Reliability of in vivo neutron activation analysis for measuring body composition: Comparisons with tracer dilution and dual-energy x-ray absorptiometry

In vivo neutron activation (IVNA) analysis has the capacity to measure several total body elements in human subjects. Although it has been considered a criterion method for the past 3 decades, the reliability of IVNA analysis has been tested only in phantom calibrations. In 5 male weight-stable patients with AIDS, total body N, Ca, CI, Na, P, and C were measured three times in 16 weeks at Brookhaven National Laboratory. With tracer dilution methods for total body water (TBW) by (H2O)-H-3 and for extracellular water (ECW) by (SO4)-S-35 and NaBr, and dual-energy x-ray absorptiometry (DXA), total body calcium (TBCa) and fat percentage were measured within 2 weeks of IVNA measurements. For comparison, tracer dilution for TBW by D2O and ECW by NaBr, plus DXA measurements, were performed three times in 5 weight-stable healthy volunteers. The reliability of the IVNA technique was very high in patients with AIDS; it ranged from 0.99 for total body chloride (TBCl) to 0.84 for total body phosphorus (TBP), and it agreed with phantom calibration results in the literature. The reliability for measuring fat percentage and TBCa by DXA was similar in patients with AIDS and in healthy volunteers. Tracer dilution for measuring TBW by (H2O)-H-3 in patients with AIDS and by D2O in healthy volunteers had a reliability score similar to those found with IVNA and DXA. The reliability scores for measuring ECW in patients with AIDS by (SO4)-S-35 and NaBr, 0.66 and 0.68, respectively, were the lowest among all measurements, whereas the reliability score for NaBr in healthy volunteers was 0.96, as with the other measurements.