Prognostic value of controlling nutritional status score (CONUT) and prognostic nutritional index (PNI) in patients with peripheral artery disease: an evidence-based analysis

Abstract

Objectives

To evaluate the controlling nutritional status score (CONUT) and the prognostic nutritional index (PNI) prognostic value in peripheral artery disease (PAD) patients through a meta-analysis.

Methods

We searched PubMed, Embase, Web of Science, and Cochrane databases up to November 2024. Mortality, major adverse cardiovascular events (MACE), amputation, and poor ulcer healing were extracted. Data were synthesized using odds ratios (OR) with 95% confidence intervals (CI). Sensitivity analysis assessed result stability and heterogeneity sources. All analyses were performed using Review Manager 5.4 and STATA 15.1.

Results

Fifteen cohort studies with 6,830 PAD patients were included. Higher CONUT scores were linked to increased mortality (12 studies, HR: 1.34; 95% CI: 1.21–1.49) and amputation risk (6 studies, OR: 1.20; 95% CI: 1.10–1.32), while higher PNI was associated with reduced mortality (3 studies, HR: 0.95; 95% CI: 0.92–0.97) and amputation risk (2 studies, OR: 0.91; 95% CI: 0.87–0.95). No significant link was found between CONUT and MACE or poor ulcer healing. Sensitivity analysis revealed instability in the association between CONUT and poor ulcer healing.

Conclusions

CONUT and PNI could predict mortality and amputation risk in PAD patients, aiding early identification of high-risk individuals. Due to the study's retrospective design and potential instability, further large-scale, multicenter prospective studies are needed to confirm these findings.

Supplementary Information

The online version contains supplementary material available at https://doi.org/10.1186/s12872-026-06139-3.

Introduction

Peripheral artery disease (PAD) affects over 200 million people worldwide, reducing functional capacity and quality of life [1]. Arteriosclerosis obliterans (ASO), a common form of PAD, is marked by chronic stenosis or occlusion of the lower limb arteries, from the aorta to the iliac arteries, impairing arterial perfusion to the limbs and organs [2]. In 2015, 236 million individuals aged 25 and older worldwide were affected by lower limb ASO [1]. While most are asymptomatic, about two-thirds progress over five years, with intermittent claudication as the most common symptom [3]. Twenty-one percent develop severe lower limb ischemia (CLI), resulting in ulcers and gangrene [4].

ASO is linked to cardiovascular events, including myocardial infarction and stroke [2]. In CLI patients, the 1-year mortality and amputation rate is around 20% [4]. Poor prognosis in ASO is associated with factors like age, severity of lower limb ischemia, and complications. Secondary factors, including ulcers, infections, and malnutrition, exacerbate the condition. Chronic pain, non-healing wounds, and complications reduce activity, leading to prolonged bed rest, poor nutrition, and deficiencies in energy, protein, and trace elements. Ischemic ulcers can cause daily protein loss of up to 100 g [5]. Nutritional issues are often masked by overweight and obesity [6]. Studies show 78% of hospitalized vascular surgery patients are malnourished [7]. Early detection and intervention are crucial.

Clear and simple diagnostic criteria for malnutrition with high specificity and sensitivity are still lacking [8, 9]. Disease-related malnutrition is a complex syndrome resulting from insufficient nutritional intake, which fails to meet the patient’s physiological needs, and systemic inflammation associated with the disease [10, 11]. Malnutrition risk scoring tools, including the Subjective Global Assessment (SGA), Nutritional Risk Screening 2002 (NRS2002), Prognostic Nutritional Index (PNI), and Controlling Nutritional Status Score (CONUT), are commonly used in clinics for screening. Subjective nutritional scoring is time-consuming and may be influenced by the investigator's expertise and the patient's cultural background. In contrast, low-cost, simple, objective scores based on blood biochemical markers are more practical. Studies show that these objective scores correlate with adverse outcomes in fatal diseases such as chronic heart failure, end-stage renal disease, and cancer [12].

