AAA is most commonly diagnosed in male among the age of 60 to 80 (33). Currently known risk factors for AAA include male, smoking, family history, advancing age, hypertension, as well as obesity (34). While most of the risk factors are also linked to DM, previous meta-analysis and epidemiological studies identified DM diagnosis as a negative predictor for AAA (35, 36). The presence, growth, and probably rupture of AAA were found to be higher among non-diabetics comparing to diabetic patients (6, 7).

The negative association between DM and AAA was also demonstrated by pooled data analysis from large population prevalence studies and smaller studies in selected populations. Prospective studies showed significantly lower number of new AAA diagnosis in DM patients (37). When diabetic patients were stratified into advanced and uncomplicated groups based on existing DM comorbidities, it was demonstrated that advanced diabetes exerts a stronger protection against AAA without rupture, with risk of AAA decreased by near a half in the advanced DM group (6). Le et al. also revealed a decreasing risk of AAA with longer diabetes duration in addition to the independent inverse association between DM and AAA. The odds ratio of AAA was 0.50 at 3–5 years of follow-up (95% CI 0.27, 0.89), which declines further to 0.37 after 12 years of follow-up (95% CI 0.19, 0.70) (38). This outcome was compatible to the conclusion of another study, which discovered no difference in aortic diameter and AAA prevalence between those newly diagnosed type 2 diabetes and those without diabetes among 65-year-old men (39). In addition, an inverse association was established between fasting serum glucose level and aortic diameter even among non-diabetic men (38).

Rabben et al. identified 4 independent AAA risk factors: smoking, hypertension, BMI >30, and DM, with DM being an inverse factor for AAA. The growth rate of AAA was slower in DM patients comparing to normoglycemic patients (34). Astrand et al. found the aortic intima-media thickness (IMT) to be markedly greater in diabetic patients comparing to healthy controls, adjusting for age and sex. The thick aortic wall in diabetes resulted in a 20% reduction in aortic wall stress, which proposed a potential protective mechanism against AAA (40). Another multicenter randomized study enrolling patients with small AAAs (4.1-5.4 cm) revealed a remarkably lower probability of aneurysm growth > 5 mm at 36 months in diabetic participants comparing to nondiabetics (40.8% versus 85.1%). For patients who did not undergo open repair at baseline, the need for repair at 30 months was also lower for diabetics. However, this study also demonstrated a higher hazard ratio (HR) for all-cause mortality at 36 months among diabetic patients, which could be attributed to higher obesity and cardiovascular disease rate (41).

A sub-analysis study of VIVA (Viborg Vascular) Randomized Screening Trial found that the median growth rate significantly slower in subjects with DM comparing to those without (1.7 versus 2.7 mm/year). When participants were stratified by glycosylated hemoglobin (HbA1C), the aorta growth was smaller in the highest HbA1c-tertile group compared with the lowest HbA1c-tertile group, adjusting for covariates (42).

Despite advancements in surgical technologies, RAAA is still the major cause of AAA-related death, resulting in approximately ninety percent of mortality rate (33). The impact of DM on development and outcome of RAAA is controversial. Overall, there are three conflicting theories regarding the relationship between DM and RAAA: no association, negative association and positive association.

A large Danish register based matched case control study by Kristensen et al. (35) included 5395 cases with ruptured AAA (RAAA) matched 1:1 with patients undergoing elective AAA repairs by sex, age, and year of diagnosis. Outcomes of the study indicated that presence of DM did not protect against the risk of RAAA or influence overall 30-day mortality. A retrospective multivariate analysis study by Gokani et al. also found no statistically significant association between DM status and risk of aneurysm rupture, adjusting for cofactors (43).

In some other studies, DM is reported to be reversely associated with RAAA. A nationwide multicenter study in France reported a significantly lower prevalence of DM in ruptured AAA comparing to non-ruptured AAA (44). However, this study also demonstrated no significant difference between DM and non-DM patients in terms of in-hospital mortality in the ruptured AAA cohort. A meta-analysis by Takagi et al. also described significantly lower prevalence/incidence of RAAA in diabetic patients (odds ratio/hazard ratio, 0.71; 95% Confidence Interval, 0.56 to 0.89; p=0.003) (45). Theivacumar et al. investigated the relationship between DM and any aortic aneurysm rupture at a single center, and concluded that aortic aneurysm rupture and aneurysm rupture-related death are less likely to occur among diabetic patients (46).

On the other hand, epidemiological research of the National Health Fund and the Central Statistical Office in Poland in 2012 found significantly higher incidence of RAAA in diabetic population, regardless of gender stratification (47). Regrettably, data limitations inhibited the evaluation of the type or duration of diabetes, as well as the relationship of RAAA with the presence of cofounding factors such as hypertension, cigarette smoking and lipid values. Based on the paradoxical research outcomes regarding this issue, further prospective epidemiological research with standardized methodology is warranted to determine the link between DM and RAAA.

