Subjects

We studied 13 crew members who agreed to participate in Flight Study E377, “Cardiac Atrophy and Diastolic Dysfunction During and After Long Duration Spaceflight: Functional Consequences for Orthostatic Intolerance, Exercise Capacity, and Risk of Cardiac Arrhythmias,” conducted by NASA. As part of this investigation, detailed imaging studies were prospectively designed and conducted on astronauts before space flight with various tilt angle settings to simulate changes with gravitational stress. All astronauts were in excellent health, as determined by comprehensive NASA class III physical examinations, but were not elite athletes. No subject smoked or used any cardiovascular medication regularly. This research conformed to the Declaration of Helsinki, and all subjects signed an informed consent form approved by the NASA Human Subjects Review Committee and institutional review boards of the principal investigators’ institutions.

Tilt Study

Experiments were performed 79 ± 19 days (median, 78 [interquartile range, 66, 94]) prior to the launching day. Subjects refrained from high-intensity exercise and took no over-the-counter medication within 24 hours of all studies; no exercise of any sort was allowed 4 hours prior to study. Subjects were studied at least 2 hours after a meal; no alcohol was allowed for 24 hours or caffeine for 12 hours prior to study, and subjects were encouraged to get at least 6 hours of sleep the previous night. Experiments were conducted in a quiet, environmentally controlled laboratory with an ambient temperature of 25°C at Johnson Space Center, Houston, TX. Four different tilt angles were used to mimic varying degrees of gravitational stress relevant for spaceflight and given by the sine of the tilt angle ( Figure 1 ): 0° (supine, simulating microgravity of space), 22° head-up tilt (0.38 G, simulating Martian gravity), 41° head-up tilt (0.66 G, simulating the approximate peak G load predicted to occur during landing on a planetary surface or near-Earth object), and 80° head-up tilt (1G, effectively full Earth gravity). Following instrumentation, subjects rested at least 15 minutes in the supine position, after which supine echocardiography was performed. At each angle above 0°, echocardiography was begun 2 minutes after reaching that position and lasted for no more than 11 minutes. Based on our pilot data, using invasive measures of cardiac filling pressure during changes in hydrostatic gradients and monitoring reflex responses like muscle sympathetic nerve activity, we have observed that filling pressure (right atrial and pulmonary capillary wedge pressure) stabilizes within 30 seconds to 1 minute, and reflex responses (blood pressure [BP], heart rate, and muscle sympathetic nerve activity) are stable within 1-2 minutes. Images from all participants were acquired in the same, progressive order (0°, 22°, 41°, and 80° head-up tilt) without returning to zero angle prior to the next set of images ( Figure 2 ). BPs were assessed at the brachial artery using a commercially available oscillometric cuff monitor (Dynamap, Critikon, Tampa, FL).

Progressive upright tilt test and gravitational force. The hydrostatic gradient from head to foot (Gz) is equivalent to the sine of the tilt angle. At supine (θ = 0°), sin θ = 0, that is, the absence of gravity in Gz direction. G , Gravitational acceleration.

Four-chamber views from an astronaut during tilt test. (A) 0°, (B) 22°, (C) 41°, and (D) 80°.

Echocardiography

Echocardiographic examinations were performed by an experienced sonographer (D.S.M.) using a commercially available ultrasound system (iE33, Philips Medical Systems, Andover, MA), equipped with two-dimensional (2D) and three-dimensional (3D) transducers. At least three cardiac cycles were acquired in each of the standard echocardiography views and stored in digital imaging and communications in medicine (DICOM) format and raw data for analysis. Analysis was done by an experienced core lab technician, who was blinded to astronaut and condition. Two-dimensional LV end-diastolic (EDV) and end-systolic (ESV) volumes and EF were determined from the biplane method of disks. Three-dimensional EDV, ESV, and EF were analyzed using commercially available software (QLAB, Philips Medical Systems). The off-line software automatically identified the cavity wall interface in the 3D space throughout the cardiac cycle. Manual adjustments were performed, when necessary, to include the papillary muscles and trabeculae in the LV cavity. Biplane left atrial (LA) volume was measured at end systole.

