Left ventricular remodeling during and after 60 days of sedentary head-down bed rest

Abstract

Short periods of weightlessness are associated with reduced stroke volume and left ventricular (LV) mass that appear rapidly and are thought to be largely dependent on plasma volume. The magnitude of these cardiac adaptations are even greater after prolonged periods of simulated weightlessness, but the time course during and the recovery from bed rest has not been previously described. We collected serial measures of plasma volume (PV, carbon monoxide rebreathing) and LV structure and function [tissue Doppler imaging, three-dimensional (3-D) and 2-D echocardiography] before, during, and up to 2 wk after 60 days of 6° head down tilt bed rest (HDTBR) in seven healthy subjects (four men, three women). By 60 days of HDTBR, PV was markedly reduced (2.7 ± 0.3 vs. 2.3 ± 0.3 liters, P < 0.001). Resting measures of LV volume and mass were ∼15% (P < 0.001) and ∼14% lower (P < 0.001), respectively, compared with pre-HDTBR values. After 3 days of reambulation, both PV and LV volumes were not different than pre-HDTBR values. However, LV mass did not recover with normalization of PV and remained 12 ± 4% lower than pre-bed rest values (P < 0.001). As previously reported, decreased PV and LV volume precede and likely contribute to cardiac atrophy during prolonged LV unloading. Although PV and LV volume recover rapidly after HDTBR, there is no concomitant normalization of LV mass. These results demonstrate that reduced LV mass in response to prolonged simulated weightlessness is not a simple effect of tissue dehydration, but rather true LV muscle atrophy that persists well into recovery.

the removal of gravitational forces by exposure to actual or simulated weightlessness (e.g., head-down-tilt bed rest, HDTBR) causes a relative shift of blood from the lower portions of the body into the thoracic region, neck, and head. This central redistribution of fluid transiently increases venous return, elevating cardiac output and stroke volume (SV) (5, 60). A compensatory cardiovascular unloading, including plasma volume loss (13, 27), decreased red blood cell mass (1, 17), and reduced systemic vascular resistance (47) acts to reduce ventricular wall stress and work load of the heart. Although these initial adjustments likely help protect against pressure overload-induced damage, it is believed that the consequent decreased preload and altered left ventricular (LV) function contribute to orthostatic intolerance (6, 39) and reduced exercise capacity (32, 40, 59) in humans upon termination of bed rest or return to Earth's gravity (1 G) after space flight.

Evidence of altered cardiac performance has been extensively reported after bed rest and space flight. Most notable is a marked decrease in LV SV at rest in response to orthostatic stress, during exercise within hours of a space mission landing (6, 7, 9, 32), and immediately after bed rest (4, 52, 53, 62). For example, using two-dimensional (2D) echocardiography to evaluate systolic and diastolic function, researchers reported that resting supine SV was at least 17% lower after 4- to 8-day missions aboard the Space Shuttle and was due, in part, to a smaller LV end-diastolic volume (9, 44). Investigations examining the mechanisms responsible for the reduction in SV following short-duration space flight have focused largely on adaptive responses that affect cardiac preload (12, 34), such as reduced plasma volume (8, 19). Results from our laboratory have confirmed these findings and further suggest that the degree of cardiac impairment depends on mission length, with larger decrements after prolonged periods of weightlessness (37, 39, 51).

The functional deficits manifested in the human heart after weightlessness likely result not only of smaller ventricular volumes but also cardiac muscle atrophy. Results from early radiographic studies and from conventional 2D echocardiography suggested that cardiac mass in some astronauts and subjects undergoing bed rest was smaller after real and simulated weightlessness (20, 28). Unfortunately, given the poor spatial resolution and geometric assumptions that each method carries, these estimates of morphological changes are considered unreliable. More recently, cardiovascular magnetic resonance (CMR) imaging has proven useful in assessing space flight- and bed rest-induced changes to the heart (14, 24, 50, 54), but its use is currently restricted to pre- and post-space flight measurements and to use during ground-based analogs. Newer matrix-array 3D echocardiography is not burdened by geometric assumptions that are inherent to 2D ultrasound technology, and analysis of LV morphology and function are comparable to that of CMR imaging (43, 65). Investigations into the direct cardiac effects of bed rest using either 3D echocardiography or CMR imaging have reported similar reductions in LV volumes (10, 14, 49). Thus 3D echocardiography may be a useful method for investigating and monitoring cardiac mass during space flight.

