Regulatory, Integrative and Comparative Physiology


The objectives of this study were as follows: 1) to measure human energy expenditure (EE) during spaceflight on a shuttle mission by using the doubly labeled water (DLW) method;2) to determine whether the astronauts were in negative energy balance during spaceflight;3) to use the comparison of change in body fat as measured by the intake DLW EE,18O dilution, and dual energy X-ray absorptiometry (DEXA) to validate the DLW method for spaceflight; and 4) to compare EE during spaceflight against that found with bed rest. Two experiments were conducted: a flight experiment (n = 4) on the 16-day 1996 life and microgravity sciences shuttle mission and a 6° head-down tilt bed rest study with controlled dietary intake (n = 8). The bed rest study was designed to simulate the flight experiment and included exercise. Two EE determinations were done before flight (bed rest), during flight (bed rest), and after flight (recovery). Energy intake and N balance were monitored for the entire period. Results were that body weight, water, fat, and energy balance were unchanged with bed rest. For the flight experiment, decreases in weight (2.6 ± 0.4 kg,P < 0.05) and N retention (−2.37 ± 0.45 g N/day,P < 0.05) were found. Dietary intake for the four astronauts was reduced in flight (3,025 ± 180 vs. 1,943 ± 179 kcal/day, P < 0.05). EE in flight was 3,320 ± 155 kcal/day, resulting in a negative energy balance of 1,355 ± 80 kcal/day (−15.7 ± 1.0 kcal ⋅ kg−1 ⋅ day−1,P < 0.05). This corresponded to a loss of 2.1 ± 0.4 kg body fat, which was within experimental error of the fat loss determined by18O dilution (−1.4 ± 0.5 kg) and DEXA (−2.4 ± 0.4 kg). All three methods showed no change in body fat with bed rest. In conclusion,1) the DLW method for measuring EE during spaceflight is valid, 2) the astronauts were in severe negative energy balance and oxidized body fat, and 3) in-flight energy (E) requirements can be predicted from the equation: E = 1.40 × resting metabolic rate + exercise.

  • astronauts
  • nitrogen balance
  • bed rest

astronauts lose weight during spaceflight even on the relatively short (<3 wk) shuttle missions (9, 15, 22, 25). Some degree of weight loss is to be expected because there is less need for the antigravity muscles (29). After the initial adaptation phase is over, the mechanism is believed to be via a reduction in the protein synthesis rate. Numerous studies on the ground of humans during bed rest and rats with their hindlimbs unweighted have shown that decreased tension on the antigravity muscles leads to a reduction in protein synthesis by the affected muscles (2, 5, 6, 14, 26).

On most missions where dietary intake has been measured, dietary intake in flight has been substantially less than preflight (11, 19). Whether this decrease in intake parallels a reduction in energy expenditure arising from a possible decrease in the energy costs of living and working in space or simply reflects a failure to maintain energy balance is not known. If astronauts were in negative energy balance, this would be a factor in promoting protein loss. We previously hypothesized that on missions with high energy requirements, astronauts are unable to maintain energy balance (22). High energy needs can occur because of the amount of exercise required and/or the extent of extravehicular activity. Exercise has been recommended as a countermeasure: aerobic to maintain the cardiovascular system and resistive to preserve muscle mass (3, 16).

The first objective of this experiment was to test this hypothesis in flight on crew members participating in the physically demanding life and microgravity sciences (LMS) mission. The LMS mission involved extensive exercise activities by the crew. The second objective was to compare the in-flight energy expenditure measurements to a comparable data set obtained on the ground during bed rest. The bed rest study included the same exercise and testing protocols as the flight study.


Two experiments were performed: a spaceflight experiment performed on the 1996 LMS mission and a bed rest study conducted at the National Aeronautics and Space Administration (NASA)-Ames Research Center in the summer of 1995. Informed consent for both sets of studies was obtained in accordance with the policies of the University of Medicine and Dentistry of New Jersey and NASA. The objective of the bed rest experiment was to model on the ground the measurements made in space. As far as possible the protocol and measurements were common to both experiments.

Submaximal and maximal exercise testing were done (30). The maximal test was a continuous test to volitional exhaustion (V˙o 2 max) consisting of four 3-min stages at 50, 100, 150, and 175 W followed by 25-W increments every 2 min until exhaustion. Tests were performed twice before bed rest [control (C)−13, C−8]/ spaceflight [launch (L)−60, L−15], during bed rest [bed rest (BR)2, BR8, BR13], and after bed rest [recovery (R)+2, R+6/spaceflight (R+4, R+8)]. In the second protocol, the submaximal test involved the same incremental sequence but was terminated at 85%V˙o 2 max as defined from the first pre-bed rest/preflight determination. Submaximal tests were only done in flight and these were done on flight days (FD) FD2, FD8, and FD13 (30).

