Cardiovascular, limb, and kidney effects of SGLT2 inhibitors in patients with peripheral artery disease: a systematic review and meta-analysis of randomized controlled trial data

Abstract

Background

Sodium–glucose cotransporter-2 inhibitors (SGLT2i) provide substantial cardiovascular benefits across a broad spectrum of patients at high atherosclerotic risk. However, evidence for their cardiovascular efficacy in patients with peripheral artery disease (PAD) remains limited. Accordingly, we conducted this systematic review and meta-analysis to synthesize available evidence and clarify the cardiovascular, limb, and kidney outcomes of SGLT2i in patients with PAD.

Methods

PubMed, Embase, and the Cochrane Library were searched from inception to 20 December 2025 for studies evaluating SGLT2i therapy in patients with PAD. The primary outcome was the composite of hospitalization for heart failure (HHF) or cardiovascular death. Secondary outcomes included HHF, cardiovascular death, amputation, composite renal outcomes, all-cause mortality, and major adverse cardiovascular events (MACE).

Results

Five studies comprising 7,275 patients with PAD were included. Among patients with PAD, SGLT2i therapy was associated with a lower risk of the composite of HHF or cardiovascular death (HR 0.73; 95% CI 0.64 to 0.83; p < 0.001), HHF (HR 0.63; 95% CI 0.51 to 0.77; p < 0.001), cardiovascular death (HR 0.83; 95% CI 0.69 to 1.00; p = 0.045) and adverse renal outcomes (HR 0.74; 95% CI 0.55 to 0.98; p = 0.038). SGLT2i was not associated with an increased risk of amputation (HR 1.17; 95% CI 0.87 to 1.56; p = 0.293). No significant reductions were observed for all-cause mortality (HR 0.86; 95% CI 0.69 to 1.08; p = 0.192) or MACE (HR 0.89; 95% CI 0.75to 1.06; p = 0.207). There was no evidence of effect modification based on PAD status (all P_interaction >0.10). However, in exploratory subgroup analyses among patients with PAD, SGLT2i treatment for ≥ 2 years was associated with a numerically greater reduction in cardiovascular death (HR 0.73; 95% CI 0.60–0.90), whereas no clear effect was observed in trials with a duration of < 2 years (P_interaction = 0.06).

Conclusions

SGLT2i-therapy is associated with a reduced risk of composite of cardiovascular and kidney outcomes in patients with PAD, without increasing the risk of amputation. These findings support a potential role of sustained long-term SGLT2i-therapy in this high-risk group.

Graphical abstract

Supplementary Information

The online version contains supplementary material available at 10.1186/s12933-026-03207-y.

Introduction

Peripheral artery disease (PAD) is a common manifestation of systemic atherosclerosis defined by stenosis or occlusion of arteries in the lower limbs [1]. It affects more than 200 million individuals worldwide and remains a major public health burden, particularly among patients with cardiovascular disease risk factors including advanced age, diabetes and smoking status [2].

Growing evidence indicates that PAD should be regarded not merely as a localized vascular pathology but as a powerful marker of diffuse systemic atherosclerosis [3]. Consistent with this concept, PAD is associated with a substantially higher risk of myocardial infarction, stroke, cardiovascular death (CV death), and vascular rehospitalization compared with isolated coronary or cerebrovascular disease, reflecting its frequent manifestation as a polyvascular atherosclerotic process involving multiple arterial beds [4, 5]. Collectively, these findings establish PAD as a high-risk systemic cardiovascular phenotype, underscoring an urgent need for novel and effective therapeutic strategies to reduce its substantial burden of adverse cardiovascular outcomes.

