Persistent deterioration of visuospatial performance in spaceflight
Study Title
Persistent deterioration of visuospatial performance in spaceflight
Organism
Homo sapiens
Project
| Project Category | Flight Program | Flight Duration | Flight Program Information Link |
|---|---|---|---|
| Spaceflight Study | International Space Station (ISS) | Half-year | n.a. |
Studied Phenomenon Property
Nervous and regulatory systems - Neuromotor control ---> Electrophysiological responses
Nervous and regulatory systems - Neuro-sensory function ---> Visuospatial performance
Nervous and regulatory systems - Cognitive performance
Investigation Attributes
Gender (human/animal research)
Male
Age (human/animal research)
54.2 (SD = 2.6)
Abstract
Although human adaptation to spaceflight has been studied for decades, little is known about its long-term effects on brain and behavior. The present study investigated visuospatial performance and associated electrophysiological responses in astronauts before, during, and after an approximately half-year long mission to the International Space Station. Here we report findings demonstrating that cognitive performance can suffer marked decrements during spaceflight. Astronauts were slower and more error-prone on orbit than on Earth, while event-related brain potentials reflected diminished attentional resources. Our study is the first to provide evidence for impaired performance during both the initial (~ 8 days) and later (~ 50 days) stages of spaceflight, without any signs of adaptation. Results indicate restricted adaptability to spaceflight conditions and calls for new research prior to deep space explorations.
| Assay Name | Protocol Description |
|---|---|
| Treatment protocol | In order to minimize fatigue, tasks were presented in 5–8 min blocks with short breaks in-between. Blocks of tasks were arranged in a way to minimize differences due to practice/fatigue building up in the course of a session. The order of task blocks was as follows: Control measurements, one block of Lines task, one block of Clock task NoFrame condition, one block of Clock task Frame condition, four blocks of Visuomotor Tracking task, one block of Clock Frame task condition, one block of Clock task NoFrame condition, and one block of Lines task. The total execution time was 70 min for each session. This paper is focused on the results of the Lines and the Clock tasks, for results regarding the Visuomotor Tracking task. The Lines and the Clock tasks were both developed for this study to assess the influence of weightlessness on the perception of spatial directions as well as attention-related ERP components.Subjects were equipped with an EEG cap and looked straight ahead at a laptop screen through a form-fitting facemask attached to a cylindrical tunnel. The tunnel excluded external visual cues and provided a circular viewing field. The screen was centered on the line of gaze at a distance of 25 cm from the eyes. The tunnel had a diameter of 22 cm. The background color of the screen was dark gray. During pre-flight and post-flight sessions, participants performed the experimental tasks in a seated position in a quiet room at a computer desk, while focusing on a computer screen through a facemask attached to a cylindrical tube. During flight, subjects performed the experiment in a quasi free-floating posture in the Columbus module of the ISS, while holding on to the tunnel-computer complex with the facemask strapped to the forehead with an adjustable belt. They had no rigid contact with the station structure during the performance of the experiment.Prior to the first session, subjects were familiarized with the procedure and practiced the experimental tasks during two sessions separated by at least six days. Subjects were trained to perform the in-flight measurements themselves with the help of a fellow crewmember. |
| Lines task | First, a simple yellow line appeared on the screen for 100 ms, presented on a light gray disk (with a diameter of 16.8 cm) . The orientation of this Reference stimulus had to be remembered. Following a 500 ms blank screen, a Probe stimulus (a blue line) was presented for 100 ms. The orientation of the Reference and the Probe stimuli could either be identical or could differ by 30 or 60 degrees with equal probability. Subjects had to indicate whether the Probe stimulus had the same orientation as the Reference stimulus by pressing a pushbutton on a gamepad as quickly as possible. If the lines had the same orientation, a button had to be pressed with the right index finger, while another button had to be pressed with the right thumb if the lines had different orientations. Subjects had 900 ms to react after Probe stimulus onset. The time between consecutive trials (stimulus onset asynchrony, SOA) was 1700 ms. In 20% of the trials, a task-irrelevant stimulus (a picture) was presented for 100 ms instead of the Probe stimulus. Subjects were instructed not to press any button in response to Irrelevant stimuli. Reference and Probe lines were both 2.1 cm in length and 0.2 cm in width. Irrelevant stimuli were various colorful, circular, fisheye pictures with a diameter of 4.7 cm showing buildings, statues, and everyday objects. The Lines task consisted of 450 trials divided into 2 blocks. In the first 10 trials of the first block, an auditory feedback (beep) was provided when the subject responded erroneously. |