Numerous tools exist for assessing nutritional risk, among which the PNI and CONUT scores are two objective scores based on blood biochemical indicators. The PNI, calculated from serum albumin levels and peripheral blood lymphocyte counts, reflects nutritional and immune status [13]. The CONUT score integrates serum albumin, total cholesterol levels, and lymphocyte counts, aiming to assess nutritional reserves and control status [14]. Clinically, both lower PNI and higher CONUT values ​​suggest poorer nutritional status or higher nutritional risk and are associated with poor prognosis in various chronic diseases [15]. However, their comprehensive prognostic value in patients with PAD has not been systematically reviewed.

Recent studies show that CONUT and PNI are linked to adverse outcomes, including all-cause mortality and amputation risk in PAD patients [13, 16–19]. However, variations in inclusion criteria, sample sizes, and cutoff values prevent consensus on their prognostic value, and no evidence-based study has combined the existing data. This meta-analysis is the first to evaluate the prognostic value of CONUT and PNI, easily detectable nutritional markers, in PAD patients. By reviewing literature and analyzing recent clinical data, it examines the role of CONUT and PNI in predicting PAD prognosis and provides evidence for developing a malnutrition-related prognostic model.

Methods

Literature search

This study adhered to the PRISMA 2020 guidelines [20] and was registered with PROSPERO (CRD42024624127). PubMed, Embase, Web of Science, and Cochrane were conducted up to November 2024, focusing on studies assessing the CONUT and PNI prognostic value in PAD patients. Literature through the following terms was searched: “Peripheral Arterial Disease”, “controlling nutritional status score”, “prognostic nutritional index”, “CONUT” and “PNI”, etc. The search strategies details in Pubmed are as follows: (("Peripheral Arterial Disease"[Mesh]) OR (((((peripheral artery disease) OR (Arteriosclerotic occlusion)) OR (arteriosclerosis obliterans)) OR (ASO)) OR (intermittent claudication))) AND ((((controlling nutritional status score) OR (CONUT)) OR (prognostic nutritional index)) OR (PNI)). We manually screened the reference lists. Two authors independently retrieved and assessed eligible articles, resolving discrepancies through discussion. The full search strategy was depicted in Table S1.

Inclusion and exclusion criteria

Inclusion criteria:

• P: Patients diagnosed with PAD.

• E: High CONUT or PNI.

• C: Low CONUT or PNI.

• O: mortality, major adverse cardiovascular events (MACE), amputation, or poor ulcer healing etc.

• S: Study design was cohort or case–control.

Study protocols, unpublished studies, non-original studies (letters, comments, abstracts, corrections, replies), and studies with insufficient data were excluded.

Data abstraction

Data abstraction was performed independently by two authors, with discrepancies resolved by another author. First author, publication year, study country, design, population, marker types, sample size, age, cut-off values, mortality, MACE, amputation, and poor ulcer healing were extracted. Corresponding authors were contacted for insufficient data. During data extraction, to obtain more robust effect estimates and control for potential confounding bias as much as possible, we prioritized extracting and including data from studies that reported multivariate-adjusted HR/OR and their 95% CI. To ensure the comparability of effect sizes across different studies, all pooled analyses were based on the logarithmic transformations of the extracted HR or OR. To clarify the control of the pooled effect sizes for confounding factors, we systematically extracted the covariates corrected for by the multivariate models in the original studies. In data extraction and analysis, this study fully respected and adopted the original authors' dichotomy definition of "high" and "low" CONUT/PNI groups. Regardless of the specific cutoff value used by each original study, as long as they explicitly reported the comparison results based on that cutoff value, we extracted and merged the data according to their definition.

Quality evaluation

Cohort study quality was assessed using the Newcastle–Ottawa Scale (NOS) [21], with studies scoring 7–9 points classified as high quality [22]. Two authors independently evaluated the studies, resolving disagreements through discussion.

