OBJECTIVE: To compare percentage of body fat (%BF) estimates from dual-energy x-ray absorptiometry (DEXA) with those derived from total body water (TBW) determination by deuterium oxide (D2O) dilution. ANIMALS: 31 healthy, adult, purebred dogs of various ages and breeds (body weight, 15 to 39 kg). PROCEDURE: The TBW was measured by D2O dilution and subsequent analysis via nuclear magnetic resonance (NMR). Blood was collected before and 2 hours after IV administration of 275 mg of D2O/kg of body weight. Plasma was separated and stored at -30 C until analysis by deuterium NMR. The DEXA scans were obtained immediately after blood collection from dogs under general anesthesia. RESULTS: Measurements of %BF by DEXA averaged 15.8% higher than calculated estimates of TBW content by D2O dilution. The linear regression of %BF by TBW content on %BF by DEXA had a slope of 1.04 and a correlation coefficient of 0.84, indicating excellent relative agreement between methods despite the significant difference in absolute agreement between the 2 methods. The average difference between methods did not differ by breed, sex, body condition score, body weight, or %BF, as measured by DEXA. CONCLUSION: Comparability of our data with those of previous studies suggest that DEXA is useful for in vivo estimation of body composition in healthy dogs. Body fat estimated by D2O dilution will be less than that determined by DEXA, despite excellent relative agreement between methods.
OBJECTIVE: To compare percentage of body fat (%BF) estimates from dual-energy x-ray absorptiometry (DEXA) with those derived from total body water (TBW) determination by deuterium oxide (D2O) dilution. ANIMALS: 31 healthy, adult, purebred dogs of various ages and breeds (body weight, 15 to 39 kg). PROCEDURE: The TBW was measured by D2O dilution and subsequent analysis via nuclear magnetic resonance (NMR). Blood was collected before and 2 hours after IV administration of 275 mg of D2O/kg of body weight. Plasma was separated and stored at -30 C until analysis by deuterium NMR. The DEXA scans were obtained immediately after blood collection from dogs under general anesthesia. RESULTS: Measurements of %BF by DEXA averaged 15.8% higher than calculated estimates of TBW content by D2O dilution. The linear regression of %BF by TBW content on %BF by DEXA had a slope of 1.04 and a correlation coefficient of 0.84, indicating excellent relative agreement between methods despite the significant difference in absolute agreement between the 2 methods. The average difference between methods did not differ by breed, sex, body condition score, body weight, or %BF, as measured by DEXA. CONCLUSION: Comparability of our data with those of previous studies suggest that DEXA is useful for in vivo estimation of body composition in healthy dogs. Body fat estimated by D2O dilution will be less than that determined by DEXA, despite excellent relative agreement between methods.