Abdominal aortic aneurysm (AAA) is a life-threatening condition characterized by permanent, localized dilation of the abdominal aorta [1]. While often asymptomatic, AAA carries a high risk of rupture, with mortality rates reaching 85–90 % upon occurrence [2]. Current international guidelines recommend routine ultrasound screening based primarily on age and sex criteria [3,4]. However, these general parameters fail to capture the heterogeneous nature of AAA and limit opportunities for personalized management strategies. Although recent studies have proposed more comprehensive screening approaches incorporating broader clinical characteristics [5], the practical implementation in primary care remains challenging due to time constraints and the burden of multi-faceted risk assessment. Consequently, blood-based biomarkers represent a promising solution for improving AAA prediction and prevention, offering objective, readily measurable indicators that could be seamlessly integrated into clinical workflows [6].

Plasma proteomics has emerged as a powerful tool for characterizing dysregulated biological processes in human diseases, providing valuable candidates for biomarker discovery [7]. However, previous proteomic studies of AAA have been constrained by methodological limitations, including case-control or cross-sectional designs, small sample sizes, and restricted protein panels [[8], [9], [10]]. A recent investigation by Ho et al. similarly leveraged the UK Biobank proteomic resource in a multi-disease analytical framework, identifying several plasma proteins associated with AAA risk among other cardiovascular outcomes [11]. While this valuable study established important predictive associations and provided a broad survey of protein-disease relationships, key questions regarding AAA-specific pathogenesis and causal mechanisms remain unresolved. In particular, the transition from protein-disease associations to causal inference and therapeutic target prioritization requires more focused investigation.

Beyond prediction, the absence of effective pharmacological therapies for AAA underscores the critical need to identify and validate therapeutic targets. Mendelian randomization (MR) analysis of plasma proteins offers a powerful approach for causal inference and therapeutic targets priorization. However, previous MR studies on AAA have been limited by their inclusion of both cis- and trans-variants as instrumental variables, increasing susceptibility to horizontal pleiotropy [12,13], and by their examination of only a limited number of proteins [12,14], leaving comprehensive causal assessment of the plasma proteome in AAA largely unexplored.

To address these limitations, we performed integrated analyses of 2911 plasma proteins in relation to incident AAA using both phenotypic and genetic data from the UK Biobank. Our study aims to: (1) identify proteins prospectively associated with incident AAA as potential biomarkers; (2) elucidate the biological pathways of significant AAA-associated proteins; (3) assess the predictive performance of selected proteins compared to the clinical risk factor model for AAA; and (4) identify causal proteins for AAA through two-sample cis-MR analyses to prioritize therapeutic targets.