For each experiment three complimentary methods for determining energy balance were employed. The first was by measuring changes in body composition using isotope dilution (H2 18O) to estimate changes in body fat content. If subjects are in positive energy balance, fat is deposited. Conversely, if energy balance is negative, body fat stores are mobilized and used to compensate for the deficit in intake. The second and potentially more accurate method for determining energy balance is from the difference between energy intake and energy expenditure measured by the doubly labeled water (DLW,2H2 18O) method. The third was by measuring body fat by dual energy X-ray absorptiometry (DEXA) before and after the spaceflight/bed rest study period.

Bed Rest Experiment

A 17-day bed rest study with 6° head-down tilt was conducted in the Clinical Research Center of the NASA-Ames Research Center with eight healthy adult males who were recruited from the local community. To compare the bed rest and flight studies, the bed rest study was designed to simulate the flight experiment as far as possible. The bed rest study included the full complement of exercise testing that was done on the shuttle payload crew but not the activity associated with living and working in space (1, 30). Except when the subjects were being tested, they were maintained at 6° head-down tilt.

As with the flight study, the bed rest study was divided into three phases, a 15-day pre-bed rest ambulatory period followed by 17 days of bed rest and ending with a 15-day recovery period. During the 47 days of the study, the subjects received all their food from the research center. Twelve daily menus were made up comprising 2,500 kcal/day and 90 g protein/day. Subjects were served weighed portions of food, and any residues were weighed. In addition, subjects were allowed limited access to a snack basket that contained fruit, cookies, some candy, and granola bars if they were hungry. An accurate record of dietary intake was kept for the entire study period. Urine was collected continuously for the 47-day period.

Three sets of DLW-energy expenditure measurements were done, two during the pre-bed rest ambulatory (control) phase (C−15–C−9, C−9–C−3), two during bed rest (BR3–BR9 and BR9–BR15), and two during the recovery phase (R+4–R+10 and R+10–R+16). Six equal isotope doses of 20 g H2 18O as a 10% solution and 8 g 99%2H2O (ICONS, Summit, NJ) were given. Two of the subjects served as undosed controls.

On isotope dosing days, two saliva samples were collected before giving the doubly labeled water and two 6–8 h after giving the isotopes. On all other days a single saliva sample was collected. Subjects were required to abstain from eating or drinking for half an hour before collecting saliva.

A whole body DEXA scanner (Lunar, DPX) located at the Ames Research Center was used to obtain body composition of the bed rest subjects. Two scans were obtained in the week before bed rest and one after 24–48 h of reambulation.

Flight Experiment

The spaceflight study was done on the four payload crew of the LMS mission. There were two overall goals of the mission. The first was to conduct a series of experiments on the response of the human musculoskeletal system to spaceflight, and the second was to perform a series of material science experiments. The mission lasted 17 days.

For the flight study we measured energy expenditure and body composition before flight, in flight, and postflight. Energy expenditure and balance were measured over blocks of 6 days. There were two consecutive 6-day blocks preflight (days L−16–L−10 and L−10–L−4), two similar consecutive 6-day blocks in flight (days FD3–FD9 and FD9–FD15) and postflight (R+4–R+9 and R+9–R+15). L−x refers to launch day minusx days; FDx is flight dayx; and R+x is return minusx days. The use of such a block scheme increases the sensitivity of the measurements by allowing for repeated measurements on each subject, an important factor when the number of test subjects is small. Daily dietary intake was monitored and daily urine collections were made for the period L−15–R+15 as previously described (22).

Briefly, the crew recorded the amount of each food item eaten. A bar code reader was used to monitor the type of food eaten. The proportion of the individual food package contents eaten were recorded by the crew using a cassette recorder. For the preflight and postflight periods, subjects collected their food each day from a NASA facility either at the Johnson Space Center (JSC) or the Kennedy Space Center (KSC). Spacelab included a system especially designed to collect, measure, and save an aliquot of each urine void. This unisex apparatus measured the mass and time of each void and then aliquoted a portion of the void into a labeled 20-ml tube. The tubes were then transferred manually by one of the crew to a freezer.

