Importance: Anorexia nervosa (AN) is commonly associated with cardiovascular complications.
Objective: To investigate the trajectories of the risk of cardiovascular conditions in a nationwide cohort of patients with AN in Taiwan.
Design, setting, and participants: From a population-based health insurance database from January 1, 2011, to December 31, 2021, this longitudinal cohort study identified patients with AN and controls through propensity score matching at a 1:10 ratio according to sex, age, urbanization level of residence, socioeconomic status, and year of diagnosis. Data were analyzed from June 27, 2023, to February 23, 2024.
Exposure: First-time diagnosis of AN by psychiatrists during the study period.
Main outcomes and measures: Incidence and risk of composite cardiovascular conditions. Kaplan-Meier curves were used to estimate the cumulative incidence of major adverse cardiovascular events (MACE) and any cardiovascular condition. With adjustment for psychiatric comorbidities, conditional Cox proportional hazards regression analyses were performed to estimate the risk of cardiovascular events, which were presented as hazard ratios (HRs) and 95% CIs, relative to the comparison group. Risks of individual cardiovascular conditions were calculated during 3 follow-up periods after AN diagnosis.
Results: The study population included 2081 patients with AN and 20 810 matched controls, for a total of 22 891 participants (mean [SD] age, 24.9 [9.9] years; 91.3% female). In total, 99 patients with AN (4.8%) had MACE vs 175 (0.8%) in controls, and 124 patients with AN (6.0%) had any cardiovascular condition vs 483 controls (2.3%). At the 5-year follow-up, the cumulative incidence rate of MACE was 4.82% (95% CI, 3.85%-6.02%) and of any cardiovascular condition was 6.19% (95% CI, 5.19%-7.53%). Compared with the control group, the AN group had significantly higher risks of MACE (adjusted HR [AHR], 3.78; 95% CI, 2.83-5.05) and any cardiovascular condition (AHR, 1.93; 95% CI, 1.54-2.41). The significantly increased risks of congestive heart failure, conduction disorder, and structural heart disease occurred in the initial follow-up period and disappeared after 60 months of follow-up. Notably, patients with AN did not have an increased risk of ischemic heart disease until after 60 months of follow-up (AHR, 3.01; 95% CI, 1.48-6.13).
Conclusions and relevance: In this national matched cohort study, increased risk of cardiovascular conditions was found in different periods after AN diagnosis. Clinicians should monitor comorbid cardiovascular conditions among patients with AN at initial presentation, during treatment, and at follow-up.