Statistical analysis of the GAB and Henderson equations for sorption isotherms of foods

GAB and Henderson equations were fitted to experimental sorption isotherms of wheat, apricot and cassava. The statistical significance of equation parameters were tested with joint interval confidence as well as differences between linear regression and non-linear regression fits. The results showed that two equations (GAB and Henderson) had a good fit (r(2) > 0.95) with respect to experimental points for linear and non-linear fits. Nevertheless, the GAB equation can predict water activity values grater than one, and this result is a thermodynamic inconsistency. The statistical analysis showed that some parameters with theoretical basis of the GAB equation had no significance. This important result shows the possibility that theoretical basis of the GAB equation must be revised.