Previous studies have demonstrated significantly lower AAA incidence and rupture rate in women comparing to men (47). Intriguingly, the protective effect of diabetes was found attenuated among female patients. In a population-based data analysis, the sex-specific analysis showed no difference in AAA incidence rates between DM and non-DM cohorts for females, whereas the incidence rate was significant lower in males with type 2 DM (6). Another subgroup analysis of a register-based matched case control study also found increased risk of AAA rupture in female patients (35). Tsai et al. hypothesized that the diminished protective effect of DM in women was prompted by higher oestrogen level (6), which needs further investment based on paradoxical evidences in previous animal (48, 49) and clinical studies (50, 51).

Metformin is the first-line oral antidiabetic medicine for mild to moderate type 2 DM patients. It is also the most well-studied hypoglycemic drug in AAA. Previous studies illustrated the negative association between metformin and development of AAA. Metformin was shown to retard the formation and growth rate of AAA (56), even in normoglycemic mice (52, 57). Golledge et al. revealed that patients with diabetes prescribed metformin were associated with significantly lower AAA repair and rupture-related mortality comparing to patients with diabetes not prescribed metformin or patients with no diabetes. In contrary, diabetic patients not prescribed metformin did not show reduced incidence of AAA events comparing to patients with no diabetes (58), which raised the question that whether the protective effect of DM comes from anti-diabetic medicines. Itoga et al. found that patients with a metformin prescription had a 0.20 mm reduction in yearly AAA enlargement comparing to those without metformin prescription, adjusting for other AAA-related variables (59). While it was hypothesized that by attenuating arterial accumulation of matrix molecules, metformin could reverse the protective mechanism of diabetes and lead to increased AAA incidences, Kristensen et al. demonstrated a statistically nonsignificant protective effect of long-term metformin against ruptured AAA (RAAA) (55).

Studies mentioned above showed the promising clinical effect of metformin in limiting AAA progression. Utilization of metformin markedly reduced the maximum aortic diameter and formation of aortic aneurysm (14). Metformin activates the AMPK signal pathway, thereby downregulating the expression of proinflammatory cytokines (IL (interleukin)-1β, IL-6, MCP (monocyte chemotactic protein)-1 and TNF (tumor necrosis factor)-α), vascular growth cytokines (VEGFA, Flt-1 and CD31) and MMPs (60). The underlying mechanisms of metformin in AAA pathology eventually lead to preservation of smooth muscle cell (SMC), suppression of matrix remodeling and inhibition of inflammation pathways (61, 62). Treatment with metformin in normoglycemic mice also showed suppression of AAA formation and progression via restoration of the perivascular adipose tissue (PVAT) and vascular endothelial function (63). Some other potential biological pathways involved in the protective mechanism of metformin against AAA include: reduction of ECM volume, augmented arterial wall matrix formation through advanced glycation end products (8) and suppression of inflammation and oxidative stress (14). The potential mechanisms of metformin are illustrated in Figure 2 .

Metformin is excreted through kidney, thereby clearance of metformin decreases during acute or chronic renal function impairment. Given the risk of contrast-induced nephropathy (CIN) after endovascular aneurysm repair (EVAR) (64) and the risk of metformin-induced lactic acidosis in patients with renal insufficiency (65), perioperative administration of metformin should be cautiously monitored. The stop and reinitiating criteria of metformin as recommended by Society for Vascular Surgery (SVS) practice guideline is presented in Table 1 .

TZD hypoglycemic agents (such as pioglitazone and rosiglitazone) were found to active peroxisome proliferator-activated receptor-γ (PPARγ). In animal studies, PPARγ was proved to balance the aortic inflammatory conditions and to delay the progression and rupture of AAA (66). TZDs are PPARγ agonists that inhibit inflammatory response and ECM remodeling, thereby ameliorating AAA development and rupture in angiotensin II (Ang II)-induced mouse model (67–70).

Que et al. discovered that pioglitazone activates PPARγ and antagonizes the nuclear factor of activated T-lymphocytes (NFAT)/NF-κB, thus decreasing the protein expression of SMC phenotypic modulation markers. Downregulation of these modulation markers prevents cell proliferation, migration, and macrophage adhesion to SMCs, which ameliorates angiotensin II-induced aortic aneurysms (70). The expression of PPARγ in bone marrow mesenchymal stem cells of AAA patients was found notably upregulated after the administration of pioglitazone (71). The role of pioglitazone is also related to the reduction of macrophages infiltration in aortic wall and retroperitoneal periaortic fat. In a study involving sixteen patients with AA (> 5 cm in diameter) awaiting open surgical repair (OSR), after 2 months of pioglitazone pretreatment, the macrophages infiltration was diminished (72).