Although previous studies have examined the effects of space flight and bed rest on LV mass and volumes, a clear time course of these adaptations has not been fully described and their recovery has been largely ignored. Serial measures of LV mass in the same subjects during bed rest have been conducted in only one study that examined five subjects after 2 and 6 wk but only three subjects after 12 wk of bed rest (49). Thus the primary objective of this study was to quantify the changes in LV volumes and mass in the same subjects during and after prolonged 6° HDTBR using 3D echocardiography. We hypothesized that without countermeasures, subjects would experience LV atrophy during 60 days of HDTBR that would partially recover during reambulation and reconditioning. Additionally, we sought to determine whether any relationship existed between changes in LV mass and changes in plasma volume during and after HDTBR. Plasma volume losses in response to short-duration space flight and pharmacologically induced diuresis may explain the majority of measured decreases in LV mass as a result of a reduction in the fluid compartment of cardiac tissue (55, 56), but reduced LV mass after prolonged bed rest and space flight likely result from the combined effects of cardiac muscle atrophy and decreased tissue fluid. Furthermore, although evidence from several studies investigating cardiac-specific consequences of short-duration bed rest and space flight suggest that most adaptations are resolved within hours of gravitational reloading, we hypothesized that LV mass would lag behind the rapid recovery of plasma volume following prolonged bed rest.

METHODS

Subject selection.

Subjects were recruited by the Johnson Space Center Human Test Subject Facility and passed health and psychological screenings in addition to a National Aeronautics and Space Administration (NASA)-modified Air Force Class III physical examination before they were accepted into the study. All subjects were without history of cardiovascular disease or metabolic disorders and had not smoked within 6 mo before the start of the study. Additionally, each subject was evaluated for recent substandard nutritional status and use of medication that might interfere with the study results. Subjects received verbal and written explanations of bed rest and test protocols before they provided written informed consent. The study protocols were reviewed and approved by the NASA Johnson Space Center Committee for the Protection of Human Subjects, the University of Texas Medical Branch (UTMB) Institutional Review Board, and the UTMB General Clinical Research Center Advisory Committee.

Study design.

The study design and general bed rest conditions were described in detail previously (38). Briefly, all subjects were housed and supervised in the Flight Analog Research Unit at the UTMB Children's Hospital. After an 11- to 14-day study lead-in for acclimation and dietary/fluid stabilization, subjects were maintained in strict 6° HDTBR for 60 days. Subjects were required to meet minimum daily water consumption (28.5 ml/kg), but allowed to drink ad libitum beyond the minimum. Each subject received a standardized diet of ∼2,500 kcal/day that averaged <3,500 mg of sodium/day. Caloric intake was adjusted as needed to maintain body weight within ±3% of the weight measured on the third day of HDTBR. Echocardiography, plasma volume, and hemodynamic data collection before bed rest and during recovery was performed with subjects in the supine position, whereas data collection during bed rest was performed in the 6° head-down position. These measurements were performed after at least 20 min of quiet rest.

After 60 days of HDTBR were completed, subjects remained in the hospital for an additional 14 days for physical reconditioning before exiting the study. During the post-HDTBR reconditioning period, subjects participated in a daily 1-h program of supervised ambulation and exercise beginning on the fourth day of recovery. The intensity, duration, and complexity of the exercises increased throughout the reconditioning period according to subject tolerance of foot tenderness and knee and ankle pain secondary to prolonged HDTBR.

Echocardiography.

Two- and three-dimensional echocardiograms (iE33; Philips Ultrasound, Bothell, WA) were performed by registered sonographers following the guidelines of the European Association of Echocardiography/American Society of Echocardiography (30) with subjects in the left lateral decubitus position while supine (before and after HDTBR) or −6° head-down (during HDTBR). Images were obtained before HDTBR (baseline), during HDTBR on bed rest day 7 (BR7), BR21, BR31, and BR60, and during recovery after 4 h after reambulation (BR+4 hr), on R+3, and on BR+13. Images were stored digitally for offline analysis using commercially available software. Each image was independently analyzed by two sonographers. For all ultrasound measures, a difference in analysis of greater than 10% between sonographers was flagged for reanalysis by a third sonographer. All cardiac measurements were collected after a standardized low-fat meal and within a 2- to 4-h window across days to minimize circadian variability.

