INTRODUCTION

Top Abstract Introduction Methods Results Discussion Appendix References

THE ENERGY COST of physical activity (EEA) has been estimated to account for 5-17% of overall energy expenditure in neonates (3, 5, 8, 9). These values are usually derived from indirect calorimetry studies, which provide measurement of total energy expenditure (TEE), resting metabolic rate (RMR), and diet-induced thermogenesis (DIT). EEA can then be estimated by TEE  (RMR + DIT). Other estimates have come from determination of O₂ consumption (O₂) in different activity states. If these estimates are valid, EEA can account for a significant portion of the daily energy expenditure in these infants, and in infants with poor growth, EEA could represent a portion of the energy balance that can be manipulated by changes in care practices.

To measure the contribution of EEA to overall energy expenditure and to determine its contribution to the variability in TEE in preterm infants, we sought to directly measure EEA. A force plate system was designed that provides real-time measurements of mechanical work. In this system the infant lies on the force plate, and the force the infant exerts on the plate is converted to work measurements. By use of this device in conjunction with indirect calorimetry, metabolic measurements (i.e., TEE, O₂, CO₂ production, and respiratory quotient) and EEA determinations were performed in enterally fed, preterm infants. TEE and EEA were also determined for different activity states. This provides the first direct measurement of EEA as a component of TEE in this population.

	METHODS

Top Abstract Introduction Methods Results Discussion Appendix References

Force Plate Design

Force Plate Calibration and Validation

Static calibration. The strain-gauge sensors were individually characterized as follows: the plate surface was divided into 50 equal 1.5 × 1.5-in. grids. Calibration masses (Rice Lake Bearing, Rice Lake, WI) ranging from 50 to 2,500 g were placed in each grid multiple times, and the force reading from each sensor was recorded. After subtraction of the unloaded static force reading from each sensor's output, the uncorrected force readings were subjected to a multiple linear regression with the expected sum of the four sensor outputs equal to the calibrated mass. The evaluation of this regression provided a scale factor for each sensor that was then used to correct the individual sensor output values.

Static validation. Static validation studies were performed to determine whether the above scale factors for each sensor accurately measured the force produced by a mass placed at any point on the plate. Precision masses ranging from 50 to 2,500 g were placed in each grid in multiple trials. These masses were chosen in that they were expected to represent the minimal and maximal masses that might be generated by the body of a preterm infant. In initial testing it was determined that the strain-gauge sensors were temperature sensitive over an environmental temperature range of 2-34°C. Therefore, all validation studies were conducted at approximately the same temperature range measured inside the infant study chamber.

Dynamic validation. A motor-pulley arrangement was used to vertically reciprocate (i.e., move up and down) a known mass over a range of velocities (see APPENDIX for apparatus design and work calculation). For each trial, different masses, mass velocities, and mass positions on the plate were tested. Data are expressed as work extrapolated to a 24-h measurement (kcal · kg¹ · day¹). Results are presented as measured work as a percentage of predicted work.

Clinical Study

Study population. The eligible study population consisted of nonventilated, stable, enterally fed, growing preterm infants who weighed 2 kg at birth and received care in the Neonatal Intensive Care Unit at the University of Colorado Health Sciences Center. Enrollment criteria excluded infants who had acute intercurrent illnesses or congenital anomalies. Infants could be the appropriate size for gestational age (AGA) or small for gestational age (SGA). Written parental consent for infant participation was obtained before the study. The protocol was approved by the Institutional Review Board at the University of Colorado Health Sciences Center.

Study methods and measurements. To determine metabolic measurements (i.e., O₂, CO₂ production, and TEE), indirect calorimetry was performed for ~6 h in each infant with use of a calorimeter designed specifically for use in preterm infants, including those who require supplemental O₂ (18). Six-hour studies were performed, since this has been determined to be the minimum measurement duration that accurately reflects daily TEE in preterm infants (1). Calorimetry results are presented as 1-min average values. Studies were conducted as previously described (18) with the infant placed in a clear, Plexiglas chamber. Humidified medical-grade compressed tank gas was used for the inspiratory gas source (40-60% relative humidity, with relative humidity stable within 10% for each subject, monitored using humidity and temperature indicator HMI 31, Vaisala, Helsinki, Finland). Ambient temperature was maintained by placing the chamber under a servo-controlled radiant warmer. Continuous electrocardiogram and respiratory rate monitoring, pulse oximetry, and axillary temperature measurements were performed. After the infant was placed in the study chamber, metabolic measurements were not recorded until after an equilibration period of 20 min. Most infants were removed from the calorimetry chamber for feeding and nursing care for 10-20 min before each feeding.

Data Analysis

Values are means ± SE. Unpaired Student's t-test was used to compare differences between groups.

