Abstract
BACKGROUND: For the relationship between obesity and venous thromboembolism (VTE), nonalcoholic fatty liver disease (recently termed metabolic dysfunction-associated steatotic liver disease) is of interest given the hepatic role in hemostasis. OBJECTIVES: We aimed to assess the association between the fatty liver index (FLI), as a proxy for nonalcoholic fatty liver disease, and VTE risk in a population-based cohort. METHODS: Data from the Tromsø 4 (1994-1995) and 6 (2007-2008) surveys were used to calculate the FLI in 9870 participants. All VTEs were recorded up to December 31, 2020. We used Cox regression to estimate hazard ratios for VTE with 95% CIs by FLI groups defined according to clinical cut-offs (<30, 30-59, and ≥60). Because waist circumference and body mass index (BMI) are main determinants for FLI calculation, we assessed the potential contribution of FLI to VTE risk beyond these body fat measures. RESULTS: During a median follow-up of 13.1 years, 507 incident VTEs occurred. Compared with the reference group (FLI < 30), the hazard ratios for VTE were 1.5 (95% CI, 1.1-1.9) and 1.8 (95% CI, 1.4-2.3) for the FLI 30-59 and ≥60 groups, respectively, in models adjusted for age, sex, alcohol intake, educational level, and physical activity. The association of FLI with VTE was no longer observed, with risk estimates close to unity, when participants were stratified by clinical categories of waist circumference and BMI. CONCLUSION: Higher values of the FLI were associated with a higher VTE risk. This association was explained by waist circumference and BMI, which reflect excessive body fat deposition and are determinants of the FLI.