Changes in Blood Biomarkers of Brain Injury and Degeneration Following Long-Duration Spaceflight

Abstract

This study investigates the longitudinal course of blood-based biomarkers representing the brain parenchyma in long-duration spaceflight.

Long-duration spaceflight has a widespread effect on human physiology. The past decade first revealed eyeball alterations, and then neuroimaging studies hinted at potentially detrimental effects on the brain.^1,2 Expansion of cerebrospinal fluid spaces occurs at the cost of the gray and white matter compartment. A neurobiological integrity assessment of the brain’s tissues after prolonged exposure to microgravity has never been conducted, to our knowledge. Therefore, we investigated the longitudinal course of blood-based biomarkers representing the brain parenchyma in long-duration spaceflight.

Methods

Blood samples were taken from cosmonauts prior to and after missions (mean duration, 169 days) on the International Space Station from 2016 to 2020. Baseline data acquisition occurred 20 days before launch. Postflight sampling occurred on 1 day, 1 week, and 21 to 25 days after landing. Single-molecule array (Simoa) immunoassay quantification of neurofilament light chain (NfL), glial fibrillary acidic protein (GFAP), total tau, and 2 amyloid-β (Aβ) proteins (Aβ₄₀ and Aβ₄₂) including hemoglobin as a control protein was performed (eMethods in the Supplement). Data on race and ethnicity were not collected. The protocol was evaluated by the Russian Ethical Board (Biomedicine Ethics Committee) at the Institute of Biomedical Problems, Moscow, and the ESA Medical Board as well as the Russian Space Agency. The local Ethical Committee of Ludwig-Maximilians University Munich, in Munich, Germany, also approved the protocol. All individuals gave written informed consent.

Statistical evaluation of the data set was performed using SPSS statistical software, version 26 (IBM). We performed nonparametric testing of the repeated measurements against baseline values with the paired Wilcoxon signed-rank test. A Spearman-Rho correlation analysis between the biomarkers with each other and with mission time (days from preflight baseline) was also conducted. All tests were 2-sided. A value of P < .05 was considered statistically significant.

Results

Five male cosmonauts (mean [SEM] age, 49.2 [2.7] years) were included. NfL was significantly elevated compared with preflight levels directly postflight, 1 week, and 3 weeks after return to Earth (Table). The longitudinal course of hemoglobin as a systemic control protein for the blood-based analysis did not show any significant change or even trend over the same time course using the identical statistical approach in this cohort. For each point, mean (SEM) values of hemoglobin (g/dL; to convert to grams per liter, multiply by 10) are as follows: postflight: 15.7 (0.6); 1 day after landing: 14.2 (0.5); 1 week after landing: 15.5 (1.2); and 21 to 25 days after landing: 14.1 (0.5). GFAP showed an increase at the end of the first week postflight and beyond (mean [SEM] preflight, 169 [43] pg/mL vs 1 week after landing, 215 [31] pg/mL; P < .08 and vs 21 to 25 days after landing, 205 [34] pg/mL; P < .08). Total tau showed a nonsignificant elevation throughout the course of the first week, followed by a significant drop below baseline 3 weeks after return (mean [SEM] preflight, 0.46 [0.06] pg/mL vs 21 to 25 days after return, 0.22 [0.06] pg/mL; P < .04). Aβ₄₀ was significantly increased throughout the follow-up period (mean [SEM] preflight, 101 [17] pg/mL vs 1 day after landing, 156 [16] pg/mL; P < .04; vs 1 week after landing, 183 [18] pg/mL; P < .04; and vs 21 to 25 days after landing, 167 [20] pg/mL; P < .04), while levels of Aβ₄₂ showed a similar but less pronounced increase (mean [SEM] preflight, 5.63 [0.7] pg/mL vs 1 day after landing, 8.43 [1.21] pg/mL; P < .08; vs 1 week after landing, 9.63 [1.4] pg/mL; P < .04; and vs 21 to 25 days after landing, 8.60 [1.4] pg/mL; P < .08) (Figure).

Table. Summary Statistics of the Blood-Based Brain Biomarkers at Different Times.

