Non-Alcoholic Fatty Liver Disease Defined by Fatty Liver Index and Incidence of Heart Failure in the Korean Population: A Nationwide Cohort Study

Abstract

Fatty liver index (FLI) is a simple and useful index that evaluates non-alcoholic fatty liver disease (NAFLD), particularly in large epidemiologic studies. Heart failure (HF) is becoming a burden to public health as the global trend toward an aging society continues. Thus, we investigated the effect of FLI on the incidence of HF using large cohort data from the Korean National Health Insurance health database. Methods and Results: A total of 7,958,538 subjects aged over 19 years without baseline HF (men = 4,142,264 and women = 3,816,274) were included. Anthropometric and biochemical measurements were evaluated. FLI scores were calculated and FLI ≥ 60 was considered as having NAFLD. Hazard ratios (HRs) and 95% confidence intervals (CIs) for HF incidence were analysed using multivariable time-dependent Cox proportional hazard models. During a mean follow up of 8.26 years, 17,104 participants developed HF. The FLI components associated with the incidence of HF and FLI showed a causal relationship with HF; the FLI ≥ 60 group had a higher HR for HF (HR 1.493; 95% CIs 1.41−1.581) than the FLI < 30 group. Subgroup analysis showed that fatty liver (FLI ≥ 60) with age ≥ 65 years or women displayed higher HR for HF than fatty liver with age < 65 or men, respectively. An increase in FLI score significantly increased the HR for HF except for those with a FLI score change from <30 to 30−60. Conclusion: NAFLD defined by FLI and increase in FLI score were associated with the incidence of HF. Further detailed prospective studies are needed.

Keywords: fatty liver index; heart failure; non-alcoholic fatty liver disease; obesity; waist circumference.

Figure 1

The incidence rate (per 1000 persons) and hazard ratio for heart failure by the components of the fatty liver index. * Statistically not significant. Adjusted for age, gender, smoking, drinking, exercise, diabetes mellitus, hypertension, dyslipidaemia, and BMI. BMI, body mass index; WC, waist circumference; TG, triglyceride; GGT, gamma-glutamyl transferase; Q, quintiles (Q1, Q2, Q3, Q4, and Q5). (a) Incidence rate for heart failure per 1000 persons, (b) Hazard ratio for the incidence of heart failure.

Figure 2

Change in fatty liver index and its hazard ratio for the incidence of heart failure during the four-year follow-up. * Statistically not significant. Adjusted for age, gender, smoking, drinking, exercise, diabetes mellitus, hypertension, dyslipidaemia, and body mass index.