Sodium–glucose cotransporter-2 inhibitors (SGLT2i), initially developed as glucose-lowering agents for type 2 diabetes (T2DM), have emerged as a transformative class of reno- and cardioprotective agents over the past decade with extensive application to reduce the risk of hospitalization for heart failure (HHF), progression of chronic kidney disease (CKD), and CV death, even in patients without diabetes [6, 7]. Beyond their metabolic effects, SGLT2i exert pleiotropic benefits through hemodynamic, anti-inflammatory, endothelial, and metabolic mechanisms that directly target pathways involved in atherosclerosis and vascular dysfunction [8–10]. Large-scale randomized controlled trials have consistently demonstrated that SGLT2i significantly reduce the risk of heart failure (HF) hospitalization and CV death across diverse patient populations, including those with and without diabetes [11, 12]. Accordingly, the 2023 European Society of Cardiology guidelines assign a Class I, Level A recommendation for patients with heart failure with reduced and mildly reduced ejection fraction [13].

Despite major advances in guideline-directed medical therapy for atherosclerotic cardiovascular disease, patients with PAD continue to experience disproportionately high rates of adverse cardiovascular and limb outcomes. Although a strong pathophysiological rationale supports the use of SGLT2i in this population—and despite the markedly elevated cardiovascular risk associated with PAD—the potential protective effects of SGLT2i have not been comprehensively evaluated in patients with PAD. Available evidence is largely confined to subgroup analyses and secondary reports from major cardiovascular outcome trials [14–18], with no prior study systematically synthesizing these data to define the clinical impact of SGLT2i specifically in PAD in regards to cardiovascular outcome. Accordingly, we conducted a systematic review and meta-analysis to comprehensively evaluate the effects of SGLT2i on cardiovascular, limb and kidney outcomes in patients with PAD.

Methods

This systematic review and meta-analysis was conducted according to the PRISMA 2020 Statement [19]. The study protocol was registered on the PROSPERO website, registration number: CRD420251244655. In this study, two authors (C.H and M.E) independently searched databases, selected studies, extracted data, and assessed the risk of bias, while the third author resolved any disagreements that arose during the process independently (A.K).

Ethical statement

Ethical approval and informed consent were not required because this meta-analysis used data from previously published studies.

Data sources and search strategy

We systematically searched publications in PubMed, Embase and The Cochrane Library from database inception to 20 December 2025. The terms used for the search included “sodium glucose co-transporter 2 inhibitors”, the names of individual SGLT-2 inhibitor drugs, and “peripheral arterial disease”. The complete search strategy in showed in Supplementary File 1. The reports were not limited to any language. Additionally, we performed a manual search of reference lists from key articles to identify other relevant studies.

Study selection criteria

Initially, the titles and abstracts of all articles retrieved from these databases were screened, and subsequently, a second screening was conducted for the full text. Literature selection was based on the following criteria. The inclusion criteria for eligible studies were as follows: [1] randomized controlled trials (RCT) or secondary analyses of RCT that reported outcomes specifically in patients with PAD; [2] studies with reports of the composite of HHF or CV death, HHF, CV death, amputation, composite renal outcomes, all-cause mortality, or major adverse cardiovascular events (MACE). Studies were excluded if they met any of the following criteria: [1] non-randomized designs, including prospective or retrospective studies; [2] studies without PAD patients; [3] studies that did not report the prespecified outcomes.

Data extraction and quality assessment

Using a predefined data extraction form, two independent reviewers (C.H. and M.E.) extracted data on the following variables: authors, year, study design, treatment arms, number of participants in each group, follow-up time of participants, and patient’s characteristics (age, sex, study population, race, comorbidities and current therapy). The primary outcome considered was the composite of HHF or CV death, as prespecified by the individual trials. Secondary outcomes included: [1] HHF [2], CV death [3], amputation (defined as any lower extremity amputation) [4], composite renal outcomes (defined according to each trial’s prespecified definition, generally included a decline in kidney function, onset of end-stage kidney disease, and renal death) [5], all-cause mortality, or [6] MACE (defined as CV death, myocardial infarction and stroke). The exact components of each study-reported composite endpoint are shown in Supplementary File 2.