Statistical analysis

Meta-analysis was performed with Review Manager 5.4.1, calculating odds ratios (OR)/hazard ratio (HR) and 95% confidence intervals (CIs) for data synthesis. Heterogeneity was assessed using the chi-squared (χ²) test (Cochran’s Q) and inconsistency index (I²) [23]. Heterogeneity was deemed high if the χ2 P-value was < 0.1 or I² > 50%. The random-effects model calculated the overall OR/HR for all outcomes. Sensitivity analysis and subgroup analysis assessed the impact of each study on the overall OR/HR for outcomes with significant heterogeneity and the sources of heterogeneity. Funnel plots in Review Manager 5.4.1 and Egger's regression tests in Stata 15.1 (Stata Corp, College Station, Texas, USA) for outcomes with over 10 studies were used to assess publication bias. A P-value < 0.05 indicated significant bias [24]. For results with publication bias, the impact of publication bias on the results is assessed using the trimming-and-filling method.

Results

Literature retrieval, study characteristics, and baseline

Literature retrieval and selection process were depicted in Fig. 1. A total of 306 studies were identified from PubMed (n = 221), Embase (n = 41), Web of Science (n = 44), and Cochrane (n = 0). After removing duplicates, 258 titles and abstracts were assessed, and 15 cohort studies involving 6,830 patients were included [13, 16–19, 25–34]. The characteristics and quality assessment of each study were shown in Table 1. Detailed NOS evaluation results for each study are available in Table S2.

Fig. 1.

Flowchart of the systematic search and selection process

Table 1.

