Cardiac adaptations to 60 day head‐down‐tilt bed rest deconditioning. Findings from the AGBRESA study

Abstract Aims Reduced physical activity increases the risk of heart failure; however, non‐invasive methodologies detecting subclinical changes in myocardial function are not available. We hypothesized that myocardial, left ventricular, systolic strain measurements could capture subtle abnormalities in myocardial function secondary to physical inactivity. Methods and results In the AGBRESA study, which assessed artificial gravity through centrifugation as potential countermeasure for space travel, 24 healthy persons (eight women) were submitted to 60 day strict −6° head‐down‐tilt bed rest. Participants were assigned to three groups of eight subjects: a control group, continuous artificial gravity training on a short‐arm centrifuge (30 min/day), or intermittent centrifugation (6 × 5 min/day). We assessed cardiac morphology, function, strain, and haemodynamics by cardiac magnetic resonance imaging (MRI) and echocardiography. We observed no differences between groups and, therefore, conducted a pooled analysis. Consistent with deconditioning, resting heart rate (∆8.3 ± 6.3 b.p.m., P < 0.0001), orthostatic heart rate responses (∆22.8 ± 19.7 b.p.m., P < 0.0001), and diastolic blood pressure (∆8.8 ± 6.6 mmHg, P < 0.0001) increased, whereas cardiac output (∆−0.56 ± 0.94 L/min, P = 0.0096) decreased during bed rest. Left ventricular mass index obtained by MRI did not change. Echocardiographic left ventricular, systolic, global longitudinal strain (∆1.8 ± 1.83%, P < 0.0001) decreased, whereas left ventricular, systolic, global MRI circumferential strain increased not significantly (∆−0.68 ± 1.85%, P = 0.0843). MRI values rapidly returned to baseline during recovery. Conclusion Prolonged head‐down‐tilt bed rest provokes changes in cardiac function, particularly strain measurements, that appear functional rather than mediated through cardiac remodelling. Thus, strain measurements are of limited utility in assessing influences of physical deconditioning or exercise interventions on cardiac function.

obtained through echocardiography or MRI, could detect subclinical changes in myocardial function secondary to bed rest deconditioning. Furthermore, we determined whether artificial gravity through short-arm centrifugation would ameliorate the response.

Methods
This study is part of the NASA/ESA/DLR 60 day À6°headdown-tilt bed rest study 'Artificial Gravity Bed Rest with European Space Agency' (AGBRESA) conducted at the DLR: envihab. The study enrolled 24 healthy persons (23-54 years, 24.3 ± 2 kg/m 2 , eight women). We obtained written informed consent prior to study entry. The study was approved by the North Rhine Medical Association Ethics Committee and prospectively registered (DRKS00015677).
The study comprised 14 day baseline, 60 day strict À6°h ead-down-tilt bed rest, and 15 day recovery. Participants were pseudorandomly distributed to a control group, daily 6 × 5 min short-arm centrifugation with 3 min breaks, or daily continuous 30 min short-arm centrifugation, each with 1 Gz at the centre of mass. Participants did not exercise, were on a controlled sodium diet, and maintained a constant body weight.
We performed echocardiographic and Doppler imaging (Vivid-IQ with M5SC-RS sector probe, GE Healthcare, Boston, Massachusetts, USA) at baseline (supine, 6 days before bed rest) and at the end of bed rest (À6°head-down-tilt, 1 day before recovery) to assess biplane end-diastolic and end-systolic volumes; mitral annulus plane systolic excursion; left ventricular, systolic, global longitudinal peak strain by speckle tracking; transmitral filling patterns [E wave, A wave, E/A, and tissue Doppler of the lateral mitral annulus (e'lat) velocities and ratio]; and stroke volume index (derived from pulsed-wave Doppler velocity-time integral of the left ventricular outflow tract, its diameter, and body surface area).
During passive orthostatic testing at the last day of baseline and on the last day of bed rest, we recorded resting heart rate and blood pressure.
Results are reported as mean ± standard deviation. We calculated group and time point effects using linear mixed-effect
Left ventricular, systolic global circumferential peak strain by cardiac MRI did not change significantly with bed rest ( Figure 3). However, following 4 day recovery, global circumferential peak strain tended to decrease compared with bed rest (P = 0.05; Figure 4). Circumferential contraction expressed as systolic strain rate and time to peak was significantly augmented at Day 56 of bed rest compared with baseline with increases in strain rate and shortened time to peak. While peak values for transmitral A wave did not change with bed rest, E was reduced such that the E/A ratio decreased. We observed a similar pattern for e'lat, whereas E/e'lat remained unchanged.
Artificial gravity through intermittent or continuous centrifugation did not abolish cardiovascular adaptations to headdown-tilt bed rest (Appendix 1).

