, due to certain logistics and administrative issues, we could only recruit 10 participants from each group. Therefore, the study was conducted on 20 participants (ten Indian [Age – 23.11 ± 2.0 years, height – 173 ± 3.0 cm, weight – 66.15 ± 5.17 kg and body mass index (BMI) – 22 ± 1.81 kg/m²] and ten Kyrgyz [Age – 20.60 ± 3.0 years, height – 176.10 ± 4.0 cm, weight – 72 ± 6 kg and BMI – 23 ± 1.60 kg/m²] soldiers) with an educational qualification of higher secondary level. All participants were lowlanders with no history of HA exposure, 6/6 visual acuity without corrective lenses and no recent eye infection, history of eye surgery or discomfort or pain in the eyes while using digital display media. The term ‘lowlanders’ essentially depicts the natives of the sea levels. The average terrestrial elevation of Indian lowlanders and Kyrgyz lowlanders is 235 m. and 800 m, respectively. The participants did not undergo any altitude training. They provided informed consent to participate in the study. The study protocol was approved by the Institutional Ethics Committee on the use of humans as the experimental subjects and the experiment conforms to the principles outlined by the deceleration of the Helsinki protocol. Indian soldiers were flown to Bishkek, Kyrgyzstan, by 4 h of a direct flight. There is a time difference of half an hour between India and Kyrgyzstan (the Kyrgyz time zone is advanced by half an hour). To counteract the effects of jet lag (if any), baseline recording was carried out on day 3 at Bishkek (800 m above sea level) in both Indian and Kyrgyz soldiers, and both groups of soldiers were then ascended by road to HA, Tuya Ashu (3,200 m above sea level) by road. After a stay of 21 days, soldiers were returned to Bishkek, where the study was conducted on day 3 [Figure 1]. Atmospheric temperature was ranging from 28°C to35°C at Bishkek and from 9°C to 25°C at Tuya Ashu, whereas inside the rooms, the temperature was maintained at 22°C ± 2°C.

Close

Figure 1:

Schematic diagram of the experimental protocol. Indian soldiers (IS) reached Bishkek by air, and baseline recordings (R) were performed on both IS and Kyrgyz soldiers (KS) on day 3 (D3). Soldiers were transported by road from Bishkek to 3,200 m altitude (Tuya Ashu) where they stayed for 21 days. High-altitude R’s were carried out on day 3 (D3), day 7 (D7), day 14 (D14), and day 21 (D21). After that, soldiers were returned from high altitude to Bishkek by road and all the parameters were recorded on day 3.

Export to PPT

Saturation of peripheral oxygen (SpO₂) was measured using a finger pulse oximeter (Model No. 3230, M/s, Nonin Medical, Inc.).

Visual search tasks

Based on the principles outlined by Wolfe (2010)^[10] and Wolfe (2015),^[16] a visual search task was designed in the Java^TM platform in which participants had to search for a target in the presence of many distractors. For the present study, (A circle with a right cut) and (Circle with a left cut) were selected as the stimuli for ‘Target’ and ‘Distractors’, respectively. The ‘Target’ and ‘Distractors’ are mirror images of each other with a minute feature of discretion, which creates a strong conflict in the brain toward accurate identification. Target and distractors had similar dimensions during presentation mode, extending a visual angle of 1.2°–1° for a distance ranging from 60 cm to 70 cm between the screen and the participant’s eye. The opening space of the target and distractors extended a visual angle of 0.6°–0.5° for a distance ranging from 60 cm to 70 cm between the screen and the participant’s eye. With a set size of 12 and 16 stimuli, two types of search displays were generated. There were 1 target and 11 distractors, and 1 target and 15 distractors under 12 and 16 stimuli conditions, respectively. A total of 16 trials were conducted, with eight trials presenting the presence of targets and the remaining eight trials presenting the absence of targets. The stimuli were randomly distributed in the search matrix with a jittering of 100 pixels. The participants were instructed to register their responses by pressing the key ‘F’ in the presence of the target and the key ‘J’ during the absence of the target. The manoeuvre of key pressing had to be accomplished as early as possible, followed by a self-decision on the presence or absence of the target. The ‘target present’ and ‘target absent’ trials appeared randomly, followed by the registering of the responses by pressing a relevant key. The trials started with an eye fixation of 500 ms on a black ‘X’ in the middle of the screen on a white background. Immediately after that, the search matrices started appearing by pressing either key ‘F’ or key ‘J’. The next search matrix appeared with a delay of 3000 ms. The details of the experimental protocol are shown in Figure 2. Both groups of participants received a clear demonstration of the entire procedure of task performance. In the case of Kyrgyz soldiers, the detailed procedure of task performance was first vividly demonstrated to a native senior researcher (who had good proficiency in English) in English. The native senior researcher then instructed and showed the task to their soldiers in the Kyrgyz language, ensuring that they understood how to complete it. Before the start of the main experiment, each participant completed at least three practice sessions to become familiar with the task-performance hardware, software and experimental guidelines.