Characteristics of included studies

Author	Region	Population	Measurement time point	No. of patients	Gender		Mean age	Mean/median BMI	Cut-off	Mean follow-up	Outcomes	Adjustment factors
					Male	Female
CONUT
Cao 2024 [14]	China	Patients With Chronic Limb-Threatening Ischemia Following Endovascular Revascularization	Preoperative	232	147	85	70.8	23.3	5	2.1 years	Mortality, MACE	Hypertension, Coronary artery disease, Diabetes, Dialysis, Technical success
Furuyama 2020 [26]	Japan	Patients with critical limb ischemia	Baseline	112	81	31	73.6	21.5	4	17.8 months	Poor ulcer healing	End-stage renal disease
Jhang 2020 [27]	China	Lower extremity arterial disease undergoing endovascular therapy	Preoperative	232	109	123	85.4	23.1	3	971 days	Mortality	Ambulatory status, Congestive heart failure, Cerebrovascular accident, C-reactive protein, Dialysis, Body mass index
Kobayashi 2022a [28]	Japan	Patients with Chronic Limb Threatening Ischaemia Having Distal Bypass	Preoperative	249	169	80	76	NA	NA	28 months	Mortality, Poor ulcer healing	Diabetes, coronary artery disease, cerebrovascular disease, Haemodialysis
Lin 2022a [6]	China	Dialysis Patients Undergoing Endovascular Therapy for Peripheral Artery Disease	Preoperative	395	172	223	68	24	2	2.2 years	Mortality, MACE	Age, sex, current smoking, dialysis vintage, body mass index, diabetes mellitus, hypertension, coronary artery disease, congestive heart failure, cerebrovascular disease, atrial fibrillation, use of antiplatelets, use of beta blockers, use of renin angiotensin aldosterone system inhibitors, and use of statins
Lin 2022b [28]	China	Dialysis Patients Undergoing Endovascular Therapy for Peripheral Artery Disease	Preoperative	395	172	223	68	24	5	2.2 years	Mortality, MACE	Age, sex, current smoking, dialysis vintage, body mass index, diabetes mellitus, hypertension, coronary artery disease, congestive heart failure, cerebrovascular disease, atrial fibrillation, use of antiplatelets, use of beta blockers, use of renin angiotensin aldosterone system inhibitors, and use of statins
Lin 2022c [29]	China	Dialysis Patients Undergoing Endovascular Therapy for Peripheral Artery Disease	Preoperative	395	172	223	68	24	9	2.2 years	Mortality, MACE	Age, sex, current smoking, dialysis vintage, body mass index, diabetes mellitus, hypertension, coronary artery disease, congestive heart failure, cerebrovascular disease, atrial fibrillation, use of antiplatelets, use of beta blockers, use of renin angiotensin aldosterone system inhibitors, and use of statins
Mii 2020 [30]	Japan	Patients Undergoing Open Bypass for Critical Limb Ischemia	Preoperative	359	122	237	74	NA	NA	969 days	Mortality, Amputation	Age, ABI, ESRD
Mine 2021 [31]	Japan	Patients with Chronic Limb-Threatening Ischemia after Endovascular Treatment	Preoperative	120	84	36	73.2	21.6	5	204 days	Amputation, Poor ulcer healing	Hemodialysis, CRP, Moderate-to-severe frailty
Mizobuchi 2019 [32]	Japan	Patients with peripheral artery disease who were undergoing endovascular therapy	Preoperative	628	434	194	69	23.3	NA	828 days	Mortality	Age, sex, BMI, LVEF, Hemoglobin, eGFR
Morisaki 2023I-a [18]	Japan	Patients with chronic limb-threatening ischemia undergoing infrainguinal surgical or endovascular revascularization	Preoperative	508	298	210	75.3	22	NA	NA	Mortality, Amputation	Age, BMI, Improved ambulatory status, endovascular therapy
Morisaki 2023I-b [19]	Japan	Patients with chronic limb-threatening ischemia undergoing infrainguinal surgical or endovascular revascularization	Preoperative	508	298	210	75.3	22	NA	NA	Mortality, Amputation	Age, BMI, Improved ambulatory status, endovascular therapy
Morisaki 2023II-a [18]	Japan	Patients with chronic limb-threatening ischemia	Preoperative	289	199	90	73.8	21.8	NA	NA	Mortality, Amputation	Hemodialysis, Non-ambulatory, Endovascular, Postoperative complication
Shiraki 2021 [33]	Japan	Patients With Chronic Limb Threatening Ischaemia Undergoing Revascularisation	Preoperative	499	338	161	73	22	NA	36 months	Mortality, Amputation	Non-ambulatory status, heart failure, and tissue loss at baseline
Yokoyama 2018a [34]	Japan	Patients With Peripheral Artery Disease Following Endovascular Therapy	Preoperative	357	287	70	74	22.2	2	1071 days	MACE	Age, FFMI, hyperlipidemia, previous IHD, CLI, eGFR, Cys, and hsCRP
Yokoyama 2018b [34]	Japan	Patients With Peripheral Artery Disease Following Endovascular Therapy	Preoperative	357	287	70	74	22.2	5	1071 days	MACE	Age, FFMI, hyperlipidemia, previous IHD, CLI, eGFR, Cys, and hsCRP
PNI
Erdogan 2024 [16]	Turkey	Patients with Chronic Limb-Threatening Ischemia Undergoing Endovascular Therapy	Preoperative	113	87	26	62.9	NA	37	6 months	Mortality	Age, Chronic kidney disease, Chronic heart failure, Amputation, Statin
Pamukcu 2021 [25]	Turkey	Patients with lower extremity peripheral artery disease	Baseline	266	217	49	66	23	NA	NA	Amputation	Hypertension, Diabetes mellitus, Sodium, Hemoglobin, Diffuse involvement
Itagaki 2024 [13]	Japan	Peripheral artery disease undergoing endovascular therapy	Preoperative	278	206	72	73.6	22.2	45	5 years	MACE	Age, sex, BMI, frailty, hypertension, dyslipidemia, diabetes mellitus, current smoker, hemodialysis, prior stroke, prior PCI or CABG, prior heart failure hospitalization, CLTI, hemoglobin, B-type natriuretic peptide, and use of statin and ACEIs and/or ARBs
Kobayashi 2022b [28]	Japan	Patients with Chronic Limb Threatening Ischaemia Having Distal Bypass	Preoperative	249	169	80	76	NA	NA	28 months	Mortality	Diabetes, coronary artery disease, cerebrovascular disease, Haemodialysis
Morisaki 2023II-b [19]	Japan	Patients with chronic limb-threatening ischemia	Preoperative	289	199	90	73.8	21.8	NA	NA	Mortality, Amputation	Hemodialysis, Non-ambulatory, Endovascular, Postoperative complication

CONUT and mortality

Results for CONUT and mortality from 12 cohort studies showed a significantly higher mortality risk in the high CONUT group (HR: 1.34; 95% CI: 1.21, 1.49; P < 0.00001). Significant heterogeneity was observed (I2 = 70%, P = 0.0001) (Fig. 2A).

