*4.3. Part 1: Optimization of Formulation Cariables by CCD*

The rivastigmine-loaded NLC formulation variables were optimized using a CCD and the data were statistically analyzed using ANOVA, evaluating the importance of the selected factors on dependent responses and the suitability of the selected design for the CQAs through the value of R squared (*R* 2 ). Different ANOVA models were tested (Supplementary Data, Section 3, Table S9) and the one that presented the closest *R* <sup>2</sup> value to 1 (0.93563) for the CQAs was selected, namely the linear/quadratic main effects + 2-ways model [89,90].