Any disagreements regarding the data extraction or risk-of-bias assessment were resolved through consensus. To evaluate the quality of studies, the version 2 of the Cochrane risk-of-bias tool (RoB 2) was applied [20]. This tool evaluates five domains: randomization process, deviations from intended interventions, missing outcome data, outcome measurement, and selection of reported results. Based on this assessment, studies were classified as having low, some concerns, or high risk of bias. Given that included studies were based on secondary analyses of RCT, we additionally appraised their credibility using the criteria proposed by Sun et al. [21]. These criteria were applied at the domain level. As the Sun’s framework does not provide a validated overall numeric score or explicit cut-offs, we report domain-level judgments rather than an overall summary score.

Statistical analysis

The primary effect size to be pooled is the risk estimate for adverse outcomes, reported as Hazard ratio (HR) and its corresponding 95% confidence intervals (CI). To evaluate whether the treatment efficacy of SGLT2i differed by PAD status, we calculated the ratio of hazard ratios (RHR) for each included study. Given expected heterogeneity in study designs and settings, we use a random effects model for all pooled analyses. The degree of heterogeneity was assessed using the I² test, with values < 30%, 30–60%, and > 60% indicating low, moderate, and high levels of heterogeneity, respectively. To explore potential sources of heterogeneity, we performed an exploratory subgroup analysis stratified by the duration of follow-up and type of SGLT2i. To evaluate the robustness of our findings, we conducted a sensitivity analysis by leave-one-out analysis. Publication bias was assessed using Begg’s funnel plot. p < 0.05 was considered to be statistically significant. The Stata 18 (StataCorp, College Station, TX, USA) software was used to conduct all statistical analysis.

Results

Study selection and characteristics

The study selection process is shown on the PRISMA flow chart (Fig. 1). A total of 1252 articles were initially identified, of which 138 remained after duplicates removal. Following title and abstract screening based on predefined selection criteria, 1101 articles were excluded. We conducted a full-text assessment of the remaining 13 studies, and 8 were excluded. Finally, 5 studies [14–18] involving a total of 7,275 patients were considered eligible and included in the meta-analysis.

Fig. 1.

PRISMA (Preferred Reporting Items for Systematic Reviews and Meta-Analyses) flow chart of the study selection process

The included articles were published between 2017 and 2023. Of the included studies, two were pre-specified subgroup analyses of RCT, while three were post-hoc analyses of pooled RCT. Regarding the specific SGLT2i used, two studies used dapagliflozin, two used empagliflozin, and one focused on canagliflozin. The study populations also varied: three studies enrolled patients with PAD and T2DM, while two targeted patients with PAD and HF across the spectrum of ejection fraction, including heart failure with reduced, mildly reduced, and preserved ejection fraction. Across all cohorts, participants had a mean age ranging from 62.0 to 72.2 years, with males accounting for the majority (66.4%–76.1%). Median follow-up duration across the studies ranged from 1.8 to 4.2 years, allowing for sufficient observation of outcomes. The study populations were predominantly White, comprising 79.6%–84.0% of participants, while representation of other racial and ethnic groups was limited. These demographic characteristics highlight the generalizability of the findings primarily to older, White male populations with the conditions studied. The average Body Mass Index was relatively high, ranging from 28.6 to 30.8 kg/m². Regarding smoking status, current smokers represented a minority (16.7% to 22.8%), whereas the proportion of current or former smokers was substantial, ranging from 57.0% to 67.8%. The baseline characteristics of the included studies are summarized in Table 1 and Supplementary File 3. According to the RoB2 assessment, two study was rated as having a low risk of bias, while the remaining three were moderate risk. The detailed item-by-item assessment for each study is presented in Fig. 2. The domain-based appraisal using the credibility criteria of Sun et al. is presented in Supplementary File 4. Across all included reports, PAD was consistently defined as a baseline characteristic, and subgroup comparisons were conducted within individual studies. In addition, the direction of treatment effects was generally consistent across studies and across related cardiovascular and renal outcomes, with no clear evidence of materially divergent effects according to PAD status. Biological rationale was also considered plausible, given the underlying pathophysiology of PAD. Taken together, these findings suggest that the available PAD evidence is supportive but should be interpreted as exploratory.

