Digital and Technology-Enabled Approaches in Dietary Assessment: Addressing Bias, Error, and Feasibility in...

Dietary intake data are essential for understanding diet-disease relationships, informing policy, and evaluating nutrition interventions. This is particularly challenging in population- and community-based research, where varying dietary patterns, motivation to participate, and practical constraints limit the feasibility of highly controlled dietary assessment methods. Advances in digital tools, including web- and smartphone-based assessment, image- and speech-based methods, wearable sensors, and artificial intelligence are transforming how dietary data are collected, processed, and analyzed. This narrative review examines how digital and technology-enabled approaches impact key sources of bias and error throughout the stages of dietary assessment, from sample selection and participant reporting to food classification and nutrient composition assignment. We also assess the feasibility of implementing these approaches in large-scale research settings by considering resource requirements, respondent burden, and scalability. Evidence supporting these technologies varies considerably across applications, populations, and settings, and relatively few approaches have been evaluated in large-scale or low-resource research contexts. Although many technologies can reduce specific sources of bias and measurement error, they may also introduce new challenges, such as selection bias and higher development or implementation costs. These trade-offs are especially pronounced in large and heterogeneous study populations. Constraints related to connectivity, device access, and technical capacity may further limit implementation in low-resource settings. While technological innovations offer significant opportunities to improve dietary assessment, selecting the appropriate assessment method requires careful consideration of the trade-offs between accuracy and feasibility in the intended research context.