|Home | About | Journals | Submit | Contact Us | Français|
Microgravity and inactivity due to prolonged bed rest have been shown to result in atrophy of spinal extensor muscles such as the multifidus, and either no atrophy or hypertrophy of flexor muscles such as the abdominal group and psoas muscle. These effects are long-lasting after bed rest and the potential effects of rehabilitation are unknown. This two-group intervention study aimed to investigate the effects of two rehabilitation programs on the recovery of lumbo-pelvic musculature following prolonged bed rest. 24 subjects underwent 60 days of head down tilt bed rest as part of the 2nd Berlin BedRest Study (BBR2-2). After bed rest, they underwent one of two exercise programs, trunk flexor and general strength (TFS) training or specific motor control (SMC) training. Magnetic resonance imaging of the lumbo-pelvic region was conducted at the start and end of bed rest and during the recovery period (14 and 90 days after re-ambulation). Cross-sectional areas (CSAs) of the multifidus, psoas, lumbar erector spinae and quadratus lumborum muscles were measured from L1 to L5. Morphological changes including disc volume, spinal length, lordosis angle and disc height were also measured. Both exercise programs restored the multifidus muscle to pre-bed-rest size, but further increases in psoas muscle size were seen in the TFS group up to 14 days after bed rest. There was no significant difference in the number of low back pain reports for the two rehabilitation groups (p = .59). The TFS program resulted in greater decreases in disc volume and anterior disc height. The SMC training program may be preferable to TFS training after bed rest as it restored the CSA of the multifidus muscle without generating potentially harmful compressive forces through the spine.
Alterations in physical activity experienced during spaceflight or prolonged bed rest have direct effects on the musculoskeletal system. Exposure to microgravity has been shown to lead to an increased incidence of low back pain (LBP) associated with abnormal lengthening of the spine , atrophy of spinal musculature , increased intervertebral disc (IVD) height and area [35, 61], and altered IVD composition. Greater than 50% of astronauts complain of LBP during space missions , and astronauts have an increased incidence of disc protrusion when compared to a general or an army aviation population . Astronauts undergo specific training programs before and after spaceflight, to try and minimise the effects of loss of functional weight bearing, and help prevent development of conditions such as LBP. LBP is also known to be a problem for a large number of people , but it is difficult to perform longitudinal studies before and after onset as development of insidious LBP may take a long time. This process is accelerated in spaceflight and prolonged bed rest, and these studies therefore offer a unique opportunity to study people before and after the onset of LBP. Bed rest studies are also useful as they allow testing of countermeasures and rehabilitation procedures both during and after periods of reduced functional weight bearing.
Countermeasures and rehabilitation procedures have traditionally aimed at reversing muscle atrophy and weakness which are assumed to be the result of LBP, spaceflight and prolonged bed rest; however, this is not always the case. While some muscles do atrophy, other muscles increase in size under these conditions. For example, laboratory studies have demonstrated the effects of induced pain [27, 32] or anticipation of pain  on muscle function in normal subjects, and demonstrated that while some muscles become inhibited (and atrophy) , other muscles increase their activation in response to induced pain .
This situation of differential atrophy between muscles also occurs in bed rest. Studies using magnetic resonance imaging (MRI) have demonstrated greater atrophy in the spinal extensor muscles than in the flexor muscles [12, 39]. The psoas muscle, a spinal flexor, has been shown not to change significantly in size during spaceflight [36, 37] and bed rest [12, 38] with other more recent studies showing increases in psoas size during bed rest [4, 21]. Recent studies of prolonged bed rest have shown differential atrophy in the paraspinal (extensor) muscles with the multifidus muscle showing greater atrophy than the lumbar erector spinae at the lower lumbar levels [4, 21]. In comparison, the abdominal flexor muscle group has been shown to increase in size during bed rest .
