Revision of calculations in the doubly labeled water method for measurement of energy expenditure in humans.

In the doubly labeled water (DLW) method for the measurement of energy expenditure in humans, the basis of the calculation for CO2 production is the difference between the products of the rate constants for the disappearance of 18O and 2H from body water (KO, and KD, respectively) and the matching isotope dilution spaces (NO and ND, respectively). Thus, omitting corrections for isotope fractionation, CO2 production = 0.5 (KONO-KDND). In this calculation, it is also customary to normalize observed NO and ND values to a fixed value for ND/NO. The increasing use of the method has resulted in the generation of substantially more information on the normal value for ND/NO than existed at the time the method was first developed, and recent work has suggested that revisions of the originally used value of 1.03 may now be deemed appropriate. Values of 1.034 or 1.0427 have recently been suggested, but when applied in energy expenditure studies these estimates would lead to significantly different expenditure measurements. It can, however, be shown from published work and direct experimental study that ND/NO values are method dependent, and for these reasons the lower revised value of 1.034 appears to be more acceptable. The possibility that particular populations may ultimately be shown to be different from 1.034 should not, however, be dismissed entirely, and for this reason we suggest that information derived in individual experiments could be used in a Bayesian fashion to generate new ND/NO estimates. The appropriate techniques are described.