Close

Figure 2:

Schematic presentation of Visual search experiment which started with ‘5’ points eye calibration followed by a 500 ms fixation on ‘X.’ The experiment started by pressing of keys ‘F’ or ‘J’ and the matrix with set sizes ‘12’ and ‘16’ started appearing, response time was registered by pressing the ‘F’ and ‘J’ keys for the target present and absent respectively.

Export to PPT

Recording of eye movements and experimental setup

Volunteers were instructed to put on headset with the eye tracking optics (Sensomotoric Instruments [SMI] iViewX Eye Tracker, M/s SMIs, Germany) ensuring the ease of wearing for an optimum calibration condition. Followed by this ‘5’ point, calibration was performed by standard protocol. The recording software ‘iViewX’ (M/s SMI) was run at a sampling rate of 250 Hz. Stimuli were delivered in a 17-inch Liquid Crystal Display (LCD) monitor (M/s Hewlett-Packard, Model No. L1710) with a 5:4 aspect ratio and 160° × 160° Horizontal × Vertical Viewing Angle. Participants’ chins were supported by a chinrest to stabilise the redundant movement of the head. The illumination level of the experiment room was maintained at 450 lux. Recording started with the onset of the search task and stopped immediately with the termination of all search trials. The recorded files of eye movement data were analysed in ‘BeGaze’ software (M/s SMI).

Parameters

The correct reaction time of ‘target present’ (RTP) and reaction time of ‘target absent’ (RTA) were registered at the end of each trial for individual soldiers. Slopes (search efficiency), that is, time spent/item and intercepts of the visual search, were calculated from ‘set size 12’ to ‘set size 16’ for all the experimental sessions. In the case of eye movement parameters, average fixation duration (AFD) and saccade latency were considered to interpret the visuospatial attention while performing the visual search.

Statistical analysis

All the parameters were checked for normal distribution by Shapiro–Wilk test. A two-way mixed-design analysis of variance (ANOVA) was conducted to observe any significant effect of experimental independent variables on the dependent variables. The between-measure independent factor was the group of soldiers, that is Indian and Kyrgyz soldiers. There were two within-measure independent factors, (1) various days of experiments (DE), that is, Baseline, on day 3 (Day 3 at HA [HA_D3]), day 7 (Day 7 at HA [HA_D7]), day 14 (Day 14 at HA [HA_D14]) and day 21 (Day 21 at HA [HA_D21]) at HA and return from HA, and (2) set size, that is 12 stimuli and 16 stimuli at which both groups of soldiers were repeatedly measured. The individual and interaction effects of independent variables were observed on the dependent variables’ fixation duration and saccade latency. In the case of reaction time, there was another within-measure independent factor, that is, Target present/absent. Bonferroni post hoc test was also applied to compare the individual pairs followed by the overall significant effect of an independent variable. A Two-sample t-test (Assuming Unequal Variances) was also carried out to observe the significance difference between Indian and Kyrgyz soldiers of mean changes for all the variables at baseline. All the tests were conducted on the Statistical Package for the Social Sciences (Version 16). The significance of the test was confirmed when P ≤ 0.05.

Reaction time

There was no significant difference for RTP under set size 12 (t₁₈ = 1.80, P = 0.087) and 16 (t₁₅ = 1.32, P = 0.206) at baseline, respectively. Similarly, in the case of RTA, the difference was found non-significant under set size 12 (t₁₈ = 1.51, P = 0.147). Whereas, under 16 set size conditions, the mean difference in RTA was found significant (t₁₆ = 3.11, P = 0.006) at baseline. Results [Table 1] of mixed-design ANOVA revealed an overall significant effect of DE (F_[3.78,64.56] = 68, P < 0.05), set size (F_[1,17] = 26.20, P < 0.05) and presence/absence of target (F_[1,17] =104.71, P < 0.05) on reaction time of visual search. Interaction effects of DE X set size (F_[2.47,42.05] =19.12, P < 0.05), DE X Target (F_[2.20,37.41] =12.50, P < 0.05) also produced significant effects on reaction time. When the ethnicity interacted with DE, that is, DE X Group (F_[5,85] = 7.20, P < 0.05) and the presence/absence of target, that is, Target X Group (F_[1,17] = 7.50, P < 0.05) significant effects were observed on reaction time.

