A food-frequency questionnaire usually asks people how often they eat a list of foods, sometimes pairing those answers with standard portion assumptions. The result can be useful, but it is easy to ask more of it than it can deliver.

In a 2026 Perspective in the European Journal of Clinical Nutrition, researchers Jimmy Chun Yu Louie and Jahar Bhowmik argue that these questionnaires work best as ranking tools. They can help separate people who tend to eat more of a food from those who tend to eat less. They are less dependable when a study treats the answers as exact measurements of one person's intake.

That distinction matters when reading research headlines. A study may find a meaningful pattern across groups without establishing a precise intake threshold for an individual. A larger sample can make an estimate more statistically stable, but it does not automatically remove systematic errors in memory, portion estimates, or the questionnaire's food list.

The useful question is whether the claim matches the measuring tool. Broad comparisons between groups may fit an FFQ well. Precise cutoffs, small differences, or changes within one person deserve more caution, especially when the questionnaire has not been calibrated against a stronger reference method. Calibration can reduce bias, but it cannot erase every source of error.

This paper is a narrative Perspective, not a new validation experiment. Its proposed reporting checklist is a prototype rather than a formal consensus standard. Other dietary assessment methods, including recalls and weighed records, also have limitations. The takeaway is not that questionnaire research is worthless. It is that the confidence and precision of the conclusion should stay in step with what the instrument can actually measure.