The2H2 18O was given preflight on days L−16 (20 g18O, 8 g2H), L−10 (20 g18O, 8 g2H), and L−4 (10 g18O, 4 g2H); in flight on days FD3 (30 g18O, 12 g2H), FD9 (36 g18O, 14 g2H), and FD16 (9 g18O, 3.6 g2H); and postflight on R+4 (20 g18O, 8 g2H), R+10 (20 g18O, 8 g2H), and R+16 (9.6 g18O, 3.8 g2H). For the pre- and postflight studies, the isotope mixtures were made up by mixing 10% H2 18O and 99%2H2O (ICONS). For the in-flight study, 70% H2 18O and 99%2H2O were used. Isotope doses were increased for the flight studies to minimize the complexity of correcting for the varying isotopic background of the drinking water from the shuttle’s fuel cells. On board, fuel cells provided the drinking water. During flight, galley water was sampled seven times to determine the background enrichment.

Two saliva samples were collected on each of the days after and preceding the administration of the isotopes. On isotope dosing days, two saliva samples were collected before giving the doubly labeled water, and two were collected 6–8 h after giving the isotopes. In addition, duplicate saliva samples were collected on several other days when permitted by the crew’s schedules. On days where duplicate saliva samples were collected, 30 or more minutes were allowed to elapse between collections. Subjects were required to abstain from eating or drinking for half an hour before collecting saliva.

Body composition was determined on the four flight crew by DEXA. The DEXA scans were done in Houston using a Hologic 2000 ∼30 and 60 days before flight and at R+2, R+10, and R+30 days after flight. The DEXA scans were analyzed for the lean body mass (LBM) and body fat. The two preflight scans were averaged for data analysis.

The three major differences between the bed rest and the flight experiment were 1) exercise to 85%V˙o 2 max was substituted for exercise to exhaustion for the flight period (5),2) the isotope doses for the ambulatory control period (C−15 and C−9), the bed rest period (BR4 and BR10), and the recovery period (R+4 and R+10) were all equal (20 g 18O, 8 g2H), and3) subjects were not given a final dose of2H2 18O at the end of each period (C−16, BR16, and R+16).

Analytic Methodology

A series of standards (range 0.05 to 0.6%18O and 0.02 to 0.24%2H) was prepared and standardized by an outside laboratory (Metabolic Solutions, Merrimack, NH). The samples were coded, randomized, and analyzed independently on three separate occasions.

18 O analyses. Saliva (1 ml) was placed in a 10-ml Vacutainer (Beckton Dickinson), and flushed for 20 s with a mixture of 10% CO2 in N2. The samples were then equilibrated at room temperature for 3–4 days in a shaking water bath.

2 H analyses. Saliva (0.4 ml) was placed in a 13-ml screw cap container (Europa, Cleveland, OH) that contained ∼10 mg of 5% platinum on a 325 mesh alumina catalyst (Sigma-Aldrich, Milwaukee, WI) in a 0.25-ml conical glass insert. Tubes were evacuated for 1.0 min while immersed in ice, and then H2 gas at 4 lbs/in.2 was added for 10 s (21). Samples were equilibrated for 4 days at room temperature. All isotope analyses were done on either a VG-SIRA-II (VG Instruments, Cheshire, UK) or a Europa 20–20 Hydra isotope ratio mass spectrometer. Standard curves were run after every 20–30 samples. All samples were run in duplicate. The isotopic enrichment of the saliva was calculated from the standard curves.

We did all of the analyses, except those for one subject from the flight experiment where some of the samples were too small for our instruments (1.0 ml needed). These samples (range 0.5–1.5 ml) were analyzed by Metabolic Solutions.

Methods of Calculation

Energy expenditure and balance. The total body water (TBW) was calculated from the mean of the two predoses and the mean of the two postisotope18O dosing values usingequation 1 TBW=(d/18.02)×(δaδi)/(δsδp) Equation 1 where d is the dose of 18O; δa is the18O enrichment; and δi, δs, and δp are the enrichments of tap water, saliva after dosing, and saliva before dosing, respectively.

LBM and body fat were calculated from the relationships LBM = TBW/0.73 and body fat = weight − LBM (8).

Isotope loss rates were calculated by regression analysis. Energy expenditure was calculated from the 1986 equation of Schoeller et al. (20)rCO2=0.481N(1.01kO1.04kH)0.0258N(kOkH) Equation 2 where rCO2 is rate of CO2 production, N is mean TBW in moles, k O is fractional H2 18O turnover rate, andk H is fractional2H2O turnover rate.

The CO2 production rate was then converted to energy expenditure in kilocalories by the Weir equation (28)E=3.941VO2+1.106VCO22.17N Equation 3 rCO2=VO2/22.4 Equation 4 RQ=VCO2/VO2 Equation 5 where Vo 2 is rate of O2 utilization in liters, Vco 2 is rate of CO2 production in liters, E is rate of O2 utilization in kilocalories per kilogram per day, and RQ is respiratory quotient.