Table 1.

Study design and key characteristics of the included studies

Study, year	Study design	Trail	Number of participants		Sex, male (%)	Age, years (mean ± SD)	Intervention	Control	Inclusion of patient population	Follow-up, year (median)
			Intervention	Control
Barraclough 2022	Post hoc analysis of pooled RCT	CANVAS, CREDENCE	1707	1452	66.40%	63.8 ± 8.7	Canagliflozin 100 mg	Placebo	T2DM with PAD	2.4–2.6
Bonaca 2020	Pre-specified analysis of RCT	DECLARE-TIMI 58	522	503	68.40%	62 (Median)	Dapagliflozin 10 mg	Placebo	T2DM with PAD	4.2
Butt 2023	Post hoc analysis of pooled RCT	DAPA-HF, DELIVER	381	428	76.10%	70.7 ± 8.8	Dapagliflozin 10 mg	Placebo	HF with PAD	1.8
Verma 2017	Subgroup analysis of RCT	EMPA-REG	982	479	69%	64.0 ± 8.5	Empagliflozin 10/25 mg	Placebo	T2DM with PAD	3.1
Verma 2023	Post hoc analysis of pooled RCT	EMPEROR-Reduced and Preserved	418	403	70.5%	72.2 ± 8.3	Empagliflozin 10 mg	Placebo	HF with PAD	1.8

NR, not reported; RCT, randomized controlled trial; SD, standard deviation

Fig. 2.

Risk of bias assessment using the RoB 2 tool

Cardiovascular, limb, and kidney outcomes

The composite of HHF or CV death (I² = 0.0%), HHF (I² = 17.51%), and CV death (I² = 21.00%) showed low heterogeneity, MACE (I² = 38.88%) and amputation (I² = 41.13%) showed moderate heterogeneity, whereas all-cause mortality (I² = 60.08%) and composite renal outcomes (I² = 63.19%) showed high heterogeneity.

The composite of HHF or CV death was reported in all five studies, involving a total of 7,275 patients. Pooled analysis showed that SGLT2i was associated with a significantly lower risk of HHF (HR 0.73; 95% CI 0.64 to 0.83; p < 0.001; Fig. 3a). HHF was consistently reported across all five studies, with SGLT2i therapy associated with a significant reduction in risk (HR 0.63; 95% CI 0.51 to 0.77; p < 0.001; Fig. 3b). Similarly, based on data from all five studies, SGLT2i treatment was significantly associated with a reduced risk of CV death (HR 0.83; 95% CI 0.69 to 1.00; p = 0.045; Fig. 3c). Amputation was reported in four studies, involving a total of 6454 patients. In the pooled analysis, SGLT2i treatment was not associated with a significant difference in the risk of amputation (HR 1.17; 95% CI 0.87 to 1.56; p = 0.293; Fig. 3d). In terms of kidney protection, the composite renal outcomes was evaluated across four studies involving 6466 patients. Treatment with SGLT2i significantly reduced the risk of composite renal outcomes compared with placebo (HR 0.74; 95% CI 0.55 to 0.98; p = 0.038; Fig. 4a). All-cause mortality was assessed across four studies involving 6250 patients, and SGLT2i therapy was not associated with a significant reduction in risk (HR 0.86; 95% CI 0.69 to 1.08; p = 0.192; Fig. 4b). Similarly, MACE was reported in 4 studies covering 6454 patients. There was no significant association between SGLT2i and the risk of MACE (HR 0.89; 95% CI 0.75 to 1.06; p = 0.207; Fig. 4c). The detailed results and corresponding forest plots for these outcomes are presented in Figs. 3 and 4.

Fig. 3.

Forest plot representing the hazard ratios (HR) in each study as well as the pooled HR and 95% Confidence interval (CI). a composite of hospitalization for heart failure or cardiovascular death, b hospitalization for heart failure, c cardiovascular death, and d amputation

Fig. 4.