Fig. 2.

Forest plots of mortality (CONUT) (A) and MACE (CONUT) (B)

CONUT and MACE

Results from 6 cohort studies showed no significant difference in MACE risk between the two groups (OR: 1.31; 95% CI: 0.95, 1.80; P = 0.10), with no significant heterogeneity (I² = 20%, P = 0.28) (Fig. 2B).

CONUT and amputation

Meta-analysis of 6 cohort studies showed a significantly higher risk of amputation in the high CONUT group (OR: 1.20; 95% CI: 1.10, 1.32; P = 0.0001). No significant heterogeneity was observed (I2 = 40%, P = 0.14) (Fig. 3A).

Fig. 3.

Forest plots of amputation (CONUT) (A), and poor ulcer healing (CONUT) (B)

CONUT and poor ulcer healing

Meta-analysis of 3 cohort studies showed no significant difference in poor ulcer healing risk between the two groups (OR: 3.11; 95% CI: 0.90, 10.76; P = 0.07), with significant heterogeneity (I² = 69%, P = 0.04) (Fig. 3B). In the study by Mine et al. [31], "poor healing" was clearly defined as failure to achieve complete epithelialization more than 1 year after revascularization (excluding patients who died or underwent major amputation during this period), and the follow-up protocol was to conduct regular follow-up every 3 months after surgery. The study by Kobayashi et al. [28] listed "wound healing" as a secondary endpoint, but did not give a precise definition of "poor healing". Generally, failure to heal within the follow-up period was considered an adverse outcome, and their follow-up was conducted through review of medical records. The study by Furuyama et al. [26] directly classified major amputation or death before achieving complete epithelialization as "ulcer healing failure", and their average follow-up time was 17.8 months.

PNI and mortality

Meta-analysis of 3 cohort studies showed a significantly lower mortality risk in the high PNI group (HR: 0.95; 95% CI: 0.92, 0.97; P = 0.0003), with no heterogeneity (I2 = 0%, P = 0.54) (Fig. 4A).

Fig. 4.

Forest plots of mortality (PNI) (A), and amputation (PNI) (B)

PNI and amputation

Meta-analysis of 2 cohort studies showed a significantly lower amputation risk in the high PNI group (OR: 0.91; 95% CI: 0.87, 0.95; P < 0.0001), with no heterogeneity (I² = 0%, P = 0.69) (Fig. 4B).

Publication bias and sensitivity analysis

We assessed publication bias for the CONUT-mortality association using funnel plots and Egger’s test. Both the funnel plot (Fig. 5A) and Egger’s test (P = 0.001, Fig. 5B) revealed significant bias. The results of the trimming-and-filling method suggest that the relationship between CONUT and mortality was not significantly affected by publication bias (HR: 1.22; 95% CI: 1.08, 1.38) (Fig. 6). We performed sensitivity analysis for mortality (CONUT) and poor ulcer healing (CONUT) outcomes by excluding each cohort study to assess their impact on the total OR. The total OR remained stable after excluding each cohort study for mortality (CONUT) (Fig. 7A). Excluding data from Lin 2022c reduced the heterogeneity of mortality (CONUT) from 70 to 40%, indicating it as the main source of heterogeneity. For poor ulcer healing (CONUT), after excluding data from Kobayashi 2022a, the association between CONUT and poor ulcer healing changed from insignificant to significant (OR: 6.08; 95% CI: 1.95, 18.94), suggesting that this outcome measure is unstable (Fig. 7B). Furthermore, after excluding data from Kobayashi 2022a, the heterogeneity of poor ulcer healing (CONUT) decreased from 69 to 0%, indicating that this study was the main cause of heterogeneity.

Fig. 5.

Funnel plots (A) and Egger’s test (B) of mortality (CONUT)

Fig. 6.

Trimming-and-filling funnel plot of the relationship between CONUT and mortality

Fig. 7.