Some of the changes in bed rest have been observed to be long-lasting in nature. It took 28 days for the psoas muscle to return to its pre-bed-rest size following re-ambulation  and changes in the multifidus muscle were still evident at 90 days after re-ambulation and return to normal activities . Electromyographic studies have also shown the persistence of motor control changes in the lumbo-pelvic musculature up to 1 year after bed rest [7–9]. Similar changes in muscle size have been reported for people with LBP for the multifidus [12, 14, 24, 25, 36–38, 60] and psoas [14, 43, 58] muscles. This differential atrophy of the muscles during bed rest has also been linked to the development of LBP [4, 28].
The effects of prolonged bed rest on the lumbar IVDs have also been studied [4, 6, 12, 28, 35]. Similar to during spaceflight, the length of the lumbar spine has been shown to increase [4, 6, 12, 28]. This is thought to be due to discs imbibing fluid in combination with flattening of the lumbar lordosis . It is important to measure these morphological changes as they may also have biomechanical consequences on the ability of the disc to distribute load  and have been associated with LBP incidence after bed rest [4, 28]. It has previously been shown that exercises performed for the trunk muscles may directly affect passive structures such as the disc , and the possible negative effects of exercises on passive structures should be considered [2, 41, 46] when developing exercise programs. This may be very relevant for people performing exercises when they have LBP associated with discal pathology, and also for astronauts when they exercise and reload their spines after being in the microgravity environment.
Prolonged bed rest has previously resulted in changes in the lumbo-pelvic muscles [6, 21] and IVDs . As these changes have been shown to persist over time [6–9], rehabilitation efforts are warranted. The aim of this study was to determine the effectiveness of two commonly adopted rehabilitation programs on recovery of size of the trunk muscles and changes in morphology of the lumbar spine following prolonged bed rest.
Twenty-four medically and psychologically healthy male subjects were recruited for the bed rest study. Full details of the study protocol have been described elsewhere . In brief, four groups of six subjects underwent bed rest at a time in 2007 and 2008. Baseline data were collected for 9 days before the commencement of the bed rest phase (BDC). All subjects underwent 60 days of 6° head down tilt bed rest (HDT 1-60) and returned for assessment at 14 and 90 days following the bed rest period (recovery, R+14 and R+90). Inclusion criteria included subjects aged between 20 and 45 years, with a height of 155–195 cm. Exclusion criteria included any addictions, regular intake of medication, chronic diseases, history of psychological disease, cardiovascular disease, muscle or bone disease, metal implants, chronic LBP, spinal operations, scoliosis and low bone mass.
The subjects participated in a trial designed to counterbalance the effect of prolonged bed rest. Subjects were paired as room-mates according to psychological criteria and the pairs were randomised to three different groups during bed rest. One group performed resistive exercises with whole body vibration during bed rest (RVE), one group performed resistive exercises only (RES) and one group did not perform any exercises during bed rest (control group, CTR). The study was approved by the institutional ethics committee. All subjects gave their informed written consent prior to participation in the study. Rehabilitation was not commenced until after the bed rest phase of the study was completed.
Two rehabilitation approaches were delivered; a trunk flexor and general strength (TFS) program and a specific motor control (SMC) training program. The SMC program consisted of training the subjects to voluntarily contract the multifidus and deep abdominal muscles using ultrasound imaging [23–25] before progressing to functional retraining in upright positions focusing on spinal position [26, 54]. Subjects worked on maintaining normal patterns of respiration and developing endurance of the deep abdominal and back muscles. Further progression involved functional retraining, including work in a forward leaning position, sitting to standing and maintaining a semi-squat position. To further increase resistance, Thera Band exercise bands (USA) were used for both the upper and lower limbs. A Flexi-bar (Germany) was also used to train endurance of the trunk muscles. This required subjects to rhythmically perturb a flexible bar while maintaining their spinal position.