For the flight samples, it was necessary to correct for the difference in enrichment between the water available inflight for drinking on the shuttle (galley water) and the water that the crew had been drinking on the ground before flight. The isotopic enrichment of the galley water (18O, 35.8 ± 0.1δ;2H, −116.6 ± 3.7δ) was considerably different from Houston water (18O, 3.5δ;2H, 26δ), so a background correction was necessary. The galley water was sampled seven times inflight and did not vary. Fluid intake averaged ∼2,000 ml/day, of which 80% was derived from galley water and the remainder from the diet (12). The mean TBW for the four subjects was ∼50 liters. Thus 1.6 liters of galley water entered the body per day, and 0.4 liters entered from food. We assumed that the small amount of food water approximated Houston water. From these data a daily correction factor was calculated. The isotope loss rates were calculated from the uncorrected and the corrected saliva2H and18O enrichments by regression analysis. A value of 0.85 was assigned to the RQ and 14 for the grams N in the Weir equation. Energy balance was calculated from the difference between dietary intake and energy expenditure.

Nitrogen balance. For the bed rest study, nitrogen balance was estimated from the difference in dietary nitrogen intake and urinary nitrogen excretion. Urinary creatinine excretion was constant for all three phases of the bed rest study. Unlike the bed rest study, urinary creatinine excretion was abnormally low for the flight phase for two of three subjects (One subject did not fully cooperate with collecting urine during the entire study and therefore was excluded). The other two sweated heavily during the in-flight exercise periods and so lost a considerable amount of water as sweat. The urinary creatinine excretion for these two subjects was only ∼70% of their preflight value. To estimate the actual urinary N excretion, the measured in-flight N excretion was normalized to the preflight creatinine for these two subjects.


Statistical significance of the data was evaluated by either ANOVA using a repeated-measures design (RMANOVA), pairedt-tests, or regression analysis, as appropriate. For all measurements where there were values for the control period, study period (bed rest or spaceflight), and the recovery period, a RMANOVA was done. The six blocks were evaluated both independently and combined into control, bed rest, and recovery periods. For N intake, excretion, and balance, a RMANOVA was done on the mean value per subject for those periods. If the ANOVA indicated significance at the P < 0.05 level or better, group differences were identified by the least significant differences test. The SAS Statistical System (SAS Institute, Cary, NC) was used for the statistical computations. Data in the text, Figs.1-4, and Tables 1-8 are means ± SE.


Bed Rest Study

Table 1 gives the age, height, and weight data for the bed rest subjects and astronauts. Six of eight subjects were dosed with2H2 18O. The other two served as undosed controls. All data reported here refer only to the six dosed subjects. There was no change in either body weight, body water, or the isotopically determined LBM or body fat with bed rest or during the recovery phase (Table2, Fig. 1). The DEXA-derived whole body lean tissue values for the bed rest subjects showed a small but significant (P < 0.01) decrease in total body lean tissue compared with pre-bed rest (pre = 57.7 ± 1.7 kg, post = 56.3 ± 1.3 kg). Body fat content was unchanged with bed rest (pre = 21.7 ± 2.4 kg, post = 23.1 ± 2.8 kg).

View this table:
Table 1.

Age, height, and weight data for bed rest subjects and astronauts

View this table:
Table 2.

Body composition before, during, and after bed rest

Fig. 1.

Comparison of body water changes between spaceflight (solid symbols) and bed rest (BR; open symbols). LMS, shuttle data; P, pre-bed rest or flight; B, bed rest; F, flight; R, recovery. * P < 0.05 vs. mean preflight by repeated-measures ANOVA.

All values for nitrogen balance, unless otherwise stated, are estimates because they are based on the excretion of nitrogen in the urine and do not include fecal nitrogen losses. These losses run ∼20–25 mg N ⋅ kg−1 ⋅ day−1(10, 13). Nitrogen retention decreased during the bed rest phase of the experiment (Table 3). If allowance for fecal N losses are made, nitrogen balance during bed rest was negative by ∼25 mg N ⋅ kg−1 ⋅ day−1. After bed rest, nitrogen retention increased and, although significantly greater than during bed rest, was not greater than the pre-bed rest control period (Table 3).

View this table:
Table 3.

Nitrogen balance before, during, and after bed rest

Although conditions were such as to encourage subjects to maintain a constant dietary intake, subjects made sufficient use of the snack basket to substantially increase their intake for the two ambulatory phases of the study by ∼250 kcal ⋅ kg−1 ⋅ day−1(Table 4). During the bed rest phase of the experiment, intake declined slightly (by ∼250 kcal). This decline in intake was matched by a parallel decline in energy expenditure (Table4). Subjects were in energy balance for all three periods (Table 4). Both the decrease in intake and the decrease in energy expenditure were statistically significant.

