Abstract Body (Do not enter title and authors here): Background:
Air pollution is a significant risk factor for type 2 diabetes (T2D). While dietary factors may help mitigate T2D risk, traditional dietary assessments may not accurately capture individual metabolic responses to diet and its potential interactions with environmental exposures. This study aims to compare the associations of metabolomics-based versus questionnaire-based EAT-Lancet diet with T2D risk and evaluate their modification effect of air pollution-related T2D risk.

Research Questions:
Can metabolomics-based dietary assessment better capture diet-disease relationships and modify air pollution-related T2D risk compared to traditional questionnaire-based assessment?

Methods:
This study included 49,625 UK Biobank participants without baseline T2D. Air pollutant concentrations (PM2.5, PM10, PM2.5-10, NO2, and NOx) were estimated using land-use regression models. An integrated air pollution score was constructed to assess their joint exposure. Diet quality was evaluated using questionnaire-based and metabolomics-based approaches to the EAT-Lancet diet. Hazard ratio (HR) and absolute risk difference (ARD) were used to estimate the associations of diet and air pollution with incident T2D. Both additive and multiplicative interactions were examined.

Results:
Higher integrated air pollution score was associated with higher T2D risk (HR: 1.10; 95% CI: 1.03-1.18). The metabolomics-based EAT-Lancet diet showed a stronger association with incident T2D (per SD increment HR: 0.88; 95% CI: 0.83-0.93) than the questionnaire-based assessment (per SD increment HR: 0.96; 95% CI: 0.91-1.02) (PZ-test=0.047). Participants with the lowest diet scores and highest air pollution exposure had a higher T2D risk than those with the highest diet and lowest pollution (HR: 1.63; 95% CI: 1.24-2.15 and ARD: 27%; 95% CI: 20%-33%). The relative excess risk due to interaction between high air pollution and low diet quality was 0.31 (95% CI: 0.03-0.58), accounting for 31% of T2D risk. No significant interactions were observed with questionnaire-based assessment.

Conclusions:
Air pollution was positively associated with T2D risk. The metabolomics-based, but not questionnaire-based, EAT-Lancet diet was inversely associated with T2D risk and significantly modified air pollution-related T2D risk.