Three-dimensional echocardiography data were acquired using an X3-1 broadband xMATRIX array transducer. The wide-angled acquisition was obtained from four consecutive cardiac cycles during held respiration with acquisition triggered to the R wave and harmonic mode off. Imaging was optimized to include the entire LV cavity within the pyramidal scan volume to ensure accurate assessment of volumes. The analysis of 3D images was performed offline using reconstructive analysis software (QLab Version 7.1; Phillips) (23). LV mass was calculated as the product of myocardial volume (the difference between the epicardial and endocardial borders) and myocardial specific density (1.05 g/ml). Additional parameters calculated from the LV volumes included SV, cardiac output (CO), and ejection fraction (EF). SV was calculated as LVEDV − LVESV, cardiac output = heart rate × SV, and EF = (LVEDV − LVESV)/LVEDV.

Tissue and pulsed-wave Doppler imaging.

Standard 2D imaging and tissue Doppler echocardiography was performed using an S5-1 transducer to assess LV function as previously performed by our laboratory (37, 54). Estimates of LV filling pressure were calculated as the ratio between early transmitral flow velocity (E) and early diastolic velocity of the mitral valve annulus (E′) (25, 45, 46).

Plasma volume.

Plasma volume was measured five times during the study (before the start of HDTBR [baseline]; at 21 [BR21], 31, [BR31], and 60 days of HDTBR [BR60], and after 3 days of reambulation [R+3]) using the carbon monoxide rebreathing technique as previously reported by our laboratory (51, 61). Briefly, the subjects breathed 100% oxygen for 2 min through a closed breathing circuit fitted with a two-way nonrebreathing valve (Model 2700; Hans Rudolph, Kansas City, MO). Carbon monoxide (60 ml) was then added to the system and was rebreathed for 10 min. Before and after injection of carbon monoxide, a 3-ml blood sample was collected for measurement of hemoglobin, carboxyhemoglobin, and hematocrit. Total blood volume, red blood cell volume, and plasma volume were calculated (35). Similar to echocardiography, plasma volume was measured in subjects while they were supine (before and after HDTBR) or −6° head-down (during HDTBR).

Cycle ergometry testing.

Peak exercise oxygen consumption (V̇o_{2 peak}) was measured using upright cycle ergometry administered four times during the course of participation in this study. The first test performed served as a familiarization session and was performed 10-12 days before the start of HDTBR. V̇o_{2 peak} was measured again 7 days before the start of HDTBR, which served as the baseline measurement. After HDTBR, testing was conducted either at the end of the first day or in the morning after 1 day of reambulation. The fourth test was conducted ∼10 days after the last day of HDTBR. The exercise test protocol consisted of 3-min stages at either 50, 75, and 100 W or 50, 100, and 150 W followed by 1-min stages increasing 25 W per stage until a subject reached volitional fatigue or he or she could no longer maintain the desired pedaling cadence.

Data analysis and statistics.

Our main dependent variables (except for plasma volume and V̇o_{2 peak}) were assessed once before HDTBR, four times during HDTBR (BR7, BR21, BR31, and BR60), and three times after HDTBR (R+4 hr, R+3, and R+13). Separate mixed-effects linear regression models were used to evaluate the effects of HDTBR on each of our dependent variables of interest using dummy-coded beta coefficients for time allowing us to compare differences relative to baseline. Bootstrap resampling was performed to improve estimates of variance given the small sample size (16). Subject body surface area, calculated using the Mosteller formula (43), was included as a fixed covariate as needed to improve model precision for 3D echocardiography measures. We used the Holm (26) correction for multiple comparisons between baseline and each bed rest and recovery time point. Summary data are expressed as means ± SE unless otherwise stated. All statistical analyses were performed using Stata/IC software (v13.1; StataCorp LP, College Station, TX) and setting two-tailed alpha to reject the null hypothesis at 0.05.

RESULTS

Study population.

Eight subjects (4 men and 4 women, age 36 ± 8 yr) volunteered to participate in this study. One subject (a woman) was excluded from the analysis due to poor acoustic windows. Selected subject characteristics of the remaining seven subjects are presented in Table 1. Plasma volume (−16%) and V̇o_{2 peak} (−31%) decreased following HDTBR (P < 0.0001 for both as measured on BR60). Mean body weight, although statistically lower than it was before bed rest, remained within ±3% of BR3 body weight.

Table 1.