View this table:
Table 4.

Energy intake, expenditure, and balance before, during, and after bed rest

Flight Experiment

Body weight was measured preflight, once in flight toward the end of the mission on FD15, and after landing on R+4, R+10, and R+16. According to the in-flight measurement at the end of the mission, all of the astronauts had lost weight by FD15, and three had lost >4 kg (Table 5). The weight loss for these three subjects persisted into the postflight phase. By the time the R+4 measurement was made, fluid balance had stabilized. The mean body weight loss using the R+4 weights was 1.9 ± 0.8 kg.

View this table:
Table 5.

Body composition data before, during, and after spaceflight for 4 astronauts

For all subjects, there was a pronounced decrease in TBW early in flight. With increasing time in orbit this loss was regained, and the body water content returned to the preflight value (Fig. 1). The single in-flight body weight measurement on FD15 agreed well with the body water decrease. The body water was regained by the time of the first postflight measurement on R+4. In contrast to the body water data obtained before and during flight, there was considerable scatter in the postflight body water data (Fig. 1).

Neither the DEXA- nor 18O isotope dilution-derived LBM measurements showed any change with spaceflight (Table 5), but the decrease in N retention with spaceflight was statistically significant (Table 6).

View this table:
Table 6.

Nitrogen balance data before, during, and after spaceflight for 4 astronauts

The isotopically determined body fat measurements showed a net loss of 4.4 ± 0.9 kg fat using the R+4 isotope dilution data and 1.4 ± 0.5 kg if the R+10 data were used. The DEXA results were similar (2.8 ± 0.4 and 2.4 ± 0.4 kg using the R+2 and R+10 DEXA data points). All of these changes are statistically significant (Table5).

Dietary intake for the four astronauts was significantly reduced (36 ± 4%, days 2-16) during spaceflight. After the first day in earth orbit, energy intake was relatively constant for the remainder of the flight period (Fig.2). Three of four subjects showed a much greater reduction in intake than the fourth subject (Table 6). The three subjects with the greatest reduction in intake were those who had had the greatest loss of weight and body fat.

Fig. 2.

Energy intake before, during, and after spaceflight.

We were unable to obtain energy expenditure values for the pre- and postflight periods because it was not possible to correct for the uncertain background water enrichment with sufficient accuracy. The reasons for the uncertainty were several, including the heavy sweat losses, replacement of water losses from a variety of nontap sources (bottled water, soda, beer, etc.), and the variable location of the crew (JSC, TX; KSC, FL, and Marshall Space Center, AL) during the study periods. The background water was different at each site. The isotope doses were not high enough (due to fiscal considerations), as they were for the in-flight phase of the experiment, to reduce the large background correction to a reasonable level.

In contrast, during the flight phase of the experiment it was possible to correct for changes in background water because accurate records of fluid intake were maintained and fluid sources were limited to galley water (80%) and food water (20%). In addition, relatively large isotope doses were administered in flight.

Making a correction for the background differences in water intake increases the calculated energy expenditure rate by ∼700 kcal/day (from 2,579 ± 121 to 3,320 ± 155 kcal/day, ∼9 kcal ⋅ kg−1 ⋅ day−1). All four astronauts were in negative energy balance during the flight period (Table 7). There was no difference between the two in-flight periods. Irrespective of whether a background correction is or is not made, the subjects were in a net energy deficit. The deficit was 1,355 ± 80 kcal/day (15.7 ± 1.4 kcal ⋅ kg−1 ⋅ day−1, Table 7). If a background correction for the galley water2H and18O enrichments is not made, the energy deficit was 594 ± 203 kcal/day (6.9 ± 2.4 kcal ⋅ kg−1 ⋅ day−1).

View this table:
Table 7.

Inflight energy expenditure data for 4 astronauts


Bed Rest

The bed rest study was designed to simulate the mission. It therefore included multiple independent experiments involving an extensive battery of physiological and psychological testing. The experiments were the same as those done on the astronauts. The overall focus of the physiological testing was on the musculoskeletal system.

The combination of providing the subjects with access to a snack basket and not compelling them to eat all the food offered introduced enough latitude for them to adjust their intake needs to the situation. The difference was primarily in the two ambulatory phases where the subjects made more extensive use of the snack basket, although on average about one food item (e.g., a glass of skimmed milk) was declined per subject per day for reasons of taste. The net result was a decrease in dietary intake for the bed rest phase of ∼250 kcal/day (Table 4).

