Development of Tools for the Evaluation of Risk of Bias in Nutrition Studies

Abstract

Dietary exposure assessment is an essential concern for evaluating risk of bias in human nutrition studies, yet existing tools do not consistently incorporate nutrition specific criteria. While the principles and methodologies of evidence-based medicine globally apply to nutrition topics, diet-related challenges in nutrition studies necessitate special consideration. These include the intricacy in obtaining comprehensive and accurate estimates of exposure, the potential for complex confounding and interactive effects of food substances, and the impact of individual baseline nutritional status on the relationship between exposure to a particular food substance and any health outcome. This thesis comprises three main sections: 1) To address the limitations of existing bias assessment tools for dietary exposure assessment, an eight step process was used to develop and test a set of valid and reliable nutrition-specific risk of bias tools for randomized controlled trials, cohort studies and case-control studies, including detailed guidance and a worksheet; 2) As a continuation of this research, foundational work for the development of a risk of bias tool in cross-sectional studies, a scoping review of tools, methodology and bias concepts was conducted to inform the future development of nutrition-specific tools for cross-sectional studies; and, 3) To apply the findings of the scoping review, the rationale and scientific approach to be used for the development, testing and establishment of valid, reliable and nutrition-specific tools applicable to descriptive and analytic cross-sectional research designs is described. The intent is to assemble a suite of risk of bias tools referred to as the NUtrition QUality Evaluation Strengthening Tools (NUQUEST) with integrated consideration for dietary exposure assessment. The availability of these new tools will foster consistent and transparent assessment of potential biases commonly encountered in human nutrition studies. Future research will expand NUQUEST to develop tools applicable to other study designs and extend testing to inform systematic refinement if needed.