There is level I evidence  to support the efficacy of this exercise approach in people with LBP  in both restoring multifidus muscle size [24, 26] and decreasing pain and disability [17, 50, 51, 55, 59] and LBP recurrence rates. The trunk component of the TFS program involved trunk strengthening exercises performed in the supine position. These exercises involved lifting the trunk and lower limbs off the floor, and included exercises such as sit-ups, diagonal sit-ups, leg lifts (single and double) and alternating arm and leg lifts. To increase resistance Thera Band exercise bands were used. These exercises have been shown to strongly activate the trunk and hip flexor muscles [1, 13, 15, 40] in co-contraction with the lumbar extensor muscles , but are also associated with the generation of large compressive forces on the spine . This form of floor exercise is used extensively to train the trunk muscles in both athletic programs (competitive sports and fitness) and rehabilitation . In addition, subjects in the TFS program performed exercises targeting the following muscles: quadratus lumborum, hip abductors and adductors, muscles of the shoulder girdle, biceps, and triceps. They also performed stretches of the hamstring, gastrocnemius, soleus, pectoralis major and minor, biceps and triceps muscles. Exercises for both groups were progressed by increasing the number of repetitions of exercises and hold times for static endurance work.
Four subjects from each bed rest trial group (n = 12) were allocated to the SMC program and the TFS program. This design, stratified for the bed rest interventions, ensured that there was equal representation of the three groups during bed rest (high load resistive exercise with whole body vibration, resistive exercise alone and control, no exercise) in each post-bed-rest training group. Subjects from each rehabilitation group were seen for equal amounts of time. From R+2 to R+7, all subjects were seen in hospital daily for individual treatment sessions lasting 30 min. From R+8 to R+14, subjects were seen for two appointments of 30-min duration in a private physiotherapy practice in Berlin. They were given a formal written home program. Between R+16 and R+29, subjects were seen for 30 min in the physiotherapy practice for four appointments, and between R+31 and R+89 for three appointments. At these times, exercises were checked and progressed as able. Overall, it was planned for each subject to be seen 15 times for rehabilitation.
MRI was performed prior to bed rest during the baseline data collection period (BDC, either 8 or 9 days prior to bed rest), at the end of bed rest (HDT55/56), at R+14 and R+90. To obtain the axial images used for measurement of muscle cross-sectional areas (CSAs), five groups of three axial slices (slice thickness: 4 mm; interslice distance: 4 mm; repetition time: 7,560 ms; echo time: 97 ms; field of view: 260 × 234 mm interpolated to 320 × 288 pixels) were positioned over the transverse process of each vertebral body from L1 to L5. Each group of slices was angulated to be parallel to the superior vertebral endplate of its vertebra (Fig. 1).
For spinal morphology, 29 sagittal images (thickness 3 mm; interslice distance: 3.3 mm; repetition time: 5,240 ms; echo time: 101 ms; field of view: 380 × 380 mm interpolated to 320 × 320 pixels) were taken to encompass the entire vertebral body and include the transverse processes of the lumbar spine in a field of view from the lower thoracic spine (typically T10) to the sacrum (Fig. 2).
To ensure measurer blinding to study timepoint, each data set was assigned a random number (http://www.random.org). ImageJ 1.38x (http://rsb.info.nih.gov/ij/) was used for MR image analyses which were conducted by the same operator. The following measures of spinal morphology were conducted in every image where the required anatomical landmarks could be delineated (Fig. 2):
Bilateral CSA measurements of the lumbar multifidus (MF), erector spinae (ES), quadratus lumborum (QL) and psoas (PS) muscles were conducted (Fig. 1). To accurately delineate MF and the more laterally placed longissimus muscle, the fascial border  separating these two muscles was used as an anatomical landmark. The CSA measures for the three MR images at each vertebral level were averaged and then left and right sides were averaged prior to further analysis.