Characteristic	Mean (SE)		P valuea	Return +1 wk		Return +3 wk
	Preflight	Return +1 d		Mean (SE)	P value	Mean (SE)	P valuea
Protein, pg/mL
Neurofilament light	11.4 (2.3)	15.2 (3)	<.04	17.2 (2.9)	<.04	15.2 (3.8)	<.04
Glial fibrillary acidic proteins	169 (43)	203 (32)	<.23	215 (31)	<.08	205 (34)	<.08
Total tau protein	0.46 (0.06)	0.60 (0.19)	<.69	0.58 (0.11)	<.35	0.22 (0.06)	<.04
Amyloid β 40	101 (17)	156 (16)	<.04	183 (18)	<.04	167 (20)	<.04
Amyloid β 42	5.63 (0.7)	8.43 (1.21)	<.08	9.63 (1.4)	<.04	8.60 (1.4)	<.08

^a P values represent comparison with preflight baseline for preceding time point column.

Figure. Timeline for Neurofilament Light Chain, Glial Fibrillary Acidic Protein, Total Tau, and Aβ₄₂/Aβ₄₀ Ratio Response.

A, Neurofilament light chain protein showed a 33% increase on day 1 and a 50% increase 1 week after return from a long-duration mission compared with preflight measurements. This may represent an axonal injury. B, Glial fibrillary acidic protein showed an increase of 20% and 27% for the same comparisons in time, pointing to a concurrent and sustained astrocytic response. C, Total tau protein as a surrogate for the neuronal tissue showed nonsignificant elevation for the first week postflight with a significant drop of more than 50% below baseline levels 3 weeks after return to Earth. D, Amyloid β (Aβ)₄₀ and Aβ₄₂ increased significantly postflight (Table) with the stronger relative elevation in Aβ_40. Therefore, the course for the ratio of the 2 amyloid proteins, Aβ₄₂/Aβ₄₀, showed a decreasing trajectory after return to Earth from a long-duration mission. This kind of downward trajectory is often viewed as unfavorable with respect to long-term brain health in the general population. The solid line represents the mean group response, while dashed lines show the individual’s parameter course. Error bars give the standard error of mean for each time point. Actual percentage change from preflight baseline indicated for the greatest deviation for the respective parameter.

Correlation analyses showed a significant association of NfL, GFAP, and Aβ₄₀ levels with each other across the participants and times. GFAP was correlated with NfL (r = 0.423, P = .04). GFAP was correlated with Aβ₄₀ (r = 0.499, P = .03), and NfL was correlated with Aβ₄₀ (r = 0.451, P = .046). Aβ₄₀ and Aβ₄₂ were correlated with each other (r = 0.885, P < .001). Each amyloid protein was also significantly correlated with the number of days from mission start (Aβ₄₀: r = 0.697, P < .001; Aβ₄₂: r = 0.672, P < .001).

Discussion

The increases in NfL and GFAP argue for an axonal disintegration process (NfL) along with astrocytic activation (GFAP) during the postflight course of a long-duration mission.³ The increases of both Aβ proteins over the entire postflight phase potentially depict an accumulative association of the cephalad fluid shift with the interstitial tissue. The findings could reflect coherent reparatory processes of intracranial pressure associations from cephalad fluid shift with the brain with subsequent restoration of the blood-brain barrier integrity.⁴ We speculate the elevation of amyloid proteins back on Earth to represent a washout phase after months of hindered protein waste clearance since albumin has been shown to remain stable or even decrease.^5,6 The effective half-life of the parameters in our study (GFAP, 24 to 48 hours; NfL, 1 to 3 weeks; Aβ₄₀, 2 to 4 hours) argues for ongoing reparatory processes postflight. The 3-way association between NfL, GFAP, and Aβ₄₀ illustrates the broadness of the brain tissue response. Our observed changes of blood-based biomarkers indicate brain injury as a previously unknown risk for humans in long-duration spaceflight. Further longitudinal studies in larger samples are needed to characterize the association of long-duration space flight with neurological health.

Supplement.

eMethods