Two-point vs multipoint sample collection for the analysis of energy expenditure by use of the doubly labeled water method.

Energy expenditure over a 2-wk period was determined by the doubly labeled water (2H2(18)O) method in nine adults. When daily samples were analyzed, energy expenditure was 2859 +/- 453 kcal/d (means +/- SD); when only the first and last time points were considered, the mean calculated energy expenditure was not significantly different (2947 +/- 430 kcal/d). An analysis of theoretical cases in which isotope flux is not constant indicates that the multipoint method can cause errors in the calculation of average isotope fluxes, but these are generally small. Simulations of the effect of analytical error indicate that increasing the number of replicates on two points reduces the impact of technical errors more effectively than does performing single analyses on multiple samples. It appears that generally there is no advantage to collecting frequent samples when the 2H2(18)O method is used to estimate energy expenditure in adult humans.