Understanding how diet affects health depends on a deceptively simple question: what are people actually eating? A new review suggests that answering that question reliably remains one of the biggest challenges in nutrition science and that no single tool can do the job on its own.
The review, published in Nature Food, brings together evidence from nutrition science, metabolomics, microbiome research and digital technologies to examine how dietary intake is currently measured and where those methods fall short. The authors argue that combining multiple approaches, rather than relying on any one method, may offer the clearest picture of real-world eating habits.
Most large nutrition studies still rely on self-reported data, such as food frequency questionnaires, diet recalls or food diaries. These tools are practical and inexpensive, but they depend heavily on memory, estimation and honesty. Research has consistently shown that people tend to misremember what they ate, underreport certain foods and struggle to estimate portion sizes, especially over long periods of time.
“These limitations make it difficult to reliably link diet with health outcomes,” said Dr. Thomas Wilson, a co-author of the review and a researcher at Aberystwyth University.
Inaccurate intake data can also make it harder for policymakers to assess how diets are changing in response to health and sustainability pressures.
To address these gaps, the review highlights emerging tools designed to reduce reliance on memory alone. These include wearable cameras that capture images of meals in real time and use computer vision and artificial intelligence to identify foods and estimate portions, as well as smartphone apps that prompt users to log foods closer to when they are consumed.
The authors also point to biomarkers of food intake, or BFIs, as a promising advance. These biomarkers are chemicals measured in blood, urine or stool that correspond to specific foods or dietary patterns. Unlike self-reports, biomarkers offer objective evidence that certain foods were consumed, though they cannot capture everything a person eats and often apply only to specific foods or nutrients.
Importantly, the review stresses that no single technology solves all measurement problems. Wearable cameras raise privacy and data management questions. Biomarkers are limited in scope and availability. Digital tools may reduce reporting errors but still depend on user engagement. Each method has strengths and weaknesses that vary by study design and population.
Rather than promoting one solution, the authors propose a flexible, integrated framework that combines self-reports, digital tools and biological measures. Such an approach could be adapted to different research settings, from tightly controlled clinical trials to large population studies.
Better dietary measurement, the authors argue, is not just a technical issue. It underpins efforts to develop more precise nutrition recommendations, understand links between diet and chronic disease and design policies that support both human and planetary health. Without more accurate intake data, even well-designed studies can produce uncertain or conflicting results.
The review also helps explain why nutrition research can feel inconsistent to the public. When diet is measured imprecisely, small errors can ripple through analyses and shape conclusions. Improving how diets are measured may not eliminate disagreement, but it could make findings clearer and more comparable.
As interest grows in personalized nutrition and data-driven dietary advice, the authors caution that better tools should complement, not replace, thoughtful study design and interpretation. Measuring what people eat more accurately is a necessary step, but it does not remove the need for careful analysis, transparency and humility about what nutrition science can and cannot yet answer.
Research contributing to this review was supported in part by the Novo Nordisk Foundation, the National Institute of Diabetes and Digestive and Kidney Diseases of the National Institutes of Health, and the Austrian Science Fund. Additional support included a British Nutrition Foundation early-career award and facilitation by the NuGO Office.