Normality, homoscedasticity and independence were tested previously on recollected data, then a two-way analysis of variance was applied to determine its interactions. If significant (P < 0.05) differences were found, Tukey tests were used to compare averages between treatments. In each carbohydrate content, protein diet contents were compared by one-way ANOVA followed by a Tukey test. All values were calculated using Minitab statistical software (version 16.1, Minitab State College, PA, USA).