A nonlinear regression-kinetic method for quantification of serum triglycerides

We report a new kinetic approach to the quantification of triglycerides (triacylglycerols) in human sera. The new approach combines a commercially available enzyme-reagent system with a multiple-linear-regression data-processing method. Values of absorbance collected at 3-s intervals between 20 and 305 s are fitted to a pseudo-first-order model to compute the total change in absorbance expected if the reaction were to proceed to equilibrium. The computed absorbance change varies linearly with triglyceride concentration between 3 and 20 mumol/L (after 250-fold dilution of sample). The pooled relative standard deviation was 7.2% for 23 serum samples with triglyceride concentrations between 3 and 20 mumol/L. Comparison of kinetic (y) and equilibrium (x) absorbance changes for the 23 sera gave a least-squares equation of y = 0.994x + 0.00066, confirming good agreement between the methods. The temperature coefficient is less than 0.2% per degree Celsius.