2.5. Statistical analysis
Data derived from antioxidant assays were fitted into an appropriate regression
model that allowed us to determine AA50 of FPP (mg/ml). Whereas data obtained
from the supplementation study was entered into Medcalc (version 11.4.2, Mariakerke,
Belgium) and Prism™ (GraphPad, version 3.0) for statistical analysis. After
omission of outliers, statistical differences within groups over time were determined
using Student’s paired samples t-test, where data was non-normal, the
non-parametric alternative Wilcoxon test was used. Statistical differences between
groups over time were determined using Student’s independent samples t-test,
where Mann–Whitney test was used for non-normal data. A two-tailed P value less
than 0.05 was considered to be significant. Significance was set at 95% CI.