Sensitivity analysis of mortality (CONUT) (A), and poor ulcer healing (CONUT) (B)

Subgroup analysis

Subgroup analyses of the relationship between CONUT and mortality were performed based on cutoff values. Results indicated that in the subgroup with a CONUT cutoff value ≥ 5, the association between high CONUT scores and mortality was statistically significant (pooled HR = 2.42, 95% CI: 1.36–4.31, P = 0.003), although moderate heterogeneity existed within this subgroup (I2 = 73%). In the subgroup with a CONUT cutoff value < 5, the association was not statistically significant (pooled HR = 1.68, 95% CI: 0.78–3.64, P = 0.18). The effect size difference between the two groups was not statistically significant (subgroup difference test P = 0.46) (Fig. 8).

Fig. 8.

Subgroup analysis of CONUT and mortality

Discussion

This meta-analysis, synthesizing data from 15 cohort studies involving 6,830 patients, provides comprehensive evidence on the prognostic utility of objective nutritional indices in PAD. The principal finding confirms that a poor nutritional status, as indicated by an elevated CONUT score or a decreased PNI, is consistently associated with an increased risk of mortality and amputation. In contrast, the current evidence does not support a significant association between CONUT scores and the risk of MACE. The relationship between CONUT and poor ulcer healing requires cautious interpretation due to limited and inconsistent data. These findings underscore the potential value of integrating simple, laboratory-based nutritional assessments into the routine prognostic evaluation of PAD patients.

This study found no correlation between CONUT and MACE. This may be because MACE is more directly associated with atherosclerotic thrombotic events, while the "malnutrition-inflammation-weakness" state reflected by CONUT/PNI may more generally affect the risk of death and amputation through "non-cardiovascular" pathways such as increased susceptibility to infection, multiple organ dysfunction, and poor postoperative recovery. Furthermore, differences in the definition of MACE among different studies may also be a contributing factor. Besides, this study found that CONUT lacked significant predictive value for poor ulcer healing, but sensitivity analysis contradicted this. Excluding the data from Kobayashi 2022a significantly reduced heterogeneity in poor ulcer healing and altered significance, suggesting that CONUT may predict poor ulcer healing. Current evidence does not support the conclusion that CONUT lacks predictive value; instead, it may have some potential. However, given the limited studies, further large-scale research is needed to confirm CONUT's predictive value for ulcer healing rates. Furthermore, the high statistical heterogeneity (I2 = 70%) found in the association between CONUT score and mortality suggests that the pooled effect size should be interpreted with caution. Although we attempted to explore its sources through subgroup analysis, we were unable to adequately quantify these potential sources of heterogeneity because most original studies did not report stratified data according to other key clinical or methodological variables (including specific disease phenotypes and details of revascularization methods). In addition, the stability of the association between CONUT and ulcer healing was challenged in sensitivity analyses. Therefore, it must be acknowledged that the inherent variability in the current evidence framework limits our ability to accurately estimate the “true” effect size. In light of this, we have significantly toned down the interpretation of our findings in the main text to avoid making overly definitive inferences. Future research urgently needs to standardize the recording and presentation of key covariates in protocol design and outcome reporting to support more reliable evidence synthesis, thereby providing a more solid foundation for the precise application of CONUT and PNI in PAD management.

Beyond confirming the overall prognostic utility, a key challenge in translating these findings into clinical practice lies in the heterogeneity of cutoff values ​​used across different studies. Our analysis included studies that defined “high” CONUT scores using thresholds ranging from as low as 2 to as high as 9. This lack of standardization underscores the current lack of consensus on the optimal cutoff point for risk stratification in the PAD population. Sensitivity analysis indicated that studies using significantly high cutoff values were a major source of heterogeneity in mortality analyses, suggesting that predictive strength may vary depending on the choice of threshold. Future research must prioritize identifying validated and potentially severity-related cutoff values ​​for CONUT and PNI in PAD patients. Establishing these criteria is crucial to ensuring the consistent and reliable application of these tools across diverse clinical settings to identify which patients will benefit most from nutritional interventions.

