Long duration spaceflight missions to the International Space Station (ISS) have widespread effects on human neurophysiology and performance. Sensory and motor deficits have been identified post-flight (1–13), and in-flight (14, 15). Cognitive changes due to spaceflight have been more difficult to identify.

Studies have reported decreased attentional capacity, decreased dual task performance, decreased spatial awareness, impaired performance on a visuospatial task, and complaints of “space fog” in microgravity (14, 16–22). Dual task performance has been shown to suffer when crewmembers are initially exposed to microgravity, but it then recovers over the course of weeks to months inflight (19, 20), suggesting that some manner of adaptation occurs. Similarly, deficits in dual-tasking postflight have been observed soon following return to Earth but recovers during the first week following long-duration missions (23). In the NASA Twins Study, the twin that went to space displayed more risk taking on the balloon analog risk test (BART; a decision making test to assess risk-taking behavior) during the entirety of his 1 year in spaceflight and also showed decreased task speed in the emotion recognition task and the digit symbol substitution task, as well as a decline in abstract matching (17). However, in a previous study we also analyzed digit symbol substitution task performance changes from pre-to postflight, finding no effects of microgravity in a larger cohort of astronauts (n = 15) that spent 6–12 months on the ISS (6). Further, we found no changes in spatial working memory performance from pre-to post-flight, but we did see alterations in brain functional connectivity (24) during task performance from pre-to postflight. This suggests that performance may be preserved with recruitment of additional compensatory neural resources post-flight, similar to what we identified post-flight in a sensorimotor adaptation task (25). In sum, some studies show cognitive performance changes with spaceflight whereas others do not. These inconsistencies could be due to a variety of factors including small sample sizes, individual differences, varying flight durations, times tested relative to g-transitions, varying tasks and cognitive domains assessed, and different task difficulty levels across studies.

We have previously assessed the effects of HDBR on cognition, finding increased performance on the card rotation task due to task practice (26). Further, we found decreased functional MRI activity during a spatial working memory task (SWM) within the right inferior frontal gyri and the left dentate nucleus. Moreover, participants that increased activation within the right angular gyrus increased spatial working memory performance, and those with a greater decrease in the inferior frontal gyri showed less declines in SWM accuracy (27). Further investigations have identified a general slowing in cognitive processing due to HDBR, particularly in sensorimotor processing speed (28, 29). Overall, these findings illustrate that even though cognitive performance may be maintained, there may still be underlying brain changes occurring with HDBR. These brain changes may reflect compensation or decreased neural efficiency as a result of HDBR. Recruitment of additional neural resources to maintain task performance levels may reduce the ability for dual tasking or to direct attention elsewhere.

HDBR involves adaptation to a new environment and sensory reweighting due to reduced foot sole and other somatosensory inputs and rotation of the body relative to the gravitational vector (30). Due to this, HDBR serves as a model environment to test integrated countermeasures such as artificial gravity (AG). This can be applied by having participants lay at the end of a short arm centrifuge spinning at a constant angular velocity. During the NASA Artificial Gravity Bed Rest – European Space Agency (AGBRESA) campaign, this velocity was adjusted for participant height and applied such that there was 1Gz at the center of mass (CoM), 2Gz at the feet and 0.3Gz at the head. AG is applied along the long axis of the body, increasing proprioceptive and somatosensory inputs, and stimulating the vestibular system, resulting in improved orthostatic tolerance relative to HDBR without AG (31–33). In a 5 day HDBR pilot study, AG reduced orthostatic intolerance with 30 min of daily AG (33). AG also reduced orthostatic intolerance in the 60 day AGBRESA campaign (76). We found that AG participants in AGBRESA performed better on the paced auditory serial addition task (PASAT, performed during active centrifugation) compared to control participants who performed the task in HDBR with no centrifugation, while other cognitive tasks showed no effects of AG (34). This was further supported by work showing learning effects (faster reaction time and increased accuracy) during simple and complex visual and auditory tasks accompanied by changes in electrocortical responses that may be related to impaired attentional processing, however, there were no counteractive effects of AG (35). Another study identified an overall slowing of cognitive performance across the AGBRESA campaign, including prolonged time to identify facial expressions during HDBR and an overall decrease in sensorimotor processing speed; these negative effects were not mitigated by AG (29).

Our primary aim was to examine whether AG applied along the long axis of the body via centrifugation would mitigate HDBR-induced changes in cognitive performance and associated brain activity. We administered the same spatial working memory and motor-cognitive dual task tests that we have previously used in our prior HDBR and spaceflight studies (26, 36–44, 77, 78). We found that during HDBR, there was increased functional brain activation during dual tasking across a large network containing frontal, parietal, cingulate, temporal and occipital cortices (45). In another HDBR study combined with elevated CO₂ we found increased dual task cost effects in the middle temporal gyrus (78). We also observed decreased activation in the right middle frontal gyrus and left cerebellum while participants performed a SWM task in HDBR. Moreover, participants that improved SWM performance decreased their brain activity throughout HDBR, suggestive of practice effects (27).