	RESULTS

Top Abstract Introduction Methods Results Discussion Appendix References

Force Plate Calibration and Validation

Static validation. Precision masses ranging from 20 to 1,000 g were placed in each grid on multiple trials. Results are shown in Table 2.

Dynamic validation. Dynamic validation results (Table 3) indicate that work recovery from the plate averaged 88.0 ± 1.1 and 97.2 ± 2.5% for work simulations of <1 and >1 kcal · kg¹ · day¹, respectively. The plate was sensitive enough to determine the caloric expenditure from respiratory activity at rest (Fig. 1). A typical 10-min activity recording is shown in Fig. 2.

View larger version (21K): [in this window] [in a new window]   Fig. 1.   Ten-second force plate recording from infant in deep sleep with no visually detected motion. Data at 128 Hz were smoothed using a 2-s rolling average. Respiratory rate in this infant averaged 48-50 breaths/min on cardiorespiratory monitor during this period of deep sleep, and heart rate was 148-160 beats/min. Respiratory pattern is identical to that detected by cardiorespiratory monitor.

View larger version (20K): [in this window] [in a new window]   Fig. 2.   Raw mass data recording over a 10-min interval from preterm infant.

Patient Studies

1

1

1

1

View larger version (38K): [in this window] [in a new window]   Fig. 3.   Energy expenditure of physical activity (top), total energy expenditure (middle), and percentage of total energy expenditure accounted for by energy expenditure of physical activity for each infant (bottom). Values are means ± SE; infants are numbered as in Table 4 (i.e., 1-24).

The mean time spent in each of five activity states and the mean caloric expenditure in each of these states are shown in Fig. 4. As has been described by other investigators (12), preterm infants in this study spent an average of 80-90% of their time in sleep (quiet sleep plus active sleep in Fig. 4, top).

View larger version (33K): [in this window] [in a new window]   Fig. 4.   Time spent in each of 5 activity states (top) and for caloric expenditure in each of these states for all infants (bottom). Values are means ± SE.

RMR in infants is generally considered to be the energy expenditure determined by indirect calorimetry while an infant is quietly sleeping 2.5-3 h after the last meal. In this study RMR averaged 63.8 ± 1.6 kcal · kg¹ · day¹. Energy expended for quiet breathing (i.e., the only visible motor movement such as seen in Fig. 1) averaged 0.32 ± 0.03 kcal · kg¹ · day¹. Not all infants actively cried during the study, but for those who did, the increase in energy expenditure above that seen in quiet sleep ranged from 12 to 55%. There was a linear correlation between the mean TEE in each activity state and the numerical assignment for each state (Fig. 5).

View larger version (19K): [in this window] [in a new window]   Fig. 5.   Relationship between total energy expenditure in each activity state and numerical assignment for each state. Values are means ± SE for all infants.

We compared our direct measurement of EEA with an estimate of EEA based on our data using one of the estimation methods reported in the literature. Freymond et al. (5) estimated EEA by subtracting the postmeal resting energy expenditure (obtained by regressing mean recorded energy expenditure vs. activity state) from the TEE and arrived at an EEA of 68.4  64.8 = 3.6 kcal · kg¹ · day¹. In our study, mean postmeal RMR for all infants was 66.2 kcal · kg¹ · day¹, and therefore EEA estimated by the technique of Freymond et al. would be 69.2  66.2 = 2.3 kcal · kg¹ · day¹. This estimate is very close to our mean direct EEA measurement of 2.4 kcal · kg¹ · day¹.

In this study, infants spent a majority of their time in the two sleeping states. These states accounted for the lowest EEA. Although some infants were gavage fed during the measurements, no work measurements were made during nursing care or feedings involving infant handling, since measured work might be secondary to the caregiver's movement that is detected by the plate and not work of the infant. To estimate the additional EEA expended during nursing care, two investigators independently performed state assignments each minute during nursing interventions in a subpopulation of the study infants (n = 5). The time spent in each state was determined, and the energy expenditure of nursing care was determined from the product of the time in each times the average daily energy expenditure for each state as determined from the study. In these five infants, care periods averaged 11 min with a range of 4-25 min. One infant slept through the care periods and gavage feeding, while at the other extreme one infant spent the majority of the 18-min care period in states 4 and 5, with an average EEA during this period of 19 kcal · kg¹ · day¹. If this latter infant received care every 3 h, several more kilocalories per kilogram per day would be added to this infant's estimated EEA.

	DISCUSSION

Top Abstract Introduction Methods Results Discussion Appendix References

We have reported the design, validation, and clinical use of a force plate system for measuring the mechanical work of movement in infants. This is the first study to directly measure the EEA in preterm infants. Previous determinations of the amount of energy expended in movement in these infants have been estimates determined by a variety of different techniques. In general, these have involved subtracting the measurable components of energy expenditure from TEE.