Nutritional status, frailty, and sarcopenia are objective indicators of a patient's general condition [35–37]. Frailty and sarcopenia, nutrition-related conditions, impact the clinical outcomes of PAD patients [38–40]. Sarcopenia is linked to higher mortality in PAD patients after endovascular revascularization [41], while the CONUT score predicts sarcopenia in cancer patients [42]. However, the CONUT score and sarcopenia are independent in their prognostic value. Previous studies indicate that PNI correlates with poor prognosis in PAD patients, with lower PNI linked to higher cumulative mortality [13, 16, 18, 25, 28]. Both CONUT and PNI are simple, cost-effective tools for assessing nutritional status, enabling continuous monitoring, early detection of malnutrition, and timely interventions. A prospective, multicenter study found that patients with improved nutritional status after revascularization had longer life expectancy, even with poor preoperative nutrition [33]. Additionally, some studies suggest that CONUT is more effective and convenient than the Geriatric Nutritional Risk Index (GNRI) for predicting survival and wound healing [28]. However, due to limited data, direct comparison of these scores requires further clinical investigation.

Malnutrition increases postoperative complications, mortality, and hospital stay, while reducing quality of life. It impairs wound healing by decreasing fibroblast proliferation, collagen formation, and angiogenesis [43]. Malnutrition weakens the immune system, reduces leukocyte activity, and increases the risk of infection and delayed healing. Albumin, a nutritional marker, predicts poor prognosis in PAD patients [44], while preoperative hypoalbuminemia is linked to higher morbidity and mortality in those undergoing lower limb bypass surgery [45]. Malnutrition reduces lymphocyte counts, impairing cellular immunity and lytic leukocyte activity. Multidisciplinary care, including resistance training and nutritional therapy, improves PAD prognosis. A simple nutritional index should be identified for risk stratification and clinical management of PAD [46]. PNI and CONUT are simple indices combining nutritional status and inflammatory factors, calculable from routine biochemistry and blood counts. This study confirms that PNI and CONUT assessments predict outcomes in PAD patients after EVT and guide clinical practice by enabling early nutritional support and intervention to improve outcomes.

This meta-analysis reveals significant correlations that elevate CONUT and PNI from simple nutritional screening to valuable prognostic tools. These findings have strong biological validity. Malnutrition, as reflected in these scores, contributes to adverse outcomes through multiple pathways: impaired immune function and increased risk of infection, impaired tissue repair and wound healing, and exacerbation of sarcopenia and weakness [47]. The practical significance of this study is that CONUT and PNI provide a rapid, inexpensive, and objective method to identify patients with PAD at the highest risk of death and amputation who may be trapped in this vicious cycle. This shifts the paradigm from general nutritional awareness to targeted risk assessment. Integrating these scores into routine clinical evaluations allows for timely referrals for comprehensive nutritional screening and paves the way for personalized, multidisciplinary management strategies [48, 49]. Therefore, the core clinical question arising from our findings is not whether nutrition matters, but how to systematically identify vulnerable populations and provide effective interventions. Future prospective studies should not only attempt to validate the prognostic value of these scores but also investigate whether CONUT/PNI-guided nutritional therapy can truly improve the hard clinical endpoints of PAD.

While our meta-analysis demonstrated a significant and consistent association between nutritional indicators and adverse outcomes, a fundamental principle must be emphasized: correlation does not equal causation. The observational design of all included studies means that no matter how strong the identified association, it cannot ultimately prove that poor nutritional status directly causes increased mortality or amputation rates. A key and plausible alternative explanation is inverse causation. Specifically, patients with more severe PADs (such as chronic limb-threatening ischemia) often experience a cascading effect—including chronic systemic inflammation, increased metabolic demand, prolonged bed rest, anorexia due to chronic pain, and significant protein loss from non-healing wounds—all of which can directly trigger or exacerbate malnutrition [50, 51]. In this context, a high CONUT or low PNI score could be both a measure of nutritional deficiency and a marker of overall disease severity and inflammatory burden. Therefore, while aggressive nutritional support remains a reasonable clinical strategy, our findings do not confirm that improving these nutritional scores directly translates to better outcomes. Future research should include well-designed prospective studies to strictly control for disease severity and ultimately interventional trials to explore whether targeted nutritional interventions can improve outcomes in high-risk patients identified by these scores.