In the present study, it was observed that target present slopes ranged from 54 to 440 ms/item [Table 2] in both groups of soldiers, irrespective of DEs. The same fact is observed in the case of target-absent slopes, which range from 430 to 1131 ms/item [Table 2]. Graphical presentation of the slope of reaction time X set size function revealed slight steepness during target-present and target-absent trials, with a negative slope in HA_D3 in both Indian and Kyrgyz soldiers ([Figures 3 and 4], except for Kyrgyz soldiers during target-absent trials). The details of pair wise comparison has been reported in Table 3 by conducting Bonferroni post hoc analysis.

Close

Figure 3:

Target present correct reaction time are plotted as a function of set size at baseline, high altitude day 3 (HA_D3), day 7 (HA_D7), day 14 (HA_D14), day 21 (HA_D21) and return from high altitude for both Indian and Kyrgyz soldiers.

Export to PPT

Close

Figure 4:

Target absent correct reaction time are plotted as a function of set size at baseline (BL), high altitude day 3 (HA_D3), day 7 (HA_D7), day 14 (HA_D14), day 21 (HA_D21) and return from high altitude (RFH) for both Indian and Kyrgyz soldiers.

Export to PPT

Average Fixation Duration (AFD)

The mean difference in AFD was neither found to be significant under the 12 set size (t14 = 1.41, P = 0.180) nor under the 16 set size (t₁₆ = 0.68, P = 0.504) at baseline. DE produced a significant effect (F_[2.63,44.75] = 6.18, P < 0.05) on the changes in fixation duration [Table 4]. Interaction between DE X set size imposed a significant effect (F_[5,85] = 1.16, P < 0.05) on fixation duration. When Group separately interacted with DE, that is, DE X Group, significant effect was observed (F_[5,85] = 2.76, P < 0.05). The details of pair wise comparison has been reported in Table 5 by conducting Bonferroni post hoc analysis

Saccade latency

No significant changes were observed in saccade latency, neither under 12 set size (t₁₂ = 0.19, P = 0.851) nor under 16 set size (t₁₃ = 0.01, P = 0.984) at baseline. DE produced a significant effect (F_[4.19,71.33] = 5.14, P < 0.05) on the changes in saccade latency [Table 4]. The individual effect between group factors was found significant (F_[1,17] = 11.00, P < 0.05). When the group interacted with DE, that is, DE X Group a significant effect was observed (F_[5,85] = 2.55, P < 0.05). The details of pair wise comparison has been reported in Table 6 by conducting Bonferroni post hoc analysis.

Arterial oxygen saturation

Results of repeated measure ANOVA revealed significant changes (F_[4,45] = 39.97, P < 0.05) in SpO₂ in Kyrgyz soldiers. SpO₂ was significantly decreased after 3 days (96 ± 0.88; P < 0.001) of ascent to HA than baseline (baseline, 98 ± 0.16). On days 7 and 14, SpO2 remained the same as on day 3, 96 ± 0.40 (P < 0.001 than baseline) and 96 ± 0.86 (P < 0.01 than baseline), respectively, which further increased to 97 ± 0.65 on day 21 (P < 0.05 than baseline) on days. In Indian, soldiers significant changes (F_[4,45] = 24.9, P < 0.05) were also observed. SpO₂ was significantly decreased after 3 days (96 ± 0.73; P < 0.001) of ascent to HA than baseline (98 ± 0.16). On days 7, 14 and 21, SpO₂ remained the same as on day 3, 96 ± 0.90 (P < 0.01 than baseline), 96 ± 1.08 (P < 0.01 than baseline) and 96 ± 1.08 (P < 0.01 than baseline), respectively.

Limitations of the study

The present study was limited by several factors, which could be improved in future research. A lower statistical power may be expected due to the small sample size. Initiation of the assessment earlier, for example immediately after 24 h of arrival at HA may also be helpful for stronger interpretation. Integration of more advanced neurological techniques such as electroencephalogram and functional magnetic resonance imaging-based imaging will be of immense guidance to demarcate the task-specific activation of the cortical regions. Hence, future studies may be directed at exploring the effects of other variants of visual search such as conjunction search, more set sizes, and an earlier onset of experiments at HA involving a larger sample. Furthermore, a detailed socioeconomic and ethnogenetic profiling of both groups will also be considered critically for future investigations.