A force plate system was designed to provide real-time measurements of mechanical work in preterm infants. This system is analogous to the floor-sized force plates that have been recently described in adult whole- room calorimeters (15, 16). The authors of these studies state that, at the time of publication, their force plate system was the most accurate method to determine mechanical work in humans over extended periods of time. Our system is based on the same principles.

With physical activity, internal (metabolic or muscular) and external (mechanical) work are performed. Virtually all mechanical work done by the body is work performed when the body moves against gravity, and "downward" force vectors, which can be detected by load cells, are created. Activities such as walking involve up-and-down movement of the extremities against gravity plus movement of the body's center of gravity, both of which generate downward forces. As such, when an infant makes horizontal "swimming" motions with the arms and legs, the infant's center of gravity changes, and this is detected by our force platform and converted to work equivalents. Occasionally, horizontal movement occurs without a change in the center of gravity if this movement is countered by an equal and opposite movement, and the summation of the mechanical work performed does not produce a net downward force. In our experience involving hundreds of hours of close infant observation, even the smallest arm or leg motions are detected by the force platform.

This force platform allows for reasonable estimates of external work. The measurement of internal work is extremely difficult, because it involves measurement of TEE by a technique such as indirect calorimetry with the subtraction of other sources of energy expenditure (i.e., mechanical work, DIT, basal metabolic rate, and energy expenditure of growth), and some of these values are indirectly derived. In this study we sought to measure only external or mechanical work. The internal work of physical activity was not, and cannot practically be, measured.

There have been several attempts to estimate energy expenditure at different levels of activity (3, 13, 14). The first detailed determination of the relationship of activity score to energy expenditure was by Freymond et al. (5). They described a curvilinear relationship (3rd-degree regression curve, r = 0.752) between the full (10-point) Bruck's activity score and energy expenditure but a linear relationship with use of a simplified scale that more closely resembled the modified scale used in our study. In our study there was a strong linear correlation between the mean TEE in each activity state and the numerical assignment for each state (Fig. 5). All activity scales have somewhat arbitrary numerical assignments. However, they are useful in assessment of activity.

The TEE in our study of 69.2 kcal · kg¹ · day¹ is comparable to the values of 66.4 kcal · kg¹ · day¹ reported by Sauer et al. (11) in infants after the 1st wk of life and 68.4 kcal · kg¹ · day¹ reported by Freymond et al. (5). The somewhat higher TEE in our study than in similar populations in other reports (1, 3, 9, 12) may have been partly due to the large number of SGA infants in this study, and as a group their TEE was higher than that of AGA infants.

The mechanical work of quiet breathing in this study was 0.32 ± 0.03 kcal · kg¹ · day¹, or 0.4% of simultaneous TEE, and in this population of healthy, growing preterm infants it had a minimal effect on overall energy expenditure. This force plate is not designed to measure work of breathing, but it would probably give a good estimate of work of combined lung and abdominal displacement involved with breathing. Although not easily done, actual work of breathing is probably best measured by measuring O₂ per unit of ventilation with the remainder of the body in a "quiet" state so that there is no contribution to O₂ by nonrespiratory muscles (10). Estimation of the work of breathing in preterm infants with respiratory distress has been problematic. In the first few days of life, there are reports of a strong correlation of O₂ with ventilatory index (19) and no correlation between the two (7). In preterm infants with more chronic lung disease, a correlation between ventilatory index and O₂ has been demonstrated, although the spontaneous breath rate did not vary (2). The small muscle mass in these infants may limit the amount of energy that can be expended with breathing.

Actual daily TEE of mechanical work is probably slightly higher than that determined by our study for several reasons. First, work recovery averaged 90%, as might be expected with a mechanical device, but this would add only 0.3 kcal · kg¹ · day¹ to the work measurement value. In addition, no work measurements were made during nursing care or feedings involving infant handling, since measured work might be caused by caregiver movement that is detected by the plate and not movement of the infant. Thus, on the basis of the observational study during nursing care, it is not unreasonable to add 1-3 kcal · kg¹ · day¹ to our measured value of 2.4 kcal · kg¹ · day¹ (depending on an infant's relative activity compared with other infants) to derive a value for EEA in this patient population.

In summary, this study is the first to directly measure the EEA in infants. The majority of variability in individual patient caloric expenditure is attributable to the stress of nursing care. For most stable preterm infants, work expended in motor activity accounts for only a small component of TEE (3.5%). However, for infants with poor growth, excessive irritability, or poor response to nursing care and interventions, energy expenditure from physical activity could have a major impact on growth.