The experiment lasted 138 days: 15 days of baseline observation and pre-adaptation, 90 days at a six-degree head-down tilt, and 33 days of recovery after the participants returned to an upright position. Radial pulse graphs were recorded at baseline, on days 7, 15, 30, 45, 60, 75, and 90 of bed rest, and on recovery days 7, 15, and 30. Cardiovascular ultrasound data were collected at corresponding stages. This repeated schedule allowed the investigators to compare not only pre- and post-bed-rest averages, but also the timing of physiological changes during the prolonged exposure and their course after normal gravity resumed.

The study was a controlled simulation, not a spaceflight, and it cannot reproduce the combined effects of radiation, isolation, confinement, disrupted circadian rhythms, and microgravity. Its participants were all healthy men aged 32.0 ± 2.6 years, and each experimental group contained only seven or eight people. The researchers randomly allocated them according to age, height, weight, peak oxygen consumption, maximum squat strength, and femoral-neck bone mineral density. Eight men formed a control group without a countermeasure. The remaining groups received aerobic exercise, resistance exercise, combined aerobic and lower-load resistance training, or combined aerobic and high-load resistance training. Because the study was conducted in healthy young men, its findings cannot automatically be generalized to women, older adults, people with cardiovascular disease, or astronauts in orbit.

Pulse waves were acquired with a digital instrument at the left radial artery, with the forearm relaxed, the palm facing upward, the elbow flexed to about 120 degrees, and breathing kept steady. After the waveform stabilized, its amplitude was maximized and a 30-second pulse graph was recorded. The investigators extracted time-domain features using a combination of an energy envelope and the Hilbert transform. Principal amplitude measures included h1, the main wave associated with large-artery compliance and left-ventricular ejection; h3, a pre-dicrotic wave linked to arterial elasticity and peripheral resistance; h4, the dicrotic notch associated with peripheral resistance and aortic-valve closure; and h5, a dicrotic wave related to large-artery compliance and aortic-valve function. Timing features described one cardiac cycle and the rapid-ejection, systolic, and diastolic phases, while ratios expressed relationships among wave heights and durations.

Most pulse characteristics changed across the baseline, bed-rest, and recovery stages. The main wave h1 rose from an average of 17.41 at baseline to 20.68 during bed rest before falling to 19.40 during recovery. The pre-dicrotic wave h3 increased from 12.02 to 13.52 and then returned toward baseline at 11.52, while the dicrotic notch h4 rose from 6.79 to 8.60 before declining to 6.51. The total cardiac-cycle duration t rose from 0.79 to 0.94 during bed rest and remained somewhat elevated at 0.90 during recovery. In contrast, h3/h1 and w/t declined, and h4/h1 rose during bed rest before falling below its baseline value during recovery. These are group-level patterns and measured differences, not evidence that every individual responded in the same way.

Ultrasound measurements showed that the clearest changes involved vascular structure, arterial wave propagation, blood flow, and resistance rather than conventional cardiac structure or function. Carotid intima-media thickness increased from 0.45 ± 0.08 mm at baseline to 0.50 ± 0.07 mm during bed rest and 0.51 ± 0.07 mm during recovery. Density-calibrated pulse-wave velocity rose from 5.20 ± 0.59 to 5.82 ± 0.83 m/s, while common-carotid pulse-wave velocity increased from 7.40 ± 1.03 to 8.14 ± 1.37 m/s. The maximum velocities measured in the left internal carotid and femoral arteries initially decreased and then rose again during recovery. Resistance indices in the internal carotid, vertebral, and femoral beds generally declined. By contrast, ejection fraction and other cardiac structural and functional measures did not differ significantly across the three broad stages, although repeated-measures analysis detected changes in some ventricular dimensions.

Canonical correlation analysis tested whether the pulse-feature set and ultrasound-indicator set moved together as two broader systems. The two reported canonical correlations were 0.741 and 0.674, both statistically significant at P < 0.05. The ultrasound variables contributing most strongly included carotid and femoral resistance indices and carotid-femoral pulse-wave velocity. Pulse features with higher contributions included h3, h4, t5, As, h3/h1, h5/h1, and w/t. Separate bivariate analyses identified recurring relationships over at least four time points, with coefficients above 0.3, between individual pulse parameters and measures such as carotid-femoral pulse-wave velocity, density-calibrated pulse-wave velocity, internal carotid resistance, and femoral resistance. These are statistical associations, not proof that pulse morphology directly causes a cardiovascular change or can diagnose a specific condition on its own.

The intervention groups followed broadly similar physiological trajectories, but their timing and fluctuations differed. The peak in h1 occurred earlier in all four exercise groups than in the control group—for example, on day 7 in the resistance-exercise group compared with day 15 in controls. Carotid intima-media thickness continued to rise in controls during early bed rest, whereas it stabilized after an initial increase in the intervention groups and subsequently declined after day 30. The combined aerobic and high-load resistance group showed the smallest intima-media fluctuations, while the resistance-only group showed the largest. The control group also had greater fluctuations across most ultrasound indicators. Because these intergroup findings were based largely on time-point trends in small groups, they indicate a possible advantage of combined training rather than establishing a causal or definitive superiority of one exercise protocol.

Taken together, the results support radial pulse-wave parameters as possible dynamic indicators of cardiovascular adaptation during prolonged head-down bed rest. The authors interpret the shared patterns as evidence that changes in vascular elasticity and local hemodynamics can be reflected in outwardly detectable pulse characteristics. The approach is technically promising because it could support repeated, noninvasive monitoring where direct cardiovascular assessment is difficult, but the current evidence comes from a simulated environment and does not establish clinical performance in space. The authors also note that recovery was not uniform: some pulse ratios remained different from baseline even after 30 upright days, indicating that cardiovascular readaptation may proceed in phases. Further studies with larger and more diverse populations, longer follow-up, and more complex environmental simulations will be needed before pulse-based monitoring can be considered a validated space-medicine tool.

Subject of Research: Assessment of cardiovascular adaptation to prolonged head-down bed rest using radial pulse-wave features and ultrasound.

Article Title: Application of radial artery signal characteristics in long-term head-down bed rest experimental cardiovascular status assessment

Article References: Ha, K., Hu, X., Liu, D., Zhang, G., Zhang, S., Zhao, S., Tu, L., Jiang, T., Ma, X., Fan, Q., Liu, J., Zhang, C., Shi, H., Xu, J., & Cui, J. (2026). Application of radial artery signal characteristics in long-term head-down bed rest experimental cardiovascular status assessment. Frontiers in Physiology, 17, Article 1961052. https://doi.org/10.3389/fphys.2026.1961052

Image Credits: AI Generated

DOI: 10.3389/fphys.2026.1961052

Keywords: head-down bed rest, simulated weightlessness, radial pulse waves, cardiovascular monitoring, arterial elasticity, exercise intervention, space medicine, vascular resistance, pulse-wave analysis, Application, radial, artery