Eur J Epidemiol. 2026 Jul 23. doi: 10.1007/s10654-026-01445-8. Online ahead of print.
ABSTRACT
Cardiovascular diseases (CVD) remain a leading cause of death, with diet as a key modifiable risk factor. While prospective cohort studies using food frequency questionnaires (FFQs) underpin dietary guidelines, FFQs are prone to measurement errors that typically attenuate true associations. Characterizing the degree of error through validation is essential to properly interpret and adjust observed diet-disease relationships. It remains unclear whether FFQ validation evidence adequately supports the specific dietary exposures analyzed in recent CVD research. Therefore, this meta-research study aimed to evaluate this gap by examining discrepancies between validated dietary factors and those used in association analyses. A systematic search of multiple databases identified cohort studies published between 2019 and 2024 using keywords related to cohort studies, CVDs, and FFQs (PROSPERO: CRD42024589050). The proportion of cohorts with discrepancies was examined, defined as cases in which the dietary variables examined in association analyses lacked direct validation in the referenced FFQ validation paper. A total of 298 cohort studies from over 50 countries were included, with 88.3% using validated FFQs (against reference methods such as 24-hour dietary recalls, dietary records or biomarkers) for dietary assessment. A total of 93 unique validation papers were referenced across all included studies. However, at the acceptable correlation threshold (r > 0.3), discrepancies were identified in 80.6% of studies, indicating that the dietary variables examined in association analyses lacked direct validation in the referenced FFQ validation paper. Correlation analysis was the most commonly used validation metric (92.5% of validation studies), with 41.9% relying solely on correlation analyses. This research highlighted a significant FFQ validation gap in diet-CVD association research. Future research should clearly report FFQ validation metrics, especially de-attenuated correlation coefficients, and ensure that the dietary exposures of interest are covered in the referenced validation study. For dietary pattern studies, whenever feasible, researchers should directly validate dietary pattern scores against reference methods, so that these assessments are supported by direct evidence of validity. PROSPERO registry number: CRD42024589050.
PMID:42489826 | DOI:10.1007/s10654-026-01445-8