Background: Previous studies suggested that aspirin inhibits liver fibrogenesis and reduces hepatic fat, potentially benefiting patients with metabolic dysfunction-associated steatotic liver disease (MASLD). However, observational studies face challenges in selecting an appropriate active comparator for a cohort study to assess the effectiveness of aspirin initiation in this population.

Objectives: To evaluate the association between the use of aspirin and risk of hepatic events in patients with MASLD.

Methods: We used health screening records from a tertiary hospital linked to healthcare claims in Korea (2013-2022) to construct a cohort of MASLD patients with no prescription of aspirin a year prior to cohort entry date (CED). The MASLD status was defined based on the presence of fatty liver on abdominal ultrasonography at the date of first health screening (CED), accompanied by an elevated HbA1c level (≥5.7%) and a fibrosis-4 index (≥1.30). Using the clone-censor-weight approach, eligible patients were replicated into two clones and assigned to either group of treatment (aspirin initiation within 180-day grace period) or control (no initiation during the 5-year follow-up). Individuals were censored if they deviate from the assigned group during follow-up. Time-varying inverse probability of censoring weights were estimated using a pooled logistic regression model. The primary outcome was a composite of hepatic decompensation events, including ascites, esophageal varices with bleeding, hepatic failure, liver transplant, or hepatocellular carcinoma. We calculated 5-year absolute risk, risk difference (RD), and risk ratio (RR) using survival probabilities derived from the cumulative baseline hazard function of weighted Cox models. We performed 300-sample non-parametric bootstrapping to obtain 95% confidence intervals (CIs).

Results: Among a total of 19,207 participants (mean [SD] age, 54.1 [10.3] years; 13,985 males [72.8%]) replicated into two clones, 1,820 (9.5%) adhered to aspirin treatment group and 17,705 (92.2%) adhered to control group. The number of hepatic events was 39 (0.20%) in the treatment group and 181 (0.94%) in the control group, corresponding to a 5-year absolute risk of hepatic events of 0.26% (95% CI, 0.14% to 0.90%) in the treatment group, and 1.10% (0.92% to 1.31%) in the control group. The use of aspirin was associated with 76% lower risk of hepatic events (RR 0.24, 95% CI 0.12-0.78) compared to control, though the absolute RD (-0.84%, -1.07% to -0.22%) was small.

Conclusions: Consistent with its proposed biological mechanism, the use of aspirin was associated with a lower risk of hepatic events in patients with MASLD. Further studies are warranted to validate the robustness of the clone-censor-weight approach in this context.