SV was determined from the velocity time integral of the LV outflow tract (LVOT) pulsed-wave Doppler. The early filling (E-wave), the peak late diastolic (A-wave) velocities, and deceleration time were assessed from transmitral flow. The peak systolic (s′), early diastolic (e′), and late diastolic (a′) mitral annular velocities were measured at both the septal and lateral annulus using spectral Doppler and averaged. The ratio of early mitral diastolic inflow velocity to early diastolic mitral annular velocity (E/e′) was calculated. The averaged values from three cardiac beats were reported.

End-systolic BP (0.9 × systolic BP) was calculated as previously described. Total afterload was defined by the effective arterial elastance (Ea; 0.9 × systolic BP/SV). Systemic vascular resistance (SVR), the nonpulsatile component of afterload, was determined by dividing mean arterial pressure/cardiac output × 79.9. Total arterial compliance (TAC) was estimated from the ratio of SV to pulse pressure. Load-independent contractility was assessed using modified single-beat method end-systolic elastance (Ees; determined from BP, SV, and preejection and total systolic periods determined on LV outflow Doppler, EF, and an estimated normalized ventricular elastance at arterial end diastole). Ventricular-arterial coupling was assessed by the coupling ratio (Ea/Ees). Stroke work was defined as SV times mean arterial pressure, and pressure-volume area was estimated as previously reported. The pressure-volume area consists of stroke work and potential energy and correlates linearly with myocardial oxygen consumption.

Myocardial Strain Analysis

Myocardial deformation analysis was performed using a commercially available speckle-tracking software (Cardiac Performance Analysis version 1.1, TomTec Imaging Systems, Unterschlessheim, Germany). DICOM images at native acoustic frame rates (48-86 frames/sec) were analyzed. Three apical views were used to obtain both peak systolic global longitudinal strain (GLS) and strain rate (GLSR), while short-axis views at the papillary muscle level were used for peak systolic global circumferential strain (GCS) and systolic strain rate (GCSR). Analyses of deformation indices were performed as described elsewhere. Briefly, the endocardial border was manually traced, and the software subsequently traced the border in the other frames automatically. In segments with poor tracking, the border was readjusted manually until adequate tracking was achieved. In fact, minor adjustment was required for the majority of the images to achieve optimal tracking quality. The deformation parameters from three cardiac cycles were averaged.

Observer Variability

To assess inter- and intraobserver variability, we randomly chose six subjects. Each of two observers (Z.B.P. and A.B.) obtained two separate measurements on each subject, for a total of four measurements per subject. Absolute intraobserver variability was calculated as a difference between the first and second measurement performed by the same observer on the same subject. Intraclass correlation coefficient was also calculated. The analogous procedure was used to assess interobserver variability, except that the two measurements obtained on the same subject were taken by different observers. Observer variabilities were expressed as average ± SD. To assess variability in volumetric assessment, we chose subjects who underwent similar exams some months apart and had 2D volumes ( n = 13) or 3D volumes ( n = 6) measured. From these, mean ± SD of the absolute difference in the volumes and the Pearson correlation coefficient are reported.

Statistical Analysis

All continuous variables are presented as mean ± SD, and categorical data are expressed as percentages, unless otherwise specified. Linear mixed model was used to handle the dependencies in repeated measurements within the same subject. Unstructured covariance matrix was chosen because it had the lowest Akaike information criterion. Missing values in biplane LA volumes were imputed from LA area from four-chamber view, systolic and diastolic BPs, and E/e’ using multiple imputation. Two-sided P < .05 was accepted as indicating statistical significance. All data were analyzed using IBM SPSS statistics version 20 (IBM, Chicago, IL) and STATA version 12 (StataCorp LP, College Station, TX).