Because of this decreased intake, the bed rest subjects were able to maintain energy balance for all three phases of the experiment. The measured dietary intake was the same as the measured energy expenditure for all six measurement periods (Table 4). The body fat measurements (by isotope dilution and DEXA) confirmed the intake−DLW energy balance measurements. There was no change in body fat throughout the study (Table 2).

The bed rest energy expenditures are greater than those reported in an earlier bed rest study by Gretebeck (24.2 ± 0.8 kcal ⋅ kg−1 ⋅ day−1; Ref. 8). We suggest that the difference (30.8 − 24.2 = 6.4 kcal ⋅ kg−1 ⋅ day−1) is because the subjects in the present study were subjected to an extensive battery of physiological and psychological testing during the bed rest period.


The age, height, and weight of the four astronauts are summarized in Table 1. The mean dietary intake in flight for the entire flight period (1,708 ± 98 kcal/day) was very much less than before flight (3,025 ± 180 kcal/day, Table 6), a reduction of about one-third. It was also substantially less than the control bed rest experiment. On a day-to-day basis, energy intake in flight was, with the exception of the first 2 days in orbit, relatively constant within a period (preflight, in flight, or postflight; Fig. 2). The decrease in intake during the first day or two in orbit was due to the transient space motion sickness. A similar phenomenon was noted on the 1991 Space Life Sciences mission-1 (SLS1) and 1993 Space Life Sciences mission-2 (SLS2; Ref. 22), where after the first 2 days in orbit, intake was constant. It suggests that in space, intake rapidly adjusts to a new set point.

There is other evidence in support of this statement. In Fig.3 we have plotted the dietary intake during the first 2 wk for all of the missions for which dietary intake data are available. The missions are Skylab, SLS1/2, and LMS. Skylab consisted of three missions, Skylabs 2, 3, and 4, with three astronauts per mission, and lasted 28, 59, and 84 days, respectively. Dietary intake was increased progressively from Skylab 2 to Skylab 4 (22). The Skylab astronauts were on a metabolic balance study and were required to eat a predetermined amount of food per day.

Fig. 3.

Comparison of energy intake during first 2 wk of spaceflight for 3 Skylab missions (Skylab 2, 28 days; Skylab 3, 56 days; and Skylab 4, 84 days), SLS1 and 2 (combined), and LMS mission (22, 29).

There was no difference in the preflight energy intakes between any of the five missions, but the in-flight intakes were very different. Two outliers have been dropped: one subject from SLS1/2 (discussed in Ref.22) and one subject (C) from this mission. Figure 3 shows the data with this subject included. The plots for the three Skylab missions show little variance, as is to be expected for a controlled diet. But the plots for SLS1/2 and LMS, where intake was ad libitum, show equally little variance, suggesting that there is some unknown factor that regulates intake on a mission. The factor is not energy need, otherwise the subjects in this study would not have been in negative energy balance.

The in-flight dietary intake was remarkably low (24.6 ± 3.3 kcal ⋅ kg−1 ⋅ day−1) on this mission, and the reduction accounts for the chronic negative energy balance found during both the early (FD3–9) and late (FD9–15) phases of the mission (Table 7). The level of dietary intake found on this mission was much lower than that on SLS1 and SLS2 (30.4 ± 1.5 kcal ⋅ kg−1 ⋅ day−1) or the much earlier Skylab (36.8 ± 1.4 kcal ⋅ kg−1 ⋅ day−1) and lower than on a composite of other shuttle missions (27.1 ± 1.4 kcal ⋅ kg−1 ⋅ day−1; Refs. 12, 19, 22). The reduction in intake occurred even though physical activity was much greater on this mission. Consequently, astronauts were in substantial negative energy balance during their time in orbit.

Comparison of Energy Balance by the Three Methods

In this study we obtained three independent estimates of energy balance during spaceflight (Table 8). The first was from the difference between energy intake and the DLW energy expenditure. The second was from the amounts of endogenous body fat and protein oxidized during the flight period. The third method is from the pre- and postflight DEXA measurements.

View this table:
Table 8.

Comparison of change in body fat with bed rest and spaceflight as measured by 3 independent methods

Intake−DLW. The negative energy balance of 1,335 kcal/day can be converted into kilograms of body fat for the entire mission by multiplying 1,335 by the duration of the mission (16 days), subtracting the contribution from protein oxidized, and dividing the result by 9,000 (caloric equivalent of 1 kg of body fat). The amount of body protein oxidized can be estimated from the change in nitrogen retention between the preflight and in-flight periods {= [47 − (−)29 mg N ⋅ kg−1 ⋅ day−1× 81 kg (body wt) × 16 (flight duration) × 6.25 × 4 (caloric equivalence of protein) × 0.001] ≃ 2,560 kcal}. The total negative energy balance for the flight period was 16 × 1,335 = 21,400 kcal. The amount contributed by fat is 21,400 − 2,560 ≃ 19,000 kcal ≃ 2.1 kg fat.

