WCRF/AICR dietary adherence-associated metabolic bridge phenotypes and incident lung cancer risk: a prospective cohort study of the UK Biobank.

BACKGROUND

Lung cancer is the leading cause of cancer-related mortality worldwide. Although the World Cancer Research Fund/American Institute for Cancer Research (WCRF/AICR) dietary recommendations are widely endorsed, their association with lung cancer risk and related circulating metabolic phenotypes remains unclear.

METHODS

We derived a UK Biobank-adapted WCRF/AICR dietary adherence score from six components. Among 91,917 participants with complete dietary, covariate, follow-up, and metabolomic data, we constructed a diet-related plasma metabolomic signature and an FDR-restricted bridge score. Cox and Fine-Gray models examined associations with incident lung cancer, with Benjamini-Hochberg correction applied to high-dimensional analyses.

RESULTS

Over follow-up, 470 incident lung cancer cases occurred. The dietary score was not significantly associated with lung cancer risk. In contrast, the metabolomic signature was associated with lower risk in both Cox (HR per 1-SD: 0.816, 95% CI: 0.752-0.885; p < 0.001) and Fine-Gray models (SHR: 0.830, 95% CI: 0.778-0.885; p < 0.001). The bridge score showed a borderline association in single-exposure Cox models (HR: 0.907, 95% CI: 0.823-1.000; p = 0.051) but was attenuated in joint models. Three plasma fatty acid measures met dual-path FDR criteria: saturated fatty acids (higher risk), linoleic acid, and polyunsaturated fatty acids (both lower risk). Sensitivity analyses excluding early events supported robustness.

CONCLUSION

The questionnaire-based dietary score was not directly associated with lung cancer risk, whereas a plasma metabolomic signature showed a consistent inverse association. Fatty acid compositional measures may represent candidate metabolic biomarkers linking diet to lung cancer risk. These associative findings suggest that metabolic profiling may complement dietary questionnaires in future studies.