Interactions between 2H and 18O natural abundance variations and DLW measurements of energy expenditure.

Appropriate corrections for 2H and 18O natural abundance are necessary in energy expenditure measurements with the doubly labeled water (DLW) method. The contribution of natural abundance variations to errors in the method is generally ignored if an appropriate dose level is given. Calculation of the appropriate dose level assumes that 2H and 18O natural abundance values are covariant and have a fixed slope. This study has investigated the between- and within-subject natural abundance variations, the latter over periods of time similar to those used in DLW experiment. Although 2H and 18O natural abundance values are covariant between subjects, the slope of their relationship is different from that of the Meteoric Water Line. Within subjects there is no such demonstrable covariance. It is concluded that: 1) the determinants of 2H and 18O natural abundance are different within and between subjects; 2) the concept of an optimal ratio of doses is not valid, and a safer strategy is probably to use a ratio > 10; 3) natural abundance variations may contribute significantly to the magnitude of error in the DLW measurements of energy expenditure.