Subjects completed a habitual physical activity questionnaire , prior to bed rest and at the end of the rehabilitation period (either 90 or 180 days after bed rest). Furthermore, from day 2 to day 7 after bed rest (R+2 to R+7) and when subjects returned to the bed rest facility 14, 30 and 90 days (R+14, R+30, R+90) after bed rest, they were asked to fill out a LBP questionnaire. Subjects were asked to report whether LBP was present, mark its location on a body chart and its intensity on a Visual Analogue Scale (VAS) . Incidence of LBP was defined as any report of pain or discomfort between the first lumbar vertebrae and the coccyx.
Linear mixed-effects models  were used to assess each of the spinal morphology variables with factors of group (TFS, SMC), study date (HDT55/56, R+14, R+90), a group × study date interaction and linear co-variates of baseline (BDC) subject age, height and weight. Random effects for each subject were permitted and were necessary allowances for heterogeneity of variance (such as due to group or study date). For disc volume and height, the additional factors of vertebral level (L1/2, L2/3, L3/4, L4/5, L5/S1), all appropriate interactions and random effects were used. An α of 0.05 was taken for statistical significance. For the muscle CSA data, separate models were constructed for each muscle and vertebral level, and a similar approach was used as per spinal morphology. Additional analyses examined the effect of subject training group during bed rest on recovery after bed rest [i.e. interaction of countermeasure group (RVE, RE or CTR) with rehabilitation group (SMC, TFS)]. The “R” statistical environment (version 2.6.1, http://www.r-project.org) was used for all analyses. Activity questionnaire data were also evaluated in a similar fashion.
For analysis of LBP incidence, Chi-square analyses were used to evaluate differences between groups in the total number of reports of LBP compared to the total number of times the questionnaire was completed from R+2 to R+90.
One subject from the TFS group finished bed rest early due to medical reasons and was therefore excluded from the rehabilitation study. Two subjects (1 from the SMC and 1 from the TFS group) failed to return for testing at R+90. The mean (SD) age, height and weight for the subjects in the two groups was as follows—TFS: 30.9 (8.5) years, 177.9 (5.5) cm and 77 (7.9) kg; SMC: 34.1 (7.2) years, 183.4 (5.8) cm and 81.6 (9.2) kg. Total habitual physical activity from questionnaires (no units) before bed rest was 9.1 (1.6) [mean (SD)] in the SMC group and 9.3 (0.9) in the TFS group (no difference between groups before bed rest; p = .80). At the end of the rehabilitation period, total physical activity scores were not significantly different (p > .27) to baseline values [SMC: 9.1 (1.3), TFS: 9.4 (1.0)]. Compliance rates for attendance at the rehabilitation sessions were similar for both groups with a compliance rate of 77.1% for the SMC group and 80.7% for the TFS group.
As analysis of muscle CSA (PS: p > 0.079, ES: p > 0.18, QL: p > .072, MF: p > .059) and spinal morphology data (disc volume: p > .33, anterior disc height: p > .033, posterior disc height: p > .39, lordosis: p = .49, spinal length: p = .63) provided limited evidence of an interaction between training group during bed rest (RVE, RE or CTR) on the effect of rehabilitation group after bed rest (SMC, TFS), the variable of training group during bed rest was removed from the further analyses presented here.
Disc volume, anterior and posterior disc height, lordosis angle and lumbar spine length changed from the end of bed rest up until the end of the rehabilitation phase (F ≥ 9.3, p ≤ .0005). At the end of bed rest, disc volume and height and spinal length were all larger and the lumbar lordosis was smaller (Table 1). In the recovery phase, all of these variables returned towards their baseline levels; however, 90 days after bed rest (R+90), disc volume [mean (SE); 4.8 (0.7)% greater on average than before bed rest], posterior disc height [4.3 (1.1)%], anterior disc height [3.6 (0.8)%], as well as spinal length [1.1 (0.2)%] remained significantly greater than at baseline (Table 1).