Subject characteristics measured before, during, and after bed rest

	Baseline	7 Days	21 Days	31 Days	60 Days	R+4 hr	R+3	R+13
Weight, kg	71.2 ± 2.8	70.5 ± 2.9*	70.0 ± 2.8*	69.8 ± 2.9*	69.9 ± 2.9*	70.2 ± 2.9*	71.2 ± 3.0*	71.4 ± 3.0
BMI, kg/m2	22.6 ± 0.9	22.4 ± 0.9*	22.2 ± 0.9*	22.2 ± 0.9*	22.2 ± 0.8*	22.3 ± 0.8*	22.6 ± 0.9	22.6 ± 0.9
SBP, mmHg	114 ± 2	111 ± 2	114 ± 2	109 ± 2	114 ± 3	114 ± 3	114 ± 3	112 ± 3
DBP, mmHg	65 ± 2	64 ± 2	69 ± 2	66 ± 3	65 ± 2	65 ± 2	69 ± 3	67 ± 3
MAP, mmHg	81 ± 1	80 ± 2	84 ± 2	80 ± 2	82 ± 2	82 ± 2	84 ± 3	82 ± 2
PV, liter	2.7 ± 0.4		2.3 ± 0.1*	2.3 ± 0.1*	2.3 ± 0.1*		2.7 ± 0.1
V̇o2peak, l/min	2.1 ± 0.1				1.4 ± 0.4*			1.8 ± 0.4

Values are means ± SE. R+4 hr, 4 hours after reambulation; R+3, 3 days after reambulation; R+13, 13 days after reambulation;

BMI, body mass index; SBP, systolic blood pressure; DBP, diastolic blood pressure; MAP, mean arterial pressure; PV, plasma volume.

^*Statistically different than before bed rest and after correction for multiple comparisons.

LV morphologic and functional responses to bed rest deconditioning.

Before bed rest, 3D echocardiographic measurements of LV mass and LVEDV, but not LVESV, were statistically lower in women compared with men. However, changes in these cardiac indices with bed rest and adjusted for body surface area were similar between sexes (group data by sex not shown, Fig. 1). Compared with before bed rest, LVEDV was significantly reduced as early as BR21 (−6%; Table 2) and continued to decline throughout the duration of HDTBR. In contrast, LVESV was not significantly different than pre-HDTBR values until BR60, when both LVEDV and LVESV were reduced by ∼15% (P < 0.001 and P = 0.004, respectively). LV mass, like LVEDV, was significantly lower by BR21 (P = 0.007), reaching a nadir of ∼14% (range, −22.9 to −8.1%; P < 0.001) or a loss of ∼20 g by BR60 (Figs. 2 and 3).

Fig. 1.

Left ventricular (LV) end systolic and LV end-diastolic volume at baseline, after 60 days of head-down tilt bed rest (HDTBR), and after 14 days of active recovery in seven subjects. Bars represent mean group values; lines represent individual subject responses.

Table 2.

Resting hemodynamic and ventricular parameters before, during, and after bed rest

	Baseline	7 Days	21 Days	31 Days	60 Days	R+4 hr	R+3	R+13
HR, bpm	68 ± 5	66 ± 2	67 ± 3	72 ± 2	74 ± 2	86 ± 5*	78 ± 3	77 ± 2
SV, ml	72.3 ± 3.8	72.5 ± 4.6	69.8 ± 3.3	67.6 ± 2.8	60.8 ± 4.4	61.7 ± 4.3	66.8 ± 3.3	76.6 ± 5.4
CO, l/min	5.0 ± 0.4	4.6 ± 0.4	4.7 ± 0.2	5.0 ± 0.2	4.5 ± 0.3	5.1 ± 0.3	5.2 ± 0.4	5.9 ± 0.4
EF, %	62.8 ± 1.4	64.9 ± 1.6	64.2 ± 0.8	64.3 ± 1.0	63.1 ± 1.9	62.8 ± 1.3	62.2 ± 1.3	64.7 ± 2.2
LVESV, ml	41.7 ± 2.6	38.7 ± 1.9	38.8 ± 1.7	37.4 ± 1.6	35.0 ± 2.1	36.2 ± 2.0	40.9 ± 2.6	42.5 ± 4.2
LVEDV, ml	115.1 ± 5.0	111.2 ± 5.4	108.7 ± 4.6*	105.2 ± 3.8*	95.8 ± 5.0*	97.9 ± 5.6*	107.7 ± 5.3	119.0 ± 8.7
LVM, g	140.4 ± 7.9	134.0 ± 6.9	126.6 ± 6.4*	124.0 ± 5.5*	120.1 ± 6.2*	121.7 ± 7.2*	124.6 ± 7.9*	132.5 ± 9.2

Values are means ± SE.

HR, heart rate; bpm, beats per minute; SV, stroke volume; CO, cardiac output; EF, ejection fraction; LVESV, left ventricular end-systolic volume; LVEDV, left ventricular end-diastolic volume; LVM, left ventricular mass.

^*Statistically different than before bed rest and after multiple comparisons correction.

Fig. 2.