Forest plot representing the hazard ratios (HR) in each study as well as the pooled ORs and 95% Confidence interval (CI). a composite renal outcomes, b all-cause mortality, and c major adverse cardiovascular events

Effect modification by PAD status

We assessed whether PAD status modified the treatment effects of SGLT2i using the RHR. Although the point estimate for HHF was numerically lower in the PAD group, this difference did not reach statistical significance (RHR 0.90; 95% CI 0.69 to 1.16; p = 0.41). Across all outcomes, pooled RHRs were close to 1 with confidence intervals crossing unity, indicating consistent treatment effects irrespective of PAD status. Detailed forest plots are provided in Supplementary File 5.

Subgroup analysis

Due to the limited data for several outcomes, we conducted exploratory subgroup analysis only for the composite of HHF or CV death, HHF and CV death. In exploratory analyses by follow-up duration, the benefits of SGLT2i consistent across both subgroups for HHF and the composite of HHF or CV death. For the composite outcome, HR were 0.71 (95% CI 0.60 to 0.84) in studies with follow-up ≥ 2 years and 0.75 (95% CI 0.61 to 0.91) in those with follow-up < 2 years. Similarly, for HHF alone, HR were 0.54 (95% CI 0.40 to 0.73) and 0.70 (95% CI 0.55 to 0.88), respectively. No significant difference was observed between these two subgroups (P_subgroup = 0.72 and 0.19; Fig. 5a and b). In the subgroup analysis for CV death, a significant reduction was observed in studies with follow-up duration of ≥ 2 year (HR 0.73; 95% CI 0.60 to 0.90). In contrast, treatment showed no significant benefit in the subgroup characterized by follow-up duration < 2 years (HR 1.01; 95% CI 0.78 to 1.31). The test for subgroup differences approached but did not reach statistical significance (P_subgroup = 0.06; Fig. 5c). In exploratory analyses by SGLT2i type, no significant differences in treatment effects were observed across SGLT2i types. For the composite of HHF or CV death, the pooled HR were 0.69 (95% CI 0.56 to 0.86) for canagliflozin, 0.76 (95% CI 0.61 to 0.95) for dapagliflozin, and 0.73 (95% CI 0.58 to 0.92) for empagliflozin, with no significant subgroup difference (P_subgroup = 0.82; Fig. 6a). For HHF, the corresponding pooled HR were 0.68 (95% CI 0.49 to 0.94), 0.64 (95% CI 0.34 to 1.19), and 0.60 (95% CI 0.44 to 0.83), respectively (P_subgroup = 0.88; Fig. 6b). For CV death, the pooled HR were 0.77 (95% CI 0.59 to 1.00), 0.95 (95% CI 0.70 to 1.28), and 0.77 (95% CI 0.44 to 1.34), respectively (P_subgroup = 0.57; Fig. 6c). The detailed results of all subgroup analyses are presented in Figs. 5 and 6.

Fig. 5.

Subgroup analysis stratified by follow-up duration. a composite of hospitalization for heart failure or cardiovascular death, b hospitalization for heart failure and c cardiovascular death

Fig. 6.

Subgroup analysis stratified by SGLT2i type. a composite of hospitalization for heart failure or cardiovascular death, b hospitalization for heart failure and c cardiovascular death

Sensitivity analysis and publication bias

Sensitivity analysis was performed by removing one study at a time and evaluating the influence of the remaining pooled data. The results for efficacy of SGLT2i on the composite of HHF or CV death and HHF were highly robust. In contrast, the association between SGLT2i and CV death was more sensitive to analytical assumptions. For amputation, composite renal outcome, all-cause mortality and MACE, the pooled results were sensitive to specific trials. The sensitivity analyses are displayed in Supplementary File 6. Corresponding funnel plots are illustrated in Supplementary File 7. There was no evidence of publication bias in the composite of HHF or CV death, HHF, cardiovascular death, amputation, composite renal outcome, all-cause mortality, or MACE (all P_Begg’s > 0.10).

