Background and aims: Inflammation has been revealed to facilitate thrombogenesis and to increase the risk of venous thromboembolism (VTE). However, limited data are available on the association between the anti-inflammatory diet and incident VTE. We conducted a cohort analysis to examine this association and to further examine whether this association is modified by smoking status, a trigger of systemic inflammation. Methods and results: We used data from two cohorts including 81,507 middle-aged and older Swedish adults without previous VTE at baseline. An empirically validated anti-inflammatory diet index (AIDI), based on 12 foods with anti-inflammatory potential and 5 foods with proinflammatory potential, was employed to estimate the anti-inflammatory potential of diet. Hazard ratios (HRs), with corresponding 95% confidence intervals (CIs), of VTE were estimated by Cox proportional hazards regression models. During a mean follow-up of 17.8-years, 5241 VTE cases were diagnosed. Compared with individuals in the lowest quartile of the AIDI (score 4), those in the highest quartile (score !8) had a 9% (95% CI, 0e17%) lower risk of VTE. The inverse association was observed in current and past smokers (HR between the two extreme quartiles, 0.80, 95% CI, 0.70e0.91) but not in never smokers (HR, 1.03, 95% CI, 0.91e1.17). French fries (HR per serving, 1.33, 95% CI, 1.06, 1.67) but no other foods included in AIDI was associated with VTE.
Conclusion:The study suggests that a consumption of foods with high anti-inflammatory potential may play a role in the prevention of VTE in smokers.