Absolute LV mass at baseline, after 60 days of HDTBR, and after 14 days of active recovery in seven subjects. Bars represent mean group values; lines represent individual subject responses.

Fig. 3.

Percent change in LV mass from baseline with 95% confidence intervals, measured across 60 days of HDTBR and 14 days of physical reconditioning using three-dimensional echocardiography. *P < 0.01 compared with values before bed rest.

SV was unchanged during the first 31 days of HDTBR largely as a function of a preserved LVESV. By BR60, however, SV was reduced by −16 ± 4% (P < 0.001). Neither CO nor EF changed during bed rest. There was a marked impairment in early and late diastolic filling velocities resulting in a significantly reduced Doppler E/A ratio (−39 ± 11%, Table 3) by the end of HDTBR. Interestingly, the mitral peak Doppler E-wave, but not A-wave, was significantly lower at BR21 and steadily fell up to BR60. The reduction in the E/A ratio by BR60 coincided with a significant increase (14.2 ± 2.4 ms) in isovolumic relaxation time (IVRT). The ratio of transmitral flow velocity (E) vs. peak annulus velocity during early filling (E′) is an estimate of LV filling pressure. E/E′ was lower by ∼19% at BR60 (P = 0.019), although it did not reach statistical significance after adjustment for multiple comparisons.

Table 3.

Tissue Doppler parameters before, during, and after bed rest

	Baseline	7 Days	21 Days	31 Days	60 Days	R+4 hr	R+3	R+13
IVCT, ms	45.6 ± 2.4	45.8 ± 2.3	51.6 ± 1.4	46.6 ± 1.3	45.3 ± 2.9	50.9 ± 3.7	48.3 ± 3.1	44.9 ± 2.4
IVRT, ms	85.3 ± 2.8	93.7 ± 3.6	95.0 ± 4.4	93.8 ± 4.7	99.5 ± 1.8*	94.0 ± 3.3	88.6 ± 2.8	86.1 ± 3.0
E, cm/s	76.0 ± 6.9	65.1 ± 5.8	61.4 ± 6.0*	59.3 ± 3.4*	53.3 ± 5.6*	54.8 ± 6.7*	73.7 ± 4.1	80.6 ± 5.0
A, cm/s	44.2 ± 2.3	43.3 ± 1.7	45.2 ± 1.6	52.2 ± 3.0	51.5 ± 1.8	52.1 ± 4.0	57.0 ± 2.3*	54.5 ± 3.2*
E/A ratio	1.72 ± 0.13	1.52 ± 0.14*	1.35 ± 0.12*	1.17 ± 0.09*	1.05 ± 0.12*	1.05 ± 0.10*	1.30 ± 0.08*	1.49 ± 0.08*
E′, cm/s	15.5 ± 1.4	13.4 ± 1.0	13.2 ± 1.3	11.9 ± 0.6*	13.6 ± 0.9	14.0 ± 0.8	13.7 ± 0.7	14.6 ± 0.8
E/E′ ratio	5.04 ± 0.32	4.97 ± 0.39	4.74 ± 0.32	5.02 ± 0.21	3.90 ± 0.24*	3.92 ± 0.47*	5.41 ± 0.17	5.54 ± 0.23

Values are means ± SE.

IVCT, isovolumic contraction time; IVRT, isovolumic relaxation time; E, peak mitral flow velocity during early filling; A, peak mitral flow during atrial systole; E′, peak annulus velocity during early filling.

^*Statistically different than before bed rest and after multiple comparisons correction.

LV morphological and functional responses to reconditioning after bed rest.

Bed rest-induced changes in LVESV and LVEDV were partially restored to pre-HDTBR volumes by R+3, coinciding with the restoration of plasma volume. Unlike the recovery of LV volumes, LV mass remained significantly lower at R+3 (P < 0.001) compared with pre-HDTBR values. By R+13, LV mass was no longer statistically different from pre-HDTBR levels (6.1 ± 4.1%, P = 0.18) (Figs. 2 and 3).

Resting heart rate increased gradually throughout HDTBR but it was not significantly higher until R+4 hr. Although CO was largely unchanged during HDTBR, it gradually increased across the recovery period. By R+13, although CO was not statistically different (P = 0.058) from before HDTBR, it was 19 ± 13% higher. At R+3, Doppler measures of peak E-wave velocity were similar to values before bed rest. Interestingly, however, peak A-wave velocity increased during recovery to the extent that the measure of late diastolic filling was highest at R+3 (P < 0.001) and remained significantly higher (difference of 10.4 ± 3.4 cm/s) at R+13. Estimates of LV filling pressure (E/E′ ratio) trended higher than before HDTBR throughout recovery but did not reach statistical significance (Table 3).