Discussion

In this comprehensive meta-analysis of RCT data including 7275 patients with PAD, treatment with SGLT2i was associated with a significant reduction in the composite of HHF or CV death, HHF, CV death and the composite renal outcomes. Importantly, SGLT2i therapy did not increase the risk of lower limb amputation, providing crucial reassurance regarding limb safety in this vulnerable cohort. These benefits were observed across diverse patient populations including individuals with PAD and concomitant HF or T2DM. In our interaction analysis, we found no evidence that PAD status influences the treatment effects of SGLT2i across outcomes. Overall, this indicating that the relative efficacy of SGLT2i was preserved irrespective of PAD status. SGLT2i did not demonstrate a significant reduction in all-cause mortality or MACE in patients with PAD. Nevertheless, the overall efficacy-to-safety result highlights their clinical value and underscores the need for further dedicated studies to explore potential long-term survival benefits in this high-risk population.

Across current clinical guidelines for PAD, SGLT2i are positioned as cardiometabolic risk-modifying therapy for patients with concomitant T2DM. The PAD guidelines of European Society of Cardiology issued 2024 endorse SGLT2i, alongside GLP-1R agonists, as a Class I, Level A recommendation for patients with T2DM and PAD [1]. Similarly, the Canadian Cardiovascular Society´s 2022 guideline and the joint guidelines from the American College of Cardiology and Heart Association issued in 2024 provide strong recommendation for the use of SGLT2i in this population [22, 23]. To our knowledge, this study represents the first meta-analysis specifically evaluating the effects of SGLT2i on cardiovascular, limb and kidney outcomes in patients with PAD. Our findings are highly consistent with current guideline recommendations and, importantly, provide consolidated evidence supporting the cardioprotective role of SGLT2i in this high-risk population. By synthesizing PAD-specific outcome data that had not previously been systematically summarized, our study offers an important evidence layer that complements and reinforces the rationale behind existing guidelines.

Our findings demonstrate that SGLT2i therapy was not associated with a significant increase in amputation risk among patients with PAD, a population at high risk for limb events. The 2017 CANVAS trial [24] initially reported an increased incidence of major adverse limb events (MALE), prompting extensive investigation into a potential class effect of SGLT2i in patients. However, current evidence suggests that the amputation risk in CANVAS may have been an outlier confounded by baseline imbalances, unmeasured risk factors, and a lack of rigorous early foot-care protocols. Accumulating evidence has since demonstrated that SGLT2i therapy is not associated with an elevated risk of amputations or MALE [25–27]. Consequently, citing this lack of reproducibility, the U.S. Food and Drug Administration officially removed the boxed warning regarding amputation risk for canagliflozin in 2020. Notably, because only a limited number of included studies reported MALE, we were unable to perform a quantitative synthesis. However, in the two studies that did report MALE [14, 15], SGLT2i were not associated with an increased risk (HR 0.93 and 1.09). Future trials with standardized limb-endpoint adjudication are needed to more definitively characterize limb safety and potential effect modification by baseline limb risk.

It is well established that a tight, bidirectional interplay exists between the cardiovascular system and the kidneys, recognized as the cardiorenal axis [28]. In patients with PAD, this pathological cardiorenal nexus is particularly pronounced [29]. As PAD represents an advanced manifestation of diffuse systemic atherosclerosis, these individuals are not only highly susceptible to incident heart failure, but they also face a substantially elevated risk of rapid kidney function decline or CKD [30]. Crucially, impaired kidney function further exacerbates systemic inflammation and vascular calcification, ultimately leading to severe lower extremity ischemia and an increased risk of amputation [29, 31]. Mechanistically, the renoprotective effects of SGLT2i are biologically plausible and extend beyond glucose lowering. By inhibiting proximal tubular sodium–glucose reabsorption, SGLT2i restoring tubuloglomerular feedback, promoting afferent arteriolar vasoconstriction, and thereby reducing intraglomerular hypertension and hyperfiltration [32]. In parallel, SGLT2i exert favorable systemic and intrarenal effects, including modest reductions in blood pressure and plasma volume, mitigation of renal congestion, improved tubular oxygen balance and energy efficiency, and attenuation of inflammatory and profibrotic signaling [33].

