Contrasting Associations of Prudent and Western Dietary Patterns with Risk of Developing Venous Thromboembolism

Background: Published studies are inconsistent about whether differences in diet are associated with risk of venous thromboembolism. We studied the association between dietary patterns and incident venous thromboembolism in a large US cohort. Methods: The Atherosclerosis Risk in Communities study followed 14,818 middle-aged males and females for incident venous thromboembolism over an average of 22 years between 1987 and 2015. Trained interviewers assessed dietary intake at visits 1 and 3, using a food frequency questionnaire. We derived 2 dietary pattern scores using principal component analysis and ascertained and verified hospitalized venous thromboembolism. In separate proportional hazards regression analyses, we examined associations of quintiles of the prudent and the Western dietary pattern scores with risk of developing non-cancer-related and total venous thromboembolism, adjusting for demographic characteristics, lifestyle factors, body mass index, and diabetes. Results: With 860 total incident venous thromboembolism events, the hazard ratios (95% confidence intervals) of incident non-cancer-related venous thromboembolism (n = 631) across quintiles of the prudent dietary pattern score were 1 (reference), 1.04 (0.81-1.32), 0.84 (0.65-1.08), 0.70 (0.53-0.91), and 0.88 (0.67-1.15), P_trend =.04. Across quintiles of the Western dietary pattern score, hazard ratios of non-cancer-related venous thromboembolism were 1 (reference), 1.13 (0.87-1.45), 1.20 (0.92-1.56), 1.03 (0.77-1.39), and 1.58 (1.13-2.21), P_trend =.04. Associations were similar for total venous thromboembolism. Conclusions: In this community-based cohort, a prudent dietary pattern was associated with a lower risk of future venous thromboembolism, whereas a Western dietary pattern was associated with a higher risk.