An international team led by Prof. John Speakman from the Shenzhen Institutes of Advanced Technology of the Chinese Academy of Sciences has developed a groundbreaking predictive model for total energy expenditure. This model combines classical statistics and machine learning to provide a more objective way to assess the accuracy of food intake records.
The study, published in Nature Food on January 13, addresses the challenges faced by nutritional epidemiology in linking dietary exposures to chronic diseases. Traditionally, evaluating dietary intake relied heavily on tools such as food frequency questionnaires, 24-hour recall interviews, and food diaries. However, these methods are prone to inaccuracies as individuals may forget or even falsify their reports, leading to dietary misreporting.
To overcome this limitation, researchers utilized the doubly-labeled water technique, an isotope-based method that directly measures an individual’s energy needs. By collecting over 6,000 measurements and employing both classical statistics and machine-learning approaches, they developed a predictive model that was validated in an additional 600 subjects. This model is currently the most accurate method for estimating energy requirements without the need for direct measurements.
Applying this model to large surveys of food intake data in the United States and the United Kingdom, researchers found that a significant portion of the records exhibited unrealistically low levels of energy intake. This suggests that current dietary instruments may be producing flawed data, requiring a reevaluation of widely held beliefs in nutrition science.
Prof. John Speakman emphasized the importance of acknowledging and addressing the flaws in existing dietary assessment methods. While it may be challenging to discard large amounts of data, continuing to publish erroneous information will only hinder progress in the field of nutrition science. Moving forward, revising outdated beliefs based on problematic methods will be essential for advancing our understanding of dietary intake and its impact on health.
For more information on this groundbreaking study, readers can refer to the research article published in Nature Food. The study, titled “Predictive equation derived from 6,497 doubly labelled water measurements enables the detection of erroneous self-reported energy intake,” provides detailed insights into the development and validation of the predictive model.
This research was conducted by a team of scientists from the Chinese Academy of Sciences and represents a significant contribution to the field of nutritional epidemiology. By challenging conventional dietary assessment methods, this study paves the way for more accurate and reliable approaches to evaluating energy expenditure and food intake.