DISCUSSION

We examined the course of changes in LV volume and mass during 60 days of strict sedentary HDTBR and throughout 14 days of ambulatory reconditioning. The findings from this study 1) confirm previous reports of altered cardiac function and LV atrophy during prolonged bed rest and extend the methods of study to 3D echocardiography using serial measurements from the same subjects; 2) demonstrate that in this population of test subjects cardiac mass and function recover within 2 wk following HDTBR when performing a progressive reconditioning program; and 3) establish that the loss in LV mass, particularly after prolonged weightlessness, cannot be fully explained by plasma volume loss alone, which had been suggested following shorter periods of weightlessness (55, 56). Resting measures of LV volumes and function in our subjects were quickly restored with the normalization of plasma volume (by R+3), but readaptation of LV mass was much slower, indicating true cardiac muscle atrophy.

Long-duration cardiac deconditioning and LV adaptation.

An important feature of the current study design was the longitudinal assessment of cardiac structural and functional adaptations over the bed rest period in the same set of subjects studied under the same conditions. Using 2D and 3D echocardiography our data demonstrate that 60 days of sedentary HDTBR induced LV dysfunction (e.g., lower E/E′, E/A, and IVRT), which was accompanied by ventricular remodeling and progressive cardiac atrophy. The morphological adaptations observed in the current study are consistent with those of previous research (Table 4) using both CMR imaging and 3D echocardiography (10, 14, 24, 49, 54), although generally, those studies obtained only a single measure at the end of bed rest. In the only other bed rest study to obtain serial measurements, researchers found that in as little as 2 wk of horizontal bed rest, five male subjects displayed a significant decrease in LVEDV, which was accompanied by an 8% reduction in LV mass by week 6; an additional 8% loss was observed in three subjects who remained in bed until week 12 (49). The role that those structural adaptations contribute to changes in LV function remains unclear. However, it has been hypothesized that these atrophic alterations underlie changes that affect LV chamber filling characteristics, including a less distensible ventricle resulting in a leftward shift on the pressure-volume curve (33), impaired diastolic suction (15), and abnormalities of myocardial deformation (29).

Table 4.

Studies quantifying changes in ventricular mass and function during bed rest without countermeasures

	Bed Rest Days	Subjects	LVM	LVEDV	SV	Bed Rest Posture	Test Posture
Perhonen et al., 2001 (49)	14	5 men	MRI	MRI	RB	Horizontal	Horizontal
Levine et al., 1997 (33)	14	11 men, 1 woman	2D	2D	RB	−6° HDT	Horizontal
Shibata et al., 2010 (53)	14	6 men, 1 woman	MRI	2D	RB	−6° HDT	Horizontal
Dorfman et al., 2008 (15)	18	14	MRI	MRI	NA	−6° HDT	Horizontal
Stenger et al., 2014 (54)	21	7 men	2D	2D	PWV	−6° HDT	Horizontal/HDT
Hastings et al., 2012 (24) and Carrick-Ranson et al., 2013 (10)	35	8 men, 1 woman	MRI	3D	RB	−6° HDT	Horizontal
Kozakova et al., 2011 (29)	35	men	2D	2D	PWV	−6° HDT	Not described
Perhonen et al., 2001 (49)	42	5 men	MRI	MRI	RB	Horizontal	Horizontal
Dorfman et al., 2007 (14)	60	8 women	MRI	MRI	NA	−6° HDT	Horizontal
Perhonen et al., 2001 (49)	84	3 men	MRI	MRI	RB	Horizontal	Horizontal

2D, two-dimensional echocardiography; 3D, three-dimensional echocardiography; MRI, magnetic resonance imaging; PWV, pulse-wave Doppler; RB, rebreathing technique; NA, not applicable.

