Objective To investigate the relationship between liver fibrosis score (LFS) and the incidence of cardiovascular disease (CVD) in rural Uyghur residents in Uyghur Autonomous Region (Xinjiang), and to provide a reference for the early detection and prevention of CVD.

Methods Using a typical sampling, we recruited 13 464 rural Uyghur residents aged ≥18 years in the 51st Regiment of the Third Division of the Xinjiang Production and Construction Corps and conducted a baseline survey in August-September 2016. Five follow-up surveys were conducted in July 2019, July 2020, August 2021, June 2022, and June 2023 among 11 863 non-pregnant and non-frequent migrant participants with complete information and no CVD at baseline. The Kaplan-Meier method was used to estimate the cumulative incidence of CVD in three groups with different grades of LFS, including fibrosis-4 score (FIB-4 score), body mass index (BMI)-aspartate aminotransferase/alanine aminotransferase(AST/ALT) ratio-diabetes score (BARD score) , and Forns index. The Cox proportional hazards regression model was used to assess the relationship between the three LFSs and CVD incidence. The Framingham CVD risk score model (Framingham model) was used to assess the predictive value of LFS for CVD risk.

Results At the end of the surveys, 423 (3.57%) of the 11 863 participants were lost to follow-up and 1 280 CVD events were identified in the 11 440 participants finally included in the analysis, with an incidence density of 17.62/1 000 person-years and a cumulative incidence of 11.19% for the total follow-up of 72 662.30 person-years and an average of 6.35±1.28 person-years for per participant. The cumulative incidences of CVD were 9.10%, 19.28%, and 27.71% for the low (<1.30), moderate (1.30–2.67), and high (>2.67) FIB-4 score groups; 7.19%, 16.45%, and 33.81% for the low (<2), moderate (2–3), and high (>3) BARD score groups; and 1.42%, 9.89%, and 26.88% for the low (<4.2), moderate (4.2–6.9), and high (>6.9) Forns index groups, respectively, with statistically significant differences in the cumulative CVD incidence among groups with different LFS of three types (all P<0.001). After adjustment for sex, age, smoking status, alcohol consumption, hypertension, diabetes, obesity, waist circumference, high-density lipoprotein cholesterol (HDL-C), and estimated glomerular filtration rate (eGFR), the multifactorial Cox proportional hazards regression model showed that the risk of CVD was higher in the moderate and high FIB-4 score groups (hazard risk HR=1.26, 95% confidence interval 95%CI: 1.09–1.45 and HR=1.91, 95%CI: 1.59–2.30) than those in the low score group; the risk of CVD was also higher in the moderate and high BARD score groups (HR=1.57, 95%CI: 1.24–1.97) and HR=2.47, 95%CI: 1.51–4.04) and Forns index groups (HR=3.00, 95%CI: 2.12–4.24 and HR=4.95, 95%CI: 3.44–7.11) than those in the low BARD score and Forns index groups, respectively. The concordance index (C-index) (95%CI) of the Framingham model analysis was 0.805 (0.794–0.815). The C-index (95%CI) after adding the FIB-4 score, BARD score, and Forns index was 0.809 (0.799–0.820), 0.810 (0.800–0.821), and 0.810 (0.799–0.820), respectively (all P<0.001). The net reclassification index (95%CI) after adding the BARD score and the Forns index was 0.239 (0.206–0.269) and 0.105 (0.070–0.137), respectively (all P<0.001). The integrated discrimination improvement index (95%CI) was 0.007 (0.003–0.011) and 0.005 (0.002–0.010), respectively (all P<0.001). The addition of the three LFSs may improve the predictive ability of the Framingham model for CVD risk.

Conclusions Elevated LFS levels may increase the risk of CVD in rural Uyghur residents in Xinjiang. The addition of LFS may improve the predictive ability of the Framingham model for CVD risk.