Cognitive and sensorimotor declines that occur in HDBR and microgravity such as those that we have described here have been argued to stem from disrupted sensory inputs, sensory reweighting, and interference from adaptation to the microgravity environment. This study is a continuation of our previous efforts (34) to investigate the use of AG to counteract cognitive and sensorimotor brain and behavior deficits due to HDBR. In that study, we assessed overall behavior along the timelines as here and found that AG had minimal effects of cognitive performance, however, those that received AG had better performance on the PASAT, while other cognitive measures were largely unaffected. Functional mobility and balance performance showed significant effects due to HDBR, but not AG, but there was a large difference between those that received AG and those that did not in balance performance (34). Here, we hypothesized that those receiving daily AG would have reduced changes in brain activity during dual tasking during HDBR as compared to controls (in HDBR but receiving AG). Moreover, we predicted that this decreased activation would be associated with better cognitive performance. This would be suggestive of greater neural efficiency in the AG group.

To assess significance, we used a family wise error (FWE) corrected p value of FWE < 0.05 and voxel extent of k˃10. Our initial analyses identified no significant or trending differences in brain activation or behavior between the two AG groups (intermittent versus continuous). We therefore pooled them together to increase statistical power.

We investigated whether AG would mitigate HDBR effects on brain changes that occur during cognitive-motor dual-tasking or during SWM task performance. We observed no effect of AG on brain changes in single tasking, nor dual-task cost. We identified an initial increase followed by decreased activation within the right pre-supplementary motor cortex that was specific to the HDBR phase during dual-tasking (Figure 3). We also identified that during the SWM task, the right precuneus, the right postcentral gyrus and the left paracentral lobule increased activation once participants had entered the HDBR environment; this effect remained constant throughout HDBR and was resolved by 10 days post-HDBR. Finally, we identified a dual-tasking brain-behavior change-change correlation in the left postcentral gyrus, in which individuals that increased their brain activation the most maintained their dual-task tapping performance relative to baseline. Taken together, we interpret these findings to reflect that 30 min of AG daily has minimal effects on brain changes underlying cognitive function during HDBR, at least for the duration and dosage used in the AGBRESA campaign.

While we did not identify any counter-active effects of AG on dual-tasking brain activation, we did see increased activation in the right pre-supplementary motor area (preSMA) for the combined groups during dual tasking performance that was specific to the HDBR period (Figure 3). This increase in activation subsided by the end of HDBR, suggesting adaptation occurred over time in HDBR. We observed similar effects in a different HDBR study (45), suggesting that this region may be particularly affected by HDBR. The preSMA has been argued to be important for movement control, particularly for complex actions and cognitive control of movement (59). This region also plays a critical role in stopping action (60–62), suggesting that HDBR may affect cognitive-motor interactions. We further identified a brain-behavior change-change correlation during dual-tasking where those that decreased their activation within the left postcentral gyrus the most maintained their tapping performance relative to baseline at the end of HDBR (Figure 5). The postcentral gyrus comprises of the somatosensory cortex, the primary region associated with sensory perception from the body (63). Interestingly, this correlation would suggest that those that were the most able to decrease activation in a region heavily involved in sensory perception region maintained their pre-HDBR motor performance. This may reflect sensory reweighting. It is not directly clear how this result relates to performance, however, since the cluster is in a region that is substantially inferior to the hand knob location.

Similar to cognitive-motor dual-tasking, AG had no effects on SWM performance and functional brain activation. All participants (AG and CTRL) increased activation in various brain regions during spatial working memory task performance after they entered the HDBR environment. Specifically, we identified increased activation within the right precuneus, the right postcentral gyrus and the left paracentral lobule (Figure 4) that persisted throughout HDBR and resolved by 10 days after participants exited the HDBR environment. The precuneus has been shown to play a role in visuo-spatial and higher cognitive function (64). The precuneus has also been argued to serve as a major node in various functional brain networks, including two subnetworks of the default mode network (65). The default mode network is heavily involved in introspection and rumination, and is considered an anti-task network, exhibiting decreased activity when participants perform a task relative to rest (66). This increased activation of the precuneus could indicate difficulty with network switching between task and default mode networks; similar effects have been seen in older adults (67). The postcentral gyrus receives somatosensory projections via the thalamus (63), as discussed above. The somatosensory cortex is primarily lateralized, with the left hemisphere receiving inputs from the right side of the body (68, 69). Here we saw increased activation in the right hemisphere, while participants were using their right hand. This implies a compensatory response, as compensation typically involves recruitment of bilateral brain regions (70). Due to the reduced somatosensory inputs that occur in HDBR (with the body largely unloaded), bilateral recruitment of the postcentral gyrus may increase the gain of somatosensory signaling. We also observed increased activation in the left paracentral lobule, another region which has been shown to process somatosensory inputs (71). This would also align with previous findings showing that HDBR results in increased functional connectivity between the posterior parietal cortex and somatosensory cortices, and AG was associated with decreased connectivity between these regions, suggesting compensatory responses in somatosensory processing due to HDBR (86). Further, those in the AG group that decreased their connectivity the most had minimal mobility declines from pre-to post-HDBR (86). Overall, these changes in brain activation patterns suggest that the HDBR environment is making this SWM task more difficult to perform, resulting in recruitment of additional brain regions. It is interesting that these are largely in somatosensory cortical regions rather than in frontoparietal areas, typically engaged for SWM tasks (72).