In general, reductions in LV volume during bed rest, particularly diastolic volume, occur rapidly and precede cardiac remodeling. Thereafter, LVEDV appears to decrease at a slower rate as bed rest duration increases, although this has not been a consistent finding. For example, Perhonen et al. (49) reported that LVEDV decreased in the first 2 wk of bed rest but did not decrease further by 6 wk. In contrast, LVEDV decreased progressively throughout 60 days of bed rest in our subjects and was paralleled by a progressive reduction in the mitral E wave. Subjects studied by Perhonen et al. (49) were not in head-down tilt during bed rest and testing, in contrast to the positioning of our subjects. Although this likely resulted in differences in ventricular filling during bed rest between the two groups of subjects and confounds direct comparisons, we had previously adopted the head-down posture for in- and end-of-bed rest tests with the assumption that this more closely represented the inflight condition of astronauts (51). Differences in the positioning of subjects during data acquisition in these two longitudinal studies may contribute to the disparate findings. Alternatively, differences in LVEDV in response to bed rest may have been related to plasma volume status. It has previously been shown that plasma volume is significantly reduced after a 1-wk period of dietary stabilization (62). Subjects who participated in our study followed the NASA standard 2-wk fluid and dietary stabilization period before entering HDTBR. Although we did not assess plasma volume loss during the stabilization period, it is likely that our subjects already had a significant loss in plasma volume at the start of HDTBR and therefore the reduction in LVEDV in the initial stages of bed rest were not as large as previously reported (3, 49).

Largely on the basis of a single study in female bed rest subjects, it has been assumed that the rate of LV atrophy in men and women during bed rest are not different from each other (14). Dorfman and colleagues (14) examined CMR measures of LV mass and volume in women after 60 days of HDTBR and found that both were significantly lower after bed rest compared with pre-bed rest values. Although the authors did not directly compare cardiac atrophy in men and women, the percent decrease in LV mass measured after 60 days of HDTBR appears to agree with the rate of loss reported in men by Perhonen et al. during 84 days of bed rest (49). The current study, in which both men and women were tested in the same bed rest conditions, appears to confirm this assertion, but further studies with a larger number of female participants may further strengthen these conclusions.

Recovery from bed rest.

Subjects participated in a progressive reconditioning program that was intended to return them to a state of conditioning in which they could be safely released from the bed rest facility. Although the overall objective of reconditioning was to allow for successful performance of activities of daily living and was the same for all subjects, the program was individualized for each subject on the basis of their own levels of strength, endurance, aerobic fitness, and neurovestibular disturbances during the recovery period. Subjects progressed from simple ambulation, balance, and body weight exercises to mild stair climbing and weightlifting exercises until the time that they were discharged. The reconditioning program was not designed specifically to promote cardiac hypertrophy, and release from the facility was not determined on the basis of achieving a particular score on any of the physiological tests. However, it is clear that recovery to pre-bed rest levels of LV mass in most subjects is not prolonged when participating in a moderate-intensity programmed reconditioning plan; LV mass after 2 wk of reambulation and reconditioning was not different than it was before bed rest even though the deconditioning during HDTBR was almost four times longer. Regaining cardiac mass likely is predicated on the restoration of cardiac work associated with normal upright activities. However, recovery of LV mass and function after bed rest might be hastened by more intense or more frequent reconditioning (2), but this has not been specifically studied in a systematic manner (31).

Plasma volume and cardiac structure and function after bed rest.

The reduction in cardiac preload, presumably caused by hypovolemia, is considered to be one of the contributing factors behind cardiovascular deconditioning during weightlessness. To examine the independent contribution of plasma volume loss to the observed reduction in LV mass, Summers and colleagues (55) used a low-salt diet combined with an intravenous infusion of furosemide to mimic the dehydration experienced during short periods of weightlessness. The authors reported a significant and positive correlation between plasma volume and LV mass, which was then used to explain the rapid improvements in 2D echocardiographic measurements of LV volume and LV wall thickness observed in Shuttle astronauts 3 days after landing (55). However, in a separate study comparing cardiac adaptations between acute furosemide-induced hypovolemia and bed rest, Perhonen and colleagues (50) showed that changes in both LVEDV and SV were significantly greater after bed rest than with hypovolemia alone. These data demonstrate that the reductions in LV volume and mass that occur as a result of weightlessness only in part are the result of tissue dehydration. Myocardial mass is responsive to changes in loading conditions associated with the transition from normal ambulatory activities to space flight and bed rest, which include reduced cardiac work as a result of decreased physical activity, lack of orthostatic stress, and lower blood pressures (47, 48). Additionally, cardiac remodeling might be induced by reduced regional wall stress resulting from an increased ventricular sphericity during weightlessness (55, 57).

The rate of LV volume and mass recovery has not been compared with that of plasma volume restoration after prolonged bed rest until now. In the present study, resting SV, along with LVEDV and LVESV, increased within 3 days following HDTBR, a rate similar to plasma volume repletion. Although LV volumes were normal soon after HDTBR ended, LV mass remained significantly lower than baseline. The prognostic significance of this slow-resolving atrophy is not apparent in the resting functional measures obtained in this study, but may contribute to the delayed recovery of V̇o_{2 peak} (31, 40).

Space flight applications.