The type of rehabilitation impacted upon the changes in disc volume and anterior disc height after bed rest (F ≥ 9.2, p ≤ .0002) with greater reductions in disc volume seen in the TFS group at R+14 [−6.1 (0.6)% compared to the end of bed rest; mean (SEM)] and R+90 [−2.7 (0.6)%] than in the SMC group [R+14: −2.4 (0.9)%, R+90: +0.3 (0.7)% compared to the end of bed rest]. Also, anterior disc height decreased more in the TFS group, and this effect was most apparent at R+14 [−5.9 (0.7)% compared to the end of bed rest vs. −1.6 (0.7)% in the SMC group; R+90: −1.3 (0.7) vs. −0.7 (0.8)% in the SMC group]. The change in posterior disc height, lordosis angle and lumbar spine length during the rehabilitation phase after bed rest was similar in both the TFS and SMC groups (F ≤ 1.5, p ≥ .24; Table 1). Vertebral level (L1/2 to L5S1) had no influence on the changes in disc height or volume over the course of the study or on the impact of the rehabilitation (F ≤ 1.6, p ≥ .14).
With the exception of the quadratus lumborum muscle at L1 and L2 (F ≤ 2.4, p ≥ .10), the CSA of all muscles changed at all vertebral levels in the period from the end of bed rest (HDT55/56) up to 90 days later (R+90; F ≥ 6.6, p ≤ .0022). For the psoas muscle, results from the ANOVA suggested a different response between the two rehabilitation groups at all vertebral levels (F > 4.6, p ≤ .0117). Psoas muscle CSA was marginally increased in the TFS group at all vertebral levels at R+14 (Table 2), while the SMC group showed decreases in CSA of the psoas muscle at all vertebral levels at this time point. At R+90, CSA of the psoas muscle had decreased further in both rehabilitation groups (Table 2). At R+90, with the exception of L5 in the TFS group, CSA of the psoas muscle was still greater than pre-bed-rest values in both subject groups, and this effect was significant at L3 and L5 in the SMC group (p ≤ .032; Table 2) and at L1 in the TFS group (p = .002; Table 2).
The CSA of the multifidus muscle increased in both groups during the rehabilitation phase and there was a different response between the two rehabilitation groups only at one vertebral level, L3 (F = 7.8, p = .0008; otherwise: F < 1.4, p > .24; Table 2). A greater increase in the CSA of the multifidus muscle was seen at the L3 vertebral level in the TFS group at both R+14 and R+90. At R+90, the CSA of the multifidus muscle was not significantly less (p > .13) than pre-bed-rest levels, though in the TFS group the CSA of this muscle was greater than pre-bed-rest levels at R+90 at L1 and L3 (p ≤ .042).
The CSA of the erector spinae muscle also increased in both rehabilitation groups after bed rest (Table 3). Ninety days after bed rest, the CSA of the erector spinae muscle was marginally (non-significantly) greater than pre-bed-rest values in the TFS group, whereas the SMC group showed marginally smaller values for this muscle at all vertebral levels except L5, and this effect reached significance at L2 (p = .046). The results of the ANOVA showed that these marginal differences between groups were not significant (F < 2.4, p ≥ .10).
The CSA of the quadratus lumborum muscle also increased after bed rest (Table 3). Ninety days after bed rest, the CSA of this muscle was significantly less than pre-bed-rest levels in the SMC group at L2, L3 and L4, whereas this was not the case in the TFS group (Table 3), but results of the ANOVA suggested the response in the two rehabilitation groups to be similar for this muscle (F ≤ 2.0, p ≥ .14).
Subjects in both the TFS and SMC groups reported LBP during the rehabilitation period. From R+2 to R+7, 5/12 (42%) of the subjects in the SMC group reported LBP, versus 6/11 (55%) in the TFS group. In the period subsequent to this (from R+14 to R+90), an additional two subjects in each group reported LBP. The SMC group reported pain on a total of 14 times (of 107 times of questionnaire completion between R+2 and R+90; 13%) and the TFS group reported LBP 17 times (of 98 questionnaire completions; 17%), though these differences were not significant (χ2 = .29, p = .59).