Although the overall kidney benefits of SGLT2i are clearly established, the composite renal outcome in our analysis should be interpreted with caution. This is primarily due to the heterogeneity in the definitions of the composite renal endpoint across the included trials. For example, regarding the threshold for estimated glomerular filtration rate (eGFR) decline, the DECLARE E-TIMI 58 trial [15] utilized a ≥ 40% reduction criterion, whereas the EMPEROR trials [17] applied a more stringent threshold of a ≥ 50% decline. Furthermore, the EMPA-REG trial [18] incorporated progression to macroalbuminuria—a relatively early clinical indicator—into its composite outcome. Notably, in the DECLARE-TIMI 58 trial [15], the prespecified cardiorenal composite outcome included not only eGFR decline and end-stage kidney disease, but also CV death. The inclusion of this cardiovascular component may have inflated the overall event rate of the composite endpoint and introduced potential assessment bias. Therefore, future studies specifically targeting PAD populations should adopt more standardized and kidney-specific hard renal endpoints to further validate the observed these renoprotective effects.

In our exploratory subgroup analysis, the reduction in the composite of HHF or CV death and HHF with SGLT2i remained consistent regardless of follow-up duration, underscoring the robustness of this benefit in the PAD population. In contrast, for CV death, we observed significant heterogeneity based on follow-up time: while long-term studies (≥ 2 years) showed significant protection, short-term studies (< 2 years) did not. However, this apparent temporal discrepancy may be confounded by overlapping with differences in inclusion criteria. Clinical studies with shorter follow-up predominantly enrolled patients with HF [16, 17], in whom the prevalence of T2DM ranged from 54.9% to 65.2%. In contrast, studies with longer follow-up [14, 15, 18] primarily enrolled patients with T2DM, among whom the prevalence of HF ranged from 10% to 23.3%. Therefore, rather than reflecting a pure effect of follow-up duration, this observation likely captures differences in the predominant clinical phenotype of the study populations, with varying burdens of HF and T2DM. Owing to the absence of individual patient–level data, the independent effects of follow-up duration and underlying patient characteristics could not be disentangled. Further high-quality trials are warranted to clarify whether the observed difference in CV death outcomes is driven by treatment duration, baseline disease phenotype, or an interaction between both.

Several pathophysiological mechanisms may explain the observed reduction in HHF and CV death in patients with PAD treated with SGLT2i. Beyond glucose-lowering, SGLT2i exert pleiotropic cardiovascular effects, including osmotic diuresis, reduction in preload and afterload [34], improved myocardial energetics, and attenuation of adverse ventricular remodeling [35]. In patients with PAD, who frequently exhibit diffuse atherosclerosis, endothelial dysfunction and microvascular impairment [36–38], SGLT2i may additionally improve vascular function and reduce systemic inflammation [8, 39]. The preferential improvement in heart failure outcomes indicates that hemodynamic and metabolic mechanisms, rather than direct anti-atherothrombotic effects, are likely the principal drivers of clinical benefit in this population. PAD remains a high-risk cardiovascular phenotype that is frequently underdiagnosed and commonly coexists with other cardiometabolic conditions, including T2DM, HF, and chronic kidney disease [37]. The findings of this meta-analysis have direct implications for clinical decision-making in patients with PAD. The consistent reduction in the composite of HHF or CV death and HHF across studies, irrespective of follow-up duration or patient population, indicates that this benefit is robust and broadly applicable. In contrast, the observed patterns in CV death suggest that the magnitude of protection likely influenced by baseline clinical characteristics or treatment duration, underscoring the complexity of cardiometabolic interactions in PAD patients with coexisting HF and T2DM. Taken together, these findings underscore the need to consider patient phenotypes when evaluating cardiovascular outcomes with SGLT2i. Prospective trials in PAD patients are required to clarify the interplay of T2DM, heart failure, and PAD, refine patient selection, and optimize therapy.