Discussion
Sixty days of strict head-down-tilt bed rest elicited cardiovascular deconditioning indicated by increases in resting and upright heart rate with reductions in left ventricular end-diastolic volume, cardiac output, and stroke volume. Yet bed rest did not lead to clinical apparent heart failure. Previous studies showed worsened cardiopulmonary fitness and orthostatic tolerance. 8 Yet we did not observe sustained reductions in left ventricular function assessed by systolic strain analysis in line with shorter duration bed rest studies. 9 Finally, myocardial mass did not change significantly, suggesting that cardiac atrophy is not a general feature during physical deconditioning and cannot be seen as risk factor for developing chronic heart failure. While we cannot exclude modest improvements in cardiovascular deconditioning, artificial gravity failed to abolish the response.

Cardiac deconditioning
Strain can be affected by intrinsic myocardial properties, cardiac loading conditions, and sympathetic drive. 10 We and others observed reductions in left ventricular end-diastolic volume with predominant long-axis diameter shortening following bed rest deconditioning. 11 The phenomenon may result from plasma volume reductions during bed rest. 12,13 Plasma volume reductions are at least in part explained by cephalad volume shifts promoting natriuretic peptide release through atrial stretch. [14][15][16][17] The left ventricle seems less compliant with a smaller stroke volume independent of the volume loss. 14 The asymmetric change in left ventricular shape likely explains differential global circumferential and longitudinal strain responses. 10 Normalization of strain and left ventricular volumes within days of recovery is consistent with loading-dependent functional changes rather than cardiac remodelling that might lead to persistent cardiac dysfunction. Left ventricular diastolic filling, which is also preload dependent, changed as well. 18,19 Similar volume   alterations have been reported during 5 and 35 days of bed rest. 13,20 Altered loading conditions may also explain the significant albeit small reduction in left ventricular ejection fraction upon recovery. Cardiac function measurements could be confounded by sympathetic activation, which is an expected physiological response to plasma volume reductions. Indeed, increases in resting heart rate and diastolic blood pressure, which we observed at the end of bed rest similar to others, 13 often occur in conditions associated with increased sympathetic drive. 21,22 Previous findings in bed rest studies support the idea that sympathetic activity is, indeed, increased. [23][24][25] Furthermore, after 21 day bed rest, plasma norepinephrine increased more with orthostasis compared with baseline. 26 We speculate that sympathetic activation may have increased circumferential strain with bed rest. Global circumferential peak strain. Circumferential peak strain measurements at baseline, end of bed rest, and end of recovery in a representative study participant. Upper panel: end-systolic cross-sectional short-axis cardiac magnetic resonance imaging images at the level just above the papillary muscles with circumferential strain overlay. Middle panel: Bull's eye view of the 16 American Heart Association (AHA) myocardial segments model with circumferential peak strain values and colour coding, where deeper blue resembles higher strain values. Lower panel: Circumferential peak strain time course over one heartbeat for the 16 AHA myocardial segments model.

Cardiac deconditioning
The main limitation of our study is the relatively small sample size limiting statistical power and detailed subgroup analyses. Yet rigorous standardization including controlled sodium intake and caloric adjustment to maintain body weight made it possible observing small but relevant physiological changes in cardiovascular function. Furthermore, participants were relatively young with low heart failure risk. Finally, longer periods of limited physical activity may be required to alter intrinsic myocardial properties and to promote interstitial fibrosis.
We conclude that 60 days of À6°head-down-tilt bed rest provoke changes in cardiac function that appear functional rather than mediated through cardiac remodelling. Additional risks such as older age or concomitant cardiovascular disease may be required to express cardiac dysfunction and consecutive chronic heart failure. Because À6°head-down-tilt bed rest is a model for weightless conditions, our findings are reassuring for human space travel. While in weightlessness, cardiopulmonary fitness and orthostatic tolerance will deteriorate in the absence of sufficient countermeasures, overt cardiac disease appears unlikely. Furthermore, our findings might have implications for patients undergoing forced bed rest in, for example, intensive care settings. Finally, our study suggests that strain measurements, as preload-dependent analysis, may be of limited utility in prospectively guiding exercise interventions in the prevention of heart failure. While deconditioning elicits plasma volume reductions and sympathetic activation, physical exercise, particularly endurance training, elicits the opposite response. 27

Cardiac magnetic resonance imaging acquisition parameters
Two-chamber, three-chamber, and four-chamber views and right ventricular long-axis view-cine