Doubly labeled water
In practice, for both practical and safety reasons, almost all recent applications of the "doubly labeled water" method use water labeled with the heavy, non-radioactive forms of the elements deuterium and oxygen-18 (O-18 or 18O), or deuterium oxide-18 (D218O). In theory, radioactive forms of hydrogen and oxygen could be used for such labeling, and this was the case in many early applications of the method.
In particular, use of the doubly labeled water method (or DLW method) yields a particular type of measurement of metabolic rate, in which the average metabolic rate of an organism is measured over a period. This is done by administering a dose of doubly labeled water, and then measuring the elimination rates of deuterium and O-18 in the subject over time through the regular sampling of heavy isotope concentrations in the body water (by sampling saliva, urine, or blood). The minimum number of samples required is two—an initial sample after the isotopes have reached equilibrium in the body, and a second sample some time later. The time between the collecting of these samples depends on size of the animal involved. In small animals the period may be as short as 24 hours, and in larger animals like adult humans, the period may be as long as 14 days. In animals this average daily metabolic rate measured by the DLW method is often also called the Field metabolic rate or FMR. The method was invented in the 1950s by Nathan Lifson and colleagues at the University of Minnesota; however, its use was restricted to small animals until the 1980s because of the high cost of the oxygen-18 isotope. Advances in mass spectrometry during the 1970s and early 1980s reduced the amount of isotope required, which made it feasible to apply the method to humans. The first application to humans was in 1982, by Dale Schoeller, over 25 years after the method was initially discovered. A complete summary of the technique is provided in a book by British biologist John Speakman.
Mechanism of the test 
The technique measures a subject's carbon dioxide production during the interval between first and last body water samples. The method depends on the details of carbon metabolism in our bodies. When cellular respiration breaks down carbon-containing molecules to release energy, carbon dioxide is released as a byproduct. Carbon dioxide contains two oxygen atoms and only one carbon atom, but food molecules such as carbohydrates do not contain enough oxygen to provide both oxygen atoms found in CO2. It turns out, one of the two oxygen atoms in CO2 is derived from body water. If the oxygen in water is labeled with 18O, then CO2 produced by respiration will contain labeled oxygen. In addition, as CO2 travels from the site of respiration through the cytoplasm of a cell, through the interstitial fluids, into the bloodstream and then to the lungs some of it is reversibly converted to bicarbonate. So, after consuming water labeled with 18O, the 18O equilibrates with the body's bicarbonate and dissolved carbon dioxide pool (through the action of the enzyme carbonic anhydrase). As carbon dioxide is exhaled, 18O is lost from the body. This was discovered by Lifson in 1949. However, 18O is also lost through body water loss (such as urine and evaporation of fluids). However, deuterium (the second label in the doubly labeled water) is lost only when body water is lost. Thus the loss of deuterium in body water over time can be used to mathematically compensate for the loss of 18O by the water-loss route. This leaves only the remaining net loss of 18O in carbon dioxide. This measurement of the amount of carbon dioxide lost is an excellent estimate for total carbon dioxide production. Once this is known, the total metabolic rate may be estimated from simplifying assumptions regarding the ratio of oxygen used in metabolism (and therefore heat generated), to carbon dioxide eliminated (see respiratory quotient). This quotient can be measured in other ways, and almost always has a value between 0.7 and 1.0, and for a mixed diet is usually about 0.8.
In lay terms:
- Metabolism can be calculated from oxygen-in/CO2-out.
- DLW ('tagged') water is traceable hydrogen (deuterium), and traceable oxygen (18O).
- The 18O leaves the body in two ways: (i) exhaled CO2, and (ii) water loss in (mostly) urine, sweat, & breath.
- BUT: the deuterium leaves only in the second way (water loss).
SO: from deuterium loss, we know how much of the tagged water left the body as water. And, since the concentration of 18O in the body's water is measured after the labeling dose is given, we also know how much of the tagged oxygen left the body in the water. (A simpler view is that the ratio of deuterium to 18O in body water is fixed, so total loss-rate of deuterium from the body multiplied by this ratio, immediately gives the loss rate of 18O in water.) Measurement of 18O dilution with time gives the total loss of this isotope by all routes (by water and respiration). Since the ratio of 18O to total water oxygen in the body is measured, we can convert 18O loss in respiration to total oxygen lost from the body's water pool via conversion to carbon dioxide. How much oxygen left the body as CO2 is the same as the CO2 produced by metabolism, since the body only produces CO2 by this route. The CO2 loss tells us the energy produced, if we know or can estimate the respiratory quotient (ratio of CO2 produced to oxygen used).