The 2nd Berlin BedRest Study has provided a unique set of circumstances to investigate the effects of two different rehabilitation programs. The advantage of this study was that a homogenous group of healthy male individuals was exposed to a controlled environment known to induce predictable changes in lumbo-pelvic muscles. Both rehabilitation programs used in this study are currently widely used in the community for people with and without LBP .
Both rehabilitation programs adopted were successful in restoring the CSA of the multifidus muscle. This is important as decreases in the CSA of the multifidus muscle have been documented not only in those who have undergone bed rest studies [4, 6, 21], and these changes have been linked to LBP incidence after bed rest . Data from astronauts [36, 37] suggest that similar changes in multifidus may occur in microgravity as well as in individuals with acute and chronic LBP [14, 24, 26, 31, 44]. While the mechanism of muscle atrophy may not be the same in these different populations, findings were promising in that they showed that rehabilitation of the multifidus muscle was successful for both groups by R+14. Changes were greatest during the period of intensive rehabilitation between R+2 and R+14, with no significant improvements seen between R+14 and R+90 (when subjects were seen less regularly and were performing predominantly home exercise programs). As the TFS training targeted the trunk flexor muscles, the improvement in CSA of the multifidus muscle in this group is most likely due to co-contraction of the abdominal muscles with the multifidus and erector spinae muscles during the exercises performed. This phenomenon has been shown in EMG studies of exercises such as the sit-up , and underlies the common belief that doing exercises such as sit-ups will simultaneously strengthen the back muscles. A possible explanation for the greater increase in multifidus CSA at the L3 vertebral level in the TFS group compared to the SMC group may relate to the location of the vertebral level (middle of the lumbar spine) being loaded by exercises which lift the trunk (maximal effect upper to mid-lumbar spine) and lift the lower limbs (maximal effect lower lumbar spine to mid-lumbar spine), though the overall higher loads developed during TFS group training may have also played a role. With regard to the SMC training, similar improvements in multifidus muscle CSA have been seen in subjects with acute LBP  and athletes with LBP  who underwent a similar training program.
For the psoas muscle, which increased in size during bed rest, there was a different result for the two rehabilitation groups. The TFS training resulted in an initial increase in the CSA of the psoas muscles from the end of bed rest up until R+14, with decreases after that period. This was not seen in the SMC group.
The increase in the psoas muscle size which was demonstrated in the TFS group was likely due to its action as a powerful hip and lumbar spine flexor. In this capacity, the psoas is capable of exerting substantial loads on the lumbar spine due to its attachment to the vertebral column . For this reason, researchers have advised that increasing psoas action by spine and hip flexion could have detrimental effects on the lumbar spine by increasing compressive forces. It has, therefore, been suggested that, when performing abdominal exercises in the supine position, exercises should focus on spine flexion alone [2, 33] rather than spine and hip flexion [2, 46]. Researchers have further advocated performing trunk flexion exercises with the hips and knees bent to (1) reduce tension in the psoas muscle , (2) reduce involvement of the hip flexors  and (3) to reduce the torque produced . However, others have shown that using this position decreased compressive forces , while another study showed that it made no difference to the induced compressive forces on the spine .
When considering the effects of the psoas and multifidus muscles working together in normal function, Quint et al.  proposed that theses muscles co-contract to increase the overall stiffness of the lumbar spinal segments. Using fixed levels of simulated muscle forces in the multifidus and psoas muscles, these researchers demonstrated this to be the case for axial torque and lateral bending, but such co-contraction acted to destabilise the lumbar segments in flexion. Thus, the situation of an imbalance between the hypertrophied psoas muscle, which induces a flexion force on the spine, and the atrophied multifidus muscle may be detrimental to the lumbar spine. Interestingly, increased psoas muscle size has also been documented in athletes with LBP when compared with athletes without LBP [43, 58].