Isotope dilution (18O dilution space). Three estimates of the change in body fat content with spaceflight were obtained from the three body water determinations, on FD15 (−4.6 ± 0.9 kg), R+4 (−4.4 ± 0.9 kg), and R+10 (−1.4 ± 0.5 kg). All three estimates showed that the astronauts lost body fat. Tissue hydration is likely to be different in flight and in the early postlanding phase because of the fluid shifts that occur, and this accounts for the unphysiologically high estimates obtained at the end of the mission and in the period after landing on FD15 and R+4 (Table 5; Ref. 11). By R+10 reasonable values were obtained for the change in body fat (Table 5). There is a good correlation between the fat lost as measured by18O dilution on R+10 and the intake − DLW energy balance calculation (Fig.4,r 2 = 0.95,P = 0.047).

Fig. 4.

Change in body fat using difference in body fat content between R+10 and preflight vs. in-flight energy deficit.

DEXA. The DEXA study also found a loss of body fat. DEXA measurements were made 90 and 60 days preflight and after flight on R+2 and R+10. All four subjects lost fat. According to the DEXA measurement on R+2, the fat lost was 2.4 ± 0.5 kg, and from the R+10 measurement, it was 2.0 ± 0.1 kg. Exact comparison between the DEXA and the two isotopically determined values is not possible because the DEXA measurements were made 1 and 2 mo before flight. It is possible that there were body composition changes during the month before flight. Weights were, however, constant for three of four subjects. For the fourth there was some weight loss between L−60 and L−30. The L−30 value was used to calculate the data in Tables 5 and 8.

There was no statistically significant difference between any of the methods (Table 8). The excellent concordance between the three independent methods and the correlation between the body fat loss as measured by isotope dilution and intake−DLW balance validates both the finding of a negative energy balance and the use of the DLW method to measure energy expenditure on astronauts in earth orbit. Finding such good agreement would be highly improbable if there was something amiss with any of the methods. These are important findings; they validate the methodologies involved for spaceflight experiments.

Nitrogen balance. The preflight N balance is in good agreement with the values found on the previous missions SLS1 and SLS2 (57.5 ± 9.1 mg N ⋅ kg−1 ⋅ day−1) or Skylab (73.9 ± 4.0 mg N ⋅ kg−1 ⋅ day−1). But the in-flight nitrogen balance was much more negative (−29.3 ± 4.9 mg N ⋅ kg−1 ⋅ day−1) than it was on either SLS1 or SLS2 (−16.3 ± 3.4 mg N ⋅ kg−1 ⋅ day−1) or Skylab (−18.5 ± 5.9 mg N ⋅ kg−1 ⋅ day−1; Ref. 21). The severe negative energy balance on LMS accounts for the greater nitrogen loss.

We previously suggested from a comparison of energy intake and nitrogen balance between Skylab and SLS1/2, that astronauts on missions that had high physical activity requirements were either unable or unwilling to eat enough to maintain energy or nitrogen balance (22). The observations from this mission where exercise requirements were much higher than normal because of the extensive testing done confirm this hypothesis. The daily exercise requirements were high, and the astronauts were unable to even approach nitrogen balance.

The bed rest study shows that, provided energy intake is adequate, protein losses are small. Two recent studies reported that exercise prevents the atrophic changes in muscle during bed rest (4, 7). It may do so during bed rest, but the spaceflight situation is different because of the negative energy balance.

The energy deficit on this mission was considerable and could not be sustained for very long. The astronauts had, on average, 20 kg of body fat (Table 5), of which ∼90% (18 kg) is available as an energy reserve (0.9 × 20 × 9,000 ≃ 160,000 kcal). The average daily energy deficit on LMS was 1,400 kcal/day. Thus, if the negative energy balance were to persist at this level by 4 mo, the astronauts would be approaching terminal starvation (160,000/1,335 ≃ 120 days). Obviously this does not happen.

