Abstract:Abstract Background Observational studies have suggested a link between panic disorder (PD) and Alzheimer disease (AD). This study aimed to identify the underlying association of PD with the risk of AD using Mendelian randomization. Methods Genetic instrumental variables (IVs) were retrieved in the genome-wide association study between PD and AD. Then, five different models, namely inverse variance weighting (IVW), weighted median, weighted mode, MR-Egger and MR-robust adjusted profile scores (MR-RAPS), were used for MR Analysis. Finally, the heterogeneity and pleiotropy of identified IVs were verified by multiple sensitivity tests. Results The Cochran’s Q test based on MR Egger and IVW showed that no evidence of heterogeneity was found in the effects of instrumental variables, so a fixed-effect model was used. IVW analysis (OR 1.000479, 95% CI [1.000147056, 1.000811539], p = 0.005) indicated that PD was associated with an increased risk of AD, and a causal association existed between them. Meanwhile, weighted median (OR 1.000513373, 95% CI [1.000052145, 1.000974814], p = 0.029) and MR-RAPS (OR 1.000510118, 95% CI [1.000148046, 1.00087232], p = 0.006) also showed the similar findings. In addition, extensive sensitivity analyses confirmed the robustness and accuracy of these results. Conclusion This investigation provides evidence of a potential causal relationship between PD and the increased risk of AD. Based on our MR results, when diagnosing and treating patients with PD, clinicians should pay more attention to their AD-related symptoms to choose therapeutic measures or minimize comorbidities. Furthermore, the development of drugs that improve both PD and AD may better treat patients with these comorbidities.

A causal relationship between panic disorder and risk of alzheimer disease: a two-sample mendelian randomization analysis