This result would support the theory that it may be preferable to use SMC and functional retraining rather than floor-based programs following bed rest. The results of increasing the size of the multifidus muscle without concurrent increase in psoas muscle size may be due to the focus on obtaining and maintaining an “s” curve (thoracic kyphosis and lumbar lordosis) in SMC training with axial loading of the spine. This principle of exercise which has been demonstrated to be effective in the rehabilitation of LBP, and now for bed rest subjects, may also apply to training programs used for astronauts.
The other main consideration of the exercises used is the potential effect on the IVD and vertebrae. Exercises in the supine position which involve lifting the trunk or lifting the legs are often prescribed with caution in those with LBP due to the potential for increased intradiscal pressure [2, 41, 46]. Evidence of this concern has been confirmed in patients with osteoporosis where a greater percentage of further wedge or compression fractures have occurred in subjects performing flexion exercises such as sit-ups . The results from the current study showed that the forces induced by the TFS program were large enough to induce changes in the anterior disc height and disc volume. While this may seem a desirable effect, it happened rapidly (by R+14), and other works [4, 28] have found a link between IVD changes during bed rest and LBP subsequent to bed rest. The incidence of LBP after bed rest was marginally (non-significantly) higher in the TFS group and hence a gradual return to more normal disc morphology may be preferable from a safety perspective.
While the tissues of the lumbar spine may be more susceptible to damage following bed rest [4, 28, 34] and spaceflight , this does not mean that these subjects should never perform higher load exercises, as the spine is exposed to high loads in everyday activities such as lifting , and even coughing and sneezing , due to forces created by the trunk muscles themselves. It may be sensible to commence training for bed rest subjects and astronauts using an approach such as SMC, with progression later on to higher load activities, when the spinal morphology has returned to normal. Another feature of appropriate exercise may be the use of weight bearing (closed chain) exercises which activate antigravity extensor muscles and apply axial loading to the skeleton and soft tissues .
The main limitation of this study was the small subject sample size, which is common to all bed rest studies due to their complex nature and expense. Despite the small sample size, significant changes were seen in the muscles of the lumbo-pelvic region in terms of muscle atrophy, muscle hypertrophy and response to rehabilitation. Although it was not possible to include a control group for the rehabilitation phase of the study due to small numbers, results from the 1st Berlin BedRest Study [6–9, 21] (where subjects did not receive rehabilitation after bed rest) showed that changes in the multifidus muscle were long-lasting without rehabilitation.
The authors wish to thank the subjects who participated in the study, and the staff of ward 18A in the Charité Campus Benjamin Franklin Hospital, Berlin, Germany. We would also like to acknowledge Professor Chris Snijders and Christine Hamilton. Christine Hamilton contributed to the development of the rehabilitation program and assisted in delivering the program to the first group of subjects. The 2nd Berlin BedRest Study was supported by grant 14431/02/NL/SH2 from the European Space Agency and grant 50WB0720 from the German Aerospace Center (DLR). The 2nd Berlin BedRest Study was also sponsored by Novotec Medical, Charité Universitätsmedizin Berlin, Siemens, Osteomedical Group, Wyeth Pharma, Servier Deutschland, P&G, Kubivent, Seca, Astra-Zeneka and General Electric. The rehabilitation project was funded by the European Space Agency. Julie A Hides was supported by a travel grant from The University of Queensland, Australia. Gunda Lambrecht was supported by a grant from the European Space agency and Daniel L. Belavý was supported by a post-doctoral fellowship from the Alexander von Humboldt Foundation. The investigations in the current study were also supported by grant number FE 468/5-1 from the German Research Foundation (DFG).
An erratum to this article can be found at http://dx.doi.org/10.1007/s00586-010-1551-2