| Clock task | At the beginning of each trial, a black number (Reference stimulus) representing clock time was presented on a light gray disk (representing the face of a digital clock) with a diameter of 1.1 cm for 100 ms. The following numbers were used as Reference stimuli: 1, 2, 4, 5, 7, 8, 10 and 11. Numbers were rotated randomly by ± 60º, ± 30º, or 0º. Following a 500 ms blank screen, a white dot (Probe stimulus) was displayed on the perimeter of an invisible circle (representing the face of an imaginary analog clock) for 50 ms. The diameter of the dot was 0.4 cm. Subjects had to indicate as quickly as possible whether the location of the dot corresponded to the digital clock time indicated by the Reference stimulus. By using a gamepad, subjects were instructed to press a button with the right index finger if the location of the dot matched the clock time and to push another button with the right thumb if the dot did not match the clock time. Subjects had 900 ms to react after the Probe stimulus onset. SOA was 1700 ms. The Probe stimulus could either match the clock time indicated by the Reference stimulus or could differ from it by 1 or 2 h with equal probability. As in the Lines task, task-irrelevant stimuli were presented instead of the Probe stimulus in 20% of the trials. Irrelevant stimuli were various fisheye pictures in response to which subjects were instructed not to press any button.The Clock task consisted of 600 trials divided into 4 blocks. In order to foster spatial orientation, a squared frame was presented around the visible workspace on the screen (Frame condition; with a 10.5 cm × 10.5 cm frame) in one half of the blocks. In the other half of the blocks, instead of a squared frame, stimuli were presented in a circle (NoFrame condition) with a diameter of 12.1 cm. In the first 10 trials of the first block of both conditions, auditory feedback (a beep) was provided when the subject pushed the wrong key. |
| EEG recording and analysis | In all in-flight and pre-flight measurements as well as in 30% percent of the post-flight measurements, EEG was recorded with the Multi-Electrodes Encephalogram Measurement Module (MEEMM, specifically created for the ISS by OHB Systems, Germany), with a sampling frequency of 1116 Hz. 58 scalp electrodes were placed according to the extended 10–20 system; the ground electrode was placed on the forehead. In the other 70% of the post-flight measurements, EEG was recorded with the ANT system (ANT Neuro, The Netherlands) with a sampling frequency of 1024 Hz. 59 scalp electrodes were placed using the extended 10–20 system while an additional scalp electrode (AFz) functioned as ground. It is worth noting that only comparisons with post-flight measurements could have been influenced by the different EEG systems, however, the two setups were compared (same day, same subjects) and no observable differences were found.Horizontal eye movements were monitored using two electrodes placed lateral to the outer canthi of each eye and vertical eye-movements were monitored with an electrode placed below the left eye. A right-ear reference was used for all recordings. EEG was analyzed with the EEGLAB toolbox. EEG was bandpass filtered offline (0.5–40 Hz, Kaiser windowed sinc FIR filter) and down-sampled to 512 Hz. Large, noisy time segments and channels were removed after visual inspection. Extended independent component analysis (ICA) was performed on individual data sets to remove eye blink artifacts from EEG recordings. ICA components representing eye blink and horizontal eye movement artifacts were identified by inspecting the component scalp maps, time courses and ERP-images (visualization of event-related signal variations across single trials) and were deleted. Missing data (channels) were interpolated using spherical spline interpolation. Processed EEG was re-referenced to the average of the signal of all electrodes and was lowpass filtered at 30 Hz (Kaiser windowed sinc FIR filter).1000 ms-wide epochs were extracted (100 ms pre-probe to 900 ms post-probe) and baseline corrected (100 to 0 ms pre-probe) in each task. Only trials with correct responses were analyzed. Epochs with a signal range exceeding 70 µV on frontal, central, and temporal channels and 100 µV on parietal and occipital channels (where alpha oscillations frequently exceed 70 µV) were discarded from the analyses. Grand-means were computed from individual averages. |
