Flavor misattribution: A novel approach to improving compliance and blinding in food-based clinical interventions

• Taste obscured for a clinical intervention by intentional flavor misattribution.
• A data-guided “masking” carrier complemented broccoli taste and aroma attributes.
• Trained descriptive analysis panelists deconstructed the broccoli sensory profile.
• Panelists matched complementary flavors using a test–retest experimental approach.
• Pineapple, lime, and ginger identified by factor analysis as optimal masking agents

BackgroundClinical trials that test food-based interventions frequently suffer from ineffective blinding of study participants which can reduce the statistical power of reported outcomes, and can lead to poor compliance. This study used descriptive sensory analysis with highly trained evaluators, and well-validated statistical techniques to develop a protocol to mask the consumption of phytochemical-rich broccoli sprout extracts (BSEs3) for the use in clinical studies seeking to address a variety of conditions.MethodsA trained sensory team identified foods and beverages that, when mixed with a BSE, showed promise in masking the extract’s flavors. Established sensory evaluation techniques were then implemented by a group of seven trained descriptive analysis panelists to deconstruct the sensory profile of each sample (BSE suspended in a delivery vehicle). The sensory characteristics were then clustered into dimensions based upon factor analysis and principal component analysis, followed by a test-retest protocol, to match complementary flavors from liquid-based food sources that would be readily available in the cultural context of our clinical test sites.ResultsClustering of sensory attributes (dimensions) was identified and was both negatively and positively associated with the perception of glucoraphanin-rich and sulforaphane-rich BSE. Four dimensions were able to explain 73% of the sample set variability. Pineapple juice was identified as a complementary flavor that was most effective in masking broccoli complex attributes, and lime and ginger were effective in masking other “harsh” or objectionable flavor components of the BSE.ConclusionEffective beverages worked by invoking “flavor misattribution”, wherein a food (broccoli extract) with an objectionable sensory characteristic was paired with a vector in which that characteristic was an acceptable component of the vector’s flavor profile. Further development of this concept with an unlimited palate could be used to develop optimal carriers for food product development and/or to refine the approach for clinical trials based upon local taste preferences.

Figure optionsDownload as PowerPoint slide