The findings of this study suggest that integrating readily available nutritional and inflammatory markers such as CONUT and PNI into existing comprehensive assessment and management systems for peripheral artery disease has significant clinical translational potential. They can serve as a valuable complement to traditional risk assessment tools (including the WIFI scale and frailty scale), jointly constructing a multidimensional and more refined risk stratification model, thereby identifying patients at extremely high risk of death and amputation earlier and more accurately. In clinical practice, both CONUT and PNI are simple, cost-effective, and objective indicators; their selection may depend on local laboratory routines, focus on specific outcomes, and clinician preferences [15]; current evidence is insufficient to determine which is absolutely superior. It must be clear that CONUT and PNI are important prognostic biomarkers, their value lying in risk warning and patient stratification, rather than direct therapeutic targets [52]. Therefore, routine monitoring of these indicators should aim to trigger comprehensive assessment of high-risk patients and initiate multidisciplinary management. However, whether improvements in the scores themselves translate into benefits for hard endpoints still needs to be validated through prospective, interventional studies. Future research should focus on exploring the effectiveness of risk-oriented intervention strategies based on these scores.

This meta-analysis has several limitations. First, all included studies were retrospective cohort studies, a design that inherently carries the risk of unmeasurable confounding bias (such as differences in treatment strategies and nutritional support measures), which may affect the reliability of the observed associations. Second, most studies were from Asia (primarily China and Japan), lacking data from populations in Europe, America, and Africa, limiting the generalizability of the findings globally. Furthermore, significant heterogeneity and publication bias were observed in the analysis of the association between CONUT and mortality, and the predictive value of CONUT for poor ulcer healing showed instability in sensitivity analysis; these methodological limitations necessitate careful interpretation of the results. In addition, because the original studies included differed in their definitions of "poor ulcer healing" (for example, some studies defined it as ulcers failing to heal or shrink to a certain proportion within a specific time after intervention), this meta-analysis failed to provide a unified standard definition. This is another limitation of this study and may introduce clinical heterogeneity when pooling data. Finally, due to limitations in the number of studies included, we were unable to conduct subgroup analyses based on PAD severity (such as intermittent claudication versus chronic limb-threatening ischemia), comorbidities, or revascularization types. These important sources of heterogeneity require further investigation through future meta-analyses of individual participant data with larger sample sizes. Despite these limitations, as the first meta-analysis to comprehensively assess the prognostic value of CONUT and PNI for patients with PAD, the results of this study highlight the importance of monitoring these nutritional indicators and provide crucial evidence for future large-scale prospective studies.

Conclusion

CONUT and PNI are accessible, cost-effective, and non-invasive nutritional markers that could predict mortality and amputation risks in PAD patients, helping identify those at risk of poor prognosis. Therefore, if a high CONUT score or low PNI score is detected clinically, clinicians should be encouraged to refer these patients for comprehensive nutritional screening and assessment, and to initiate a multidisciplinary team intervention to consider individualized nutritional support and rehabilitation plans. However, CONUT and PNI are prognostic biomarkers, and their value lies in identifying high-risk patients for enhanced monitoring and management, rather than being direct therapeutic targets. Due to the potential instability and retrospective design of this study, further large-scale, multicenter prospective studies are needed to assess their prognostic value in PAD patients.

Supplementary Information

Supplementary Material 1. ^{(85.9KB, docx)}

Supplementary Material 2. ^{(21.5KB, docx)}

Authors’ contributions

All authors contributed to the study conception and design. **Ying He**: Conceptualization, Methodology, Software, Writing- Original draft, Data curation, Visualization were performed; **Xiaoci He, Meng Liu, Wei Gao and Yang Liu**: Investigation, Writing—Original Draft, Writing—Reviewing and Editing were performed; **Shuyun Guo**: Conceptualization, Supervision, Project administration were performed. All authors read and approved the final manuscript.

Funding

Medical Science Research Project of Hebei (20200967).

Data availability

The data used to support the findings of this study are included within the article.

Declarations

Ethics approval and consent to participate

Not applicable.

Consent for publication

Not applicable.

Competing interests

The authors declare no competing interests.