| Data analysis | To eliminate reactions reflecting fast guesses, only correct responses with a duration greater than 200 ms after Probe offset were included in the analyses of reaction time. Median reaction time was calculated for each session, stimulus type, and subject in both tasks.Accuracy was calculated as the percentage of correct button presses for Probe stimuli in both tasks. To assess the effect of spaceflight on both reaction time and accuracy, two-factor [Task (Lines, Clock) × Session (1:9)] rANOVAs were conducted. In order to evaluate the effect of visual frame in the Clock task, separate two-factor [Visual Frame (Frame, NoFrame) × Session (1:9)] rANOVAs were performed on both reaction time and accuracy.The ERP analysis focused on the P3a ERP component elicited by Irrelevant stimuli and the P3b component elicited by Probe stimuli. As the P3a component elicited by Lines and Clock Irrelevant stimuli showed similar waveforms and peak latencies, P3a peak latency was determined on the overall mean ERP waveforms averaged across subjects, sessions, and tasks at the Cz electrode site within the range of 280–480 ms after the onset of Irrelevant stimuli (collapsed localizer method). P3a amplitude was evaluated in a 100 ms wide time window centered at averaged peak latency. Mean amplitude values were analyzed using Session (1:9) × Task (Lines, Clock) × Region (Frontal: F3, Fz, F4; Central: C3, Cz, C4; Parietal: P3, Pz, P4) × Laterality (Left: F3, C3, P3; Midline: Fz, Cz, Pz; Right: F4, C4, P4) rANOVA. To assess the effect of visual frame in the Clock task, a separate rANOVA was conducted with Session (1:9) × Visual Frame (Frame, NoFrame) × Region (Frontal: F3, Fz, F4, Central: C3, Cz, C4, Parietal: P3, Pz, P4) × Laterality (Left: F3, C3, P3, Midline: Fz, Cz, Pz, Right: F4, C4, P4) as independent factors.P3b peak latencies and peak amplitudes were computed for Probe stimuli. Since P3b latency was significantly different between the Lines and the Clock task, latency ranges for mean amplitude measurements were defined separately for the two tasks. Using the collapsed localizer method, P3b peak latencies were identified in the grand-mean ERP waveforms averaged across subjects and sessions for the Lines and Clock tasks separately at the CPz electrode site within 350–700 ms after the onset of Probe stimuli. P3b peak latency was 468.7 ms for the Lines, and 568.4 ms for the Clock task. P3b amplitude was evaluated in a 100 ms time window centered at peak latency. Mean amplitude values were analyzed using a Session (1:9) × Task (Lines, Clock) × Region (Frontal: F3, Fz, F4; Central: C3, Cz, C4; Centroparietal: CP3, CPz, CP4; Parietal: P3, Pz, P4) × Laterality (Left: F3, C3, CP3, P3; Midline: Fz, Cz, CPz, Pz; Right: F4, C4, CP4, P4) rANOVA. The effect of visual frame in the Clock task was evaluated using Session (1:9) × Visual Frame (Frame, NoFrame) × Region (Frontal: F3, Fz, F4; Central: C3, Cz, C4; Centroparietal: CP3, CPz, CP4; Parietal: P3, Pz, P4) × Laterality (Left: F3, C3, CP3, P3; Midline: Fz, Cz, CPz, Pz; Right: F4, C4, CP4, P4) rANOVA.Greenhouse–Geisser correction was applied for all repeated measures with greater than 1 degree of freedom. Uncorrected degrees of freedom and corrected p values are reported. Partial eta squared (η2p) was computed as an estimate of effect size.In case the omnibus ANOVA showed significant Session main or cross effects, planned contrasts were performed to compare sessions grouped into in-flight, early and late post-flight with the pre-flight baseline. The difference between the two in-flight measurements was also investigated with contrasts. The same set of contrasts was applied on all outcome variables. Any effects involving the Region or Laterality factors in the ERP analysis were treated as post-hoc effects and were specified by Tukey-HSD tests. Given that tests of normality have low power with small sample size, all a priori contrasts were checked with a non-parametric alternative (Friedman test), which lead to parallel results in all cases. |
Study outcome: Impaired visuospatial task performance in space
Task performance in the Lines and the Clock task. (A) Mean reaction times (corrected for practice effect). Reaction times were significantly slower during space travel compared to pre-flight. Reactions remained slower during early post-flight, but returned to pre-flight levels for late post-flight sessions. (B) Mean task accuracy. Accuracy decreased in space in the Clock task. Despite the similar trend, accuracy in the Lines task remained unchanged in space. Accuracy returned to pre-flight levels for late post-flight sessions in both tasks. Error bars represent within-subjects standard error of mean (SEM). *P < 0.05.
Involved assay protocols: Treatment protocol
Study outcome: In both tasks P3a was evoked by Irrelevant stimuli. P3a amplitudes decreased considerably in spaceflight.
Grand mean ERP elicited by Irrelevant stimuli, presented on the Cz electrode.
Involved assay protocols: Treatment protocol
Study outcome: P3b was evoked by Probe stimuli in both the Lines and Clock task.
Grand mean ERP elicited by Probe stimuli, presented on the CPz electrode.
Involved assay protocols: Treatment protocol
Study outcome: Similarly to P3a, P3b amplitudes also diminished during spaceflight.
Amplitude of P3a and P3b elicited by Irrelevant and Probe stimuli, respectively, averaged over electrodes. P3a and P3b amplitudes showed similar spaceflight related alterations, as both components decreased significantly during in-flight compared to pre-flight. No differences were present between the two in-flight sessions. Compared to pre-flight, amplitudes remained decreased throughout the post-flight period. Error bars represent within-subjects SEM22. **p < 0.01; *p < 0.05.
Involved assay protocols: Treatment protocol