274. Phenolic Composition, Not Total Phenolic Concentration, Predicts Total Antioxidant Capacity as Measured by the QUENCHER-CUPRAC Assay in Pet Food Fiber Ingredients.

Abstract Phenolic compounds are a structurally diverse class of plant-derived phytochemicals commonly associated with cell wall material and dietary fiber. Interest in phenolics has grown due to their antioxidant properties and potential roles in mitigating oxidative stress in both foods and in vivo. Despite their relevance, limited information is available on the phenolic composition and antioxidant capacity of fiber ingredients commonly incorporated into pet foods. The objective of this study was to characterize phenolic profiles and total antioxidant capacity (TAC) in a variety of fiber ingredients and evaluate the relationships between them. Thirteen cryogenically ground ingredients were analyzed for phenolic composition via HPLC-UV using a 15-compound targeted panel, and TAC was quantified via the QUENCHER-CUPRAC method. Because phenolic and TAC values were right-skewed, Spearman’s rank correlations were used to test associations between TAC and (1) the sum total of measured phenolic concentration and (2) individual phenolic compounds. Principal component analysis (PCA; correlation matrix; z-scaled variables) was performed to characterize the multivariate structure of the phenolic data, and TAC was projected as a supplementary quantitative variable to evaluate relationships with principal component (PC) vectors without influencing their construction. Linear regression was subsequently used to assess the significance and predictive strength of PCs associated with TAC. Two ingredients were excluded from analysis due to the absence of any detectable compounds. Additionally, one compound was not detected in any samples and was removed from analysis. Total phenolic concentration was not associated with TAC (ρ = 0.02, p = 0.96). Additionally, no individual phenolic compound exhibited a significant pairwise association with TAC (p > 0.05). PCA revealed a strong multivariate phenolic structure, with the first four PCs capturing ∼94% of total variance. The loadings of PC1 (42.1%) were evenly distributed between protocatechuic acid, vanillin, caffeic acid, syringaldehyde, p-coumaric acid, and cinnamic acid, whereas PC2 (30.3%) was driven by chlorogenic acid, vanillic acid, p-hydroxybenzaldehyde, and sinapic acid. Conversely, PC3 (13.5%) was loaded heavily with syringic acid and naringin, whereas PC4 (8.3%) was loaded heavily with gallic acid, with a moderate negative loading from ferulic acid. TAC showed moderate correlations with PC2 (r = 0.63) and PC4 (r = 0.62). When included in a linear regression model, both PC2 and PC4 were significant predictors (p < 0.01), and jointly explained a large proportion of TAC variability (adj. R2 = 0.725, p = 0.002). Among this dataset, compounds positively associated with PC2 and PC4 appear to be more related to TAC. These data suggest that phenolic composition, rather than total concentration, drives antioxidant potential in fiber ingredients. Consequently, diet formulations aimed at enhancing antioxidant functionality may benefit from targeted phenolic profiles, rather than maximizing for total phenolic content.