Understanding the cardiovascular changes associated with long-duration space flight to design and implement better countermeasure strategies that preserve astronaut cardiac structure and function are of paramount importance to NASA and its partnering space agencies. Given that structural and functional changes of the myocardium likely contribute to the decrease in orthostatic tolerance (6, 33, 39) and exercise capacity (32, 40) after both short- and long-duration space flight, preventing cardiac atrophy may have profound implications for the performance of mission-critical tasks during exploration missions, on extraterrestrial surfaces, and upon return to Earth. The aggregate of results from this study and others [(14, 24, 49, 53), see Fig. 4] suggest that cardiac functional and morphological adaptations to space flight likely are progressive, and thus countermeasures should be regularly applied throughout the mission. Although it is clear that LV volume and mass can be maintained with prescription of appropriate countermeasures during bed rest (14, 24, 53), and from this study that reambulation and reconditioning can contribute to recovery, it has yet to be proven that LV mass and function can be restored during space flight after a period of reduced participation in countermeasures or inactivity due to mission-specific constraints, failure of exercise hardware, or injury that precludes regular participation in exercise countermeasures.

Fig. 4.

Mean percent change in LV mass and LV end-diastolic volume in bed rest studies of varying duration. See Table 4 for details about each study with regard to bed rest duration, sex of subjects, and measurement methodology.

Additionally, the results from the current study support the use of high-resolution 3D echocardiography for future in-flight medical evaluations and research needs. Although 3D echocardiography is not currently available on the International Space Station, unlike 2D echocardiography (11), it is not burdened by geometric assumptions and therefore provides more accurate estimates of LV mass and volumes, which are comparable with CMR imaging (42, 63). Three-dimensional echocardiography in a compact portable form will overcome many of the technical and cost issues (e.g., volume of space needed, electric power consumption) associated with flying a CMR imager on the International Space Station or during space exploration beyond low Earth orbit (36). Moreover, use of this technology will make it possible to track cardiac adaptations during space flight, to modify or refine countermeasures, and to obtain meaningful results from studies involving fewer subjects, which is desirable for space flight research. Three-dimensional echocardiography has the advantage of having lower intraobserver and interobserver variability than traditional 2D echocardiography (21), and guidance by an ultrasound expert on the ground communicating directly with the astronaut in orbit as they step through an ultrasound examination has been used successfully to acquire on-orbit ultrasound vascular and cardiac images (18, 22).

Study limitations.

First, the number of subjects was relatively small, primarily because of the cost of performing an HDTBR study. To improve our power to detect differences over time, we used multilevel mixed-effects linear regression. Multilevel modeling allows the detection of fixed group differences (analogous to standard regression) while accounting for the variation of within-subject (random) effects such as differences in intercepts and slopes. Second, our study population consisted of both men and women who, on average, were not well conditioned according to their initial low absolute V̇o_{2 peak} values. Although 60 days of HDTBR resulted in further deconditioning and significant losses of LV mass in these subjects, it is likely that even greater remodeling would be observed in more fit subjects, as has been observed in other physiological measures (31). As such, reductions in LV mass observed in this study may not be representative of the expected changes in astronauts, who typically are more fit (40, 41), if they were unable to perform exercise countermeasures during a space flight mission.

Conclusions.

Reductions in cardiac mass and function are progressive during long-duration bed rest in subjects who are not performing countermeasures, and there are no apparent differences in the responses between women and men. Recovery of LV volumes and function are rapid, and appear to be linked to plasma volume recovery, while restoration of LV mass lags and is not recovered in normal healthy subjects until 2 wk of reambulation.

GRANTS

This work was funded by National Aeronautics and Space Administration Human Research Program Grant BRC-2003-0000-0543 and in part by National Center for Advancing Translational Sciences Grant 1UL-1RR-029876-01.

DISCLOSURES

No conflicts of interest, financial or otherwise, are declared by the authors.

AUTHOR CONTRIBUTIONS

C.M.W., D.S.M., S.M.C.L., M.B.S., and S.H.P. conception and design of research; C.M.W., D.S.M., S.M.C.L., M.B.S., and S.H.P. performed experiments; C.M.W. and D.S.M. analyzed data; C.M.W., D.S.M., S.M.C.L., M.B.S., and S.H.P. interpreted results of experiments; C.M.W. and S.M.C.L. prepared figures; C.M.W. drafted manuscript; C.M.W., D.S.M., S.M.C.L., M.B.S., and S.H.P. edited and revised manuscript; C.M.W., D.S.M., S.M.C.L., M.B.S., and S.H.P. approved final version of manuscript.