New information on nutrition
A lot of nutritional research data may have to be redone
A recent study published in Nature Food attracted considerable attention for suggesting that prior survey-based studies linking self-reported diets to health outcomes may be fundamentally flawed. The main issue lies in significant inaccuracies in how study participants report their food intake. By applying a validated equation to two large datasets—the National Diet and Nutrition Survey and the National Health and Nutrition Examination Survey—the researchers found that the level of misreporting exceeded 50%. As misreporting increased, the macronutrient composition derived from these dietary reports became systematically biased, potentially leading to spurious associations between specific diet components and body mass index.¹
A recent Science commentary also examined the implications of this study for the field of nutrition.² One expert cited in the article remarked, “we got to try and use new technologies to do better.” In other words, a significant portion of what we currently believe about the relationship between specific foods and diseases is likely inaccurate. This study helps explain why “diet data” have often produced contradictory findings.
A good example may be a recent article in the European Heart Journal, which concluded that drinking coffee in the morning may be more strongly associated with a reduced risk of mortality than drinking coffee later in the day.³ While the general finding by these Chinese researchers—that coffee consumption is associated with lower mortality—is interesting, we would caution against placing too much weight on this study, especially in light of the aforementioned concerns about dietary data accuracy.
References
1. Bajunaid et al., Nature Food 2025;6:58-71
2. Offord C. Science 2025;387:6732:352
3. Wang et al., Europ Heart J. 2025