Authors acknowledge several limitations of the current meta-analysis. First, although all included studies were derived from RCT, the PAD-specific results were obtained from pre-specified subgroup analyses or post-hoc analyses rather than from trials primarily designed for patients with PAD. Therefore, these findings should be interpreted with caution, as such analyses possess generally inferior evidential weight compared to primary trial outcomes. Second, the relatively small number of included studies may have reduced the statistical power of subgroup analyses and contributed to greater uncertainty in the estimated effects. Third, PAD represents a heterogeneous spectrum, and differences in disease severity may influence treatment effects. Due to the lack of granular data, stratification by PAD severity was not feasible. Finally, heterogeneity in study populations, outcome definitions, and follow-up duration may have influenced effect estimates, particularly for composite renal outcomes and MACE.

Conclusion

In summary, this meta-analysis of RCT data demonstrates that SGLT2i therapy is associated with meaningful reductions in the composite of HHF or CV death, HHF, CV death, and adverse kidney outcomes among patients with PAD, without increasing the risk of amputation. No significant differences were observed in MACE or all-cause mortality. Our analyses indicate that the overall efficacy of SGLT2i remains highly consistent irrespective of PAD status. However, as all included analyses were post-hoc or pre-specified subgroup analyses rather than primary trial endpoints in PAD-specific populations, these findings should be interpreted with caution. Collectively, these findings represent the most comprehensive PAD-specific evidence to date supporting the cardiovascular and kidney benefits of SGLT2i, notably without major impact on amputation risk. This underscores the clinical value of SGLT2i in patients with PAD, a population at exceptionally high cardiovascular risk that has historically been underrepresented in major trials. Dedicated RCT enrolling patients with PAD as a primary study population are warranted to confirm these observations and to better define optimal treatment strategies in this high-risk group.

Supplementary Information

Below is the link to the electronic supplementary material.

Abbreviations

CI

Confidence intervals

CKD

Chronic kidney disease

CV death

Cardiovascular death

eGFR

Estimated glomerular filtration rate

HF

Heart failure

HHF

Hospitalization for heart failure

HR

Hazard ratio

MALE

Major adverse limb events

PAD

Peripheral artery disease

RHR

Ratio of hazard ratios

SGLT2i

Sodium–glucose cotransporter-2 inhibitors

T2DM

Type 2 diabetes

Author contributions

C.H.: Conceptualization, Methodology, Validation, Formal analysis, Investigation, Data Curation, Writing - Original Draft, Visualization, M.E.: Investigation, Data Curation, Writing - Original Draft. R.S.: Investigation, Writing - Original Draft. Q.Y.: Investigation. F.P.: Writing - Review & Editing. T.A.Z.: Writing - Review & Editing. G.T.S.: Writing - Review & Editing. P.P.: Writing - Review & Editing. A.K.: Writing - Review & Editing, Conceptualization, Supervision, Validation. B.K.P.: Writing - Review & Editing, Conceptualization, Supervision, Project administration. All authors read and approved the final manuscript.

Funding

None.

Data availability

All data generated or analysed during this study are included in this published article and its supplementary information files. The datasets analysed during the current study are available from the original published studies.

Declarations

Ethics approval and consent to participate

Not applicable.

Consent for publication

Not applicable.

Competing interests

Author F.P. serves as an Associate Editor of this journal. Author T.A.Z. reports research grants from the Austrian Science Funds and Boehringer Ingelheim; honoraria for serving on advisory boards from Bayer AG, Boehringer Ingelheim, and Eli Lilly and Company; personal fees from Alkem Lab. Ltd, AstraZeneca, Bayer AG, Boehringer Ingelheim, Eli Lilly and Company, NovoNordisk, and Sun Pharmaceutical Industries. All these funding sources and disclosures are unrelated to the present study. The remaining authors declare that they have no competing interests.