Isotope-specific fractionation corrections improve doubly labelled water estimates of energy expenditure in mice

There has been debate over the appropriate equations for calculation of daily energy expenditure (DEE) in doubly labelled water (DLW) studies. Previous studies that have used tritium as the hydrogen label have generally not corrected for fractionation effects, while studies using deuterium generally have. Some validation data are available to support both positions, but as yet very little work has compared the different approaches in animals simultaneously injected with both isotopes. We compared the metabolizable energy intake (MEI) of mice ( Mus musculus , outbred MF1, n = 15) averaged over 5 days, to the calculated DEE measured on day 6 using the DLW method, for mice injected simultaneously with tritium, deuterium and 18 oxygen. This allowed us to compare in the same animals the validity of the contrasting equations. Using the deuterium data and the equations that correct for deuterium fractionation, the average difference between DEE and MEI was 0.6 ± 11.4%. Using tritium data and equations that account for tritium fractionation, the average difference was 1.2 ± 11.6%. The discrepancy was greater if fractionation effects were ignored. Using the deuterium data, the discrepancy averaged 2.5 ± 11.4% and using the tritium data it averaged 3.7 ± 11.7%. Applying the tritium data in the equations that correct for deuterium fractionation resulted in an intermediate discrepancy of 1.9 ± 11.6%. Using the difference between deuterium and tritium washouts to estimate the fractionated water loss produced physiologically impossible individual estimates in 13 of 15 mice. Mice ate less food and lost body mass on the day when the DLW method was applied compared to the previous 5 days, but energy expenditure was apparently unaffected.