Practical isotope administration 
Doubly labeled water may be administered by injection, or orally (the usual route in humans). Since the isotopes will be diluted in body water, there is no need to administer them in a state of high isotopic purity, no need to employ water in which all or even most atoms are heavy atoms, or even to begin with water which is doubly labeled. Nor is it necessary to administer exactly one atom of 18O for every two atoms of deuterium. This matter in practice is governed by the economics of buying 18O enriched water, and the sensitivity of the mass-spectrographic equipment available.
In practice, doses of doubly labeled water for metabolic work are prepared by simply mixing a dose of deuterium oxide (heavy water) (90 to 99%) with a second dose of H218O, which is water which has been separately enriched with 18O (though usually not to a high level, since doing this would be expensive, and unnecessary for this use), but otherwise contains normal hydrogen. The mixed water sample then contains both types of heavy atoms, in a far higher degree than normal water, and is now "doubly labeled." The free interchange of hydrogens between water molecules (via normal ionization) in liquid water ensures that the pools of oxygen and hydrogen in any sample of water (including the body's pool of water) will be separately equilibrated in a short time with any dose of added heavy isotope(s).
The doubly labeled water method is particularly useful for measuring average metabolic rate (Field metabolic rate) over relatively long periods of time (a few days or weeks), in subjects for which other types of direct or indirect calorimetric measurements of metabolic rate would be difficult or impossible. For example, the technique can measure the metabolism of animals in the wild state, with the technical problems being related mainly to how to administer the dose of isotope, and collect several samples of body water at later times to check for differential isotope elimination.
Most animal studies involve capturing the subject animals and injecting them, then holding them for a variable period before the first blood sample has been collected. This period depends on the size of the animal involved and varies between 30 minutes for very small animals to 6 hours for much larger animals. In both animals and humans, the test is made more accurate if a single determination of respiratory quotient has been made for the organism eating the standard diet at the time of measurement, since this value changes relatively little (and more slowly) compared with the much larger metabolic rate changes related to thermoregulation and activity.
Because the heavy hydrogen and oxygen isotopes used in the standard doubly labeled water measurement are non-radioactive, and also non-toxic in the doses used (see heavy water), the doubly labeled water measurement of mean metabolic rate has been used extensively in human volunteers, and even in infants and pregnant women. The technique has been used on over 200 species of wild animals (mostly birds, mammals and some reptiles). Applications of the method to animals have been reviewed.
- Lifson, N., Gordon, G.B. and McClintock, R. (1955) Measurement of total carbon dioxide production by means of D218O. J. Appl. Physiol., 7, 704–710.
- Lifson, N. and McClintock R. (1966) Theory of use of the turnover rates of body water for measuring energy and material balance. J. Theor. Biol., 12, 46–74.
- Speakman JR (October 1998). "The history and theory of the doubly labeled water technique". Am. J. Clin. Nutr. 68 (4): 932S–938S. PMID 9771875.
- Schoeller, D.A. and van Santen, E. (1982) Measurement of energy expenditure in humans by doubly labelled water. J. Appl. Physiol., 53, 955–959.
- Speakman, J.R., Doubly Labelled Water: Theory and Practice. Springer Scientific publishers. ISBN 0-412-63780-4 ISBN 978-0412637803 416pp)
- Lifson, N., Gordon, G.B., Visscher, M.B. and Nier, A.O. (1949) The fate of utilized molecular oxygen and the source of the oxygen of respiratory carbon dioxide, studied with the aid of heavy oxygen. J. Biol. Chem. A , 180, 803–811.
- Jones PJ, Winthrop AL, Schoeller DA, et al. (March 1987). "Validation of doubly labeled water for assessing energy expenditure in infants". Pediatr. Res. 21 (3): 242–6. doi:10.1203/00006450-198703000-00007. PMID 3104873.
- Heini A, Schutz Y, Diaz E, Prentice AM, Whitehead RG, Jéquier E (July 1991). "Free-living energy expenditure measured by two independent techniques in pregnant and nonpregnant Gambian women". Am. J. Physiol. 261 (1 Pt 1): E9–17. PMID 1858878.
- Speakman, JR (2000) The cost of living: Field metabolic rates of small mammals. Advances in Ecological Research 30: 177–297
- Nagy, KA (2005) Field metabolic rates and body size. Journal of Experimental Biology 208, 1621–1625.