Humans can and do adapt to chronic energy deficits. Adaptive processes are both metabolic and discretionary. An example of the former is the reduction in macronutrient substrate cycling (∼7%), of which the major component is the reduction in protein turnover (24) and, of the latter, a reduction in voluntary activity (27). On the recent shuttle/Mir missions, energy intake for some of the astronauts was low; for five of six subjects for whom we have data, the mean in-flight energy intake was 26.3 ± 2.3 kcal ⋅ kg−1 ⋅ day−1(n = 6; Ref. 23). The Mir crews adapted by reducing the amount of exercise done. Compliance with the mandated exercise regimens has been poor on the shuttle/Mir missions. If the energy costs of spaceflight without exercise are ∼31 kcal ⋅ kg−1 ⋅ day−1, then the approximate daily energy deficit on Mir would have been 80 (kg) × 4 = 240 kcal/day. This degree of negative energy balance would be sustainable for ∼160,000/240 ≃ 660 days. Nevertheless even with accommodation to the reduced intake at some point, metabolic processes will become compromised and the well known consequences of severe undernutrition would occur. Chronic malnutrition is not a healthy state, especially for people in exotic environments where they are required to function efficiently.

Predicting Energy Needs

Astronauts do not have to be in precise energy balance, but they need to be close to it and certainly closer than they were on this mission. It is not necessary that a predictive equation gives exact values for energy expenditure, because a small deficit in intake would have little effect. A large deficit, such as occurred on this mission, would be health-threatening were it to occur on a long-duration mission. For whatever reason, the normal mechanisms for maintaining energy balance do not seem to operate effectively in flights with high levels of physical activity.

In view of the potentially serious negative consequences of a prolonged energy deficit there is a need for an equation for predicting energy needs of an individual astronaut. It is not possible to use a single value for estimating energy needs for all subjects. In 13 subjects, Lane et al. (12) found a range of between 28 and 47 kcal ⋅ kg−1 ⋅ day−1(12). Although the mean value was similar to the World Health Organization estimate for moderate activity, there was no correlation between the calculated and the measured energy expenditure rate (r 2 = 0.02). The need is for an equation that can be applied to the individual subject.

Equation 6 is proposedEnergy needs Equation 6 =(a)×BMR+exercise+EVAactivity where BMR is the basal metabolic rate, exercise is the amount of exercise (treadmill, cycle, etc.) done, and EVA is the energy costs of any extravehicular activity. With recent improvements in space suits, the energy costs of EVA activity are low (H. W. Lane, unpublished observations).

The database for estimating the value of (a), the activity factor, is the LMS data on the four subjects who exercised and the SLS1/2 data for the subjects who did not exercise (Table 9). The activity factor excludes any contribution from programmed exercise, which, because it is so variable, is treated as a separate variable in the equation. There are two data points from SLS2. The first is the group mean for nine subjects. We previously inferred that they were in or close to energy balance (intake = 30.4 ± 1.5 kcal ⋅ kg−1 ⋅ day−1). An energy balance study was done on one subject on that mission, and the result is in agreement with the estimate. Energy expenditure (by DLW) was 28.5 kcal ⋅ kg−1 ⋅ day−1, and energy intake was 29.1 kcal ⋅ kg−1 ⋅ day−1. RMR data were available from SLS2 but not LMS, so we used the equation of Owen et al. (17, 18) to calculate the RMR from the LBM for the pooled SLS1/2 subjects. LBM data were available for the other five subjects from both DEXA and isotope dilution. The data in Table 9 use the isotope dilution data. The results are not different if the DEXA values are used. The calculated value of the slope (a) is 1.39 ± 0.037, so the value of (a) inequation 6 is 1.4 (r 2 = 0.88, P < 0.04).

View this table:
Table 9.

Database for evaluating equation 6

From this study, we can conclude the following.1) The subjects on this mission were in severe negative energy balance and showed the expected consequences of a negative energy balance, namely weight loss, body fat loss, and protein loss. 2) The use of the DLW method for measuring energy expenditure during spaceflight is valid.3) Tissue hydration levels take several days to return to the preflight state after spaceflight.4) Energy expenditure and, hence, energy requirements can be predicted from the equation: Energy needs = 1.40 × BMR + exercise + EVA activity.


We thank the numerous people in the Space and Life Sciences Directorate of the NASA-LBJ Space Center at Houston and Lockheed-Martin Government Services Division for implementing this experiment. Special thanks are due to the bed rest subjects and the crew of the LMS mission for their cooperation.


  • Address for reprint requests and other correspondence: T. P. Stein, Dept. of Surgery, Univ. of Medicine and Dentistry of New Jersey, School of Osteopathic Medicine, 2 Medical Center Drive, Stratford, NJ 08084 (E-mail: tpstein{at}

  • This work was supported by NASA contract NAS 9–18775.

  • The costs of publication of this article were defrayed in part by the payment of page charges. The article must therefore be hereby marked “advertisement” in accordance with 18 U.S.C. §1734 solely to indicate this fact.


View Abstract