Heated dorsal hand vein sampling for metabolic studies: a reappraisal.

The purpose of the current study was to determine the accuracy of a given heated dorsal hand vein (HDHV) measurement in predicting a simultaneous arterial measurement and to validate this technique for use in stable isotope studies. Twenty catheterizations of the femoral artery, femoral vein, and a dorsal hand vein were performed in 13 healthy male subjects. Simultaneous blood samples were obtained from all three sites during primed continuous infusions of L-[1-13C]leucine (Leu) and L-[ring-2H5]phenylalanine (Phe) in the postabsorptive state, with or without intravenous glucose infusion. Relationships were examined by linear regression analysis, with 95% prediction intervals for femoral arterial values determined using the HDHV-derived values as independent variable. Glucose concentrations and isotopic enrichments of ketoisocaproate (KIC), Leu, and Phe were similar in HDHV- and arterial-derived blood, with slopes between 0.9414 and 1.0008, intercepts not different from zero, and r2 values of 0.7613 or greater (P < 0.05). Intercepts for KIC, Leu, and Phe concentrations all were different from zero (P < 0.05), and slopes ranged between 0.7560 and 0.8625. For each analysis the HDHV sample correlated better with the femoral arterial sample than with the femoral venous sample. These data support the use of HDHV sampling as a surrogate for direct arterial sampling but document significant limitations in the technique.