The UNIVARIATE procedure (SAS Inst. Inc., Cary, NC)
was used to confirm the homogeneity of variance and also
to analyze for outliers, but no outliers were identified. To
compare differences among diets, data were subjected to
ANOVA using the MIXED procedure of SAS. The fixed effect
of the diet (n¼3) and the random effect of the experimental
period (n¼3) and pig (n¼6) were included in the
main model. Treatment means were calculated using the
LSMEANS statement and means were separated using
the PDIFF option. Least squares means were compared
using the Tukey–Kramer adjustment. The differences were
considered significant if Po0.05 and were considered a
trend if the P-value was between 0.05 and 0.10.