A New Approach to Measure Energy Expenditure in the Neonate

Energy expenditure in the neonate is higher when expressed per unit of body weight than at any other time in the life cycle (¹). Although predictive equations may be useful in healthy children, there is in fact marked variation in energy expenditure in the ill child such that direct measurement is necessary, for individual patients, which then allows for adjusting their nutrition management so as to provide optimal care (²). Patients with inflammation may show as much as a 50% increase over the predicted measurement, while those with muscle atrophy may show as much as a 40% decrease from the predicted measurement (²). There is a large body of literature in which the energy expenditure of neonates has been measured by indirect calorimetry mostly open circuit (ventilated hood) indirect calorimetry (^{1, 3}), which measures oxygen consumption (FO₂) and carbon dioxide production (FCO₂). From these values energy expenditure is calculated using standard equations (¹). It is however possible to determine energy expenditure from FO₂ or from FCO₂ alone (^{1, 4, 5}), provided the macronutrient balance (protein/fat/carbohydrate) of the feeding is known. From the macronutrient balance can be calculated the “food quotient,” which is the ratio of CO₂ produced to O₂ consumed, if the food is completely oxidized to water and CO₂. It is then assumed that this food quotient is representative of the FO₂ and FCO₂ resulting from oxidation of the feedings provided (in this case TPN). This is analogous to the RQ, which is the ratio of FCO₂ to FO₂. The food quotient can be used to substitute for FO₂ when missing (and vice versa) (^{4, 5}). In the neonate who requires supplemental oxygen or who is on a ventilator, indirect calorimetry has its limitations (^4–8). A particular problem with neonates on a ventilator is that cuffed endotracheal tubes are not normally used and hence there is a variable leak, which would result in an under-estimation of energy expenditure (⁴). Yet as Shew and colleagues (⁵) point out, it is important to determine the energy needs of these critically ill neonates. With this in mind they have evaluated the use of the [¹³C] bicarbonate tracer technique as a possible new method to determine the energy expenditure of neonates on ventilators. This new technique was independently introduced by two groups, one from the United Kingdom (⁹) and the other from North America (¹⁰) and validated in adults. It was demonstrated that the isotopic dilution of ¹³CO₂ derived from administration of NaH¹³CO₃ could be used as an estimate of CO₂ production rate. As mentioned above if the food quotient of the diet is known, it is possible to convert the CO₂ production rate into a measured energy expenditure.

Shew et al. (⁵) compared the isotope dilution technique with ventilated hood (respiratory gas exchange) in eight surgical neonates who were on TPN feedings. They do show clinically acceptable relationship between the “gold standard” of indirect calorimetry and the isotope dilution method. One of the problems with the isotope dilution method is that bicarbonate excretion in breath is delayed by retention of the bicarbonate in various body pools. In the present experiment the estimated recovery of label was 67.6% (range 54.5–74.6). Earlier Van Aerde et al. (¹¹) showed that the isotope recovery is dependent on the energy intake, varying from 69.6 to 83.8% of infused ¹³C-bicarbonate. Furthermore, the recovery depends on whether the isotope is given (as it is in the present experiment) as a continuous infusion versus a single bolus dose. Bolus dose studies in human neonates show recoveries of 57 ± 10% (¹²), lower than those seen in the present study or in the study of Van Aerde et al. (¹¹). Fundamentally, CO₂ production rate, determined by the isotope dilution method, is the dose of ¹³C-bicarbonate infused divided by the enrichment of ¹³CO₂ in breath at the isotopic plateau, with a correction for the mass of the isotope infused. Because the dose infused is a constant, variation in the proportion of label recovered is likely to be a source of error. Indeed the authors point to this as a limitation of the earlier work by Kien, who did measure CO₂ production in enterally fed preterm infants using the isotope dilution technique (¹³), but did not correct for variable label recovery, preferring to use a constant to account for percent label recovery. Shew et al. (⁵) are proposing the isotope dilution method for use in babies on ventilators to avoid having to do complete collection of respiratory gases. However, it appears that they will also have to use a constant to account for percent label recovery in ventilated neonates.

It is important also to note that energy expenditure correlates with energy intake and similarly FCO₂ also correlates with energy intake. In an earlier study (¹¹) where the energy intakes varied more widely (37–113 kcal/day) than they do in the current paper, the relationship between the percent of the isotope excreted and FCO₂ was weaker than it is in the present study; specifically the correlation coefficient was 0.65 (r² = 0.42). Whereas in the present study (⁵), they report a correlation of 0.94 (r² = 0.881). The present study is able to predict energy expenditure, from ¹³CO₂ appearance in breath, with a coefficient of variance of 12%, which is comparable to the coefficient of variance for indirect calorimetry in their study of 10%. Whereas in the earlier study with it's larger range in energy, intake, more than twice the variance in the prediction of FCO₂ from ¹³CO₂ production, was reported (¹¹). Therefore a potential limitation of using ¹³CO₂ appearance rates in determining energy expenditure might be the level of energy infused. There is also a concern about the non-protein energy source, that is glucose only, versus glucose plus lipid. In babies fed glucose only as a source of non-protein energy, their RQ (FCO₂/FO₂) was >1 due to lipogenesis from glucose (¹⁴). This was associated with an increased energy expenditure in the group fed glucose only compared with feeding a mixture of glucose plus lipid as a source of non-protein energy. The change in FO₂ however, was significantly smaller than that in FCO₂; hence, this might also affect the proportional recovery of label in ¹³CO₂. It would appear that these variables, that is, the energy level fed and the fat-carbohydrate mixture fed parenterally, may well affect proportional label recovery and hence the ¹³C bicarbonate tracer estimate of energy expenditure.

Using cuffed endotracheal tubes, it is possible to obtain constant and reliable estimates of energy expenditure using indirect calorimetry in adults on a ventilator (¹⁵). Furthermore, several groups have reported their experience in using indirect calorimetry in neonates on ventilators (^{4, 6–8}). One of these groups (⁴) chose to only measure FCO₂ as a way around the problem of gas leaks. The isotope dilution technique is a novel idea but has a number of potential errors mentioned above. Because of gas leak problems, indirect calorimetry in ventilated babies also has its problems and potential errors. It would be useful to explore which approach has the lowest error, isotope dilution or indirect calorimetry in ventilated patients, particularly human neonates. In this regard the work of Lucas et al. (⁴) is of particular note because they showed that recovery of exhaled CO₂ was virtually complete in their ventilated neonates. During inspiration a leak was detected; however, this only affected the FO₂ estimate. Conversely during expiration there was no detectable leak of CO₂ around the tube, hence they were able to obtain accurate estimates of FCO₂. Therefore it should be possible to use the approach taken by Lucas et al. (⁴) to measure directly FCO₂ in ventilated babies and compare it with the estimate obtained from isotope dilution.

A final point worth being mentioning is that blood ¹³C-bicarbonate enrichment can be measured in adults, instead of having to collect breath CO₂ (¹⁶). Hence it is possible to consider using blood instead of breath to measure ¹³C-bicarbonate enrichment in neonates. However, blood volume limitations might be an issue since currently approximately 2 mL of whole blood is needed.

In conclusion it is possible to measure energy expenditure in parenterally fed neonates using ¹³C-bicarbonate dilution. However, since the present study only included eight infants who were clinically stable and received a narrow range of energy intake, further work is needed before the method can be generalized to all infants as an alternative to indirect calorimetry, let alone sick infants on ventilators.