PPARγ activation by rosiglitazone was also reported to reduce the availability of inflammatory mediators, thereby influencing aneurysm formation and AT1a Ang II receptor expression (66, 67). Rosiglitazone can reduce the expression of TNF-α and MMP-9 in abdominal aortic aneurysm wall and retroperitoneal abdominal aortic fat (66). By increasing the production of collagen, rosiglitazone also leads to thickening of the aortic wall to reduce aortic dilation and late aneurysm rupture (66). Moreover, rosiglitazone also plays a potential protective role in the formation and rupture of aneurysms through the inhibitory effect on c-Jun N-terminal kinase (JNK) phosphorylation and toll-like receptor 4 (TLR4) expression at the site of lesion formation (68).

DPP-4 is a glycoprotein involved in cleavage and inactivation of a variety of substrates, including incretins such as glucagon-like peptide1 (GLP-1) (4). The expression of DPP-4 in AAA media and adventitia is positively correlated with typical aneurysmal disease processes, including numerous immune responses, ECM degradation and peptidase activity, angiogenesis and reactive oxygen species (73). DPP-4 also increases plasmin levels by binding to adenosine deaminase, which results in degradation of ECM and activation of MMPs via activation of plasminogen-2 (74). Moreover, DPP-4 and the DPP-4-like enzyme attractin were found to induce inflammation cascades that are critical to AAA development (4).

DPP-4 inhibitors (such as sitagliptin, alogliptin and linagliptin) delay GLP-1 degradation via DPP-4 activity suppression, thereby improving glucose control (11). DPP-4Is also antagonize functions of DPP-4 in previous mentioned AAA pathological pathways, which consequently reduces the formation and development of AAA (4, 74, 75). in a rat-based AAA model by Bao et al., alogliptin administrations in both low-does (1 mg/kg/d) and high-dose (3 mg/kg/d) groups were associated with significant reduction of reactive oxygen species (ROS) expression comparing to the control group on day 7. However, the effect became only significant in high-dose group on day 28. All the other observed effects such as decreased MMP level and dilation in aortic aneurysm wall were also more prominent in the high-dose group. This dose-dependent pattern could be further explored in clinical study (76). Lu et al. investigated the function of another DPP-4I sitagliptin in Ang II-infused mice. Their results demonstrated that sitagliptin may attenuate AAA formation by restraining microphage filtration, MMPs production as well as elastin destruction (77). Other DPP4-Is, teneligliptin and vildagliptin also showed similar protection mechanisms in AAA formation and development (4, 78).

GLP-1 is the main insulin-stimulating hormone that induces insulin secretion after nutrient intake. It involves in islet beta cell proliferation and survival, glucagon secretion regulation, and gastrointestinal motility control, leading to the development of effective pharmacological treatment against DM and obesity (75). The positive effects of GLP-1RAs on AAA formation is similar to DPP-4Is, as they share same targets in pathological pathways of AAA (75).

In animal-based AAA models, GLP-1RAs (such as liraglutide and lixisenatide) inhibit AAA development through ECM preservation and through antioxidant and anti-inflammatory effects (75). Lixisenatide attenuates ROS expression and oxidative DNA damage, which in turn retards the inflammatory process of macrophage filtration, pro-inflammatory cytokine release, and eventual MMPs expression (79). Lu et al. also proved that liraglutide reduced Ang II-treated ROS production in U937 human mononuclear cell lines, confirming its antioxidative effect (77). Lixisenatide was also found to decrease levels of extracellular signal-regulated kinase (ERK), which plays a crucial role in the regulation of MMP secretion, thereby ameliorates aortic dilatation (79).

SGLT-2I is a novel class of hypoglycemic agent that has been investigated beyond its anti-glycemic indication. SGLT-2Is, such as dapagliflozin and empagliflozin, were found to improve the cardiovascular outcomes in patients with heart failure. In an Ang II-induced dissecting AAA mouse model, cotreatment with empagliflozin resulted in significant reduction in maximal suprarenal aortic diameter. Immunohistochemistry study further confirmed that empagliflozin disrupted elastin degradation, neovascularization, and macrophage infiltration in the AAA formation process (80). p38 mitogen-activated protein kinase (p38 MAPK) was previously proved to promote MMP production, while nuclear factor-κB (NF-κB) was known to upregulate cytokines/chemokines in the vascular wall, leading to AAA growth. Empagliflozin blunted p38 MAPK and NF-κB phosphorylation, thereby restricting AAA growth (80). Another study discovered that dapagliflozin markedly impairs medial SMC loss and alleviated aneurysmal aortic expansion in mice treated with intra-aortic porcine pancreatic elastase (PPE) infusion (54).