These cognitive assessments were included as part of a large HDBR campaign with funding from American, German and European space associations (NASA, DLR & ESA; AGBRESA). The overall aim was to determine whether AG would serve as an effective integrated countermeasure for mitigating many HDBR changes that occur. In this campaign, another group reported that HDBR resulted in cognitive performance declines, with no benefit of AG (28). We found largely similar results, with the exception of cognitive task performance during AG centrifugation. Those that received AG had better cognitive performance on the Paced Auditory Serial Addition Task (PASAT) (34) than those that performed it during HDBR alone. This, taken in conjunction with the present findings, suggests that AG does not have a strong impact on cognition, but it could be useful for cognitive tasks performed during centrifugation.

The literature suggests that the primary advantages of AG are benefits to sensory and motor behavior. We have shown that AG can increase neural efficiency in visuomotor rotation tasks (83); further, we observed that those who decreased brain activity the most during task performance showed the least HDBR associated performance declines. We also found that the AG group showed greater decreases in resting state functional connectivity with the posterior parietal cortex, a major hub of sensory integration, than controls (86). Further, those in the AG group that showed the largest decreases in connectivity had minimal mobility declines from before to after HDBR (86). We have also shown that HDBR results in balance declines; however, those in the AG group had less declines, as evident by the large effect size even though not statistically significant (34). Thus, while AG ameliorates some of the sensorimotor effects of HDBR, the dosage of 30 min daily may be insufficient to see counter-active effects in all aspects of cognition. Moreover, the centrifugation velocity resulted in 1 Gz at the heart, but only 0.3 Gz at the level of the head. Increasing the dosage to reach levels of 1Gz at the head may result in more pronounced effects of AG on cognition and the brain. However, this would also increase Gz at the heart and the feet and may not be optimal for function in those areas. Other doses/durations of AG may have greater effects on neurocognitive function. It should be kept in mind, however, that findings in HDBR do not completely map onto those observed in the microgravity environment [for review see: Barkaszi et al. (73)].

The sample size of this study is limited, particularly if the AG groups are kept separate. This makes it difficult to identify more nuanced group differences, particularly when investigating cognitive changes with HDBR as effect sizes are generally not large. Moreover, the time spent in HDBR for this campaign was 60 days, whereas a typical mission to the ISS lasts approximately 6 months. We do not currently understand whether or how the duration of HDBR scales to the duration of microgravity, particularly in terms of neurocognitive changes. Sixty days of HDBR does result in many similar effects to microgravity on the musculoskeletal and cardiovascular systems, though. Moreover, the centrifugation velocity in this study resulted in 1 Gz at the CoM, approximately 2 Gz at the foot, and approximately one third Gz at the head. It is possible that the lack of effects here are due to the level of AG at the head not being sufficient to mitigate structural and functional brain changes from HDBR. Future studies should investigate greater levels of Gz at the head to potentially increase neurological benefits. It is also important to note, that the HDBR environment may increase levels of physiological stress, such as with sleep disturbances that could have impacts brain connectivity and cognitive task performance (74, 75). Ultimately, while AG does not appear to have a clear effect on cognitive function in HDBR, further examination is warranted with a greater emphasis on larger sample sizes and varying AG doses.

Here, we investigated whether AG is an effective countermeasure for HDBR induced performance changes in cognitive-motor dual tasking and spatial working memory. We identified no effects of AG on either task, instead finding increased activation in somatosensory brain regions for all participants after they entered the HDBR environment. Further, we identified a brain-behavior correlation in dual-tasking where those that decreased their activation the most in the left postcentral gyrus had the highest level of performance. Overall, these findings would suggest that AG, as applied in this study, is not sufficient to counteract functional brain changes that occur in HDBR, and possibly spaceflight.